Compare commits
| Author | SHA1 | Date | |
|---|---|---|---|
|
|
b0784aa131 |
No files matched your search
@@ -1,23 +1,11 @@
|
||||
aclocal.m4
|
||||
config.cache
|
||||
config.guess
|
||||
config.h
|
||||
config.h.in
|
||||
config.log
|
||||
config.status
|
||||
config.sub
|
||||
configure
|
||||
libtool
|
||||
ltmain.sh
|
||||
ltconfig
|
||||
Makefile
|
||||
Makefile.in
|
||||
mono-*.tar.gz
|
||||
stamp-h
|
||||
config.cache
|
||||
stamp-h.in
|
||||
mono.pc
|
||||
mono.spec
|
||||
autom4te.cache
|
||||
mint.pc
|
||||
stamp-h1
|
||||
config.h.in
|
||||
config.status
|
||||
aclocal.m4
|
||||
configure
|
||||
stamp-h
|
||||
libtool
|
||||
Makefile.in
|
||||
config.log
|
||||
@@ -1,3 +1 @@
|
||||
Miguel de Icaza ([email protected])
|
||||
Paolo Molaro ([email protected])
|
||||
Dietmar Maurer ([email protected])
|
||||
@@ -1,485 +0,0 @@
|
||||
|
||||
The Mono runtime is licensed under the terms of the GNU
|
||||
Library General Public License, version 2.
|
||||
|
||||
GNU LIBRARY GENERAL PUBLIC LICENSE
|
||||
Version 2, June 1991
|
||||
|
||||
Copyright (C) 1991 Free Software Foundation, Inc.
|
||||
675 Mass Ave, Cambridge, MA 02139, USA
|
||||
Everyone is permitted to copy and distribute verbatim copies
|
||||
of this license document, but changing it is not allowed.
|
||||
|
||||
[This is the first released version of the library GPL. It is
|
||||
numbered 2 because it goes with version 2 of the ordinary GPL.]
|
||||
|
||||
Preamble
|
||||
|
||||
The licenses for most software are designed to take away your
|
||||
freedom to share and change it. By contrast, the GNU General Public
|
||||
Licenses are intended to guarantee your freedom to share and change
|
||||
free software--to make sure the software is free for all its users.
|
||||
|
||||
This license, the Library General Public License, applies to some
|
||||
specially designated Free Software Foundation software, and to any
|
||||
other libraries whose authors decide to use it. You can use it for
|
||||
your libraries, too.
|
||||
|
||||
When we speak of free software, we are referring to freedom, not
|
||||
price. Our General Public Licenses are designed to make sure that you
|
||||
have the freedom to distribute copies of free software (and charge for
|
||||
this service if you wish), that you receive source code or can get it
|
||||
if you want it, that you can change the software or use pieces of it
|
||||
in new free programs; and that you know you can do these things.
|
||||
|
||||
To protect your rights, we need to make restrictions that forbid
|
||||
anyone to deny you these rights or to ask you to surrender the rights.
|
||||
These restrictions translate to certain responsibilities for you if
|
||||
you distribute copies of the library, or if you modify it.
|
||||
|
||||
For example, if you distribute copies of the library, whether gratis
|
||||
or for a fee, you must give the recipients all the rights that we gave
|
||||
you. You must make sure that they, too, receive or can get the source
|
||||
code. If you link a program with the library, you must provide
|
||||
complete object files to the recipients so that they can relink them
|
||||
with the library, after making changes to the library and recompiling
|
||||
it. And you must show them these terms so they know their rights.
|
||||
|
||||
Our method of protecting your rights has two steps: (1) copyright
|
||||
the library, and (2) offer you this license which gives you legal
|
||||
permission to copy, distribute and/or modify the library.
|
||||
|
||||
Also, for each distributor's protection, we want to make certain
|
||||
that everyone understands that there is no warranty for this free
|
||||
library. If the library is modified by someone else and passed on, we
|
||||
want its recipients to know that what they have is not the original
|
||||
version, so that any problems introduced by others will not reflect on
|
||||
the original authors' reputations.
|
||||
|
||||
Finally, any free program is threatened constantly by software
|
||||
patents. We wish to avoid the danger that companies distributing free
|
||||
software will individually obtain patent licenses, thus in effect
|
||||
transforming the program into proprietary software. To prevent this,
|
||||
we have made it clear that any patent must be licensed for everyone's
|
||||
free use or not licensed at all.
|
||||
|
||||
Most GNU software, including some libraries, is covered by the ordinary
|
||||
GNU General Public License, which was designed for utility programs. This
|
||||
license, the GNU Library General Public License, applies to certain
|
||||
designated libraries. This license is quite different from the ordinary
|
||||
one; be sure to read it in full, and don't assume that anything in it is
|
||||
the same as in the ordinary license.
|
||||
|
||||
The reason we have a separate public license for some libraries is that
|
||||
they blur the distinction we usually make between modifying or adding to a
|
||||
program and simply using it. Linking a program with a library, without
|
||||
changing the library, is in some sense simply using the library, and is
|
||||
analogous to running a utility program or application program. However, in
|
||||
a textual and legal sense, the linked executable is a combined work, a
|
||||
derivative of the original library, and the ordinary General Public License
|
||||
treats it as such.
|
||||
|
||||
Because of this blurred distinction, using the ordinary General
|
||||
Public License for libraries did not effectively promote software
|
||||
sharing, because most developers did not use the libraries. We
|
||||
concluded that weaker conditions might promote sharing better.
|
||||
|
||||
However, unrestricted linking of non-free programs would deprive the
|
||||
users of those programs of all benefit from the free status of the
|
||||
libraries themselves. This Library General Public License is intended to
|
||||
permit developers of non-free programs to use free libraries, while
|
||||
preserving your freedom as a user of such programs to change the free
|
||||
libraries that are incorporated in them. (We have not seen how to achieve
|
||||
this as regards changes in header files, but we have achieved it as regards
|
||||
changes in the actual functions of the Library.) The hope is that this
|
||||
will lead to faster development of free libraries.
|
||||
|
||||
The precise terms and conditions for copying, distribution and
|
||||
modification follow. Pay close attention to the difference between a
|
||||
"work based on the library" and a "work that uses the library". The
|
||||
former contains code derived from the library, while the latter only
|
||||
works together with the library.
|
||||
|
||||
Note that it is possible for a library to be covered by the ordinary
|
||||
General Public License rather than by this special one.
|
||||
|
||||
GNU LIBRARY GENERAL PUBLIC LICENSE
|
||||
TERMS AND CONDITIONS FOR COPYING, DISTRIBUTION AND MODIFICATION
|
||||
|
||||
0. This License Agreement applies to any software library which
|
||||
contains a notice placed by the copyright holder or other authorized
|
||||
party saying it may be distributed under the terms of this Library
|
||||
General Public License (also called "this License"). Each licensee is
|
||||
addressed as "you".
|
||||
|
||||
A "library" means a collection of software functions and/or data
|
||||
prepared so as to be conveniently linked with application programs
|
||||
(which use some of those functions and data) to form executables.
|
||||
|
||||
The "Library", below, refers to any such software library or work
|
||||
which has been distributed under these terms. A "work based on the
|
||||
Library" means either the Library or any derivative work under
|
||||
copyright law: that is to say, a work containing the Library or a
|
||||
portion of it, either verbatim or with modifications and/or translated
|
||||
straightforwardly into another language. (Hereinafter, translation is
|
||||
included without limitation in the term "modification".)
|
||||
|
||||
"Source code" for a work means the preferred form of the work for
|
||||
making modifications to it. For a library, complete source code means
|
||||
all the source code for all modules it contains, plus any associated
|
||||
interface definition files, plus the scripts used to control compilation
|
||||
and installation of the library.
|
||||
|
||||
Activities other than copying, distribution and modification are not
|
||||
covered by this License; they are outside its scope. The act of
|
||||
running a program using the Library is not restricted, and output from
|
||||
such a program is covered only if its contents constitute a work based
|
||||
on the Library (independent of the use of the Library in a tool for
|
||||
writing it). Whether that is true depends on what the Library does
|
||||
and what the program that uses the Library does.
|
||||
|
||||
1. You may copy and distribute verbatim copies of the Library's
|
||||
complete source code as you receive it, in any medium, provided that
|
||||
you conspicuously and appropriately publish on each copy an
|
||||
appropriate copyright notice and disclaimer of warranty; keep intact
|
||||
all the notices that refer to this License and to the absence of any
|
||||
warranty; and distribute a copy of this License along with the
|
||||
Library.
|
||||
|
||||
You may charge a fee for the physical act of transferring a copy,
|
||||
and you may at your option offer warranty protection in exchange for a
|
||||
fee.
|
||||
|
||||
2. You may modify your copy or copies of the Library or any portion
|
||||
of it, thus forming a work based on the Library, and copy and
|
||||
distribute such modifications or work under the terms of Section 1
|
||||
above, provided that you also meet all of these conditions:
|
||||
|
||||
a) The modified work must itself be a software library.
|
||||
|
||||
b) You must cause the files modified to carry prominent notices
|
||||
stating that you changed the files and the date of any change.
|
||||
|
||||
c) You must cause the whole of the work to be licensed at no
|
||||
charge to all third parties under the terms of this License.
|
||||
|
||||
d) If a facility in the modified Library refers to a function or a
|
||||
table of data to be supplied by an application program that uses
|
||||
the facility, other than as an argument passed when the facility
|
||||
is invoked, then you must make a good faith effort to ensure that,
|
||||
in the event an application does not supply such function or
|
||||
table, the facility still operates, and performs whatever part of
|
||||
its purpose remains meaningful.
|
||||
|
||||
(For example, a function in a library to compute square roots has
|
||||
a purpose that is entirely well-defined independent of the
|
||||
application. Therefore, Subsection 2d requires that any
|
||||
application-supplied function or table used by this function must
|
||||
be optional: if the application does not supply it, the square
|
||||
root function must still compute square roots.)
|
||||
|
||||
These requirements apply to the modified work as a whole. If
|
||||
identifiable sections of that work are not derived from the Library,
|
||||
and can be reasonably considered independent and separate works in
|
||||
themselves, then this License, and its terms, do not apply to those
|
||||
sections when you distribute them as separate works. But when you
|
||||
distribute the same sections as part of a whole which is a work based
|
||||
on the Library, the distribution of the whole must be on the terms of
|
||||
this License, whose permissions for other licensees extend to the
|
||||
entire whole, and thus to each and every part regardless of who wrote
|
||||
it.
|
||||
|
||||
Thus, it is not the intent of this section to claim rights or contest
|
||||
your rights to work written entirely by you; rather, the intent is to
|
||||
exercise the right to control the distribution of derivative or
|
||||
collective works based on the Library.
|
||||
|
||||
In addition, mere aggregation of another work not based on the Library
|
||||
with the Library (or with a work based on the Library) on a volume of
|
||||
a storage or distribution medium does not bring the other work under
|
||||
the scope of this License.
|
||||
|
||||
3. You may opt to apply the terms of the ordinary GNU General Public
|
||||
License instead of this License to a given copy of the Library. To do
|
||||
this, you must alter all the notices that refer to this License, so
|
||||
that they refer to the ordinary GNU General Public License, version 2,
|
||||
instead of to this License. (If a newer version than version 2 of the
|
||||
ordinary GNU General Public License has appeared, then you can specify
|
||||
that version instead if you wish.) Do not make any other change in
|
||||
these notices.
|
||||
|
||||
Once this change is made in a given copy, it is irreversible for
|
||||
that copy, so the ordinary GNU General Public License applies to all
|
||||
subsequent copies and derivative works made from that copy.
|
||||
|
||||
This option is useful when you wish to copy part of the code of
|
||||
the Library into a program that is not a library.
|
||||
|
||||
4. You may copy and distribute the Library (or a portion or
|
||||
derivative of it, under Section 2) in object code or executable form
|
||||
under the terms of Sections 1 and 2 above provided that you accompany
|
||||
it with the complete corresponding machine-readable source code, which
|
||||
must be distributed under the terms of Sections 1 and 2 above on a
|
||||
medium customarily used for software interchange.
|
||||
|
||||
If distribution of object code is made by offering access to copy
|
||||
from a designated place, then offering equivalent access to copy the
|
||||
source code from the same place satisfies the requirement to
|
||||
distribute the source code, even though third parties are not
|
||||
compelled to copy the source along with the object code.
|
||||
|
||||
5. A program that contains no derivative of any portion of the
|
||||
Library, but is designed to work with the Library by being compiled or
|
||||
linked with it, is called a "work that uses the Library". Such a
|
||||
work, in isolation, is not a derivative work of the Library, and
|
||||
therefore falls outside the scope of this License.
|
||||
|
||||
However, linking a "work that uses the Library" with the Library
|
||||
creates an executable that is a derivative of the Library (because it
|
||||
contains portions of the Library), rather than a "work that uses the
|
||||
library". The executable is therefore covered by this License.
|
||||
Section 6 states terms for distribution of such executables.
|
||||
|
||||
When a "work that uses the Library" uses material from a header file
|
||||
that is part of the Library, the object code for the work may be a
|
||||
derivative work of the Library even though the source code is not.
|
||||
Whether this is true is especially significant if the work can be
|
||||
linked without the Library, or if the work is itself a library. The
|
||||
threshold for this to be true is not precisely defined by law.
|
||||
|
||||
If such an object file uses only numerical parameters, data
|
||||
structure layouts and accessors, and small macros and small inline
|
||||
functions (ten lines or less in length), then the use of the object
|
||||
file is unrestricted, regardless of whether it is legally a derivative
|
||||
work. (Executables containing this object code plus portions of the
|
||||
Library will still fall under Section 6.)
|
||||
|
||||
Otherwise, if the work is a derivative of the Library, you may
|
||||
distribute the object code for the work under the terms of Section 6.
|
||||
Any executables containing that work also fall under Section 6,
|
||||
whether or not they are linked directly with the Library itself.
|
||||
|
||||
6. As an exception to the Sections above, you may also compile or
|
||||
link a "work that uses the Library" with the Library to produce a
|
||||
work containing portions of the Library, and distribute that work
|
||||
under terms of your choice, provided that the terms permit
|
||||
modification of the work for the customer's own use and reverse
|
||||
engineering for debugging such modifications.
|
||||
|
||||
You must give prominent notice with each copy of the work that the
|
||||
Library is used in it and that the Library and its use are covered by
|
||||
this License. You must supply a copy of this License. If the work
|
||||
during execution displays copyright notices, you must include the
|
||||
copyright notice for the Library among them, as well as a reference
|
||||
directing the user to the copy of this License. Also, you must do one
|
||||
of these things:
|
||||
|
||||
a) Accompany the work with the complete corresponding
|
||||
machine-readable source code for the Library including whatever
|
||||
changes were used in the work (which must be distributed under
|
||||
Sections 1 and 2 above); and, if the work is an executable linked
|
||||
with the Library, with the complete machine-readable "work that
|
||||
uses the Library", as object code and/or source code, so that the
|
||||
user can modify the Library and then relink to produce a modified
|
||||
executable containing the modified Library. (It is understood
|
||||
that the user who changes the contents of definitions files in the
|
||||
Library will not necessarily be able to recompile the application
|
||||
to use the modified definitions.)
|
||||
|
||||
b) Accompany the work with a written offer, valid for at
|
||||
least three years, to give the same user the materials
|
||||
specified in Subsection 6a, above, for a charge no more
|
||||
than the cost of performing this distribution.
|
||||
|
||||
c) If distribution of the work is made by offering access to copy
|
||||
from a designated place, offer equivalent access to copy the above
|
||||
specified materials from the same place.
|
||||
|
||||
d) Verify that the user has already received a copy of these
|
||||
materials or that you have already sent this user a copy.
|
||||
|
||||
For an executable, the required form of the "work that uses the
|
||||
Library" must include any data and utility programs needed for
|
||||
reproducing the executable from it. However, as a special exception,
|
||||
the source code distributed need not include anything that is normally
|
||||
distributed (in either source or binary form) with the major
|
||||
components (compiler, kernel, and so on) of the operating system on
|
||||
which the executable runs, unless that component itself accompanies
|
||||
the executable.
|
||||
|
||||
It may happen that this requirement contradicts the license
|
||||
restrictions of other proprietary libraries that do not normally
|
||||
accompany the operating system. Such a contradiction means you cannot
|
||||
use both them and the Library together in an executable that you
|
||||
distribute.
|
||||
|
||||
7. You may place library facilities that are a work based on the
|
||||
Library side-by-side in a single library together with other library
|
||||
facilities not covered by this License, and distribute such a combined
|
||||
library, provided that the separate distribution of the work based on
|
||||
the Library and of the other library facilities is otherwise
|
||||
permitted, and provided that you do these two things:
|
||||
|
||||
a) Accompany the combined library with a copy of the same work
|
||||
based on the Library, uncombined with any other library
|
||||
facilities. This must be distributed under the terms of the
|
||||
Sections above.
|
||||
|
||||
b) Give prominent notice with the combined library of the fact
|
||||
that part of it is a work based on the Library, and explaining
|
||||
where to find the accompanying uncombined form of the same work.
|
||||
|
||||
8. You may not copy, modify, sublicense, link with, or distribute
|
||||
the Library except as expressly provided under this License. Any
|
||||
attempt otherwise to copy, modify, sublicense, link with, or
|
||||
distribute the Library is void, and will automatically terminate your
|
||||
rights under this License. However, parties who have received copies,
|
||||
or rights, from you under this License will not have their licenses
|
||||
terminated so long as such parties remain in full compliance.
|
||||
|
||||
9. You are not required to accept this License, since you have not
|
||||
signed it. However, nothing else grants you permission to modify or
|
||||
distribute the Library or its derivative works. These actions are
|
||||
prohibited by law if you do not accept this License. Therefore, by
|
||||
modifying or distributing the Library (or any work based on the
|
||||
Library), you indicate your acceptance of this License to do so, and
|
||||
all its terms and conditions for copying, distributing or modifying
|
||||
the Library or works based on it.
|
||||
|
||||
10. Each time you redistribute the Library (or any work based on the
|
||||
Library), the recipient automatically receives a license from the
|
||||
original licensor to copy, distribute, link with or modify the Library
|
||||
subject to these terms and conditions. You may not impose any further
|
||||
restrictions on the recipients' exercise of the rights granted herein.
|
||||
You are not responsible for enforcing compliance by third parties to
|
||||
this License.
|
||||
|
||||
11. If, as a consequence of a court judgment or allegation of patent
|
||||
infringement or for any other reason (not limited to patent issues),
|
||||
conditions are imposed on you (whether by court order, agreement or
|
||||
otherwise) that contradict the conditions of this License, they do not
|
||||
excuse you from the conditions of this License. If you cannot
|
||||
distribute so as to satisfy simultaneously your obligations under this
|
||||
License and any other pertinent obligations, then as a consequence you
|
||||
may not distribute the Library at all. For example, if a patent
|
||||
license would not permit royalty-free redistribution of the Library by
|
||||
all those who receive copies directly or indirectly through you, then
|
||||
the only way you could satisfy both it and this License would be to
|
||||
refrain entirely from distribution of the Library.
|
||||
|
||||
If any portion of this section is held invalid or unenforceable under any
|
||||
particular circumstance, the balance of the section is intended to apply,
|
||||
and the section as a whole is intended to apply in other circumstances.
|
||||
|
||||
It is not the purpose of this section to induce you to infringe any
|
||||
patents or other property right claims or to contest validity of any
|
||||
such claims; this section has the sole purpose of protecting the
|
||||
integrity of the free software distribution system which is
|
||||
implemented by public license practices. Many people have made
|
||||
generous contributions to the wide range of software distributed
|
||||
through that system in reliance on consistent application of that
|
||||
system; it is up to the author/donor to decide if he or she is willing
|
||||
to distribute software through any other system and a licensee cannot
|
||||
impose that choice.
|
||||
|
||||
This section is intended to make thoroughly clear what is believed to
|
||||
be a consequence of the rest of this License.
|
||||
|
||||
12. If the distribution and/or use of the Library is restricted in
|
||||
certain countries either by patents or by copyrighted interfaces, the
|
||||
original copyright holder who places the Library under this License may add
|
||||
an explicit geographical distribution limitation excluding those countries,
|
||||
so that distribution is permitted only in or among countries not thus
|
||||
excluded. In such case, this License incorporates the limitation as if
|
||||
written in the body of this License.
|
||||
|
||||
13. The Free Software Foundation may publish revised and/or new
|
||||
versions of the Library General Public License from time to time.
|
||||
Such new versions will be similar in spirit to the present version,
|
||||
but may differ in detail to address new problems or concerns.
|
||||
|
||||
Each version is given a distinguishing version number. If the Library
|
||||
specifies a version number of this License which applies to it and
|
||||
"any later version", you have the option of following the terms and
|
||||
conditions either of that version or of any later version published by
|
||||
the Free Software Foundation. If the Library does not specify a
|
||||
license version number, you may choose any version ever published by
|
||||
the Free Software Foundation.
|
||||
|
||||
14. If you wish to incorporate parts of the Library into other free
|
||||
programs whose distribution conditions are incompatible with these,
|
||||
write to the author to ask for permission. For software which is
|
||||
copyrighted by the Free Software Foundation, write to the Free
|
||||
Software Foundation; we sometimes make exceptions for this. Our
|
||||
decision will be guided by the two goals of preserving the free status
|
||||
of all derivatives of our free software and of promoting the sharing
|
||||
and reuse of software generally.
|
||||
|
||||
NO WARRANTY
|
||||
|
||||
15. BECAUSE THE LIBRARY IS LICENSED FREE OF CHARGE, THERE IS NO
|
||||
WARRANTY FOR THE LIBRARY, TO THE EXTENT PERMITTED BY APPLICABLE LAW.
|
||||
EXCEPT WHEN OTHERWISE STATED IN WRITING THE COPYRIGHT HOLDERS AND/OR
|
||||
OTHER PARTIES PROVIDE THE LIBRARY "AS IS" WITHOUT WARRANTY OF ANY
|
||||
KIND, EITHER EXPRESSED OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, THE
|
||||
IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
|
||||
PURPOSE. THE ENTIRE RISK AS TO THE QUALITY AND PERFORMANCE OF THE
|
||||
LIBRARY IS WITH YOU. SHOULD THE LIBRARY PROVE DEFECTIVE, YOU ASSUME
|
||||
THE COST OF ALL NECESSARY SERVICING, REPAIR OR CORRECTION.
|
||||
|
||||
16. IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN
|
||||
WRITING WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MAY MODIFY
|
||||
AND/OR REDISTRIBUTE THE LIBRARY AS PERMITTED ABOVE, BE LIABLE TO YOU
|
||||
FOR DAMAGES, INCLUDING ANY GENERAL, SPECIAL, INCIDENTAL OR
|
||||
CONSEQUENTIAL DAMAGES ARISING OUT OF THE USE OR INABILITY TO USE THE
|
||||
LIBRARY (INCLUDING BUT NOT LIMITED TO LOSS OF DATA OR DATA BEING
|
||||
RENDERED INACCURATE OR LOSSES SUSTAINED BY YOU OR THIRD PARTIES OR A
|
||||
FAILURE OF THE LIBRARY TO OPERATE WITH ANY OTHER SOFTWARE), EVEN IF
|
||||
SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
|
||||
DAMAGES.
|
||||
|
||||
END OF TERMS AND CONDITIONS
|
||||
|
||||
Appendix: How to Apply These Terms to Your New Libraries
|
||||
|
||||
If you develop a new library, and you want it to be of the greatest
|
||||
possible use to the public, we recommend making it free software that
|
||||
everyone can redistribute and change. You can do so by permitting
|
||||
redistribution under these terms (or, alternatively, under the terms of the
|
||||
ordinary General Public License).
|
||||
|
||||
To apply these terms, attach the following notices to the library. It is
|
||||
safest to attach them to the start of each source file to most effectively
|
||||
convey the exclusion of warranty; and each file should have at least the
|
||||
"copyright" line and a pointer to where the full notice is found.
|
||||
|
||||
<one line to give the library's name and a brief idea of what it does.>
|
||||
Copyright (C) <year> <name of author>
|
||||
|
||||
This library is free software; you can redistribute it and/or
|
||||
modify it under the terms of the GNU Library General Public
|
||||
License as published by the Free Software Foundation; either
|
||||
version 2 of the License, or (at your option) any later version.
|
||||
|
||||
This library is distributed in the hope that it will be useful,
|
||||
but WITHOUT ANY WARRANTY; without even the implied warranty of
|
||||
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
|
||||
Library General Public License for more details.
|
||||
|
||||
You should have received a copy of the GNU Library General Public
|
||||
License along with this library; if not, write to the Free
|
||||
Software Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
|
||||
|
||||
Also add information on how to contact you by electronic and paper mail.
|
||||
|
||||
You should also get your employer (if you work as a programmer) or your
|
||||
school, if any, to sign a "copyright disclaimer" for the library, if
|
||||
necessary. Here is a sample; alter the names:
|
||||
|
||||
Yoyodyne, Inc., hereby disclaims all copyright interest in the
|
||||
library `Frob' (a library for tweaking knobs) written by James Random Hacker.
|
||||
|
||||
<signature of Ty Coon>, 1 April 1990
|
||||
Ty Coon, President of Vice
|
||||
|
||||
That's all there is to it!
|
||||
@@ -1,221 +1 @@
|
||||
AUTOMAKE_OPTIONS = foreign
|
||||
ACLOCAL_AMFLAGS = -I .
|
||||
|
||||
SUBDIRS = $(libgc_dir) mono $(ikvm_jni_dir) docs data runtime scripts man samples support web
|
||||
|
||||
# Keep in sync with SUBDIRS
|
||||
## 'tools' is not normally built
|
||||
DIST_SUBDIRS = libgc mono ikvm-jni docs data runtime scripts man samples support web tools
|
||||
|
||||
EXTRA_DIST= mono.pc.in mono.spec.in mint.pc.in
|
||||
|
||||
pkgconfigdir = $(libdir)/pkgconfig
|
||||
if JIT_SUPPORTED
|
||||
pkgconfig_DATA= mono.pc mint.pc
|
||||
else
|
||||
pkgconfig_DATA= mint.pc
|
||||
endif
|
||||
DISTCLEANFILES= mono.pc mint.pc
|
||||
|
||||
#
|
||||
# The following targets are used to compile and install
|
||||
# mono and mcs without running into the out-of-sync
|
||||
# problems by doing the compilation in the right order.
|
||||
#
|
||||
|
||||
mcs_topdir=$(top_srcdir)/../mcs
|
||||
|
||||
.PHONY: bootstrap faststrap fasterstrap bootstrap-clean bootstrap-check get-monolite-latest monolite-bootstrap
|
||||
.PHONY: tmpinst-dir tmpinst-mcs-basic-build tmpinst-mcs-short-build tmpinst-mcs-full-build tmpinst-monolite
|
||||
.PHONY: mono-do-runtime-only mcs-do-basic-build mcs-do-short-build mcs-do-full-build mcs-do-clean-profiles mcs-do-run-test
|
||||
.PHONY: populate-runtime-subdir
|
||||
|
||||
if USE_JIT
|
||||
mono_runtime = mono/mini/mono
|
||||
else
|
||||
mono_runtime = mono/interpreter/mint
|
||||
endif
|
||||
|
||||
# Minimal set of directories to build, and files used to run the 'mcs' compiler
|
||||
MCS_BASIC_DIRS = jay mcs class/corlib class/System
|
||||
MCS_BASIC_FILES = \
|
||||
$(mcs_topdir)/mcs/mcs.exe \
|
||||
$(mcs_topdir)/class/lib/default/mscorlib.dll \
|
||||
$(mcs_topdir)/class/lib/default/System.dll
|
||||
|
||||
# Set of directories and files the 'mcs' compiler can use
|
||||
## We build class/System twice to resolve the cyclic dependency of class/System.XML
|
||||
MCS_DIRS = $(MCS_BASIC_DIRS) class/System.XML class/System class/Mono.CSharp.Debugger
|
||||
MCS_FILES = $(MCS_BASIC_FILES) \
|
||||
$(mcs_topdir)/class/lib/default/System.Xml.dll \
|
||||
$(mcs_topdir)/class/lib/default/Mono.CSharp.Debugger.dll
|
||||
|
||||
tmpinst = _tmpinst
|
||||
|
||||
if PLATFORM_WIN32
|
||||
|
||||
populate_profiles = default:net_1_1
|
||||
|
||||
# bootstraps using the .NET C# compiler
|
||||
bootstrap: tmpinst-dir mono-do-runtime-only
|
||||
$(MAKE) fasterstrap
|
||||
|
||||
else
|
||||
|
||||
populate_profiles = default:net_1_1 net_2_0:net_2_0
|
||||
|
||||
bootstrap: tmpinst-mcs-basic-build mono-do-runtime-only
|
||||
$(MAKE) tmpinst-mcs-short-build
|
||||
$(MAKE) fasterstrap
|
||||
|
||||
endif
|
||||
|
||||
faststrap: tmpinst-mcs-short-build
|
||||
$(MAKE) mono-do-runtime-only
|
||||
$(MAKE) fasterstrap
|
||||
|
||||
fasterstrap: tmpinst-mcs-full-build
|
||||
$(MAKE) populate-runtime-subdir
|
||||
$(MAKE)
|
||||
|
||||
bootstrap-clean: clean mcs-do-clean-profiles
|
||||
bootstrap-check: check mcs-do-run-test
|
||||
|
||||
# building with monolite
|
||||
|
||||
get-monolite-latest:
|
||||
wget -O $(srcdir)/../monolite-latest.tar.gz http://www.go-mono.com/daily/monolite-latest.tar.gz
|
||||
|
||||
monolite-bootstrap: tmpinst-monolite mono-do-runtime-only
|
||||
rm -f $(MCS_FILES)
|
||||
$(MAKE) tmpinst-mcs-short-build
|
||||
$(MAKE) fasterstrap
|
||||
|
||||
# internal targets
|
||||
|
||||
tmpinst-mcs-basic-build: mcs-do-basic-build tmpinst-dir
|
||||
mv -f $(MCS_BASIC_FILES) $(tmpinst)/lib
|
||||
|
||||
tmpinst-mcs-short-build: mcs-do-short-build tmpinst-dir
|
||||
mv -f $(MCS_FILES) $(tmpinst)/lib
|
||||
|
||||
tmpinst-mcs-full-build: mcs-do-full-build tmpinst-dir
|
||||
cp -f $(MCS_FILES) $(tmpinst)/lib
|
||||
|
||||
tmpinst-monolite: tmpinst-dir
|
||||
-rm -f $(tmpinst)/lib/mcs.exe $(tmpinst)/lib/*.dll
|
||||
-rm -fr $(tmpinst)/monolite-*
|
||||
srcdir=`cd $(srcdir) && pwd` && cd $(tmpinst) && ( gzip -d -c $$srcdir/../monolite-latest.tar.gz | tar xf - )
|
||||
mv -f $(tmpinst)/monolite-*/mcs.exe $(tmpinst)/monolite-*/*.dll $(tmpinst)/lib
|
||||
rm -fr $(tmpinst)/monolite-*
|
||||
|
||||
# Build enough of 'mono' to run the stage1 mcs.exe
|
||||
mono-do-runtime-only: $(CONFIG_HEADER)
|
||||
for dir in $(libgc_dir) mono data; do \
|
||||
(cd $$dir && $(MAKE)) || exit 1 ; \
|
||||
done
|
||||
|
||||
mcs-do-basic-build:
|
||||
rm -f $(MCS_FILES)
|
||||
for dir in $(MCS_BASIC_DIRS); do \
|
||||
(cd $(mcs_topdir)/$$dir && $(MAKE) PROFILE=default USE_BOOT_COMPILE=yes BOOTSTRAP_MCS="mcs -d:BOOTSTRAP_WITH_OLDLIB") || exit 1 ; \
|
||||
done
|
||||
|
||||
mcs-do-short-build:
|
||||
tmpinst=`cd $(tmpinst) && pwd` ; \
|
||||
PATH=$$tmpinst/bin:$$PATH; export PATH ; \
|
||||
for dir in $(MCS_DIRS); do \
|
||||
(cd $(mcs_topdir)/$$dir && $(MAKE) PROFILE=default) || exit 1 ; \
|
||||
done
|
||||
|
||||
mcs-do-full-build:
|
||||
tmpinst=`cd $(tmpinst) && pwd` ; \
|
||||
PATH=$$tmpinst/bin:$$PATH ; export PATH ; \
|
||||
cd $(mcs_topdir) && $(MAKE) all-profiles
|
||||
|
||||
mcs-do-clean-profiles:
|
||||
cd $(mcs_topdir) && $(MAKE) clean-profiles
|
||||
|
||||
mcs-do-run-test: tmpinst-dir
|
||||
tmpinst=`cd $(tmpinst) && pwd` ; \
|
||||
PATH=$$tmpinst/bin:$$PATH ; export PATH ; \
|
||||
cd $(mcs_topdir) && $(MAKE) PROFILE=default run-test
|
||||
|
||||
populate-runtime-subdir:
|
||||
test -n '$(populate_profiles)'
|
||||
@srcdir=`cd $(srcdir) && pwd`; tmpinst=`pwd`/$(tmpinst) ; \
|
||||
PATH=$$tmpinst/bin:$$PATH ; export PATH ; \
|
||||
for profile in $(populate_profiles); do \
|
||||
( set fnord `echo $$profile | sed 's,:, ,g'` && \
|
||||
echo "cd $(mcs_topdir) && $(MAKE) PROGRAM_INSTALL_DIR=$$srcdir/runtime PROFILE=$$2 LIBRARY_INSTALL_DIR=$$srcdir/runtime/$$3 prefix=$$tmpinst/prefix install" && \
|
||||
cd $(mcs_topdir) && \
|
||||
$(MAKE) PROGRAM_INSTALL_DIR=$$srcdir/runtime PROFILE=$$2 LIBRARY_INSTALL_DIR=$$srcdir/runtime/$$3 prefix=$$tmpinst/prefix install ) || exit 1 ; \
|
||||
done
|
||||
find $(tmpinst)/prefix -type f -path '*/jay*' -exec rm -f '{}' ';'
|
||||
test -z "`find $(tmpinst)/prefix -type f -print`"
|
||||
rm -rf $(tmpinst)/prefix
|
||||
|
||||
tmpinst-dir:
|
||||
mkdir -p $(tmpinst) $(tmpinst)/bin $(tmpinst)/lib $(tmpinst)/etc/mono
|
||||
$(MAKE) tmpinst-dir-contents
|
||||
|
||||
tmpinst-dir-contents: $(tmpinst)/bin/mono $(tmpinst)/bin/mcs $(tmpinst)/etc/mono/machine.config
|
||||
|
||||
if PLATFORM_WIN32
|
||||
|
||||
$(tmpinst)/bin/mono: $(srcdir)/Makefile.am
|
||||
echo '#! /bin/sh' > $@
|
||||
echo 'builddir="'`pwd`'"; win_builddir="'`cygpath -w -a .`'"' >> $@
|
||||
echo 'tmpinst="$$win_builddir\\$(tmpinst)"' >> $@
|
||||
echo 'if test -z "$$MONO_PATH" ; then MONO_PATH="$$tmpinst\\lib"; else MONO_PATH="$${MONO_PATH};$$tmpinst\\lib" ; fi' >> $@
|
||||
echo 'MONO_CFG_DIR="$$tmpinst\\etc"' >> $@
|
||||
echo 'export MONO_PATH MONO_CFG_DIR' >> $@
|
||||
echo 'exec "$$builddir/libtool" --mode=execute "$$builddir/mono.exe" "$$@"' >> $@
|
||||
chmod +x $@
|
||||
|
||||
else
|
||||
|
||||
$(tmpinst)/bin/mono: $(srcdir)/Makefile.am
|
||||
echo '#! /bin/sh' > $@
|
||||
echo 'builddir="'`pwd`'"; tmpinst=$$builddir/$(tmpinst)' >> $@
|
||||
echo 'if test -z "$$MONO_PATH" ; then MONO_PATH=$$tmpinst/lib; else MONO_PATH=$${MONO_PATH}:$$tmpinst/lib ; fi' >> $@
|
||||
echo 'MONO_CFG_DIR=$$tmpinst/etc' >> $@
|
||||
echo 'export MONO_PATH MONO_CFG_DIR' >> $@
|
||||
echo 'exec "$$builddir/libtool" --mode=execute "$$builddir/$(mono_runtime)" --config "$$builddir/data/config" "$$@"' >> $@
|
||||
chmod +x $@
|
||||
|
||||
endif
|
||||
|
||||
$(tmpinst)/bin/mcs: $(srcdir)/Makefile.am
|
||||
(t=`cd $(tmpinst) && pwd`; echo '#! /bin/sh'; echo 'exec "'"$$t/bin/mono"'" "'"$$t/lib/mcs.exe"'" "$$@"') > $@
|
||||
chmod +x $@
|
||||
|
||||
$(tmpinst)/etc/mono/machine.config: $(srcdir)/data/machine.config
|
||||
rm -f $@
|
||||
srcdir=`cd $(srcdir) && pwd`; cd $(tmpinst)/etc/mono && $(LN_S) $$srcdir/data/machine.config machine.config
|
||||
|
||||
# Obsolete target -- don't use.
|
||||
.PHONY: fullbuild
|
||||
fullbuild: bootstrap
|
||||
$(MAKE) install
|
||||
echo "*** 'fullbuild' is an obsolete target. Use 'make bootstrap && make install'."
|
||||
exit 1
|
||||
|
||||
win32getdeps:
|
||||
wget http://www.go-mono.com/archive/pkgconfig-0.11-20020310.zip
|
||||
wget http://www.go-mono.com/archive/glib-2.0.4-20020703.zip
|
||||
wget http://www.go-mono.com/archive/glib-dev-2.0.4-20020703.zip
|
||||
wget http://www.go-mono.com/archive/libiconv-1.7.zip
|
||||
wget http://www.go-mono.com/archive/libiconv-dev-1.7.zip
|
||||
wget http://www.go-mono.com/archive/libintl-0.10.40-20020101.zip
|
||||
wget http://www.jroith.de/nsis4cygwin.zip
|
||||
unzip -n -d / pkgconfig-0.11-20020310.zip
|
||||
unzip -n -d / glib-2.0.4-20020703.zip
|
||||
unzip -n -d / glib-dev-2.0.4-20020703.zip
|
||||
unzip -n -d / libiconv-1.7.zip
|
||||
unzip -n -d / libiconv-dev-1.7.zip
|
||||
unzip -n -d / libintl-0.10.40-20020101.zip
|
||||
unzip -n -d / nsis4cygwin.zip
|
||||
|
||||
win32setup:
|
||||
makensis /DMILESTONE=$(VERSION) /DSOURCE_INSTALL_DIR=$(SOURCE_INSTALL_DIR) /DBUILDNUM=$(BUILDNUM) monowiz.win32.nsi
|
||||
SUBDIRS = mono doc
|
||||
@@ -1,256 +1,8 @@
|
||||
This is MonoNet.
|
||||
|
||||
This is Mono.
|
||||
Ximian's portable implementation of .NET
|
||||
|
||||
1. Installation
|
||||
2. Using Mono
|
||||
3. Directory Roadmap
|
||||
|
||||
1. Compilation and Installation
|
||||
===============================
|
||||
|
||||
a. Build Requirements
|
||||
---------------------
|
||||
|
||||
To build Mono, you will need the following components:
|
||||
|
||||
* pkg-config
|
||||
|
||||
Available from: http://www.freedesktop.org/Software/pkgconfig
|
||||
|
||||
* glib 2.0
|
||||
|
||||
Available from: http://www.gtk.org/
|
||||
|
||||
Optional dependencies:
|
||||
|
||||
* ICU library
|
||||
|
||||
http://oss.software.ibm.com/icu/index.html
|
||||
|
||||
You will need this one to get complete support for the international
|
||||
features of the .NET Framework.
|
||||
|
||||
* Cairo and libgdiplus
|
||||
|
||||
If you want to get support for System.Drawing, you will need to get
|
||||
both Libgdiplus and Cairo.
|
||||
|
||||
b. Building the Software
|
||||
------------------------
|
||||
|
||||
If you obtained this package as an officially released tarball,
|
||||
this is very simple, use configure and make:
|
||||
|
||||
./configure --prefix=/usr/local
|
||||
make
|
||||
make install
|
||||
|
||||
Mono supports a JIT engine on x86, SPARC and PowerPC systems.
|
||||
The various commands that ship with Mono default to the JIT engine
|
||||
on x86 and SPARC, to turn it on for PPC systems, use the --with-jit=yes
|
||||
command line option to configure.
|
||||
|
||||
MacOS X Users: you will need to download the latest Boehm GC
|
||||
Alpha release for garbage collection to work properly.
|
||||
|
||||
If you obtained this as a snapshot, you will need an existing
|
||||
Mono installation. To upgrade your installation, unpack both
|
||||
mono and mcs:
|
||||
|
||||
tar xzf mcs-XXXX.tar.gz
|
||||
tar xzf mono-XXXX.tar.gz
|
||||
mv mono-XXX mono
|
||||
mv mcs-XXX mcs
|
||||
cd mono
|
||||
./autogen.sh --prefix=/usr/local
|
||||
make bootstrap
|
||||
|
||||
c. Building the software from CVS
|
||||
---------------------------------
|
||||
|
||||
If you are building the software from CVS, make sure that you
|
||||
have up-to-date mcs and mono sources:
|
||||
|
||||
cvs co mono mcs
|
||||
|
||||
Then, go into the mono directory, and configure:
|
||||
|
||||
cd mono
|
||||
./autogen.sh --prefix=/usr/local
|
||||
|
||||
Depending on whether you have an existing Mono installation or
|
||||
not, you can try the following:
|
||||
|
||||
i. If you have an existing Mono installation
|
||||
-----------------------------------------
|
||||
|
||||
First verify that you have a working installation:
|
||||
|
||||
echo 'class X { static void Main () { System.Console.Write("OK");}}' > x.cs
|
||||
|
||||
Compile:
|
||||
|
||||
mcs x.cs
|
||||
|
||||
And run:
|
||||
|
||||
mono x.exe
|
||||
|
||||
If you get the output `OK' and no errors, and you have mono
|
||||
version 0.31 or later, continue. Otherwise, you can try option
|
||||
(ii) below.
|
||||
|
||||
You are ready to start your CVS upgrade. Compile with
|
||||
|
||||
make bootstrap
|
||||
|
||||
This will automatically go into the mcs/ tree and build the
|
||||
binaries there, and copy them into the appropriate
|
||||
sub-directories of mono/runtime/.
|
||||
|
||||
Now, go to step (iii) below.
|
||||
|
||||
ii. If you don't have a working Mono installation
|
||||
---------------------------------------------
|
||||
|
||||
If you don't have a working Mono installation, an obvious choice
|
||||
is to install the latest released packages of 'mono' for your
|
||||
distribution and go back to step (i).
|
||||
|
||||
You can also try a slightly more risky approach that should work
|
||||
almost all the time.
|
||||
|
||||
This works by first getting the latest version of the 'monolite'
|
||||
distribution, which contains just enough to run the 'mcs'
|
||||
compiler. You do this by
|
||||
|
||||
make get-monolite-latest
|
||||
|
||||
This should place a monolite-latest.tar.gz in the parent
|
||||
directory of the 'mono' source tree. You can then run:
|
||||
|
||||
make monolite-bootstrap
|
||||
|
||||
This will automatically gunzip and untar the tarball, place the
|
||||
files appropriately, and then complete the bootstrap.
|
||||
|
||||
iii. Testing and Installation
|
||||
------------------------
|
||||
|
||||
You can run the mono and mcs testsuites with the command:
|
||||
|
||||
make -k bootstrap-check
|
||||
|
||||
Expect to find several testsuite failures, especially in the
|
||||
mcs/ tree. As a sanity check, you can compare the failures you
|
||||
got with
|
||||
|
||||
http://go-mono.com/tests/displayTestResults.php
|
||||
|
||||
You can now install it:
|
||||
|
||||
make install
|
||||
|
||||
Failure to follow these steps will result in a broken installation.
|
||||
|
||||
iv. Other useful "bootstrap"-like facilities
|
||||
----------------------------------------
|
||||
|
||||
If you have a CVS snapshot that you keep updating periodically,
|
||||
and/or do your development in, you may try using a couple of
|
||||
specialty make targets that may be slightly faster than a
|
||||
"make bootstrap".
|
||||
|
||||
You can try a two-stage bootstrap with:
|
||||
|
||||
make faststrap
|
||||
|
||||
This assumes that you have already run a "make bootstrap" on the
|
||||
tree before.
|
||||
|
||||
If you want to avoid waiting even for that, you can try the
|
||||
riskier one-stage build:
|
||||
|
||||
make fasterstrap
|
||||
|
||||
This should be attempted only if you're sure that the sources of
|
||||
the mcs compiler itself and the sources of the libraries used by
|
||||
it have not been changed.
|
||||
|
||||
2. Using Mono
|
||||
=============
|
||||
|
||||
Once you have installed the software, you can run a few programs:
|
||||
|
||||
* runtime engine
|
||||
|
||||
mono program.exe
|
||||
or
|
||||
mint program.exe
|
||||
|
||||
* C# compiler
|
||||
|
||||
mcs program.cs
|
||||
|
||||
* CIL Disassembler
|
||||
|
||||
monodis program.exe
|
||||
|
||||
See the man pages for mono(1), mint(1), monodis(1) and mcs(2)
|
||||
for further details.
|
||||
|
||||
3. Directory Roadmap
|
||||
====================
|
||||
|
||||
doc/
|
||||
Contains the web site contents.
|
||||
|
||||
docs/
|
||||
Technical documents about the Mono runtime.
|
||||
|
||||
data/
|
||||
Configuration files installed as part of the Mono runtime.
|
||||
|
||||
mono/
|
||||
The core of the Mono Runtime.
|
||||
|
||||
metadata/
|
||||
The object system and metadata reader.
|
||||
|
||||
jit/
|
||||
The Just in Time Compiler.
|
||||
|
||||
dis/
|
||||
CIL executable Disassembler
|
||||
|
||||
cli/
|
||||
Common code for the JIT and the interpreter.
|
||||
|
||||
io-layer/
|
||||
The I/O layer and system abstraction for
|
||||
emulating the .NET IO model.
|
||||
|
||||
cil/
|
||||
Common Intermediate Representation, XML
|
||||
definition of the CIL bytecodes.
|
||||
|
||||
interp/
|
||||
Interpreter for CLI executables.
|
||||
|
||||
arch/
|
||||
Architecture specific portions.
|
||||
|
||||
man/
|
||||
|
||||
Manual pages for the various Mono commands and programs.
|
||||
|
||||
scripts/
|
||||
|
||||
Scripts used to invoke Mono and the corresponding program.
|
||||
|
||||
runtime/
|
||||
|
||||
A directory holding a pre-compiled version of the Mono
|
||||
runtime.
|
||||
|
||||
----------
|
||||
|
||||
To edit the go-mono website, see doc/README.
|
||||
@@ -1,37 +0,0 @@
|
||||
#undef HAVE_INET_PTON
|
||||
#undef HAVE_INET_ATON
|
||||
#undef HAVE_MSG_NOSIGNAL
|
||||
#undef HAVE_SOL_IP
|
||||
#undef HAVE_SOL_TCP
|
||||
#undef HAVE_SOL_IPV6
|
||||
#undef HAVE_IP_PKTINFO
|
||||
#undef HAVE_STRUCT_IP_MREQN
|
||||
#undef HAVE_STRUCT_IP_MREQ
|
||||
#undef HAVE_PTHREAD_MUTEX_TIMEDLOCK
|
||||
#undef USE_MONO_MUTEX
|
||||
#undef HAVE_LARGE_FILE_SUPPORT
|
||||
#undef HAVE_NEW_ICONV
|
||||
#undef NAME_DEV_RANDOM
|
||||
#undef HAVE_CRYPT_RNG
|
||||
#undef HAVE_BOEHM_GC
|
||||
#undef USE_INCLUDED_LIBGC
|
||||
#undef DISABLE_SHARED_HANDLES
|
||||
#undef HAVE_TIMEZONE
|
||||
#undef HAVE_TM_GMTOFF
|
||||
#undef MONO_SIZEOF_SUNPATH
|
||||
#undef MONO_ZERO_ARRAY_LENGTH
|
||||
#undef PTHREAD_POINTER_ID
|
||||
#undef HAVE_GC_H
|
||||
#undef HAVE_GC_GCJ_MALLOC
|
||||
#undef HAVE_GC_ENABLE
|
||||
#undef WITH_BUNDLE
|
||||
#undef HAVE_GETHOSTBYNAME2_R
|
||||
#undef HAVE_ICU
|
||||
#undef HAVE_KW_THREAD
|
||||
#undef HAVE_SIGNBIT
|
||||
#undef USE_MACH_SEMA
|
||||
#undef HAVE_SOCKLEN_T
|
||||
#undef HAVE_TRUNCL
|
||||
#undef HAVE_AINTL
|
||||
#undef HAVE_SCANDIR
|
||||
#undef HAVE_WORKING_SIGALTSTACK
|
||||
@@ -1,142 +1,6 @@
|
||||
#!/bin/sh
|
||||
# Run this to generate all the initial makefiles, etc.
|
||||
# Ripped off from GNOME macros version
|
||||
|
||||
DIE=0
|
||||
|
||||
srcdir=`dirname $0`
|
||||
test -z "$srcdir" && srcdir=.
|
||||
|
||||
if [ -n "$MONO_PATH" ]; then
|
||||
# from -> /mono/lib:/another/mono/lib
|
||||
# to -> /mono /another/mono
|
||||
for i in `echo ${MONO_PATH} | tr ":" " "`; do
|
||||
i=`dirname ${i}`
|
||||
if [ -n "{i}" -a -d "${i}/share/aclocal" ]; then
|
||||
ACLOCAL_FLAGS="-I ${i}/share/aclocal $ACLOCAL_FLAGS"
|
||||
fi
|
||||
if [ -n "{i}" -a -d "${i}/bin" ]; then
|
||||
PATH="${i}/bin:$PATH"
|
||||
fi
|
||||
done
|
||||
export PATH
|
||||
fi
|
||||
|
||||
(autoconf --version) < /dev/null > /dev/null 2>&1 || {
|
||||
echo
|
||||
echo "**Error**: You must have \`autoconf' installed to compile Mono."
|
||||
echo "Download the appropriate package for your distribution,"
|
||||
echo "or get the source tarball at ftp://ftp.gnu.org/pub/gnu/"
|
||||
DIE=1
|
||||
}
|
||||
|
||||
if [ -z "$LIBTOOL" ]; then
|
||||
LIBTOOL=`which glibtool 2>/dev/null`
|
||||
if [ ! -x "$LIBTOOL" ]; then
|
||||
LIBTOOL=`which libtool`
|
||||
fi
|
||||
fi
|
||||
|
||||
(grep "^AM_PROG_LIBTOOL" $srcdir/configure.in >/dev/null) && {
|
||||
($LIBTOOL --version) < /dev/null > /dev/null 2>&1 || {
|
||||
echo
|
||||
echo "**Error**: You must have \`libtool' installed to compile Mono."
|
||||
echo "Get ftp://ftp.gnu.org/pub/gnu/libtool-1.2d.tar.gz"
|
||||
echo "(or a newer version if it is available)"
|
||||
DIE=1
|
||||
}
|
||||
}
|
||||
|
||||
grep "^AM_GNU_GETTEXT" $srcdir/configure.in >/dev/null && {
|
||||
grep "sed.*POTFILES" $srcdir/configure.in >/dev/null || \
|
||||
(gettext --version) < /dev/null > /dev/null 2>&1 || {
|
||||
echo
|
||||
echo "**Error**: You must have \`gettext' installed to compile Mono."
|
||||
echo "Get ftp://alpha.gnu.org/gnu/gettext-0.10.35.tar.gz"
|
||||
echo "(or a newer version if it is available)"
|
||||
DIE=1
|
||||
}
|
||||
}
|
||||
|
||||
(automake --version) < /dev/null > /dev/null 2>&1 || {
|
||||
echo
|
||||
echo "**Error**: You must have \`automake' installed to compile Mono."
|
||||
echo "Get ftp://ftp.gnu.org/pub/gnu/automake-1.3.tar.gz"
|
||||
echo "(or a newer version if it is available)"
|
||||
DIE=1
|
||||
NO_AUTOMAKE=yes
|
||||
}
|
||||
|
||||
|
||||
# if no automake, don't bother testing for aclocal
|
||||
test -n "$NO_AUTOMAKE" || (aclocal --version) < /dev/null > /dev/null 2>&1 || {
|
||||
echo
|
||||
echo "**Error**: Missing \`aclocal'. The version of \`automake'"
|
||||
echo "installed doesn't appear recent enough."
|
||||
echo "Get ftp://ftp.gnu.org/pub/gnu/automake-1.3.tar.gz"
|
||||
echo "(or a newer version if it is available)"
|
||||
DIE=1
|
||||
}
|
||||
|
||||
if test "$DIE" -eq 1; then
|
||||
exit 1
|
||||
fi
|
||||
|
||||
if test -z "$*"; then
|
||||
echo "**Warning**: I am going to run \`configure' with no arguments."
|
||||
echo "If you wish to pass any to it, please specify them on the"
|
||||
echo \`$0\'" command line."
|
||||
echo
|
||||
fi
|
||||
|
||||
case $CC in
|
||||
xlc )
|
||||
am_opt=--include-deps;;
|
||||
esac
|
||||
|
||||
|
||||
if grep "^AM_PROG_LIBTOOL" configure.in >/dev/null; then
|
||||
if test -z "$NO_LIBTOOLIZE" ; then
|
||||
echo "Running libtoolize..."
|
||||
${LIBTOOL}ize --force --copy
|
||||
fi
|
||||
fi
|
||||
|
||||
echo "Running aclocal $ACLOCAL_FLAGS ..."
|
||||
aclocal $ACLOCAL_FLAGS || {
|
||||
echo
|
||||
echo "**Error**: aclocal failed. This may mean that you have not"
|
||||
echo "installed all of the packages you need, or you may need to"
|
||||
echo "set ACLOCAL_FLAGS to include \"-I \$prefix/share/aclocal\""
|
||||
echo "for the prefix where you installed the packages whose"
|
||||
echo "macros were not found"
|
||||
exit 1
|
||||
}
|
||||
|
||||
if grep "^AM_CONFIG_HEADER" configure.in >/dev/null; then
|
||||
echo "Running autoheader..."
|
||||
autoheader || { echo "**Error**: autoheader failed."; exit 1; }
|
||||
fi
|
||||
|
||||
echo "Running automake --gnu $am_opt ..."
|
||||
automake --add-missing --gnu $am_opt ||
|
||||
{ echo "**Error**: automake failed."; exit 1; }
|
||||
echo "Running autoconf ..."
|
||||
autoconf || { echo "**Error**: autoconf failed."; exit 1; }
|
||||
|
||||
if test -d $srcdir/libgc; then
|
||||
echo Running libgc/autogen.sh ...
|
||||
(cd $srcdir/libgc ; NOCONFIGURE=1 ./autogen.sh "$@")
|
||||
echo Done running libgc/autogen.sh ...
|
||||
fi
|
||||
|
||||
|
||||
conf_flags="--enable-maintainer-mode --enable-compile-warnings" #--enable-iso-c
|
||||
|
||||
if test x$NOCONFIGURE = x; then
|
||||
echo Running $srcdir/configure $conf_flags "$@" ...
|
||||
$srcdir/configure $conf_flags "$@" \
|
||||
&& echo Now type \`make\' to compile $PKG_NAME || exit 1
|
||||
else
|
||||
echo Skipping configure process.
|
||||
fi
|
||||
libtoolize --automake
|
||||
automake -a
|
||||
autoheader
|
||||
aclocal $ACLOCAL_FLAGS
|
||||
autoconf
|
||||
./configure $*
|
||||
@@ -1,4 +0,0 @@
|
||||
Makefile.in
|
||||
Makefile
|
||||
config
|
||||
.cvsignore
|
||||
@@ -1,12 +0,0 @@
|
||||
monodir = $(sysconfdir)/mono
|
||||
|
||||
EXTRA_DIST = config.in \
|
||||
machine.config \
|
||||
DefaultWsdlHelpGenerator.aspx \
|
||||
browscap.ini
|
||||
|
||||
mono_DATA = config \
|
||||
machine.config \
|
||||
DefaultWsdlHelpGenerator.aspx \
|
||||
browscap.ini
|
||||
|
||||
@@ -1,10 +0,0 @@
|
||||
<configuration>
|
||||
<dllmap dll="cygwin1.dll" target="@LIBC@" />
|
||||
<dllmap dll="libc" target="@LIBC@" />
|
||||
<dllmap dll="libintl" target="@INTL@" />
|
||||
<dllmap dll="libxslt.dll" target="libxslt@libsuffix@" />
|
||||
<dllmap dll="libmySQL.dll" target="libmysqlclient@libsuffix@" />
|
||||
<dllmap dll="odbc32.dll" target="libodbc@libsuffix@" />
|
||||
<dllmap dll="oci" target="clntsh" />
|
||||
<dllmap dll="db2cli" target="libdb2_36@libsuffix@"/>
|
||||
</configuration>
|
||||
@@ -1,216 +0,0 @@
|
||||
<?xml version="1.0" encoding="utf-8"?>
|
||||
|
||||
<configuration>
|
||||
|
||||
<configSections>
|
||||
<section name="appSettings" type="System.Configuration.NameValueFileSectionHandler, System, Version=1.0.5000.0, Culture=neutral, PublicKeyToken=b77a5c561934e089" />
|
||||
<section name="system.diagnostics" type="System.Diagnostics.DiagnosticsConfigurationHandler, System, Version=1.0.5000.0, Culture=neutral, PublicKeyToken=b77a5c561934e089" />
|
||||
<section name="system.runtime.remoting" type="System.Configuration.IgnoreSectionHandler, System, Version=1.0.5000.0, Culture=neutral, PublicKeyToken=b77a5c561934e089" allowLocation="false"/>
|
||||
<sectionGroup name="system.web">
|
||||
<section name="httpHandlers"
|
||||
type="System.Web.Configuration.HttpHandlersSectionHandler, System.Web, Version=1.0.5000.0, Culture=neutral, PublicKeyToken=b03f5f7f11d50a3a" />
|
||||
<section name="httpModules"
|
||||
type="System.Web.Configuration.HttpModulesConfigurationHandler, System.Web, Version=1.0.5000.0, Culture=neutral, PublicKeyToken=b03f5f7f11d50a3a" />
|
||||
<section name="machineKey"
|
||||
type="System.Web.Configuration.MachineKeyConfigHandler, System.Web, Version=1.0.5000.0, Culture=neutral, PublicKeyToken=b03f5f7f11d50a3a" />
|
||||
<section name="authentication"
|
||||
type="System.Web.Configuration.AuthenticationConfigHandler, System.Web, Version=1.0.5000.0, Culture=neutral, PublicKeyToken=b03f5f7f11d50a3a" />
|
||||
<section name="authorization"
|
||||
type="System.Web.Configuration.AuthorizationConfigHandler, System.Web, Version=1.0.5000.0, Culture=neutral, PublicKeyToken=b03f5f7f11d50a3a" />
|
||||
<section name="globalization"
|
||||
type="System.Web.Configuration.GlobalizationConfigurationHandler, System.Web, Version=1.0.5000.0, Culture=neutral, PublicKeyToken=b03f5f7f11d50a3a" />
|
||||
<section name="sessionState"
|
||||
type="System.Web.SessionState.SessionStateSectionHandler, System.Web, Version=1.0.5000.0, Culture=neutral, PublicKeyToken=b03f5f7f11d50a3a" />
|
||||
<section name="trace"
|
||||
type="System.Web.Configuration.TraceConfigurationHandler, System.Web, Version=1.0.5000.0, Culture=neutral, PublicKeyToken=b03f5f7f11d50a3a" />
|
||||
<section name="webServices"
|
||||
type="System.Web.Services.Configuration.WebServicesConfigurationSectionHandler, System.Web.Services, Version=1.0.5000.0 Culture=neutral, PublicKeyToken=b03f5f7f11d50a3a" />
|
||||
<section name="webControls"
|
||||
type="System.Web.Configuration.WebControlsSectionHandler, System.Web, Version=1.0.5000.0, Culture=neutral, PublicKeyToken=b03f5f7f11d50a3a" />
|
||||
<section name="compilation"
|
||||
type="System.Web.Configuration.CompilationConfigurationHandler, System.Web, Version=1.0.5000.0, Culture=neutral, PublicKeyToken=b03f5f7f11d50a3a" />
|
||||
<section name="httpRuntime"
|
||||
type="System.Web.Configuration.HttpRuntimeConfigurationHandler, System.Web, Version=1.0.5000.0, Culture=neutral, PublicKeyToken=b03f5f7f11d50a3a" />
|
||||
<section name="customErrors"
|
||||
type="System.Web.Configuration.CustomErrorsConfigHandler, System.Web, Version=1.0.5000.0, Culture=neutral, PublicKeyToken=b03f5f7f11d50a3a" />
|
||||
<section name="pages"
|
||||
type="System.Web.Configuration.PagesConfigurationHandler, System.Web, Version=1.0.5000.0, Culture=neutral" />
|
||||
<section name="clientTarget"
|
||||
type="System.Web.Configuration.ClientTargetSectionHandler, System.Web, Version=1.0.5000.0, Culture=neutral" />
|
||||
</sectionGroup>
|
||||
<sectionGroup name="system.net">
|
||||
<section name="settings"
|
||||
type="System.Net.Configuration.NetConfigurationHandler, System, Version=1.0.5000.0, Culture=neutral, PublicKeyToken=b77a5c561934e089" />
|
||||
<section name="connectionManagement"
|
||||
type="System.Net.Configuration.ConnectionManagementHandler, System, Version=1.0.5000.0, Culture=neutral, PublicKeyToken=b77a5c561934e089" />
|
||||
<section name="authenticationModules"
|
||||
type="System.Net.Configuration.NetAuthenticationModuleHandler, System, Version=1.0.5000.0, Culture=neutral, PublicKeyToken=b77a5c561934e089" />
|
||||
<section name="defaultProxy"
|
||||
type="System.Net.Configuration.DefaultProxyHandler, System, Version=1.0.5000.0, Culture=neutral, PublicKeyToken=b77a5c561934e089" />
|
||||
<section name="webRequestModules"
|
||||
type="System.Net.Configuration.WebRequestModuleHandler, System, Version=1.0.5000.0, Culture=neutral, PublicKeyToken=b77a5c561934e089" />
|
||||
</sectionGroup>
|
||||
<section name="system.drawing" type="System.Configuration.NameValueSectionHandler, System, Version=1.0.5000.0, Culture=neutral, PublicKeyToken=b77a5c561934e089" />
|
||||
</configSections>
|
||||
<system.net>
|
||||
<connectionManagement>
|
||||
<add address="*" maxconnection="2" />
|
||||
</connectionManagement>
|
||||
<authenticationModules>
|
||||
<add type="System.Net.BasicClient, System, Version=1.0.5000.0, Culture=neutral, PublicKeyToken=b77a5c561934e089" />
|
||||
<add type="System.Net.DigestClient, System, Version=1.0.5000.0, Culture=neutral, PublicKeyToken=b77a5c561934e089" />
|
||||
<add type="System.Net.NtlmClient, System, Version=1.0.5000.0, Culture=neutral, PublicKeyToken=b77a5c561934e089" />
|
||||
</authenticationModules>
|
||||
<defaultProxy>
|
||||
<proxy />
|
||||
<!-- bypassonlocal and proxyaddress are ok -->
|
||||
<!-- usessystemdefault is not supported -->
|
||||
</defaultProxy>
|
||||
<webRequestModules>
|
||||
<add prefix="http" type="System.Net.HttpRequestCreator, System, Version=1.0.5000.0, Culture=neutral, PublicKeyToken=b77a5c561934e089" />
|
||||
<add prefix="https" type="System.Net.HttpRequestCreator, System, Version=1.0.5000.0, Culture=neutral, PublicKeyToken=b77a5c561934e089" />
|
||||
<add prefix="file" type="System.Net.FileWebRequestCreator, System, Version=1.0.5000.0, Culture=neutral, PublicKeyToken=b77a5c561934e089" />
|
||||
</webRequestModules>
|
||||
<settings>
|
||||
<ipv6 enabled="false"/>
|
||||
</settings>
|
||||
</system.net>
|
||||
<system.web>
|
||||
<httpHandlers>
|
||||
<add verb="*" path="Trace.axd" type="System.Web.Handlers.TraceHandler, System.Web, Version=1.0.5000.0, Culture=neutral, PublicKeyToken=b03f5f7f11d50a3a" />
|
||||
<add verb="*" path="*.aspx" type="System.Web.UI.PageHandlerFactory, System.Web, Version=1.0.5000.0, Culture=neutral, PublicKeyToken=b03f5f7f11d50a3a" />
|
||||
<add verb="*" path="*.asmx" validate="false" type="System.Web.Services.Protocols.WebServiceHandlerFactory, System.Web.Services, Version=1.0.5000.0, Culture=neutral, PublicKeyToken=b03f5f7f11d50a3a" />
|
||||
<add verb="*" path="*.ashx" type="System.Web.UI.SimpleHandlerFactory, System.Web, Version=1.0.5000.0, Culture=neutral, PublicKeyToken=b03f5f7f11d50a3a" />
|
||||
<add verb="GET" path="WebResource.axd" type="System.Web.Handlers.AssemblyResourceLoader, System.Web, Version=1.0.5000.0, Culture=neutral, PublicKeyToken=b03f5f7f11d50a3a" />
|
||||
<add verb="*" path="*.asax" type="System.Web.HttpForbiddenHandler, System.Web, Version=1.0.5000.0, Culture=neutral, PublicKeyToken=b03f5f7f11d50a3a" />
|
||||
<add verb="*" path="*.ascx" type="System.Web.HttpForbiddenHandler, System.Web, Version=1.0.5000.0, Culture=neutral, PublicKeyToken=b03f5f7f11d50a3a" />
|
||||
<add verb="*" path="*.config" type="System.Web.HttpForbiddenHandler, System.Web, Version=1.0.5000.0, Culture=neutral, PublicKeyToken=b03f5f7f11d50a3a" />
|
||||
<add verb="*" path="*.cs" type="System.Web.HttpForbiddenHandler, System.Web, Version=1.0.5000.0, Culture=neutral, PublicKeyToken=b03f5f7f11d50a3a" />
|
||||
<add verb="*" path="*.rem" type="System.Runtime.Remoting.Channels.Http.HttpRemotingHandlerFactory, System.Runtime.Remoting, Culture=neutral, PublicKeyToken=b77a5c561934e089" validate="false" />
|
||||
<add verb="*" path="*.soap" type="System.Runtime.Remoting.Channels.Http.HttpRemotingHandlerFactory, System.Runtime.Remoting, Culture=neutral, PublicKeyToken=b77a5c561934e089" validate="false" />
|
||||
<add verb="GET,HEAD" path="*" type="System.Web.StaticFileHandler, System.Web, Version=1.0.5000.0, Culture=neutral, PublicKeyToken=b03f5f7f11d50a3a" />
|
||||
<add verb="*" path="*" type="System.Web.HttpMethodNotAllowedHandler, System.Web, Version=1.0.5000.0, Culture=neutral, PublicKeyToken=b03f5f7f11d50a3a" />
|
||||
</httpHandlers>
|
||||
<httpModules>
|
||||
<add name="FormsAuthentication" type="System.Web.Security.FormsAuthenticationModule, System.Web, Version=1.0.5000.0, Culture=neutral, PublicKeyToken=b03f5f7f11d50a3a" />
|
||||
<add name="UrlAuthorization" type="System.Web.Security.UrlAuthorizationModule, System.Web, Version=1.0.5000.0, Culture=neutral, PublicKeyToken=b03f5f7f11d50a3a" />
|
||||
<add name="Session" type="System.Web.SessionState.SessionStateModule, System.Web, Version=1.0.5000.0, Culture=neutral, PublicKeyToken=b03f5f7f11d50a3a" />
|
||||
<add name="OutputCache" type="System.Web.Caching.OutputCacheModule, System.Web, Version=1.0.5000.0, Culture=neutral, PublicKeyToken=b03f5f7f11d50a3a" />
|
||||
</httpModules>
|
||||
<authentication mode="Forms">
|
||||
<forms name=".MONOAUTH" loginUrl="login.aspx" protection="All" timeout="30" path="/">
|
||||
<credentials passwordFormat="Clear">
|
||||
<!--<user name="gonzalo" password="gonz"/>-->
|
||||
</credentials>
|
||||
</forms>
|
||||
</authentication>
|
||||
<machineKey validationKey="AutoGenerate" decryptionKey="AutoGenerate" validation="SHA1" />
|
||||
<globalization requestEncoding="utf-8"
|
||||
responseEncoding="utf-8"
|
||||
fileEncoding="utf-8"/>
|
||||
<!--
|
||||
culture="en-US"
|
||||
uiculture="en-US" />
|
||||
-->
|
||||
<sessionState mode="InProc" />
|
||||
<webServices>
|
||||
<protocols>
|
||||
<add name="HttpSoap"/>
|
||||
<add name="HttpPost"/>
|
||||
<add name="HttpGet"/>
|
||||
<add name="Documentation"/>
|
||||
</protocols>
|
||||
<wsdlHelpGenerator href="DefaultWsdlHelpGenerator.aspx" />
|
||||
</webServices>
|
||||
<webControls clientScriptsLocation="/web_scripts" />
|
||||
<compilation debug="false" defaultLanguage="c#">
|
||||
<compilers>
|
||||
<compiler language="cs;c#;csharp" extension=".cs" warningLevel="1" compilerOptions=""
|
||||
type="Microsoft.CSharp.CSharpCodeProvider, System, Version=1.0.5000.0, Culture=neutral, PublicKeyToken=b77a5c561934e089" />
|
||||
<compiler language="vb;vbs;visualbasic;vbscript" extension=".vb"
|
||||
warningLevel="1" compilerOptions=""
|
||||
type="Microsoft.VisualBasic.VBCodeProvider, System, Version=1.0.5000.0, Culture=neutral, PublicKeyToken=b77a5c561934e089" />
|
||||
</compilers>
|
||||
<assemblies>
|
||||
<!--<add assembly="mscorlib" /> -->
|
||||
<add assembly="System" />
|
||||
<add assembly="System.Xml" />
|
||||
<add assembly="System.Data" />
|
||||
<add assembly="System.Web" />
|
||||
<add assembly="System.Web.Services" />
|
||||
<add assembly="System.Drawing" />
|
||||
<add assembly="*" /> <!-- Add assemblies in bin directory -->
|
||||
</assemblies>
|
||||
</compilation>
|
||||
<httpRuntime executionTimeout="90"
|
||||
maxRequestLength="4096"
|
||||
useFullyQualifiedRedirectUrl="false"
|
||||
minFreeThreads="8"
|
||||
minLocalRequestFreeThreads="4"
|
||||
appRequestQueueLimit="100" />
|
||||
<pages buffer="true"
|
||||
enableSessionState="true"
|
||||
autoEventWireup="true"
|
||||
validateRequest="true" />
|
||||
<clientTarget>
|
||||
<add alias="ie5" userAgent="Mozilla/4.0 (compatible; MSIE 5.5; Windows NT 4.0)" />
|
||||
<add alias="ie4" userAgent="Mozilla/4.0 (compatible; MSIE 4.0; Windows NT 4.0)" />
|
||||
<add alias="uplevel" userAgent="Mozilla/4.0 (compatible; MSIE 4.0; Windows NT 4.0)" />
|
||||
<add alias="downlevel" userAgent="Unknown" />
|
||||
</clientTarget>
|
||||
</system.web>
|
||||
|
||||
<system.runtime.remoting>
|
||||
<application>
|
||||
<channels>
|
||||
<channel ref="http client" displayName="http client (delay loaded)" delayLoadAsClientChannel="true" />
|
||||
<channel ref="tcp client" displayName="tcp client (delay loaded)" delayLoadAsClientChannel="true" />
|
||||
</channels>
|
||||
</application>
|
||||
<channels>
|
||||
<channel id="http" type="System.Runtime.Remoting.Channels.Http.HttpChannel, System.Runtime.Remoting, Version=1.0.5000.0, Culture=neutral, PublicKeyToken=b77a5c561934e089" />
|
||||
<channel id="http client" type="System.Runtime.Remoting.Channels.Http.HttpClientChannel, System.Runtime.Remoting, Version=1.0.5000.0, Culture=neutral, PublicKeyToken=b77a5c561934e089" />
|
||||
<channel id="http server" type="System.Runtime.Remoting.Channels.Http.HttpServerChannel, System.Runtime.Remoting, Version=1.0.5000.0, Culture=neutral, PublicKeyToken=b77a5c561934e089" />
|
||||
<channel id="tcp" type="System.Runtime.Remoting.Channels.Tcp.TcpChannel, System.Runtime.Remoting, Version=1.0.5000.0, Culture=neutral, PublicKeyToken=b77a5c561934e089" />
|
||||
<channel id="tcp client" type="System.Runtime.Remoting.Channels.Tcp.TcpClientChannel, System.Runtime.Remoting, Version=1.0.5000.0, Culture=neutral, PublicKeyToken=b77a5c561934e089" />
|
||||
<channel id="tcp server" type="System.Runtime.Remoting.Channels.Tcp.TcpServerChannel, System.Runtime.Remoting, Version=1.0.5000.0, Culture=neutral, PublicKeyToken=b77a5c561934e089" />
|
||||
</channels>
|
||||
<channelSinkProviders>
|
||||
<clientProviders>
|
||||
<formatter id="soap" type="System.Runtime.Remoting.Channels.SoapClientFormatterSinkProvider, System.Runtime.Remoting, Version=1.0.5000.0, Culture=neutral, PublicKeyToken=b77a5c561934e089" />
|
||||
<formatter id="binary" type="System.Runtime.Remoting.Channels.BinaryClientFormatterSinkProvider, System.Runtime.Remoting, Version=1.0.5000.0, Culture=neutral, PublicKeyToken=b77a5c561934e089" />
|
||||
</clientProviders>
|
||||
<serverProviders>
|
||||
<formatter id="soap" type="System.Runtime.Remoting.Channels.SoapServerFormatterSinkProvider, System.Runtime.Remoting, Version=1.0.5000.0, Culture=neutral, PublicKeyToken=b77a5c561934e089" />
|
||||
<formatter id="binary" type="System.Runtime.Remoting.Channels.BinaryServerFormatterSinkProvider, System.Runtime.Remoting, Version=1.0.5000.0, Culture=neutral, PublicKeyToken=b77a5c561934e089" />
|
||||
<provider id="wsdl" type="System.Runtime.Remoting.MetadataServices.SdlChannelSinkProvider, System.Runtime.Remoting, Version=1.0.5000.0, Culture=neutral, PublicKeyToken=b77a5c561934e089" />
|
||||
</serverProviders>
|
||||
</channelSinkProviders>
|
||||
</system.runtime.remoting>
|
||||
|
||||
<appSettings>
|
||||
<!--<add key="yourkey" value="your value" /> -->
|
||||
<!--<remove key="a key defined higher in the hierarchy" /> -->
|
||||
<!--<clear/> Removes all defined settings -->
|
||||
</appSettings>
|
||||
<system.diagnostics>
|
||||
<trace autoflush="false" indentsize="4" />
|
||||
</system.diagnostics>
|
||||
<system.drawing>
|
||||
</system.drawing>
|
||||
|
||||
<strongNames>
|
||||
<pubTokenMapping>
|
||||
<!-- ECMA key -->
|
||||
<map Token="b77a5c561934e089" PublicKey="002400000480000094000000060200000024000052534131000400000100010079159977d2d03a8e6bea7a2e74e8d1afcc93e8851974952bb480a12c9134474d04062447c37e0e68c080536fcf3c3fbe2ff9c979ce998475e506e8ce82dd5b0f350dc10e93bf2eeecf874b24770c5081dbea7447fddafa277b22de47d6ffea449674a4f9fccf84d15069089380284dbdd35f46cdff12a1bd78e4ef0065d016df" />
|
||||
<!-- Microsoft (final) key -->
|
||||
<map Token="b03f5f7f11d50a3a" PublicKey="002400000480000094000000060200000024000052534131000400000100010079159977d2d03a8e6bea7a2e74e8d1afcc93e8851974952bb480a12c9134474d04062447c37e0e68c080536fcf3c3fbe2ff9c979ce998475e506e8ce82dd5b0f350dc10e93bf2eeecf874b24770c5081dbea7447fddafa277b22de47d6ffea449674a4f9fccf84d15069089380284dbdd35f46cdff12a1bd78e4ef0065d016df" />
|
||||
<!-- Microsoft (Web Service Enhancement) key -->
|
||||
<map Token="31bf3856ad364e35" PublicKey="002400000480000094000000060200000024000052534131000400000100010079159977d2d03a8e6bea7a2e74e8d1afcc93e8851974952bb480a12c9134474d04062447c37e0e68c080536fcf3c3fbe2ff9c979ce998475e506e8ce82dd5b0f350dc10e93bf2eeecf874b24770c5081dbea7447fddafa277b22de47d6ffea449674a4f9fccf84d15069089380284dbdd35f46cdff12a1bd78e4ef0065d016df" />
|
||||
<!-- IBM (DB2 Data Provider) key -->
|
||||
<map Token="7c307b91aa13d208" PublicKey="002400000480000094000000060200000024000052534131000400000100010079159977d2d03a8e6bea7a2e74e8d1afcc93e8851974952bb480a12c9134474d04062447c37e0e68c080536fcf3c3fbe2ff9c979ce998475e506e8ce82dd5b0f350dc10e93bf2eeecf874b24770c5081dbea7447fddafa277b22de47d6ffea449674a4f9fccf84d15069089380284dbdd35f46cdff12a1bd78e4ef0065d016df" />
|
||||
</pubTokenMapping>
|
||||
</strongNames>
|
||||
</configuration>
|
||||
|
||||
|
||||
@@ -0,0 +1,3 @@
|
||||
Makefile.in
|
||||
Makefile
|
||||
all-docs
|
||||
@@ -0,0 +1,16 @@
|
||||
WEB_FILES=
|
||||
c-sharp class-library contact contributing documentation download \
|
||||
faq gcc-frontend ideas index makefile passport rationale resources \
|
||||
roadmap runtime status team testing thanks tools
|
||||
|
||||
OTHERS= pending resources-pending todo
|
||||
|
||||
EXTRA_DIST = $(WEB_FILES) README $(OTHERS)
|
||||
|
||||
all-docs: $(WEB_FILES)
|
||||
cat rationale roadmap c-sharp tools class-library \
|
||||
runtime documentation download faq contributing \
|
||||
resources status > all-docs
|
||||
|
||||
webit:
|
||||
(cd web; make && make push)
|
||||
@@ -0,0 +1,17 @@
|
||||
To edit the website appearance:
|
||||
|
||||
cd doc/web
|
||||
edit template.html.in
|
||||
make
|
||||
|
||||
To edit website content:
|
||||
|
||||
cd doc (this directory)
|
||||
edit files carefully
|
||||
cd web
|
||||
make
|
||||
|
||||
To publish changes:
|
||||
|
||||
cd web
|
||||
make push
|
||||
@@ -0,0 +1,151 @@
|
||||
* MCS: The Ximian C# compiler
|
||||
|
||||
MCS began as an experiment to learn the features of C# by
|
||||
writing a large C# program. MCS is currently able to parse C#
|
||||
programs and create an internal tree representation of the
|
||||
program. MCS can parse itself.
|
||||
|
||||
Work is progressing quickly on various fronts in the C#
|
||||
compiler. Recently I started using the System.Reflection API
|
||||
to load system type definitions and avoid self-population of
|
||||
types in the compiler and dropped my internal Type
|
||||
representation in favor of using the CLI's System.Type.
|
||||
|
||||
** Phases of the compiler
|
||||
|
||||
The compiler has a number of phases:
|
||||
|
||||
<ul>
|
||||
* Lexical analyzer: hand-coded lexical analyzer that
|
||||
provides tokens to the parser.
|
||||
|
||||
* The Parser: the parser is implemented using Jay (A
|
||||
Berkeley Yacc port to Java, that I ported to C#).
|
||||
The parser does minimal work and syntax checking,
|
||||
and only constructs a parsed tree.
|
||||
|
||||
Each language element gets its own class. The code
|
||||
convention is to use an uppercase name for the
|
||||
language element. So a C# class and its associated
|
||||
information is kept in a "Class" class, a "struct"
|
||||
in a "Struct" class and so on. Statements derive
|
||||
from the "Statement" class, and Expressions from the
|
||||
Expr class.
|
||||
|
||||
* Parent class resolution: before the actual code
|
||||
generation, we need to resolve the parents and
|
||||
interfaces for interface, classe and struct
|
||||
definitions.
|
||||
|
||||
* Semantic analysis: since C# can not resolve in a
|
||||
top-down pass what identifiers actually mean, we
|
||||
have to postpone this decision until the above steps
|
||||
are finished.
|
||||
|
||||
* Code generation: nothing done so far, but I do not
|
||||
expect this to be hard, as I will just use
|
||||
System.Reflection.Emit to generate the code.
|
||||
|
||||
</ul>
|
||||
|
||||
<a name="tasks">
|
||||
** Current pending tasks
|
||||
|
||||
Simple tasks:
|
||||
|
||||
<ul>
|
||||
* Array declarations are currently being ignored,
|
||||
|
||||
* PInvoke declarations are not supported.
|
||||
|
||||
* Pre-processing is not supported.
|
||||
|
||||
* Attribute declarations and passing currently ignored.
|
||||
|
||||
* Compiler does not pass around line/col information from tokenizer for error reporting.
|
||||
|
||||
* Jay does not work correctly with `error'
|
||||
productions, making parser errors hard to point. It
|
||||
would be best to port the Bison-To-Java compiler to
|
||||
become Bison-to-C# compiler ([email protected]
|
||||
might have more information)
|
||||
</ul>
|
||||
|
||||
Critical tasks:
|
||||
|
||||
<ul>
|
||||
* Resolve "base" classes and "base" interfaces for
|
||||
classes, structs and interfaces.
|
||||
|
||||
Once this is done, we can actually do the semantic
|
||||
analysis, because otherwise we do not know who our
|
||||
parents are.
|
||||
</ul>
|
||||
|
||||
Interesting tasks:
|
||||
|
||||
<ul>
|
||||
* Finishing the JB port from Java to C#. If you are
|
||||
interested in working on this, please contact Brian
|
||||
Jepson (bjepson at oreilly d-o-t com).
|
||||
|
||||
More on JB at: <a href="http://www.cs.colorado.edu/~dennis/software/jb.html">
|
||||
http://www.cs.colorado.edu/~dennis/software/jb.html</a>
|
||||
|
||||
JB will allow us to move from the Berkeley Yacc
|
||||
based Jay to a Bison-based compiler (better error
|
||||
reporting and recovery).
|
||||
</ul>
|
||||
|
||||
** Questions and Answers
|
||||
|
||||
Q: Why not write a C# front-end for GCC?
|
||||
|
||||
A: I wanted to learn about C#, and this was an exercise in this
|
||||
task. The resulting compiler is highly object-oriented, which has
|
||||
lead to a very nice, easy to follow and simple implementation of
|
||||
the compiler.
|
||||
|
||||
I found that the design of this compiler is very similar to
|
||||
Guavac's implementation.
|
||||
|
||||
Targeting the CIL/MSIL byte codes would require to re-architecting
|
||||
GCC, as GCC is mostly designed to be used for register machines.
|
||||
|
||||
The GCC Java engine that generates Java byte codes cheats: it does
|
||||
not use the GCC backend; it has a special backend just for Java, so
|
||||
you can not really generate Java bytecodes from the other languages
|
||||
supported by GCC.
|
||||
|
||||
Q: If your C# compiler is written in C#, how do you plan on getting
|
||||
this working on a non-Microsoft environment.
|
||||
|
||||
We will do this through an implementation of the CLI Virtual
|
||||
Execution System for Unix (our JIT engine).
|
||||
|
||||
Q: Do you use Bison?
|
||||
|
||||
A: No, currently I am using Jay which is a port of Berkeley Yacc to
|
||||
Java that I later ported to C#. This means that error recovery is
|
||||
not as nice as I would like to, and for some reason error
|
||||
productions are not being caught.
|
||||
|
||||
In the future I want to port one of the Bison/Java ports to C# for
|
||||
the parser.
|
||||
|
||||
Q: How do I compile it?
|
||||
|
||||
A: Compiling MCS currently requires you to run my port of <a
|
||||
href="http://primates.ximian.com/~miguel/code/jay.cs.tar.gz">Jay to
|
||||
C#</a> on a Unix system to generate the parser, and then you need
|
||||
to use Microsoft's .NET csc.exe compiler to compile the compiler.
|
||||
|
||||
You only need to compile the compiler compiler (C code), the
|
||||
samples are Java samples that I did not port, and you do not need
|
||||
them.
|
||||
|
||||
It might be simple to port Jay.cs to Windows, but I have not tried
|
||||
this.
|
||||
|
||||
You might also want to look at the <a href="faq.html#gcc">GCC</a>
|
||||
section on the main FAQ
|
||||
@@ -0,0 +1,174 @@
|
||||
* The Class Library
|
||||
|
||||
The Class Library should be compatible with Microsoft's .NET
|
||||
implementation.
|
||||
|
||||
Please see the <a href="class-status.html">Class Status</a>
|
||||
page for a status of who is working on which classes.
|
||||
|
||||
We will write as much code as possible in C#. We may need to
|
||||
interface with code written in C to gain access to the
|
||||
functionality of libraries like libart, Gtk+, and libc.
|
||||
|
||||
** Contributing
|
||||
|
||||
We welcome contributions to the the Class Library. To get
|
||||
started, check the status page for information about which
|
||||
APIs are being worked on, and how to get in touch with
|
||||
individual maintainers.
|
||||
|
||||
If you want to work on a class, first check the <a
|
||||
href="download.html">Classes Distribution</a> to see if it is
|
||||
not implemented yet, if not, check the <a
|
||||
href="class-status.html">Class Status</a> to see if someone is
|
||||
already working on it, and maybe contact them.
|
||||
|
||||
If nobody is working on it, mail <a
|
||||
href="mailto:[email protected]">[email protected]</a>
|
||||
with the class you want to implement and CC <a
|
||||
href="mailto:[email protected]">[email protected]</a>
|
||||
|
||||
** Layout
|
||||
|
||||
The Class Library resides in the `mcs' module in the directoy
|
||||
`class'.
|
||||
|
||||
Each directory in the directory represents the assembly where
|
||||
the code belongs to, and inside each directory we divide the
|
||||
code based on the namespace they implement.
|
||||
|
||||
There are two cases when we should consider portability: when
|
||||
we are dealing with a couple of classes only that differ from
|
||||
system to system (Consider System.Net and System.IO for Win32
|
||||
and Unix). In those cases we will just place the files for
|
||||
example on <t>corlib/System/System.IO/Unix-Console.cs</t> and
|
||||
<t>corlib/System/System.IO/Win32-Console.cs</t>.
|
||||
|
||||
For classes that might differ more (for example, the
|
||||
implementation of Windows.Forms), we might have different
|
||||
directories altogether: <t>System.Windows.Forms/Win32</t>,
|
||||
<t>System.Windows.Forms/Gtk+</t> and
|
||||
<t>System.Windows.Forms/Cocoa</t>.
|
||||
|
||||
** Using existing components from GNOME.
|
||||
|
||||
Our current plan is to implement the GUI tools on top of
|
||||
Gtk+. The only obstacle here is that applications from Windows
|
||||
might expect to be able to pull the HWND property from the
|
||||
widgets and use PInvoke to call Windows functions.
|
||||
|
||||
** Class Library and Win32 dependencies.
|
||||
|
||||
There are a few spots where the Win32 foundation is exposed to
|
||||
the class library (for example, the HDC and HWND properties in
|
||||
the GDI+). Casual inspection suggests that these can be
|
||||
safely mapped to Gdk's GC and GdkWindow pointers without
|
||||
breaking anything.
|
||||
|
||||
The only drawback is that support for PInvoke of Win32 code
|
||||
won't be available. An alternate solution would be to use
|
||||
portions of Wine, or even to use Wine as our toolkit.
|
||||
|
||||
*** Initial GDI+ and WinForms implementation
|
||||
|
||||
The initial implementation will use Gtk+ as the underlying
|
||||
toolkit. Since GTK+ has already been ported to many windowing
|
||||
systems other than X (including frame buffer, Win32, and BeOS)
|
||||
its use should cover most applications for most users.
|
||||
|
||||
*** Database access
|
||||
|
||||
We will implement ADO.NET functionality by reusing <a
|
||||
href="http://www.gnome-db.org">GNOME-DB</a>. This is an ideal
|
||||
choice, since GNOME-DB was implemented precisely to provide an
|
||||
ADO-like system for GNOME.
|
||||
|
||||
*** Component Integration
|
||||
|
||||
We will provide a new namespace to use GNOME specific features
|
||||
as well as a namespace to host Bonobo interfaces and classes
|
||||
in Mono.
|
||||
|
||||
** Licensing
|
||||
|
||||
The class library will be licensed under the terms of the GNU
|
||||
LGPL. Some people have pointed out that the plain LGPL is
|
||||
troublesome for embedded use of the Mono class libraries. So
|
||||
we are considering to use the GPL with a special exception
|
||||
(like the <a
|
||||
href="http://www.gnu.org/software/classpath/classpath.html">GNU
|
||||
Classpath</a> project did.
|
||||
|
||||
The exception to the GPL would be:
|
||||
|
||||
<i>The library is distributed under the terms of the GNU General
|
||||
Public License with the following exception:
|
||||
|
||||
If you link this library against your own program, then you do not
|
||||
need to release the source code for that program. However, any
|
||||
changes that you make to the library itself, or to any native
|
||||
methods upon which the library relies, must be re-distributed in
|
||||
accordance with the terms of the GPL.</i>
|
||||
|
||||
If you are going to contribute, please keep in mind that we
|
||||
might require you to agree that Ximian might adjust the
|
||||
license to enable the use of the class libraries on embedded
|
||||
systems or to develop proprietary applications using Mono.
|
||||
|
||||
We suggest that you assign the copyright of your work to the
|
||||
GNOME Foundation or the Free Software Foundation to simplify
|
||||
defending the code in case it is used inappropiately.
|
||||
|
||||
** Class Library testing
|
||||
|
||||
We need to write regression tests that will verify
|
||||
the correctness of the class library, compiler, and JIT
|
||||
engine.
|
||||
|
||||
** Coding conventions
|
||||
|
||||
Please follow the conventions on the ECMA specification (On
|
||||
the Annex Partition) for your coding your libraries.
|
||||
|
||||
Use 8 space tabs for writing your code (hopefully we can keep
|
||||
this consistent). If you are modifying someone else's code, try
|
||||
to keep the coding style similar.
|
||||
|
||||
For a rationale on 8 space tabs, read Linus Torvald's Coding
|
||||
Style guidelines in the Linux kernel source for a rationale.
|
||||
|
||||
*** Missing implementation bits
|
||||
|
||||
If you implement a class and you are missing implementation bits,
|
||||
please put in the code the word "TODO" and a description of what
|
||||
is missing to be implemented.
|
||||
|
||||
*** Tagging buggy code
|
||||
|
||||
If there is a bug in your implementation tag the problem by using
|
||||
the word "FIXME" in the code, together with a description of the
|
||||
problem.
|
||||
|
||||
Do not use XXX or obscure descriptions, because otherwise people
|
||||
will not be able to understand what you mean.
|
||||
|
||||
*** Tagging Lame specs
|
||||
|
||||
Sometimes the specification will be lame (consider Version.ToString (fieldCount)
|
||||
where there is no way of knowing how many fields are available, making the API
|
||||
not only stupid, but leading to unreliable code).
|
||||
|
||||
In those cases, use the keyword "LAMESPEC".
|
||||
|
||||
** FAQ
|
||||
|
||||
Frequently asked questions about the class library:
|
||||
|
||||
Q: I am writing a new class that overrides one of the system classes,
|
||||
and I am getting a 1595 warning from the compiler. Should we use a
|
||||
different namespace?
|
||||
|
||||
A: There is a quick solution to the problem, you can pass the command
|
||||
line argument /nowarn:1595 and this will effectively let you use
|
||||
your implementation of the code, while overriding the ones from the
|
||||
system assemblies.
|
||||
@@ -0,0 +1,27 @@
|
||||
* Status of the various pieces of the class library
|
||||
|
||||
You can browse the status of the class library and see who has
|
||||
registered to work on what parts of the system. These list
|
||||
work-in-progress components currently.
|
||||
|
||||
<ul>
|
||||
* Per <a href="class-status/per-assembly.html">Assembly</a>
|
||||
|
||||
* Per <a href="class-status/per-completion.html">Completion</a>
|
||||
|
||||
* Per <a href="class-status/per-maintainer.html">Maintainer</a>
|
||||
|
||||
* Per <a href="class-status/per-unimplemented.html">Unimplemented</a>
|
||||
</ul>
|
||||
|
||||
A better mechanism will soon be in place, I just did this
|
||||
really quickly, as we already got one duplicated implementation.
|
||||
|
||||
You can also download the XML <a
|
||||
href="class-status/maintainers.xml">maintainers</a> file that
|
||||
contains the actual maintainers list.
|
||||
|
||||
I am sure we can generate better reports, if you are
|
||||
interested in contributing, let <a
|
||||
href="mailto:[email protected]">me</a> know.
|
||||
|
||||
@@ -0,0 +1,16 @@
|
||||
* Contacting the Mono team.
|
||||
|
||||
You can contact the Mono Team by sending e-mail to `<a
|
||||
href="mailto:[email protected]">mailto:[email protected]</a>'.
|
||||
|
||||
You can contact the general forum of discussion by sending
|
||||
e-mail to <a href="mailto:[email protected]">[email protected]</a>
|
||||
|
||||
You can contact me (Miguel de Icaza) by sending e-mail to <a
|
||||
href="mailto:[email protected]">[email protected]</a>. My web
|
||||
page is <a
|
||||
href="http://primates.ximian.com/~miguel">http://primates.ximian.com/~miguel"</a>
|
||||
|
||||
You can also <a
|
||||
href="http://www.ximian.com/about/contact.php3">reach Ximian.</a>
|
||||
|
||||
@@ -0,0 +1,53 @@
|
||||
* Contributing to the Mono project
|
||||
|
||||
There are many ways in which you can help in the Mono project:
|
||||
|
||||
<ul>
|
||||
* <b>Programmers:</b> You can work on a free
|
||||
implementation of the <a
|
||||
href="class-library.html">class libraries</a>, the
|
||||
<a href="runtime.html">runtime engine</a>, <a
|
||||
href="tools.html">the tools</a>, the <a
|
||||
href="testing.html">testing framework</a>
|
||||
|
||||
* <b>Writers:</b> You can help us with writing <a
|
||||
href="documentation.html">documentation</a>.
|
||||
</ul>
|
||||
|
||||
Those are just broad things that need to be worked on, but
|
||||
something that would help tremendously would be to help with
|
||||
small duties in the project that need to be addressed.
|
||||
|
||||
** Small tasks
|
||||
|
||||
A few smaller tasks are here, dropped in no particular order:
|
||||
<ul>
|
||||
* <b>MCS compilation process:</b> Currently MCS does
|
||||
not build with a single `make' command. This should
|
||||
be fixed (this being part of the `Joel Test' that
|
||||
software has to pass).
|
||||
|
||||
* <b>Mono/doc and web site:</b> They need to be
|
||||
packaged up in the official `distribution'
|
||||
|
||||
* <b>Enthusiasts: </b>
|
||||
</ul>
|
||||
|
||||
* Special note
|
||||
|
||||
If you have looked at Microsoft's implementation of .NET or
|
||||
their shared source code, you may not be able to contribute
|
||||
to Mono. Details will follow when we know more about this.
|
||||
|
||||
In general be careful when you are implementing free software
|
||||
and you have access to proprietary code. We need to make sure
|
||||
that we are not using someone else's copyrighted code
|
||||
accidentally.
|
||||
|
||||
Please do not use the <b>ildasm</b> program to disassemble
|
||||
proprietary code when you are planning to reimplement a class
|
||||
for Mono. If you have done this, we might not be able to use
|
||||
your code.
|
||||
|
||||
Please stick to published documentation for implementing any
|
||||
classes.
|
||||
@@ -0,0 +1,50 @@
|
||||
* Documentation
|
||||
|
||||
Although most of the concepts from Microsoft.NET can
|
||||
be applied to the completed Mono platform, we do need to
|
||||
have a complete set of free documentation written specifically
|
||||
for Mono.
|
||||
|
||||
The documentation license we have chosen is the GNU Free
|
||||
Documentation License (FDL), the standard for most documents
|
||||
in the free software world.
|
||||
|
||||
We need documentation on a number of topics:
|
||||
|
||||
<ul>
|
||||
|
||||
* The development tools (compilers, assembler tools,
|
||||
language reference, design time features).
|
||||
|
||||
* Frequently Asked Question compilations.
|
||||
|
||||
* HOWTO documents.
|
||||
|
||||
* The Class Libraries
|
||||
|
||||
* Tutorials on Mono and the specifics of running it.
|
||||
|
||||
* A guide to Mono as compared to the Microsoft.NET
|
||||
Framework SDK
|
||||
|
||||
</ul>
|
||||
|
||||
** Class Library documentation
|
||||
|
||||
When contributing to the Class Library effort, please use the
|
||||
inline XML documentation tags to document your classes so we
|
||||
can automatically generate the documentation from the class
|
||||
libraries.
|
||||
|
||||
If you provide examples, please do not embed them into the
|
||||
source code, as that will make the source code harder to read
|
||||
and maintain. Instead, put examples for your code into a
|
||||
subdirectory of the class libraries. Make your sample
|
||||
code a full standalone application that people can compile.
|
||||
Ideally the Mono documentation browser will let you edit, modify
|
||||
and run the sample programs.
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
@@ -0,0 +1,44 @@
|
||||
* Software Availability
|
||||
|
||||
The Virtual Execution System is available in package `mono'.
|
||||
Currently this contains a metadata library and the
|
||||
disassembler. Please reffer to our <a
|
||||
href="runtime.html">Runtime</a> description for more details
|
||||
on this part of the project.
|
||||
|
||||
The code for the C# compiler as well as the language error
|
||||
test suite and the class library are in the `mcs' package, we
|
||||
will move this later into `mono' itself.
|
||||
|
||||
The mcs compiler requires the `jay' parser generator, this is
|
||||
available in module `jay'.
|
||||
|
||||
** Sources
|
||||
|
||||
Currently we will be providing CVS snapshots every day, until
|
||||
we make the first tarball release.
|
||||
|
||||
<ul>
|
||||
<a name="july-14">
|
||||
<b>July 14th, 2001</b>
|
||||
<ul>
|
||||
* <a href="archive/jay.cs.tar.gz">archive/jay.cs.tar.gz</a>: Sources to Jay.
|
||||
* <a href="archive/mcs-12-Jul-2001.tar.gz">mcs-12-Jul-2001.tar.gz</a>: CVS snapshot.
|
||||
* <a href="archive/mono-0.3.tar.gz">mono-0.3.tar.gz</a>: Packaged Source Code.
|
||||
</ul>
|
||||
|
||||
<a name="july-8">
|
||||
<b>July 8th, 2001</b>
|
||||
|
||||
<ul>
|
||||
* <a href="archive/jay.cs.tar.gz">archive/jay.cs.tar.gz</a>: Sources.
|
||||
* <a href="archive/mcs-08-Jul-2001.tar.gz">mcs-08-Jul-2001.tar.gz</a>: CVS Snapshot
|
||||
* <a href="archive/mono-08-Jul-2001.tar.gz">mono-08-Jul-2001.tar.gz</a>: CVS Snapshot
|
||||
</ul>
|
||||
</ul>
|
||||
|
||||
** CVS
|
||||
|
||||
We are trying to figure out where to put our CVS repository.
|
||||
We are debating between the GNOME CVS or SourceForge. Watch
|
||||
this spot.
|
||||
@@ -0,0 +1,609 @@
|
||||
** Basics
|
||||
|
||||
Q: Is Mono the same as Microsoft's .NET initiative?
|
||||
|
||||
A: It is not.
|
||||
|
||||
.NET is a company-wide initiative at Microsoft that
|
||||
encompasses many different areas. The .NET development framework,
|
||||
Passport, Biztalk, new server products, and anything that is
|
||||
remotely connected to .NET gets the ".NET-stamping" treatment.
|
||||
Some components of Microsoft's .NET initiative have been announced
|
||||
and some others are in the works.
|
||||
|
||||
Mono is a project to implement several technologies developed by
|
||||
Microsoft that have now been submitted to the ECMA Standards Body.
|
||||
|
||||
Q: What technologies are included in Mono?
|
||||
|
||||
A: Mono contains a number of components useful for building new
|
||||
software:
|
||||
|
||||
<ul>
|
||||
* A Common Language Infrastructure (CLI) virtual
|
||||
machine that contains a class loader, Just-in-time
|
||||
compiler, and a garbage collecting runtime.
|
||||
|
||||
* A class library that can work with any language
|
||||
which works on the CLR.
|
||||
|
||||
* A compiler for the C# language. In the future we
|
||||
might work on other compilers that target the Common
|
||||
Language Runtime.
|
||||
|
||||
Windows has compilers that target the
|
||||
virtual machine from
|
||||
<a href="http://msdn.microsoft.com/net/thirdparty/default.asp#lang">a
|
||||
number of languages:</a> Managed C++, Java Script,
|
||||
Eiffel, Component Pascal, APL, Cobol, Oberon, Perl,
|
||||
Python, Scheme, Smalltalk, Standard ML, Haskell,
|
||||
Mercury and Oberon.
|
||||
</ul>
|
||||
|
||||
The CLR and the Common Type System (CTS) enables applications and
|
||||
libraries to be written in a number of languages. Classes and
|
||||
methods created in one language can be used from a different
|
||||
language.
|
||||
|
||||
This means for example that if you define a class to do algebraic
|
||||
manipulation in C#, that class can be reused from any other
|
||||
language that supports the CLI. You could create a class in C#,
|
||||
subclass it in C++ and instantiate it in an Eiffel program.
|
||||
|
||||
A single object system, threading system, class libraries, and
|
||||
garbage collection system can be shared across all these languages.
|
||||
|
||||
Q: Where can I find the specification for these technologies?
|
||||
|
||||
A: You can find the work-in-progress documentation from the T3G ECMA
|
||||
group here:
|
||||
|
||||
<a href="http://www.dotnetexperts.com">http://www.dotnetexperts.com</a>
|
||||
|
||||
Q: Will you implement the .NET Framework SDK class libraries?
|
||||
|
||||
A: Yes, we will be implementing the APIs of the .NET Framework SDK
|
||||
class libraries.
|
||||
|
||||
Q: What does Mono stand for?
|
||||
|
||||
A: Mono is the word for `Monkey' in Spanish. We like monkeys.
|
||||
|
||||
Q: When will you ship it?
|
||||
|
||||
A: We do not know when the code will be shipped. The more
|
||||
contributions we get to the project, the sooner it will ship.
|
||||
|
||||
A rough estimate is that we might be able to run our C# compiler on
|
||||
Linux by the end of the year. That means running the Windows
|
||||
Executable generated by a Microsoft .NET compiler on the Linux
|
||||
platform.
|
||||
|
||||
We expect that doing GUI applications will require more work on the
|
||||
class libraries. That could take another six months.
|
||||
|
||||
Q: How can I contribute?
|
||||
|
||||
A: Check the <a href="contributing.html">contributing</a> section.
|
||||
|
||||
Q: You guys should innovate instead of copying.
|
||||
|
||||
A: In this particular case, we see a clear advantage in the platform
|
||||
and we are interested in using the features of the CLI on open source systems.
|
||||
|
||||
We have decided that we should spend our limited resources towards
|
||||
implementing an existing specification instead of designing and
|
||||
implementing our own.
|
||||
|
||||
Designing and implementing our own would be possible, but it doesn't make
|
||||
sense to do that just because the specification comes from a
|
||||
proprietary vendor.
|
||||
|
||||
** Ximian
|
||||
|
||||
Q: Why is Ximian working on .NET?
|
||||
|
||||
A: We are interested in providing the best tools for programmers to
|
||||
develop applications for Free Operating Systems.
|
||||
|
||||
For more information, read the project <a
|
||||
href="rationale.html">rationale</a> page.
|
||||
|
||||
Q: Will Ximian be able to take on a project of this size?
|
||||
|
||||
A: Ximian will not be able to taken on the whole project on its own.
|
||||
Mono will be a free software/open source community project, that is
|
||||
the only way we can hope to implement something of this size. You
|
||||
can <a href="contributing.html">contribute</a> to this effort.
|
||||
|
||||
Q: What pieces will Ximian be working on?
|
||||
|
||||
A: We will focus on building a development and execution
|
||||
environment.
|
||||
|
||||
The idea is to get Mono to a state of that would allow
|
||||
third parties to actually be able to use it real-world development.
|
||||
|
||||
Q: Why does Ximian even care?
|
||||
|
||||
A: We like the features that the CLI and its related technologies
|
||||
bring to the table. An exciting addition to the developer toolkit.
|
||||
The goal of Mono is to bring this technology to non-Windows
|
||||
platforms (although we hope Mono will also run on Windows, for
|
||||
debugging and comparative purposes).
|
||||
|
||||
|
||||
** Licensing
|
||||
|
||||
Q: Will I be able to write proprietary applications that run with
|
||||
Mono?
|
||||
|
||||
A: Yes. We will make sure that the licensing will allow developers to
|
||||
create proprietary applications.
|
||||
|
||||
Q: What license is Mono on?
|
||||
|
||||
A: The C# Compiler is released under the terms of the GPL. The class
|
||||
libraries will be under the LGPL or the GPL with a special
|
||||
exception. The runtime libraries are under the LGPL.
|
||||
|
||||
Here is the exception for the class libraries:
|
||||
|
||||
The library is distributed under the terms of the GNU General
|
||||
Public License with the following exception:
|
||||
|
||||
If you link this library against your own program, then you do not
|
||||
need to release the source code for that program. However, any
|
||||
changes that you make to the library itself, or to any native
|
||||
methods upon which the library relies, must be re-distributed in
|
||||
accordance with the terms of the GPL.
|
||||
|
||||
This is similar in spirit to <a
|
||||
href="http://www.gnu.org/software/classpath/classpath.html">GNU
|
||||
Classpath.</a>
|
||||
|
||||
Q: Will you accept code under the XXX License?
|
||||
|
||||
A: If the XXX License is compatible with the license we use in that
|
||||
specific piece of code, then yes. If you want to use the BSD license, make
|
||||
sure you use the BSD license without the advertisement clause (The
|
||||
`Ousterhout License').
|
||||
|
||||
** Mono and .NET
|
||||
|
||||
Q: If applications use Mono, does that mean that I have to pay a service fee?
|
||||
|
||||
A: No. Mono is not related to Microsoft's initiative of
|
||||
software-as-a-service.
|
||||
|
||||
Q: If you implement .NET, will I depend on Microsoft Passport to run my software?
|
||||
|
||||
A: No. The .NET Framework is a runtime infrastructure and collection
|
||||
of class libraries. Passport may be required to access certain web
|
||||
services written for that framework, but only if the programmer
|
||||
chooses Passport as the authentication mechanism.
|
||||
|
||||
Q: Is .NET just a bunch of marketing slogans?
|
||||
|
||||
A: Although the `.NET initiative' is still quite nebulous, The .NET Framework
|
||||
has been available for some time. Mono is not an implementation of the .NET
|
||||
initiative, just the development framework.
|
||||
|
||||
Q: What is a 100% .NET application?
|
||||
|
||||
A: A `100% .NET application' is one that only uses the APIs defined
|
||||
under the System namespace and does not use PInvoke. These
|
||||
applications would in theory run unmodified on Windows, Linux,
|
||||
HP-UX, Solaris, MacOS X and others.
|
||||
|
||||
Q: But Microsoft will release a port of the real thing under the
|
||||
`Shared Source' license, why bother with anything else?
|
||||
|
||||
A: The Shared Source implementation will not be usable for commercial
|
||||
purposes. We are working towards an implementation that will grant
|
||||
a number of rights to recipients: use for any purpose,
|
||||
redistribution, modification, and redistribution of modifications.
|
||||
|
||||
This is what we call <a
|
||||
href="http://www.gnu.org/philosophy/free-sw.html">Free Software</a>
|
||||
|
||||
** Passport
|
||||
|
||||
Q: Is this a free implementation of Passport?
|
||||
|
||||
A: No. Passport is part of Microsoft's Hailstorm initiative. Mono
|
||||
is just a runtime, a compiler and a set of class libraries.
|
||||
|
||||
Q: Will the System.Web.Security.PassportIdentity class, mean
|
||||
that my software will depend on Passport?
|
||||
|
||||
A: No. That just means that applications might use that API to
|
||||
contact a Passport site.
|
||||
|
||||
As long as your application does not use Passport, you will not
|
||||
need Passport.
|
||||
|
||||
It might even be possible to implement that class with
|
||||
a set of dummy functions, or use an alternate Passport implementation.
|
||||
|
||||
We do not know at this time whether the Passport protocol is
|
||||
documented and whether we will be able to talk to
|
||||
passport.com
|
||||
|
||||
Q: What is your opinion?
|
||||
|
||||
A: You can read my personal <a href="passport.html">opinion on
|
||||
passport</a>.
|
||||
|
||||
** Mono and Windows
|
||||
|
||||
Q: Will Mono allow me to run Microsoft Office on Linux?
|
||||
|
||||
A: No, it will not. Microsoft Office is a Windows application. To
|
||||
run Windows applications on Intel Unix systems refer to <a
|
||||
href="http://www.winehq.com">the Wine Project</a>
|
||||
|
||||
** GNOME
|
||||
|
||||
Q: How is this related to GNOME?
|
||||
|
||||
A: In a number of ways:
|
||||
|
||||
* Mono will use existing
|
||||
components that have been developed for GNOME when it makes
|
||||
sense. For example on X systems, we will use Gtk+ and
|
||||
Libart to implement Winforms and the Drawing2D API.
|
||||
|
||||
For database access, we will use LibGDA (not really
|
||||
depending on GNOME, but related to).
|
||||
|
||||
* This project was born out of the need of providing improved
|
||||
tools for the GNOME community.
|
||||
|
||||
* We would like to add support to our CLR implementation to
|
||||
deal with GObjects (in GNOME 1.x, they are called
|
||||
GtkObjects), and allow Mono developers to provide GObjects
|
||||
or use and extend existing GObjects.
|
||||
|
||||
Q: Has the GNOME Foundation or the GNOME team adopted Mono?
|
||||
|
||||
A: Mono is too new to be adopted by those groups. We hope that the
|
||||
tools that we will provide will be adopted by free software
|
||||
programmers including the GNOME Foundation members and the GNOME
|
||||
project generally.
|
||||
|
||||
Q: Should GNOME programmers switch over to Mono?
|
||||
|
||||
A: Mono will not be ready even within the next six months, and a
|
||||
complete implementation is probably one year away.
|
||||
|
||||
We encourage GNOME developers to continue using the existing tools,
|
||||
libraries and components. Improvements made to GNOME will have an
|
||||
impact on Mono, as they will provide the "backend" for various
|
||||
classes.
|
||||
|
||||
Q: Will Mono include compatibility with Bonobo components?
|
||||
|
||||
A: Yes, we will provide a set of classes for implementing and using
|
||||
Bonobo components from within Mono.
|
||||
|
||||
** Mono and the Web
|
||||
|
||||
Q: Is Mono a way of running Java applets?
|
||||
|
||||
A: No.
|
||||
|
||||
** Web Services
|
||||
|
||||
Q: Is Mono just a new way of writing Web Services?
|
||||
|
||||
A: No.
|
||||
|
||||
Q: If this implements the SDK classes, will I be able to write and
|
||||
execute .NET Web Services with this?
|
||||
|
||||
A: Yes, you will.
|
||||
|
||||
When the project is finished, you will be able to use the same
|
||||
technologies that are available through the .NET Framework SDK on
|
||||
Windows to write Web Services.
|
||||
|
||||
Q: What about Soup?
|
||||
|
||||
A: Soup is a library for GNOME applications to create SOAP server and
|
||||
SOAP clients. You can browse the source code for soup using <a
|
||||
href="http://cvs.gnome.org/bonsai">GNOME's Bonsai</a>
|
||||
|
||||
Q: Can I use CORBA?
|
||||
|
||||
A: Yes. The CLI contains enough information about a class that
|
||||
exposing it to other RPC systems (like CORBA) is really simple, and
|
||||
does not even require support from an object.
|
||||
|
||||
We will be implementing CORBA interoperation as an extension to the
|
||||
Mono classes so that we can integrate with Bonobo, just like
|
||||
Microsoft provides COM interoperation classes and support
|
||||
mechanisms.
|
||||
|
||||
Q: Can I serialize my objects to other things other than XML?
|
||||
|
||||
A: Yes, although the serializing tools have not yet been planned, and
|
||||
you would probably have to implement them yourself.
|
||||
|
||||
** Development Tools
|
||||
|
||||
Q: Will it be possible to use the CLI features without using bytecodes
|
||||
or the JIT?
|
||||
|
||||
A: Yes. The CLI engine will be made available as a shared library.
|
||||
The garbage collection engine, the threading abstraction, the
|
||||
object system, the dynamic type code system and the JIT will be
|
||||
available for C developers to integreate with their applications if
|
||||
they wish to do so.
|
||||
|
||||
Q: Will you have new development tools?
|
||||
|
||||
A: Hopefully Free Software enthusiasts will contribute tools to
|
||||
improve the developer environment. These tools could be developed
|
||||
initially using Microsoft implementation of the CLI and then
|
||||
executed later with Mono.
|
||||
|
||||
** Mono and Java
|
||||
|
||||
Q: What about using Java? After all there are many languages that
|
||||
target the Java VM.
|
||||
|
||||
A: You can get very good tools for doing Java development on free
|
||||
systems right now. <a href="http://www.redhat.com">Red Hat</a> has
|
||||
contributed a <a href="http://gcc.gnu.org">GCC</a> <a
|
||||
href="http://gcc.gnu.org/java">frontend for Java</a> that can take
|
||||
Java sources or Java byte codes and generate native executables; <a
|
||||
href="http://www.transvirtual.com">Transvirtual</a> has implemented
|
||||
<a href="http://www.kaffe.org">Kaffe</a> a JIT engine for Java;
|
||||
Intel also has a Java VM called <a
|
||||
href="http://www.intel.com/research/mrl/orp">ORP</a>.
|
||||
|
||||
The JVM is not designed to be a general purpose virtual machine.
|
||||
The Common Intermediate Language (CIL), on the other hand, is
|
||||
designed to be a target for a
|
||||
wide variety of programming languages, and has a set of rules
|
||||
designed to be optimal for JITers.
|
||||
|
||||
|
||||
Q: What kind of rules make the Common Intermediate Language useful for
|
||||
JITers?
|
||||
|
||||
A: The main rule is that the stack in the CLI is not a general purpose
|
||||
stack. You are not allowed to use it for other purposes than
|
||||
computing values and passing arguments to functions or return
|
||||
values.
|
||||
|
||||
At any given call or return instruction, the types on the stack
|
||||
have to be the same independently of the flow of execution of your
|
||||
code.
|
||||
|
||||
Q: I heard that the CIL is ideal for JITing and not efficient for
|
||||
interpreters, is this the case?
|
||||
|
||||
A: The CIL is better suited to be JITed than JVM byte codes, but you
|
||||
can interpret them as trivially as you can interpret JVM byte
|
||||
codes.
|
||||
|
||||
** Extending Mono
|
||||
|
||||
Q: Would you allow other classes other than those in the
|
||||
specification?
|
||||
|
||||
A: Yes. The Microsoft class collection is very big, but it is by no
|
||||
means complete. It would be nice to have a port of `Camel' (the
|
||||
Mail API used by Evolution inspired by Java Mail) for Mono
|
||||
applications.
|
||||
|
||||
You might also want to look into implementing CORBA for Mono. Not
|
||||
only because it would be useful, but because it sounds like a fun
|
||||
thing to do, given the fact that the CLI is such a type rich
|
||||
system.
|
||||
|
||||
For more information on extending Mono, see our <a
|
||||
href="ideas.html">ideas</a> page.
|
||||
|
||||
** Mono and portability
|
||||
|
||||
Q: Will Mono only work on Linux?
|
||||
|
||||
A: Currently, we are doing our work on Linux-based systems and
|
||||
Windows. We do not expect many Linux-isms in the code, so it
|
||||
should be easy to port Mono to other UNIX variants.
|
||||
|
||||
Q: What about Mono on non X-based systems?
|
||||
|
||||
A: Our main intention at Ximian is to be able to develop GNOME
|
||||
applications with Mono, but if you are interested in providing a
|
||||
port of the Winform classes to other platforms (frame buffer or
|
||||
MacOS X for example), we would gladly integrate them, as long
|
||||
they are under a Free Software License.
|
||||
|
||||
** Reusing existing Code
|
||||
|
||||
Q: What projects will you reuse or build upon?
|
||||
|
||||
A: We want to get Mono in the hands of programmers soon. We are
|
||||
interested in reusing existing open source software.
|
||||
|
||||
Q: What about Intel's research JIT framework, ORP?
|
||||
|
||||
A: At this time, we are investigating whether we can use elements of
|
||||
ORP for Mono. ORP is a research JIT engine that has a clear
|
||||
defined API that splits the JIT from the GC system and the actual
|
||||
byte code implementation. It is a research product.
|
||||
|
||||
Q: What about using GNU Lightning?
|
||||
|
||||
A: We are also researching <a
|
||||
href="http://www.gnu.org/software/lightning/lightning.html">GNU
|
||||
Lightning</a>.
|
||||
|
||||
** Ximian and Microsoft
|
||||
|
||||
Q: Didn't Miguel de Icaza say that `Unix Sucks'?
|
||||
|
||||
A: Yes, he did, as a catch phrase in his opening remark on the Ottawa
|
||||
Linux Symposium. His talk focused on various ways to improve Unix.
|
||||
|
||||
There is a paper describing some ways to improve Unix at:
|
||||
|
||||
<a href="http://primates.ximian.com/~miguel/bongo-bong.html">
|
||||
http://primates.ximian.com/~miguel/bongo-bong.html</a>
|
||||
|
||||
Q: Didn't Ximian's Miguel work for Microsoft?
|
||||
|
||||
A: Actually, Nat Friedman (Ximian's co-founder) did work as an
|
||||
intern for Microsoft for one summer but Miguel did not.
|
||||
|
||||
Q: Did Nat and Miguel meet at Microsoft?
|
||||
|
||||
A: They met online on the Linux IRC network; They met in person for
|
||||
the first time in 1997.
|
||||
|
||||
** Mono and Microsoft
|
||||
|
||||
Q: How can you expect Mono to compete with Microsoft, wont this
|
||||
require an effort too large?
|
||||
|
||||
A: You are right. Mono will never become a reality without the help
|
||||
of other contributors. Ximian is a small company that can not
|
||||
finish Mono alone. We will be working with members of the
|
||||
community to deliver the product.
|
||||
|
||||
Q: Is Microsoft and Corel involved in the Mono implementation?
|
||||
|
||||
A: No, they are not.
|
||||
|
||||
Q: Are you writing Mono from the ECMA specs?
|
||||
|
||||
A: yes, we are writing them from the ECMA specs and the published
|
||||
materials in print about .NET
|
||||
|
||||
** Acronyms
|
||||
|
||||
Q: What is the difference between CLR (Common Language Runtime) and
|
||||
CLI (Common Language Infrastructure)?
|
||||
|
||||
A: CLI is the specification of an execution system. The Microsoft
|
||||
implementation of this specification is named CLR.
|
||||
|
||||
Unless we come up with our own acronym, we could just call ours
|
||||
also CLR, just because it would do exactly the same thing the
|
||||
Microsoft implementation does.
|
||||
|
||||
<a name="gcc">
|
||||
** Mono and GCC
|
||||
|
||||
Q: Should someone work on a GCC front-end to C#?
|
||||
|
||||
A: I would love if someone does, and we would love to help anyone that
|
||||
takes on that task, but we do not have the time or expertise to
|
||||
build a C# compiler with the GCC engine. I find it a lot more fun
|
||||
personally to work on C# on a C# compiler, which has an intrinsic
|
||||
beauty.
|
||||
|
||||
Q: Should someone make a GCC backend that will generate CIL images?
|
||||
|
||||
A: I would love to see a backend to GCC that generates CIL images. It
|
||||
would provide a ton of free compilers that would generate CIL
|
||||
code. This is something that people would want to look into
|
||||
anyways for Windows interoperation in the future.
|
||||
|
||||
Q: What about making a front-end to GCC that takes CIL images and
|
||||
generates native code?
|
||||
|
||||
A: I would love to see this, specially since GCC supports this same
|
||||
feature for Java Byte Codes. You could use the metadata library
|
||||
from Mono to read the byte codes (ie, this would be your
|
||||
"front-end") and generate the trees that get passed to the
|
||||
optimizer.
|
||||
|
||||
Ideally our implementation of the CLI will be available as a shared
|
||||
library that could be linked with your application as its runtime
|
||||
support.
|
||||
|
||||
Q: But would this work around the GPL in the GCC compiler and allow
|
||||
people to work on non-free front-ends?
|
||||
|
||||
A: People can already do this by targeting the JVM byte codes (there
|
||||
are about 130 compilers for various languages that target the JVM).
|
||||
|
||||
Q: Why are you writing a JIT engine instead of a front-end to GCC?
|
||||
|
||||
A: The JIT engine and runtime engine will be able to execute CIL
|
||||
executables generated on Windows.
|
||||
|
||||
** Mono and GNU
|
||||
|
||||
Q: Is Mono part of the GNU project?
|
||||
|
||||
A: Mono is a Ximian project, that is also part of the GNU project.
|
||||
|
||||
** Mono and Portability
|
||||
|
||||
Q: Will Mono work on other variants of Unix?
|
||||
|
||||
A: Yes. We do not expect to add any gratuitous incompatibilities.
|
||||
|
||||
Q: Will Mono run on Windows?
|
||||
|
||||
A: Hopefully yes. Currently some parts of Mono only run on Windows
|
||||
(the C# compiler is a .NET executable) and other parts have only
|
||||
been compiled on Linux, but work on Windows with Cygwin.
|
||||
|
||||
Q: Will Mono depend on GNOME?
|
||||
|
||||
A: It will depend only if you are using a particular assembly (for
|
||||
example, for doing GUI applications). If you are just interested
|
||||
in Mono for implementing a `Hello World Enterprise P2P Web
|
||||
Service', you will not need any GNOME component.
|
||||
|
||||
** Performance
|
||||
|
||||
Q: How fast will be Mono?
|
||||
|
||||
A: We can not predict the future, but a conservative estimate is that
|
||||
it would be at least `as fast as other JIT engines'.
|
||||
|
||||
Now, wishfully thinking I hope that we will ship various JITs with
|
||||
Mono just like Microsoft has done. A fast JITer when maximum
|
||||
performance is not needed, but fast load times are important; And
|
||||
an optimizing JITer that would be slower at generating code but
|
||||
produce more optimal output.
|
||||
|
||||
The CIL has some advantages over the Java byte code: it is really
|
||||
an intermediate representation and there are a number of
|
||||
restrictions on how you can emit CIL code that simplify creating
|
||||
better JIT engines.
|
||||
|
||||
For example, on the CIL the stack is not really an abstraction
|
||||
available for the code generator to use at will: it is just a way
|
||||
of creating a postfix representation of the parsed tree. At any
|
||||
given call point or return point, the contents of the stack are
|
||||
expected to contain the same object types independently of how the
|
||||
instructions was reached.
|
||||
|
||||
** Assorted questions
|
||||
|
||||
Q: You say that the CLI allows multiple languages to execute on the
|
||||
same environment. Isn't this the purpose of CORBA?
|
||||
|
||||
A: CORBA can be used for this purpose, but think of CORBA mostly as a
|
||||
inter-process (local machine or remote machine) communication
|
||||
system.
|
||||
|
||||
Think of the CLI to be targeted to interoperation of languages in
|
||||
the same process, and CORBA as being a nifty addition for
|
||||
out-of-process.
|
||||
|
||||
Sure, we have used CORBA for in-proc components, but the lack of an
|
||||
ABI for in-proc components does not translate into a good framework
|
||||
for in-proc systems
|
||||
|
||||
|
||||
|
||||
@@ -0,0 +1,9 @@
|
||||
* The GCC front-end
|
||||
|
||||
The GCC front-end will accept input in a binary file with
|
||||
codes in the Common Intermediate Language (CIL), and generate
|
||||
native code.
|
||||
|
||||
This will allow pre-compilation and full optimization to take
|
||||
place before a program is executed.
|
||||
|
||||
@@ -0,0 +1,40 @@
|
||||
* Ideas
|
||||
|
||||
Here are a few ideas of tools, classes and projects that you
|
||||
could start. More are forthcoming.
|
||||
|
||||
<a name="classes">
|
||||
** Classes
|
||||
|
||||
<ul>
|
||||
* Implement a JXTA protocol implementation:
|
||||
<a href="http://www.jxta.org">http://www.jxta.org</a>
|
||||
|
||||
* Implement a Mail API, similar to Camel or JavaMail (Camel has
|
||||
significant architecture features that are required on a real
|
||||
mailer).
|
||||
|
||||
You can check the current C
|
||||
<a href="http://cvs.gnome.org/bonsai/rview.cgi?dir=evolution%2Fcamel">
|
||||
Camel implementation</a>.
|
||||
|
||||
* Interfacing to Multimedia systems. You might want
|
||||
to look into the Quicktime API. I know <a
|
||||
href="mailto:[email protected]">Vladimir</a> has
|
||||
researched the problem before
|
||||
</ul>
|
||||
|
||||
<a name="projects">
|
||||
** Projects
|
||||
|
||||
<ul>
|
||||
* Implement an xmlStorageSystem for the CLI:
|
||||
<a href="http://www.soapware.org/xmlStorageSystem">
|
||||
http://www.soapware.org/xmlStorageSystem</a>
|
||||
|
||||
* Build a CORBA interoperability engine for the CLR.
|
||||
You do not need to do all of the work, just talking
|
||||
the protocol will get us a long way (<a
|
||||
href="http://www.omg.org">The OMG site</a> has the CORBA specs).
|
||||
|
||||
</ul>
|
||||
@@ -0,0 +1,39 @@
|
||||
** Jul 14, 2001
|
||||
|
||||
A <a
|
||||
href="http://mail.ximian.com/archives/public/mono-list/2001-July/000399.html">new
|
||||
release</a> of the
|
||||
runtime, compiler and classes has been made. Get it <a href="download.html#july-14">here</a>
|
||||
|
||||
** Jul 12, 2001
|
||||
|
||||
I keep getting questions about my opinion on Passport, even when
|
||||
Mono has <b>nothing</b> to do with it. I finally <a
|
||||
href="passport.html">wrote something.</a>
|
||||
|
||||
** Jul 9, 2001
|
||||
|
||||
Today <a href="http://www.ximian.com">Ximian</a> announced the
|
||||
launch of the Mono
|
||||
project, an effort to create a free software
|
||||
(sometimes called open source) implementation of the .NET
|
||||
Development Framework.
|
||||
|
||||
If you have questions about the
|
||||
project, please read our list of <a
|
||||
href="faq.html">Frequently Asked Questions</a>.
|
||||
|
||||
You might also want to <a href="download.html">Download the
|
||||
source</a> for our work so far.
|
||||
|
||||
You might want to <a
|
||||
href="resources.html#mailing">subscribe</a> to our mono-list
|
||||
and mono-announce-list
|
||||
|
||||
You can contact the team at: <a
|
||||
href="mailto:[email protected]">[email protected]</a>
|
||||
|
||||
** O'Reilly
|
||||
|
||||
Brian posted a story on <a
|
||||
href="http://www.oreillynet.com/dotnet">O'Reilly Network .NET</a>
|
||||
@@ -0,0 +1,231 @@
|
||||
* Microsoft Hailstorm and Passport
|
||||
|
||||
Microsoft Passport is a centralized database hosted by
|
||||
Microsoft that enhances the consumer experience with the Web
|
||||
by providing a single logon system that they can use across a
|
||||
number of participant web sites.
|
||||
|
||||
As you might know by now from our extensive <a
|
||||
href="faq.html">FAQ</a>, the Mono project has nothing to do
|
||||
with Microsoft Hailstorm or <a
|
||||
href="http://www.passport.com">Microsoft Passport.</a>
|
||||
|
||||
Still a lot of people have asked us our opinion on them.
|
||||
|
||||
** Passport
|
||||
|
||||
Passport is important not because of it being breaktrough
|
||||
technologically speaking, but because the company is in a
|
||||
position to drive most people toward being suscribers of it.
|
||||
|
||||
At the time of this writing passport is required to use the
|
||||
free mail service <a href="http://www.hotmail.com">Hotmail</a>
|
||||
to get customized support for the <a
|
||||
href="http://www.msn.com">MSN portal</a>, <a
|
||||
href="http://msdn.microsoft.com">Microsoft Developers
|
||||
Network</a> and according to the original announcement from
|
||||
Microsoft <a href="http://www.americanexpress.com">American
|
||||
Express</a> and <a href="http://www.ebay.com">EBay</a> will be
|
||||
adopting it.
|
||||
|
||||
There is already a <a
|
||||
href="http://www.passport.com/Directory/Default.asp?PPDir=C&lc=1033">Large
|
||||
list</a> of participating sites.
|
||||
|
||||
There are many current users of it and Microsoft will be
|
||||
driving more users towards Passport as it <a
|
||||
href="http://news.cnet.com/news/0-1003-200-6343275.html">integrates
|
||||
it</a> in their upcoming release of Windows.
|
||||
|
||||
Microsoft has also <a
|
||||
href="http://www.passport.com/Business/JoinPassportNetwork.asp?lc=1033">developed
|
||||
a toolkit</a> to enable current web merchants to integrate
|
||||
their services with passport.
|
||||
|
||||
To the end user, there is a clear benefit: they only have to
|
||||
log into a single network and not remember multiple passwords
|
||||
across sites on the internet. Companies that adopt passport
|
||||
will have a competition advantage over those that dont.
|
||||
Microsoft lists a list of <a
|
||||
href="http://www.passport.com/Business/Default.asp?lc=1033">benefits</a>
|
||||
to companies.
|
||||
|
||||
|
||||
** The problems of Passport
|
||||
|
||||
There are a number of concerns that different groups have over
|
||||
Passport. Sometimes I have some, sometimes I do not. But
|
||||
overall, consumers and businesses can have better solutions.
|
||||
|
||||
<ul>
|
||||
* <b>Single Point of Failure:</b> As more services and
|
||||
components depend on remote servers, functionality can
|
||||
grind to a halt if there is a failure on the
|
||||
centralized Passport system.
|
||||
|
||||
Such a failure was predicted, and we recently <a
|
||||
href="http://news.cnet.com/news/0-1005-200-6473003.html">witnessed</a>
|
||||
got a lot of people worried.
|
||||
|
||||
The outgage lasted for seven days. Think what this
|
||||
could do to your business.
|
||||
|
||||
* <b>Trust:</b> Not everyone trusts Microsoft to keep
|
||||
their information confidential. Concerns are not only
|
||||
at the corporate level policy, but also the fact that
|
||||
the source code for Microsoft products is not
|
||||
available, means that trojans or worms could be built
|
||||
into the products by malicious engineers. This is not
|
||||
unheard of, as the <a
|
||||
href="http://slashdot.org/articles/00/04/14/0619206.shtml">Microsoft
|
||||
Internet Server</a> had a trojan horse built into that
|
||||
allowed anyone that knew about this to control any
|
||||
server running IIS.
|
||||
|
||||
* <b>Security:</b> With a centralized system like
|
||||
Passport, imagine the repercussions of a malicious
|
||||
hacker gaining access to the Passport database.
|
||||
Personal information and credit card information about
|
||||
almost everyone using a computer could be stored there.
|
||||
|
||||
Hackers have already <a
|
||||
href="http://slashdot.org/articles/00/10/27/1147248.shtml">broken
|
||||
into Microsoft</a> in the past. And the company was
|
||||
unable to figure out for how long their systems had been hacked.
|
||||
</ul>
|
||||
|
||||
Microsoft might or might not realize this. The idea behind
|
||||
Passport is indeed a good one (I can start to get rid of my
|
||||
file that keeps track of the 30 logins and passwords or so
|
||||
that I use across the various services on the net myself).
|
||||
|
||||
** Alternatives to Microsoft Passport
|
||||
|
||||
An alternative to Microsoft Passport needs to take the above
|
||||
problems into consideration. Any solution of the form `We
|
||||
will just have a competing offering' will not work.
|
||||
|
||||
The system thus has to be:
|
||||
|
||||
<ul>
|
||||
* <b>Distributed:</b> The entire authentication
|
||||
system should not create an internet `blackout' in the
|
||||
case of failure.
|
||||
|
||||
* <b>Allow for multiple registrars:</b> Users should
|
||||
be able to choose a registrar (their banks, local
|
||||
phone company, service provider, Swiss bank, or any
|
||||
other entity they trust.
|
||||
|
||||
* <b>Mandate good security measures:</b> As a
|
||||
principle, only Open Source software should be used
|
||||
for servers in the registrar, and they should conform
|
||||
to a standard set of tools and software that can be
|
||||
examined by third parties.
|
||||
</ul>
|
||||
|
||||
An implementation of this protocol could use the DNS or a
|
||||
DNS-like setup to distribute the information of users with the
|
||||
possibility of replicating and caching public information
|
||||
about the user.
|
||||
|
||||
For instant messaging (another pieces of the Hailstorm bit),
|
||||
you want to use a non-centralized system like Sun's <a
|
||||
href="http://www.jxta.org">JXTA</a>. Some people mailed me to
|
||||
mention Jabber as a messaging platform.
|
||||
|
||||
It could also just use the user e-mail address as the `key' to
|
||||
choose the registrar (msn.com, hotmail.com -> passport.com;
|
||||
aol.com -> aol.passport.com; you get the idea).
|
||||
|
||||
The <a
|
||||
href="http://www.soapware.org/xmlStorageSystem">xmlStorage</a>
|
||||
idea from <a href="http://www.scripting.com">Dave Winer</a>
|
||||
could be used to store the information.
|
||||
|
||||
A toolkit for various popular web servers could be provided,
|
||||
authenticated and should be open sourced (for those of you who
|
||||
think that a binary program would give more security and would
|
||||
prevent people from tampering: you are wrong. You can always
|
||||
use a proxy system that "behaves" like the binary, and passes
|
||||
information back and forth from the real program, and snoops
|
||||
in-transit information).
|
||||
|
||||
Good cryptographers need to be involved in this problem to
|
||||
figure out the details and the possible insecure pieces of a
|
||||
proposal like this.
|
||||
|
||||
** Implementation: In short
|
||||
|
||||
To keep it short: <b>DNS, JXTA, xmlStorage.</b>
|
||||
|
||||
|
||||
** Deploying it
|
||||
|
||||
The implementation of such a system should be a pretty
|
||||
straightforward tasks once security cryptographers have
|
||||
designed such a beast.
|
||||
|
||||
The major problems are:
|
||||
|
||||
<ul>
|
||||
* <b>People might just not care:</b> In a poll to US
|
||||
citizens a couple of decades ago, it was found that
|
||||
most people did not care about the rights they were
|
||||
given by the Bill of Rights, which lead to a number of
|
||||
laws to be passed in the US that eliminated most of
|
||||
the rights people had.
|
||||
|
||||
* <b>The industry will move way too slow:</b>
|
||||
Microsoft's implementation is out in the open now: it
|
||||
is being deployed, and soon it will be insinuated to
|
||||
many, many users. The industry needs to get together
|
||||
soon if they care about this issue.
|
||||
|
||||
By the time the industry reacts, it might be too
|
||||
late.
|
||||
</ul>
|
||||
|
||||
** Passport and Mono
|
||||
|
||||
The .NET class libraries includes a Passport class that
|
||||
applications might use to authenticate with Passport. Since
|
||||
we do not have information at this point on the exact protocol
|
||||
of Passport, it is not even feasible to implement it.
|
||||
|
||||
If at some point the information is disclosed, it could be
|
||||
implemented.
|
||||
|
||||
If a competing system to Passport existed, we could probably
|
||||
hide all the authentication information to use a number of
|
||||
different passport-like systems.
|
||||
|
||||
If a user does not want to use Passport at all, he could
|
||||
always turn it off (or completely remove the class from the
|
||||
library). After all, this is free software.
|
||||
|
||||
Currently, we are too far from the point where this is a real
|
||||
issue.
|
||||
|
||||
** Disclaimer
|
||||
|
||||
This is just a group of personal thoughts of mine that I have
|
||||
placed here because I get asked this question a lot lately.
|
||||
The views of this page are not a statement from my employer
|
||||
(Ximian, Inc).
|
||||
|
||||
This is not part of Mono. We are not trying to deal with this
|
||||
problem.
|
||||
|
||||
Nat Friedman (Ximian's co-founder) has his own ideas on how a
|
||||
competing system to Passport could be designed, but I will let
|
||||
<a href="http://www.nat.org/gym">him</a> post his own story.
|
||||
|
||||
** Other Alternatives
|
||||
|
||||
Some people have pointed out <a
|
||||
href="http://www.xns.org">XNS</a>
|
||||
|
||||
Send comments to me: Miguel de Icaza (<a
|
||||
href="mailto:[email protected]">[email protected]</a>)
|
||||
|
||||
@@ -0,0 +1,14 @@
|
||||
** Microsoft and GNU and Linux.
|
||||
|
||||
Q: Does this mean that Microsoft is better than Linux?
|
||||
|
||||
A: Many of us are working on <a
|
||||
href="http://www.gnu.org/philosophy/free-sw.html">free software<a>
|
||||
and want to have an <a href="http://www.opensource.org">open
|
||||
source</a> environment that we can change, modify, improve, learn
|
||||
from, and share with others. Some of us also think that this will
|
||||
lead on the long run to better software: more efficient, faster,
|
||||
more robust and more.
|
||||
|
||||
We are willing to take good ideas from any source they come from.
|
||||
|
||||
@@ -0,0 +1,173 @@
|
||||
|
||||
* The Mono Project
|
||||
|
||||
** Background.
|
||||
|
||||
The GNOME project goal was to bring missing technologies to
|
||||
Unix and make it competitive in the current market place for
|
||||
desktop applications. We also realized early on that language
|
||||
independence was important, and that is why GNOME APIs were
|
||||
coded using a standard that allowed the APIs to be easily
|
||||
wrapped for other languages. Our APIs are available to most
|
||||
programming languages on Unix (Perl, Python, Scheme, C++,
|
||||
Objective-C, Ada).
|
||||
|
||||
Later on we decided to use better methods for encapsulating
|
||||
our APIs, and we started to use CORBA to define interfaces to
|
||||
components. We complemented it with policy and a set of
|
||||
standard GNOME interfaces for easily creating reusable,
|
||||
language independent components, controls and compound
|
||||
documents. This technology is known as <a
|
||||
href="http://www.ximian.com/tech/bonobo.php3">Bonobo</a>.
|
||||
Interfaces to Bonobo exist for C, Perl, Python, and
|
||||
Java.
|
||||
|
||||
CORBA is good when you define coarse interfaces, and most
|
||||
Bonobo interfaces are coarse. The only problem is that
|
||||
Bonobo/CORBA interfaces are not good for small interfaces.
|
||||
For example, an XML parsing Bonobo/CORBA component would be
|
||||
inefficient compared to a C API.
|
||||
|
||||
** Another explanation
|
||||
|
||||
I recently explained our motivations to Dave Winer, and he posted
|
||||
it <a
|
||||
href="http://scriptingnews.userland.com/stories/storyReader$1275">here</a>
|
||||
|
||||
** Microsoft's .NET
|
||||
|
||||
The Microsoft .NET initiative is confusing because it is a
|
||||
company wide effort that ranges from development tools to end
|
||||
user applications. .NET is a branding formative that
|
||||
has been applied to:
|
||||
|
||||
<ul>
|
||||
* The .NET development platform, a new platform for
|
||||
writing software.
|
||||
|
||||
* Web services.
|
||||
|
||||
* Microsoft Server Applications.
|
||||
|
||||
* New tools that use the new development platform.
|
||||
|
||||
* Hailstorm, the Passport centralized single-signon
|
||||
system that is being integrated into Windows XP.
|
||||
|
||||
</ul>
|
||||
|
||||
Mono is an implementation of the .NET development platform.
|
||||
|
||||
** The Common Language Infrastructure platform.
|
||||
|
||||
Microsoft has created a new development platform. The
|
||||
highlights of this new development platform are:
|
||||
|
||||
<ul>
|
||||
* A runtime environment that provides garbage
|
||||
collection, threading and a virtual machine
|
||||
specification (The Virtual Execution System, VES)
|
||||
|
||||
* A comprehensive class library.
|
||||
|
||||
* A new language, C#. Very similar to Java, C#
|
||||
allows programmers to use all the features available
|
||||
on the .NET runtime.
|
||||
|
||||
* A language specification that compilers can
|
||||
follow if they want to generate classes and code
|
||||
that can interoperate with other programming
|
||||
languages (The Common Language Specification: CLS)
|
||||
</ul>
|
||||
|
||||
The Common Language Infrastructure platform is similar to the
|
||||
goals we had in GNOME of giving language independence to
|
||||
programmers. It is more mature, documented, larger in scope,
|
||||
and has a consistent design.
|
||||
|
||||
Any API that is written using a CLS provider language can be
|
||||
used by any language that is a CLS consumer. Compilers
|
||||
generate code in a format called Common Intermediate Language
|
||||
(CIL) which is an intermediate representation of a compiled
|
||||
program and is easy to compile to native code or compiled
|
||||
using Just-in-Time (JIT) engines. The restrictions placed by
|
||||
the runtime on the CIL byte codes ensures that it is possible
|
||||
to do a good job at optimizing the code in a JIT compiler.
|
||||
|
||||
There is not really a lot of innovation in this platform: we
|
||||
have seen all of these concepts before, and we are all
|
||||
familiar with how these things work.
|
||||
|
||||
What makes the Common Language Infrastructure development
|
||||
platform interesting is that it is a good mix of technologies
|
||||
that have been nicely integrated.
|
||||
|
||||
The .NET development platform is essentially a new foundation
|
||||
for program development that gives Microsoft a room to grow
|
||||
for the coming years.
|
||||
|
||||
** ECMA standards.
|
||||
|
||||
Microsoft has submitted the
|
||||
specifications of C#, the runtime, the metadata and the
|
||||
other various bits of the .NET development platform to the
|
||||
<a href="http://www.ecma.ch">ECMA</a> for standarization.
|
||||
|
||||
You can get a copy of the specifications submitted to ECMA
|
||||
from: <a href="http://www.dotnetexperts.com/ecma">http://www.dotnetexperts.com/ecma</a>
|
||||
|
||||
** Mono: an Open Source Common Language Infrastructure implementation.
|
||||
|
||||
Ximian has begun work on Mono, a project that aims to bring
|
||||
the Common Language Infrastructure platform to free systems.
|
||||
|
||||
When the GNU project was launched, they picked the best
|
||||
operating system that was available out there, and they
|
||||
began to clone it: Unix.
|
||||
|
||||
The .NET development platform is a very rich, powerful, and
|
||||
well designed platform that would help improve the free
|
||||
software development platform. Just like the GNU project
|
||||
began to clone Unix sixteen years ago, we will be cloning the
|
||||
.NET development platform because it is a great platform to
|
||||
build on.
|
||||
|
||||
** What makes up Mono?
|
||||
|
||||
There are various pieces that will make up Mono:
|
||||
|
||||
<ul>
|
||||
* A C# compiler.
|
||||
|
||||
* The Virtual Execution System: that will have the
|
||||
Just-in-Time compiler, garbage collector, loader,
|
||||
threading engine.
|
||||
|
||||
A byte code interpreter will be provided for quickly
|
||||
porting Mono to new systems and debugging the JIT
|
||||
purposes, but it is not intended to be the ideal
|
||||
execution environment.
|
||||
|
||||
* An implemenation of the .NET class library.
|
||||
|
||||
* Visual development tools.
|
||||
|
||||
* A CIL GCC frontend.
|
||||
</ul>
|
||||
|
||||
** Why use GNOME components?
|
||||
|
||||
GNOME is an umbrella project that consists of infrastructural
|
||||
components (GUI toolkit, XML libraries, CORBA implementation,
|
||||
printing architecture, imaging system), a desktop environment,
|
||||
and productivity applications.
|
||||
|
||||
The GNOME infrastructural components can be used to quickly
|
||||
implement various pieces of the class libraries without reinventing
|
||||
the wheel, and since all those components are licensed under
|
||||
the terms of the GNU LGPL it is a perfect fit.
|
||||
|
||||
Libart will be used to implement the Drawing.2D API; Gtk+ and
|
||||
the GNOME libraries will be used to implement the WinForms
|
||||
API and of course Glib and libxml will be used in various
|
||||
places.
|
||||
@@ -0,0 +1,109 @@
|
||||
|
||||
* Resources
|
||||
|
||||
There are a number of resources available for those of you who
|
||||
want to contribute to the Mono project. Here are a few links.
|
||||
|
||||
If you want to send suggestions for links, address them to <a
|
||||
mailto="[email protected]">[email protected]</a>.
|
||||
|
||||
** ECMA Documentation.
|
||||
|
||||
You can get the documentation for the ECMA specs from a number of sites:
|
||||
<ul>
|
||||
* <a href="http://msdn.microsoft.com/net/ecma">At MSDN</a>
|
||||
* <a href="http://www.dotnetexperts.com">Dot Net Experts</a>
|
||||
* <a href="http://developer.intel.com/software/idap/ecma">Intel</a>
|
||||
* <a href="http://lightning.csse.monash.edu.au/.net/CLI">Monash University</a>
|
||||
</ul>
|
||||
|
||||
These contain specifications for the assembler, the metadata,
|
||||
byte codes supported by the CLI virtual machine, the C#
|
||||
language and the core class libraries.
|
||||
|
||||
For details on the .NET class libraries, you can
|
||||
visit the Microsoft's Developer Network:
|
||||
|
||||
<ul>
|
||||
* <a href="http://msdn.microsoft.com/library/default.asp?url=/library/en-us/cpref/html/cpref_start.asp">.NET Framework Class Library</a>
|
||||
</ul>
|
||||
|
||||
You can also get this information if you install the Beta2
|
||||
release of the .NET Framework.
|
||||
|
||||
<a name="mailing">
|
||||
** Mailing Lists
|
||||
|
||||
There are a number of mailing lists for Mono
|
||||
|
||||
<ul>
|
||||
|
||||
* <b><a
|
||||
href="http://mail.ximian.com/mailman/listinfo/mono-list">mono-list:</a></b>
|
||||
The general Mono discussion list.
|
||||
* <b><a
|
||||
href="http://mail.ximian.com/mailman/listinfo/mono-announce-list">mono-announce-list:</a></b>
|
||||
Announcements of Mono developments.
|
||||
* <b><a
|
||||
href="http://mail.ximian.com/mailman/listinfo/mono-docs-list">mono-docs-list:</a></b>
|
||||
Discussion on the documentation of Mono.
|
||||
* <b><a
|
||||
href="http://discuss.develop.com/dotnet.html">Dotnet mailing
|
||||
list at Develop.com:</a></b> The guys at Develop Mentor run
|
||||
this general purpose mailing list.
|
||||
</ul>
|
||||
|
||||
** Discussion Groups.
|
||||
|
||||
<ul>
|
||||
|
||||
* <a href="http://www.oreillynet.com">O'Reilly
|
||||
Network</a> has a <a
|
||||
href="http://www.oreillynet.com/dotnet">section devoted to
|
||||
.NET</a>
|
||||
* <a
|
||||
href="http://msdn.microsoft.com/newsgroups">MSDN</a> also
|
||||
lists various newsgroups related to .NET</ul>
|
||||
</ul>
|
||||
|
||||
** Other .NET related projects
|
||||
|
||||
There are a number of related projects to Mono:
|
||||
|
||||
<ul>
|
||||
* <a
|
||||
href="http://www.icsharpcode.net/OpenSource/SD/default.asp">Sharp
|
||||
Develop:</a> an IDE for the C# language written in C#.
|
||||
|
||||
* <a
|
||||
href="http://www.southern-storm.com.au/portable_net.html">Portable.NET:</a>
|
||||
|
||||
* <a href="http://nunit.sourceforge.net">NUnit:</a> A
|
||||
testing framework for .NET classes.
|
||||
|
||||
* <a href="http://www.kaffe.org">Kaffe:</a> A popular
|
||||
Free Software JIT engine for Java.
|
||||
|
||||
* <a href="http://www.intel.com/research/mrl/orp">ORP:</a> A research
|
||||
JIT/VM/GC system from Intel.
|
||||
</ul>
|
||||
|
||||
** GNOME Documentation
|
||||
|
||||
Documnetation on GNOME, and the GNOME APIs is available from
|
||||
the <a href="http://developer.gnome.org">developer</a> site at
|
||||
GNOME:
|
||||
|
||||
<ul>
|
||||
* <a href="http://developer.gnome.org/doc/API/">GNOME
|
||||
API documentation</a>
|
||||
|
||||
* <a href="http://developer.gnome.org/doc/books">GNOME
|
||||
Online books</a>
|
||||
|
||||
* <A
|
||||
href="http://developer.gnome.org/arch/">Architecture Overview</a>
|
||||
</ul>
|
||||
|
||||
|
||||
|
||||
@@ -0,0 +1,30 @@
|
||||
** MacOS Documentation
|
||||
|
||||
** Assembly Language Manuals online
|
||||
Intel
|
||||
MIPS
|
||||
SPARC
|
||||
|
||||
** Microsoft
|
||||
msdn.microsoft.com/net
|
||||
Research.microsoft.com
|
||||
|
||||
** Related Technologies
|
||||
|
||||
<ul>
|
||||
|
||||
* The CLI allows people to create Web Services using the SOAP
|
||||
protocol. SOAP is based on XML, XML schemas an the HTTP
|
||||
protocol.
|
||||
<ul>
|
||||
* XML specification.
|
||||
* XML Namespaces.
|
||||
* XML Schemas.
|
||||
* SOAP Specification.
|
||||
</ul>
|
||||
</ul>
|
||||
|
||||
** Compiler Information
|
||||
GCC
|
||||
GCC Sample front-end tutorial
|
||||
|
||||
@@ -0,0 +1,12 @@
|
||||
* Roadmap
|
||||
|
||||
We are working on the following three projects at Ximian:
|
||||
|
||||
The C# Compiler (mcs/mcs)
|
||||
|
||||
A .NET compatible Class Library (mcs/class)
|
||||
|
||||
The JIT/interpreter (mono)
|
||||
|
||||
|
||||
|
||||
@@ -0,0 +1,133 @@
|
||||
The MonoNet runtime
|
||||
|
||||
The MonoNet runtime will implement the JIT engine (and a byte
|
||||
code interpreter for quickly porting to new systems), the
|
||||
class loader, the garbage collector, threading system and
|
||||
metadata access libraries.
|
||||
|
||||
Currently the runtime contains the beginning of an image
|
||||
loader and metadata access entry points. Since Beta2 has been
|
||||
now released, it is possible to resume work using the ECMA
|
||||
specs and testing with Beta2-generated executables.
|
||||
|
||||
The runtime core will be implemented in C, in a library
|
||||
"libMonoVES.so".
|
||||
|
||||
** Executing MSIL/CIL images
|
||||
|
||||
The code will load an executable and map the references to
|
||||
external assemblies to our own version of the assemblies on
|
||||
GNU/Linux.
|
||||
|
||||
Our roadmap looks like this, this has been updated as of
|
||||
<b>Jul 15, 2001</b>:
|
||||
|
||||
<ul>
|
||||
|
||||
* Milestone 1: Fully read and parse all CIL byte-codes
|
||||
and metadata tokens (ie, a disassembler).
|
||||
|
||||
* Milestone 2: Complete an interpreter for CIL byte
|
||||
codes. This interpreter can be used temporarly to
|
||||
run CIL byte code on a system where no JIT is
|
||||
available.
|
||||
|
||||
* Milestone 3: Define an <i>lburg</i> instruction
|
||||
selector for the JITer for Intel. Although slower
|
||||
at JITing than a streaming JITer, it generates
|
||||
better code. The same grammar can later be used for
|
||||
the stream jitter.
|
||||
|
||||
* Milestone 4: Implement JITer.
|
||||
|
||||
* Milestone 5: Port of the JITer to non IA32 systems.
|
||||
</ul>
|
||||
|
||||
A setup similar to the Kaffe JIT engine can be used to
|
||||
layout the code to support non-IA32 architectures. Our work
|
||||
will be focused on getting a IA32 version running first.
|
||||
|
||||
The JIT engine should work on Linux and Win32, although you
|
||||
will need to install the CygWin32 development tools to get a
|
||||
Unix-like compilation environment.
|
||||
|
||||
** JIT Engine (<b>updated, Jul 14th, 2001</b>)
|
||||
|
||||
We will be using a code-generator generator approach for our
|
||||
JITer. Given the properties of CIL byte codes, we can take
|
||||
full advantage of a real instruction selector for our code
|
||||
generator.
|
||||
|
||||
There are a couple of books that deal with this technique: "A
|
||||
Retargetable C Compiler" and "Advanced Compiler Design and
|
||||
Implementation" are good references. You can also get a
|
||||
technical description of <a
|
||||
href="http://research.microsoft.com/copyright/accept.asp?path=http://www.research.microsoft.com/~drh/pubs/iburg.pdf&pub=ACM">lbrug</a>
|
||||
|
||||
Previously we had looked at a number of JIT engines and tools,
|
||||
but they would not take full advantage of the CIL properties:
|
||||
|
||||
<ul>
|
||||
* <a
|
||||
href="http://www.intel.com/research/mrl/orp/">ORP</a>
|
||||
|
||||
* <a
|
||||
href="http://www.gnu.org/software/lightning/">GNU
|
||||
Lightning</a>
|
||||
|
||||
* href="http://www.eecs.harvard.edu/~nr/toolkit/">NJ Machine
|
||||
Toolkit</a>.).
|
||||
|
||||
* VCODE.
|
||||
</ul>
|
||||
|
||||
** Garbage Collection
|
||||
|
||||
We have decided to implement a generational tracing garbage
|
||||
collector, which is very similar to the one being used by
|
||||
.NET. For an introduction to the garbage collection system
|
||||
used by Microsoft's CLR implementation, you can read this book
|
||||
on <a
|
||||
href="http://www.amazon.com/exec/obidos/ASIN/0471941484/o/qid=992556433/sr=2-1/ref=aps_sr_b_1_1/103-5866388-0492603">Garbage
|
||||
Collection.</a>
|
||||
|
||||
Although using a conservative garbage collector like Bohem's
|
||||
would work, all the type information is available at runtime,
|
||||
so we can actually implement a better collector than a
|
||||
conservative collector.
|
||||
|
||||
<ul>
|
||||
* Garbage collection list and FAQ:<br>
|
||||
<a href="http://www.iecc.com/gclist/">http://www.iecc.com/gclist/</a>
|
||||
|
||||
* The Microsoft .NET Garbage Collection Implementation:<br>
|
||||
<a href="http://msdn.microsoft.com/library/default.asp?url=/library/en-us/dnmag00/html/GCI.asp">http://msdn.microsoft.com/library/default.asp?url=/library/en-us/dnmag00/html/GCI.asp</a>
|
||||
<a href="http://msdn.microsoft.com/library/default.asp?url=/library/en-us/dnmag00/html/GCI.asp">http://msdn.microsoft.com/library/default.asp?url=/library/en-us/dnmag00/html/GCI.asp</a>
|
||||
</ul>
|
||||
|
||||
** Useful links
|
||||
|
||||
Paolo Molaro found a few interesting links:
|
||||
|
||||
<ul>
|
||||
|
||||
* On compilation of stack-based languages:<br>
|
||||
<a href="http://www.complang.tuwien.ac.at/projects/rafts.html">
|
||||
http://www.complang.tuwien.ac.at/projects/rafts.html</a>
|
||||
|
||||
* A paper on fast JIT compilation of a stack-based language:<br>
|
||||
<a href="http://www.research.microsoft.com/~cwfraser/pldi99codegen.pdf">
|
||||
http://www.research.microsoft.com/~cwfraser/pldi99codegen.pdf</a>
|
||||
|
||||
* Vmgen generates much of the code for efficient virtual machine (VM)
|
||||
interpreters from simple descriptions of the VM instructions:<br>
|
||||
<a href="http://www.complang.tuwien.ac.at/anton/vmgen/">
|
||||
http://www.complang.tuwien.ac.at/anton/vmgen</a>
|
||||
</ul>
|
||||
|
||||
** PInvoke
|
||||
|
||||
PInvoke will be supported, and will be used to wrap Unix API
|
||||
calls, these in turn are required for reusing some of the
|
||||
GNOME libraries that will reduce the work we have to do to
|
||||
deliver a complete class library.
|
||||
@@ -0,0 +1,44 @@
|
||||
* Project Status
|
||||
|
||||
|
||||
Pieces of Mono that have been implemented:
|
||||
|
||||
<ul>
|
||||
* C# compiler: The C# parser can now generate parse
|
||||
trees. Next up: semantic analysis, compiler lever optimizations
|
||||
and code generation.
|
||||
|
||||
* Metadata library: Can currently parse and load
|
||||
information from .NET modules (executables and DLL
|
||||
files). It has a few bugs outstanding (table
|
||||
dimension computation is incorrect).
|
||||
|
||||
Parsing of Exception tables is missing.
|
||||
|
||||
It lacks Token to (Table, Index) mapping, but this
|
||||
should be trivial to write.
|
||||
|
||||
* Disassembler: Can disassemble .NET modules. Still
|
||||
lacking exception handling as well as useful debugging
|
||||
tools (hex dumping, token dumping).
|
||||
|
||||
* Class Libraries: Only a few classes have been
|
||||
implemented.
|
||||
</ul>
|
||||
|
||||
Tasks on the critical path:
|
||||
|
||||
<ul>
|
||||
* Method and Field resolution.
|
||||
|
||||
* Bytecode interpreter: Not implemented yet.
|
||||
|
||||
* Simple JIT: Not implemented yet.
|
||||
|
||||
* Garbage collection engine.
|
||||
</ul>
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
@@ -0,0 +1,21 @@
|
||||
* Testing
|
||||
|
||||
All classes in Mono libraries should have comprehensive unit test
|
||||
suites to go with them. Unit testing is a software engineering
|
||||
methodology that makes it easier to build correct code. Every
|
||||
method in every class should have a set of tests to verify
|
||||
that they work correctly. Mono also needs a testing framework
|
||||
to make it easy to write and run lots of tests.
|
||||
|
||||
Try <a href="http://nunit.sourceforge.net">NUnit</a>
|
||||
|
||||
Why do unit testing? It becomes simple to run automated tests
|
||||
for the whole library. Unit tests are a safety net - you can
|
||||
change part of the code and verify that you haven't broken
|
||||
anything. Ideally, tests are written before the actual library
|
||||
code itself. And every time a bug is discovered, a test should
|
||||
be written to demonstrate the bug and its fix. Then, if
|
||||
you ever reintroduce the bug, you will know immediately. For
|
||||
more info, read <a
|
||||
href="http://nunit.sourceforge.net/doc/testinfected/testing.html">
|
||||
JUnit Test Infected: Programmers Love Writing Tests</a>.
|
||||
@@ -0,0 +1,6 @@
|
||||
* Thanks
|
||||
|
||||
We would like to thank Tim O'Reilly, Brian Jepson and Nathan
|
||||
Torkington for their help.
|
||||
|
||||
Dave Winer for provided interesting comments and a to read.
|
||||
@@ -0,0 +1,25 @@
|
||||
* Tools
|
||||
|
||||
We need a number of tools to make people productive using a
|
||||
Mono-based solution. Some of these tools can be developed on
|
||||
Windows before Mono is fully finished.
|
||||
|
||||
** Debugger
|
||||
|
||||
We will need a debugging API to debug CLI applications and
|
||||
then a debugger component that can be used in an IDE
|
||||
environment.
|
||||
|
||||
** Integrated Development Environment
|
||||
|
||||
There is already a project to create a C# development
|
||||
environment (SharpDevelop). People could work with the
|
||||
SharpDevelop hackers to produce a unified development environment.
|
||||
|
||||
** Help Browser
|
||||
|
||||
We need a good help browser that can be used to browse all the
|
||||
Mono documentation.
|
||||
|
||||
|
||||
|
||||
@@ -0,0 +1,2 @@
|
||||
*.src
|
||||
*.html
|
||||
@@ -0,0 +1,17 @@
|
||||
0,Home,index.html,index.src
|
||||
1,FAQ,faq.html,faq.src
|
||||
0,Mono,rationale.html,rationale.src
|
||||
1,Runtime,runtime.html,runtime.src
|
||||
1,Classes,class-library.html,class-library.src
|
||||
2,Class Status,class-status.html,class-status.src
|
||||
1,C# Compiler,c-sharp.html,c-sharp.src
|
||||
1,Status,status.html,status.src
|
||||
0,Contributing,contributing.html,contributing.src
|
||||
1,Documentation,documentation.html,documentation.src
|
||||
1,Test Suite,testing.html,testing.src
|
||||
1,Tools,tools.html,tools.src
|
||||
0,Download,download.html,download.src
|
||||
0,Resources,resources.html,resources.src
|
||||
1,Ideas,ideas.html,ideas.src
|
||||
1,Passport,passport.html,passport.src
|
||||
0,Contact,contact.html,contact.src
|
||||
@@ -0,0 +1,24 @@
|
||||
#!/usr/bin/perl
|
||||
$q = 1;
|
||||
|
||||
while (<>){
|
||||
chop;
|
||||
if (/^\* (.*)$/){
|
||||
print "<h1>$1</h1>\n";
|
||||
} elsif (/^\*\* (.*)$/) {
|
||||
print "<h2>$1</h2>\n";
|
||||
} elsif (/^\*\*\* (.*)$/) {
|
||||
print "<h3>$1</h3>\n";
|
||||
} elsif (/^$/) {
|
||||
print "<p>\n";
|
||||
} elsif (/^\t\t\* (.*)$/) {
|
||||
print "<li>$1\n";
|
||||
} elsif (/^Q: (.*)$/){
|
||||
print "<p><a name=\"q$q\"><b>Question $q:</b> $1\n";
|
||||
$q++;
|
||||
} elsif (/^A: (.*)$/){
|
||||
print "$1\n";
|
||||
} else {
|
||||
print "$_\n";
|
||||
}
|
||||
}
|
||||
|
Before Width: | Height: | Size: 4.8 KiB After Width: | Height: | Size: 4.8 KiB |
|
Before Width: | Height: | Size: 4.5 KiB After Width: | Height: | Size: 4.5 KiB |
|
After Width: | Height: | Size: 3.3 KiB |
|
Before Width: | Height: | Size: 4.4 KiB After Width: | Height: | Size: 4.4 KiB |
|
Before Width: | Height: | Size: 49 B After Width: | Height: | Size: 49 B |
|
Before Width: | Height: | Size: 156 B After Width: | Height: | Size: 157 B |
@@ -0,0 +1,39 @@
|
||||
SOURCES= \
|
||||
../contributing \
|
||||
../class-library \
|
||||
../class-status \
|
||||
../contact \
|
||||
../c-sharp \
|
||||
../documentation \
|
||||
../download \
|
||||
../faq \
|
||||
../gcc-frontend \
|
||||
../index \
|
||||
../ideas \
|
||||
../passport \
|
||||
../rationale \
|
||||
../resources \
|
||||
../roadmap \
|
||||
../runtime \
|
||||
../status \
|
||||
../testing \
|
||||
../tools
|
||||
|
||||
all:
|
||||
-mkdir site
|
||||
for i in $(SOURCES); do \
|
||||
perl htmlify $$i > `basename $$i`.src; \
|
||||
done;
|
||||
|
||||
perl process.pl commands template.html.in .
|
||||
|
||||
clean:
|
||||
for i in $(SOURCES); do \
|
||||
rm -f `basename $$i`.src `basename $$i`.html; \
|
||||
done;
|
||||
|
||||
push:
|
||||
scp *.html [email protected]:/web/cvsmodules/mono
|
||||
|
||||
push2:
|
||||
scp *.html primates:public_html/xxx
|
||||
@@ -0,0 +1,73 @@
|
||||
#!/usr/bin/perl
|
||||
#
|
||||
# Author:
|
||||
# Sean MacIsaac
|
||||
#
|
||||
|
||||
use strict;
|
||||
|
||||
my $full_expand = 1;
|
||||
my @template;
|
||||
my $n;
|
||||
|
||||
if ($#ARGV != 2) {
|
||||
print "process.pl command_file template_file directory_prefix\n";
|
||||
exit ();
|
||||
}
|
||||
|
||||
my $menu = "";
|
||||
|
||||
open COMMANDS, $ARGV[0] || die "Can not open $ARGV[0]";
|
||||
while (<COMMANDS>) {
|
||||
chop;
|
||||
my @command = split /,/;
|
||||
if ($command[0] != -1) {
|
||||
$menu .= "\t\t";
|
||||
if ($command[0] == 0){
|
||||
$menu .= "<tr><td valign=\"top\" class=\"navi\"><a class=\"navi\"";
|
||||
} else {
|
||||
$menu .= "<tr><td valign=\"top\" class=\"subnavi\">  <a class=\"subnavi\"";
|
||||
}
|
||||
$menu .= "HREF=\"$command[2]\">$command[1]</A></td></tr>\n\n";
|
||||
}
|
||||
}
|
||||
close COMMANDS;
|
||||
|
||||
open TEMPLATE, $ARGV[1] || die "Can not open $ARGV[1]";
|
||||
while (<TEMPLATE>) {
|
||||
push @template, $_;
|
||||
}
|
||||
close TEMPLATE;
|
||||
|
||||
open COMMANDS, $ARGV[0] || die "Can not open $ARGV[0]";
|
||||
while (<COMMANDS>) {
|
||||
chop;
|
||||
my @command = split /,/;
|
||||
|
||||
$n = $ARGV[2] . "/" . $command[2];
|
||||
open OUTPUT, ">" . $n || die "Can not create $n";
|
||||
|
||||
my $content = "";
|
||||
open INPUT, $command[3] || die "Can not open $command[3]";
|
||||
while (<INPUT>) {
|
||||
$content .= $_;
|
||||
}
|
||||
close INPUT;
|
||||
|
||||
my $line;
|
||||
my $temp;
|
||||
my $tit;
|
||||
my $title;
|
||||
|
||||
$tit = $command[1];
|
||||
foreach $line (@template) {
|
||||
$temp = $line;
|
||||
$title = "$tit / Mono";
|
||||
$temp =~ s/#TITLE#/$title/;
|
||||
$temp =~ s/#CONTENT#/$content/;
|
||||
$temp =~ s/#MENU#/$menu/;
|
||||
print OUTPUT $temp;
|
||||
}
|
||||
|
||||
close OUTPUT;
|
||||
}
|
||||
@@ -0,0 +1,78 @@
|
||||
<html>
|
||||
<head>
|
||||
<title>#TITLE#</title>
|
||||
<style type="text/css">
|
||||
<!--
|
||||
body { font-family: "trebuchet ms", lucida, verdana, helvetica;
|
||||
background-image: url("images/bgsquares.gif");
|
||||
background-attachment: fixed; }
|
||||
body, td, table { font-family: "trebuchet ms", lucida, verdana, helvetica;
|
||||
font-size: 12px; }
|
||||
|
||||
.navi { font-size: 14px; background: #444444; }
|
||||
.subnavi { font-size: 12px; }
|
||||
.footnote { font-size: 10px; color: #aaaaaa; }
|
||||
|
||||
a.navi { color: #ffffff; text-decoration: none; font-weight: bold; }
|
||||
a.navi:visited { color: #cccccc; }
|
||||
a.navi:hover { color: #ee9900; text-decoration: underline; }
|
||||
|
||||
a.subnavi { color: #ffffff; text-decoration: none; font-weight: bold; }
|
||||
a.subnavi:visited { color: #cccccc; }
|
||||
a.subnavi:hover { color: #ee9900; text-decoration: underline; }
|
||||
|
||||
// -->
|
||||
</style>
|
||||
</head>
|
||||
<body bgcolor="#555555" text="#000000">
|
||||
|
||||
<table cellpadding="0" cellspacing="0" border="0" width="100%">
|
||||
<tr>
|
||||
<td><img src="images/pixel.gif"></td><!-- left border -->
|
||||
<td colspan="4">
|
||||
<a href="/"><img src="images/mono.gif" border="0"></a></td>
|
||||
<td><img src="images/pixel.gif"></td><!-- right border -->
|
||||
</tr>
|
||||
<tr>
|
||||
<td><img src="images/pixel.gif" width="1" height="1"></td>
|
||||
<td colspan="3" bgcolor="black"><img src="images/pixel.gif" height="2"></td>
|
||||
<td bgcolor="black"><img src="images/pixel.gif" width="1"></td>
|
||||
<td><img src="images/pixel.gif"></td>
|
||||
</tr>
|
||||
<tr>
|
||||
<td width="100"><img src="images/pixel.gif"></td>
|
||||
<td valign="top">
|
||||
<table cellpadding="2" valign="top" cellspacing="0" border="0">
|
||||
#MENU#
|
||||
</table>
|
||||
</td>
|
||||
<td bgcolor="black" width="1"><img src="images/pixel.gif" width="1"></td>
|
||||
<td bgcolor="white">
|
||||
<table cellpadding="16">
|
||||
<tr><td>
|
||||
#CONTENT#
|
||||
</td></tr>
|
||||
</table>
|
||||
</td>
|
||||
<td bgcolor="black"><img src="images/pixel.gif" width="1"></td>
|
||||
<td width="100"><img src="images/pixel.gif"></td>
|
||||
</tr>
|
||||
<tr>
|
||||
<td colspan="2">
|
||||
<img src="images/pixel.gif"></td>
|
||||
<td colspan="2" bgcolor="black"><img src="images/pixel.gif" height="1"></td>
|
||||
<td bgcolor="black"><img src="images/pixel.gif" width="1"></td>
|
||||
<td><img src="images/pixel.gif"></td>
|
||||
</tr>
|
||||
|
||||
<td colspan="2"></td>
|
||||
<td colspan="2" align="center">
|
||||
<a class="footnote" href="mailto: [email protected]">[email protected]</a>
|
||||
</td>
|
||||
<td colspan="2"></td>
|
||||
</tr>
|
||||
</tr>
|
||||
</table>
|
||||
|
||||
</body>
|
||||
</html>
|
||||
@@ -1,2 +0,0 @@
|
||||
Makefile
|
||||
Makefile.in
|
||||
@@ -1,15 +0,0 @@
|
||||
EXTRA_DIST = \
|
||||
aot-compiler.txt \
|
||||
assembly-bundle \
|
||||
embedded-api \
|
||||
exceptions \
|
||||
gc-issues \
|
||||
jit-thoughts \
|
||||
jit-trampolines \
|
||||
local-regalloc.txt \
|
||||
mini-doc.txt \
|
||||
mini-porting.txt \
|
||||
opcode-decomp.txt \
|
||||
object-layout \
|
||||
unmanaged-calls \
|
||||
stack-alignment
|
||||
@@ -1,525 +0,0 @@
|
||||
|
||||
Here "abc" stays for "array bounds check", or "array bound checks", or
|
||||
some combination of the two...
|
||||
|
||||
------------------------------------------------------------------------------
|
||||
USAGE
|
||||
|
||||
Simply use the "abcrem" optimization invoking mono.
|
||||
|
||||
To see if bound checks are actually removed, use "mono -v" and grep for
|
||||
"ARRAY-ACCESS" in the output, there should be a message for each check
|
||||
that has been removed.
|
||||
|
||||
To trace the algorithm execution, use "-v -v -v", and be prepared to be
|
||||
totally submersed by debugging messages...
|
||||
|
||||
------------------------------------------------------------------------------
|
||||
EFFECTIVENESS
|
||||
|
||||
The abc removal code can now always remove bound checks from "clean"
|
||||
array scans in loops, and generally anyway there are clear conditions that
|
||||
already state that the index is "safe".
|
||||
|
||||
To be clearer, and give an idea of what the algorithm can and cannot do
|
||||
without describing it in detail... keep in mind that only "redunant" checks
|
||||
cannot be removed. By "redundant", I mean "already explicitly checked" in
|
||||
the method code.
|
||||
|
||||
Unfortunately, analyzing complex expressions is not so easy (see below for
|
||||
details), so the capabilities of this "abc remover" are limited.
|
||||
|
||||
These are simple guidelines:
|
||||
- Only simple comparisons between variables are understood.
|
||||
- Only increment-decrement operations are handled.
|
||||
- "switch" statements are not yet handled.
|
||||
|
||||
This means that code like this will be handled well:
|
||||
|
||||
for (int i = 0; i < a.Length; i++) {
|
||||
a [i] = .....
|
||||
}
|
||||
|
||||
The "i" variable could be declared out of the "for", the "for" could be a
|
||||
"while", and maybe even implemented with "goto", the array could be scanned
|
||||
in reverse order, and everything would still work.
|
||||
I could have a temporary variable storing the array length, and check on
|
||||
it inside the loop, and the abc removal would still occurr, like this:
|
||||
|
||||
int i = 0;
|
||||
int l = a.Length;
|
||||
while ( i < l ) {
|
||||
a [i] ......
|
||||
}
|
||||
|
||||
or this:
|
||||
|
||||
int l = a.Length;
|
||||
for (int i = l; i > 0; i--) {
|
||||
a [i] = .....
|
||||
}
|
||||
|
||||
|
||||
But just something like this
|
||||
|
||||
for (int i = 0; i < (a.Length -1); i++) .....
|
||||
|
||||
or like this
|
||||
|
||||
for (int i = 0; i < a.Length; i += 2) .....
|
||||
|
||||
would not work, the check would stay there, sorry :-(
|
||||
|
||||
Just to make you understand how things are tricky... this would work!
|
||||
|
||||
int limit = a.Length - 1;
|
||||
for (int i = 0; i < limit; i++) {
|
||||
a [i] = .....
|
||||
}
|
||||
|
||||
|
||||
A detailed explanation of the reason why things are done like this is given
|
||||
below...
|
||||
|
||||
All in all, the compiling phase is generally non so fast, and the abc removed
|
||||
are so few, that there is hardly a performance gain in using the "abcrem"
|
||||
flag in the compiling options for short programs like the ones I tried.
|
||||
|
||||
Anyway, various bound checks *are* removed, and this is good :-)
|
||||
|
||||
------------------------------------------------------------------------------
|
||||
THE ALGORITHM
|
||||
|
||||
Unfortunately, this time google has not been my friend, and I did not
|
||||
find an algorithm I could reasonably implement in the time frame I had,
|
||||
so I had to invent one. The two cleasses of algorithms I found were either
|
||||
simply an analisys of variable ranges (which is not very different from
|
||||
what I'm doing), or they were based on type systems richer than the .NET
|
||||
one (particularly, they required integer types to have an explicit range,
|
||||
which could be related to array ranges). By the way, it seems that the
|
||||
gcj people do not have an array bound check removal optimization (they
|
||||
just have a compiler flag that unconditionally removes all the checks,
|
||||
but this is not what we want).
|
||||
|
||||
This array bound check removal (abc removal) algorithm is based on
|
||||
symbolic execution concepts. I handle the removal like an "unreachable
|
||||
code elimination" (and in fact the optimization could be extended to remove
|
||||
also other unreachable sections of code, due to branches that "always go the
|
||||
same way").
|
||||
|
||||
In symbolic execution, variables do not have actual (numeric) values, but
|
||||
instead symbolic expressions (like "a", or "x+y").
|
||||
Also, branch conditions are handled like symbolic conditions (like "i<k"),
|
||||
which state relations between variable values.
|
||||
|
||||
The SSA representation inside mini is somewhat close to a symbolic
|
||||
representation of the execution of the compiled method.
|
||||
Particularly, the representation of variable values is exactly a symbolic
|
||||
one. It is enough to find all CEE_STIND_* instructions which store to a
|
||||
local variable, and their second argument is exactly the variable value.
|
||||
Actually, "cfg->vars [<variable-index>]->def" should contain exactly
|
||||
those store instructions, and the "get_variable_value_from_ssa_store"
|
||||
function extracts the variable value from there.
|
||||
|
||||
On the other hand, the conditions under which each basic block is
|
||||
executed are not made fully explicit.
|
||||
|
||||
However, it is not difficult to make them so.
|
||||
Each BB that has more than one exit BB, in practice must end either with
|
||||
a conditional branch instruction or with a switch instruction.
|
||||
In the first case, the BB has exactly two exit BBs, and their execution
|
||||
conditions are easy to get from the condition of the branch (see the
|
||||
"get_relation_from_branch_instruction" function, and expecially the end of
|
||||
"analyze_block" in abcremoval.c.
|
||||
If there is a switch, the jump condition of every exit BB is the equality
|
||||
of the switch argument with the particular index associated with its case
|
||||
(but the current implementation does not handle switch statements yet).
|
||||
|
||||
These individual conditions are in practice associated to each arc that
|
||||
connects BBs in the CFG (with the simple case that unconditional branches
|
||||
have a "TRUE" condition, because they always happen).
|
||||
|
||||
So, for each BB, its *proper* entry condition is the union of all the
|
||||
conditions associated to arcs that enter the BB. The "union" is like a
|
||||
logical "or", in the sense that either of the condition could be true,
|
||||
they are not necessarily all true. This means that if I can enter a BB
|
||||
in two ways, and in one case I know that "x>0", and in the other that
|
||||
"x==0", actually in the BB I know that "x>=0", which is a weaker
|
||||
condition (the union of the two).
|
||||
|
||||
Also, the *complete* entry condition for a BB is the "intersection" of all
|
||||
the entry conditions of its dominators. This is true because each dominator
|
||||
is the only way to reach the BB, so the entry condition of each dominator
|
||||
must be true if the control flow reached the BB. This translates to the
|
||||
logical "and" of all the "proper" conditions of the BBs met walking up in the
|
||||
dominator tree. So, if one says "x>0", and another "x==1", then I know
|
||||
that "x==1", which is a stronger condition (the intersection of the two).
|
||||
Note that, if the two conditions were "x>0" and "x==0", then the block would
|
||||
be unreachable (the intersection is empty), because some branch is impossible.
|
||||
|
||||
Another observation is that, inside each BB, every variable is subject to the
|
||||
complete entry condition of that very same BB, and not the one in which it is
|
||||
defined (with the "complete entry condition" being the thing I defined before,
|
||||
sorry if these terms "proper" and "complete" are strange, I found nothing
|
||||
better).
|
||||
This happens because the branch conditions are related to the control flow.
|
||||
I can define "i=a", and if I am in a BB where "a>0", then "i>0", but not
|
||||
otherwise.
|
||||
|
||||
So, intuitively, if the available conditions say "i>=0", and i is used as an
|
||||
index in an array access, then the lower bound check can be omitted.
|
||||
If the condition also says "(i>=0)&&(i<array.length)", the abc removal can
|
||||
occur.
|
||||
|
||||
So, a complete solution to the problem of abc removal would be the following:
|
||||
for each array access, build a system of equations containing:
|
||||
[1] all the symbolic variable definitions
|
||||
[2] the complete entry condition of the BB in which the array access occurs
|
||||
[3] the two "goal functions" ("index >=0" and "index < array.length")
|
||||
If the system is valid for *each possible* variable value, then the goal
|
||||
functions are always true, and the abc can be removed.
|
||||
|
||||
All this discussion is useful to give a precise specification to the problem
|
||||
we are trying to solve.
|
||||
The trouble is that, in the general case, the resulting system of equations
|
||||
is like a predicate in first order logic, which is semi-decidable, and its
|
||||
general solution is anyway too complex to be attempted in a JIT compiler
|
||||
(which should not contain a full fledged theorem prover).
|
||||
|
||||
Therefore, we must cut some corner.
|
||||
|
||||
|
||||
By the way, there is also another big problem, which is caused by "recursive"
|
||||
symbolic definitions. These definition can (and generally do) happen every
|
||||
time there is a loop. For instance, in the following piece of code
|
||||
|
||||
for ( int i = 0; i < array.length; i++ ) {
|
||||
Console.WriteLine( "array [i] = " + array [i] );
|
||||
}
|
||||
|
||||
one of the definitions of i is a PHI that can be either 0 or "i + 1".
|
||||
|
||||
Now, we know that mathematically "i = i + 1" does not make sense, and in
|
||||
fact symbolic values are not "equations", they are "symbolic definitions".
|
||||
|
||||
The symbolic value of i is a generic "n", where "n" is the number of
|
||||
iterations of the loop, but this is terrible to handle (and in more complex
|
||||
examples the symbolic definition of i simply cannot be written, because i is
|
||||
calculated in an iterative way).
|
||||
|
||||
However, the definition "i = i + 1" tells us something about i: it tells us
|
||||
that i "grows". So (from the PHI definition) we know that i is either 0, or
|
||||
"grows". This is enough to tell that "i>=0", which is what we want!
|
||||
It is important to note that recursive definitions can only occurr inside
|
||||
PHI definitions, because actually a variable cannot be defined *only* in terms
|
||||
of itself!
|
||||
|
||||
|
||||
At this point, I can explain which corners I want to cut to make the
|
||||
problem solvable. It will not remove all the abc that could theoretically
|
||||
be removed, but at least it will work.
|
||||
|
||||
The easiest way to cut corners is to only handle expressions which are
|
||||
"reasonably simple", and ignore the rest.
|
||||
Keep in mind that ignoring an expression is not harmful in itself.
|
||||
The algorithm will be simply "less powerful", because it will ignore
|
||||
conditions that could have caused to the removal of an abc, but will
|
||||
not remove checks "by mistake" (so the resulting code will be in any case
|
||||
correct).
|
||||
|
||||
In a first attempt, we can consider only conditions that have the simple
|
||||
possible form, which is "(compare simpleExpression simpleExpression)",
|
||||
where "simpleExpression" is either "(ldind.* local[*])" or
|
||||
"(ldlen (ldind.ref local[*]))" (or a constant, of course).
|
||||
|
||||
We can also "summarize" variable definitions, keeping only what is relevant
|
||||
to know: if they are greater or smaller than zero or other variables.
|
||||
|
||||
One particular note on PHI functions: they work (obviously) like the logical
|
||||
"or" of their definitions, and therefore are equivalent to the "logical or"
|
||||
of the summarization of their definitions.
|
||||
|
||||
About recursive definitions (which, believe me, are the worst thing in all
|
||||
this mess), we handle only "monotonic" ones. That is, we try to understand
|
||||
if the recursive definition (which, as we said above, must happen because
|
||||
of a loop) always "grows" or "gets smaller". In all other cases, we decide
|
||||
we cannot handle it.
|
||||
|
||||
So, the algorithm to optimize one method could be something like:
|
||||
|
||||
[1] Build the SSA representation.
|
||||
[2] Analyze each BB. Particularly, do the following:
|
||||
[2a] summarize its exit conditions
|
||||
[2b] summarize all the variable definitions occurring in the BB
|
||||
[2c] check if it contains array accesses (only as optimization, so that
|
||||
step 3 will be performed only when necessary)
|
||||
[3] Perform the removal. For each BB that contains array accesses, do:
|
||||
[3a] build an evaluation area where for each variable we keep track
|
||||
of its relations with zero and with each other variable.
|
||||
[3b] populate the evaluation area with the initial data (the variable
|
||||
definitions and the complete entry condition of the BB)
|
||||
[3c] propagate the relations between variables (so that from "a<b" and
|
||||
"b<c" we know that "a<c") in the evaluation area
|
||||
[3d] for each array access in the BB, use the evaluation area to "match
|
||||
conditions", to see if the goal functions are satisfied
|
||||
|
||||
|
||||
------------------------------------------------------------------------------
|
||||
LOW LEVEL DESIGN DECISIONS
|
||||
|
||||
One of the problems I had to solve was the representation of the relations
|
||||
between symbolic values.
|
||||
The "canonical" solution would be to have a complex, potentially recursive
|
||||
data structure, somewhat similar to MonoInst, and use it all the time (maybe
|
||||
MonoInst could be used "as is").
|
||||
The problem is that, at this point, the software would have to "play" with
|
||||
these structures, for instance to deduce "b < a-1" from "a > b+1", because we
|
||||
are interested in what value b has, and not a. And maybe the software should
|
||||
be able to do this without modifying the original condition, because it was
|
||||
used elsewhere (or could be reused).
|
||||
Then, the software should also be able to apply logical relations to
|
||||
conditions (like "(a>0)||(c<1)"), and manipulate them, too.
|
||||
|
||||
In my opinion, an optimizer written in this way risks to be too complex, and
|
||||
its performance not acceptable in a JIT compiler.
|
||||
Therefore, I decided to represent values in a "summarized" way.
|
||||
For instance, "x = a + 1" becomes "x > a" (and, by the way, "x = x + 1"
|
||||
would look like "x > x", which means that "x grows").
|
||||
So, each variable value is simply a "relation" of the variable with other
|
||||
variables or constants. Anything more complex is ignored.
|
||||
Relations are represented with flags (EQ, LT and GT), so that it is easy
|
||||
to combine them logically (it is enough to perform the bitwise operations
|
||||
on the flags). The condition (EQ|LT|GT) means we know nothing special, any
|
||||
value is possible. The empty condition would mean an unreachable statement
|
||||
(because no value is acceptable).
|
||||
|
||||
There is still the problem of identifying variables, and where to store all
|
||||
these relations. After some thinking, I decided that a "brute force" approach
|
||||
is the easier, and probably also the fastest. In practice, I keep in memory
|
||||
an array of "variable relations", where for each variable I record its
|
||||
relation with the constants 0 and 1 and an array of its relations with all
|
||||
other variables. The evaluation area, therefore, looks like a square matrix
|
||||
as large as the number of variables. I *know* that the matrix is rather sparse
|
||||
(in the sense that only few of its cells are significant), but I decided that
|
||||
this data structure was the simplest and more efficient anyway. After all,
|
||||
each cell takes just one byte, and any other data structure must be some
|
||||
kind of associative array (like a g_hash_table). Such a data structue is
|
||||
difficult to use with a MonoMemPool, and in the end is slower than a plain
|
||||
array, and has the storage overhead of all the pointers... in the end, maybe
|
||||
the full matrix is not a waste at all.
|
||||
<note>
|
||||
After a few benchmarks, I clearly see that the matrix works well if the
|
||||
variables are not too much, otherwise it gets "too sparse", and much time
|
||||
is wasted in the propagation phase. Maybe the data structure could be changed
|
||||
after a more careful evaluation of the problem...
|
||||
</note>
|
||||
|
||||
With these assumptions, all the memory used to represent values, conditions
|
||||
and the evaluation area can be allocated in advance (from a MonoMemPool), and
|
||||
used for all the analysis of one method. Moreover, the performance of logical
|
||||
operations is high (they are simple bitwise operations on bytes).
|
||||
<note>
|
||||
For the propagation phase, there was a function I could not easily code
|
||||
with bitwise operations. As that phase is a real pain for the performance
|
||||
(the complexity is quadratic, and it must be executed iteratively until
|
||||
nothing changes), I decided to code it in perl, and generate a precomputed
|
||||
table then used by the C code (just 64 bytes). I think this is faster than
|
||||
all those branches... Really, I could think again to phases 3c and 3d (see
|
||||
above), maybe there is a better way to do them.
|
||||
</note>
|
||||
|
||||
Of course not all expressions can be handled by the summarization, so the
|
||||
optimizer is not as effective as a full blown theorem prover...
|
||||
|
||||
Another thing related to design... the implementation is as less intrusive
|
||||
as possible.
|
||||
I "patched" the main files to add an option for abc removal, and handle it
|
||||
just like any other option, and then wrote all the relevant code in two
|
||||
separated files (one header for data structures, and a C file for the code).
|
||||
If the patch turns out to be useful also for other pourposes (like a more
|
||||
generic "unreachable code elimination", that detects branches that will never
|
||||
be taken using the analysis of variable values), it will not be difficult
|
||||
to integrate the branch summarizations into the BasicBlock data structures,
|
||||
and so on.
|
||||
|
||||
One last thing... in some places, I knowingly coded in an inefficient way.
|
||||
For instance, "remove_abc_from_block" does something only if the flag
|
||||
"has_array_access_instructions" is true, but I call it anyway, letting it
|
||||
do nothing (calling it conditionally would be better). There are several
|
||||
things like this in the code, but I was in a hurry, and preferred to make
|
||||
the code more readable (it is already difficult to understand as it is).
|
||||
|
||||
------------------------------------------------------------------------------
|
||||
WHEN IS ABC REMOVAL POSSIBLE? WHEN IS IT USEFUL?
|
||||
|
||||
Warning! Random ideas ahead!
|
||||
|
||||
In general, managed code should be safe, so abc removal is possible only if
|
||||
there already is some branch in the method that does the check.
|
||||
So, one check on the index is "mandatory", the abc removal can only avoid a
|
||||
second one (which is useless *because* of the first one).
|
||||
If in the code there is not a conditional branch that "avoids" to access
|
||||
an array out of its bounds, in general the abc removal cannot occur.
|
||||
|
||||
It could happen, however, that one method is called from another one, and
|
||||
the check on the index is performed in the caller.
|
||||
|
||||
For instance:
|
||||
|
||||
void
|
||||
caller( int[] a, int n ) {
|
||||
if ( n <= a.Length ) {
|
||||
for ( int i = 0; i < n; i++ ) {
|
||||
callee( a, i );
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
void callee( int[] a, int i ) {
|
||||
Console.WriteLine( "a [i] = " + a [i] );
|
||||
}
|
||||
|
||||
|
||||
In this case, it should be possible to have two versions of the callee, one
|
||||
without checks, to be called from methods that do the check themselves, and
|
||||
another to be called otherwise.
|
||||
This kind of optimization could be profile driven: if one method is executed
|
||||
several times, wastes CPU time for bound checks, and is mostly called from
|
||||
another one, it could make sense to analyze the situation in detail.
|
||||
|
||||
However, one simple way to perform this optimization would be to inline
|
||||
the callee (so that the abc removal algorithm can see both methods at once).
|
||||
|
||||
The biggest benefits of abc removal can be seen for array accesses that
|
||||
occur "often", like the ones inside inner loops. This is why I tried to
|
||||
handle "recursive" variable definitions, because without them you cannot
|
||||
optimize inside loops, which is also where you need it most.
|
||||
|
||||
Another possible optimization (alwais related to loops) is the following:
|
||||
|
||||
void method( int[] a, int n ) {
|
||||
for ( int i = 0; i < n; i++ ) {
|
||||
Console.WriteLine( "a [i] = " + a [i] );
|
||||
}
|
||||
}
|
||||
|
||||
In this case, the check on n is missing, so the abc could not be removed.
|
||||
However, this would mean that i will be checked twice at each loop iteration,
|
||||
one against n, and the other against a.Length. The code could be transformed
|
||||
like this:
|
||||
|
||||
void method( int[] a, int n ) {
|
||||
if ( n < a.Length ) {
|
||||
for ( int i = 0; i < n; i++ ) {
|
||||
Console.WriteLine( "a [i] = " + a [i] ); // Remove abc
|
||||
}
|
||||
} else {
|
||||
for ( int i = 0; i < n; i++ ) {
|
||||
Console.WriteLine( "a [i] = " + a [i] ); // Leave abc
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
This could result in performance improvements (again, probably this
|
||||
optimization should be profile driven).
|
||||
|
||||
------------------------------------------------------------------------------
|
||||
OPEN ISSUES
|
||||
|
||||
There are several issues that should still be addressed...
|
||||
|
||||
One is related to aliasing. For now, I decided to operate only on local
|
||||
variables, and ignore aliasing problems (and alias analisys has not yet
|
||||
been implemented anyway).
|
||||
Actually, I identified the local arrays with their length, because I
|
||||
totally ignore the contents of arrays (and objects in general).
|
||||
|
||||
Also, in several places in the code I only handle some cases, and ignore
|
||||
other more complex ones, which could anyway work (there are comments where
|
||||
this happens). Anyway, this is not harmful (the code is not as effective
|
||||
as it could be).
|
||||
|
||||
Another possible improvement is the explicit handling of all constants.
|
||||
For now, code like
|
||||
|
||||
void
|
||||
method () {
|
||||
int[] a = new int [16];
|
||||
for ( int i = 0; i < 16; i++ ) {
|
||||
a [i] = i;
|
||||
}
|
||||
}
|
||||
|
||||
is not handled, because the two constants "16" are lost when variable
|
||||
relations are summarized (actually they are not lost, they are stored in the
|
||||
summarized values, but they are not yet used correctly).
|
||||
|
||||
The worst thing, anyway, is that for now I fail completely in distinguishing
|
||||
between signed and unsigned variables/operations/conditions, and between
|
||||
different integer sizes (in bytes). I did this on pourpose, just for lack of
|
||||
time, but this can turn out to be terribly wrong.
|
||||
The problem is caused by the fact that I handle arithmetical operations and
|
||||
conditions as "ideal" operations, but actually they can overflow and/or
|
||||
underflow, giving "surprising" results.
|
||||
|
||||
For instance, look at the following two methods:
|
||||
|
||||
public static void testSignedOverflow()
|
||||
{
|
||||
Console.WriteLine( "Testing signed overflow..." );
|
||||
int[] ia = new int[70000];
|
||||
int i = 1;
|
||||
while ( i <ia.Length )
|
||||
{
|
||||
try
|
||||
{
|
||||
Console.WriteLine( " i = " + i );
|
||||
ia[i] = i;
|
||||
}
|
||||
catch ( IndexOutOfRangeException e )
|
||||
{
|
||||
Console.WriteLine( "Yes, we had an overflow (i = " + i + ")" );
|
||||
}
|
||||
i *= 60000;
|
||||
}
|
||||
Console.WriteLine( "Testing signed overflow done." );
|
||||
}
|
||||
|
||||
public static void testUnsignedOverflow()
|
||||
{
|
||||
Console.WriteLine( "Testing unsigned overflow..." );
|
||||
int[] ia = new int[70000];
|
||||
uint i = 1;
|
||||
while ( i <ia.Length )
|
||||
{
|
||||
try
|
||||
{
|
||||
Console.WriteLine( " i = " + i );
|
||||
ia[i] = (int) i;
|
||||
}
|
||||
catch ( IndexOutOfRangeException e )
|
||||
{
|
||||
Console.WriteLine( "Yes, we had an overflow (i = " + i + ")" );
|
||||
}
|
||||
i *= 60000;
|
||||
}
|
||||
Console.WriteLine( "Testing unsigned overflow done." );
|
||||
}
|
||||
|
||||
In the "testSignedOverflow" method, the exception will be thrown, while
|
||||
in the "testUnsignedOverflow" everything will go on smoothly.
|
||||
The problem is that the test "i <ia.Length" is a signed comparison in the
|
||||
first case, and will not exit from the loop if the index is negative.
|
||||
Therefore, in the first method, the abc should not be removed.
|
||||
|
||||
In practice, the abc removal code should try to predict if any given
|
||||
expression could over/under-flow, and act accordingly.
|
||||
For now, I made so that the only accepted expressions are plain increments
|
||||
and decrements of variables (which cannot over/under-flow without being
|
||||
trapped by the existing conditions). Anyway, this issue should be analyzed
|
||||
better.
|
||||
By the way, in summarizations there is a field that keeps track of the
|
||||
relation with "1", which I planned to use exactly to summarize products and
|
||||
quotients, but it is never used (as I said, only increment and decrement
|
||||
operations are properly summarized).
|
||||
|
||||
@@ -1,44 +0,0 @@
|
||||
Mono Ahead Of Time Compiler
|
||||
===========================
|
||||
|
||||
The new mono JIT has sophisticated optimization features. It uses SSA and has a
|
||||
pluggable architecture for further optimizations. This makes it possible and
|
||||
efficient to use the JIT also for AOT compilation.
|
||||
|
||||
|
||||
* file format: We use the native object format of the platform. That way it is
|
||||
possible to reuse existing tools like objdump and the dynamic loader. All we
|
||||
need is a working assembler, i.e. we write out a text file which is then
|
||||
passed to gas (the gnu assembler) to generate the object file.
|
||||
|
||||
* file names: we simply add ".so" to the generated file. For example:
|
||||
basic.exe -> basic.exe.so
|
||||
corlib.dll -> corlib.dll.so
|
||||
|
||||
* staring the AOT compiler: mini --aot assembly_name
|
||||
|
||||
The following things are saved in the object file:
|
||||
|
||||
* version infos:
|
||||
|
||||
* native code: this is labeled with method_XXXXXXXX: where XXXXXXXX is the
|
||||
hexadecimal token number of the method.
|
||||
|
||||
* additional informations needed by the runtime: For example we need to store
|
||||
the code length and the exception tables. We also need a way to patch
|
||||
constants only available at runtime (for example vtable and class
|
||||
addresses). This is stored i a binary blob labeled method_info_XXXXXXXX:
|
||||
|
||||
PROBLEMS:
|
||||
|
||||
- all precompiled methods must be domain independent, or we add patch infos to
|
||||
patch the target doamin.
|
||||
|
||||
- the main problem is how to patch runtime related addresses, for example:
|
||||
|
||||
- current application domain
|
||||
- string objects loaded with LDSTR
|
||||
- address of MonoClass data
|
||||
- static field offsets
|
||||
- method addreses
|
||||
- virtual function and interface slots
|
||||
@@ -1,57 +0,0 @@
|
||||
|
||||
HOWTO bundle assemblies inside the mono runtime.
|
||||
Paolo Molaro ([email protected])
|
||||
|
||||
* Intent
|
||||
|
||||
Bundling assemblies inside the mono runtime may be useful for a number
|
||||
of reasons:
|
||||
|
||||
* creating a standalone complete runtime that can be more easily
|
||||
distributed
|
||||
|
||||
* having an application run against a known set of assemblies
|
||||
that has been tested
|
||||
|
||||
Of course, there are drawbacks, too: if there has been fixes
|
||||
to the assemblies, replacing them means recompiling the
|
||||
runtime as well and if there are other mono apps, unless they
|
||||
use the same mono binary, there will be less opportunities for
|
||||
the operating system to optimize memory usage. So use this
|
||||
feature only when really needed.
|
||||
|
||||
* Creating the Bundle
|
||||
|
||||
To bundle a set of assemblies, you need to create a file that
|
||||
lists the assembly names and the relative files. Empty lines
|
||||
and lines starting with # are ignored:
|
||||
|
||||
== cut cut ==
|
||||
# Sample bundle template
|
||||
mscorlib: /path/to/mscorlib/assembly.dll
|
||||
myapp: /path/to/myapp.exe
|
||||
== cut cut ==
|
||||
|
||||
Next you need to build the mono runtime using a special configure option:
|
||||
|
||||
./configure --with-bundle=/path/to/bundle/template
|
||||
|
||||
The path to the template should be an absolute path.
|
||||
|
||||
The script metadata/make-bundle.pl will take the specifie
|
||||
assemblies and embed them inside the runtime where the loading
|
||||
routines can find them before searching for them on disk.
|
||||
|
||||
* Open Issues
|
||||
|
||||
There are still two issues to solve:
|
||||
|
||||
* config files: sometimes they are needed but they are
|
||||
not yet bundled inside the library ()
|
||||
|
||||
* building with the included libgc makes it not
|
||||
possible to build a mono binary statically linked to
|
||||
libmono: this needs to be fixed to make bundles
|
||||
really useful.
|
||||
|
||||
|
||||
@@ -1,271 +0,0 @@
|
||||
|
||||
Embedding the Mono runtime, preliminary version
|
||||
Miguel de Icaza ([email protected]),
|
||||
Paolo Molaro ([email protected])
|
||||
|
||||
This document describes how to embed the Mono runtime in your
|
||||
application, and how to invoke CIL methods from C, and how to
|
||||
invoke C code from CIL. Both the JIT and interpreter can be
|
||||
embedded in very similar ways so most of what is described
|
||||
here can be used in either case.
|
||||
|
||||
* Embedding the runtime.
|
||||
|
||||
Embedding the runtime consists of various steps:
|
||||
|
||||
* Compiling and linking the Mono runtime
|
||||
|
||||
* Initializing the Mono runtime
|
||||
|
||||
* Optionally expose C code to the C#/CIL universe.
|
||||
|
||||
These are discussed in detail next.
|
||||
|
||||
** Compiling and Linking
|
||||
|
||||
To embed the runtime, you have to link your code against the
|
||||
Mono runtime libraries. To do this, you want to pass the
|
||||
flags returned by pkg-config to your compiler:
|
||||
|
||||
pkg-config --cflags --libs mono
|
||||
|
||||
is used to get the flags for the JIT runtime and
|
||||
|
||||
pkg-config --cflags --libs mint
|
||||
|
||||
for the interpreted runtime.
|
||||
|
||||
Like this:
|
||||
|
||||
gcc sample.c `pkg-config --cflags --libs mono`
|
||||
|
||||
You can separate the compilation flags from the linking flags, for
|
||||
instance, you can use the following macros in your makefile:
|
||||
|
||||
CFLAGS=`pkg-config --cflags mono`
|
||||
LDFLAGS=`pkg-config --libs mono`
|
||||
|
||||
** Initializing the Mono runtime
|
||||
|
||||
To initialize the JIT runtime, call mono_jit_init, like this:
|
||||
|
||||
#include <mono/mini/jit.h>
|
||||
|
||||
MonoDomain *domain;
|
||||
|
||||
domain = mono_jit_init ("domain-name");
|
||||
|
||||
For the interpreted runtime use mono_interp_init instead:
|
||||
|
||||
#include <mono/interpreter/embed.h>
|
||||
|
||||
MonoDomain *domain;
|
||||
|
||||
domain = mono_interp_init ("domain-name");
|
||||
|
||||
That will return a MonoDomain where your code will be
|
||||
executed. You can create multiple domains. Each domain is
|
||||
isolated from the other domains and code in one domain will
|
||||
not interfere with code in other domains. This is useful if
|
||||
you want to host different applications in your program.
|
||||
|
||||
Then you can load an assembly containing code into the domain:
|
||||
|
||||
MonoAssembly *assembly;
|
||||
|
||||
assembly = mono_domain_assembly_open (domain, "file.dll");
|
||||
if (!assembly)
|
||||
error ();
|
||||
|
||||
In the above example, the contents of `file.dll' will be
|
||||
loaded into the domain. This only loads the code, but it will
|
||||
not execute anything yet. You can replace `file.dll' with
|
||||
another transport file, like `file.exe'
|
||||
|
||||
To start executing code, you must invoke a method in the
|
||||
assembly, or if you have provided a static Main method (an
|
||||
entry point), you can use the convenience function:
|
||||
|
||||
retval = mono_jit_exec (domain, assembly, argc - 1, argv + 1);
|
||||
|
||||
or when using the interpreter use:
|
||||
|
||||
retval = mono_interp_exec (domain, assembly, argc - 1, argv + 1);
|
||||
|
||||
If you want to invoke a different method, look at the
|
||||
`Invoking Methods in the CIL universe' section later on.
|
||||
|
||||
** Shutting down the runtime
|
||||
|
||||
To shutdown the Mono runtime, you have to clean up all the
|
||||
domains that were created, use this function:
|
||||
|
||||
mono_jit_cleanup (domain);
|
||||
|
||||
Or in the case of the interpreted runtime use:
|
||||
|
||||
mono_interp_cleanup (domain);
|
||||
|
||||
** Applications that use threads.
|
||||
|
||||
The Boehm GC system needs to catch your calls to the pthreads
|
||||
layer, so in each file where you use pthread.h you should
|
||||
include the <gc/gc.h> file.
|
||||
|
||||
* Exposing C code to the CIL universe
|
||||
|
||||
The Mono runtime provides two mechanisms to expose C code to
|
||||
the CIL universe: internal calls and native C code. Internal
|
||||
calls are tightly integrated with the runtime, and have the
|
||||
least overhead, as they use the same data types that the
|
||||
runtime uses.
|
||||
|
||||
The other option is to use the Platform Invoke (P/Invoke) to
|
||||
call C code from the CIL universe, using the standard P/Invoke
|
||||
mechanisms.
|
||||
|
||||
To register an internal call, use this call in the C code:
|
||||
|
||||
mono_add_internal_call ("Hello::Sample", sample);
|
||||
|
||||
Now, you need to declare this on the C# side:
|
||||
|
||||
using System;
|
||||
using System.Runtime.CompilerServices;
|
||||
|
||||
class Hello {
|
||||
[MethodImplAttribute(MethodImplOptions.InternalCall)]
|
||||
extern static string Sample ();
|
||||
}
|
||||
|
||||
Since this routine returns a string, here is the C definition:
|
||||
|
||||
static MonoString*
|
||||
Sample ()
|
||||
{
|
||||
return mono_string_new (mono_domain_get (), "Hello!");
|
||||
}
|
||||
|
||||
Notice that we have to return a `MonoString', and we use the
|
||||
`mono_string_new' API call to obtain this from a string.
|
||||
|
||||
* Invoking Methods in the CIL universe
|
||||
|
||||
Calling a method in the CIL universe from C requires a number of steps:
|
||||
|
||||
* Obtaining the MonoMethod handle to the method.
|
||||
|
||||
* The method invocation.
|
||||
|
||||
** Obtaining a MonoMethod
|
||||
|
||||
To get a MonoMethod there are several ways.
|
||||
|
||||
You can get a MonoClass (the structure representing a type)
|
||||
using:
|
||||
|
||||
MonoClass *
|
||||
mono_class_from_name (MonoImage *image, const char* name_space, const char *name);
|
||||
|
||||
and then loop in the returned class method array until you get
|
||||
the one you're looking for. There are examples of such
|
||||
searches as static functions in several C files in
|
||||
metadata/*.c: we need to expose one through the API and remove
|
||||
the duplicates.
|
||||
|
||||
The other, simpler, way is to use the functions in
|
||||
debug-helpers.h: there are examples of their use in monograph,
|
||||
mint and the jit as well. You basically use a string
|
||||
description of the method, like:
|
||||
|
||||
"System.Object:GetHashCode()"
|
||||
|
||||
and create a MonoMethodDesc out of it with:
|
||||
|
||||
MonoMethodDesc* mono_method_desc_new (const char *name, gboolean include_namespace);
|
||||
|
||||
You can then use:
|
||||
|
||||
MonoMethod* mono_method_desc_search_in_class (MonoMethodDesc *desc, MonoClass *klass);
|
||||
MonoMethod* mono_method_desc_search_in_image (MonoMethodDesc *desc, MonoImage *image);
|
||||
|
||||
to search for the method in a class or in an image. You would
|
||||
tipically do this just once at the start of the program and
|
||||
store the result for reuse somewhere.
|
||||
|
||||
** Invoking a Method
|
||||
|
||||
There are two functions to call a managed method:
|
||||
|
||||
MonoObject*
|
||||
mono_runtime_invoke (MonoMethod *method, void *obj, void **params,
|
||||
MonoObject **exc);
|
||||
and
|
||||
MonoObject*
|
||||
mono_runtime_invoke_array (MonoMethod *method, void *obj, MonoArray *params,
|
||||
MonoObject **exc);
|
||||
|
||||
obj is the 'this' pointer, it should be NULL for static
|
||||
methods, a MonoObject* for object instances and a pointer to
|
||||
the value type for value types.
|
||||
|
||||
These functions can be used in both the JIT and the interpreted
|
||||
environments.
|
||||
|
||||
The params array contains the arguments to the method with the
|
||||
same convention: MonoObject* pointers for object instances and
|
||||
pointers to the value type otherwise. The _invoke_array
|
||||
variant takes a C# object[] as the params argument (MonoArray
|
||||
*params): in this case the value types are boxed inside the
|
||||
respective reference representation.
|
||||
|
||||
From unmanaged code you'll usually use the
|
||||
mono_runtime_invoke() variant.
|
||||
|
||||
Note that this function doesn't handle virtual methods for
|
||||
you, it will exec the exact method you pass: we still need to
|
||||
expose a function to lookup the derived class implementation
|
||||
of a virtual method (there are examples of this in the code,
|
||||
though).
|
||||
|
||||
You can pass NULL as the exc argument if you don't want to
|
||||
catch exceptions, otherwise, *exc will be set to the exception
|
||||
thrown, if any. if an exception is thrown, you can't use the
|
||||
MonoObject* result from the function.
|
||||
|
||||
If the method returns a value type, it is boxed in an object
|
||||
reference.
|
||||
|
||||
We have plans for providing an additional method that returns
|
||||
an unmanaged->managed thunk like this:
|
||||
|
||||
void* mono_method_get_unmanaged_thunk (MonoMethod *method);
|
||||
|
||||
You'll be able to store the returned pointer in a function
|
||||
pointer with the proper signature and call that directly from
|
||||
C:
|
||||
|
||||
typedef gint32 (*GetHashCode) (MonoObject *obj);
|
||||
|
||||
GetHashCode func = mono_method_get_unmanaged_thunk (System_Object_GetHashCode_method);
|
||||
|
||||
gint32 hashvalue = func (myobject);
|
||||
|
||||
It may not be possible to manage exceptions in that case,
|
||||
though. I need to think more about it.
|
||||
|
||||
** Threading issues
|
||||
|
||||
If your application creates threads on its own, and you want them to
|
||||
be able to call code into the CIL universe with Mono, you have to
|
||||
register the thread with Mono before issuing the call.
|
||||
|
||||
To do so, call the mono_thread_attach() function before you execute
|
||||
any managed code from the thread
|
||||
|
||||
* Samples
|
||||
|
||||
See the sample programs in mono/sample/embed for examples of
|
||||
embedding the Mono runtime in your application.
|
||||
|
||||
|
||||
@@ -1,110 +0,0 @@
|
||||
Exception Implementation in the Mono Runtime
|
||||
Dietmar Maurer ([email protected])
|
||||
(C) 2001 Ximian, Inc.
|
||||
|
||||
Exception implementation (jit):
|
||||
===============================
|
||||
|
||||
Stack unwinding:
|
||||
================
|
||||
|
||||
We record the code address (start_address, size) of all methods. That way it is
|
||||
possible to map an instruction pointer (IP) to the method information needed
|
||||
for unwinding the stack:
|
||||
|
||||
We also save a Last Managed Frame (LMF) structure at each call from managed to
|
||||
unmanaged code. That way we can recover from exceptions inside unmanaged code.
|
||||
|
||||
void handle_exception ((struct sigcontext *ctx, gpointer obj)
|
||||
{
|
||||
if (ctx->bp < mono_end_of_stack) {
|
||||
/* unhandled exception */
|
||||
abort ();
|
||||
}
|
||||
|
||||
info = mono_jit_info_table_find (mono_jit_info_table, ctx->ip);
|
||||
|
||||
if (info) { // we are inside managed code
|
||||
|
||||
if (ch = find_catch_handler ())
|
||||
execute_catch_handler (ch, ctx, obj);
|
||||
|
||||
execute_all_finally_handler ();
|
||||
|
||||
// restore register, including IP and Frame pointer
|
||||
ctx = restore_caller_saved_registers_from_ctx (ji, ctx);
|
||||
|
||||
// continue unwinding
|
||||
handle_exception (ctx, obj);
|
||||
|
||||
} else {
|
||||
|
||||
lmf = get_last_managed_frame ();
|
||||
|
||||
// restore register, including IP and Frame pointer
|
||||
ctx = restore_caller_saved_registers_from_lmf (ji, lmf);
|
||||
|
||||
// continue unwinding
|
||||
handle_exception (ctx, obj);
|
||||
}
|
||||
}
|
||||
|
||||
|
||||
Code generation:
|
||||
================
|
||||
|
||||
leave: is simply translated into a branch to the target. If the leave
|
||||
instruction is inside a finally block (but not inside another handler)
|
||||
we call the finally handler before we branch to the target.
|
||||
|
||||
finally/endfinally, filter/endfilter: is translated into subroutine ending with
|
||||
a "return" statement. The subroutine does not save EBP, because we need access
|
||||
to the local variables of the enclosing method. Its is possible that
|
||||
instructions inside those handlers modify the stack pointer, thus we save the
|
||||
stack pointer at the start of the handler, and restore it at the end. We have
|
||||
to use a "call" instruction to execute such finally handlers. This makes it
|
||||
also possible to execute them inside the stack unwinding code. The exception
|
||||
object for filters is passed in a local variable (cfg->exvar).
|
||||
|
||||
throw: we first save all regs into a sigcontext struct and then call the stack
|
||||
unwinding code.
|
||||
|
||||
catch handler: catch hanlders are always called from the stack unwinding
|
||||
code. The exception object is passed in a local variable (cfg->exvar).
|
||||
|
||||
gcc support for Exceptions
|
||||
==========================
|
||||
|
||||
gcc supports exceptions in files compiled with the -fexception option. gcc
|
||||
generates DWARF exceptions tables in that case, so it is possible to unwind the
|
||||
stack. The method to read those exception tables is contained in libgcc.a, and
|
||||
in newer versions of glibc (glibc 2.2.5 for example), and it is called
|
||||
__frame_state_for(). Another usable glibc function is backtrace_symbols() which
|
||||
returns the function name corresponding to a code address.
|
||||
|
||||
We dynamically check if those features are available using g_module_symbol(),
|
||||
and we use them only when available. If not available we use the LMF as
|
||||
fallback.
|
||||
|
||||
Using gcc exception information prevents us from saving the LMF at each native
|
||||
call, so this is a way to speed up native calls. This is especially valuable
|
||||
for internal calls, because we can make sure that all internal calls are
|
||||
compiled with -fexceptions (we compile the whole mono runtime with that
|
||||
option).
|
||||
|
||||
All native function are able to call function without exception tables, and so
|
||||
we are unable to restore all caller saved registers if an exception is raised
|
||||
in such function. Well, its possible if the previous function already saves all
|
||||
registers. So we only omit the the LMF if a function has an exception table
|
||||
able to restore all caller saved registers.
|
||||
|
||||
One problem is that gcc almost never saves all caller saved registers, because
|
||||
it is just unnecessary in normal situations. But there is a trick forcing gcc
|
||||
to save all register, we just need to call __builtin_unwind_init() at the
|
||||
beginning of a function. That way gcc generates code to save all caller saved
|
||||
register on the stack.
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
@@ -1,173 +0,0 @@
|
||||
Rules for opening shared files
|
||||
==============================
|
||||
|
||||
File is already open, with share set to none:
|
||||
|
||||
Can not open again
|
||||
|
||||
|
||||
File is already open for reading, with read share:
|
||||
|
||||
Can open for reading only, share must include read (can have write too)
|
||||
|
||||
|
||||
File is already open for reading, with write share:
|
||||
|
||||
Can open for writing only, share must include read (can have write too)
|
||||
|
||||
|
||||
File is already open for reading, with read + write share:
|
||||
|
||||
Can open for read, writing or both, share must include read (can have write too)
|
||||
|
||||
|
||||
File is already open for writing, with read share:
|
||||
|
||||
Can open for reading only, share must include write (can have read too)
|
||||
|
||||
|
||||
File is already open for writing, with write share:
|
||||
|
||||
Can open for writing only, share must include write (can have read too)
|
||||
|
||||
|
||||
File is already open for writing, with read + write share:
|
||||
|
||||
Can open for reading, writing or both, share must include write (can have read too)
|
||||
|
||||
|
||||
File is already open for reading + writing, with read share:
|
||||
|
||||
Can open for reading only, share must be read + write
|
||||
|
||||
|
||||
File is already open for reading + writing, with write share:
|
||||
|
||||
Can open for for writing only, share must be read + write
|
||||
|
||||
|
||||
File is already open for reading + writing, with read + write share:
|
||||
|
||||
Can open for read, writing or both, share must be read + write
|
||||
|
||||
|
||||
|
||||
Executive Summary
|
||||
-----------------
|
||||
|
||||
Second open must have access within first share, must set second share to at least first access
|
||||
|
||||
|
||||
|
||||
|
||||
Documenting code
|
||||
----------------
|
||||
|
||||
|
||||
#include <stdio.h>
|
||||
#include <windows.h>
|
||||
|
||||
int access[] = {
|
||||
GENERIC_READ,
|
||||
GENERIC_WRITE,
|
||||
GENERIC_READ | GENERIC_WRITE
|
||||
};
|
||||
|
||||
char *access_names[] = {
|
||||
"G_READ",
|
||||
"G_WRITE",
|
||||
"G_READ|G_WRITE"
|
||||
};
|
||||
|
||||
int share[] = {
|
||||
FILE_SHARE_READ,
|
||||
FILE_SHARE_WRITE,
|
||||
FILE_SHARE_READ | FILE_SHARE_WRITE
|
||||
};
|
||||
|
||||
char *share_names[] = {
|
||||
"SHARE_READ",
|
||||
"SHARE_WRITE",
|
||||
"SHARE_READ|SHARE_WRITE"
|
||||
};
|
||||
|
||||
void lockfiles(int access1, int share1, int access2, int share2)
|
||||
{
|
||||
HANDLE h1, h2;
|
||||
BOOL ret;
|
||||
|
||||
if (access2 == 0 && share2 == 0) {
|
||||
printf("\n");
|
||||
printf("%22.22s\n%22.22s", access_names[access1], share_names[share1]);
|
||||
}
|
||||
|
||||
h1 = CreateFile("lockedfile",
|
||||
access[access1],
|
||||
share[share1],
|
||||
NULL, OPEN_ALWAYS, FILE_ATTRIBUTE_NORMAL, NULL);
|
||||
if (h1 == INVALID_HANDLE_VALUE) {
|
||||
printf("Open1 failed: %d\n", GetLastError());
|
||||
return;
|
||||
}
|
||||
|
||||
h2 = CreateFile("lockedfile",
|
||||
access[access2],
|
||||
share[share2],
|
||||
NULL, OPEN_ALWAYS, FILE_ATTRIBUTE_NORMAL, NULL);
|
||||
if (h2 == INVALID_HANDLE_VALUE) {
|
||||
printf(" %4.4s", "");
|
||||
} else {
|
||||
printf(" %4.4s", "OK");
|
||||
CloseHandle(h2);
|
||||
}
|
||||
|
||||
CloseHandle(h1);
|
||||
}
|
||||
|
||||
int main(int argc, char **argv)
|
||||
{
|
||||
int i, j, k, l;
|
||||
|
||||
printf("\t\t\t\t\t\t\tSecond Open\n");
|
||||
printf("%22.22s G_RE G_RE G_RE G_WR G_WR G_WR G_RW G_RW G_RW\n", "");
|
||||
printf("%22.22s S_RE S_WR S_RW S_RE S_WR S_RW S_RE S_WR S_RW", "First open --v ");
|
||||
for (i = 0; i < 3; i++) {
|
||||
for (j = 0; j < 3; j++) {
|
||||
for (k = 0; k < 3; k++) {
|
||||
for (l = 0; l < 3; l++) {
|
||||
lockfiles(i, j, k, l);
|
||||
}
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
return(0);
|
||||
}
|
||||
|
||||
|
||||
|
||||
Code output
|
||||
-----------
|
||||
|
||||
Second Open
|
||||
G_RE G_RE G_RE G_WR G_WR G_WR G_RW G_RW G_RW
|
||||
First open --v S_RE S_WR S_RW S_RE S_WR S_RW S_RE S_WR S_RW
|
||||
G_READ
|
||||
SHARE_READ OK OK
|
||||
G_READ
|
||||
SHARE_WRITE OK OK
|
||||
G_READ
|
||||
SHARE_READ|SHARE_WRITE OK OK OK OK OK OK
|
||||
G_WRITE
|
||||
SHARE_READ OK OK
|
||||
G_WRITE
|
||||
SHARE_WRITE OK OK
|
||||
G_WRITE
|
||||
SHARE_READ|SHARE_WRITE OK OK OK OK OK OK
|
||||
G_READ|G_WRITE
|
||||
SHARE_READ OK
|
||||
G_READ|G_WRITE
|
||||
SHARE_WRITE OK
|
||||
G_READ|G_WRITE
|
||||
SHARE_READ|SHARE_WRITE OK OK OK
|
||||
|
||||
@@ -1,31 +0,0 @@
|
||||
* GC issues you need to be careful about when hacking on the mono runtime.
|
||||
|
||||
mono currently uses the Boehm garbage collection library. This is a conservative
|
||||
GC package: this means that the memory is searched for possible memory references
|
||||
to chunks allocated with the GC. Not all the memory is searched, but only the memory
|
||||
allocated with the GC itself (unless you use the 'atomic' variant of the allocation
|
||||
routines), the stack and global variables. Also, if the last reference to an object
|
||||
is stored in an area of themory that is not searched, the object may be freed resulting
|
||||
in memory corruption ind segfaults.
|
||||
|
||||
In particular, memory allocated with system's malloc() is not searched, so you need to be
|
||||
careful NOT to store object references in memory allocated with malloc() (unless you are sure
|
||||
that the object reference will be live at the same time in some other area under the control
|
||||
of the GC (on the stack or in a global variable, for example). Since g_malloc()
|
||||
ultimately calls system malloc() the data structures in GLib are not safe to
|
||||
use to store object references.
|
||||
|
||||
Occasionally, you'll need to store some object reference from C code: in this case,
|
||||
you must make sure that the store location is searched by the GC. You can safely
|
||||
use thread local storage areas, global variables, stack variables. If you need a more
|
||||
complicated data structure, you can use a hash table: MonoGHashTable.
|
||||
This hash table has the same interface as GHashTable from GLib, just stick the "mono_"
|
||||
prefix in function calls and the "Mono" prefix in the hash table type name.
|
||||
This hash table ensures that object references stored in it are tracked by the GC, as long
|
||||
as the pointer to the hash is tracked itself.
|
||||
|
||||
Other data structures that are allocated with the GC and are safe to use to store pointers
|
||||
to GC-allocated memory, are MonoDomain (keeps track of many domain specfic objects)
|
||||
and MonoVTable (referenced by MonoDomain: keeps track of the static data of a type
|
||||
since that can hold references to objects).
|
||||
|
||||
@@ -1,84 +0,0 @@
|
||||
Internal Call Topics
|
||||
|
||||
* Introduction
|
||||
|
||||
The Common Language Infrastructure allows for methods to be
|
||||
implemented in unmanaged code. Unlike the Platform Invocation
|
||||
services which provide marshalling and unmarshalling of data
|
||||
from managed to unmanaged and viceversa the Internal calls do
|
||||
not perform any kind of marshalling.
|
||||
|
||||
* Basic Type mapping
|
||||
|
||||
The following lists how the C# types are exposed to the C API.
|
||||
|
||||
C# type C type
|
||||
-----------------------------
|
||||
char gunichar2
|
||||
bool MonoBoolean
|
||||
sbyte signed char
|
||||
byte guchar
|
||||
short gint16
|
||||
ushort guint16
|
||||
int gint32
|
||||
uint guint32
|
||||
long gint64
|
||||
ulong guint64
|
||||
IntPtr/UIntPtr gpointer
|
||||
object MonoObject*
|
||||
string MonoString*
|
||||
|
||||
* Pointers
|
||||
|
||||
For ref and out paramaters you'll use the corresponding
|
||||
pointer type.
|
||||
|
||||
So if you have a C# type listed as "ref int", you should use
|
||||
"int *" in your implementation.
|
||||
|
||||
* Arrays
|
||||
|
||||
Arrays of any type must be described with a MonoArray* and the
|
||||
elements must be accessed with the mono_array_* macros.
|
||||
|
||||
* Other Structures
|
||||
|
||||
Any other type that has a matching C structure representation,
|
||||
should use a pointer to the struct instead of a generic
|
||||
MonoObject pointer.
|
||||
|
||||
* Instance Methods.
|
||||
|
||||
Instance methods that are internal calls will receive as first argument
|
||||
the instance object, so you must account for it in the C method signature:
|
||||
|
||||
[MethodImplAttribute(MethodImplOptions.InternalCall)]
|
||||
public extern override int GetHashCode ();
|
||||
|
||||
becomes:
|
||||
|
||||
gint32 ves_icall_System_String_GetHashCode (MonoString *this);
|
||||
|
||||
* How to hook internal calls with the runtime
|
||||
|
||||
Once you require an internal call in corlib, you need to
|
||||
create a C implementation for it and register it in a static
|
||||
table in metadata/icall.c. Add an entry in the table like:
|
||||
|
||||
"System.String::GetHashCode", ves_icall_System_String_GetHashCode,
|
||||
|
||||
Note that you need to include the full namespace.name of the
|
||||
class. If there are overloaded methods, you need also to
|
||||
specify the signature of _all_ of them:
|
||||
|
||||
[MethodImplAttribute(MethodImplOptions.InternalCall)]
|
||||
public extern override void DoSomething ();
|
||||
[MethodImplAttribute(MethodImplOptions.InternalCall)]
|
||||
public extern override void DoSomething (bool useful);
|
||||
|
||||
should be mapped with:
|
||||
|
||||
"Namespace.ClassName::DoSomething()", ves_icall_Namespace_ClassName_DoSomething,
|
||||
"Namespace.ClassName::DoSomething(bool)", ves_icall_Namespace_ClassName_DoSomething_bool,
|
||||
|
||||
|
||||
@@ -1,54 +0,0 @@
|
||||
Mono Intermediate Representation Documentation
|
||||
|
||||
Operation: OP_LOCALLOC
|
||||
inst_left:
|
||||
The value
|
||||
ins->inst_left = *sp;
|
||||
ins->cil_code = ip;
|
||||
|
||||
This is the equivalent of alloca() in a C implementation: it
|
||||
allocates a bunch of bytes on the stack dynamically. Local
|
||||
variables are allocated statically in the stack instead.
|
||||
|
||||
It is implemented by moving the stack pointer according to the
|
||||
platform ABI: you may need to keep a specific frame register
|
||||
to address local variables when LOCALLOC is used (the PPC code
|
||||
does this, for example: ifyou're doing the s390 port I suggest
|
||||
starting from that)
|
||||
|
||||
Operation: OP_CALL_HANDLER.
|
||||
|
||||
cil_code:
|
||||
Points to the CIL where the LEAVE or LEAVE_S
|
||||
instruction was found.
|
||||
|
||||
inst_target_bb:
|
||||
The target basic block for this invocation.
|
||||
|
||||
This instruction must call the target basic block.
|
||||
|
||||
Operation: OP_CISINST
|
||||
|
||||
Implements the CEE_MONO_CISINST opcode that is decoded from
|
||||
the CIL stream.
|
||||
|
||||
This opcode takes as input an object reference and a class,
|
||||
and leaves on the assigned register:
|
||||
|
||||
0. if the object is an instance of the class,
|
||||
1. if the object is not instance of the class,
|
||||
2. if the object is a proxy whose type cannot be determined.
|
||||
|
||||
Operation: OP_CCASTCLASS
|
||||
|
||||
Implements the CEE_MONO_CCASTCLASS opcode that is decoded from
|
||||
the CIL stream.
|
||||
|
||||
This opcode takes as input an object reference and a class,
|
||||
and leaves on the assigned register:
|
||||
|
||||
0. if the object is an instance of the class,
|
||||
1. if the object is a proxy whose type cannot be determined
|
||||
|
||||
throws an InvalidCastException exception otherwise.
|
||||
|
||||
@@ -1,64 +0,0 @@
|
||||
Register Allocation
|
||||
===================
|
||||
|
||||
The current JIT implementation uses a tree matcher to generate code. We use a
|
||||
simple algorithm to allocate registers in trees, and limit the number of used
|
||||
temporary register to 4 when evaluating trees. So we can use 2 registers for
|
||||
global register allocation.
|
||||
|
||||
Register Allocation for Trees
|
||||
=============================
|
||||
|
||||
We always evaluate trees from left to right. When there are no more registers
|
||||
available we need to spill values to memory. Here is the simplified algorithm.
|
||||
|
||||
gboolean
|
||||
tree_allocate_regs (tree, exclude_reg)
|
||||
{
|
||||
if (!tree_allocate_regs (tree->left, -1))
|
||||
return FALSE;
|
||||
|
||||
if (!tree_allocate_regs (tree->right, -1)) {
|
||||
|
||||
tree->left->spilled == TRUE;
|
||||
|
||||
free_used_regs (tree->left);
|
||||
|
||||
if (!tree_allocate_regs (tree->right, tree->left->reg))
|
||||
return FALSE;
|
||||
}
|
||||
|
||||
free_used_regs (tree->left);
|
||||
free_used_regs (tree->right);
|
||||
|
||||
/* try to allocate a register (reg != exclude_reg) */
|
||||
if ((tree->reg = next_free_reg (exclude_reg)) != -1)
|
||||
return TRUE;
|
||||
|
||||
return FALSE;
|
||||
}
|
||||
|
||||
The emit routing actually spills the registers:
|
||||
|
||||
tree_emit (tree)
|
||||
{
|
||||
|
||||
tree_emit (tree->left);
|
||||
|
||||
if (tree->left->spilled)
|
||||
save_reg (tree->left->reg);
|
||||
|
||||
tree_emit (tree->right);
|
||||
|
||||
if (tree->left->spilled)
|
||||
restore_reg (tree->left->reg);
|
||||
|
||||
|
||||
emit_code (tree);
|
||||
}
|
||||
|
||||
|
||||
Global Register Allocation
|
||||
==========================
|
||||
|
||||
TODO.
|
||||
@@ -1,125 +0,0 @@
|
||||
Just some thoughts for the JITer:
|
||||
|
||||
General issues:
|
||||
===============
|
||||
|
||||
We are designing a JIT compiler, so we have to consider two things:
|
||||
|
||||
- the quality of the generated code
|
||||
- the time needed to generate that code
|
||||
|
||||
The current approach is to keep the JITer as simple as possible, and thus as
|
||||
fast as possible. The generated code quality will suffer from that.
|
||||
|
||||
Register Allocation:
|
||||
====================
|
||||
|
||||
With lcc you can assign a fixed register to a tree before register
|
||||
allocation. For example this is needed by call, which return the value always
|
||||
in EAX on x86. The current implementation works without such system, due to
|
||||
special forest generation.
|
||||
|
||||
Different Register Sets:
|
||||
========================
|
||||
|
||||
Most processors have more that one register set, at least one for floating
|
||||
point values, and one for integers. Should we support architectures with more
|
||||
that two sets? Does someone knows such an architecture?
|
||||
|
||||
64bit Integer Values:
|
||||
=====================
|
||||
|
||||
I can imagine two different implementation. On possibility would be to treat
|
||||
long (64bit) values simply like any other value type. This implies that we
|
||||
call class methods for ALU operations like add or sub. Sure, this method will
|
||||
be be a bit inefficient.
|
||||
|
||||
The more performant solution is to allocate two 32bit registers for each 64bit
|
||||
value. We add a new non terminal to the monoburg grammar called "lreg". The
|
||||
register allocation routines takes care of this non terminal and allocates two
|
||||
32 bit registers for them.
|
||||
|
||||
Forest generation:
|
||||
==================
|
||||
|
||||
Consider the following code:
|
||||
|
||||
OPCODE: STACK LOCALS
|
||||
LDLOC.0 (5) [5,0]
|
||||
LDC.1 (5,1) [5,0]
|
||||
STLOC.0 (5) [1,0]
|
||||
STLOC.1 () [1,5]
|
||||
|
||||
A simple forest generation generates:
|
||||
|
||||
STLOC.0(LDC.1)
|
||||
STLOC.1(LDLOC.0)
|
||||
|
||||
Which is wrong, since it stores the wrong value (1 instead of 5). Instead we
|
||||
must generate something like:
|
||||
|
||||
STLOC.TMP(LDLOC.0)
|
||||
STLOC.0(LDC.1)
|
||||
STLOC.1(LDLOC.TMP)
|
||||
|
||||
Where STLOC.TMP saves the value into a new temporary variable.
|
||||
|
||||
We also need a similar solution for basic block boundaries when the stack depth
|
||||
is not zero. We can simply save those values to new temporary values. Consider
|
||||
the following basic block with one instruction:
|
||||
|
||||
LDLOC.1
|
||||
This should generate a tree like:
|
||||
|
||||
STLOC.TMP(LDLOC.1) Please notice that an intelligent register allocator can
|
||||
still allocate registers for those new variables.
|
||||
|
||||
DAG handling:
|
||||
=============
|
||||
|
||||
Monoburg can't handle DAGs, instead we need real trees as input for
|
||||
the code generator. So we have two problems:
|
||||
|
||||
1.) DUP instruction: This one is obvious - we need to store the value
|
||||
into a temporary variable to solve the problem.
|
||||
|
||||
2.) function calls: Chapter 12.8, page 343 of "A retargetable C compiler"
|
||||
explains that: "because listing a call node will give it a hidden reference
|
||||
from the code list". I don't understand that (can someone explain that?), but
|
||||
there is another reason to save return values to temporaries: Consider the
|
||||
following code:
|
||||
|
||||
x = f(y) + g(z); // all functions return integers
|
||||
|
||||
We could generate such a tree for this expression: STLOC(ADD(CALL,CALL))
|
||||
|
||||
The problem is that both calls returns the value in the same register,
|
||||
so it is non trivial to generate code for that tree. We must copy one
|
||||
register into another one, which make register allocation more complex.
|
||||
The easier solution is store the result of function calls to
|
||||
temporaries. This leads to the following forest:
|
||||
|
||||
STLOC(CALL)
|
||||
STLOC(CALL)
|
||||
STLOC(ADD (LDLOC, LDLOC))
|
||||
|
||||
This is what lcc is doing, if I understood 12.8, page 342, 343?
|
||||
|
||||
Possible Optimisations:
|
||||
=======================
|
||||
|
||||
Miguel said ORP does some optimisation on IL level, for example moving array
|
||||
bounds checking out of loops:
|
||||
|
||||
for (i = 0; i < N; i++) { check_range (a, i); a [i] = X; }
|
||||
|
||||
id transformed to:
|
||||
|
||||
if (in_range (a, 0, N)) { for (i = 0; i < N; i++) a[i] = X; }
|
||||
else for (i = 0; i < N; i++) { check_range (a, i); a [i] = X; }
|
||||
|
||||
The "else" is only to keep original semantics (exception handling).
|
||||
|
||||
We need loop detection logic in order to implement this (dominator tree).
|
||||
|
||||
AFAIK CACAO also implements this.
|
||||
@@ -1,52 +0,0 @@
|
||||
Author: Dietmar Maurer ([email protected])
|
||||
(C) 2001 Ximian, Inc.
|
||||
|
||||
Howto trigger JIT compilation
|
||||
=============================
|
||||
|
||||
The JIT translates CIL code to native code on a per method basis. For example
|
||||
if you have this simple program:
|
||||
|
||||
public class Test {
|
||||
public static void Main () {
|
||||
System.Console.WriteLine ("Hello");
|
||||
}
|
||||
}
|
||||
|
||||
the JIT first compiles the Main function. Unfortunately Main() contains another
|
||||
reference to System.Console.WriteLine(), so the JIT also needs the address for
|
||||
WriteLine() to generate a call instruction.
|
||||
|
||||
The simplest solution would be to JIT compile System.Console.WriteLine()
|
||||
to generate that address. But that would mean that we JIT compile half of our
|
||||
class library at once, since WriteLine() uses many other classes and function,
|
||||
and we have to call the JIT for each of them. Even worse there is the
|
||||
possibility of cyclic references, and we would end up in an endless loop.
|
||||
|
||||
Thus we need some kind of trampoline function for JIT compilation. Such a
|
||||
trampoline first calls the JIT compiler to create native code, and then jumps
|
||||
directly into that code. Whenever the JIT needs the address of a function (to
|
||||
emit a call instruction) it uses the address of those trampoline functions.
|
||||
|
||||
One drawback of this approach is that it requires an additional indirection. We
|
||||
always call the trampoline. Inside the trampoline we need to check if the
|
||||
method is already compiled or not, and when not compiled we start JIT
|
||||
compilation. After that we call the code. This process is quite time consuming
|
||||
and shows very bad performance.
|
||||
|
||||
The solution is to add some logic to the trampoline function to detect from
|
||||
where it is called. It is then possible for the JIT to patch the call
|
||||
instruction in the caller, so that it directly calls the JIT compiled code
|
||||
next time.
|
||||
|
||||
Implementation for x86
|
||||
======================
|
||||
|
||||
emit-x86.c (arch_create_jit_trampoline): return the JIT trampoline function
|
||||
|
||||
emit-x86.c (x86_magic_trampoline): contains the code to detect the caller and
|
||||
patch the call instruction.
|
||||
|
||||
emit-x86.c (arch_compile_method): JIT compile a method
|
||||
|
||||
|
||||
@@ -1,208 +0,0 @@
|
||||
|
||||
* Proposal for the local register allocator
|
||||
|
||||
The local register allocator deals with allocating registers
|
||||
for temporaries inside a single basic block, while the global
|
||||
register allocator is concerned with method-wide allocation of
|
||||
variables.
|
||||
The global register allocator uses callee-saved register for it's
|
||||
purpouse so that there is no need to save and restore these registers
|
||||
at call sites.
|
||||
|
||||
There are a number of issues the local allocator needs to deal with:
|
||||
*) some instructions expect operands in specific registers (for example
|
||||
the shl instruction on x86, or the call instruction with thiscall
|
||||
convention, or the equivalent call instructions on other architectures,
|
||||
such as the need to put output registers in %oX on sparc)
|
||||
*) some instructions deliver results only in specific registers (for example
|
||||
the div instruction on x86, or the call instructionson on almost all
|
||||
the architectures).
|
||||
*) it needs to know what registers may be clobbered by an instruction
|
||||
(such as in a method call)
|
||||
*) it should avoid excessive reloads or stores to improve performance
|
||||
|
||||
While which specific instructions have limitations is architecture-dependent,
|
||||
the problem shold be solved in an arch-independent way to reduce code duplication.
|
||||
The register allocator will be 'driven' by the arch-dependent code, but it's
|
||||
implementation should be arch-independent.
|
||||
|
||||
To improve the current local register allocator, we need to
|
||||
keep more state in it than the current setup that only keeps busy/free info.
|
||||
|
||||
Possible state information is:
|
||||
|
||||
free: the resgister is free to use and it doesn't contain useful info
|
||||
freeable: the register contains data loaded from a local (there is
|
||||
also info about _which_ local it contains) as a result from previous
|
||||
instructions (like, there was a store from the register to the local)
|
||||
moveable: it contains live data that is needed in a following instruction, but
|
||||
the contents may be moved to a different register
|
||||
busy: the register contains live data and it is placed there because
|
||||
the following instructions need it exactly in that register
|
||||
allocated: the register is used by the global allocator
|
||||
|
||||
The local register allocator will have the following interfaces:
|
||||
|
||||
int get_register ();
|
||||
Searches for a register in the free state. If it doesn't find it,
|
||||
searches for a freeable register. Sets the status to moveable.
|
||||
Looking for a 'free' register before a freeable one should allow for
|
||||
removing a few redundant loads (though I'm still unsure if such
|
||||
things should be delegated entirely to the peephole pass).
|
||||
|
||||
int get_register_force (int reg);
|
||||
Returns 'reg' if it is free or freeable. If it is moveable, it moves it
|
||||
to another free or freeable register.
|
||||
Sets the status of 'reg' to busy.
|
||||
|
||||
void set_register_freeable (int reg);
|
||||
Sets the status of 'reg' to freeable.
|
||||
|
||||
void set_register_free (int reg);
|
||||
Sets the status of 'reg' to free.
|
||||
|
||||
void will_clobber (int reg);
|
||||
Spills the register to the stack. Sets the status to freeable.
|
||||
After the clobbering has occurred, set the status to free.
|
||||
|
||||
void register_unspill (int reg);
|
||||
Un-spills register reg and sets the status to moveable.
|
||||
|
||||
FIXME: how is the 'local' information represented? Maybe a MonoInst* pointer.
|
||||
|
||||
Note: the register allocator will insert instructions in the basic block
|
||||
during it's operation.
|
||||
|
||||
* Examples
|
||||
|
||||
Given the tree (on x86 the right argument to shl needs to be in ecx):
|
||||
|
||||
store (local1, shl (local1, call (some_arg)))
|
||||
|
||||
At the start of the basic block, the registers are set to the free state.
|
||||
The sequence of instructions may be:
|
||||
instruction register status -> [%eax %ecx %edx]
|
||||
start free free free
|
||||
eax = load local1 mov free free
|
||||
/* call clobbers eax, ecx, edx */
|
||||
spill eax free free free
|
||||
call mov free free
|
||||
/* now eax contains the right operand of the shl */
|
||||
mov %eax -> %ecx free busy free
|
||||
un-spill mov busy free
|
||||
shl %cl, %eax mov free free
|
||||
|
||||
The resulting x86 code is:
|
||||
mov $fffc(%ebp), %eax
|
||||
mov %eax, $fff0(%ebp)
|
||||
push some_arg
|
||||
call func
|
||||
mov %eax, %ecx
|
||||
mov $fff0(%ebp), %eax
|
||||
shl %cl, %eax
|
||||
|
||||
Note that since shl could operate directly on memory, we could have:
|
||||
|
||||
push some_arg
|
||||
call func
|
||||
mov %eax, %ecx
|
||||
shl %cl, $fffc(%ebp)
|
||||
|
||||
The above example with loading the operand in a register is just to complicate
|
||||
the example and show that the algorithm should be able to handle it.
|
||||
|
||||
Let's take another example with the this-call call convention (the first argument
|
||||
is passed in %ecx).
|
||||
In this case, will_clobber() will be called only on %eax and %edx, while %ecx
|
||||
will be allocated with get_register_force ().
|
||||
Note: when a register is allocated with get_register_force(), it should be set
|
||||
to a different state as soon as possible.
|
||||
|
||||
store (local1, shl (local1, this-call (local1)))
|
||||
|
||||
instruction register status -> [%eax %ecx %edx]
|
||||
start free free free
|
||||
eax = load local1 mov free free
|
||||
/* force load in %ecx */
|
||||
ecx = load local1 mov busy free
|
||||
spill eax free busy free
|
||||
call mov free free
|
||||
/* now eax contains the right operand of the shl */
|
||||
mov %eax -> %ecx free busy free
|
||||
un-spill mov busy free
|
||||
shl %cl, %eax mov free free
|
||||
|
||||
What happens when a register that we need to allocate with get_register_force ()
|
||||
contains an operand for the next instruction?
|
||||
|
||||
instruction register status -> [%eax %ecx %edx]
|
||||
eax = load local0 mov free free
|
||||
ecx = load local1 mov mov free
|
||||
get_register_force (ecx) here.
|
||||
We have two options:
|
||||
mov %ecx, %edx
|
||||
or:
|
||||
spill %ecx
|
||||
The first option is way better (and allows the peephole pass to
|
||||
just load the value in %edx directly, instead of loading first to %ecx).
|
||||
This doesn't work, though, if the instruction clobbers the %edx register
|
||||
(like in a this-call). So, we first need to clobber the registers
|
||||
(so the state of %ecx changes to freebale and there is no issue
|
||||
with get_register_force ()).
|
||||
What if an instruction both clobbers a register and requires it as
|
||||
an operand? Lets' take the x86 idiv instruction as an example: it
|
||||
requires the dividend in edx:eax and returns the result in eax,
|
||||
with the modulus in edx.
|
||||
|
||||
store (local1, div (local1, local2))
|
||||
|
||||
instruction register status -> [%eax %ecx %edx]
|
||||
eax = load local0 mov free free
|
||||
will_clobber eax, edx free mov free
|
||||
force mov %ecx, %eax busy free free
|
||||
set %edx busy free busy
|
||||
idiv mov free free
|
||||
|
||||
Note: edx is set to free after idiv, because the modulus is not needed
|
||||
(if it was a rem, eax would have been freed).
|
||||
If we load the divisor before will_clobber(), we'll have to spill
|
||||
eax and reload it later. If we load it just after the idiv, there is no issue.
|
||||
In any case, the algorithm should give the correct results and allow the operation.
|
||||
|
||||
Working recursively on the isntructions there shouldn't be huge issues
|
||||
with this algorithm (though, of course, it's not optimal and it may
|
||||
introduce excessive spills or register moves). The advantage over the current
|
||||
local reg allocator is that:
|
||||
1) the number of spills/moves would be smaller anyway
|
||||
2) a separate peephole pass could be able to eliminate reg moves
|
||||
3) we'll be able to remove the 'forced' spills we currently do with
|
||||
the return value of method calls
|
||||
|
||||
* Issues
|
||||
|
||||
How to best integrate such a reg allocator with the burg stuff.
|
||||
|
||||
Think about a call os sparc with two arguments: they got into %o0 and %o1
|
||||
and each of them sets the register as busy. But what if the values to put there
|
||||
are themselves the result of a call? %o0 is no problem, but for all the
|
||||
next argument n the above algorithm would spill all the 0...n-1 registers...
|
||||
|
||||
* Papers
|
||||
|
||||
More complex solutions to the local register allocator problem:
|
||||
http://dimacs.rutgers.edu/TechnicalReports/abstracts/1997/97-33.html
|
||||
|
||||
Combining register allocation and instruction scheduling:
|
||||
http://citeseer.nj.nec.com/motwani95combining.html
|
||||
|
||||
More on LRA euristics:
|
||||
http://citeseer.nj.nec.com/liberatore97hardness.html
|
||||
|
||||
Linear-time optimal code scheduling for delayedload architectures
|
||||
http://www.cs.wisc.edu/~fischer/cs701.f01/inst.sched.ps.gz
|
||||
|
||||
Precise Register Allocation for Irregular Architectures
|
||||
http://citeseer.nj.nec.com/kong98precise.html
|
||||
|
||||
Allocate registers first to subtrees that need more of them.
|
||||
http://www.upb.de/cs/ag-kastens/compii/folien/comment401-409.2.pdf
|
||||
@@ -1,771 +0,0 @@
|
||||
|
||||
A new JIT compiler for the Mono Project
|
||||
|
||||
Miguel de Icaza (miguel@{ximian.com,gnome.org}),
|
||||
Paolo Molaro (lupus@{ximian.com,debian.org})
|
||||
|
||||
|
||||
* Abstract
|
||||
|
||||
Mini is a new compilation engine for the Mono runtime. The
|
||||
new engine is designed to bring new code generation
|
||||
optimizations, portability and pre-compilation.
|
||||
|
||||
In this document we describe the design decisions and the
|
||||
architecture of the new compilation engine.
|
||||
|
||||
* Introduction
|
||||
|
||||
Mono is a Open Source implementation of the .NET Framework: it
|
||||
is made up of a runtime engine that implements the ECMA Common
|
||||
Language Infrastructure (CLI), a set of compilers that target
|
||||
the CLI and a large collection of class libraries.
|
||||
|
||||
This article discusses the new code generation facilities that
|
||||
have been added to the Mono runtime.
|
||||
|
||||
First we discuss the overall architecture of the Mono runtime,
|
||||
and how code generation fits into it; Then we discuss the
|
||||
development and basic architecture of our first JIT compiler
|
||||
for the ECMA CIL framework. The next section covers the
|
||||
objectives for the work on the new JIT compiler, then we
|
||||
discuss the new features available in the new JIT compiler,
|
||||
and finally a technical description of the new code generation
|
||||
engine.
|
||||
|
||||
* Architecture of the Mono Runtime
|
||||
|
||||
The Mono runtime is an implementation of the ECMA Common
|
||||
Language Infrastructure (CLI), whose aim is to be a common
|
||||
platform for executing code in multiple languages.
|
||||
|
||||
Languages that target the CLI generate images that contain
|
||||
code in high-level intermediate representation called the
|
||||
"Common Intermediate Language". This intermediate language is
|
||||
rich enough to allow for programs and pre-compiled libraries
|
||||
to be reflected. The execution environment allows for an
|
||||
object oriented execution environment with single inheritance
|
||||
and multiple interface implementations.
|
||||
|
||||
This runtime provides a number of services for programs that
|
||||
are targeted to it: Just-in-Time compilation of CIL code into
|
||||
native code, garbage collection, thread management, I/O
|
||||
routines, single, double and decimal floating point,
|
||||
asynchronous method invocation, application domains, and a
|
||||
framework for building arbitrary RPC systems (remoting) and
|
||||
integration with system libraries through the Platform Invoke
|
||||
functionality.
|
||||
|
||||
The focus of this document is on the services provided by the
|
||||
Mono runtime to transform CIL bytecodes into code that is
|
||||
native to the underlying architecture.
|
||||
|
||||
The code generation interface is a set of macros that allow a
|
||||
C programmer to generate code on the fly, this is done
|
||||
through a set of macros found in the mono/jit/arch/ directory.
|
||||
These macros are used by the JIT compiler to generate native
|
||||
code.
|
||||
|
||||
The platform invocation code is interesting, as it generates
|
||||
CIL code on the fly to marshal parameters, and then this
|
||||
code is in turned processed by the JIT engine.
|
||||
|
||||
* Previous Experiences
|
||||
|
||||
Mono has built a JIT engine, which has been used to bootstrap
|
||||
Mono since January, 2002. This JIT engine has reasonable
|
||||
performance, and uses an tree pattern matching instruction
|
||||
selector based on the BURS technology. This JIT compiler was
|
||||
designed by Dietmar Maurer, Paolo Molaro and Miguel de Icaza.
|
||||
|
||||
The existing JIT compiler has three phases:
|
||||
|
||||
* Re-creation of the semantic tree from CIL
|
||||
byte-codes.
|
||||
|
||||
* Instruction selection, with a cost-driven
|
||||
engine.
|
||||
|
||||
* Code generation and register allocation.
|
||||
|
||||
It is also hooked into the rest of the runtime to provide
|
||||
services like marshaling, just-in-time compilation and
|
||||
invocation of "internal calls".
|
||||
|
||||
This engine constructed a collection of trees, which we
|
||||
referred to as the "forest of trees", this forest is created by
|
||||
"hydrating" the CIL instruction stream.
|
||||
|
||||
The first step was to identify the basic blocks on the method,
|
||||
and computing the control flow graph (cfg) for it. Once this
|
||||
information was computed, a stack analysis on each basic block
|
||||
was performed to create a forest of trees for each one of
|
||||
them.
|
||||
|
||||
So for example, the following statement:
|
||||
|
||||
int a, b;
|
||||
...
|
||||
b = a + 1;
|
||||
|
||||
Which would be represented in CIL as:
|
||||
|
||||
ldloc.0
|
||||
ldc.i4.1
|
||||
add
|
||||
stloc.1
|
||||
|
||||
After the stack analysis would create the following tree:
|
||||
|
||||
(STIND_I4 ADDR_L[EBX|2] (
|
||||
ADD (LDIND_I4 ADDR_L[ESI|1])
|
||||
CONST_I4[1]))
|
||||
|
||||
This tree contains information from the stack analysis: for
|
||||
instance, notice that the operations explicitly encode the
|
||||
data types they are operating on, there is no longer an
|
||||
ambiguity on the types, because this information has been
|
||||
inferred.
|
||||
|
||||
At this point the JIT would pass the constructed forest of
|
||||
trees to the architecture-dependent JIT compiler.
|
||||
|
||||
The architecture dependent code then performed register
|
||||
allocation (optionally using linear scan allocation for
|
||||
variables, based on life analysis).
|
||||
|
||||
Once variables had been assigned, a tree pattern matching with
|
||||
dynamic programming is used (the tree pattern matcher is
|
||||
custom build for each architecture, using a code
|
||||
generator: monoburg). The instruction selector used cost
|
||||
functions to select the best instruction patterns.
|
||||
|
||||
The instruction selector is able to produce instructions that
|
||||
take advantage of the x86 instruction indexing instructions
|
||||
for example.
|
||||
|
||||
One problem though is that the code emitter and the register
|
||||
allocator did not have any visibility outside the current
|
||||
tree, which meant that some redundant instructions were
|
||||
generated. A peephole optimizer with this architecture was
|
||||
hard to write, given the tree-based representation that is
|
||||
used.
|
||||
|
||||
This JIT was functional, but it did not provide a good
|
||||
architecture to base future optimizations on. Also the
|
||||
line between architecture neutral and architecture
|
||||
specific code and optimizations was hard to draw.
|
||||
|
||||
The JIT engine supported two code generation modes to support
|
||||
the two optimization modes for applications that host multiple
|
||||
application domains: generate code that will be shared across
|
||||
application domains, or generate code that will not be shared
|
||||
across application domains.
|
||||
|
||||
* Objectives of the new JIT engine.
|
||||
|
||||
We wanted to support a number of features that were missing:
|
||||
|
||||
* Ahead-of-time compilation.
|
||||
|
||||
The idea is to allow developers to pre-compile their code
|
||||
to native code to reduce startup time, and the working
|
||||
set that is used at runtime in the just-in-time compiler.
|
||||
|
||||
Although in Mono this has not been a visible problem, we
|
||||
wanted to pro-actively address this problem.
|
||||
|
||||
When an assembly (a Mono/.NET executable) is installed in
|
||||
the system, it would then be possible to pre-compile the
|
||||
code, and have the JIT compiler tune the generated code
|
||||
to the particular CPU on which the software is
|
||||
installed.
|
||||
|
||||
This is done in the Microsoft.NET world with a tool
|
||||
called ngen.exe
|
||||
|
||||
* Have a good platform for doing code optimizations.
|
||||
|
||||
The design called for a good architecture that would
|
||||
enable various levels of optimizations: some
|
||||
optimizations are better performed on high-level
|
||||
intermediate representations, some on medium-level and
|
||||
some at low-level representations.
|
||||
|
||||
Also it should be possible to conditionally turn these on
|
||||
or off. Some optimizations are too expensive to be used
|
||||
in just-in-time compilation scenarios, but these
|
||||
expensive optimizations can be turned on for
|
||||
ahead-of-time compilations or when using profile-guided
|
||||
optimizations on a subset of the executed methods.
|
||||
|
||||
* Reduce the effort required to port the Mono code
|
||||
generator to new architectures.
|
||||
|
||||
For Mono to gain wide adoption in the Unix world, it is
|
||||
necessary that the JIT engine works in most of today's
|
||||
commercial hardware platforms.
|
||||
|
||||
* Features of the new JIT engine.
|
||||
|
||||
The new JIT engine was architected by Dietmar Maurer and Paolo
|
||||
Molaro, based on the new objectives.
|
||||
|
||||
Mono provides a number of services to applications running
|
||||
with the new JIT compiler:
|
||||
|
||||
* Just-in-Time compilation of CLI code into native code.
|
||||
|
||||
* Ahead-of-Time compilation of CLI code, to reduce
|
||||
startup time of applications.
|
||||
|
||||
A number of software development features are also available:
|
||||
|
||||
* Execution time profiling (--profile)
|
||||
|
||||
Generates a report of the times consumed by routines,
|
||||
as well as the invocation times, as well as the
|
||||
callers.
|
||||
|
||||
* Memory usage profiling (--profile)
|
||||
|
||||
Generates a report of the memory usage by a program
|
||||
that is ran under the Mono JIT.
|
||||
|
||||
* Code coverage (--coverage)
|
||||
|
||||
* Execution tracing.
|
||||
|
||||
People who are interested in developing and improving the Mini
|
||||
JIT compiler will also find a few useful routines:
|
||||
|
||||
* Compilation times
|
||||
|
||||
This is used to time the execution time for the JIT
|
||||
when compiling a routine.
|
||||
|
||||
* Control Flow Graph and Dominator Tree drawing.
|
||||
|
||||
These are visual aids for the JIT developer: they
|
||||
render representations of the Control Flow graph, and
|
||||
for the more advanced optimizations, they draw the
|
||||
dominator tree graph.
|
||||
|
||||
This requires Dot (from the graphwiz package) and Ghostview.
|
||||
|
||||
* Code generator regression tests.
|
||||
|
||||
The engine contains support for running regression
|
||||
tests on the virtual machine, which is very helpful to
|
||||
developers interested in improving the engine.
|
||||
|
||||
* Optimization benchmark framework.
|
||||
|
||||
The JIT engine will generate graphs that compare
|
||||
various benchmarks embedded in an assembly, and run the
|
||||
various tests with different optimization flags.
|
||||
|
||||
This requires Perl, GD::Graph.
|
||||
|
||||
* Flexibility
|
||||
|
||||
This is probably the most important component of the new code
|
||||
generation engine. The internals are relatively easy to
|
||||
replace and update, even large passes can be replaced and
|
||||
implemented differently.
|
||||
|
||||
* New code generator
|
||||
|
||||
Compiling a method begins with the `mini_method_to_ir' routine
|
||||
that converts the CIL representation into a medium
|
||||
intermediate representation.
|
||||
|
||||
The mini_method_to_ir routine performs a number of operations:
|
||||
|
||||
* Flow analysis and control flow graph computation.
|
||||
|
||||
Unlike the previous version, stack analysis and control
|
||||
flow graphs are computed in a single pass in the
|
||||
mini_method_to_ir function, this is done for performance
|
||||
reasons: although the complexity increases, the benefit
|
||||
for a JIT compiler is that there is more time available
|
||||
for performing other optimizations.
|
||||
|
||||
* Basic block computation.
|
||||
|
||||
mini_method_to_ir populates the MonoCompile structure
|
||||
with an array of basic blocks each of which contains
|
||||
forest of trees made up of MonoInst structures.
|
||||
|
||||
* Inlining
|
||||
|
||||
Inlining is no longer restricted to methods containing
|
||||
one single basic block, instead it is possible to inline
|
||||
arbitrary complex methods.
|
||||
|
||||
The heuristics to choose what to inline are likely going
|
||||
to be tuned in the future.
|
||||
|
||||
* Method to opcode conversion.
|
||||
|
||||
Some method call invocations like `call Math.Sin' are
|
||||
transformed into an opcode: this transforms the call
|
||||
into a semantically rich node, which is later inline
|
||||
into an FPU instruction.
|
||||
|
||||
Various Array methods invocations are turned into
|
||||
opcodes as well (The Get, Set and Address methods)
|
||||
|
||||
* Tail recursion elimination
|
||||
|
||||
Basic blocks ****
|
||||
|
||||
The MonoInst structure holds the actual decoded instruction,
|
||||
with the semantic information from the stack analysis.
|
||||
MonoInst is interesting because initially it is part of a tree
|
||||
structure, here is a sample of the same tree with the new JIT
|
||||
engine:
|
||||
|
||||
(stind.i4 regoffset[0xffffffd4(%ebp)]
|
||||
(add (ldind.i4 regoffset[0xffffffd8(%ebp)])
|
||||
iconst[1]))
|
||||
|
||||
This is a medium-level intermediate representation (MIR).
|
||||
|
||||
Some complex opcodes are decomposed at this stage into a
|
||||
collection of simpler opcodes. Not every complex opcode is
|
||||
decomposed at this stage, as we need to preserve the semantic
|
||||
information during various optimization phases.
|
||||
|
||||
For example a NEWARR opcode carries the length and the type of
|
||||
the array that could be used later to avoid type checking or
|
||||
array bounds check.
|
||||
|
||||
There are a number of operations supported on this
|
||||
representation:
|
||||
|
||||
* Branch optimizations.
|
||||
|
||||
* Variable liveness.
|
||||
|
||||
* Loop optimizations: the dominator trees are
|
||||
computed, loops are detected, and their nesting
|
||||
level computed.
|
||||
|
||||
* Conversion of the method into static single assignment
|
||||
form (SSA form).
|
||||
|
||||
* Dead code elimination.
|
||||
|
||||
* Constant propagation.
|
||||
|
||||
* Copy propagation.
|
||||
|
||||
* Constant folding.
|
||||
|
||||
Once the above optimizations are optionally performed, a
|
||||
decomposition phase is used to turn some complex opcodes into
|
||||
internal method calls. In the initial version of the JIT
|
||||
engine, various operations on longs are emulated instead of
|
||||
being inlined. Also the newarr invocation is turned into a
|
||||
call to the runtime.
|
||||
|
||||
At this point, after computing variable liveness, it is
|
||||
possible to use the linear scan algorithm for allocating
|
||||
variables to registers. The linear scan pass uses the
|
||||
information that was previously gathered by the loop nesting
|
||||
and loop structure computation to favor variables in inner
|
||||
loops. This process updates the basic block `nesting' field
|
||||
which is later used during liveness analysis.
|
||||
|
||||
Stack space is then reserved for the local variables and any
|
||||
temporary variables generated during the various
|
||||
optimizations.
|
||||
|
||||
** Instruction selection
|
||||
|
||||
At this point, the BURS instruction selector is invoked to
|
||||
transform the tree-based representation into a list of
|
||||
instructions. This is done using a tree pattern matcher that
|
||||
is generated for the architecture using the `monoburg' tool.
|
||||
|
||||
Monoburg takes as input a file that describes tree patterns,
|
||||
which are matched against the trees that were produced by the
|
||||
engine in the previous stages.
|
||||
|
||||
The pattern matching might have more than one match for a
|
||||
particular tree. In this case, the match selected is the one
|
||||
whose cost is the smallest. A cost can be attached to each
|
||||
rule, and if no cost is provided, the implicit cost is one.
|
||||
Smaller costs are selected over higher costs.
|
||||
|
||||
The cost function can be used to select particular blocks of
|
||||
code for a given architecture, or by using a prohibitive high
|
||||
number to avoid having the rule match.
|
||||
|
||||
The various rules that our JIT engine uses transform a tree of
|
||||
MonoInsts into a list of monoinsts:
|
||||
|
||||
+-----------------------------------------------------------+
|
||||
| Tree List |
|
||||
| of ===> Instruction selection ===> of |
|
||||
| MonoInst MonoInst. |
|
||||
+-----------------------------------------------------------+
|
||||
|
||||
During this process various "types" of MonoInst kinds
|
||||
disappear and turned into lower-level representations. The
|
||||
JIT compiler just happens to reuse the same structure (this is
|
||||
done to reduce memory usage and improve memory locality).
|
||||
|
||||
The instruction selection rules are split in a number of
|
||||
files, each one with a particular purpose:
|
||||
|
||||
inssel.brg
|
||||
Contains the generic instruction selection
|
||||
patterns.
|
||||
|
||||
inssel-x86.brg
|
||||
Contains x86 specific rules.
|
||||
|
||||
inssel-ppc.brg
|
||||
Contains PowerPC specific rules.
|
||||
|
||||
inssel-long32.brg
|
||||
burg file for 64bit instructions on 32bit architectures.
|
||||
|
||||
inssel-long.brg
|
||||
burg file for 64bit architectures.
|
||||
|
||||
inssel-float.brg
|
||||
burg file for floating point instructions
|
||||
|
||||
For a given build, a set of those files would be included.
|
||||
For example, for the build of Mono on the x86, the following
|
||||
set is used:
|
||||
|
||||
inssel.brg inssel-x86.brg inssel-long32.brg inssel-float.brg
|
||||
|
||||
** Native method generation
|
||||
|
||||
The native method generation has a number of steps:
|
||||
|
||||
* Architecture specific register allocation.
|
||||
|
||||
The information about loop nesting that was
|
||||
previously gathered is used here to hint the
|
||||
register allocator.
|
||||
|
||||
* Generating the method prolog/epilog.
|
||||
|
||||
* Optionally generate code to introduce tracing facilities.
|
||||
|
||||
* Hooking into the debugger.
|
||||
|
||||
* Performing any pending fixups.
|
||||
|
||||
* Code generation.
|
||||
|
||||
*** Code Generation
|
||||
|
||||
The actual code generation is contained in the architecture
|
||||
specific portion of the compiler. The input to the code
|
||||
generator is each one of the basic blocks with its list of
|
||||
instructions that were produced in the instruction selection
|
||||
phase.
|
||||
|
||||
During the instruction selection phase, virtual registers are
|
||||
assigned. Just before the peephole optimization is performed,
|
||||
physical registers are assigned.
|
||||
|
||||
A simple peephole and algebraic optimizer is ran at this
|
||||
stage.
|
||||
|
||||
The peephole optimizer removes some redundant operations at
|
||||
this point. This is possible because the code generation at
|
||||
this point has visibility into the basic block that spans the
|
||||
original trees.
|
||||
|
||||
The algebraic optimizer performs some simple algebraic
|
||||
optimizations that replace expensive operations with cheaper
|
||||
operations if possible.
|
||||
|
||||
The rest of the code generation is fairly simple: a switch
|
||||
statement is used to generate code for each of the MonoInsts,
|
||||
in the mono/mini/mini-ARCH.c files, the method is called
|
||||
"mono_arch_output_basic_block".
|
||||
|
||||
We always try to allocate code in sequence, instead of just using
|
||||
malloc. This way we increase spatial locality which gives a massive
|
||||
speedup on most architectures.
|
||||
|
||||
*** Ahead of Time compilation
|
||||
|
||||
Ahead-of-Time compilation is a new feature of our new
|
||||
compilation engine. The compilation engine is shared by the
|
||||
Just-in-Time (JIT) compiler and the Ahead-of-Time compiler
|
||||
(AOT).
|
||||
|
||||
The difference is on the set of optimizations that are turned
|
||||
on for each mode: Just-in-Time compilation should be as fast
|
||||
as possible, while Ahead-of-Time compilation can take as long
|
||||
as required, because this is not done at a time critical
|
||||
time.
|
||||
|
||||
With AOT compilation, we can afford to turn all of the
|
||||
computationally expensive optimizations on.
|
||||
|
||||
After the code generation phase is done, the code and any
|
||||
required fixup information is saved into a file that is
|
||||
readable by "as" (the native assembler available on all
|
||||
systems). This assembly file is then passed to the native
|
||||
assembler, which generates a loadable module.
|
||||
|
||||
At execution time, when an assembly is loaded from the disk,
|
||||
the runtime engine will probe for the existence of a
|
||||
pre-compiled image. If the pre-compiled image exists, then it
|
||||
is loaded, and the method invocations are resolved to the code
|
||||
contained in the loaded module.
|
||||
|
||||
The code generated under the AOT scenario is slightly
|
||||
different than the JIT scenario. It generates code that is
|
||||
application-domain relative and that can be shared among
|
||||
multiple thread.
|
||||
|
||||
This is the same code generation that is used when the runtime
|
||||
is instructed to maximize code sharing on a multi-application
|
||||
domain scenario.
|
||||
|
||||
* SSA-based optimizations
|
||||
|
||||
SSA form simplifies many optimization because each variable
|
||||
has exactly one definition site. This means that each
|
||||
variable is only initialized once.
|
||||
|
||||
For example, code like this:
|
||||
|
||||
a = 1
|
||||
..
|
||||
a = 2
|
||||
call (a)
|
||||
|
||||
Is internally turned into:
|
||||
|
||||
a1 = 1
|
||||
..
|
||||
a2 = 2
|
||||
call (a2)
|
||||
|
||||
In the presence of branches, like:
|
||||
|
||||
if (x)
|
||||
a = 1
|
||||
else
|
||||
a = 2
|
||||
|
||||
call (a)
|
||||
|
||||
The code is turned into:
|
||||
|
||||
if (x)
|
||||
a1 = 1;
|
||||
else
|
||||
a2 = 2;
|
||||
a3 = phi (a1, a2)
|
||||
call (a3)
|
||||
|
||||
All uses of a variable are "dominated" by its definition
|
||||
|
||||
This representation is useful as it simplifies the
|
||||
implementation of a number of optimizations like conditional
|
||||
constant propagation, array bounds check removal and dead code
|
||||
elimination.
|
||||
|
||||
* Register allocation.
|
||||
|
||||
Global register allocation is performed on the medium
|
||||
intermediate representation just before instruction selection
|
||||
is performed on the method. Local register allocation is
|
||||
later performed at the basic-block level on the
|
||||
|
||||
Global register allocation uses the following input:
|
||||
|
||||
1) set of register-sized variables that can be allocated to a
|
||||
register (this is an architecture specific setting, for x86
|
||||
these registers are the callee saved register ESI, EDI and
|
||||
EBX).
|
||||
|
||||
2) liveness information for the variables
|
||||
|
||||
3) (optionally) loop info to favor variables that are used in
|
||||
inner loops.
|
||||
|
||||
During instruction selection phase, symbolic registers are
|
||||
assigned to temporary values in expressions.
|
||||
|
||||
Local register allocation assigns hard registers to the
|
||||
symbolic registers, and it is performed just before the code
|
||||
is actually emitted and is performed at the basic block level.
|
||||
A CPU description file describes the input registers, output
|
||||
registers, fixed registers and clobbered registers by each
|
||||
operation.
|
||||
|
||||
* BURG Code Generator Generator
|
||||
|
||||
monoburg was written by Dietmar Maurer. It is based on the
|
||||
papers from Christopher W. Fraser, Robert R. Henry and Todd
|
||||
A. Proebsting: "BURG - Fast Optimal Instruction Selection and
|
||||
Tree Parsing" and "Engineering a Simple, Efficient Code
|
||||
Generator Generator".
|
||||
|
||||
The original BURG implementation is unable to work on DAGs, instead only
|
||||
trees are allowed. Our monoburg implementations is able to generate tree
|
||||
matcher which works on DAGs, and we use this feature in the new
|
||||
JIT. This simplifies the code because we can directly pass DAGs and
|
||||
don't need to convert them to trees.
|
||||
|
||||
* Adding IL opcodes: an excercise (from a post by Paolo Molaro)
|
||||
|
||||
mini.c is the file that read the IL code stream and decides
|
||||
how any single IL instruction is implemented
|
||||
(mono_method_to_ir () func), so you always have to add an
|
||||
entry to the big switch inside the function: there are plenty
|
||||
of examples in that file.
|
||||
|
||||
An IL opcode can be implemented in a number of ways, depending
|
||||
on what it does and how it needs to do it.
|
||||
|
||||
Some opcodes are implemented using a helper function: one of
|
||||
the simpler examples is the CEE_STELEM_REF implementation.
|
||||
|
||||
In this case the opcode implementation is written in a C
|
||||
function. You will need to register the function with the jit
|
||||
before you can use it (mono_register_jit_call) and you need to
|
||||
emit the call to the helper using the mono_emit_jit_icall()
|
||||
function.
|
||||
|
||||
This is the simpler way to add a new opcode and it doesn't
|
||||
require any arch-specific change (though it's limited to what
|
||||
you can do in C code and the performance may be limited by the
|
||||
function call).
|
||||
|
||||
Other opcodes can be implemented with one or more of the already
|
||||
implemented low-level instructions.
|
||||
|
||||
An example is the OP_STRLEN opcode which implements
|
||||
String.Length using a simple load from memory. In this case
|
||||
you need to add a rule to the appropriate burg file,
|
||||
describing what are the arguments of the opcode and what is,
|
||||
if any, it's 'return' value.
|
||||
|
||||
The OP_STRLEN case is:
|
||||
|
||||
reg: OP_STRLEN (reg) {
|
||||
MONO_EMIT_LOAD_MEMBASE_OP (s, tree, OP_LOADI4_MEMBASE, state->reg1,
|
||||
state->left->reg1, G_STRUCT_OFFSET (MonoString, length));
|
||||
}
|
||||
|
||||
The above means: the OP_STRLEN takes a register as an argument
|
||||
and returns its value in a register. And the implementation
|
||||
of this is included in the braces.
|
||||
|
||||
The opcode returns a value in an integer register
|
||||
(state->reg1) by performing a int32 load of the length field
|
||||
of the MonoString represented by the input register
|
||||
(state->left->reg1): before the burg rules are applied, the
|
||||
internal representation is based on trees, so you get the
|
||||
left/right pointers (state->left and state->right
|
||||
respectively, the result is stored in state->reg1).
|
||||
|
||||
This instruction implementation doesn't require arch-specific
|
||||
changes (it is using the MONO_EMIT_LOAD_MEMBASE_OP which is
|
||||
available on all platforms), and usually the produced code is
|
||||
fast.
|
||||
|
||||
Next we have opcodes that must be implemented with new low-level
|
||||
architecture specific instructions (either because of performance
|
||||
considerations or because the functionality can't get implemented in
|
||||
other ways).
|
||||
|
||||
You also need a burg rule in this case, too. For example,
|
||||
consider the OP_CHECK_THIS opcode (used to raise an exception
|
||||
if the this pointer is null). The burg rule simply reads:
|
||||
|
||||
stmt: OP_CHECK_THIS (reg) {
|
||||
mono_bblock_add_inst (s->cbb, tree);
|
||||
}
|
||||
|
||||
Note that this opcode does not return a value (hence the
|
||||
"stmt") and it takes a register as input.
|
||||
|
||||
mono_bblock_add_inst (s->cbb, tree) just adds the instruction
|
||||
(the tree variable) to the current basic block (s->cbb). In
|
||||
mini this is the place where the internal representation
|
||||
switches from the tree format to the low-level format (the
|
||||
list of simple instructions).
|
||||
|
||||
In this case the actual opcode implementation is delegated to
|
||||
the arch-specific code. A low-level opcode needs an entry in
|
||||
the machine description (the *.md files in mini/). This entry
|
||||
describes what kind of registers are used if any by the
|
||||
instruction, as well as other details such as constraints or
|
||||
other hints to the low-level engine which are architecture
|
||||
specific.
|
||||
|
||||
cpu-pentium.md, for example has the following entry:
|
||||
|
||||
checkthis: src1:b len:3
|
||||
|
||||
This means the instruction uses an integer register as a base
|
||||
pointer (basically a load or store is done on it) and it takes
|
||||
3 bytes of native code to implement it.
|
||||
|
||||
Now you just need to provide the low-level implementation for
|
||||
the opcode in one of the mini-$arch.c files, in the
|
||||
mono_arch_output_basic_block() function. There is a big switch
|
||||
here too. The x86 implementation is:
|
||||
|
||||
case OP_CHECK_THIS:
|
||||
/* ensure ins->sreg1 is not NULL */
|
||||
x86_alu_membase_imm (code, X86_CMP, ins->sreg1, 0, 0);
|
||||
break;
|
||||
|
||||
If the $arch-codegen.h header file doesn't have the code to
|
||||
emit the low-level native code, you'll need to write that as
|
||||
well.
|
||||
|
||||
Complex opcodes with register constraints may require other
|
||||
changes to the local register allocator, but usually they are
|
||||
not needed.
|
||||
|
||||
* Future
|
||||
|
||||
Profile-based optimization is something that we are very
|
||||
interested in supporting. There are two possible usage
|
||||
scenarios:
|
||||
|
||||
* Based on the profile information gathered during
|
||||
the execution of a program, hot methods can be compiled
|
||||
with the highest level of optimizations, while bootstrap
|
||||
code and cold methods can be compiled with the least set
|
||||
of optimizations and placed in a discardable list.
|
||||
|
||||
* Code reordering: this profile-based optimization would
|
||||
only make sense for pre-compiled code. The profile
|
||||
information is used to re-order the assembly code on disk
|
||||
so that the code is placed on the disk in a way that
|
||||
increments locality.
|
||||
|
||||
This is the same principle under which SGI's cord program
|
||||
works.
|
||||
|
||||
The nature of the CIL allows the above optimizations to be
|
||||
easy to implement and deploy. Since we live and define our
|
||||
universe for these things, there are no interactions with
|
||||
system tools required, nor upgrades on the underlying
|
||||
infrastructure required.
|
||||
|
||||
Instruction scheduling is important for certain kinds of
|
||||
processors, and some of the framework exists today in our
|
||||
register allocator and the instruction selector to cope with
|
||||
this, but has not been finished. The instruction selection
|
||||
would happen at the same time as local register allocation. <
|
||||
@@ -1,424 +0,0 @@
|
||||
Mono JIT porting guide.
|
||||
Paolo Molaro ([email protected])
|
||||
|
||||
* Introduction
|
||||
|
||||
This documents describes the process of porting the mono JIT
|
||||
to a new CPU architecture. The new mono JIT has been designed
|
||||
to make porting easier though at the same time enable the port
|
||||
to take full advantage from the new architecture features and
|
||||
instructions. Knowledge of the mini architecture (described in
|
||||
the mini-doc.txt file) is a requirement for understanding this
|
||||
guide, as well as an earlier document about porting the mono
|
||||
interpreter (available on the web site).
|
||||
|
||||
There are six main areas that a port needs to implement to
|
||||
have a fully-functional JIT for a given architecture:
|
||||
|
||||
1) instruction selection
|
||||
2) native code emission
|
||||
3) call conventions and register allocation
|
||||
4) method trampolines
|
||||
5) exception handling
|
||||
6) minor helper methods
|
||||
|
||||
To take advantage of some not-so-common processor features
|
||||
(for example conditional execution of instructions as may be
|
||||
found on ARM or ia64), it may be needed to develop an
|
||||
high-level optimization, but doing so is not a requirement for
|
||||
getting the JIT to work.
|
||||
|
||||
We'll see in more details each of the steps required, note,
|
||||
though, that a new port may just as well start from a
|
||||
cut&paste of an existing port to a similar architecture (for
|
||||
example from x86 to amd64, or from powerpc to sparc).
|
||||
|
||||
The architecture specific code is split from the rest of the
|
||||
JIT, for example the x86 specific code and data is all
|
||||
included in the following files in the distribution:
|
||||
|
||||
mini-x86.h mini-x86.c
|
||||
inssel-x86.brg
|
||||
cpu-pentium.md
|
||||
tramp-x86.c
|
||||
exceptions-x86.c
|
||||
|
||||
I suggest a similar split for other architectures as well.
|
||||
|
||||
Note that this document is still incomplete: some sections are
|
||||
only sketched and some are missing, but the important info to
|
||||
get a port going is already described.
|
||||
|
||||
|
||||
* Architecture-specific instructions and instruction selection.
|
||||
|
||||
The JIT already provides a set of instructions that can be
|
||||
easily mapped to a great variety of different processor
|
||||
instructions. Sometimes it may be necessary or advisable to
|
||||
add a new instruction that represent more closely an
|
||||
instruction in the architecture. Note that a mini instruction
|
||||
can be used to represent also a short sequence of CPU
|
||||
low-level instructions, but note that each instruction
|
||||
represents the minimum amount of code the instruction
|
||||
scheduler will handle (i.e., the scheduler won't schedule the
|
||||
instructions that compose the low-level sequence as individual
|
||||
instructions, but just the whole sequence, as an indivisible
|
||||
block).
|
||||
|
||||
New instructions are created by adding a line in the
|
||||
mini-ops.h file, assigning an opcode and a name. To specify
|
||||
the input and output for the instruction, there are two
|
||||
different places, depending on the context in which the
|
||||
instruction gets used.
|
||||
|
||||
If the instruction is used in the tree representation, the
|
||||
input and output types are defined by the BURG rules in the
|
||||
*.brg files (the usual non-terminals are 'reg' to represent a
|
||||
normal register, 'lreg' to represent a register or two that
|
||||
hold a 64 bit value, freg for a floating point register).
|
||||
|
||||
If an instruction is used as a low-level CPU instruction, the
|
||||
info is specified in a machine description file. The
|
||||
description file is processed by the genmdesc program to
|
||||
provide a data structure that can be easily used from C code
|
||||
to query the needed info about the instruction.
|
||||
|
||||
As an example, let's consider the add instruction for both x86
|
||||
and ppc:
|
||||
|
||||
x86 version:
|
||||
add: dest:i src1:i src2:i len:2 clob:1
|
||||
ppc version:
|
||||
add: dest:i src1:i src2:i len:4
|
||||
|
||||
Note that the instruction takes two input integer registers on
|
||||
both CPU, but on x86 the first source register is clobbered
|
||||
(clob:1) and the length in bytes of the instruction differs.
|
||||
|
||||
Note that integer adds and floating point adds use different
|
||||
opcodes, unlike the IL language (64 bit add is done with two
|
||||
instructions on 32 bit architectures, using a add that sets
|
||||
the carry and an add with carry).
|
||||
|
||||
A specific CPU port may assign any meaning to the clob field
|
||||
for an instruction since the value will be processed in an
|
||||
arch-specific file anyway.
|
||||
|
||||
See the top of the existing cpu-pentium.md file for more info
|
||||
on other fields: the info may or may not be applicable to a
|
||||
different CPU, in this latter case the info can be ignored.
|
||||
|
||||
The code in mini.c together with the BURG rules in inssel.brg,
|
||||
inssel-float.brg and inssel-long32.brg provides general
|
||||
purpose mappings from the tree representation to a set of
|
||||
instructions that should be easily implemented in any
|
||||
architecture. To allow for additional arch-specific
|
||||
functionality, an arch-specific BURG file can be used: in this
|
||||
file arch-specific instructions can be selected that provide
|
||||
better performance than the general instructions or that
|
||||
provide functionality that is needed by the JIT but that
|
||||
cannot be expressed in a general enough way.
|
||||
|
||||
As an example, x86 has the special instruction "push" to make
|
||||
it easier to implement the default call convention (passing
|
||||
arguments on the stack): almost all the other architectures
|
||||
don't have such an instruction (and don't need it anyway), so
|
||||
we added a special rule in the inssel-x86.brg file for it.
|
||||
|
||||
So, one of the first things needed in a port is to write a
|
||||
cpu-$(arch).md machine description file and fill it with the
|
||||
needed info. As a start, only a few instructions can be
|
||||
specified, like the ones required to do simple integer
|
||||
operations. The default rules of the instruction selector will
|
||||
emit the common instructions and so we're ready to go for the
|
||||
next step in porting the JIT.
|
||||
|
||||
|
||||
*) Native code emission
|
||||
|
||||
Since the first step in porting mono to a new CPU is to port
|
||||
the interpreter, there should be already a file that allows
|
||||
the emission of binary native code in a buffer for the
|
||||
architecture. This file should be placed in the
|
||||
|
||||
mono/arch/$(arch)/
|
||||
|
||||
directory.
|
||||
|
||||
The bulk of the code emission happens in the mini-$(arch).c
|
||||
file, in a function called mono_arch_output_basic_block
|
||||
(). This function takes a basic block, walks the list of
|
||||
instructions in the block and emits the binary code for each.
|
||||
Optionally a peephole optimization pass is done on the basic
|
||||
block, but this can be left for later, when the port actually
|
||||
works.
|
||||
|
||||
This function is very simple, there is just a big switch on
|
||||
the instruction opcode and in the corresponding case the
|
||||
functions or macros to emit the binary native code are
|
||||
used. Note that in this function the lengths of the
|
||||
instructions are used to determine if the buffer for the code
|
||||
needs enlarging.
|
||||
|
||||
To complete the code emission for a method, a few other
|
||||
functions need implementing as well:
|
||||
|
||||
mono_arch_emit_prolog ()
|
||||
mono_arch_emit_epilog ()
|
||||
mono_arch_patch_code ()
|
||||
|
||||
mono_arch_emit_prolog () will emit the code to setup the stack
|
||||
frame for a method, optionally call the callbacks used in
|
||||
profiling and tracing, and move the arguments to their home
|
||||
location (in a caller-save register if the variable was
|
||||
allocated to one, or in a stack location if the argument was
|
||||
passed in a volatile register and wasn't allocated a
|
||||
non-volatile one). caller-save registers used by the function
|
||||
are saved in the prolog as well.
|
||||
|
||||
mono_arch_emit_epilog () will emit the code needed to return
|
||||
from the function, optionally calling the profiling or tracing
|
||||
callbacks. At this point the basic blocks or the code that was
|
||||
moved out of the normal flow for the function can be emitted
|
||||
as well (this is usually done to provide better info for the
|
||||
static branch predictor). In the epilog, caller-save
|
||||
registers are restored if they were used.
|
||||
|
||||
Note that, to help exception handling and stack unwinding,
|
||||
when there is a transition from managed to unmanaged code,
|
||||
some special processing needs to be done (basically, saving
|
||||
all the registers and setting up the links in the Last Managed
|
||||
Frame structure).
|
||||
|
||||
When the epilog has been emitted, the upper level code
|
||||
arranges for the buffer of memory that contains the native
|
||||
code to be copied in an area of executable memory and at this
|
||||
point, instructions that use relative addressing need to be
|
||||
patched to have the right offsets: this work is done by
|
||||
mono_arch_patch_code ().
|
||||
|
||||
|
||||
* Call conventions and register allocation
|
||||
|
||||
To account for the differences in the call conventions, a few functions need to
|
||||
be implemented.
|
||||
|
||||
mono_arch_allocate_vars () assigns to both arguments and local
|
||||
variables the offset relative to the frame register where they
|
||||
are stored, dead variables are simply discarded. The total
|
||||
amount of stack needed is calculated.
|
||||
|
||||
mono_arch_call_opcode () is the function that more closely
|
||||
deals with the call convention on a given system. For each
|
||||
argument to a function call, an instruction is created that
|
||||
actually puts the argument where needed, be it the stack or a
|
||||
specific register. This function can also re-arrange th order
|
||||
of evaluation when multiple arguments are involved if needed
|
||||
(like, on x86 arguments are pushed on the stack in reverse
|
||||
order). The function needs to carefully take into accounts
|
||||
platform specific issues, like how structures are returned as
|
||||
well as the differences in size and/or alignment of managed
|
||||
and corresponding unmanaged structures.
|
||||
|
||||
The other chunk of code that needs to deal with the call
|
||||
convention and other specifics of a CPU, is the local register
|
||||
allocator, implemented in a function named
|
||||
mono_arch_local_regalloc (). The local allocator deals with a
|
||||
basic block at a time and basically just allocates registers
|
||||
for temporary values during expression evaluation, spilling
|
||||
and unspilling as necessary.
|
||||
|
||||
The local allocator needs to take into account clobbering
|
||||
information, both during simple instructions and during
|
||||
function calls and it needs to deal with other
|
||||
architecture-specific weirdnesses, like instructions that take
|
||||
inputs only in specific registers or output only is some.
|
||||
|
||||
Some effort will be put later in moving most of the local
|
||||
register allocator to a common file so that the code can be
|
||||
shared more for similar, risc-like CPUs. The register
|
||||
allocator does a first pass on the instructions in a block,
|
||||
collecting liveness information and in a backward pass on the
|
||||
same list performs the actual register allocation, inserting
|
||||
the instructions needed to spill values, if necessary.
|
||||
|
||||
When this part of code is implemented, some testing can be
|
||||
done with the generated code for the new architecture. Most
|
||||
helpful is the use of the --regression command line switch to
|
||||
run the regression tests (basic.cs, for example).
|
||||
|
||||
Note that the JIT will try to initialize the runtime, but it
|
||||
may not be able yet to compile and execute complex code:
|
||||
commenting most of the code in the mini_init() function in
|
||||
mini.c is needed to let the JIT just compile the regression
|
||||
tests. Also, using multiple -v switches on the command line
|
||||
makes the JIT dump an increasing amount of information during
|
||||
compilation.
|
||||
|
||||
|
||||
* Method trampolines
|
||||
|
||||
To get better startup performance, the JIT actually compiles a
|
||||
method only when needed. To achieve this, when a call to a
|
||||
method is compiled, we actually emit a call to a magic
|
||||
trampoline. The magic trampoline is a function written in
|
||||
assembly that invokes the compiler to compile the given method
|
||||
and jumps to the newly compiled code, ensuring the arguments
|
||||
it received are passed correctly to the actual method.
|
||||
|
||||
Before jumping to the new code, though, the magic trampoline
|
||||
takes care of patching the call site so that next time the
|
||||
call will go directly to the method instead of the
|
||||
trampoline. How does this all work?
|
||||
|
||||
mono_arch_create_jit_trampoline () creates a small function
|
||||
that just preserves the arguments passed to it and adds an
|
||||
additional argument (the method to compile) before calling the
|
||||
generic trampoline. This small function is called the specific
|
||||
trampoline, because it is method-specific (the method to
|
||||
compile is hard-code in the instruction stream).
|
||||
|
||||
The generic trampoline saves all the arguments that could get
|
||||
clobbered and calls a C function that will do two things:
|
||||
|
||||
*) actually call the JIT to compile the method
|
||||
*) identify the calling code so that it can be patched to call directly
|
||||
the actual method
|
||||
|
||||
If the 'this' argument to a method is a boxed valuetype that
|
||||
is passed to a method that expects just a pointer to the data,
|
||||
an additional unboxing trampoline will need to be inserted as
|
||||
well.
|
||||
|
||||
|
||||
* Exception handling
|
||||
|
||||
Exception handling is likely the most difficult part of the
|
||||
port, as it needs to deal with unwinding (both managed and
|
||||
unmanaged code) and calling catch and filter blocks. It also
|
||||
needs to deal with signals, because mono takes advantage of
|
||||
the MMU in the CPU and of the operation system to handle
|
||||
dereferences of the NULL pointer. Some of the function needed
|
||||
to implement the mechanisms are:
|
||||
|
||||
mono_arch_get_throw_exception () returns a function that takes
|
||||
an exception object and invokes an arch-specific function that
|
||||
will enter the exception processing. To do so, all the
|
||||
relevant registers need to be saved and passed on.
|
||||
|
||||
mono_arch_handle_exception () this function takes the
|
||||
exception thrown and a context that describes the state of the
|
||||
CPU at the time the exception was thrown. The function needs
|
||||
to implement the exception handling mechanism, so it makes a
|
||||
search for an handler for the exception and if none is found,
|
||||
it follows the unhandled exception path (that can print a
|
||||
trace and exit or just abort the current thread). The
|
||||
difficulty here is to unwind the stack correctly, by restoring
|
||||
the register state at each call site in the call chain,
|
||||
calling finally, filters and handler blocks while doing so.
|
||||
|
||||
As part of exception handling a couple of internal calls need
|
||||
to be implemented as well.
|
||||
|
||||
ves_icall_get_frame_info () returns info about a specific
|
||||
frame.
|
||||
|
||||
mono_jit_walk_stack () walks the stack and calls a callback with info for
|
||||
each frame found.
|
||||
|
||||
ves_icall_get_trace () return an array of StackFrame objects.
|
||||
|
||||
** Code generation for filter/finally handlers
|
||||
|
||||
Filter and finally handlers are called from 2 different locations:
|
||||
|
||||
1.) from within the method containing the exception clauses
|
||||
2.) from the stack unwinding code
|
||||
|
||||
To make this possible we implement them like subroutines,
|
||||
ending with a "return" statement. The subroutine does not save
|
||||
the base pointer, because we need access to the local
|
||||
variables of the enclosing method. Its is possible that
|
||||
instructions inside those handlers modify the stack pointer,
|
||||
thus we save the stack pointer at the start of the handler,
|
||||
and restore it at the end. We have to use a "call" instruction
|
||||
to execute such finally handlers.
|
||||
|
||||
The MIR code for filter and finally handlers looks like:
|
||||
|
||||
OP_START_HANDLER
|
||||
...
|
||||
OP_END_FINALLY | OP_ENDFILTER(reg)
|
||||
|
||||
OP_START_HANDLER: should save the stack pointer somewhere
|
||||
OP_END_FINALLY: restores the stack pointers and returns.
|
||||
OP_ENDFILTER (reg): restores the stack pointers and returns the value in "reg".
|
||||
|
||||
** Calling finally/filter handlers
|
||||
|
||||
There is a special opcode to call those handler, its called
|
||||
OP_CALL_HANDLER. It simple emits a call instruction.
|
||||
|
||||
Its a bit more complex to call handler from outside (in the
|
||||
stack unwinding code), because we have to restore the whole
|
||||
context of the method first. After that we simply emit a call
|
||||
instruction to invoke the handler. Its usually possible to use
|
||||
the same code to call filter and finally handlers (see
|
||||
arch_get_call_filter).
|
||||
|
||||
** Calling catch handlers
|
||||
|
||||
Catch handlers are always called from the stack unwinding
|
||||
code. Unlike finally clauses or filters, catch handler never
|
||||
return. Instead we simply restore the whole context, and
|
||||
restart execution at the catch handler.
|
||||
|
||||
** Passing Exception objects to catch handlers and filters.
|
||||
|
||||
We use a local variable to store exception objects. The stack
|
||||
unwinding code must store the exception object into this
|
||||
variable before calling catch handler or filter.
|
||||
|
||||
* Minor helper methods
|
||||
|
||||
A few minor helper methods are referenced from the arch-independent code.
|
||||
Some of them are:
|
||||
|
||||
*) mono_arch_cpu_optimizations ()
|
||||
This function returns a mask of optimizations that
|
||||
should be enabled for the current CPU and a mask of
|
||||
optimizations that should be excluded, instead.
|
||||
|
||||
*) mono_arch_regname ()
|
||||
Returns the name for a numeric register.
|
||||
|
||||
*) mono_arch_get_allocatable_int_vars ()
|
||||
Returns a list of variables that can be allocated to
|
||||
the integer registers in the current architecture.
|
||||
|
||||
*) mono_arch_get_global_int_regs ()
|
||||
Returns a list of caller-save registers that can be
|
||||
used to allocate variables in the current method.
|
||||
|
||||
*) mono_arch_instrument_mem_needs ()
|
||||
*) mono_arch_instrument_prolog ()
|
||||
*) mono_arch_instrument_epilog ()
|
||||
Functions needed to implement the profiling interface.
|
||||
|
||||
|
||||
* Writing regression tests
|
||||
|
||||
Regression tests for the JIT should be written for any bug
|
||||
found in the JIT in one of the *.cs files in the mini
|
||||
directory. Eventually all the operations of the JIT should be
|
||||
tested (including the ones that get selected only when some
|
||||
specific optimization is enabled).
|
||||
|
||||
|
||||
* Platform specific optimizations
|
||||
|
||||
An example of a platform-specific optimization is the peephole
|
||||
optimization: we look at a small window of code at a time and
|
||||
we replace one or more instructions with others that perform
|
||||
better for the given architecture or CPU.
|
||||
|
||||
@@ -1,98 +0,0 @@
|
||||
=pod
|
||||
|
||||
=head1 Internal design document for the mono_handle_d
|
||||
|
||||
This document is designed to hold the design of the mono_handle_d and
|
||||
not as an api reference.
|
||||
|
||||
=head2 Primary goal and purpose
|
||||
|
||||
The mono_handle_d is a process which takes care of the (de)allocation
|
||||
of scratch shared memory and handles (of files, threads, mutexes,
|
||||
sockets etc. see L<WapiHandleType>) and refcounts of the
|
||||
filehandles. It is designed to be run by a user and to be fast, thus
|
||||
minimal error checking on input is done and will most likely crash if
|
||||
given a faulty package. No effort has been, or should be, made to have
|
||||
the daemon talking to machine of different endianness/size of int.
|
||||
|
||||
=head2 How to start the daemon
|
||||
|
||||
To start the daemon you either run the mono_handle_d executable or try
|
||||
to attach to the shared memory segment via L<_wapi_shm_attach> which
|
||||
will start a daemon if one does not exist.
|
||||
|
||||
=head1 Internal details
|
||||
|
||||
The daemon works by opening a socket and listening to clients. These
|
||||
clients send packages over the socket complying to L<struct
|
||||
WapiHandleRequest>.
|
||||
|
||||
=head2 Possible requests
|
||||
|
||||
=over
|
||||
|
||||
=item WapiHandleRequest_New
|
||||
|
||||
Find a handle in the shared memory segment that is free and allocate
|
||||
it to the specified type. To destroy use
|
||||
L</WapiHandleRequest_Close>. A L<WapiHandleResponse> with
|
||||
.type=WapiHandleResponseType_New will be sent back with .u.new.handle
|
||||
set to the handle that was allocated. .u.new.type is the type that was
|
||||
requested.
|
||||
|
||||
=item WapiHandleRequestType_Open
|
||||
|
||||
Increase the ref count of an already created handle. A
|
||||
L<WapiHandleResponse> with .type=WapiHandleResponseType_Open will be sent
|
||||
back with .u.new.handle set to the handle, .u.new.type is set to the
|
||||
type of handle this is.
|
||||
|
||||
=item WapiHandleRequestType_Close
|
||||
|
||||
Decrease the ref count of an already created handle. A
|
||||
L<WapiHandleResponse> with .type=WapiHandleResponseType_Close will be
|
||||
sent back with .u.close.destroy set to TRUE if ref count for this
|
||||
client reached 0.
|
||||
|
||||
=item WapiHandleRequestType_Scratch
|
||||
|
||||
Allocate a shared memory area of size .u.scratch.length in bytes. A
|
||||
L<WapiHandleResponse> with .type=WapiHandleResponseType_Scratch will be
|
||||
sent back with .u.scratch.idx set to the index into the shared
|
||||
memory's scratch area where to memory begins. (works just like
|
||||
malloc(3))
|
||||
|
||||
=item WapiHandleRequestType_Scratch
|
||||
|
||||
Deallocate a shared memory area, this must have been allocated before
|
||||
deallocating. A L<WapiHandleResponse> with
|
||||
.type=WapiHandleResponseType_ScratchFree will be sent back (works just
|
||||
like free(3))
|
||||
|
||||
=back
|
||||
|
||||
=head1 Why a daemon
|
||||
|
||||
From an email:
|
||||
|
||||
Dennis: I just have one question about the daemon... Why does it
|
||||
exist? Isn't it better performancewise to just protect the shared area
|
||||
with a mutex when allocation a new handle/shared mem segment or
|
||||
changing refcnt? It will however be a less resilient to clients that
|
||||
crash (the deamon cleans up ref'd handles if socket closes)
|
||||
|
||||
Dick: It's precisely because with a mutex the shared memory segment
|
||||
can be left in a locked state. Also, it's not so easy to clean up
|
||||
shared memory without it (you can't just mark it deleted when creating
|
||||
it, because you can't attach any more readers to the same segment
|
||||
after that). I did some minimal performance testing, and I don't
|
||||
think the daemon is particularly slow.
|
||||
|
||||
|
||||
=head1 Authors
|
||||
|
||||
Documentaion: Dennis Haney
|
||||
|
||||
Implementation: Dick Porter
|
||||
|
||||
=cut
|
||||
@@ -1,63 +0,0 @@
|
||||
Author: Dietmar Maurer ([email protected])
|
||||
(C) 2001 Ximian, Inc.
|
||||
|
||||
Object and VTable layout
|
||||
========================
|
||||
|
||||
The first pointer inside an Object points to a MonoVtable structure. Objects
|
||||
also contains a MonoThreadsSync structure which is used by the mono Thread
|
||||
implementation.
|
||||
|
||||
typedef struct {
|
||||
MonoVTable *vtable;
|
||||
MonoThreadsSync synchronisation;
|
||||
|
||||
/* object specific data goes here */
|
||||
} MonoObject;
|
||||
|
||||
The MonoVtable contains the vtable, interface offsets and a pointer to static
|
||||
class data. Each object/vtable belongs to exactly one AppDomain.
|
||||
|
||||
typedef struct {
|
||||
MonoClass *klass;
|
||||
MonoDomain *domain;
|
||||
gpointer *interface_offsets;
|
||||
/* a pointer to static data */
|
||||
gpointer data;
|
||||
/* the variable sized vtable is included at the end */
|
||||
gpointer vtable [0];
|
||||
} MonoVTable;
|
||||
|
||||
The MonoClass contains domain independent Class infos.
|
||||
|
||||
typedef struct {
|
||||
/* various class infos */
|
||||
MonoClass *parent;
|
||||
const char *name;
|
||||
const char *name_space;
|
||||
...
|
||||
} MonoClass;
|
||||
|
||||
|
||||
Calling virtual functions:
|
||||
==========================
|
||||
|
||||
Each MonoMethod (if virtual) has an associated slot, which is an index into the
|
||||
VTable. So we can use the following code to compute the address of a virtual
|
||||
function:
|
||||
|
||||
method_addr = object->vtable->vtable [method->slot];
|
||||
|
||||
|
||||
Calling interface methods:
|
||||
==========================
|
||||
|
||||
Each interface class is associated with an unique ID. The following code
|
||||
computes the address of an interface function:
|
||||
|
||||
method_addr = *(object->vtable->interface_offsets [interface_id] + method->slot*4);
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
@@ -1,113 +0,0 @@
|
||||
|
||||
* How to handle complex IL opcodes in an arch-independent way
|
||||
|
||||
Many IL opcodes are very simple: add, ldind etc.
|
||||
Such opcodes can be implemented with a single cpu instruction
|
||||
in most architectures (on some, a group of IL instructions
|
||||
can be converted to a single cpu op).
|
||||
There are many IL opcodes, though, that are more complex, but
|
||||
can be expressed as a series of trees or a single tree of
|
||||
simple operations. Such simple operations are architecture-independent.
|
||||
It makes sense to decompose such complex IL instructions in their
|
||||
simpler equivalent so that we gain in several ways:
|
||||
*) porting effort is easier, because only the simple instructions
|
||||
need to be implemented in arch-specific code
|
||||
*) we could apply BURG rules to the trees and do pattern matching
|
||||
on them to optimize the expressions according to the host cpu
|
||||
|
||||
The issue is: where do we do such conversion from coarse opcodes to
|
||||
simple expressions?
|
||||
|
||||
* Doing the conversion in method_to_ir ()
|
||||
|
||||
Some of these conversions can certainly be done in method_to_ir (),
|
||||
but it's not always easy to decide which are better done there and
|
||||
which in a different pass.
|
||||
For example, let's take ldlen: in the mono implementation, ldlen
|
||||
can be simply implemented with a load from a fixed position in the
|
||||
array object:
|
||||
|
||||
len = [reg + maxlen_offset]
|
||||
|
||||
However, ldlen carries also semantics information: the result is the
|
||||
length of the array, and since in the CLR arrays are of fixed size,
|
||||
this information can be useful to later do bounds check removal.
|
||||
If we convert this opcode in method_to_ir () we lost some useful
|
||||
information for further optimizations.
|
||||
|
||||
In some other ways, decomposing an opcode in method_to_ir() may
|
||||
allow for better optimizations later on (need to come up with an
|
||||
example here ...).
|
||||
|
||||
* Doing the conversion in inssel.brg
|
||||
|
||||
Some conversion may be done inside the burg rules: this has the
|
||||
disadvantage that the instruction selector is not run again on
|
||||
the resulting expression tree and we could miss some optimization
|
||||
(this is what effectively happens with the coarse opcodes in the old
|
||||
jit). This may also interfere with an efficient local register allocator.
|
||||
It may be possible to add an extension in monoburg that allows a rule
|
||||
such as:
|
||||
|
||||
recheck: LDLEN (reg) {
|
||||
create an expression tree representing LDLEN
|
||||
and return it
|
||||
}
|
||||
|
||||
When the monoburg label process gets back a recheck, it will run
|
||||
the labeling again on the resulting expression tree.
|
||||
If this is possible at all (and in an efficient way) is a
|
||||
question for dietmar:-)
|
||||
It should be noted, though, that this may not always work, since
|
||||
some complex IL opcodes may require a series of expression trees
|
||||
and handling such cases in monoburg could become quite hairy.
|
||||
For example, think of opcode that need to do multiple actions on the
|
||||
same object: this basically means a DUP...
|
||||
On the other end, if a complex opcode needs a DUP, monoburg doesn't
|
||||
actually need to create trees if it emits the instructions in
|
||||
the correct sequence and maintains the right values in the registers
|
||||
(usually the values that need a DUP are not changed...). How
|
||||
this integrates with the current register allocator is not clear, since
|
||||
that assigns registers based on the rule, but the instructions emitted
|
||||
by the rules may be different (this already happens with the current JIT
|
||||
where a MULT is replaced with lea etc...).
|
||||
|
||||
* Doing it in a separate pass.
|
||||
|
||||
Doing the conversion in a separate pass over the instructions
|
||||
is another alternative. This can be done right after method_to_ir ()
|
||||
or after the SSA pass (since the IR after the SSA pass should look
|
||||
almost like the IR we get back from method_to_ir ()).
|
||||
|
||||
This has the following advantages:
|
||||
*) monoburg will handle only the simple opcodes (makes porting easier)
|
||||
*) the instruction selection will be run on all the additional trees
|
||||
*) it's easier to support coarse opcodes that produce multiple expression
|
||||
trees (and apply the monoburg selector on all of them)
|
||||
*) the SSA optimizer will see the original opcodes and will be able to use
|
||||
the semantic info associated with them
|
||||
|
||||
The disadvantage is that this is a separate pass on the code and
|
||||
it takes time (how much has not been measured yet, though).
|
||||
|
||||
With this approach, we may also be able to have C implementations
|
||||
of some of the opcodes: this pass would insert a function call to
|
||||
the C implementation (for example in the cases when first porting
|
||||
to a new arch and implemenating some stuff may be too hard in asm).
|
||||
|
||||
* Extended basic blocks
|
||||
|
||||
IL code needs a lot of checks, bounds checks, overflow checks,
|
||||
type checks and so on. This potentially increases by a lot
|
||||
the number of basic blocks in a control flow graph. However,
|
||||
all such blocks end up with a throw opcode that gives control to the
|
||||
exception handling mechanism.
|
||||
After method_to_ir () a MonoBasicBlock can be considered a sort
|
||||
of extended basic block where the additional exits don't point
|
||||
to basic blocks in the same procedure (at least when the method
|
||||
doesn't have exception tables).
|
||||
We need to make sure the passes following method_to_ir () can cope
|
||||
with such kinds of extended basic blocks (especially the passes
|
||||
that we need to apply to all the methods: as a start, we could
|
||||
skip SSA optimizations for methods with exception clauses...)
|
||||
|
||||
@@ -1,16 +0,0 @@
|
||||
<h1>Mono 1.0 Release Notes</h1>
|
||||
|
||||
<h2>What does Mono Include</h2>
|
||||
|
||||
<h2>Missing functionality</h2>
|
||||
|
||||
<p>COM support.
|
||||
|
||||
<p>EnterpriseServices are non-existant.
|
||||
|
||||
<p>Windows.Forms is only available as a preview, it is not
|
||||
completed nor stable.
|
||||
|
||||
<h3>Assembly: System.Drawing</h3>
|
||||
|
||||
<p>System.Drawing.Printing is not supported.
|
||||
@@ -1,26 +0,0 @@
|
||||
Runtime support for Remoting
|
||||
============================
|
||||
|
||||
The runtime supports a special objects called "TransparentProxy". You can
|
||||
create objects of this type by calling GetTransparentProxy() on a "RealProxy"
|
||||
object.
|
||||
|
||||
LDFLD/STFLD for transparent proxies
|
||||
===================================
|
||||
|
||||
Access to fields must be redirected to the remote object. System.Object has
|
||||
some special methods for that:
|
||||
|
||||
void FieldGetter (string typeName, string fieldName, ref object val);
|
||||
|
||||
void FieldSetter (string typeName, string fieldName, object val);
|
||||
|
||||
This methods are never called on actual object. The are only used to pack
|
||||
LDFLD/STFLD operations into method call messages, which are then passed to the
|
||||
RealProxy::Invoke() method.
|
||||
|
||||
There are two helper methods which can be used by the JIT and the interpreter
|
||||
to convert LDFLD/STFLD operations into messages and then call
|
||||
RealProxy::Invoke(): mono_store_remote_field() and mono_load_remote_field().
|
||||
|
||||
|
||||
@@ -1,33 +0,0 @@
|
||||
Size and alignment requirements of stack values
|
||||
===============================================
|
||||
|
||||
P ... System.IntPtr
|
||||
I1 ... System.Int8
|
||||
I2 ... System.Int16
|
||||
I4 ... System.Int32
|
||||
I8 ... System.Int64
|
||||
F ... System.Single
|
||||
D ... System.Double
|
||||
LD ... native long double
|
||||
|
||||
-----------------------------------------------------------
|
||||
ARCH | P | I1 | I2 | I4 | I8 | F | D | LD |
|
||||
-----------------------------------------------------------
|
||||
X86 | 4/4 | 4/4 | 4/4 | 4/4 | 8/4 | 4/4 | 8/4 |12/4 |
|
||||
-----------------------------------------------------------
|
||||
X86/W32 | 4/4 | 4/4 | 4/4 | 4/4 | 8/4 | 4/4 | 8/4 |12/4 |
|
||||
-----------------------------------------------------------
|
||||
ARM | 4/4 | 4/4 | 4/4 | 4/4 | 8/4 | 4/4 | 8/4 | 8/4 |
|
||||
-----------------------------------------------------------
|
||||
M68K | 4/4 | 4/4 | 4/4 | 4/4 | 8/4 | 4/4 | 8/4 |12/4 |
|
||||
-----------------------------------------------------------
|
||||
ALPHA | 8/8 | 8/8 | 8/8 | 8/8 | 8/8 | 8/8 | 8/8 | 8/8 |
|
||||
-----------------------------------------------------------
|
||||
SPARC | 4/4 | 4/4 | 4/4 | 4/4 | 8/8 | 4/4 | 8/8 |16/8 |
|
||||
-----------------------------------------------------------
|
||||
SPARC64 | 8/8 | 8/8 | 8/8 | 8/8 | 8/8 | 8/8 | 8/8 |16/16|
|
||||
-----------------------------------------------------------
|
||||
MIPS | 4/4 | 4/4 | 4/4 | 4/4 | ?/? | 4/4 | 8/8 | 8/8 |
|
||||
-----------------------------------------------------------
|
||||
| | | | | | | | |
|
||||
-----------------------------------------------------------
|
||||
@@ -1,156 +0,0 @@
|
||||
|
||||
Threading in Mono
|
||||
=================
|
||||
|
||||
0. Terminology
|
||||
--------------
|
||||
|
||||
"Main thread" - The initial OS-native thread that the
|
||||
application started with.
|
||||
|
||||
"Helper thread" - A native thread created internally
|
||||
by the runtime, such as the finalizer thread, or an
|
||||
asynchronous delegate invocation thread. These
|
||||
threads can run managed code.
|
||||
|
||||
"Primary CLR thread" - The native thread that called
|
||||
the Main() method when executing an assembly.
|
||||
|
||||
"Secondary CLR thread" - A native thread created by a
|
||||
program that instantiates a System.Threading.Thread object
|
||||
and calls its Start() method.
|
||||
|
||||
1. Thread exit behaviour in the standalone mono runtime
|
||||
-------------------------------------------------------
|
||||
|
||||
The correct behaviour of the runtime should be:
|
||||
|
||||
a) If Main() returns, the runtime should wait for all
|
||||
foreground secondary CLR threads to finish. The
|
||||
wording in the class documentation states: "Once all
|
||||
foreground threads belonging to a process have
|
||||
terminated, the common language runtime ends the
|
||||
process by invoking Abort on any background threads
|
||||
that are still alive." Testing seems to indicate that
|
||||
the background thread can't cancel the Abort by
|
||||
catching the ThreadAbortException and calling
|
||||
ResetAbort here. Indeed, not even the finally block
|
||||
seems to be executed.
|
||||
|
||||
b) if any of the primary CLR thread, a secondary CLR
|
||||
thread or a helper thread calls
|
||||
System.Environment.Exit(), the application should
|
||||
terminate immediately without waiting for foreground
|
||||
primary or secondary CLR threads to finish.
|
||||
|
||||
c) if the primary CLR thread throws an uncaught
|
||||
exception, the application should terminate
|
||||
immediately without waiting for secondary CLR threads
|
||||
to finish. This might be implemented internally by
|
||||
pretending that all the running secondary CLR threads
|
||||
are background threads.
|
||||
|
||||
d) if a secondary CLR thread throws an uncaught
|
||||
exception that thread should terminate and all other
|
||||
threads should continue to execute.
|
||||
|
||||
e) if a helper thread throws an uncaught exception and
|
||||
that thread happens to be the GC finalizer thread,
|
||||
testing seems to indicate that the exception stack
|
||||
trace is displayed as normal, and the exception is
|
||||
then ignored (as though there is a try {} catch{}
|
||||
around all finalizers that just prints the stack
|
||||
trace.) Calling Abort() on the GC finalizer thread
|
||||
also does not cause it to exit: it behaves as though
|
||||
the ThreadAbortException is caught and ResetAbort is
|
||||
called. Asynchronous delegate helper threads should
|
||||
behave as secondary CLR threads, but uncaught
|
||||
exceptions should be rethrown on the thread that calls
|
||||
EndInvoke().
|
||||
|
||||
|
||||
The difficulties happen with cases b) and c):
|
||||
|
||||
The current implementation of
|
||||
System.Environment.Exit() calls exit(2) directly,
|
||||
which is rather unfriendly: it prevents any runtime
|
||||
cleanup, statistics gathering, etc. and is pretty
|
||||
obnoxious to embedded code.
|
||||
|
||||
The current exception handling code calls ExitThread()
|
||||
(emulated with pthread_exit() in the io-layer) if an
|
||||
exception is not caught.
|
||||
|
||||
When called from the main thread, both POSIX
|
||||
pthread_exit() and w32 ExitThread() block if there are
|
||||
other threads still running (in the w32 case, if there
|
||||
are other foreground threads still running; threads
|
||||
can set as background.) If the main thread is also
|
||||
the primary CLR thread, then the application will
|
||||
block until all other threads (including helper
|
||||
threads) terminate. Some helper threads will not
|
||||
terminate until specifically told to by the runtime:
|
||||
for example, the GC finalizer thread needs to run
|
||||
until all of the primary and secondary CLR threads
|
||||
have finished.
|
||||
|
||||
Also, if the main thread is also the primary CLR
|
||||
thread, the runtime loses the opportunity to do any
|
||||
cleaning up. Adding a special case to call exit(2)
|
||||
instead of ExitThread() in the primary CLR thread
|
||||
suffers from the same problems as
|
||||
System.Environment.Exit() calling exit(2).
|
||||
|
||||
|
||||
The simple solution is to run the primary CLR thread
|
||||
in a new native thread, leaving the main thread free
|
||||
for housekeeping duties. There still needs to be some
|
||||
special handling for the case where the primary CLR
|
||||
thread fails to catch an exception: the secondary CLR
|
||||
threads then need to be terminated.
|
||||
|
||||
When the primary and secondary CLR threads have all
|
||||
terminated, the helper threads can be killed off and
|
||||
the runtime can clean itself up and exit.
|
||||
|
||||
|
||||
|
||||
2. Thread initialisation
|
||||
------------------------
|
||||
|
||||
Threads have to undergo some initialisation before
|
||||
managed code can be executed. A
|
||||
System.Threading.Thread object must be created, and
|
||||
the thread details need to be stored so that the
|
||||
threads can be managed later. The JIT needs to record
|
||||
the last managed frame stack pointer in a TLS slot,
|
||||
and the current Thread object is also recorded.
|
||||
|
||||
New threads created by managed calls to
|
||||
System.Threading.Thread methods will have all needed
|
||||
initialisation performed. Threads created by the
|
||||
runtime with calls to mono_thread_create() will too.
|
||||
Existing threads can be passed to the runtime; these
|
||||
must call mono_thread_attach() before any CLR code can
|
||||
be executed on that thread.
|
||||
|
||||
|
||||
3. Constraints on embedding the Mono runtime
|
||||
--------------------------------------------
|
||||
|
||||
The discussion above concerning application behaviour
|
||||
in the event of threads terminating, whether by
|
||||
returning from the start function, throwing uncaught
|
||||
exceptions or by calling System.Environment.Exit(),
|
||||
only really applies to the standalone Mono runtime.
|
||||
|
||||
An embedding application should specify what behaviour
|
||||
is required when, for example,
|
||||
System.Environment.Exit() is called. The application
|
||||
is also responsible for its own thread management, and
|
||||
it should be prepared for any of the primary CLR
|
||||
thread or secondary CLR threads to terminate at any
|
||||
time. The application should also take into account
|
||||
that the runtime will create helper threads as needed,
|
||||
as this may cause pthread_exit() or ExitThread() to
|
||||
block indefinitely, as noted above.
|
||||
@@ -1,135 +0,0 @@
|
||||
Author: Dietmar Maurer ([email protected])
|
||||
(C) 2001 Ximian, Inc.
|
||||
|
||||
More about PInvoke and Internal calls
|
||||
=====================================
|
||||
|
||||
1.) What is PInvoke
|
||||
|
||||
PInvoke stands for Platform Invoke. It is possible to call functions contained
|
||||
in native shared libraries, for example you can declare:
|
||||
|
||||
[DllImport("cygwin1.dll", EntryPoint="puts"]
|
||||
public static extern int puts (string name);
|
||||
|
||||
If you then call "puts(...)" it invokes the native "puts" functions in
|
||||
"cygwin1.dll". It is also possible to specify several marshalling attributes
|
||||
for the arguments, for example you can specify that they puts() function expect
|
||||
ts the string in Ansi encoding by setting the CharSet attribute field:
|
||||
|
||||
[DllImport("cygwin1.dll", EntryPoint="puts", CharSet=CharSet.Ansi)]
|
||||
public static extern int puts (string name);
|
||||
|
||||
2.) What are internal calls
|
||||
|
||||
Some class library functions are implemented in C, because it is either not
|
||||
possible to implement them in C# or because of performance gains. Internal
|
||||
functions are contained in the mono executable itself. Here is an example form
|
||||
our array implementation:
|
||||
|
||||
[MethodImplAttribute(MethodImplOptions.InternalCall)]
|
||||
public extern int GetRank ();
|
||||
|
||||
If you call this GetRank() function it invokes
|
||||
ves_icall_System_Array_GetRank() inside the mono runtime.
|
||||
|
||||
If you write your own runtime environment you can add internal calls with
|
||||
mono_add_internal_call().
|
||||
|
||||
|
||||
2.) Runtime considerations
|
||||
|
||||
Invoking native (unmanaged) code has several implications:
|
||||
|
||||
- We need to handle exceptions inside unmanaged code. The JIT simply saves some
|
||||
informations at each transition from managed to unmanaged code (in a linked
|
||||
list), called Last Managed Frame (LMF). If an exception occurs the runtime
|
||||
first looks if the exception was inside managed code. If not there must be a
|
||||
LMF entry which contains all necessary information to unwind the stack.
|
||||
|
||||
Creation of those LMF structure clearly involves some overhead, so calling
|
||||
into unmanaged code is not as cheap as it looks like at first glance. Maybe
|
||||
we can introduce a special attribute to avoid the creation of LMF on internal
|
||||
call methods that cant raise exceptions.
|
||||
|
||||
- PInvoke has the possibility to convert argument types. For example Strings
|
||||
are marshalled as Char*. So each String argument is translated into a
|
||||
char*. The encoding is specified in the CharSet of the DllImport attribute.
|
||||
|
||||
|
||||
3.) When/how does the runtime call unmanaged PInvoke code
|
||||
|
||||
- LDFTN, CALLI, Delegate::Invoke, Delegate::BeginInvoke: We must generate
|
||||
wrapper code when we load the function with LDFTN, so that all arguments are
|
||||
marshalled in the right format. We also need to save/restore the LMF.
|
||||
|
||||
- MethodBase::Invoke (runtime invoke): We need to marshal all arguments in
|
||||
they right format and save/restore the LMF
|
||||
|
||||
- CALL: We need to marshal all arguments in they right format and save/restore
|
||||
the LMF
|
||||
|
||||
The easiest way to implement this is to always create a wrapper function for
|
||||
PInvoke calls, which takes care of argument marshalling and LMF save/restore.
|
||||
|
||||
4.) When/how does the runtime call unmanaged internal calls
|
||||
|
||||
We don't need to convert any arguments, so we need only take care of the LMF
|
||||
structure.
|
||||
|
||||
- LDFTN, CALLI, Delegate::Invoke, Delegate::BeginInvoke: We must generate
|
||||
wrapper code when we load the function with LDFTN which saves/restores the
|
||||
LMF.
|
||||
|
||||
- MethodBase::Invoke (runtime invoke): We need to save/restore the LMF.
|
||||
|
||||
- CALL: We need to save/restore the LMF.
|
||||
|
||||
- CALLVIRT (through the vtable): We must generate wrapper code to save/restore
|
||||
the LMF.
|
||||
|
||||
Please notice that we can call internal function with CALLVIRT, i.e. we can
|
||||
call those function through a VTable. But we cant know in advance if a vtable
|
||||
slot contains an internal call or managed code. So again it is best to generate
|
||||
a wrapper functions for internal calls in order to save/restore the LMF.
|
||||
|
||||
Unfortunately we need to push all arguments 2 times, because we have to save
|
||||
the LMF, and the LMF is currently allocated on the stack. So the stack looks
|
||||
like:
|
||||
|
||||
--------------------
|
||||
| method arguments |
|
||||
--------------------
|
||||
| LMF |
|
||||
--------------------
|
||||
| copied arguments |
|
||||
--------------------
|
||||
|
||||
AFAIK this is the way ORP works. Another way is to allocate the LMF not on the
|
||||
stack, but then we have additional overhead to allocate/free LMF structures
|
||||
(and another call to arch_get_lmf_addr).
|
||||
|
||||
Maybe it is possible to avoid this addiotional copy for internal calls by
|
||||
including the LMF in the C function signature. Lets say we hav a puts()
|
||||
function which is a internal call:
|
||||
|
||||
ves_icall_puts (MonoString *string);
|
||||
|
||||
If we simply modify that to include the LMF we can avoid to copy all arguments:
|
||||
|
||||
ves_icall_puts (MonoLMF lmf, MonoString *string);
|
||||
|
||||
But this depends somehow on the calling conventions, and I don't know if that
|
||||
works on all plattforms?
|
||||
|
||||
|
||||
5.) What is stored in the LMF
|
||||
|
||||
- all caller saved registers (since we can trust unmanaged code)
|
||||
- the instruction pointer of the last managed instruction
|
||||
- a MonoMethod pointer for the unmanaged function
|
||||
- the address of the thread local lfm_addr pointer (to avoid another call to
|
||||
arch_get_lmf_addr when restoring LMF)
|
||||
|
||||
The LMF is allocated on the stack, so we also know the stack position for
|
||||
stack unwinding.
|
||||
@@ -1,3 +0,0 @@
|
||||
Makefile.in
|
||||
Makefile
|
||||
.deps
|
||||
@@ -1,24 +0,0 @@
|
||||
2004-06-24 David Waite <[email protected]>
|
||||
|
||||
* mono-jni.c: change to C90-style comments from C99/C++-style
|
||||
|
||||
Fri Jun 18 13:46:05 CEST 2004 Paolo Molaro <[email protected]>
|
||||
|
||||
* mono-jni.c: updates for the API cleanups.
|
||||
|
||||
2004-05-05 Zoltan Varga <[email protected]>
|
||||
|
||||
* mono-jni.c: Implement some missing JNI functions.
|
||||
|
||||
2004-04-13 Zoltan Varga <[email protected]>
|
||||
|
||||
* mono-jni.c: Use <stdarg.h> instead of <varargs.h>.
|
||||
|
||||
2004-04-12 Zoltan Varga <[email protected]>
|
||||
|
||||
* mono-jni.c: Fix compilation under gcc 2.95 + fix warnings.
|
||||
|
||||
2004-04-07 Zoltan Varga <[email protected]>
|
||||
|
||||
* mono-jni.c Makefile.am: New files.
|
||||
|
||||
@@ -1,15 +0,0 @@
|
||||
|
||||
INCLUDES = \
|
||||
-I$(top_srcdir) \
|
||||
$(IKVM_JNI_CFLAGS) \
|
||||
$(GMODULE_CFLAGS) \
|
||||
$(GLIB_CFLAGS)
|
||||
|
||||
if JIT_SUPPORTED
|
||||
|
||||
lib_LTLIBRARIES = libmono-ikvm-jni.la
|
||||
endif
|
||||
|
||||
libmono_ikvm_jni_la_SOURCES = mono-jni.c
|
||||
libmono_ikvm_jni_la_LIBADD = $(top_builddir)/mono/mini/libmono.la
|
||||
|
||||
@@ -1,23 +0,0 @@
|
||||
aclocal.m4
|
||||
config.cache
|
||||
config.guess
|
||||
config.h
|
||||
config.h.in
|
||||
config.log
|
||||
config.status
|
||||
config.sub
|
||||
configure
|
||||
libtool
|
||||
ltmain.sh
|
||||
ltconfig
|
||||
Makefile
|
||||
Makefile.in
|
||||
stamp-h
|
||||
stamp-h.in
|
||||
*.lo
|
||||
.deps
|
||||
.libs
|
||||
*.la
|
||||
autom4te.cache
|
||||
libgc-mono.pc
|
||||
libgc-mono-*.tar.gz
|
||||
@@ -1,623 +0,0 @@
|
||||
|
||||
|
||||
/******************************************************************
|
||||
|
||||
AmigaOS-spesific routines for GC.
|
||||
This file is normally included from os_dep.c
|
||||
|
||||
******************************************************************/
|
||||
|
||||
|
||||
#if !defined(GC_AMIGA_DEF) && !defined(GC_AMIGA_SB) && !defined(GC_AMIGA_DS) && !defined(GC_AMIGA_AM)
|
||||
# include "gc_priv.h"
|
||||
# include <stdio.h>
|
||||
# include <signal.h>
|
||||
# define GC_AMIGA_DEF
|
||||
# define GC_AMIGA_SB
|
||||
# define GC_AMIGA_DS
|
||||
# define GC_AMIGA_AM
|
||||
#endif
|
||||
|
||||
|
||||
#ifdef GC_AMIGA_DEF
|
||||
|
||||
# ifndef __GNUC__
|
||||
# include <exec/exec.h>
|
||||
# endif
|
||||
# include <proto/exec.h>
|
||||
# include <proto/dos.h>
|
||||
# include <dos/dosextens.h>
|
||||
# include <workbench/startup.h>
|
||||
|
||||
#endif
|
||||
|
||||
|
||||
|
||||
|
||||
#ifdef GC_AMIGA_SB
|
||||
|
||||
/******************************************************************
|
||||
Find the base of the stack.
|
||||
******************************************************************/
|
||||
|
||||
ptr_t GC_get_stack_base()
|
||||
{
|
||||
struct Process *proc = (struct Process*)SysBase->ThisTask;
|
||||
|
||||
/* Reference: Amiga Guru Book Pages: 42,567,574 */
|
||||
if (proc->pr_Task.tc_Node.ln_Type==NT_PROCESS
|
||||
&& proc->pr_CLI != NULL) {
|
||||
/* first ULONG is StackSize */
|
||||
/*longPtr = proc->pr_ReturnAddr;
|
||||
size = longPtr[0];*/
|
||||
|
||||
return (char *)proc->pr_ReturnAddr + sizeof(ULONG);
|
||||
} else {
|
||||
return (char *)proc->pr_Task.tc_SPUpper;
|
||||
}
|
||||
}
|
||||
|
||||
#if 0 /* old version */
|
||||
ptr_t GC_get_stack_base()
|
||||
{
|
||||
extern struct WBStartup *_WBenchMsg;
|
||||
extern long __base;
|
||||
extern long __stack;
|
||||
struct Task *task;
|
||||
struct Process *proc;
|
||||
struct CommandLineInterface *cli;
|
||||
long size;
|
||||
|
||||
if ((task = FindTask(0)) == 0) {
|
||||
GC_err_puts("Cannot find own task structure\n");
|
||||
ABORT("task missing");
|
||||
}
|
||||
proc = (struct Process *)task;
|
||||
cli = BADDR(proc->pr_CLI);
|
||||
|
||||
if (_WBenchMsg != 0 || cli == 0) {
|
||||
size = (char *)task->tc_SPUpper - (char *)task->tc_SPLower;
|
||||
} else {
|
||||
size = cli->cli_DefaultStack * 4;
|
||||
}
|
||||
return (ptr_t)(__base + GC_max(size, __stack));
|
||||
}
|
||||
#endif
|
||||
|
||||
|
||||
#endif
|
||||
|
||||
|
||||
#ifdef GC_AMIGA_DS
|
||||
/******************************************************************
|
||||
Register data segments.
|
||||
******************************************************************/
|
||||
|
||||
void GC_register_data_segments()
|
||||
{
|
||||
struct Process *proc;
|
||||
struct CommandLineInterface *cli;
|
||||
BPTR myseglist;
|
||||
ULONG *data;
|
||||
|
||||
int num;
|
||||
|
||||
|
||||
# ifdef __GNUC__
|
||||
ULONG dataSegSize;
|
||||
GC_bool found_segment = FALSE;
|
||||
extern char __data_size[];
|
||||
|
||||
dataSegSize=__data_size+8;
|
||||
/* Can`t find the Location of __data_size, because
|
||||
it`s possible that is it, inside the segment. */
|
||||
|
||||
# endif
|
||||
|
||||
proc= (struct Process*)SysBase->ThisTask;
|
||||
|
||||
/* Reference: Amiga Guru Book Pages: 538ff,565,573
|
||||
and XOper.asm */
|
||||
if (proc->pr_Task.tc_Node.ln_Type==NT_PROCESS) {
|
||||
if (proc->pr_CLI == NULL) {
|
||||
myseglist = proc->pr_SegList;
|
||||
} else {
|
||||
/* ProcLoaded 'Loaded as a command: '*/
|
||||
cli = BADDR(proc->pr_CLI);
|
||||
myseglist = cli->cli_Module;
|
||||
}
|
||||
} else {
|
||||
ABORT("Not a Process.");
|
||||
}
|
||||
|
||||
if (myseglist == NULL) {
|
||||
ABORT("Arrrgh.. can't find segments, aborting");
|
||||
}
|
||||
|
||||
/* xoper hunks Shell Process */
|
||||
|
||||
num=0;
|
||||
for (data = (ULONG *)BADDR(myseglist); data != NULL;
|
||||
data = (ULONG *)BADDR(data[0])) {
|
||||
if (((ULONG) GC_register_data_segments < (ULONG) &data[1]) ||
|
||||
((ULONG) GC_register_data_segments > (ULONG) &data[1] + data[-1])) {
|
||||
# ifdef __GNUC__
|
||||
if (dataSegSize == data[-1]) {
|
||||
found_segment = TRUE;
|
||||
}
|
||||
# endif
|
||||
GC_add_roots_inner((char *)&data[1],
|
||||
((char *)&data[1]) + data[-1], FALSE);
|
||||
}
|
||||
++num;
|
||||
} /* for */
|
||||
# ifdef __GNUC__
|
||||
if (!found_segment) {
|
||||
ABORT("Can`t find correct Segments.\nSolution: Use an newer version of ixemul.library");
|
||||
}
|
||||
# endif
|
||||
}
|
||||
|
||||
#if 0 /* old version */
|
||||
void GC_register_data_segments()
|
||||
{
|
||||
extern struct WBStartup *_WBenchMsg;
|
||||
struct Process *proc;
|
||||
struct CommandLineInterface *cli;
|
||||
BPTR myseglist;
|
||||
ULONG *data;
|
||||
|
||||
if ( _WBenchMsg != 0 ) {
|
||||
if ((myseglist = _WBenchMsg->sm_Segment) == 0) {
|
||||
GC_err_puts("No seglist from workbench\n");
|
||||
return;
|
||||
}
|
||||
} else {
|
||||
if ((proc = (struct Process *)FindTask(0)) == 0) {
|
||||
GC_err_puts("Cannot find process structure\n");
|
||||
return;
|
||||
}
|
||||
if ((cli = BADDR(proc->pr_CLI)) == 0) {
|
||||
GC_err_puts("No CLI\n");
|
||||
return;
|
||||
}
|
||||
if ((myseglist = cli->cli_Module) == 0) {
|
||||
GC_err_puts("No seglist from CLI\n");
|
||||
return;
|
||||
}
|
||||
}
|
||||
|
||||
for (data = (ULONG *)BADDR(myseglist); data != 0;
|
||||
data = (ULONG *)BADDR(data[0])) {
|
||||
# ifdef AMIGA_SKIP_SEG
|
||||
if (((ULONG) GC_register_data_segments < (ULONG) &data[1]) ||
|
||||
((ULONG) GC_register_data_segments > (ULONG) &data[1] + data[-1])) {
|
||||
# else
|
||||
{
|
||||
# endif /* AMIGA_SKIP_SEG */
|
||||
GC_add_roots_inner((char *)&data[1],
|
||||
((char *)&data[1]) + data[-1], FALSE);
|
||||
}
|
||||
}
|
||||
}
|
||||
#endif /* old version */
|
||||
|
||||
|
||||
#endif
|
||||
|
||||
|
||||
|
||||
#ifdef GC_AMIGA_AM
|
||||
|
||||
#ifndef GC_AMIGA_FASTALLOC
|
||||
|
||||
void *GC_amiga_allocwrapper(size_t size,void *(*AllocFunction)(size_t size2)){
|
||||
return (*AllocFunction)(size);
|
||||
}
|
||||
|
||||
void *(*GC_amiga_allocwrapper_do)(size_t size,void *(*AllocFunction)(size_t size2))
|
||||
=GC_amiga_allocwrapper;
|
||||
|
||||
#else
|
||||
|
||||
|
||||
|
||||
|
||||
void *GC_amiga_allocwrapper_firsttime(size_t size,void *(*AllocFunction)(size_t size2));
|
||||
|
||||
void *(*GC_amiga_allocwrapper_do)(size_t size,void *(*AllocFunction)(size_t size2))
|
||||
=GC_amiga_allocwrapper_firsttime;
|
||||
|
||||
|
||||
/******************************************************************
|
||||
Amiga-spesific routines to obtain memory, and force GC to give
|
||||
back fast-mem whenever possible.
|
||||
These hacks makes gc-programs go many times faster when
|
||||
the amiga is low on memory, and are therefore strictly necesarry.
|
||||
|
||||
-Kjetil S. Matheussen, 2000.
|
||||
******************************************************************/
|
||||
|
||||
|
||||
|
||||
/* List-header for all allocated memory. */
|
||||
|
||||
struct GC_Amiga_AllocedMemoryHeader{
|
||||
ULONG size;
|
||||
struct GC_Amiga_AllocedMemoryHeader *next;
|
||||
};
|
||||
struct GC_Amiga_AllocedMemoryHeader *GC_AMIGAMEM=(struct GC_Amiga_AllocedMemoryHeader *)(int)~(NULL);
|
||||
|
||||
|
||||
|
||||
/* Type of memory. Once in the execution of a program, this might change to MEMF_ANY|MEMF_CLEAR */
|
||||
|
||||
ULONG GC_AMIGA_MEMF = MEMF_FAST | MEMF_CLEAR;
|
||||
|
||||
|
||||
/* Prevents GC_amiga_get_mem from allocating memory if this one is TRUE. */
|
||||
#ifndef GC_AMIGA_ONLYFAST
|
||||
BOOL GC_amiga_dontalloc=FALSE;
|
||||
#endif
|
||||
|
||||
#ifdef GC_AMIGA_PRINTSTATS
|
||||
int succ=0,succ2=0;
|
||||
int nsucc=0,nsucc2=0;
|
||||
int nullretries=0;
|
||||
int numcollects=0;
|
||||
int chipa=0;
|
||||
int allochip=0;
|
||||
int allocfast=0;
|
||||
int cur0=0;
|
||||
int cur1=0;
|
||||
int cur10=0;
|
||||
int cur50=0;
|
||||
int cur150=0;
|
||||
int cur151=0;
|
||||
int ncur0=0;
|
||||
int ncur1=0;
|
||||
int ncur10=0;
|
||||
int ncur50=0;
|
||||
int ncur150=0;
|
||||
int ncur151=0;
|
||||
#endif
|
||||
|
||||
/* Free everything at program-end. */
|
||||
|
||||
void GC_amiga_free_all_mem(void){
|
||||
struct GC_Amiga_AllocedMemoryHeader *gc_am=(struct GC_Amiga_AllocedMemoryHeader *)(~(int)(GC_AMIGAMEM));
|
||||
struct GC_Amiga_AllocedMemoryHeader *temp;
|
||||
|
||||
#ifdef GC_AMIGA_PRINTSTATS
|
||||
printf("\n\n"
|
||||
"%d bytes of chip-mem, and %d bytes of fast-mem where allocated from the OS.\n",
|
||||
allochip,allocfast
|
||||
);
|
||||
printf(
|
||||
"%d bytes of chip-mem were returned from the GC_AMIGA_FASTALLOC supported allocating functions.\n",
|
||||
chipa
|
||||
);
|
||||
printf("\n");
|
||||
printf("GC_gcollect was called %d times to avoid returning NULL or start allocating with the MEMF_ANY flag.\n",numcollects);
|
||||
printf("%d of them was a success. (the others had to use allocation from the OS.)\n",nullretries);
|
||||
printf("\n");
|
||||
printf("Succeded forcing %d gc-allocations (%d bytes) of chip-mem to be fast-mem.\n",succ,succ2);
|
||||
printf("Failed forcing %d gc-allocations (%d bytes) of chip-mem to be fast-mem.\n",nsucc,nsucc2);
|
||||
printf("\n");
|
||||
printf(
|
||||
"Number of retries before succeding a chip->fast force:\n"
|
||||
"0: %d, 1: %d, 2-9: %d, 10-49: %d, 50-149: %d, >150: %d\n",
|
||||
cur0,cur1,cur10,cur50,cur150,cur151
|
||||
);
|
||||
printf(
|
||||
"Number of retries before giving up a chip->fast force:\n"
|
||||
"0: %d, 1: %d, 2-9: %d, 10-49: %d, 50-149: %d, >150: %d\n",
|
||||
ncur0,ncur1,ncur10,ncur50,ncur150,ncur151
|
||||
);
|
||||
#endif
|
||||
|
||||
while(gc_am!=NULL){
|
||||
temp=gc_am->next;
|
||||
FreeMem(gc_am,gc_am->size);
|
||||
gc_am=(struct GC_Amiga_AllocedMemoryHeader *)(~(int)(temp));
|
||||
}
|
||||
}
|
||||
|
||||
#ifndef GC_AMIGA_ONLYFAST
|
||||
|
||||
/* All memory with address lower than this one is chip-mem. */
|
||||
|
||||
char *chipmax;
|
||||
|
||||
|
||||
/*
|
||||
* Allways set to the last size of memory tried to be allocated.
|
||||
* Needed to ensure allocation when the size is bigger than 100000.
|
||||
*
|
||||
*/
|
||||
size_t latestsize;
|
||||
|
||||
#endif
|
||||
|
||||
|
||||
/*
|
||||
* The actual function that is called with the GET_MEM macro.
|
||||
*
|
||||
*/
|
||||
|
||||
void *GC_amiga_get_mem(size_t size){
|
||||
struct GC_Amiga_AllocedMemoryHeader *gc_am;
|
||||
|
||||
#ifndef GC_AMIGA_ONLYFAST
|
||||
if(GC_amiga_dontalloc==TRUE){
|
||||
// printf("rejected, size: %d, latestsize: %d\n",size,latestsize);
|
||||
return NULL;
|
||||
}
|
||||
|
||||
// We really don't want to use chip-mem, but if we must, then as little as possible.
|
||||
if(GC_AMIGA_MEMF==(MEMF_ANY|MEMF_CLEAR) && size>100000 && latestsize<50000) return NULL;
|
||||
#endif
|
||||
|
||||
gc_am=AllocMem((ULONG)(size + sizeof(struct GC_Amiga_AllocedMemoryHeader)),GC_AMIGA_MEMF);
|
||||
if(gc_am==NULL) return NULL;
|
||||
|
||||
gc_am->next=GC_AMIGAMEM;
|
||||
gc_am->size=size + sizeof(struct GC_Amiga_AllocedMemoryHeader);
|
||||
GC_AMIGAMEM=(struct GC_Amiga_AllocedMemoryHeader *)(~(int)(gc_am));
|
||||
|
||||
// printf("Allocated %d (%d) bytes at address: %x. Latest: %d\n",size,tot,gc_am,latestsize);
|
||||
|
||||
#ifdef GC_AMIGA_PRINTSTATS
|
||||
if((char *)gc_am<chipmax){
|
||||
allochip+=size;
|
||||
}else{
|
||||
allocfast+=size;
|
||||
}
|
||||
#endif
|
||||
|
||||
return gc_am+1;
|
||||
|
||||
}
|
||||
|
||||
|
||||
|
||||
|
||||
#ifndef GC_AMIGA_ONLYFAST
|
||||
|
||||
/* Tries very hard to force GC to find fast-mem to return. Done recursively
|
||||
* to hold the rejected memory-pointers reachable from the collector in an
|
||||
* easy way.
|
||||
*
|
||||
*/
|
||||
#ifdef GC_AMIGA_RETRY
|
||||
void *GC_amiga_rec_alloc(size_t size,void *(*AllocFunction)(size_t size2),const int rec){
|
||||
void *ret;
|
||||
|
||||
ret=(*AllocFunction)(size);
|
||||
|
||||
#ifdef GC_AMIGA_PRINTSTATS
|
||||
if((char *)ret>chipmax || ret==NULL){
|
||||
if(ret==NULL){
|
||||
nsucc++;
|
||||
nsucc2+=size;
|
||||
if(rec==0) ncur0++;
|
||||
if(rec==1) ncur1++;
|
||||
if(rec>1 && rec<10) ncur10++;
|
||||
if(rec>=10 && rec<50) ncur50++;
|
||||
if(rec>=50 && rec<150) ncur150++;
|
||||
if(rec>=150) ncur151++;
|
||||
}else{
|
||||
succ++;
|
||||
succ2+=size;
|
||||
if(rec==0) cur0++;
|
||||
if(rec==1) cur1++;
|
||||
if(rec>1 && rec<10) cur10++;
|
||||
if(rec>=10 && rec<50) cur50++;
|
||||
if(rec>=50 && rec<150) cur150++;
|
||||
if(rec>=150) cur151++;
|
||||
}
|
||||
}
|
||||
#endif
|
||||
|
||||
if (((char *)ret)<=chipmax && ret!=NULL && (rec<(size>500000?9:size/5000))){
|
||||
ret=GC_amiga_rec_alloc(size,AllocFunction,rec+1);
|
||||
// GC_free(ret2);
|
||||
}
|
||||
|
||||
return ret;
|
||||
}
|
||||
#endif
|
||||
|
||||
|
||||
/* The allocating-functions defined inside the amiga-blocks in gc.h is called
|
||||
* via these functions.
|
||||
*/
|
||||
|
||||
|
||||
void *GC_amiga_allocwrapper_any(size_t size,void *(*AllocFunction)(size_t size2)){
|
||||
void *ret,*ret2;
|
||||
|
||||
GC_amiga_dontalloc=TRUE; // Pretty tough thing to do, but its indeed necesarry.
|
||||
latestsize=size;
|
||||
|
||||
ret=(*AllocFunction)(size);
|
||||
|
||||
if(((char *)ret) <= chipmax){
|
||||
if(ret==NULL){
|
||||
//Give GC access to allocate memory.
|
||||
#ifdef GC_AMIGA_GC
|
||||
if(!GC_dont_gc){
|
||||
GC_gcollect();
|
||||
#ifdef GC_AMIGA_PRINTSTATS
|
||||
numcollects++;
|
||||
#endif
|
||||
ret=(*AllocFunction)(size);
|
||||
}
|
||||
#endif
|
||||
if(ret==NULL){
|
||||
GC_amiga_dontalloc=FALSE;
|
||||
ret=(*AllocFunction)(size);
|
||||
if(ret==NULL){
|
||||
WARN("Out of Memory! Returning NIL!\n", 0);
|
||||
}
|
||||
}
|
||||
#ifdef GC_AMIGA_PRINTSTATS
|
||||
else{
|
||||
nullretries++;
|
||||
}
|
||||
if(ret!=NULL && (char *)ret<=chipmax) chipa+=size;
|
||||
#endif
|
||||
}
|
||||
#ifdef GC_AMIGA_RETRY
|
||||
else{
|
||||
/* We got chip-mem. Better try again and again and again etc., we might get fast-mem sooner or later... */
|
||||
/* Using gctest to check the effectiviness of doing this, does seldom give a very good result. */
|
||||
/* However, real programs doesn't normally rapidly allocate and deallocate. */
|
||||
// printf("trying to force... %d bytes... ",size);
|
||||
if(
|
||||
AllocFunction!=GC_malloc_uncollectable
|
||||
#ifdef ATOMIC_UNCOLLECTABLE
|
||||
&& AllocFunction!=GC_malloc_atomic_uncollectable
|
||||
#endif
|
||||
){
|
||||
ret2=GC_amiga_rec_alloc(size,AllocFunction,0);
|
||||
}else{
|
||||
ret2=(*AllocFunction)(size);
|
||||
#ifdef GC_AMIGA_PRINTSTATS
|
||||
if((char *)ret2<chipmax || ret2==NULL){
|
||||
nsucc++;
|
||||
nsucc2+=size;
|
||||
ncur0++;
|
||||
}else{
|
||||
succ++;
|
||||
succ2+=size;
|
||||
cur0++;
|
||||
}
|
||||
#endif
|
||||
}
|
||||
if(((char *)ret2)>chipmax){
|
||||
// printf("Succeeded.\n");
|
||||
GC_free(ret);
|
||||
ret=ret2;
|
||||
}else{
|
||||
GC_free(ret2);
|
||||
// printf("But did not succeed.\n");
|
||||
}
|
||||
}
|
||||
#endif
|
||||
}
|
||||
|
||||
GC_amiga_dontalloc=FALSE;
|
||||
|
||||
return ret;
|
||||
}
|
||||
|
||||
|
||||
|
||||
void (*GC_amiga_toany)(void)=NULL;
|
||||
|
||||
void GC_amiga_set_toany(void (*func)(void)){
|
||||
GC_amiga_toany=func;
|
||||
}
|
||||
|
||||
#endif // !GC_AMIGA_ONLYFAST
|
||||
|
||||
|
||||
void *GC_amiga_allocwrapper_fast(size_t size,void *(*AllocFunction)(size_t size2)){
|
||||
void *ret;
|
||||
|
||||
ret=(*AllocFunction)(size);
|
||||
|
||||
if(ret==NULL){
|
||||
// Enable chip-mem allocation.
|
||||
// printf("ret==NULL\n");
|
||||
#ifdef GC_AMIGA_GC
|
||||
if(!GC_dont_gc){
|
||||
GC_gcollect();
|
||||
#ifdef GC_AMIGA_PRINTSTATS
|
||||
numcollects++;
|
||||
#endif
|
||||
ret=(*AllocFunction)(size);
|
||||
}
|
||||
#endif
|
||||
if(ret==NULL){
|
||||
#ifndef GC_AMIGA_ONLYFAST
|
||||
GC_AMIGA_MEMF=MEMF_ANY | MEMF_CLEAR;
|
||||
if(GC_amiga_toany!=NULL) (*GC_amiga_toany)();
|
||||
GC_amiga_allocwrapper_do=GC_amiga_allocwrapper_any;
|
||||
return GC_amiga_allocwrapper_any(size,AllocFunction);
|
||||
#endif
|
||||
}
|
||||
#ifdef GC_AMIGA_PRINTSTATS
|
||||
else{
|
||||
nullretries++;
|
||||
}
|
||||
#endif
|
||||
}
|
||||
|
||||
return ret;
|
||||
}
|
||||
|
||||
void *GC_amiga_allocwrapper_firsttime(size_t size,void *(*AllocFunction)(size_t size2)){
|
||||
atexit(&GC_amiga_free_all_mem);
|
||||
chipmax=(char *)SysBase->MaxLocMem; // For people still having SysBase in chip-mem, this might speed up a bit.
|
||||
GC_amiga_allocwrapper_do=GC_amiga_allocwrapper_fast;
|
||||
return GC_amiga_allocwrapper_fast(size,AllocFunction);
|
||||
}
|
||||
|
||||
|
||||
#endif //GC_AMIGA_FASTALLOC
|
||||
|
||||
|
||||
|
||||
/*
|
||||
* The wrapped realloc function.
|
||||
*
|
||||
*/
|
||||
void *GC_amiga_realloc(void *old_object,size_t new_size_in_bytes){
|
||||
#ifndef GC_AMIGA_FASTALLOC
|
||||
return GC_realloc(old_object,new_size_in_bytes);
|
||||
#else
|
||||
void *ret;
|
||||
latestsize=new_size_in_bytes;
|
||||
ret=GC_realloc(old_object,new_size_in_bytes);
|
||||
if(ret==NULL && GC_AMIGA_MEMF==(MEMF_FAST | MEMF_CLEAR)){
|
||||
/* Out of fast-mem. */
|
||||
#ifdef GC_AMIGA_GC
|
||||
if(!GC_dont_gc){
|
||||
GC_gcollect();
|
||||
#ifdef GC_AMIGA_PRINTSTATS
|
||||
numcollects++;
|
||||
#endif
|
||||
ret=GC_realloc(old_object,new_size_in_bytes);
|
||||
}
|
||||
#endif
|
||||
if(ret==NULL){
|
||||
#ifndef GC_AMIGA_ONLYFAST
|
||||
GC_AMIGA_MEMF=MEMF_ANY | MEMF_CLEAR;
|
||||
if(GC_amiga_toany!=NULL) (*GC_amiga_toany)();
|
||||
GC_amiga_allocwrapper_do=GC_amiga_allocwrapper_any;
|
||||
ret=GC_realloc(old_object,new_size_in_bytes);
|
||||
#endif
|
||||
}
|
||||
#ifdef GC_AMIGA_PRINTSTATS
|
||||
else{
|
||||
nullretries++;
|
||||
}
|
||||
#endif
|
||||
}
|
||||
if(ret==NULL){
|
||||
WARN("Out of Memory! Returning NIL!\n", 0);
|
||||
}
|
||||
#ifdef GC_AMIGA_PRINTSTATS
|
||||
if(((char *)ret)<chipmax && ret!=NULL){
|
||||
chipa+=new_size_in_bytes;
|
||||
}
|
||||
#endif
|
||||
return ret;
|
||||
#endif
|
||||
}
|
||||
|
||||
#endif //GC_AMIGA_AM
|
||||
|
||||
|
||||
@@ -1,88 +0,0 @@
|
||||
# Makefile for Borland C++ 5.5 on NT
|
||||
# If you have the Borland assembler, remove "-DUSE_GENERIC"
|
||||
#
|
||||
bc= c:\Borland\BCC55
|
||||
bcbin= $(bc)\bin
|
||||
bclib= $(bc)\lib
|
||||
bcinclude= $(bc)\include
|
||||
|
||||
gcinclude1 = $(bc)\gc6.2\include
|
||||
gcinclude2 = $(bc)\gc6.2\cord
|
||||
|
||||
cc= $(bcbin)\bcc32
|
||||
rc= $(bcbin)\brc32
|
||||
lib= $(bcbin)\tlib
|
||||
link= $(bcbin)\ilink32
|
||||
cflags= -O2 -R -v- -vi -H -H=gc.csm -I$(bcinclude);$(gcinclude1);$(gcinclude2) -L$(bclib) \
|
||||
-w-pro -w-aus -w-par -w-ccc -w-rch -a4 -D__STDC__=0
|
||||
#defines= -DSILENT
|
||||
defines= -DSILENT -DALL_INTERIOR_POINTERS -DUSE_GENERIC -DNO_GETENV -DJAVA_FINALIZATION -DGC_OPERATOR_NEW_ARRAY
|
||||
|
||||
.c.obj:
|
||||
$(cc) @&&|
|
||||
$(cdebug) $(cflags) $(cvars) $(defines) -o$* -c $*.c
|
||||
|
|
||||
|
||||
.cpp.obj:
|
||||
$(cc) @&&|
|
||||
$(cdebug) $(cflags) $(cvars) $(defines) -o$* -c $*.cpp
|
||||
|
|
||||
|
||||
.rc.res:
|
||||
$(rc) -i$(bcinclude) -r -fo$* $*.rc
|
||||
|
||||
XXXOBJS= XXXalloc.obj XXXreclaim.obj XXXallchblk.obj XXXmisc.obj \
|
||||
XXXmach_dep.obj XXXos_dep.obj XXXmark_rts.obj XXXheaders.obj XXXmark.obj \
|
||||
XXXobj_map.obj XXXblacklst.obj XXXfinalize.obj XXXnew_hblk.obj \
|
||||
XXXdbg_mlc.obj XXXmalloc.obj XXXstubborn.obj XXXdyn_load.obj \
|
||||
XXXtypd_mlc.obj XXXptr_chck.obj XXXgc_cpp.obj XXXmallocx.obj
|
||||
|
||||
OBJS= $(XXXOBJS:XXX=)
|
||||
|
||||
all: gctest.exe cord\de.exe test_cpp.exe
|
||||
|
||||
$(OBJS) test.obj: include\private\gc_priv.h include\private\gc_hdrs.h include\gc.h include\private\gcconfig.h MAKEFILE
|
||||
|
||||
gc.lib: $(OBJS)
|
||||
del gc.lib
|
||||
$(lib) $* @&&|
|
||||
$(XXXOBJS:XXX=+)
|
||||
|
|
||||
|
||||
gctest.exe: tests\test.obj gc.lib
|
||||
$(cc) @&&|
|
||||
$(cflags) -W -e$* tests\test.obj gc.lib
|
||||
|
|
||||
|
||||
cord\de.obj cord\de_win.obj: include\cord.h include\private\cord_pos.h cord\de_win.h \
|
||||
cord\de_cmds.h
|
||||
|
||||
cord\de.exe: cord\cordbscs.obj cord\cordxtra.obj cord\de.obj cord\de_win.obj \
|
||||
cord\de_win.res gc.lib
|
||||
$(cc) @&&|
|
||||
$(cflags) -W -e$* cord\cordbscs.obj cord\cordxtra.obj \
|
||||
cord\de.obj cord\de_win.obj gc.lib
|
||||
|
|
||||
$(rc) cord\de_win.res cord\de.exe
|
||||
|
||||
gc_cpp.obj: include\gc_cpp.h include\gc.h
|
||||
|
||||
gc_cpp.cpp: gc_cpp.cc
|
||||
copy gc_cpp.cc gc_cpp.cpp
|
||||
|
||||
test_cpp.cpp: tests\test_cpp.cc
|
||||
copy tests\test_cpp.cc test_cpp.cpp
|
||||
|
||||
test_cpp.exe: test_cpp.obj include\gc_cpp.h include\gc.h gc.lib
|
||||
$(cc) @&&|
|
||||
$(cflags) -W -e$* test_cpp.obj gc.lib
|
||||
|
|
||||
|
||||
scratch:
|
||||
-del *.obj *.res *.exe *.csm cord\*.obj cord\*.res cord\*.exe cord\*.csm
|
||||
|
||||
clean:
|
||||
del gc.lib
|
||||
del *.obj
|
||||
del tests\test.obj
|
||||
|
||||
@@ -1,185 +0,0 @@
|
||||
2004-07-27 John Merryweather Cooper <[email protected]>
|
||||
# configure.in: Fix compiler flags selection to be correct for
|
||||
both FreeBSD 4.x (aka -STABLE) and FreeBSD 5.x (aka -CURRENT).
|
||||
|
||||
2004-07-01 Zoltan Varga <[email protected]>
|
||||
|
||||
* include/install-sh doc/install-sh: New files.
|
||||
|
||||
Thu Jun 24 15:39:00 CEST 2004 Paolo Molaro <[email protected]>
|
||||
|
||||
* alloc.c: use Hans' fix for bug #59557.
|
||||
|
||||
2004-06-18 Dick Porter <[email protected]>
|
||||
|
||||
* darwin_stop_world.c: Make the world actually stop. This seems
|
||||
to fix the finalizer thread signalling deadlock in 58706.
|
||||
|
||||
Sat Jun 5 19:38:20 CEST 2004 Paolo Molaro <[email protected]>
|
||||
|
||||
* alloc.c: temporary fix for bug #59557.
|
||||
|
||||
Tue May 25 13:41:08 EDT 2004 Paolo Molaro <[email protected]>
|
||||
|
||||
* darwin_stop_world.c: more thread died races fixes.
|
||||
|
||||
2004-05-21 Martin Baulig <[email protected]>
|
||||
|
||||
* pthread_stop_world.c: Merged the changes from linux_threads.c.
|
||||
|
||||
* linux_threads.c: Removed, this file no longer exists in the
|
||||
latest upstream version.
|
||||
|
||||
2004-05-20 Martin Baulig <[email protected]>
|
||||
|
||||
* libgc.patch: Removed. There's no need to have this in a
|
||||
separate file, just use CVS.
|
||||
|
||||
2004-05-20 Jackson Harper <[email protected]>
|
||||
|
||||
* Makefile.am: Remove tests from EXTRA_DIST.
|
||||
|
||||
Wed May 19 18:57:30 CEST 2004 Paolo Molaro <[email protected]>
|
||||
|
||||
* darwin_stop_world.c: deal with threads that die before they can be
|
||||
suspended (should solce the thread_info failed aborts).
|
||||
|
||||
Wed May 19 18:42:23 CEST 2004 Paolo Molaro <[email protected]>
|
||||
|
||||
* configure.in, Makefile.am: update correctly from the libgc sources:
|
||||
this excludes linux_threads.c from the build: Martin will have to port
|
||||
his changes to the new code, since linux_threads.c doesn't exist
|
||||
anymore in the upstream sources.
|
||||
|
||||
2004-05-19 Raja R Harinath <[email protected]>
|
||||
|
||||
* configure.in ($enable_embed_check): New internal option that
|
||||
will be passed by mono's configure.
|
||||
(THREADS): Use $with_libgc_threads, which is passed by mono's
|
||||
configure.
|
||||
|
||||
2004-05-18 Martin Baulig <[email protected]>
|
||||
|
||||
Importing Boehm GC 6.3 alpha 6.
|
||||
|
||||
2004-03-24 Ben Maurer <[email protected]>
|
||||
|
||||
* gcj_mlc.c: Do finalization.
|
||||
|
||||
This is a patch from Hans Boehm that fixes a bug where
|
||||
GC_gcj_malloc never does finalization.
|
||||
|
||||
See:
|
||||
http://www.hpl.hp.com/hosted/linux/mail-archives/gc/2004-March/000359.html
|
||||
http://bugzilla.ximian.com/show_bug.cgi?id=55644
|
||||
|
||||
Fixes #55644.
|
||||
|
||||
I am checking this in at Zoltan's request.
|
||||
|
||||
2003-11-07 Zoltan Varga <[email protected]>
|
||||
|
||||
* autogen.sh: Fix previous patch and silence noisy GNU which.
|
||||
|
||||
* autogen.sh: Applied patch from Peter Teichman ([email protected]).
|
||||
Use 'glibtool' instead of 'libtool' on OSX.
|
||||
|
||||
2003-10-01 Pedro Martínez Juliá <[email protected]>
|
||||
|
||||
* Makefile.am: Change lib_LTLIBRARIES to noinst_LTLIBRARIES because
|
||||
of libmonogc is included inside libmono and without this change it
|
||||
breaks down and not include it.
|
||||
|
||||
2003-09-26 Zoltan Varga <[email protected]>
|
||||
|
||||
* mips_sgi_mach_dep.S: Remove since it is a duplicate of .s and it
|
||||
breaks cvs checkout under windows.
|
||||
|
||||
Wed Sep 17 17:29:11 CEST 2003 Paolo Molaro <[email protected]>
|
||||
|
||||
* configure.in: disable ALL_INTERIOR_POINTERS: managed pointers
|
||||
are not allowed in the heap, so this should be safe and it helps
|
||||
reducing false positives.
|
||||
|
||||
2003-08-28 Dick Porter <[email protected]>
|
||||
|
||||
* pthread_support.c:
|
||||
* aix_irix_threads.c: Applied bug fix from alpha libgc
|
||||
|
||||
2003-08-26 Martin Baulig <[email protected]>
|
||||
|
||||
Imported Boehm GC 6.2.
|
||||
|
||||
2003-06-20 Martin Baulig <[email protected]>
|
||||
|
||||
* include/gc.h: Fixed compiler warnings.
|
||||
|
||||
2003-06-09 Martin Baulig <[email protected]>
|
||||
|
||||
* include/libgc-mono-debugger.h: New header file. This is
|
||||
installed as <mono/private/libgc-mono-debugger.h>, but it's a
|
||||
private header files and may only be used by the Mono Debugger.
|
||||
|
||||
* Makefile.am: Build libmonogc.la and install it. For some
|
||||
reason, using a libtool convenience archive doesn't work here.
|
||||
|
||||
2003-06-08 Martin Baulig <[email protected]>
|
||||
|
||||
* libgc-mono.pc.in: Removed.
|
||||
|
||||
* Makefile.am: Don't install anything.
|
||||
|
||||
* configure.in: We're now embedded in mono and can't be built as a
|
||||
stand-alone module any longer.
|
||||
|
||||
2003-05-17 Martin Baulig <[email protected]>
|
||||
|
||||
* libgc.patch, libgc.log: Added patch against the original version
|
||||
and a ChangeLog entry.
|
||||
|
||||
2003-05-17 Martin Baulig <[email protected]>
|
||||
|
||||
* misc.c, os_dep.c, include/private/gc_priv.h: Back to the
|
||||
original version.
|
||||
|
||||
* linux_threads.c: Use the `gc_thread_vtable' internally in this
|
||||
file instead of calling it from the header files.
|
||||
|
||||
2003-05-16 Martin Baulig <[email protected]>
|
||||
|
||||
* include/gc.h (GCThreadFunctions): Removed the `pthread_*'
|
||||
function since we don't need them here.
|
||||
|
||||
* include/gc_pthread_redirects.h: Put back the original version.
|
||||
|
||||
2003-04-08 Alp Toker <[email protected]>
|
||||
|
||||
* configure.in: fix automake 1.6 compatibility without breaking
|
||||
previous versions (AM_PROG_AS)
|
||||
|
||||
2003-03-31 Martin Baulig <[email protected]>
|
||||
|
||||
* include/gc.h (GCThreadFunctions): New type.
|
||||
(gc_thread_vtable): New global variable.
|
||||
|
||||
* include/gc_pthread_redirects.h: Use the `gc_thread_vtable'
|
||||
vtable for the redirects instead of calling the methods directly.
|
||||
This allows us to override them at runtime.
|
||||
|
||||
2003-03-31 Martin Baulig <[email protected]>
|
||||
|
||||
* include/Makefile.am: Install gc_pthread_redirects.h.
|
||||
|
||||
* include/gc.h: `#include <gc/gc_pthread_redirects.h>', not
|
||||
`#include "gc_pthread_redirects.h"'.
|
||||
|
||||
2003-03-31 Martin Baulig <[email protected]>
|
||||
|
||||
* configure.in: Call the package `libgc-mono'.
|
||||
|
||||
* libgc-mono.pc.in: Added pkg-config files.
|
||||
|
||||
2003-03-31 Martin Baulig <[email protected]>
|
||||
|
||||
This is a modified version of Boehm GC 6.1.
|
||||
|
||||
@@ -1,141 +0,0 @@
|
||||
#
|
||||
# OS/2 specific Makefile for the EMX environment
|
||||
#
|
||||
# You need GNU Make 3.71, gcc 2.5.7, emx 0.8h and GNU fileutils 3.9
|
||||
# or similar tools. C++ interface and de.exe weren't tested.
|
||||
#
|
||||
# Rename this file "Makefile".
|
||||
#
|
||||
|
||||
# Primary targets:
|
||||
# gc.a - builds basic library
|
||||
# c++ - adds C++ interface to library and include directory
|
||||
# cords - adds cords (heavyweight strings) to library and include directory
|
||||
# test - prints porting information, then builds basic version of gc.a, and runs
|
||||
# some tests of collector and cords. Does not add cords or c++ interface to gc.a
|
||||
# cord/de.exe - builds dumb editor based on cords.
|
||||
CC= gcc
|
||||
CXX=g++
|
||||
# Needed only for "make c++", which adds the c++ interface
|
||||
|
||||
CFLAGS= -O -DALL_INTERIOR_POINTERS -DSILENT
|
||||
# Setjmp_test may yield overly optimistic results when compiled
|
||||
# without optimization.
|
||||
# -DSILENT disables statistics printing, and improves performance.
|
||||
# -DCHECKSUMS reports on erroneously clear dirty bits, and unexpectedly
|
||||
# altered stubborn objects, at substantial performance cost.
|
||||
# -DFIND_LEAK causes the collector to assume that all inaccessible
|
||||
# objects should have been explicitly deallocated, and reports exceptions
|
||||
# -DSOLARIS_THREADS enables support for Solaris (thr_) threads.
|
||||
# (Clients should also define SOLARIS_THREADS and then include
|
||||
# gc.h before performing thr_ or GC_ operations.)
|
||||
# -DALL_INTERIOR_POINTERS allows all pointers to the interior
|
||||
# of objects to be recognized. (See gc_private.h for consequences.)
|
||||
# -DSMALL_CONFIG tries to tune the collector for small heap sizes,
|
||||
# usually causing it to use less space in such situations.
|
||||
# Incremental collection no longer works in this case.
|
||||
# -DDONT_ADD_BYTE_AT_END is meaningful only with
|
||||
# -DALL_INTERIOR_POINTERS. Normally -DALL_INTERIOR_POINTERS
|
||||
# causes all objects to be padded so that pointers just past the end of
|
||||
# an object can be recognized. This can be expensive. (The padding
|
||||
# is normally more than one byte due to alignment constraints.)
|
||||
# -DDONT_ADD_BYTE_AT_END disables the padding.
|
||||
|
||||
AR= ar
|
||||
RANLIB= ar s
|
||||
|
||||
# Redefining srcdir allows object code for the nonPCR version of the collector
|
||||
# to be generated in different directories
|
||||
srcdir = .
|
||||
VPATH = $(srcdir)
|
||||
|
||||
OBJS= alloc.o reclaim.o allchblk.o misc.o mach_dep.o os_dep.o mark_rts.o headers.o mark.o obj_map.o blacklst.o finalize.o new_hblk.o dyn_load.o dbg_mlc.o malloc.o stubborn.o checksums.o typd_mlc.o ptr_chck.o mallocx.o
|
||||
|
||||
CORD_OBJS= cord/cordbscs.o cord/cordxtra.o cord/cordprnt.o
|
||||
|
||||
CORD_INCLUDE_FILES= $(srcdir)/gc.h $(srcdir)/cord/cord.h $(srcdir)/cord/ec.h \
|
||||
$(srcdir)/cord/cord_pos.h
|
||||
|
||||
# Libraries needed for curses applications. Only needed for de.
|
||||
CURSES= -lcurses -ltermlib
|
||||
|
||||
# The following is irrelevant on most systems. But a few
|
||||
# versions of make otherwise fork the shell specified in
|
||||
# the SHELL environment variable.
|
||||
SHELL= bash
|
||||
|
||||
SPECIALCFLAGS =
|
||||
# Alternative flags to the C compiler for mach_dep.c.
|
||||
# Mach_dep.c often doesn't like optimization, and it's
|
||||
# not time-critical anyway.
|
||||
|
||||
all: gc.a gctest.exe
|
||||
|
||||
$(OBJS) test.o: $(srcdir)/gc_priv.h $(srcdir)/gc_hdrs.h $(srcdir)/gc.h \
|
||||
$(srcdir)/gcconfig.h $(srcdir)/gc_typed.h
|
||||
# The dependency on Makefile is needed. Changing
|
||||
# options such as -DSILENT affects the size of GC_arrays,
|
||||
# invalidating all .o files that rely on gc_priv.h
|
||||
|
||||
mark.o typd_mlc.o finalize.o: $(srcdir)/include/gc_mark.h $(srcdir)/include/private/gc_pmark.h
|
||||
|
||||
gc.a: $(OBJS)
|
||||
$(AR) ru gc.a $(OBJS)
|
||||
$(RANLIB) gc.a
|
||||
|
||||
cords: $(CORD_OBJS) cord/cordtest.exe
|
||||
$(AR) ru gc.a $(CORD_OBJS)
|
||||
$(RANLIB) gc.a
|
||||
cp $(srcdir)/cord/cord.h include/cord.h
|
||||
cp $(srcdir)/cord/ec.h include/ec.h
|
||||
cp $(srcdir)/cord/cord_pos.h include/cord_pos.h
|
||||
|
||||
gc_cpp.o: $(srcdir)/gc_cpp.cc $(srcdir)/gc_cpp.h
|
||||
$(CXX) -c -O $(srcdir)/gc_cpp.cc
|
||||
|
||||
c++: gc_cpp.o $(srcdir)/gc_cpp.h
|
||||
$(AR) ru gc.a gc_cpp.o
|
||||
$(RANLIB) gc.a
|
||||
cp $(srcdir)/gc_cpp.h include/gc_cpp.h
|
||||
|
||||
mach_dep.o: $(srcdir)/mach_dep.c
|
||||
$(CC) -o mach_dep.o -c $(SPECIALCFLAGS) $(srcdir)/mach_dep.c
|
||||
|
||||
mark_rts.o: $(srcdir)/mark_rts.c
|
||||
$(CC) -o mark_rts.o -c $(CFLAGS) $(srcdir)/mark_rts.c
|
||||
|
||||
cord/cordbscs.o: $(srcdir)/cord/cordbscs.c $(CORD_INCLUDE_FILES)
|
||||
$(CC) $(CFLAGS) -c $(srcdir)/cord/cordbscs.c -o cord/cordbscs.o
|
||||
|
||||
cord/cordxtra.o: $(srcdir)/cord/cordxtra.c $(CORD_INCLUDE_FILES)
|
||||
$(CC) $(CFLAGS) -c $(srcdir)/cord/cordxtra.c -o cord/cordxtra.o
|
||||
|
||||
cord/cordprnt.o: $(srcdir)/cord/cordprnt.c $(CORD_INCLUDE_FILES)
|
||||
$(CC) $(CFLAGS) -c $(srcdir)/cord/cordprnt.c -o cord/cordprnt.o
|
||||
|
||||
cord/cordtest.exe: $(srcdir)/cord/cordtest.c $(CORD_OBJS) gc.a
|
||||
$(CC) $(CFLAGS) -o cord/cordtest.exe $(srcdir)/cord/cordtest.c $(CORD_OBJS) gc.a
|
||||
|
||||
cord/de.exe: $(srcdir)/cord/de.c $(srcdir)/cord/cordbscs.o $(srcdir)/cord/cordxtra.o gc.a
|
||||
$(CC) $(CFLAGS) -o cord/de.exe $(srcdir)/cord/de.c $(srcdir)/cord/cordbscs.o $(srcdir)/cord/cordxtra.o gc.a $(CURSES)
|
||||
|
||||
clean:
|
||||
rm -f gc.a tests/test.o gctest.exe output-local output-diff $(OBJS) \
|
||||
setjmp_test mon.out gmon.out a.out core \
|
||||
$(CORD_OBJS) cord/cordtest.exe cord/de.exe
|
||||
-rm -f *~
|
||||
|
||||
gctest.exe: tests/test.o gc.a
|
||||
$(CC) $(CFLAGS) -o gctest.exe tests/test.o gc.a
|
||||
|
||||
# If an optimized setjmp_test generates a segmentation fault,
|
||||
# odds are your compiler is broken. Gctest may still work.
|
||||
# Try compiling setjmp_t.c unoptimized.
|
||||
setjmp_test.exe: $(srcdir)/setjmp_t.c $(srcdir)/gc.h
|
||||
$(CC) $(CFLAGS) -o setjmp_test.exe $(srcdir)/setjmp_t.c
|
||||
|
||||
test: setjmp_test.exe gctest.exe
|
||||
./setjmp_test
|
||||
./gctest
|
||||
make cord/cordtest.exe
|
||||
cord/cordtest
|
||||
@@ -1,154 +0,0 @@
|
||||
/*
|
||||
MacOS.c
|
||||
|
||||
Some routines for the Macintosh OS port of the Hans-J. Boehm, Alan J. Demers
|
||||
garbage collector.
|
||||
|
||||
<Revision History>
|
||||
|
||||
11/22/94 pcb StripAddress the temporary memory handle for 24-bit mode.
|
||||
11/30/94 pcb Tracking all memory usage so we can deallocate it all at once.
|
||||
02/10/96 pcb Added routine to perform a final collection when
|
||||
unloading shared library.
|
||||
|
||||
by Patrick C. Beard.
|
||||
*/
|
||||
/* Boehm, February 15, 1996 2:55 pm PST */
|
||||
|
||||
#include <Resources.h>
|
||||
#include <Memory.h>
|
||||
#include <LowMem.h>
|
||||
#include <stdio.h>
|
||||
#include <stdlib.h>
|
||||
#include <string.h>
|
||||
|
||||
#include "gc.h"
|
||||
#include "gc_priv.h"
|
||||
|
||||
// use 'CODE' resource 0 to get exact location of the beginning of global space.
|
||||
|
||||
typedef struct {
|
||||
unsigned long aboveA5;
|
||||
unsigned long belowA5;
|
||||
unsigned long JTSize;
|
||||
unsigned long JTOffset;
|
||||
} *CodeZeroPtr, **CodeZeroHandle;
|
||||
|
||||
void* GC_MacGetDataStart()
|
||||
{
|
||||
CodeZeroHandle code0 = (CodeZeroHandle)GetResource('CODE', 0);
|
||||
if (code0) {
|
||||
long belowA5Size = (**code0).belowA5;
|
||||
ReleaseResource((Handle)code0);
|
||||
return (LMGetCurrentA5() - belowA5Size);
|
||||
}
|
||||
fprintf(stderr, "Couldn't load the jump table.");
|
||||
exit(-1);
|
||||
return 0;
|
||||
}
|
||||
|
||||
/* track the use of temporary memory so it can be freed all at once. */
|
||||
|
||||
typedef struct TemporaryMemoryBlock TemporaryMemoryBlock, **TemporaryMemoryHandle;
|
||||
|
||||
struct TemporaryMemoryBlock {
|
||||
TemporaryMemoryHandle nextBlock;
|
||||
char data[];
|
||||
};
|
||||
|
||||
static TemporaryMemoryHandle theTemporaryMemory = NULL;
|
||||
static Boolean firstTime = true;
|
||||
|
||||
void GC_MacFreeTemporaryMemory(void);
|
||||
|
||||
Ptr GC_MacTemporaryNewPtr(size_t size, Boolean clearMemory)
|
||||
{
|
||||
static Boolean firstTime = true;
|
||||
OSErr result;
|
||||
TemporaryMemoryHandle tempMemBlock;
|
||||
Ptr tempPtr = nil;
|
||||
|
||||
tempMemBlock = (TemporaryMemoryHandle)TempNewHandle(size + sizeof(TemporaryMemoryBlock), &result);
|
||||
if (tempMemBlock && result == noErr) {
|
||||
HLockHi((Handle)tempMemBlock);
|
||||
tempPtr = (**tempMemBlock).data;
|
||||
if (clearMemory) memset(tempPtr, 0, size);
|
||||
tempPtr = StripAddress(tempPtr);
|
||||
|
||||
// keep track of the allocated blocks.
|
||||
(**tempMemBlock).nextBlock = theTemporaryMemory;
|
||||
theTemporaryMemory = tempMemBlock;
|
||||
}
|
||||
|
||||
# if !defined(SHARED_LIBRARY_BUILD)
|
||||
// install an exit routine to clean up the memory used at the end.
|
||||
if (firstTime) {
|
||||
atexit(&GC_MacFreeTemporaryMemory);
|
||||
firstTime = false;
|
||||
}
|
||||
# endif
|
||||
|
||||
return tempPtr;
|
||||
}
|
||||
|
||||
extern word GC_fo_entries;
|
||||
|
||||
static void perform_final_collection()
|
||||
{
|
||||
unsigned i;
|
||||
word last_fo_entries = 0;
|
||||
|
||||
/* adjust the stack bottom, because CFM calls us from another stack
|
||||
location. */
|
||||
GC_stackbottom = (ptr_t)&i;
|
||||
|
||||
/* try to collect and finalize everything in sight */
|
||||
for (i = 0; i < 2 || GC_fo_entries < last_fo_entries; i++) {
|
||||
last_fo_entries = GC_fo_entries;
|
||||
GC_gcollect();
|
||||
}
|
||||
}
|
||||
|
||||
|
||||
void GC_MacFreeTemporaryMemory()
|
||||
{
|
||||
# if defined(SHARED_LIBRARY_BUILD)
|
||||
/* if possible, collect all memory, and invoke all finalizers. */
|
||||
perform_final_collection();
|
||||
# endif
|
||||
|
||||
if (theTemporaryMemory != NULL) {
|
||||
long totalMemoryUsed = 0;
|
||||
TemporaryMemoryHandle tempMemBlock = theTemporaryMemory;
|
||||
while (tempMemBlock != NULL) {
|
||||
TemporaryMemoryHandle nextBlock = (**tempMemBlock).nextBlock;
|
||||
totalMemoryUsed += GetHandleSize((Handle)tempMemBlock);
|
||||
DisposeHandle((Handle)tempMemBlock);
|
||||
tempMemBlock = nextBlock;
|
||||
}
|
||||
theTemporaryMemory = NULL;
|
||||
|
||||
# if !defined(SILENT) && !defined(SHARED_LIBRARY_BUILD)
|
||||
fprintf(stdout, "[total memory used: %ld bytes.]\n",
|
||||
totalMemoryUsed);
|
||||
fprintf(stdout, "[total collections: %ld.]\n", GC_gc_no);
|
||||
# endif
|
||||
}
|
||||
}
|
||||
|
||||
#if __option(far_data)
|
||||
|
||||
void* GC_MacGetDataEnd()
|
||||
{
|
||||
CodeZeroHandle code0 = (CodeZeroHandle)GetResource('CODE', 0);
|
||||
if (code0) {
|
||||
long aboveA5Size = (**code0).aboveA5;
|
||||
ReleaseResource((Handle)code0);
|
||||
return (LMGetCurrentA5() + aboveA5Size);
|
||||
}
|
||||
fprintf(stderr, "Couldn't load the jump table.");
|
||||
exit(-1);
|
||||
return 0;
|
||||
}
|
||||
|
||||
#endif /* __option(far_data) */
|
||||
@@ -1,886 +0,0 @@
|
||||
(This file must be converted with BinHex 4.0)
|
||||
|
||||
:$deKBe"bEfTPBh4c,R0TG!"6594%8dP8)3#3"&)e!!!"4UiT8dP8)3!(!!"50A*
|
||||
|
||||
-BA8#ZJ#3!aB"#3d0#'GM,MBi5bkjBf038%-ZZ3#3%)Zi!*!8"@`!N!6rN!4069"
|
||||
|
||||
568e$3`%!UbqAD+X`19S!!!Ba!!!,*J!!!F%!!!-PfTmj1`#3"PET)d31)LTH6H4
|
||||
|
||||
#*AqG5b5HI*)QjY$IIb00%ReTJSi6rG$jG(bZ,"Rc,9Umf[IRj)6FZ-j`GfGR)#!
|
||||
|
||||
m-#qLqB#cj'G%46qffB3q8AppLXKc+P&*il4FMJMq3N32r[U,(PlSNdrQm-J(4!p
|
||||
|
||||
jK)NHmKJSHY!,&chS$4)pk%8mL3I)B0'$AU6S3'q)k%%[5[5J&ffa#68)0ZM&#T!
|
||||
|
||||
!*fHC-2dFZ3i83[Vr[4Xh'+DNQrm'J)rrpqe%ST`,FeVi6b,*qHH")4eQc28NFMN
|
||||
|
||||
ZT*m,L"Y%-`pdAk6RLHDaeVV0a,,@P(4UUK66rUM'8bf91llS("lTh81)MBQ+4*q
|
||||
|
||||
rfHENEhD)Ke#3!09'M%bL[P1+G88fa$3e)5Gpf0kARpBf*6eIH*0`ZBHR%ii"PbN
|
||||
|
||||
+D&*)688M)Sm$Bm[cCdDjh2YIjmAc`(TVpi*Vka((A*&Yl@'LTSH1M*AMP#,2[A$
|
||||
|
||||
(FHA@S"dL4dER#3b!EfBYem(C9P5iGH"a-bb-AL(F"bb-AL,F6)%a9pJUL,(hf%B
|
||||
|
||||
TeQb["X5ib4DQXV!-fa6&mZf&3,(C&UDd-((SpeMBEIB`8Zc,BcZR3A5'X+jYj$'
|
||||
|
||||
6)6HVV+R[!`#3!`X!(E@*MFQ%R4d"))`m[3JM[c)bBS54Tj'M(AP+MK&f%VD5SdG
|
||||
|
||||
SANFB@3Rqc$Am83(+)`"G(D%A'9!bBQ6!b)b4Sq3SH8D1NDGNX$)bBi54!51--$*
|
||||
|
||||
Kj0L!M"KKK"dC--,)-h+-6#KKC-$)-F)NamL!!Z06#X!!b&%bBUHp8RcN'%%6!b,
|
||||
|
||||
i!!kV"`"DLHFaK*!!"Ym4K,,2i2X4c[,`c5!GIPf!ZcNi'8'VfJFpSfdpq+CY$8j
|
||||
|
||||
-V'f-DZr2[36#1(ael5hmfT@1cSU66D5pqDSA89pdTP-`Z[jj6T&!PmZBFZjal"&
|
||||
|
||||
5iG6#blE$+&kLh#QZ118&(0T1J(hZ,9)5MJ9ic*qPI!ac'RJ96QMZjSbkMq()Ui6
|
||||
|
||||
B+f,,#'N1icbM4N"aaBr1`3Z9U'8RY'XAiVXFKp#&k2D5Be%VCdh4%,+2QS'b"Q2
|
||||
|
||||
%0PNT4rE#%kTUFqYDM56bVjfe!p8MqmL)1VmjVkJY`U[*$&*L3AMSpB@LCQ*U&l%
|
||||
|
||||
T+3890rL,V9klFN*4@f0UTf8Z&&afN!"4GC6G8p3fN9$4+4[-@DAeK%lej"@eAAL
|
||||
|
||||
eU@&4[Tm28%mqqUkS(F+VDa#lB&'rlRAllRP&l460Qc,)MHR$jMh@$8Y4Xc'e`cd
|
||||
|
||||
ZE2AUUiH+fK96feb$epq&'RAQeLG&lCDjmP+"Kr8k9#qp'eI8RPf[6R$dS+$UcqI
|
||||
|
||||
ELYSV[*ETFL&j[@lr803qd9I2A#bi4Vei3*d[+@Urk*!!&abe0HTVm%44"i4A6JN
|
||||
|
||||
c(2I!kjRl6a9e813DK"A6p(LjRZZGaGH+1L5SiBT[(6ekd2*ILMSXU(l)#m3QMDB
|
||||
|
||||
V+QTG!r*NG#RQai#DNh4,l0&!Ie`dYi98Y1%1A$5hKP4,`d9cHdKP'LkD@q4hYC*
|
||||
|
||||
%dfdLeCCNN@i9UIBNLh5l5(8N68qhM&4R`d9cfdKP'bkD@dHU+qe&XRfNZSqc10j
|
||||
|
||||
#8Me*&ZNfNZT0hSYd+dP&ri-FGM6G6P,p5D,rPNT0`dQLk5+6'NLb5"HDe'$L)Pe
|
||||
|
||||
X8N2bj-Z'$r$6-$NZjLGC)1lB-"jQSff@[ak%LJ[rI#%p2ddAGREN(@"V+,S6CI!
|
||||
|
||||
I!!!0$3KRBbj38%-ZZ@0M8&"$,VN!N"#$BJ#3%4B!!!d'!*!%rj!%68e38Ne33d-
|
||||
|
||||
"!+X[PfqV-$P*!!!'-3!!&UB!!!(&!!!&C80(jji!N!BMM#0%$L)UANhN3L9rV@9
|
||||
|
||||
B`f#c2p$XpAVVCc-[`k20Y5bJ+CTHPScj`Z'!lmr*#EPaRH(ZcR!J!!rqMKG"q)#
|
||||
|
||||
cj'G%46qffB3q8Aqp4R6FA83PM6`KUjaYD&IlZ@jDrY"pk[b&AZrdH*kFbb9PM*S
|
||||
|
||||
`4Kh$c8Lf0bVe+Y`Q$amM6mc%*C1(jF&1bFSdGIlLpc*04b#X&D8[&6R%+-#6HhJ
|
||||
|
||||
kX"#A+Bp6%6RGkB&kM%'jh$ZLmam[1Irq,r82rGM"5H4bh1ZB+b"Z%&-pD)5CL9(
|
||||
|
||||
AP(4UUK6$!(lkH+UPFXFARF-MIHHMXf!5Nd%SZYRQj'pfL)G3N!$94X#(q25G8U`
|
||||
|
||||
VXL'QU3Njk8[phV2@0Q92J#d6rA2N1["[!%c(M4X-8p,0IcYJf2lRBmD2c)*RQEF
|
||||
|
||||
68m'9jqq*MjHTji&GqDp$kh501r9fqVPJe4iQDRS)L!)ELqiX08i#@40jpP1+F@p
|
||||
|
||||
iC&))L)Qq4Bk-cK-i*h`cDlN1cMBUbZA3+beKhX*-&UD`X%ME%F91fHB3BaCC''Y
|
||||
|
||||
KNba-C@(,"-40Yl"l,#c8`YCDf%#"XGD%F4m3'*i'k"iah[Ddam+k"Xd3eV@02'B
|
||||
|
||||
bj'D90I9p!!!-q)[jAU2HhQ[NiCQC&f(Ne`JR!hlN1''4Sjc`)hcL5IK+f(@8(q&
|
||||
|
||||
(1&Nj2XreTBI[M!0dGB4'MK01#CFF2c,JK"*1MNZ1(q&(11@5ii5EKimF*ja``Np
|
||||
|
||||
#bA(#bBL6BpQ6jq5imT-m2mQ!dq2N'H&2RT2M%Nii'6$J,PF!#N#jGS3IS9Uba%G
|
||||
|
||||
'A-)*8[#%!j-9'#r3@EpUPQ9+NL6$ldj*kVS6INIK@`*q'q$hGRJCPb,`pUJm(fQ
|
||||
|
||||
3!#mGrdQqe$Nm22hkJ2cerNp"i3$m4Z62S5YA40V([V`MbHF@)QPT2IN@3@$ceHm
|
||||
|
||||
I&dT3GqF9K,'&&8[6LKMTbQ6@-*%bJE#4RM,b'FA*'VC5`0BBdTa"@aNXM#)mU'"
|
||||
|
||||
N@d@XSIKMMiMh#RbbLSjLT49GG9"F84)Q8QfN&![N1hK"A'V5F,,dJIF@+`iNJEb
|
||||
|
||||
H-(5Nar84j!"*Q54MH+j&08dYQc,(ipT9I+aFqIQc-XP313&803UUPPD4*+UAIlj
|
||||
|
||||
$U+jMAP1QUSfEYV2Qp4HKfZ#TYQTCT)hEaCbp+ZXH0"m5USfHDV1HbL4cCT@41rr
|
||||
|
||||
5+d+eL4&+'hR90)iLRp$LYcm)e5McQN@UMR#&$kKqr%eHU-DBejbUCC-k+P4N5r%
|
||||
|
||||
Iha+Uc5aj)kVfm*'ej*8Dali5ULfHDLah-l$Zfer1#G9@6l8TTf*r,RKTZ2#Q8'h
|
||||
|
||||
MA2&i%MYq(0aCicHKfPlfDYLeJ3*FFEG3l@"HmfJbqFrdHU&IU+jRHE95BmQFkJF
|
||||
|
||||
29)qp)93hX!aCGLfYP0!jSEU4HF9)-e8M9rADGfC4U(BbVVC66+8XR2Hj2RAmGk'
|
||||
|
||||
kLDNk8`@p0[6F"hrG,e3h`kmm(BhDMQjBm@`ejDH1pG)YbUXYM'Y'5aD`-H(VPZ)
|
||||
|
||||
,*i6A,Nqe)D1Y'5@UV@HM3VAE)a3$3MT+9jAGa)HI#%*E@9ie+jmf-PA9dY#66`Z
|
||||
|
||||
[fkMA!l&$eZ3)bP996crcal6`ZRdT$9NG0S#+V([`rRZ&eae,A%dMGB2V4H%9YPL
|
||||
|
||||
LfZ3B194,NC[ik!QKZSYlaE"deVc1$3[9(XVeFJIG0T,9**@'AVXJZ2Db$%'!,$a
|
||||
|
||||
e+d2+8SES`Z&RD1(C`m,VlM*Aj)cP#M@ZlJI#Djp(U28`fl)VL9dKY+IXeFM!HRJ
|
||||
|
||||
MVc0#YCpj6@!,M0VrHYh,CMQN!FBjl1ZVEPhjaCK)``"6,6JiU@@ekMjdmEEPI@M
|
||||
|
||||
3DpXKj3pi+f`LFFpIUPrF058)N4X)f4ZQ*P5c1[&!pGhC4i@Ue2BCE"bRL&haLRk
|
||||
|
||||
Thb#ZUK&ZK-Kc9k4Z-[QKhdaf&1KhN!#*#IdZ-XfJhdPQ)I6l#![SYjD'HXp$hdA
|
||||
|
||||
f$1LhNlN-r4DbV8$I8iS[RSEqj#URqY@$9b3dJG1XG))%khUHJMX,Vh896Z%"I%B
|
||||
|
||||
PFK1MejpP2[@,$LpbTe[Q%h#[hhai0BBHF+r-MrTeL9G6k!!IKHa1rmf2qMf,9c6
|
||||
|
||||
d)%I[5Hq$1hVVq60(`H@-9fb&cfkb$BBDc1-Ck@@#jrVH%0cXH$@cIK[C#F&2Q9X
|
||||
|
||||
[qpl(HTpEQ9F`KqVA3&iYS3Pl6#ARpIXMVpCP6[+ma`PkbJPkbJPkbJPkbJPkbJP
|
||||
|
||||
kbJPkbJPkbJPk1MHKTlbJTlbJpqGlF2RNe4CD`1XDTfUZEYjDHE@[F0T$,KbK"Vc
|
||||
|
||||
mA!9AAPiGS3Qjm[HQi+l-LraVj'p1i3&mcNKce1@eZ4pFX(PY@1(66rD18)Im"eF
|
||||
|
||||
YAJ1K#AYcK92peXpVBfM#AZAIKi*r&r$U$"h)dkhp2[JI!kp0S3GjhdZZV))A!43
|
||||
|
||||
jH4kk(TLQKF4pTXhHI!ITRb%hcX3KfeN#**1EI54a"'@Z8(9Dm%D@b"Y#qhm!N!-
|
||||
|
||||
0!!PRBfaTBLda,VPM8&"$,VN!N"#ah3#3%!9X!!!I``#3"2q3"&"56dT,38K-!3#
|
||||
|
||||
TY1))Uc!eD!!!@F-!N!B563#3"2$I!*!)22J1`2KbNQaPEr+hGEX``Jk!Vpa0&eT
|
||||
|
||||
RDl*eSGZ&%EEAc@iGG+hAYBDRapHZd6ETQH'lV2AbpMVJ4lN,ck0G4lMb)fcKAQi
|
||||
|
||||
*AeLhm1)VRfPGM,"Zi8pBG1%a3VYZi@m,@rM#2'iAfhjHacE,K"[bJGYB,ZcNP&#
|
||||
|
||||
"$cqJ[fRG`SmXR'aMC-H6r-)AXTaNHE+Fj"HkN!"0"R[G!H4jITB&`!(!dKX"PZ#
|
||||
|
||||
Z+PX+S(dCS&YGZI3,cN3L+P4H)V5R@D3p,54$JD"3'!j')mhRcl%mUJ)9e2PVUaF
|
||||
|
||||
j[6lNX)ll!4,jajb6UrZK!hSTX[caD`$ZIHl,pdeVm&EaLeKG-YjQB6AKT)84pF,
|
||||
|
||||
kB$+55%ID`b-4QF0T19ckfSl,d['15$X-4cTr0"2!dIR5%1j[S4JQa0,J4lT!pkc
|
||||
|
||||
"EjcQ2ZmmNDF36,1DH)X!8($N3ihbR+mcX1GC!E!0fi)+ra)rCUL`#HU&V9)ke`6
|
||||
|
||||
IhTB!b&RK%B!&4fA8Ecr8+8IBcr)4Z8L+$bmVaA0$-Lr)$3+SMf0Xkh!%1L(hiM$
|
||||
|
||||
H56i!P'Q(V3ZXrmCRE,f[6f'0N!"Z$E6%fl(AqCL20Ka-#kRdjh`qA&CRACe[!5i
|
||||
|
||||
+PSiKjh)6PJM4H$#5%&U%HF#GqF0F$MM6fH)T68dFSQ!hQ*["e3hGME'TS#e`Fmq
|
||||
|
||||
Sl`'0qRTZMfEcM@b8M`(hV,a,kqB4N8iZ[4Sh5b!9ddQpT9YP#5UK!NX`BDbr,"E
|
||||
|
||||
!TME)X#08Bm,*$)fP2Ci@G1bTGUbETe@@q%4QL60h[2d5)BQGX-U5,*6)q)99'NX
|
||||
|
||||
bP3a1pJZTH#BC&"!P%4'5XP`!Fm82LidDE@#h&eejC#m'cSQd"k1C&S(CD`*"Va"
|
||||
|
||||
S%C+TmmkE6aJ*6S3kTd8)4GS&PNjQ"#DY1419T&!JQT+cV-0*5@'9$$5+K-58Y"%
|
||||
|
||||
N8Ea'&)q3!*!!UeBZ'qd'!&14D",LQVJ'$qTI1DUU3$%0cAD!e9HMkl`KaGAASBj
|
||||
|
||||
TJ#pMhSb5Rq0c+LJ3l3LJkD2dcrJM2Q%3Kh&mZL-JR(&m+L$L-)j29b,%B4br8)j
|
||||
|
||||
X!Y$j4ZUh`)[eI!A!R(d!4AHG`LH[d[f@re6*b2mAI`)H5F0aI+2XYq2iC)+N`6M
|
||||
|
||||
qC$b5"Z2ij,N%KHI*24K!$k@Plm*Hm'Rd8-bci0h@*rK6m%JDM[-[aZ1Nhq+IKNH
|
||||
|
||||
UJA&mE-V&'KM(2a129!2Mq2,5(2qIrSHmNfTSR2rTH+3D'XHRfL81irM8FE,Ep4r
|
||||
|
||||
eTUeM[5Ra8bilkJJ6f!)lF0e(0'p*Cke+2Nq9ccEjh#UIZq6c&[RmM(3ZV*!!cL0
|
||||
|
||||
k&5l"Jp4$Ilc)-m$9BDMqeV0m$l6LhM(EAX9A,10lG,aR)2GNb6Sm29&b0@CfmMd
|
||||
|
||||
&Mr!pHLh'hX&p"qiPVV#h)jIcaN(YAHVY!-im,lH&lp&Fc$pX!KD$+,qKqbMQh",
|
||||
|
||||
@BjDAX[M-KFF0&bH!le%r'GC@E`LVXP9mKXdeG)3QcED[U18Vq4jY2c-fD8XFl$a
|
||||
|
||||
Jb0pEdXPRCYXVR!e1c(f%qF`GKAUQcPT3T6E-YjCF2GYHhq#[aqa0'*p@XJl4r*8
|
||||
|
||||
qM(Fa(e1(MAb2DUZDVTq-SD2mJ+kFAj*ldAQmX-KFQf"C5i,E1fA&P2jHj`!8*c4
|
||||
|
||||
Cbq,eU+LUqmriLrQ-H$8"RJ(GXC,YKXYCKk(M!EcN!3MV-HG3b@DB@MEAd"P5,9[
|
||||
|
||||
2CjDYplkH1ckr$1D5aNf'jH[,p0ehXaPCKe@(eI0#11SC',UQT)X9K3qD(G8hK#c
|
||||
|
||||
C@GQUfADhU*AQPE#2X"A&i-9KaAUdDe$"bpQU)@mfJNfL,U61YQ4RBFiKFac+[hC
|
||||
|
||||
Y@49Fi(Ye4UjKII9Fl[b`UM[(Ca+6ZhF[@mq`0Seer)R3*#Y$$IcK`pPc%EI6FKZ
|
||||
|
||||
I`IV"'%bLZK'Mdl!5jqQ+3J!feU'k*f(FZf(EGY@@N!!CGAmMqd9@CrDD68d'jf(
|
||||
|
||||
3TlQV6AYhAEJlGh4$epjV3bSqBiDXKA!BPjeTVUYp1pI,DPfESAK1"2eSD[B-elh
|
||||
|
||||
H#"KCEIFl0K-Um0E-CFr[,$HC6Hhc`fDr-eb-HmN5*`iSE-8)!#TL+mfKpUV"jrc
|
||||
|
||||
$X6fMXIlRYZ5'5$I94YXX-&C(`""L$Dkf)VmVe*%)GZr'mh(#3i3EqlYKNKblRf*
|
||||
|
||||
'9fi`h"aV43`ejERI0DPfA"MDB``XX)HHa#bYS3h1c!hCcPlQ0+mDh0Yr`mEU8Hk
|
||||
|
||||
YrAmUXCIMj8SFBkA%6iNVCjRI%C(IMj&E3@l3G[C&a#hGId-rBQbXrT)c0e6q'2p
|
||||
|
||||
eC)89`[fJmPd62,qrh"5fBCA-$%rb1d1R5hbj`ddQ1G,60%Q1l'T#EqB1)110@)h
|
||||
|
||||
%i!95M+ekEiM0HfqSHM1k9UQY&%V$jTQPB&VZFVm*4FmG"[Acbff$#qbZ,a3IKUr
|
||||
|
||||
B"VZ2A1J-[B%elK$paa&k8Z63JaakNVNdL$c1fP%+A`QGIJ'bm6iH0ZklkX(0S"E
|
||||
|
||||
8jP*3Mb,[3pbE@&fLD'2RS@ZY1`pG"kj1X1j#2R9*X*QX*TAMbYcVef*YX2)T6FA
|
||||
|
||||
Q@D$Hf'AE5@VBGSP+2*elSqN#9T4Gc"`I)"SMr!P3K8hPL)Se--@E+!*#j8qBAdA
|
||||
|
||||
F)f`H'*JMT!TSH@V*`'V2IZI1K@DpeEljYRXA2YJ9eU,IcfjLaVQJjXS%LTUELM'
|
||||
|
||||
UNU1Q*M@HTVX(FV[-AA`QqadqFr3i9[JU81PlSB$r%d$A3iqhZfXV+KG!GjBeeU(
|
||||
|
||||
[-cfI+9deX0(XqqDqeeCrEqGcqm6iUPf$i$#AQd`B@p0rSjJ6NR2d'hX'fX5-"MQ
|
||||
|
||||
MU,pRS%(-F-NCDZeUk[$*BA*h$2XG9RaZHj-D6bq3!1YJC6AD61@QEFZ@lXi09,[
|
||||
|
||||
#3r`40LMRE"V0'C!!FecYKJh1Q(D[`hN%90BLbX@@Y!c8C8j3QmY!ApD)[GhVGTJ
|
||||
|
||||
**CcApF6MTA!ZjkemqUrh9AKG,PI[cVeVI+q#h6`$QIm$kKcXmZ"@c&ph+[pbaRf
|
||||
|
||||
+-2[6I1-)JqV1YQR9UpZ-&Cd9Uc'6i5P6JCdV6"8c-TKV%$1eQ*@af2(L22GJCe"
|
||||
|
||||
VaTDFcfaEffcXh1Pef-$Pm$Vic)0VQmqbL$(+mRVQJpGcr8kVcZZakIJ-9F5"VJ2
|
||||
|
||||
A)XVacTfpDfd&ZhSY"9l2XleH6rpD3Epa6E1D10FlQJjH!G34SPGS&qM3*fC3Pe2
|
||||
|
||||
L`2L%lVY,CV!*T39qcpXH[fHHVQRU'%UAhk2&Qk`VKaD[,i2ZHk`cX2[6K&iQRrQ
|
||||
|
||||
lbPXmS@QX)1Y!&RH`da"Y"8BfPYDc4GPC#3lV4AhlG+E(2&HTGaMM!VD)&65CaPL
|
||||
|
||||
Dr4lQB&J09`k9kE(,mhf[0f[T[[2#[mfpH2-6*6k4bk,U5Z`kcd%Ia$UcfEZ2Z!G
|
||||
|
||||
1&'%PEF2B1aKl$'0hBH`R',X1BjX`pP1-h6AD-aHa8TJD0Z"T@[KdIJ$5L*0!R+1
|
||||
|
||||
)NmCi#mDEj(J5i`fS4KaV[49[Y[ASjjGJCfSIkdaR)f+)e-#cLpMMH4iTJQFE+B$
|
||||
|
||||
RFiN4RXfXNFpBZGXAc[3QM,G2Yh*CMh@3!(q8lFE6#ID-P'YZ"AefKT9M99N2Re%
|
||||
|
||||
Z5UJ[cKd0UjR$Y@%N5eQr[bVdDANH1X3[2[#XjcJ0%Se1!jKa'U#f[M%BE`p&`TC
|
||||
|
||||
@-mfEF*1J""c`J'Sc4b0!`0Q1cH9X!e(3aCl!)H`k4qIhpfYS1)*',+EMMLJR'JM
|
||||
|
||||
*XAVRp4,L3*6EFHJLENI+bThcfZ@BBX$BV8U1Sr-@+@iljX&F'M+D6*J-'5#(%1k
|
||||
|
||||
[1&EhlT'("@L3!%(&RA-a6V0,2#9X9%3D8*&8fT'k`V(k5V),NCZX$kh*MY@GDYV
|
||||
|
||||
4Y-8%c[bAlh!l-U6&69c*e@N4Mj-C)C2d+XbiMLZjUSJ3--Aq8HQ-$[R0RcMaPa8
|
||||
|
||||
e&lLqlpUj[TGS[iMVqri'VZr9AUl[KhZi[J-YA0r"GUl[d&eFhq'YA0rr0h*pEml
|
||||
|
||||
RqYlHa2Ap"212)[Ba!pGh2-6e$Gc+p3dqbr80[FMe`hbZAjA&I4IA2aN0'##DQ-I
|
||||
|
||||
F0B%8$M1bX*!!6V&dUi!$KD&N2-DNDAZFBic&F2BrKF2r6-!j%"D+4)8c'q,aD,f
|
||||
|
||||
3!-3j51B9SJP@RdlLA(j+(8X++A@L25E3BD9ki@,HV9l@i1F0$6KDbP$RC(bL'2*
|
||||
|
||||
%ikP8)(QCZL15MXe30%"dDAVbI)DMURqBCV&i5b4dfDrbrk!LN!!@@#SGL#9B+*j
|
||||
|
||||
N3JH#Y3HLV#@5r"fhhq@IS5Jp9LM&BLQF6+PSMTk2cbS%9c)KQ@5a90K#Sf4N5PN
|
||||
|
||||
S5M[3da4hiQK)k+XiA(ND$YpSYSe-m)LIZ,6N5rL%!p$M"e)Z2G@JJJ8FXU,((EM
|
||||
|
||||
pQ)@$C4*&(*ZN6`SqKSGP)q02Q+F@[iqA@RaFJFBHbCM4qfMF%h!%89`D('LN6e`
|
||||
|
||||
k'KDkIh4i5)XM8r4*4)JcM9hKZ+)%Kcj2Rl4%aj+pAcSALTmN,qQmF&6[3Z`$k*0
|
||||
|
||||
%H%M18RJEF-b22R&0qM&+6,@P[&-a!BIik*1U!BGKe64B611lY)`iBNHI9"S+Ab9
|
||||
|
||||
l)JjKd5HT3V25,H+!P%`9Z`rkT%9kNCS1THY!pHQ6Q&%@$8)T99L%Sfhd5H*hI$J
|
||||
|
||||
64C28Y,C`Djl#m$6b!XGfTmrR*X8$d@L`Y6QkdK+%4i(E8[b59GP&,"cqQPC3ih4
|
||||
|
||||
MlA''N6k&X1iVfl4IfC%6%hNG3kaD8[4Nmd+LGcpXR+[Xb-XNFZZYEkLS`Q4G+Yd
|
||||
|
||||
5L413!'S-T`$1NR'U9P55`+R)+U%aM8!K9-"b-+[Xk$GR5FTkh)hN*rJB5@-L'EP
|
||||
|
||||
%j(6IK+GdbSlH-e9"XT!!TkM$335*3-%BFqd`miD+#P4)M`VKJ,5STAS-5DFJ,A9
|
||||
|
||||
lRF6mdQ"V)#Q+K-c,[YUNl&M9XNEZ@PkXmY(k8'eCj+P3G[5T%69*)e+cY5@CqV"
|
||||
|
||||
#$%SP0969B)9`fR3N*L#-jAfF#50kqURL8%pU-)M3+FmipZBILqkTH!E9YJip)aj
|
||||
|
||||
%`mKhi"GMeDhkeqSZq1IU*VIi[,SeRcM3"dM$M['C$j!!BhcZ!m11mCN2&2k,$aK
|
||||
|
||||
qi32[Hr5%Rh[d,hX-I&T(k6&F2UIBBc4(!m'9d93k(d+2NBr*-djj`D*SpBJAZ,f
|
||||
|
||||
9j!86F'3iZ$+9LDAqShqJf[jh,cLPbr2V[SPKZ8BUA*j'UT'@jR"M,2UIAFerUC*
|
||||
|
||||
hbU&Hqqk24KaUB492qKV`$C4!&+Z"V#$rQ"GJ24rmKPrCa6X4KAZ0c$d@5+lmTal
|
||||
|
||||
hVejS(qNI[*91V#iSP&p#b,2@2paR1A6E52mJe6FBBMJ1dGJL*2+9p3qIhj!![Bp
|
||||
|
||||
M('C8fB"h)XK)5,I&%TpfThIZ`BHa&(9Vm2+9kL#QA,kQIZdYiIaLYrARRVV2f2q
|
||||
|
||||
YNG[k'UGr%8DeBN-EK0EmEAlarTd(p5,rIHIa&j&hIpETLXk#R@jbC@-b,9jkj$[
|
||||
|
||||
SG20dc3jaep#MG,*Rm*9,kClGd#jFfLM2Qq@TmibVrRcNcU2@95h1CX5Efl"&%5r
|
||||
|
||||
8mURGV@U5ZdHGS,k4EYRemG4[EPCrFjZ4PqYQYFV$Li`LB4cI%5Ak4CIabTc4cV5
|
||||
|
||||
Z`5pfTSPdXM(B'Xb,d*RQlCVl-6rbfNK(iUpddhemB9))4J14@"k%hM42efh'efl
|
||||
|
||||
%*i192U1qBE',qSa81Y2F(%qfjbIV-mbRlM2Dk!QiiGN-X@CeBXhQjHJG2R%#l)P
|
||||
|
||||
%*m$r!"'46R)DGS+2k[XNTp(qiGGq@r81$FI)IYZ`[)lZM!cTba)YbQKh2VHq(T'
|
||||
|
||||
iYATPahXMf583L9i#-b!5'SA3JP$LMk5FV"eL5P&e,)!2AM(fqq[&rAqqJEX3ZJ0
|
||||
|
||||
4GUAcq1#I[$MlrpXrj3jb$ZiY+2BkkdRM@qKR3r"mcb,mia%m2lM89dZ[Vqh!-,f
|
||||
|
||||
QqNbpVjjZ29qJCq04M`2d!b+N'UT5MqGLqX832%q[Aej$mA2Gr%)2D,J,T!VQVUK
|
||||
|
||||
`%6jhAB9V+HAI4,rjJHFl+Pb,m4eQEZZ5@KrPp5aF@N9GqC2+ql1S&YkPdTmG6Gr
|
||||
|
||||
!qEV`09U+&4c&223NLQNk-DpALZNdR1mDqVXNM'QAB`crlBKL%mp(M*G"*FCZ`&J
|
||||
|
||||
DZ&cZG*Ki-f,J@mmLMhX`*R29E-FB[Qe,XDNr4DlPFZc[1GrDKlkqQYkKeBBaYUl
|
||||
|
||||
YEqK(@E3aM+N[HKM14ThU%2X*Hb(-`McNHXhpB"3j2BDaPJB6I!Ne%&qEaD`r`V`
|
||||
|
||||
YU-G"k"3ar)MaKKaEKl'$NQC6hd1-Lq4B$Q0G-XB+e-BRajCJ,+'*V3bd4NrqAp,
|
||||
|
||||
B[bJT[kddmXG*R(e#AIa5)9RRT[cr!`!!$3!*Cf0XD@)Y-LkjBe"33bkj!*!3qL)
|
||||
|
||||
!N"!0"J!!,h3!N!6rN!438Np+5d&)6!%!UE6L#+X`0A!!!#*k!*!'$d%!N!43[J#
|
||||
|
||||
3#1j"$F$iCXbcEQ9ffFS2dS@*jbZl63NYVcACZY$0##1XPDZ$V[@ke[$dmVQ6K5h
|
||||
|
||||
FYGEmE+(Rmc@246PGf0D9hF)@VNAi`VhS`KGM(GQA+lmmdfiI)f`c`Tq`63P23V[
|
||||
|
||||
Y`VEH`KHqX)9f(@(E*!Zrf-)@IZi)AhKXi3[E,M3j*432"&!HrHaD@&$M#f(,qq3
|
||||
|
||||
@XL1hN!$"3Rk6AcKCb%+1%di@J&@""TeG+a&(42abSQ*m9@@VL(4[%29TUPEGj%S
|
||||
|
||||
NfN09'd1a&"q0T8,*F(-`0#85E)pZZ-eZrEB+Z[80G6A,A6ir2'5jYd$i*mlPdrI
|
||||
|
||||
-@8-1XA6I6r6dUG[h&cAjUSAPI(dbhQEPDb0*+mqX6fN-*U1*9$3@'8GN$c0%(%0
|
||||
|
||||
GelfTH&Fd4Q0)jLrR%MNc2aM&pcf8d``Y,Ak!B(cHb*GQH1E2Phb'JLQq0Yi5)P*
|
||||
|
||||
IZ&DMccNrDX`mDiN1BLbSE&MC!)B+3p!!(FM4Z3"pmf##5,64Fd39&fA9Eck6N4(
|
||||
|
||||
q-Kr+TK`qGQ`-&dGPAb51%'Q'J"dB3bK$iZYMHPIm%$'QJ`j8f2l6cq5j@TmTYD&
|
||||
|
||||
8Dh0,2)CCjkGqG*&J+Y5CqU@IDmIQUUrh9q!`X*4GG$59b(1#DBYLrXT3Hc`B6B4
|
||||
|
||||
D3NZ)Zr'(SNLFq4ETPX+0#01J@-c9Mci&E"ETe"lZK'B2D682F5pVpcl#6cM0`cF
|
||||
|
||||
VIh2RdI%LA6N'$6l@jXi1I@kfp+LX3395@i-*Bq1p(FdBDS-m*N)0#&FB@QXXRJV
|
||||
|
||||
TqHr&d$F[UDca!YiDjchaf-C3%T1`bTUFNM26%1V@@T1GbH#dKP"R2*d-KU#5L)D
|
||||
|
||||
5FVQ)&NXr0"XEY)Prh,6j`NN!Fk+aB(Zk*F3lDTZ$[P"c5bMC1Arq8UD4i#5T15f
|
||||
|
||||
KF$3@iP2*G)M2RB8&#LRFh0iTXfaMT'5S@aDD8))aK6DZ*"9[2BV(P+51c4hG,L+
|
||||
|
||||
c53S*k44Xa8Acmd49U9R$Xk-p6,4P'e,Rh4bZH3"e6"(G$Pjab5Ikh&MNk*3JKBH
|
||||
|
||||
am`[rd,p4KJ)IdrpGAkQ!SYrdArSB+K6p(4q-kaYR%DeiK@MHTTrT+airpFpf(!c
|
||||
|
||||
C6D6hMrH[fSGq[SpSi@NLdj2ApC8!q05rrM0pH5A%p,FGr*AqP!RpYPrTjl,kIr)
|
||||
|
||||
Mrc0p)kiXJcl9Cb(1%'6hP`BRQ0MP'EU4U`lF@CCrSLp0(%#3!"HAp98B52*lSGq
|
||||
|
||||
&ZrfkrM3CD5@kEp'%2R+m!*ldPFM#f(9p0R-`C#rdT5&)cLr`#Kk#rMULrlIXZ[j
|
||||
|
||||
d'6P$Y0N+!(Y!54rDdc&h'$"brDYqB3l4$[hhr$0$4PE$2eXNb2ieb2fErJLM)1T
|
||||
|
||||
RZCa*(rQIH68r2Xk[*I+#iKreEj!!r52r-kc1XRmYjSpI3ai@B(RaKIqI,BSqG$#
|
||||
|
||||
E'MkH69X[ckB'iJEe$Qi`RhhAFB-&cq&lKKZFKRc"-D9m50)#'Z6Fp%2+jFLffS0
|
||||
|
||||
N5Tj%4@C5"GI&cC(ZFcD,h$e838lFZmM*m-eX'F$dP%A,,mqff[SF8$&N-KPiM91
|
||||
|
||||
9NF2XSa0J@f1fH(J8"hGPCVYkTSRLJ,V55r6R486P'%J,"U5PdFrVi(p*UM20Z#1
|
||||
|
||||
AjGIGE[0r"EdLeqdcjp[mNSplX,Y)hCYJ5aj0I@@G*jb-Gm65lHf-'iiR1d+aG!I
|
||||
|
||||
M4Q-YACfKpTEfZ,40CpQLY-XkZ5B+lNFp6BS(cVppFXHLm)JE3biI%jRZ4TD29iR
|
||||
|
||||
SY!R1P$QEBbjeBD*lqi'1GccMbIje'bEC1H@a56dI1a@*I@9pEqBF-qYcdaaAM`b
|
||||
|
||||
5FjP9B(QLVT*e4Aa$'kXN*T*FX[j[jrbLXcJ8Me@X&Eh%AL-JTT!!Gd4B3#S&rjI
|
||||
|
||||
6(0UBDSje*M'BT4+G-9BhC9*@-5jcH$[1@!XpJKl'$ZGDCHXmRb03ICB4reapCC!
|
||||
|
||||
!(Mqj("6&rGSNfp+B@FQGKfZV'cfXb6ZLR8&V%2h"l5[mJ8hjJPR%eT0&kPUA"r-
|
||||
|
||||
MPcHq*D-)FI[,GTp4[[$$5jiqJ&BGP+G#UkjaI6!H#dFM9NbNa28pDebXI1(,,(N
|
||||
|
||||
ED'bUV!CChjPULFDCN!"U8NG00mXke@ZV@1Ge4VY$ke-3#PpeT"PAmJT`"+9)V,N
|
||||
|
||||
pTl6IHLkVI,'RZ6PAIkpR2HXM[+GCRdK'0dVZpqGr6kpmXC'CT5KCd3'NL33K%LA
|
||||
|
||||
eT(2pQ21Q5[3dR+GDX116UUkC9$)S5UXm2KGcINq`Y6NTP421bhiMS(ba5j&Vj+N
|
||||
|
||||
6f#aTQ1JNeElPhNVPLj`GVbDV%DYQDdZbmeS[j5Xpee4GLelLG+PS4`JbeUXka[&
|
||||
|
||||
k0V$H4$f6H2FMHFHjNP0bI"Sd(Fh4'2DERk5`R-%10TmaEFjrI`$I68b$mrG)kq6
|
||||
|
||||
aHBBP*&LlQC0%8Xl9HQQfr9b!L@&XcMHPT*eJ*QI3,1Ibj`$iNqZ&q@YbPJ1Ha&!
|
||||
|
||||
Tc3P+,rc(E-IjIaGE%9QEH@4l"'92bccba&FiN!#)&l6[jHikPAbI*GrYmVe9[[I
|
||||
|
||||
)phhbr86Z2U8bGeIk!)'b%TGV)mAiNDCMGeGHc9GI%IUT&GqZ"BjUSA+ed+mA[-2
|
||||
|
||||
LXC)(FAZaC"ZB'D&IrCc3Ep!"HarI&r!YF8GmAD,SLj2'YmVA4CaPLEK2k0IH*6a
|
||||
|
||||
V*Vk$fS9GI4I"H5aL!-[(@%*ka9$HA3N5qMA()VUDA4&9YPT)mi[cZX*6&cM@eJP
|
||||
|
||||
93VpZN!!h"R3P6RiqmI$[+mN)k3@15PH6#pcRH,qPD`T@&9NVUY3'[UeNf`)(%Um
|
||||
|
||||
4l0h!LdSHK&T$P4pi$qrR04'Md+mkS'(0E3aI&)EejF*+mAAAd"56T5l"Ckd*lZ6
|
||||
|
||||
dYG-("ec$9*M3CUehlN4&9Aer+0`PT+AR#H3GeRp3FMK[%pq9er8Y223JLKM!HEY
|
||||
|
||||
N,mdU@jbA#DY@la65UhIkhK'(PTE4BPEM30kDR@@'[UIiiUc6TNIh["CTp`k2hPr
|
||||
|
||||
5`jXLjbc1QSI$eZbmE28#KdHUPIB[)RkQV95-AKqV@,pZ+bUiLHmHp@@M''(eB8f
|
||||
|
||||
f*6X2R,FYF5Vrc4ePeE6)rfDaf,5cCM&h@d69*`VTa,5qikYhmZK0Ble`+6c9aU-
|
||||
|
||||
'$C(cf9ZKQl&q68LMIi$490Bh%PU%6PbL0f'aB1Hl9(X5aT1l$Kj@l3YE82GhXer
|
||||
|
||||
JkbdqLcQ3!1Fk6iB8YmemmZL+iq,&A6dRGi493YT#@5[6iERXA%YphBr&!El1[CF
|
||||
|
||||
+&dD44l1b0lLIpNA*b0Ie[@mhS`,[c9hpkT&bXm8F@aUa0,JLKIL@V(3KLJm!)8*
|
||||
|
||||
&l+8LDUmD1G8`KVdmJ3fHfLH1XVUTHZhcb&J6TE``hq4Z-c@i`ef*B0pah)HB(K3
|
||||
|
||||
H'HbMU6,f$BBChH*)C%0(+c3dM1IjL9Re`SV`bmEQ#NIi'&Lk[$Dk84behl,DCHN
|
||||
|
||||
H16RiF'r0K2I@`Gr,ZCIaFJ8(9XVm+EKbPreGN!$mr6@mUF84qbhVQ,I8i-1$d1L
|
||||
|
||||
YqD*,(#erAVJEVY!Kh&Y92c(6UfI+c4%lZQ4ZC'U$+c`cjjFl(c$,5(pJUS`F$5#
|
||||
|
||||
EZE0`h)YZC!jHBaAMZcmFjCGm1&U$M9+Ne&j+T4(,h&)bVh&lrSC-Tmk6jY8epT%
|
||||
|
||||
+KrZQ`[0dKhfNlm)+9rKGp,K6bKpRq*MNS4mHqT0LLL3I0lp35RH%Cbk#'pph)mE
|
||||
|
||||
6[h0S,fP#'NXTD5D86d2hbhap`Y5EHAZ(lFME$j!!1d1fSr"6Rb5lf@C@BB2jcJl
|
||||
|
||||
d"Pmq29"SQ8HDhKll%9B0qe'T%Lq*l`B@mDEXREcc)d9M9,K%USLj(+VSJHQqK)Q
|
||||
|
||||
BUR$*mLCd,r",+)phKPA01S'YCFRQb(lRkmXX"TYMlpHHARDS*k*$hLm)m'`$`C@
|
||||
|
||||
&''S*&!*9bDJjS-&YYQGB2'VT%G,Cl`MTLd2Sm'j5'3C),I`f)I@3!2%1,)HU+UJ
|
||||
|
||||
[bkq[4qlc"L&GfMhFDr(rrZQrf[,p)kG15hMhd4&b@XV0CQ"E"aq41''CBqMY(fk
|
||||
|
||||
6'%db`c6B2p`N-G`b3k2E`LC4PM$L%f0jKiiA$`FdZ,h'8JHGYGjZ,MFIA,hUZ$K
|
||||
|
||||
Fiik-#KIi%CQcHi)c,(2FXEaGVJlG5DIV!UPX*XE&5&T'QM)AD5aPC#KEMpRZ(3F
|
||||
|
||||
@d#@FcrhLGd[T9XjApG)IRkldZGhZJ5-RYrVI*)HP'-lr3A8KTMck#[J2AZG[`VV
|
||||
|
||||
Jha3@r)a[((G3NfNVUYR5CUc-9'i"NmFYABR*P@C*M$5iH4*6"eEDLVfl+"l+"(8
|
||||
|
||||
@M14#qZ$f$FE-%Cr66QkRcbQN$fhIF,09`KM,jee+2Zp$4fakRpHZ&p+X)mlfR0d
|
||||
|
||||
"PD(-NB(YG[A4!D[DjheP`1FGh"ibp'lGS''H'jf"FrF4Q`L4&ES+2A+LQ%dj*8l
|
||||
|
||||
JqAe2P46cqDAU"Zq2[3hH*IV!V%Q9RJD[$Y[IcD0hlLbM[MffBNarf[!E,'IqV1S
|
||||
|
||||
aElL)9fHGF2%%2`0UDi(dPMEbbl2c%Kck4I2iE0i!RV[80kDaL&r1U`2Q5CH@"Lr
|
||||
|
||||
[j0%0QdI,$*Mbr0mIb&Vl[VlL6mAA(hfaa#pj@9j6KDPc$R)3I@Chp&h`$&mbSC-
|
||||
|
||||
1!RXIf22!RJ6fYm!H!,BEf0m"Hh*LCMEaT63VNSGE8@5Q-%`Tk#5JFa%k+H!Y`!-
|
||||
|
||||
bRJ6HK'V%dHZYf,SBN!$R'c'C1LBRd`93$,0Ui1jQlR&I`LU#Zje9!2GEQ52F,Ia
|
||||
|
||||
k)@hM(PmfejF`2MlEaQ@pYK(Kfraah#la*h*F5bXCXX8fMUr1HS@dXLKKFl&i-D,
|
||||
|
||||
KRHjGikbVar'Y9la$l2RB6pmR,LdS'+0CVLaC,H`"dT@r%Z!F2cScr3P3LVMhU0$
|
||||
|
||||
RDQ6lXmIBIJ6h2FZaT-(pd#Tr(GX$[`!BEfIS4+1rNEepHBe0*1LCXfaR!QFkYKh
|
||||
|
||||
"[C!!E89`RpfiTTEKYhU%C9l5FSYb1eVZ[NShdqFHU(5[B[`[Xmd%lNp8ZZr%``V
|
||||
|
||||
Z`-Sk2q2e,eY9c6DeamCH2MPq""hf),AJ0Z`'mAk4BHU,`2"fN@(D$$6B3eKJHLe
|
||||
|
||||
ijh+BEJhfCmrNX"X@BR0iMP35pJI3b"!RLM2TKUm#`jj4mR%B@%X1Qrhh`&k8X3q
|
||||
|
||||
"I82'4(M5h,f&[F[64H#l[1e2f"XKA3FdhPMh,0f#,XX(PR*-SARJ23cXC6*+rTj
|
||||
|
||||
($GBeQHQ,U+Ad,JkXA`G[(hJpP*%d'S#PC1a"B'rNDPDX"RC'a[6!hT)eeX&I3XE
|
||||
|
||||
f-%rDMYpUEQfrmLafmJQYmYTfr+%XjmL[Mpm65YCl'2rr!!d!#'GMG'9cG#kjZ@0
|
||||
|
||||
38%-ZZ3#3%%0D!*!3(m-!!%+&!*!%rj!%8&*25NY"5%`"!+QdiJLV-$9B!!"5l3#
|
||||
|
||||
3"K+K!*!%$I3!N!Me"!i!pCQCc1abX2*Ef-,&mj8EA@KjV4fRQfkf--,fZP@[Eld
|
||||
|
||||
Z$dq2VmN'A5Bp-hbAY9lHAJFXfQdl+AG,Z2)ME*&GEJRrA-libQIDl@-,fic`*fc
|
||||
|
||||
6K5HKhAEKE`YIq-)mEQiRK(pXXmb@iapGq-+kKCfFELT3q1c,IZ&ZXPf1@pl#b%)
|
||||
|
||||
ffjdZC,)F@FK#&m,)B+r,!D4[CPq-FBbaqZ@-eH&@A,@%-I9,M(@V+THFE3i'I@,
|
||||
|
||||
PFV%p`R[E)f,)lA5*'SmV)SBMaKm`"H(DkkSAQQdeb1%*lP8%I"Kcj(3rX&H6m0M
|
||||
|
||||
IZTkaqjrj`UCT$PZ9X*!!V`m&fSamV5GNj#ReR!CAb"Z-H0XpDBqF`ePa(%eGaiT
|
||||
|
||||
)S-2EcP+HcTr1B+bXmm9Kh'q$6Mf`X[$"KF4R$RhYV2*CXk3m49H%V`fdL)`T"cl
|
||||
|
||||
J+-2j13Fpcq@-E8&E8'&IE%H%!Ne3,pZF#1HDf2Hf""Q,&l1('*Yr8%EphJ1GXSF
|
||||
|
||||
r%JrNr)3rGBV*(aq@mf,a)FC8Kq$ER2+`6KCr)B9h0"r'+0,%0Xm[rQdqSqFB2cQ
|
||||
|
||||
eBU69f4*S4krcbhc8LClZG$iIR'*cIAh0I"abUXM3iXkAEq$(ilQ,49r!j3f+,H)
|
||||
|
||||
maNhp56c112ejNK@"P6JkPXIB&fjK8aKcR!drZX6iG+jqq&li[TdQiqM4U(!CR@&
|
||||
|
||||
rGU+(,&FBA8QAdZJ+kKT@q*eSAPdm1Mm9!Sj'C"RE!a%aQhqm(IAaK-)B'-FE!ha
|
||||
|
||||
jS(fj'%,(Uc#'FK,*f-@9@FC3113DEaI$J@M)*3)Pk"9$i'!+Qm`pccf[0,(*#J2
|
||||
|
||||
h%ZcNS8*JE#k(6ij38,[0q$[cVaRB"FIjhRDA,pSLmUCDTmXQ1P[%8(M@V%X))mK
|
||||
|
||||
*81HhL'j[ZmK(3P'46jb,ab@$h%jI@)iU6J@&a*8bd!J5%NZ'TC%NDKY",5%K9lA
|
||||
|
||||
%%1kQ%f8Z9IE(4kQ5X*9Mq!UPK%dirih2+53-k[E(m!QELQ!-Rl#ccq$6B)6Z-I`
|
||||
|
||||
FQ(52iC0Hd6f'2a&QlKPm`YDG`5GX%V)aI-*'%r+rq)3prJ`qB9260)C2f"21i"-
|
||||
|
||||
feI!B2QRI@@I`#A[5'Ic*-1NH`dIV+GeMrFY8Q(52j8mG(mdXar#TGUKe(X1R`pq
|
||||
|
||||
T1G'EYSlfTT4IFZ446jL-RfpLA2G!eYX*@kf3!1dTXPdLfkfbh5AE'fAlbB5G8j'
|
||||
|
||||
`4rJkCZFXKT(SUhpj-0jKc0+KVIl1dd)2DmAG-GY8*93X&AUb"HYJr,'#0E!H,EJ
|
||||
|
||||
1NCe#Mr)KS8HMKZmGh)rJ,V"iE"haZ#h!9,BPYJl''HE&0`Sp@9F+$qSClfFqB9h
|
||||
|
||||
h3F6FlY%JbNC43[653pSVJdcS86hQ89H[mbKL98+8Rk[YF1I00PeH*e3+2HTqAYH
|
||||
|
||||
N,LMMCc%HqGX+1SASE&1&f@&'l%0mMD%M4m1VBND`e)EiiS,VCTXD(2B'40m'rl5
|
||||
|
||||
#08#c9pE!hmAAm#U26ZK4E&E48%VR2LJ-CTF+Lq-[Q!rPj"[UJRc-'14f6EKm3Rq
|
||||
|
||||
[HC!!63aQaBb,eS*44IHY`T9#9"TN-1YJpRX&fl4AmahDMZpMp-1B4i1Br38Ef*5
|
||||
|
||||
LZGT1Yf,T@L'kG+hYpILK5iVBA1+i5A[CfL*0plhmp&KCF6DUCir(CadF[VkJLmr
|
||||
|
||||
hl$189GrN0XCQaUTQQmSPVV*HpY33GT)apN++X4le+M"i0Epbf"EcSZR0GUYL,E'
|
||||
|
||||
CL0P[#,$5,pp39-AQe,`b2HjB@cfAZmLMk)i,dH$ilTe,er+S69fpF0LG9mb$!l[
|
||||
|
||||
R31a#i(BDla#LU"ri@"l9MH5GKNUFPjh[CUb%le$F&p6Y@VGPQf+Mf`$HhiaG`0F
|
||||
|
||||
EE!CpNpCmJ'NLh(AkA6XZh4NrZ+jVe`eZK4!eX*L4F(JZ0X03ArHcH#pICpR!*Pl
|
||||
|
||||
XK4j0L8ffh'rc-KeIere1L4i-[$eMkE2E5r8'IIXP(S2Gl*Q)Zf#a'@X,Qq&K$)b
|
||||
|
||||
8&-E"[@,S'A[+pp5)VrqCMI&KiNfa[Q3Qde9lQGE01baYqAD,Zb2SkYi*qa$K!H(
|
||||
|
||||
QrQk@*rZq5ckG*6lNDIDh!N0&FHA[kK@2A1Tq5ZHFEh)rKLLeYSe0M3qAR,I8E&J
|
||||
|
||||
jY+[rT[A9)lQhp[p4)R[CAjVd`eG)q5Ap59[1Ed$+lfq3!*Xb2P4bhK@8@k6rTRj
|
||||
|
||||
JV+rq[$NqA2U`m"9NK3VKAUem9mqHIDj8lbP"PFc`j0R0lNQ*I,N$6AVCdp18*hY
|
||||
|
||||
f0%'EZEh)H$fUN6,B3ica+pmIjZHp2ebp!DT9@&,)#Mf''B9-IjQPr#f@rm`"TRV
|
||||
|
||||
fXT+Kq5E,f4-2X#q@$(82A'Tf[iND,j2dTmcpQ*4$$h,S#F8M6-VMR%F+f4IGNqB
|
||||
|
||||
J'pZ22,VGhpLkJDP%PD'3!+P'N!"h!rF@[MkB[ljcr`h&frIIb#bGV(J(mUN2X4*
|
||||
|
||||
pX9j4GNhmp4Y3'hcTK+D*KTP-YEkVC$Za8E*$BZ+*q*Y0FrMmf#+ql$LLcLXFCJU
|
||||
|
||||
2[K5SU)%*YQ!q)e6KX1%9i!l`mjL@,h-VR'U"@M4@E)Vpm1i&"NfaDF-GpbrBfZ9
|
||||
|
||||
43qpR0r'kZ8c&&BRN0640K&FKHr90+PMRPJr'GaLkK'MXKd,di#&8q%UQd23bTI"
|
||||
|
||||
9"Y@$aT[+kbSUjl2Z'0pB$phR08+dF1AJHN20YhDrGZhcfjrC,IPAlKKLCBC5[4k
|
||||
|
||||
q9Idh5c&Z18Dc[QH`6BT`b"(jr6f$$LR#)NHSe0H#a(a5Q2KG+Ee$aFHh0DPJl5(
|
||||
|
||||
93@8ePZK,p9Z@,YNC(kbfH)D&!Aj)MVPY*'C3MV'dDpHCrHTGCHB"TLM1TeLdU%9
|
||||
|
||||
-9@4Q+N-4da3eSVGlhF4QX!,1CRRd4iAX3Xj@qF4Il+k`@5b@hZfl9Y@m`Nb'kFM
|
||||
|
||||
m(e%[4TI(rJ6aDdl'AmecRb,-rM4HPmkJZV0Y@[@eEEU+cSTV%FR$LPDJFf96T)J
|
||||
|
||||
SBV95T"T4851Qcr(ieNkAfS!@ABKZ@GfXkpaZ+bYKPM*EQ4$GZVVj(+2NSbLEp4*
|
||||
|
||||
QXhjcHh'fc9U5,85T)[CflEd"+)FkYrHZ,P(Zk$8UEGDRHfh@rY@LC[fUCKAPh&$
|
||||
|
||||
@Y1rVM$T#D)9kIMCdBMTe139Pm1GfheX`RFmY90UY2l2DVI1bQkD-SR6CVHVV',Y
|
||||
|
||||
QH0(D)YCpAr&dG(pClTG)CrkkmRDVHaU[M*8KLl[iXi"f16cV#a[iKE'C33leSVV
|
||||
|
||||
cA&k$1%ZK,B8aKer)+j[dSeNDl&DqM%FeA$0FT%'A9r0mEmcBIIHPIa9riGZ2&Y4
|
||||
|
||||
)Z5bXVN6AH6jd%(9@BZSH+"mmR)p+fJ,I1r!p$0mpm2dGI$I#GaYmI`rI25-pFcj
|
||||
|
||||
Ib+CiY,#QH5B*Jb`#R#"`$J)R!Rm,r%fb2`5r!f`%81ZYQ*CVS1I,dCQD4M[6f8"
|
||||
|
||||
d%aZ`,C3pl(R%#1`5BJ$fKC34E!2I+%5,Z6XAc,!&GAHH@mc&V-9$`JriRE!1mdm
|
||||
|
||||
QBJfY6"1EAXca96'V%%d15UJ[MKrdU2JbblTde+I(r2fRV)GU*0F[GKFZ'6FZ&@C
|
||||
|
||||
!@&e$S`1V*BfZ3,[Ekc'f'QM#1TGaI6mfFAd[dRd&lTYa2mhe[DcQqPkGarAYVFD
|
||||
|
||||
pRq[EGj!!kh[Gb2@pdFVerHebVZqYjlLqJ6bZladIehI`(Ul[(a4Fhf(J[@rMqRk
|
||||
|
||||
qJHZ,jh2ph!,FAqIkPGrNqY@YA,rQDG`$A2piD5R$)dE#I+49a0+%1a6`miQp3Qa
|
||||
|
||||
bq2hBFJaMcC%A-H[Lh9kI1084#2JDa"!f3ALEk![b$C%30K$$+Rp)$+Z#lAk4M'@
|
||||
|
||||
U"BZ%FY95Keh3%Y-m5!m&aNNZUbm3$MY$+e3GhSKrHRQY-ib9%UaRb2XM&r&Bb[Q
|
||||
|
||||
$#1m2Y(MG+riPr[FUR"'4$dHFrL$[$S4iX30Jl8iIhq)0r5khhm926M)p@LJ6T9)
|
||||
|
||||
i'P,4l,[)jI1kP[&L+-6l`aiMMHaaP!k@(kR(!$5jIF64)2HV9c"fkm2Bb8M[NA,
|
||||
|
||||
5*ahe$KKB9T9'TSPBKI4**`H4UR2Kk*+M&9J[`FHC*Q&NUD#pVUA83F[45Jadk'0
|
||||
|
||||
F3Yf1$dpTM65,Hfl&AGM3!#1U'a&eQabGKF82I&eA%c-D$%HjjT%"U4TMFAb*[&A
|
||||
|
||||
h)@)HETXFRBf&$h`V0NVHj1U3!,`K#cY(qL511H*j`3MI14L%iN0H')LU%pY@kEb
|
||||
|
||||
e@+I!ap@!&jDr$K6[395bNR+a,%&ISM6!LST@Uj*V5MUX3Y#A)"$4+kM@NKY`il$
|
||||
|
||||
S30pF$R`T#q@S*(BHeKMSieHp#Flf)`,0AQTaDcb@&2)PHQQ)5fb5Xdb1cXF+!Vj
|
||||
|
||||
N8DB2,Ic5f4Kjid'T!M!XRlE0,$48%8&NcjVeLhiPLG[pfVbedR#BF'qX0CFl+(-
|
||||
|
||||
SP#2N$)DCki1*FLTMEYAMF%qMfLlECUkT+5IZR$kIUlACYmcS)YhC12(&iZ3YB9'
|
||||
|
||||
@5Q5*+ZHdkID)X$BCAmp+hXKTKT6AHm#U3r4C*hSQB(BrU*ZE[*&EJ[hH"NF&f1H
|
||||
|
||||
b`j%@Ei"`&+-i5TRYhSDUbbZ*lE"hTGJB!9#%@0JA5pj3Yh-5l&V,'fQFRq0a03C
|
||||
|
||||
$hZ956TYb(mp1hP#k+8NN)bQBbZ-#L*FT4c0ATc*h9&5!)3dB`XSCTF08SdMC5D3
|
||||
|
||||
Pj6BcCAk9Up8CNNK#jN9IDNVH8!QCSr)k39+0G(N`aFD&eSVN$99-XdNF%CZY,D(
|
||||
|
||||
`"a@L69D5SkS@&F+T)ekr#"MM-CcF0*pfUMM`5Hd-*A450pjlk`mPT8VU"Y9h0R3
|
||||
|
||||
Mi#,4b)#J'D-9V[Mh#PIqZX**-8jAH0BrUp"aT*4UR0)#8Sh6@T!!8Se6@T!!maX
|
||||
|
||||
Yd(kN"FGd1[HIG2TA[3DH,8Mf'TBDXp4V02ZFVQ8q2,U3!#'KemM%T"XRp@#KVcU
|
||||
|
||||
Y"q@f5Y+$A#aMZCD&Srj`4S3qiL3hckljPY445pa8@+b09#FYcCj'[bpc@BGcr'Q
|
||||
|
||||
!%69iq@)m[C*8URU(RG4!'ib%'PfYVS`*8j,-6"h[aReIXbG[D8k5c,e@cYh[$#h
|
||||
|
||||
lT)pilFFr65[(JLU"+N',p`QF2Y40KM[Pq2-plHN1e&CT4R@a((P61@0C"rU4'Q`
|
||||
|
||||
blVmMh8FNDTaTr9MRD@`4JjR-qSM6-pGM1,T84T8160L3!*%BDI-(2jh'hIh8YR5
|
||||
|
||||
r8BZ42Y@"2cR5GhfQ,m$+0,B(FZ(*qFCchdR[JG5Dl3[K98[0EFBhc6Jf!k'Hj$p
|
||||
|
||||
R)(rUIIG)ebZT#lVHd,,'8%3DJQ5UfdlEP"@LKiU5A8P9!ff@U2hH-(@biF`FQ[(
|
||||
|
||||
KV+6++NJeiI9JS(a#A@K@FPTGe,p@Pj4QR&)AdSc6kT,5M&2U3T15dqU5QT4mULl
|
||||
|
||||
T5FPrl#eaeipXJ`L95k4YN!"fmDV'M(FlXp`hrMJpBDZc9%XlCB(Q0M6#dJJhdpT
|
||||
|
||||
%2bZdFd30'KTT[d-6#2rA22prCQFCZHEjar[pNj2C69PYp)K@DM)V+8'fT!3C%RU
|
||||
|
||||
0$!Sc%%F&0K8NII&jQb@NScQPp1@%DKc0DD4,rDbV-ccd@PV(lCAPY$H4%a*G2UI
|
||||
|
||||
ARl'MdM)(c3+5MpDF8)f1Rr4*kNc)faB*9I4DMcVDlZfJPej1UXfAEck8RMde1"C
|
||||
|
||||
Ci0@')p(QjN#S(A*Mr%a[J*8"E)T3G!%pL5YhHBl+"RVj4bhpa)5,Y@G#d)*M[FH
|
||||
|
||||
rp@3IGap(N9*kF+TlbrUSQrlA5IIaD[aidXeYj&CVNMH83&CM+!&9RaC+%&Q"[`%
|
||||
|
||||
!PM5C'9(,)ph(*fUTr9!YMqT9DV2iP&iGfErj4+r'r8D[mMkHFibb02iMPNjf1PA
|
||||
|
||||
[d("$VLh(CI8d(p1LX&VN*cJbP(8k[pfF2kE#ZPqTX(51-%LC%ZXU[a22)[*i8[E
|
||||
|
||||
rZJ[cIcUGL4G#pHMBk,e2kCF0VX,2PP#E5Iik[#T1$qmHrqXJc[6'Fa2`XLUETTM
|
||||
|
||||
$*YV-$D3cYp12%m#qEb(qhJ$feL8eGE5PqJMF0!YqXU&'QZAY39+9b(8[r8`"-MX
|
||||
|
||||
Ah$6![T!!ITF!pTb'bfV*EbNA&PMaKL[H#UA+i@kTX"!qGeH&C3R&EkCI&X"$k6d
|
||||
|
||||
9PN9@f#m[VUY"R%+aB%N90%@4PhahPUZj([c3IkY-$A%eUr''+[Q8"m(LQS3[kcE
|
||||
|
||||
1G+!PiF[1j8b6mBiYqG4I![EZK'rFji"Ab"55leDmdYV+9*,[$[MHa&2kj,XIH(K
|
||||
|
||||
90KkIa-Ep'I$!Tj5(&h&2b4cN`,G2pSf$$kqZ5Vi*m(hh+pHLCV(B#pqMEAp*2`L
|
||||
|
||||
K$S-ce482X[1!F4&mDd`jE#EL`-(e-DD6q,X(FCd12IXm1+#IdU#-2SFi1q)HB*d
|
||||
|
||||
54KI`ANVie'C`8jVJFZTNa%85A%ip'ebqP1"bkZr$jj-acJ0'8-Di!,i@'@-Q-2E
|
||||
|
||||
*q68KTiMXZ`ja[9RqCFj@hp%rG"RpQjINMlqNrpQ&-qA@"ki53rAP&2rr!!!0$3p
|
||||
|
||||
YGbpRBh4PFh3Z0MK,,VN!N""453#3%#pd!!"+8`#3"2q3"%e08&*0680$!3#V,jH
|
||||
|
||||
ZUc!jB!!!"M%!!"R%!!!"V3!!"E(*MaZS!*!'[VXM4!iL+Pj0j%)PIdhl9fbRBC!
|
||||
|
||||
!DR1(JAFp3hUJ2KNcZ@(k&LeHlIYc*cMM1X2GRCf"!*`N(81C&iAQNTm4&Ifii1"
|
||||
|
||||
EpGII4h6#PiP+'R-jb[e$&IeM12rA3hh-XBk+D2XK9#@U!P9e!@eRU22XRT!!%ar
|
||||
|
||||
%6jaP3[FjFKhiIjQ@hidE$&25cAm$`-IrIXai*1U*jZd88q%pXX1%F$M`RNJbAQS
|
||||
|
||||
ih%%N0J*@A""6p[pE#%1,cL9X%K8j[Z%i38$F)*'R%8!QpTQQT&06TCMf4amme9+
|
||||
|
||||
jii[1iC(HE43E%aa#QlrCjZ4[GSL(8*!!e8D-E"#r6LR@&GN3aF6F'028K*cdTGk
|
||||
|
||||
aT$fkUhhK6F,P(Tj11!CFTLJ+QQSXDINp,M$RL-+Cm9q6j"VK+Hr'rhrjXB16b1@
|
||||
|
||||
iec&AC&Z,)bAP)A[QZNkT`brFF9bj0@L(b*(4H3)$i*YCbh9`YK90aj%$0a!Gm&!
|
||||
|
||||
,de[B3!XlC'%$"-Eme,D0'(Z229-8DlB`9Q$FC!Y6@9L'KA%@PQm[")V0YM#PKBP
|
||||
|
||||
$[mI#m!L#i#MfjAH50i4eE512Q3bj@@90I4m!N!--!'XcXfpJlh2Ij$4lRaZHF-P
|
||||
|
||||
a`Tr-D)4&@%FjIAiV9hi5rZ3i@3NqRhV5`hI'm8m[3MNjENHi%AjN`!NMR"`rbB$
|
||||
|
||||
bTrc)FA,m$%r*F51Fm*03FTa`FTa`-Q#%%hlN'4R`Pa`RA(+FF+mMamRa)mq2m$2
|
||||
|
||||
#bB!#GjN8B'@Y6-+0iUpN*rl)-F)*2m)*8[#%!j-9H"9SN!!()1QkKK#+`Hm@K$S
|
||||
|
||||
HJ&m,rN[#E`hmIJLEJ,q0bk)PQTCS@&q4J@q@4d"9U,FU)md-(0Yrf-'kLSC3Ech
|
||||
|
||||
QTZ6PDfM!,6kXTJh48"8c3%-B$Af2ZR8CG9Ip2$-35k-p#&9[4Zd)$4`EE%%G46!
|
||||
|
||||
,R0"9-23T99CN34j4,-#2%@HJ4P(6T'aDQa#N[iMDX5G2a3J5j8hqU`G8AI)J-HU
|
||||
|
||||
[2pc+8DXTel3Q5K1DDDe`rC'MeMLS#5QV5"2QC-jFKV@(Y,XiDUf$'TI6Q941+fY
|
||||
|
||||
NIrEXmabeMLSdTZC&6Ae8m48krm8h(,@HFXdUSU`BRMk!q[lRHBlD3,RQ4#QENT@
|
||||
|
||||
#"cXRI2X+4ie6jif)dMfM+mkEUadrc9%E(G5'h+TKlGFqRHHS#3He,LFDrPe`h($
|
||||
|
||||
QCBlDa(3e*P+'jG["RP9riDM0,PI9V"`8d09SikJYP'YH1C5kHVfHlZ'SDkKIpI4
|
||||
|
||||
i+LIkaJ28)bpbe,88e9!N694cCG6ZNqFjkMUUN!"T6DE6ZT(h&AViKGmikRVU"NX
|
||||
|
||||
TAdR(H9q1FY4@bY@D,XL9SfF2rY6286HiPp,*+'9G,aJIFG50p#Uce14Gj3Y'd81
|
||||
|
||||
Ek"h5cFV&)blrQ+1f8B8b8UTJU&0#eN-9cVh+8GXGe*U-j!-kU)P6p4b9*UB'dj*
|
||||
|
||||
PCDb-E#IIrF$K4qBkCkfIRK)eFi@ZrFEXr4ae-h@$T1I(e%`C&K,!AUi3T&L#1U`
|
||||
|
||||
I'P&bCG3h(rRp#Fje+d8&50fBrKHeFp&j@4Q5M3GV$pea1eGSfk+(0$9pa80R1GF
|
||||
|
||||
ZCkfce*a5FDbGI1mKMRSpifUSq482fFRj!BlD6Id+#UPkaDr(MfcMU0YGVSSeRLY
|
||||
|
||||
8Z0V[F05H43q4)19lk0aM"lL(GMKViS"LkT1'T(MH+rPeTkZ3!*U"!([&H8FjkLl
|
||||
|
||||
+0@RS306mKfX[64ZJ+`31D"5@fGUaCaUiVRd8Y@!C+5NVP42Ef6h&a0E[S,D5e*Z
|
||||
|
||||
k$e*4k[,4R"1qUq@S0cKV-k$Hk86c@fiEqT2V*rYSlLHcfePEppip1YM9Hl2Del9
|
||||
|
||||
2!&`"@TQ,U#F1&[Z''jdelZ4b1(ZHmdimH"0(45eR)(&!*q9f)f6q6PCX0VTUBad
|
||||
|
||||
IAd$pf!@`[ik1Br'KUlR)+fakrN"cHF(H36)2h%jb&H(+NrX0&jMF9VMIj$*$&L)
|
||||
|
||||
T"p)0cLf`Yq1%"AXR6JQ`Yq'FKMf0GB,GdbRXPYLiZ+lq4#IBL8k`%jeJ*cV"6R5
|
||||
|
||||
#RHJ%1p%*GU)6l%3Rf)P1h%qc#+[@Y15RS-eL8qhT&"fJcd&k4dVkK,dC'pb'AVi
|
||||
|
||||
MRZjKXmB'HccD3(IrcJ8G(KYmfk)&p1R"5Hkrqa'fKQc`$Bdfm0&Ek'dF5*Cm&25
|
||||
|
||||
6E"T+qQc(16M5i"iI4FpKHCCb3p#-XSR6I3[1YF$(e@dVrAm(hAhGA,f#1a4fVQ`
|
||||
|
||||
D)a0bM1IcX19PNiJXd-QrQrjp$rTP0Nh4$ljDEE6C0*GdfSPEQNJ$[AaI"9dkQjE
|
||||
|
||||
)"&rjZ5PSlpQXL6c)65I42'&jkHi((6HE659pGY(F%GhJrk#CBp-AQC!!QcfG`RF
|
||||
|
||||
BE0C'2GbTm18(Qh@4"hI+cbI"'a-fkb-2I05,Qq*VI86`ZS90Dq6"IEUNPpZrZ6d
|
||||
|
||||
IkmP@hp@`f9$5UmK,"LjZ2dGjKIdd'pTRSrf,Re6[[[HdcbYXX0R3aK[KcVI)#mr
|
||||
|
||||
A-dm"R8jJFcLjAc2T0r!1Xr%Ph(NRKdhm"Y1PM9qd9#9(PFc#![X)[SNKr!e@jAm
|
||||
|
||||
!N!-0$3pYGbpRBh4PFh3Z8&"$,VN!N"!4c!#3%%+&!*!)rj!%68e38Ne33d-"!+X
|
||||
|
||||
[PkkV-$P&!!!'-3!!'Z!!!!'T!!!&bE5F%03!N!B"fL0%$L)UANhN3L9r6IYAE+G
|
||||
|
||||
KN!"UFiH"Gce$HU!q'61jBIV#iB$[cjhJM1X-GhH'!`%ib6Q'-Lm+c58r)bVkFF(
|
||||
|
||||
"YqU[[irS4$#9MENFjIkKL[iaR2rVS6lQ@%G&Y2d3UK*9JDUkJ,Bce(Pf6fJm&6R
|
||||
|
||||
b2Z8HRJiXa'A+ir""h#2TreqK*11PKX-G4'@dI[MrP@fl(cXiL9b1Haec4BbeKmP
|
||||
|
||||
aeJj"iNA$iL1d#Y1J+HR89#QQrG%86l98l[LLFhLNlhad)NaL2JK&0pZFr-d1m4!
|
||||
|
||||
+XYS)fcSm[diTeKAC%-A8h"M6e)5Fp+AHXD3p1ZNm1FY%rabj$[`E!$0bi`E$P26
|
||||
|
||||
rG@!p"$aQr-JXH*CjLX,-Um9UPGj1-5VH)fY@`*(4VHaDSf,&r6CPrlq&--R1K6X
|
||||
|
||||
*#r!9a`Q#"HZ@0$hdcLR&Z$Fm-LN%a%6I)NG'j`NF&EkCY9`(CaX9iFL4(fpK!IC
|
||||
|
||||
B8#c-*P,XP1dG-@D4KE%@0XR#9"C'PcdhhF,ZXE"3#eYVB3-&a[CDHNU"FB-@YXI
|
||||
|
||||
#PPJD!bcX5f0T(aH0)DaV'hR-C-M0+Q[Uq``!Da0l'f3fmMSr"jhCCQZ%N3NRNdf
|
||||
|
||||
14LJRP"rPR[a@3Sqr%8D1NjAJmk5Hp2#G-Ic6Le"1MJm)Pachb(2###I(6c*J%8k
|
||||
|
||||
j%8k1RiHRj,J46[K*+$P11$P11"P3`JNrmS`-q)!-Z'6!D6eKj2L4C`f-F$+J`(8
|
||||
|
||||
Q"D$m9QE4e,T1r"&qK,q%%k6J#3FQ+c!qS%%HJ+LU#N)S",rE%'S`i2Fjq"D$ha,
|
||||
|
||||
iI4qf+2P[K53BJQi)Q['&0I#IjBQL)Y4CP"42pjcUHm,'ZSf'8'HBF--Ck@qdLS0
|
||||
|
||||
b3K-d'HXH'L*+S#ZS9C93Dp(hThY##E32SH*'Y!@KRP2p0@MV!TJ"6QM*DZUi,'%
|
||||
|
||||
T+JeJ!r"$PM03TD!SBLUKM%E&Qd60d0-c)3Z*mVDqK3&9&I13!!6eTfr[iUM&P'Y
|
||||
|
||||
#%F4446G@Z(l(88YXe)LB`Z+S2TE@Pf(0!mTp(,A84Uf3!",*H&STD'4qmZcc(,@
|
||||
|
||||
-+M3XC`4&IJbl#Phql%Z1'UCF8eL3!,#@e`G3hrdd`e(,+GHd)+EL%XlQBDHLAlh
|
||||
|
||||
-85ZSm`B%mB'K&HG0PBjFj+L90QTjHXf`jUXI6h28L)eDPKBdpblBE[Mm*BjDaA6
|
||||
|
||||
94Z1DiGV"R*4rj+M9$PGCAcdSS+ZfQD2@8+iCI$4qqhSpdmj4DkPIeF4)2#fiaJ2
|
||||
|
||||
8Bbpbe(889F1L)XMT!QVVKFXFYBiUT"YMLB5UC9b&(RhKCikkRVV"`)8VDEQKZf-
|
||||
|
||||
V4kfRA)f*V,4kp-cqUedFGB0c+Hf-8Y$eQ[B"4pe)Vc*,6IQVI%eEm0!QHSG8IIA
|
||||
|
||||
L@5lrN!#MEUCQhhrr8(p,Ec3@kie#4,V"pIbK)a`9-T!!GBlk-`E@KJ84,f%LG'i
|
||||
|
||||
f[T!!!'KLUKNie$XiPM(N0&lQH[KU'dGYS"j+L['X*Sp(8hPGSl0R1'UMM9U5&&e
|
||||
|
||||
!'c8b%qDSN!"L,3rTKL6Ki3+b'A[l2CZI0G[Y06`a,LMk#PhcpFQ(1'S6pDZSCSE
|
||||
|
||||
PC!&fUR[r&Uj3-eASS(Td!+F,U1H1r2)8jpT#83&5e5EZS1kBGej+JZb9Kc82h(-
|
||||
|
||||
h9kKehN1+R,MPS8ZFDjZpPPPABF@aCZbG`abeRA(9j-b+KmcBG!p(lD"q"B9NGG@
|
||||
|
||||
[CimeFp5G$PGXM+5cUec0YcMUVRN2@9(2pG$Xii2F3jhf'KR%ZMUQL6M[[CaIGcX
|
||||
|
||||
+b8Q)f,HFGj+MGP'Z-8d&S[SrA2I32!5k3L5#cQ1CV4NkAXjer4pehS"JT*BMJmh
|
||||
|
||||
eq+jHMVUAFXdD@Pa-LB8NHQRI3K)PI3p-0D6jHqhb!-,0lkJIrAq#kpTYIkZh1S$
|
||||
|
||||
iJj!!%H0,"hSUqR8TjiTU6d$LH!3qd"l'QVp5(*Z0MQj%N5IR8$IK#2YVk#b4%AU
|
||||
|
||||
KAhRjVG*[D*cA0T*HB1mJp`hf9R+*B@mR9a,f0R*MBGp1mJVX655"`0j)XK,X1mL
|
||||
|
||||
pKlf"*+irG2*l,$B1E[#"6T2S$,#X@[56ejba+FlV&"bJcm2dMZ6dm6Xk0U4jAES
|
||||
|
||||
MHGhp&Sp0DH#"lZkGmrT#0Q@"!rVX)TRXhr[K0j4X`S%(2RS$[3RXDKCpj(@KE-T
|
||||
|
||||
cqZ`NL6E3i"kI4160319LE["D@$B9G'mQ#4Ai1,Ued1qGG(GeFr6blT!!hqqbU3a
|
||||
|
||||
-b$&jRrZ0-TY)B)&1lYedll[ACE1T#Rl`e9TlcUBkTp0ZdVF%-H4Z[lGR8a1Bi#X
|
||||
|
||||
h0hN["GM8"KlNTJYSfQ*jrHjlI6UE66PpZQMZ#'l`[pH@XGNEQ*!!Qr-kq@mqf+`
|
||||
|
||||
,HVK6rLX60R@""hI+c5IHHaBfk`-2I,5(G,jrpK(H5aSfpB%(pqQkANlrj[4mV#G
|
||||
|
||||
EHm2$CN01V`9H%R"aqMR+bhpj`iDqe%&p8bAIR!qTj%[4$kpFY(MK'lcmYcPXk&Z
|
||||
|
||||
H1lcmlTi0lIT[mPVbJIFUL!elGjRM4BM8c8"+#$@"@kr%qK5GrJGH8d5JeDSp%6Z
|
||||
|
||||
S`aY94TZmpLQ+$H(Nh"cl%r`RK-KrL#Vr!3#3!aq$!!!"!*!$!43!N!-8!*!$-Tr
|
||||
|
||||
lRLe!rr#`!,K[$#eZd!6rm2rdd"lm`FAKdkSV8FY+$deKBe"bEfTPBh4c,R0TG!)
|
||||
|
||||
!N!06594%8dP8)3#3"P0*9%46593K!*!BUc!jI3!!8M8!!!&'"1"2l'mDG@6JrHc
|
||||
|
||||
K@5U#NI*HN@GK!Z"2kQ`FG&2UN!"S!!,L@5[48(adA`CdC!EJ6qj[8hJS!!EJEHl
|
||||
|
||||
LEe5!)D$!FJC1ANl!*IrX51FI-#D`jL63G!*&0K!+1Li!&Ri!)VX-S"lbUKQJ(Z`
|
||||
|
||||
3!+SDI!$!#3ZT8,aIE!!!Q$!'8!6"aG!!N!-3!#X!"3%B!J#3"`-!N!-"!*!$!43
|
||||
|
||||
!N!-8!*!$-J$j(l!@#J#3!a`!-J!!8f9dC`#3!`S!!2rr!*!&q@G%'@B:
|
||||
|
||||
@@ -1,91 +0,0 @@
|
||||
/*
|
||||
MacOS_Test_config.h
|
||||
|
||||
Configuration flags for Macintosh development systems.
|
||||
|
||||
Test version.
|
||||
|
||||
<Revision History>
|
||||
|
||||
11/16/95 pcb Updated compilation flags to reflect latest 4.6 Makefile.
|
||||
|
||||
by Patrick C. Beard.
|
||||
*/
|
||||
/* Boehm, November 17, 1995 12:05 pm PST */
|
||||
|
||||
#ifdef __MWERKS__
|
||||
|
||||
// for CodeWarrior Pro with Metrowerks Standard Library (MSL).
|
||||
// #define MSL_USE_PRECOMPILED_HEADERS 0
|
||||
#include <ansi_prefix.mac.h>
|
||||
#ifndef __STDC__
|
||||
#define __STDC__ 0
|
||||
#endif
|
||||
|
||||
#endif
|
||||
|
||||
// these are defined again in gc_priv.h.
|
||||
#undef TRUE
|
||||
#undef FALSE
|
||||
|
||||
#define ALL_INTERIOR_POINTERS // follows interior pointers.
|
||||
//#define SILENT // want collection messages.
|
||||
//#define DONT_ADD_BYTE_AT_END // no padding.
|
||||
//#define SMALL_CONFIG // whether to a smaller heap.
|
||||
#define NO_SIGNALS // signals aren't real on the Macintosh.
|
||||
#define USE_TEMPORARY_MEMORY // use Macintosh temporary memory.
|
||||
|
||||
// CFLAGS= -O -DNO_SIGNALS -DALL_INTERIOR_POINTERS -DSILENT
|
||||
//
|
||||
//LIBGC_CFLAGS= -O -DNO_SIGNALS -DSILENT \
|
||||
// -DREDIRECT_MALLOC=GC_malloc_uncollectable \
|
||||
// -DDONT_ADD_BYTE_AT_END -DALL_INTERIOR_POINTERS
|
||||
// Flags for building libgc.a -- the last two are required.
|
||||
//
|
||||
// Setjmp_test may yield overly optimistic results when compiled
|
||||
// without optimization.
|
||||
// -DSILENT disables statistics printing, and improves performance.
|
||||
// -DCHECKSUMS reports on erroneously clear dirty bits, and unexpectedly
|
||||
// altered stubborn objects, at substantial performance cost.
|
||||
// Use only for incremental collector debugging.
|
||||
// -DFIND_LEAK causes the collector to assume that all inaccessible
|
||||
// objects should have been explicitly deallocated, and reports exceptions.
|
||||
// Finalization and the test program are not usable in this mode.
|
||||
// -DSOLARIS_THREADS enables support for Solaris (thr_) threads.
|
||||
// (Clients should also define SOLARIS_THREADS and then include
|
||||
// gc.h before performing thr_ or GC_ operations.)
|
||||
// This is broken on nonSPARC machines.
|
||||
// -DALL_INTERIOR_POINTERS allows all pointers to the interior
|
||||
// of objects to be recognized. (See gc_priv.h for consequences.)
|
||||
// -DSMALL_CONFIG tries to tune the collector for small heap sizes,
|
||||
// usually causing it to use less space in such situations.
|
||||
// Incremental collection no longer works in this case.
|
||||
// -DLARGE_CONFIG tunes the collector for unusually large heaps.
|
||||
// Necessary for heaps larger than about 500 MB on most machines.
|
||||
// Recommended for heaps larger than about 64 MB.
|
||||
// -DDONT_ADD_BYTE_AT_END is meaningful only with
|
||||
// -DALL_INTERIOR_POINTERS. Normally -DALL_INTERIOR_POINTERS
|
||||
// causes all objects to be padded so that pointers just past the end of
|
||||
// an object can be recognized. This can be expensive. (The padding
|
||||
// is normally more than one byte due to alignment constraints.)
|
||||
// -DDONT_ADD_BYTE_AT_END disables the padding.
|
||||
// -DNO_SIGNALS does not disable signals during critical parts of
|
||||
// the GC process. This is no less correct than many malloc
|
||||
// implementations, and it sometimes has a significant performance
|
||||
// impact. However, it is dangerous for many not-quite-ANSI C
|
||||
// programs that call things like printf in asynchronous signal handlers.
|
||||
// -DGC_OPERATOR_NEW_ARRAY declares that the C++ compiler supports the
|
||||
// new syntax "operator new[]" for allocating and deleting arrays.
|
||||
// See gc_cpp.h for details. No effect on the C part of the collector.
|
||||
// This is defined implicitly in a few environments.
|
||||
// -DREDIRECT_MALLOC=X causes malloc, realloc, and free to be defined
|
||||
// as aliases for X, GC_realloc, and GC_free, respectively.
|
||||
// Calloc is redefined in terms of the new malloc. X should
|
||||
// be either GC_malloc or GC_malloc_uncollectable.
|
||||
// The former is occasionally useful for working around leaks in code
|
||||
// you don't want to (or can't) look at. It may not work for
|
||||
// existing code, but it often does. Neither works on all platforms,
|
||||
// since some ports use malloc or calloc to obtain system memory.
|
||||
// (Probably works for UNIX, and win32.)
|
||||
// -DNO_DEBUG removes GC_dump and the debugging routines it calls.
|
||||
// Reduces code size slightly at the expense of debuggability.
|
||||
@@ -1,89 +0,0 @@
|
||||
/*
|
||||
MacOS_config.h
|
||||
|
||||
Configuration flags for Macintosh development systems.
|
||||
|
||||
<Revision History>
|
||||
|
||||
11/16/95 pcb Updated compilation flags to reflect latest 4.6 Makefile.
|
||||
|
||||
by Patrick C. Beard.
|
||||
*/
|
||||
/* Boehm, November 17, 1995 12:10 pm PST */
|
||||
|
||||
#ifdef __MWERKS__
|
||||
|
||||
// for CodeWarrior Pro with Metrowerks Standard Library (MSL).
|
||||
// #define MSL_USE_PRECOMPILED_HEADERS 0
|
||||
#include <ansi_prefix.mac.h>
|
||||
#ifndef __STDC__
|
||||
#define __STDC__ 0
|
||||
#endif
|
||||
|
||||
#endif /* __MWERKS__ */
|
||||
|
||||
// these are defined again in gc_priv.h.
|
||||
#undef TRUE
|
||||
#undef FALSE
|
||||
|
||||
#define ALL_INTERIOR_POINTERS // follows interior pointers.
|
||||
#define SILENT // no collection messages.
|
||||
//#define DONT_ADD_BYTE_AT_END // no padding.
|
||||
//#define SMALL_CONFIG // whether to use a smaller heap.
|
||||
#define NO_SIGNALS // signals aren't real on the Macintosh.
|
||||
#define USE_TEMPORARY_MEMORY // use Macintosh temporary memory.
|
||||
|
||||
// CFLAGS= -O -DNO_SIGNALS -DSILENT -DALL_INTERIOR_POINTERS
|
||||
//
|
||||
//LIBGC_CFLAGS= -O -DNO_SIGNALS -DSILENT \
|
||||
// -DREDIRECT_MALLOC=GC_malloc_uncollectable \
|
||||
// -DDONT_ADD_BYTE_AT_END -DALL_INTERIOR_POINTERS
|
||||
// Flags for building libgc.a -- the last two are required.
|
||||
//
|
||||
// Setjmp_test may yield overly optimistic results when compiled
|
||||
// without optimization.
|
||||
// -DSILENT disables statistics printing, and improves performance.
|
||||
// -DCHECKSUMS reports on erroneously clear dirty bits, and unexpectedly
|
||||
// altered stubborn objects, at substantial performance cost.
|
||||
// Use only for incremental collector debugging.
|
||||
// -DFIND_LEAK causes the collector to assume that all inaccessible
|
||||
// objects should have been explicitly deallocated, and reports exceptions.
|
||||
// Finalization and the test program are not usable in this mode.
|
||||
// -DSOLARIS_THREADS enables support for Solaris (thr_) threads.
|
||||
// (Clients should also define SOLARIS_THREADS and then include
|
||||
// gc.h before performing thr_ or GC_ operations.)
|
||||
// This is broken on nonSPARC machines.
|
||||
// -DALL_INTERIOR_POINTERS allows all pointers to the interior
|
||||
// of objects to be recognized. (See gc_priv.h for consequences.)
|
||||
// -DSMALL_CONFIG tries to tune the collector for small heap sizes,
|
||||
// usually causing it to use less space in such situations.
|
||||
// Incremental collection no longer works in this case.
|
||||
// -DLARGE_CONFIG tunes the collector for unusually large heaps.
|
||||
// Necessary for heaps larger than about 500 MB on most machines.
|
||||
// Recommended for heaps larger than about 64 MB.
|
||||
// -DDONT_ADD_BYTE_AT_END is meaningful only with
|
||||
// -DALL_INTERIOR_POINTERS. Normally -DALL_INTERIOR_POINTERS
|
||||
// causes all objects to be padded so that pointers just past the end of
|
||||
// an object can be recognized. This can be expensive. (The padding
|
||||
// is normally more than one byte due to alignment constraints.)
|
||||
// -DDONT_ADD_BYTE_AT_END disables the padding.
|
||||
// -DNO_SIGNALS does not disable signals during critical parts of
|
||||
// the GC process. This is no less correct than many malloc
|
||||
// implementations, and it sometimes has a significant performance
|
||||
// impact. However, it is dangerous for many not-quite-ANSI C
|
||||
// programs that call things like printf in asynchronous signal handlers.
|
||||
// -DGC_OPERATOR_NEW_ARRAY declares that the C++ compiler supports the
|
||||
// new syntax "operator new[]" for allocating and deleting arrays.
|
||||
// See gc_cpp.h for details. No effect on the C part of the collector.
|
||||
// This is defined implicitly in a few environments.
|
||||
// -DREDIRECT_MALLOC=X causes malloc, realloc, and free to be defined
|
||||
// as aliases for X, GC_realloc, and GC_free, respectively.
|
||||
// Calloc is redefined in terms of the new malloc. X should
|
||||
// be either GC_malloc or GC_malloc_uncollectable.
|
||||
// The former is occasionally useful for working around leaks in code
|
||||
// you don't want to (or can't) look at. It may not work for
|
||||
// existing code, but it often does. Neither works on all platforms,
|
||||
// since some ports use malloc or calloc to obtain system memory.
|
||||
// (Probably works for UNIX, and win32.)
|
||||
// -DNO_DEBUG removes GC_dump and the debugging routines it calls.
|
||||
// Reduces code size slightly at the expense of debuggability.
|
||||
@@ -1,9 +0,0 @@
|
||||
/*
|
||||
dataend.c
|
||||
|
||||
A hack to get the extent of global data for the Macintosh.
|
||||
|
||||
by Patrick C. Beard.
|
||||
*/
|
||||
|
||||
long __dataend;
|
||||
@@ -1,9 +0,0 @@
|
||||
/*
|
||||
datastart.c
|
||||
|
||||
A hack to get the extent of global data for the Macintosh.
|
||||
|
||||
by Patrick C. Beard.
|
||||
*/
|
||||
|
||||
long __datastart;
|
||||
@@ -1,107 +0,0 @@
|
||||
#-----------------------------------------------------------------------------#
|
||||
|
||||
# Makefile.DLLs, version 0.4.
|
||||
|
||||
# Contributed by Fergus Henderson.
|
||||
|
||||
# This Makefile contains rules for creating DLLs on Windows using gnu-win32.
|
||||
|
||||
#-----------------------------------------------------------------------------#
|
||||
|
||||
# This rule creates a `.def' file, which lists the symbols that are exported
|
||||
# from the DLL. We use `nm' to get a list of all the exported text (`T')
|
||||
# symbols and data symbols -- including uninitialized data (`B'),
|
||||
# initialized data (`D'), read-only data (`R'), and common blocks (`C').
|
||||
%.def: %.a
|
||||
echo EXPORTS > $@
|
||||
nm $< | grep '^........ [BCDRT] _' | sed 's/[^_]*_//' >> $@
|
||||
|
||||
# We need to use macros to access global data:
|
||||
# the user of the DLL must refer to `foo' as `(*__imp_foo)'.
|
||||
# This rule creates a `_globals.h' file, which contains macros
|
||||
# for doing this.
|
||||
|
||||
SYM_PREFIX = $(firstword $(SYM_PREFIX-$*) $*)
|
||||
DLL_MACRO = $(SYM_PREFIX)_USE_DLL
|
||||
IMP_MACRO = $(SYM_PREFIX)_IMP
|
||||
GLOBAL_MACRO = $(SYM_PREFIX)_GLOBAL
|
||||
|
||||
%_globals.h: %.a
|
||||
echo "/* automatically generated by Makefile.DLLs */" > $@
|
||||
echo "#if defined(__GNUC__) && defined(_WIN32) \\" >> $@
|
||||
echo " && defined($(DLL_MACRO))" >> $@
|
||||
echo "# define $(IMP_MACRO)(name) __imp_##name" >> $@
|
||||
echo "# define $(GLOBAL_MACRO)(name) (*$(IMP_MACRO)(name))" >> $@
|
||||
echo "#else" >> $@
|
||||
echo "# define $(GLOBAL_MACRO)(name) name" >> $@
|
||||
echo "#endif" >> $@
|
||||
echo "" >> $@
|
||||
for sym in `nm $< | grep '^........ [BCDR] _' | sed 's/[^_]*_//'`; do \
|
||||
echo "#define $$sym $(GLOBAL_MACRO)($$sym)" >> $@; \
|
||||
done
|
||||
|
||||
# This rule creates the export object file (`foo.exp') which contains the
|
||||
# jump table array; this export object file becomes part of the DLL.
|
||||
# This rule also creates the import library (`foo_dll.a') which contains small
|
||||
# stubs for all the functions exported by the DLL which jump to them via the
|
||||
# jump table. Executables that will use the DLL must be linked against this
|
||||
# stub library.
|
||||
%.exp %_dll.a : %.def
|
||||
dlltool $(DLLTOOLFLAGS) $(DLLTOOLFLAGS-$*) \
|
||||
--def $< \
|
||||
--dllname $*.dll \
|
||||
--output-exp $*.exp \
|
||||
--output-lib $*_dll.a
|
||||
|
||||
# The `sed' commands below are to convert DOS-style `C:\foo\bar'
|
||||
# pathnames into Unix-style `//c/foo/bar' pathnames.
|
||||
CYGWIN32_LIBS = $(shell echo \
|
||||
-L`dirname \`gcc -print-file-name=libgcc.a | \
|
||||
sed -e 's@^\\\\([A-Za-z]\\\\):@//\\\\1@g' -e 's@\\\\\\\\@/@g' \` ` \
|
||||
-L`dirname \`gcc -print-file-name=libcygwin.a | \
|
||||
sed -e 's@^\\\\([A-Za-z]\\\\):@//\\\\1@g' -e 's@\\\\\\\\@/@g' \` ` \
|
||||
-L`dirname \`gcc -print-file-name=libkernel32.a | \
|
||||
sed -e 's@^\\\\([A-Za-z]\\\\):@//\\\\1@g' -e 's@\\\\\\\\@/@g' \` ` \
|
||||
-lgcc -lcygwin -lkernel32 -lgcc)
|
||||
|
||||
RELOCATABLE=yes
|
||||
|
||||
ifeq "$(strip $(RELOCATABLE))" "yes"
|
||||
|
||||
# to create relocatable DLLs, we need to do two passes
|
||||
%.dll: %.exp %.a dll_fixup.o dll_init.o
|
||||
$(LD) $(LDFLAGS) $(LDFLAGS-$*) --dll -o $*.base \
|
||||
-e _dll_entry@12 dll_init.o \
|
||||
dll_fixup.o $*.exp $*.a \
|
||||
$(LDLIBS) $(LDLIBS-$*) \
|
||||
$(CYGWIN32_LIBS)
|
||||
$(LD) $(LDFLAGS) $(LDFLAGS-$*) --dll --base-file $*.base -o $@ \
|
||||
-e _dll_entry@12 dll_init.o \
|
||||
dll_fixup.o $*.exp $*.a \
|
||||
$(LDLIBS) $(LDLIBS-$*) \
|
||||
$(CYGWIN32_LIBS)
|
||||
rm -f $*.base
|
||||
else
|
||||
|
||||
%.dll: %.exp %.a dll_fixup.o dll_init.o
|
||||
$(LD) $(LDFLAGS) $(LDFLAGS-$*) --dll -o $@ \
|
||||
-e _dll_entry@12 dll_init.o \
|
||||
dll_fixup.o $*.exp $*.a \
|
||||
$(LDLIBS) $(LDLIBS-$*) \
|
||||
$(CYGWIN32_LIBS)
|
||||
|
||||
endif
|
||||
|
||||
# This black magic piece of assembler needs to be linked in in order to
|
||||
# properly terminate the list of imported DLLs.
|
||||
dll_fixup.s:
|
||||
echo '.section .idata$$3' > dll_fixup.s
|
||||
echo '.long 0,0,0,0, 0,0,0,0' >> dll_fixup.s
|
||||
|
||||
# This bit is necessary to provide an initialization function for the DLL.
|
||||
dll_init.c:
|
||||
echo '__attribute__((stdcall))' > dll_init.c
|
||||
echo 'int dll_entry(int handle, int reason, void *ptr)' >> dll_init.c
|
||||
echo '{return 1; }' >> dll_init.c
|
||||
|
||||
dont_throw_away: dll_fixup.o dll_init.o
|
||||
@@ -1,142 +0,0 @@
|
||||
# Copyright (c) 1999-2001 by Red Hat, Inc. All rights reserved.
|
||||
#
|
||||
# THIS MATERIAL IS PROVIDED AS IS, WITH ABSOLUTELY NO WARRANTY EXPRESSED
|
||||
# OR IMPLIED. ANY USE IS AT YOUR OWN RISK.
|
||||
#
|
||||
# Permission is hereby granted to use or copy this program
|
||||
# for any purpose, provided the above notices are retained on all copies.
|
||||
# Permission to modify the code and to distribute modified code is granted,
|
||||
# provided the above notices are retained, and a notice that the code was
|
||||
# modified is included with the above copyright notice.
|
||||
#
|
||||
# Original author: Tom Tromey
|
||||
# Severely truncated by Hans-J. Boehm
|
||||
|
||||
## Process this file with automake to produce Makefile.in.
|
||||
|
||||
## FIXME: `make dist' in this directory will not currently work. Many
|
||||
## files that should be in the distribution are not mentioned in this
|
||||
## Makefile.am.
|
||||
|
||||
AUTOMAKE_OPTIONS = foreign
|
||||
|
||||
SUBDIRS = include doc
|
||||
|
||||
noinst_LTLIBRARIES = libmonogc.la
|
||||
|
||||
EXTRA_DIST =
|
||||
## more items will be succesively added below
|
||||
|
||||
if POWERPC_DARWIN
|
||||
asm_libgc_sources = powerpc_darwin_mach_dep.s
|
||||
else
|
||||
asm_libgc_sources =
|
||||
endif
|
||||
|
||||
libmonogc_la_SOURCES = allchblk.c alloc.c blacklst.c checksums.c dbg_mlc.c \
|
||||
dyn_load.c finalize.c gc_dlopen.c gcj_mlc.c headers.c aix_irix_threads.c \
|
||||
malloc.c mallocx.c mark.c mark_rts.c misc.c new_hblk.c \
|
||||
obj_map.c os_dep.c pcr_interface.c ptr_chck.c real_malloc.c reclaim.c \
|
||||
solaris_pthreads.c solaris_threads.c specific.c stubborn.c typd_mlc.c \
|
||||
backgraph.c win32_threads.c \
|
||||
pthread_support.c pthread_stop_world.c darwin_stop_world.c \
|
||||
$(asm_libgc_sources)
|
||||
|
||||
|
||||
# Include THREADLIBS here to ensure that the correct versions of
|
||||
# linuxthread semaphore functions get linked:
|
||||
libmonogc_la_LIBADD = @addobjs@ $(THREADLIBS) $(UNWINDLIBS)
|
||||
libmonogc_la_DEPENDENCIES = @addobjs@
|
||||
libmonogc_la_LDFLAGS = -version-info 1:2:0
|
||||
|
||||
EXTRA_libmonogc_la_SOURCES = alpha_mach_dep.S \
|
||||
mips_sgi_mach_dep.s mips_ultrix_mach_dep.s powerpc_darwin_mach_dep.s \
|
||||
rs6000_mach_dep.s sparc_mach_dep.S sparc_netbsd_mach_dep.s \
|
||||
sparc_sunos4_mach_dep.s ia64_save_regs_in_stack.s
|
||||
|
||||
EXTRA_DIST += alpha_mach_dep.S mips_sgi_mach_dep.s sparc_mach_dep.S
|
||||
|
||||
AM_CFLAGS = @GC_CFLAGS@
|
||||
|
||||
## FIXME: relies on internal code generated by automake.
|
||||
all_objs = @addobjs@ $(libgc_la_OBJECTS)
|
||||
$(all_objs) : include/private/gcconfig.h include/private/gc_priv.h \
|
||||
include/private/gc_hdrs.h include/gc.h include/gc_gcj.h include/gc_mark.h
|
||||
|
||||
## FIXME: we shouldn't have to do this, but automake forces us to.
|
||||
.s.lo:
|
||||
## We use -Wp,-P to strip #line directives. Irix `as' chokes on
|
||||
## these.
|
||||
$(LTCOMPILE) -Wp,-P -x assembler-with-cpp -c $<
|
||||
|
||||
## We have our own definition of LTCOMPILE because we want to use our
|
||||
## CFLAGS, not those passed in from the top level make.
|
||||
LTCOMPILE = $(LIBTOOL) --mode=compile $(CC) $(DEFS) $(INCLUDES) \
|
||||
-I$(top_srcdir)/include $(AM_CPPFLAGS) $(CPPFLAGS) \
|
||||
$(AM_CFLAGS) $(MY_CFLAGS) $(GC_CFLAGS)
|
||||
LINK = $(LIBTOOL) --mode=link $(CC) $(AM_CFLAGS) $(MY_CFLAGS) $(LDFLAGS) -o $@
|
||||
|
||||
# Work around what appears to be a GNU make bug handling MAKEFLAGS
|
||||
# values defined in terms of make variables, as is the case for CC and
|
||||
# friends when we are called from the top level Makefile.
|
||||
AM_MAKEFLAGS = \
|
||||
"AR_FLAGS=$(AR_FLAGS)" \
|
||||
"CC_FOR_BUILD=$(CC_FOR_BUILD)" \
|
||||
"CFLAGS=$(CFLAGS)" \
|
||||
"CXXFLAGS=$(CXXFLAGS)" \
|
||||
"CFLAGS_FOR_BUILD=$(CFLAGS_FOR_BUILD)" \
|
||||
"CFLAGS_FOR_TARGET=$(CFLAGS_FOR_TARGET)" \
|
||||
"INSTALL=$(INSTALL)" \
|
||||
"INSTALL_DATA=$(INSTALL_DATA)" \
|
||||
"INSTALL_PROGRAM=$(INSTALL_PROGRAM)" \
|
||||
"INSTALL_SCRIPT=$(INSTALL_SCRIPT)" \
|
||||
"LDFLAGS=$(LDFLAGS)" \
|
||||
"LIBCFLAGS=$(LIBCFLAGS)" \
|
||||
"LIBCFLAGS_FOR_TARGET=$(LIBCFLAGS_FOR_TARGET)" \
|
||||
"MAKE=$(MAKE)" \
|
||||
"MAKEINFO=$(MAKEINFO) $(MAKEINFOFLAGS)" \
|
||||
"PICFLAG=$(PICFLAG)" \
|
||||
"PICFLAG_FOR_TARGET=$(PICFLAG_FOR_TARGET)" \
|
||||
"SHELL=$(SHELL)" \
|
||||
"EXPECT=$(EXPECT)" \
|
||||
"RUNTEST=$(RUNTEST)" \
|
||||
"RUNTESTFLAGS=$(RUNTESTFLAGS)" \
|
||||
"exec_prefix=$(exec_prefix)" \
|
||||
"infodir=$(infodir)" \
|
||||
"libdir=$(libdir)" \
|
||||
"prefix=$(prefix)" \
|
||||
"tooldir=$(tooldir)" \
|
||||
"AR=$(AR)" \
|
||||
"AS=$(AS)" \
|
||||
"CC=$(CC)" \
|
||||
"CXX=$(CXX)" \
|
||||
"LD=$(LD)" \
|
||||
"LIBCFLAGS=$(LIBCFLAGS)" \
|
||||
"NM=$(NM)" \
|
||||
"PICFLAG=$(PICFLAG)" \
|
||||
"RANLIB=$(RANLIB)" \
|
||||
"DESTDIR=$(DESTDIR)"
|
||||
|
||||
MAKEOVERRIDES=
|
||||
|
||||
# files used by makefiles other than Makefile.am
|
||||
#
|
||||
EXTRA_DIST += add_gc_prefix.c gcname.c if_mach.c if_not_there.c hpux_test_and_clear.s pc_excludes gc.mak MacOS.c MacProjects.sit.hqx mach_dep.c setjmp_t.c threadlibs.c AmigaOS.c version.h Mac_files/datastart.c Mac_files/dataend.c Mac_files/MacOS_config.h Mac_files/MacOS_Test_config.h
|
||||
|
||||
# cord package
|
||||
#
|
||||
EXTRA_DIST += cord/cordbscs.c cord/cordtest.c cord/de.c cord/de_win.c cord/de_win.ICO cord/cordprnt.c cord/cordxtra.c cord/de_cmds.h cord/de_win.h cord/de_win.RC
|
||||
|
||||
# this is an auxiliary shell file used by Makefile and Makefile.direct
|
||||
#
|
||||
CONFIG_STATUS_DEPENDENCIES = $(srcdir)/configure.host
|
||||
|
||||
# :FIXME: why do we distribute this one???
|
||||
#
|
||||
EXTRA_DIST += libtool.m4 configure.host
|
||||
|
||||
#
|
||||
# :GOTCHA: GNU make rule for making .s out of .S is flawed,
|
||||
# it will not remove dest if building fails
|
||||
.S.s:
|
||||
if $(CPP) $< >$@ ; then :; else rm -f $@; fi
|
||||
@@ -1,680 +0,0 @@
|
||||
# This is the original manually generated Makefile. It may still be used
|
||||
# to build the collector.
|
||||
#
|
||||
# Primary targets:
|
||||
# gc.a - builds basic library
|
||||
# c++ - adds C++ interface to library
|
||||
# cords - adds cords (heavyweight strings) to library
|
||||
# test - prints porting information, then builds basic version of gc.a,
|
||||
# and runs some tests of collector and cords. Does not add cords or
|
||||
# c++ interface to gc.a
|
||||
# cord/de - builds dumb editor based on cords.
|
||||
ABI_FLAG=
|
||||
# ABI_FLAG should be the cc flag that specifies the ABI. On most
|
||||
# platforms this will be the empty string. Possible values:
|
||||
# +DD64 for 64-bit executable on HP/UX.
|
||||
# -n32, -n64, -o32 for SGI/MIPS ABIs.
|
||||
|
||||
AS_ABI_FLAG=$(ABI_FLAG)
|
||||
# ABI flag for assembler. On HP/UX this is +A64 for 64 bit
|
||||
# executables.
|
||||
|
||||
CC=cc $(ABI_FLAG)
|
||||
CXX=g++ $(ABI_FLAG)
|
||||
AS=as $(AS_ABI_FLAG)
|
||||
# The above doesn't work with gas, which doesn't run cpp.
|
||||
# Define AS as `gcc -c -x assembler-with-cpp' instead.
|
||||
|
||||
# Redefining srcdir allows object code for the nonPCR version of the collector
|
||||
# to be generated in different directories.
|
||||
srcdir= .
|
||||
VPATH= $(srcdir)
|
||||
|
||||
CFLAGS= -O -I$(srcdir)/include -DATOMIC_UNCOLLECTABLE -DNO_SIGNALS -DNO_EXECUTE_PERMISSION -DSILENT -DALL_INTERIOR_POINTERS
|
||||
|
||||
# To build the parallel collector on Linux, add to the above:
|
||||
# -DGC_LINUX_THREADS -DPARALLEL_MARK -DTHREAD_LOCAL_ALLOC
|
||||
# To build the parallel collector in a static library on HP/UX,
|
||||
# add to the above:
|
||||
# -DGC_HPUX_THREADS -DPARALLEL_MARK -DTHREAD_LOCAL_ALLOC -D_POSIX_C_SOURCE=199506L
|
||||
# To build the thread-safe collector on Tru64, add to the above:
|
||||
# -pthread -DGC_OSF1_THREADS
|
||||
|
||||
# HOSTCC and HOSTCFLAGS are used to build executables that will be run as
|
||||
# part of the build process, i.e. on the build machine. These will usually
|
||||
# be the same as CC and CFLAGS, except in a cross-compilation environment.
|
||||
# Note that HOSTCFLAGS should include any -D flags that affect thread support.
|
||||
HOSTCC=$(CC)
|
||||
HOSTCFLAGS=$(CFLAGS)
|
||||
|
||||
# For dynamic library builds, it may be necessary to add flags to generate
|
||||
# PIC code, e.g. -fPIC on Linux.
|
||||
|
||||
# Setjmp_test may yield overly optimistic results when compiled
|
||||
# without optimization.
|
||||
|
||||
# These define arguments influence the collector configuration:
|
||||
# -DSILENT disables statistics printing, and improves performance.
|
||||
# -DFIND_LEAK causes GC_find_leak to be initially set.
|
||||
# This causes the collector to assume that all inaccessible
|
||||
# objects should have been explicitly deallocated, and reports exceptions.
|
||||
# Finalization and the test program are not usable in this mode.
|
||||
# -DGC_SOLARIS_THREADS enables support for Solaris (thr_) threads.
|
||||
# (Clients should also define GC_SOLARIS_THREADS and then include
|
||||
# gc.h before performing thr_ or dl* or GC_ operations.)
|
||||
# Must also define -D_REENTRANT.
|
||||
# -DGC_SOLARIS_PTHREADS enables support for Solaris pthreads.
|
||||
# (Internally this define GC_SOLARIS_THREADS as well.)
|
||||
# -DGC_IRIX_THREADS enables support for Irix pthreads. See README.irix.
|
||||
# -DGC_HPUX_THREADS enables support for HP/UX 11 pthreads.
|
||||
# Also requires -D_REENTRANT or -D_POSIX_C_SOURCE=199506L. See README.hp.
|
||||
# -DGC_LINUX_THREADS enables support for Xavier Leroy's Linux threads.
|
||||
# see README.linux. -D_REENTRANT may also be required.
|
||||
# -DGC_OSF1_THREADS enables support for Tru64 pthreads. Untested.
|
||||
# -DGC_FREEBSD_THREADS enables support for FreeBSD pthreads. Untested.
|
||||
# Appeared to run into some underlying thread problems.
|
||||
# -DGC_DARWIN_THREADS enables support for Mac OS X pthreads. Untested.
|
||||
# -DGC_DGUX386_THREADS enables support for DB/UX on I386 threads.
|
||||
# See README.DGUX386.
|
||||
# -DGC_WIN32_THREADS enables support for win32 threads. That makes sense
|
||||
# for this Makefile only under Cygwin.
|
||||
# -DGC_THREADS should set the appropriate one of the above macros.
|
||||
# It assumes pthreads for Solaris.
|
||||
# -DALL_INTERIOR_POINTERS allows all pointers to the interior
|
||||
# of objects to be recognized. (See gc_priv.h for consequences.)
|
||||
# Alternatively, GC_all_interior_pointers can be set at process
|
||||
# initialization time.
|
||||
# -DSMALL_CONFIG tries to tune the collector for small heap sizes,
|
||||
# usually causing it to use less space in such situations.
|
||||
# Incremental collection no longer works in this case.
|
||||
# -DLARGE_CONFIG tunes the collector for unusually large heaps.
|
||||
# Necessary for heaps larger than about 500 MB on most machines.
|
||||
# Recommended for heaps larger than about 64 MB.
|
||||
# -DDONT_ADD_BYTE_AT_END is meaningful only with -DALL_INTERIOR_POINTERS or
|
||||
# GC_all_interior_pointers = 1. Normally -DALL_INTERIOR_POINTERS
|
||||
# causes all objects to be padded so that pointers just past the end of
|
||||
# an object can be recognized. This can be expensive. (The padding
|
||||
# is normally more than one byte due to alignment constraints.)
|
||||
# -DDONT_ADD_BYTE_AT_END disables the padding.
|
||||
# -DNO_SIGNALS does not disable signals during critical parts of
|
||||
# the GC process. This is no less correct than many malloc
|
||||
# implementations, and it sometimes has a significant performance
|
||||
# impact. However, it is dangerous for many not-quite-ANSI C
|
||||
# programs that call things like printf in asynchronous signal handlers.
|
||||
# This is on by default. Turning it off has not been extensively tested with
|
||||
# compilers that reorder stores. It should have been.
|
||||
# -DNO_EXECUTE_PERMISSION may cause some or all of the heap to not
|
||||
# have execute permission, i.e. it may be impossible to execute
|
||||
# code from the heap. Currently this only affects the incremental
|
||||
# collector on UNIX machines. It may greatly improve its performance,
|
||||
# since this may avoid some expensive cache synchronization.
|
||||
# -DGC_NO_OPERATOR_NEW_ARRAY declares that the C++ compiler does not support
|
||||
# the new syntax "operator new[]" for allocating and deleting arrays.
|
||||
# See gc_cpp.h for details. No effect on the C part of the collector.
|
||||
# This is defined implicitly in a few environments. Must also be defined
|
||||
# by clients that use gc_cpp.h.
|
||||
# -DREDIRECT_MALLOC=X causes malloc to be defined as alias for X.
|
||||
# Unless the following macros are defined, realloc is also redirected
|
||||
# to GC_realloc, and free is redirected to GC_free.
|
||||
# Calloc and strdup are redefined in terms of the new malloc. X should
|
||||
# be either GC_malloc or GC_malloc_uncollectable, or
|
||||
# GC_debug_malloc_replacement. (The latter invokes GC_debug_malloc
|
||||
# with dummy source location information, but still results in
|
||||
# properly remembered call stacks on Linux/X86 and Solaris/SPARC.
|
||||
# It requires that the following two macros also be used.)
|
||||
# The former is occasionally useful for working around leaks in code
|
||||
# you don't want to (or can't) look at. It may not work for
|
||||
# existing code, but it often does. Neither works on all platforms,
|
||||
# since some ports use malloc or calloc to obtain system memory.
|
||||
# (Probably works for UNIX, and win32.) If you build with DBG_HDRS_ALL,
|
||||
# you should only use GC_debug_malloc_replacement as a malloc
|
||||
# replacement.
|
||||
# -DREDIRECT_REALLOC=X causes GC_realloc to be redirected to X.
|
||||
# The canonical use is -DREDIRECT_REALLOC=GC_debug_realloc_replacement,
|
||||
# together with -DREDIRECT_MALLOC=GC_debug_malloc_replacement to
|
||||
# generate leak reports with call stacks for both malloc and realloc.
|
||||
# This also requires the following:
|
||||
# -DREDIRECT_FREE=X causes free to be redirected to X. The
|
||||
# canonical use is -DREDIRECT_FREE=GC_debug_free.
|
||||
# -DIGNORE_FREE turns calls to free into a noop. Only useful with
|
||||
# -DREDIRECT_MALLOC.
|
||||
# -DNO_DEBUGGING removes GC_dump and the debugging routines it calls.
|
||||
# Reduces code size slightly at the expense of debuggability.
|
||||
# -DJAVA_FINALIZATION makes it somewhat safer to finalize objects out of
|
||||
# order by specifying a nonstandard finalization mark procedure (see
|
||||
# finalize.c). Objects reachable from finalizable objects will be marked
|
||||
# in a sepearte postpass, and hence their memory won't be reclaimed.
|
||||
# Not recommended unless you are implementing a language that specifies
|
||||
# these semantics. Since 5.0, determines only only the initial value
|
||||
# of GC_java_finalization variable.
|
||||
# -DFINALIZE_ON_DEMAND causes finalizers to be run only in response
|
||||
# to explicit GC_invoke_finalizers() calls.
|
||||
# In 5.0 this became runtime adjustable, and this only determines the
|
||||
# initial value of GC_finalize_on_demand.
|
||||
# -DATOMIC_UNCOLLECTABLE includes code for GC_malloc_atomic_uncollectable.
|
||||
# This is useful if either the vendor malloc implementation is poor,
|
||||
# or if REDIRECT_MALLOC is used.
|
||||
# -DHBLKSIZE=ddd, where ddd is a power of 2 between 512 and 16384, explicitly
|
||||
# sets the heap block size. Each heap block is devoted to a single size and
|
||||
# kind of object. For the incremental collector it makes sense to match
|
||||
# the most likely page size. Otherwise large values result in more
|
||||
# fragmentation, but generally better performance for large heaps.
|
||||
# -DUSE_MMAP use MMAP instead of sbrk to get new memory.
|
||||
# Works for Solaris and Irix.
|
||||
# -DUSE_MUNMAP causes memory to be returned to the OS under the right
|
||||
# circumstances. This currently disables VM-based incremental collection.
|
||||
# This is currently experimental, and works only under some Unix,
|
||||
# Linux and Windows versions.
|
||||
# -DMMAP_STACKS (for Solaris threads) Use mmap from /dev/zero rather than
|
||||
# GC_scratch_alloc() to get stack memory.
|
||||
# -DPRINT_BLACK_LIST Whenever a black list entry is added, i.e. whenever
|
||||
# the garbage collector detects a value that looks almost, but not quite,
|
||||
# like a pointer, print both the address containing the value, and the
|
||||
# value of the near-bogus-pointer. Can be used to identifiy regions of
|
||||
# memory that are likely to contribute misidentified pointers.
|
||||
# -DKEEP_BACK_PTRS Add code to save back pointers in debugging headers
|
||||
# for objects allocated with the debugging allocator. If all objects
|
||||
# through GC_MALLOC with GC_DEBUG defined, this allows the client
|
||||
# to determine how particular or randomly chosen objects are reachable
|
||||
# for debugging/profiling purposes. The gc_backptr.h interface is
|
||||
# implemented only if this is defined.
|
||||
# -DGC_ASSERTIONS Enable some internal GC assertion checking. Currently
|
||||
# this facility is only used in a few places. It is intended primarily
|
||||
# for debugging of the garbage collector itself, but could also
|
||||
# -DDBG_HDRS_ALL Make sure that all objects have debug headers. Increases
|
||||
# the reliability (from 99.9999% to 100% mod. bugs) of some of the debugging
|
||||
# code (especially KEEP_BACK_PTRS). Makes -DSHORT_DBG_HDRS possible.
|
||||
# Assumes that all client allocation is done through debugging
|
||||
# allocators.
|
||||
# -DSHORT_DBG_HDRS Assume that all objects have debug headers. Shorten
|
||||
# the headers to minimize object size, at the expense of checking for
|
||||
# writes past the end of an object. This is intended for environments
|
||||
# in which most client code is written in a "safe" language, such as
|
||||
# Scheme or Java. Assumes that all client allocation is done using
|
||||
# the GC_debug_ functions, or through the macros that expand to these,
|
||||
# or by redirecting malloc to GC_debug_malloc_replacement.
|
||||
# (Also eliminates the field for the requested object size.)
|
||||
# occasionally be useful for debugging of client code. Slows down the
|
||||
# collector somewhat, but not drastically.
|
||||
# -DSAVE_CALL_COUNT=<n> Set the number of call frames saved with objects
|
||||
# allocated through the debugging interface. Affects the amount of
|
||||
# information generated in leak reports. Only matters on platforms
|
||||
# on which we can quickly generate call stacks, currently Linux/(X86 & SPARC)
|
||||
# and Solaris/SPARC and platforms that provide execinfo.h.
|
||||
# Default is zero. On X86, client
|
||||
# code should NOT be compiled with -fomit-frame-pointer.
|
||||
# -DSAVE_CALL_NARGS=<n> Set the number of functions arguments to be
|
||||
# saved with each call frame. Default is zero. Ignored if we
|
||||
# don't know how to retrieve arguments on the platform.
|
||||
# -DCHECKSUMS reports on erroneously clear dirty bits, and unexpectedly
|
||||
# altered stubborn objects, at substantial performance cost.
|
||||
# Use only for debugging of the incremental collector.
|
||||
# -DGC_GCJ_SUPPORT includes support for gcj (and possibly other systems
|
||||
# that include a pointer to a type descriptor in each allocated object).
|
||||
# Building this way requires an ANSI C compiler.
|
||||
# -DUSE_I686_PREFETCH causes the collector to issue Pentium III style
|
||||
# prefetch instructions. No effect except on X86 Linux platforms.
|
||||
# Assumes a very recent gcc-compatible compiler and assembler.
|
||||
# (Gas prefetcht0 support was added around May 1999.)
|
||||
# Empirically the code appears to still run correctly on Pentium II
|
||||
# processors, though with no performance benefit. May not run on other
|
||||
# X86 processors? In some cases this improves performance by
|
||||
# 15% or so.
|
||||
# -DUSE_3DNOW_PREFETCH causes the collector to issue AMD 3DNow style
|
||||
# prefetch instructions. Same restrictions as USE_I686_PREFETCH.
|
||||
# Minimally tested. Didn't appear to be an obvious win on a K6-2/500.
|
||||
# -DUSE_PPC_PREFETCH causes the collector to issue PowerPC style
|
||||
# prefetch instructions. No effect except on PowerPC OS X platforms.
|
||||
# Performance impact untested.
|
||||
# -DGC_USE_LD_WRAP in combination with the old flags listed in README.linux
|
||||
# causes the collector some system and pthread calls in a more transparent
|
||||
# fashion than the usual macro-based approach. Requires GNU ld, and
|
||||
# currently probably works only with Linux.
|
||||
# -DTHREAD_LOCAL_ALLOC defines GC_local_malloc(), GC_local_malloc_atomic()
|
||||
# and GC_local_gcj_malloc(). Needed for gc_gcj.h interface. These allocate
|
||||
# in a way that usually does not involve acquisition of a global lock.
|
||||
# Currently requires -DGC_LINUX_THREADS, but should be easy to port to
|
||||
# other pthreads environments. Recommended for multiprocessors.
|
||||
# -DUSE_COMPILER_TLS causes thread local allocation to use compiler-supported
|
||||
# "__thread" thread-local variables. This is the default in HP/UX. It
|
||||
# may help performance on recent Linux installations. (It failed for
|
||||
# me on RedHat 8, but appears to work on RedHat 9.)
|
||||
# -DPARALLEL_MARK allows the marker to run in multiple threads. Recommended
|
||||
# for multiprocessors. Currently requires Linux on X86 or IA64, though
|
||||
# support for other Posix platforms should be fairly easy to add,
|
||||
# if the thread implementation is otherwise supported.
|
||||
# -DNO_GETENV prevents the collector from looking at environment variables.
|
||||
# These may otherwise alter its configuration, or turn off GC altogether.
|
||||
# I don't know of a reason to disable this, except possibly if the
|
||||
# resulting process runs as a privileged user?
|
||||
# -DUSE_GLOBAL_ALLOC. Win32 only. Use GlobalAlloc instead of
|
||||
# VirtualAlloc to allocate the heap. May be needed to work around
|
||||
# a Windows NT/2000 issue. Incompatible with USE_MUNMAP.
|
||||
# See README.win32 for details.
|
||||
# -DMAKE_BACK_GRAPH. Enable GC_PRINT_BACK_HEIGHT environment variable.
|
||||
# See README.environment for details. Experimental. Limited platform
|
||||
# support. Implies DBG_HDRS_ALL. All allocation should be done using
|
||||
# the debug interface.
|
||||
# -DSTUBBORN_ALLOC allows allocation of "hard to change" objects, and thus
|
||||
# makes incremental collection easier. Was enabled by default until 6.0.
|
||||
# Rarely used, to my knowledge.
|
||||
# -DHANDLE_FORK attempts to make GC_malloc() work in a child process fork()ed
|
||||
# from a multithreaded parent. Currently only supported by pthread_support.c.
|
||||
# (Similar code should work on Solaris or Irix, but it hasn't been tried.)
|
||||
# -DTEST_WITH_SYSTEM_MALLOC causes gctest to allocate (and leak) large chunks
|
||||
# of memory with the standard system malloc. This will cause the root
|
||||
# set and collected heap to grow significantly if malloced memory is
|
||||
# somehow getting traced by the collector. This has no impact on the
|
||||
# generated library; it only affects the test.
|
||||
# -DPOINTER_MASK=0x... causes candidate pointers to be ANDed with the
|
||||
# given mask before being considered. If either this or the following
|
||||
# macro is defined, it will be assumed that all pointers stored in
|
||||
# the heap need to be processed this way. Stack and register pointers
|
||||
# will be considered both with and without processing.
|
||||
# These macros are normally needed only to support systems that use
|
||||
# high-order pointer tags. EXPERIMENTAL.
|
||||
# -DPOINTER_SHIFT=n causes the collector to left shift candidate pointers
|
||||
# by the indicated amount before trying to interpret them. Applied
|
||||
# after POINTER_MASK. EXPERIMENTAL. See also the preceding macro.
|
||||
#
|
||||
|
||||
CXXFLAGS= $(CFLAGS)
|
||||
AR= ar
|
||||
RANLIB= ranlib
|
||||
|
||||
|
||||
OBJS= alloc.o reclaim.o allchblk.o misc.o mach_dep.o os_dep.o mark_rts.o headers.o mark.o obj_map.o blacklst.o finalize.o new_hblk.o dbg_mlc.o malloc.o stubborn.o checksums.o solaris_threads.o aix_irix_threads.o pthread_support.o pthread_stop_world.o darwin_stop_world.o typd_mlc.o ptr_chck.o mallocx.o solaris_pthreads.o gcj_mlc.o specific.o gc_dlopen.o backgraph.o win32_threads.o
|
||||
|
||||
CSRCS= reclaim.c allchblk.c misc.c alloc.c mach_dep.c os_dep.c mark_rts.c headers.c mark.c obj_map.c pcr_interface.c blacklst.c finalize.c new_hblk.c real_malloc.c dyn_load.c dbg_mlc.c malloc.c stubborn.c checksums.c solaris_threads.c aix_irix_threads.c pthread_support.c pthread_stop_world.c darwin_stop_world.c typd_mlc.c ptr_chck.c mallocx.c solaris_pthreads.c gcj_mlc.c specific.c gc_dlopen.c backgraph.c win32_threads.c
|
||||
|
||||
CORD_SRCS= cord/cordbscs.c cord/cordxtra.c cord/cordprnt.c cord/de.c cord/cordtest.c include/cord.h include/ec.h include/private/cord_pos.h cord/de_win.c cord/de_win.h cord/de_cmds.h cord/de_win.ICO cord/de_win.RC
|
||||
|
||||
CORD_OBJS= cord/cordbscs.o cord/cordxtra.o cord/cordprnt.o
|
||||
|
||||
SRCS= $(CSRCS) mips_sgi_mach_dep.s rs6000_mach_dep.s alpha_mach_dep.S \
|
||||
sparc_mach_dep.S include/gc.h include/gc_typed.h \
|
||||
include/private/gc_hdrs.h include/private/gc_priv.h \
|
||||
include/private/gcconfig.h include/private/gc_pmark.h \
|
||||
include/gc_inl.h include/gc_inline.h include/gc_mark.h \
|
||||
threadlibs.c if_mach.c if_not_there.c gc_cpp.cc include/gc_cpp.h \
|
||||
gcname.c include/weakpointer.h include/private/gc_locks.h \
|
||||
gcc_support.c mips_ultrix_mach_dep.s include/gc_alloc.h \
|
||||
include/new_gc_alloc.h include/gc_allocator.h \
|
||||
include/javaxfc.h sparc_sunos4_mach_dep.s sparc_netbsd_mach_dep.s \
|
||||
include/private/solaris_threads.h include/gc_backptr.h \
|
||||
hpux_test_and_clear.s include/gc_gcj.h \
|
||||
include/gc_local_alloc.h include/private/dbg_mlc.h \
|
||||
include/private/specific.h powerpc_darwin_mach_dep.s \
|
||||
include/leak_detector.h include/gc_amiga_redirects.h \
|
||||
include/gc_pthread_redirects.h ia64_save_regs_in_stack.s \
|
||||
include/gc_config_macros.h include/private/pthread_support.h \
|
||||
include/private/pthread_stop_world.h include/private/darwin_semaphore.h \
|
||||
include/private/darwin_stop_world.h $(CORD_SRCS)
|
||||
|
||||
DOC_FILES= README.QUICK doc/README.Mac doc/README.MacOSX doc/README.OS2 \
|
||||
doc/README.amiga doc/README.cords doc/debugging.html \
|
||||
doc/README.dj doc/README.hp doc/README.linux doc/README.rs6000 \
|
||||
doc/README.sgi doc/README.solaris2 doc/README.uts \
|
||||
doc/README.win32 doc/barrett_diagram doc/README \
|
||||
doc/README.contributors doc/README.changes doc/gc.man \
|
||||
doc/README.environment doc/tree.html doc/gcdescr.html \
|
||||
doc/README.autoconf doc/README.macros doc/README.ews4800 \
|
||||
doc/README.DGUX386 doc/README.arm.cross doc/leak.html \
|
||||
doc/scale.html doc/gcinterface.html doc/README.darwin \
|
||||
doc/simple_example.html
|
||||
|
||||
TESTS= tests/test.c tests/test_cpp.cc tests/trace_test.c \
|
||||
tests/leak_test.c tests/thread_leak_test.c
|
||||
|
||||
GNU_BUILD_FILES= configure.in Makefile.am configure acinclude.m4 \
|
||||
libtool.m4 install-sh configure.host Makefile.in \
|
||||
aclocal.m4 config.sub config.guess \
|
||||
include/Makefile.am include/Makefile.in \
|
||||
doc/Makefile.am doc/Makefile.in \
|
||||
ltmain.sh mkinstalldirs depcomp missing
|
||||
|
||||
OTHER_MAKEFILES= OS2_MAKEFILE NT_MAKEFILE NT_THREADS_MAKEFILE gc.mak \
|
||||
BCC_MAKEFILE EMX_MAKEFILE WCC_MAKEFILE Makefile.dj \
|
||||
PCR-Makefile SMakefile.amiga Makefile.DLLs \
|
||||
digimars.mak Makefile.direct NT_STATIC_THREADS_MAKEFILE
|
||||
# Makefile and Makefile.direct are copies of each other.
|
||||
|
||||
OTHER_FILES= Makefile setjmp_t.c callprocs pc_excludes \
|
||||
MacProjects.sit.hqx MacOS.c \
|
||||
Mac_files/datastart.c Mac_files/dataend.c \
|
||||
Mac_files/MacOS_config.h Mac_files/MacOS_Test_config.h \
|
||||
add_gc_prefix.c gc_cpp.cpp \
|
||||
version.h AmigaOS.c \
|
||||
$(TESTS) $(GNU_BUILD_FILES) $(OTHER_MAKEFILES)
|
||||
|
||||
CORD_INCLUDE_FILES= $(srcdir)/include/gc.h $(srcdir)/include/cord.h \
|
||||
$(srcdir)/include/ec.h $(srcdir)/include/private/cord_pos.h
|
||||
|
||||
UTILS= if_mach if_not_there threadlibs
|
||||
|
||||
# Libraries needed for curses applications. Only needed for de.
|
||||
CURSES= -lcurses -ltermlib
|
||||
|
||||
# The following is irrelevant on most systems. But a few
|
||||
# versions of make otherwise fork the shell specified in
|
||||
# the SHELL environment variable.
|
||||
SHELL= /bin/sh
|
||||
|
||||
SPECIALCFLAGS = -I$(srcdir)/include
|
||||
# Alternative flags to the C compiler for mach_dep.c.
|
||||
# Mach_dep.c often doesn't like optimization, and it's
|
||||
# not time-critical anyway.
|
||||
# Set SPECIALCFLAGS to -q nodirect_code on Encore.
|
||||
|
||||
all: gc.a gctest
|
||||
|
||||
LEAKFLAGS=$(CFLAGS) -DFIND_LEAK
|
||||
|
||||
BSD-pkg-all: bsd-libgc.a bsd-libleak.a
|
||||
|
||||
bsd-libgc.a:
|
||||
$(MAKE) CFLAGS="$(CFLAGS)" clean c++-t
|
||||
mv gc.a bsd-libgc.a
|
||||
|
||||
bsd-libleak.a:
|
||||
$(MAKE) -f Makefile.direct CFLAGS="$(LEAKFLAGS)" clean c++-nt
|
||||
mv gc.a bsd-libleak.a
|
||||
|
||||
BSD-pkg-install: BSD-pkg-all
|
||||
${CP} bsd-libgc.a libgc.a
|
||||
${INSTALL_DATA} libgc.a ${PREFIX}/lib
|
||||
${INSTALL_DATA} gc.h gc_cpp.h ${PREFIX}/include
|
||||
${INSTALL_MAN} doc/gc.man ${PREFIX}/man/man3/gc.3
|
||||
|
||||
pcr: PCR-Makefile include/private/gc_private.h include/private/gc_hdrs.h \
|
||||
include/private/gc_locks.h include/gc.h include/private/gcconfig.h \
|
||||
mach_dep.o $(SRCS)
|
||||
$(MAKE) -f PCR-Makefile depend
|
||||
$(MAKE) -f PCR-Makefile
|
||||
|
||||
$(OBJS) tests/test.o dyn_load.o dyn_load_sunos53.o: \
|
||||
$(srcdir)/include/private/gc_priv.h \
|
||||
$(srcdir)/include/private/gc_hdrs.h $(srcdir)/include/private/gc_locks.h \
|
||||
$(srcdir)/include/gc.h $(srcdir)/include/gc_pthread_redirects.h \
|
||||
$(srcdir)/include/private/gcconfig.h $(srcdir)/include/gc_typed.h \
|
||||
$(srcdir)/include/gc_config_macros.h Makefile
|
||||
# The dependency on Makefile is needed. Changing
|
||||
# options such as -DSILENT affects the size of GC_arrays,
|
||||
# invalidating all .o files that rely on gc_priv.h
|
||||
|
||||
mark.o typd_mlc.o finalize.o ptr_chck.o: $(srcdir)/include/gc_mark.h $(srcdir)/include/private/gc_pmark.h
|
||||
|
||||
specific.o pthread_support.o: $(srcdir)/include/private/specific.h
|
||||
|
||||
solaris_threads.o solaris_pthreads.o: $(srcdir)/include/private/solaris_threads.h
|
||||
|
||||
dbg_mlc.o gcj_mlc.o: $(srcdir)/include/private/dbg_mlc.h
|
||||
|
||||
tests/test.o: tests $(srcdir)/tests/test.c
|
||||
$(CC) $(CFLAGS) -c $(srcdir)/tests/test.c
|
||||
mv test.o tests/test.o
|
||||
|
||||
tests:
|
||||
mkdir tests
|
||||
|
||||
base_lib gc.a: $(OBJS) dyn_load.o $(UTILS)
|
||||
echo > base_lib
|
||||
rm -f dont_ar_1
|
||||
./if_mach SPARC SUNOS5 touch dont_ar_1
|
||||
./if_mach SPARC SUNOS5 $(AR) rus gc.a $(OBJS) dyn_load.o
|
||||
./if_mach M68K AMIGA touch dont_ar_1
|
||||
./if_mach M68K AMIGA $(AR) -vrus gc.a $(OBJS) dyn_load.o
|
||||
./if_not_there dont_ar_1 $(AR) ru gc.a $(OBJS) dyn_load.o
|
||||
./if_not_there dont_ar_1 $(RANLIB) gc.a || cat /dev/null
|
||||
# ignore ranlib failure; that usually means it doesn't exist, and isn't needed
|
||||
|
||||
cords: $(CORD_OBJS) cord/cordtest $(UTILS)
|
||||
rm -f dont_ar_3
|
||||
./if_mach SPARC SUNOS5 touch dont_ar_3
|
||||
./if_mach SPARC SUNOS5 $(AR) rus gc.a $(CORD_OBJS)
|
||||
./if_mach M68K AMIGA touch dont_ar_3
|
||||
./if_mach M68K AMIGA $(AR) -vrus gc.a $(CORD_OBJS)
|
||||
./if_not_there dont_ar_3 $(AR) ru gc.a $(CORD_OBJS)
|
||||
./if_not_there dont_ar_3 $(RANLIB) gc.a || cat /dev/null
|
||||
|
||||
gc_cpp.o: $(srcdir)/gc_cpp.cc $(srcdir)/include/gc_cpp.h $(srcdir)/include/gc.h Makefile
|
||||
$(CXX) -c $(CXXFLAGS) $(srcdir)/gc_cpp.cc
|
||||
|
||||
test_cpp: $(srcdir)/tests/test_cpp.cc $(srcdir)/include/gc_cpp.h gc_cpp.o $(srcdir)/include/gc.h \
|
||||
base_lib $(UTILS)
|
||||
rm -f test_cpp
|
||||
./if_mach HP_PA HPUX $(CXX) $(CXXFLAGS) -o test_cpp $(srcdir)/tests/test_cpp.cc gc_cpp.o gc.a -ldld `./threadlibs`
|
||||
./if_not_there test_cpp $(CXX) $(CXXFLAGS) -o test_cpp $(srcdir)/tests/test_cpp.cc gc_cpp.o gc.a `./threadlibs`
|
||||
|
||||
c++-t: c++
|
||||
./test_cpp 1
|
||||
|
||||
c++-nt: c++
|
||||
@echo "Use ./test_cpp 1 to test the leak library"
|
||||
|
||||
c++: gc_cpp.o $(srcdir)/include/gc_cpp.h test_cpp
|
||||
rm -f dont_ar_4
|
||||
./if_mach SPARC SUNOS5 touch dont_ar_4
|
||||
./if_mach SPARC SUNOS5 $(AR) rus gc.a gc_cpp.o
|
||||
./if_mach M68K AMIGA touch dont_ar_4
|
||||
./if_mach M68K AMIGA $(AR) -vrus gc.a gc_cpp.o
|
||||
./if_not_there dont_ar_4 $(AR) ru gc.a gc_cpp.o
|
||||
./if_not_there dont_ar_4 $(RANLIB) gc.a || cat /dev/null
|
||||
./test_cpp 1
|
||||
echo > c++
|
||||
|
||||
dyn_load_sunos53.o: dyn_load.c
|
||||
$(CC) $(CFLAGS) -DSUNOS53_SHARED_LIB -c $(srcdir)/dyn_load.c -o $@
|
||||
|
||||
# SunOS5 shared library version of the collector
|
||||
sunos5gc.so: $(OBJS) dyn_load_sunos53.o
|
||||
$(CC) -G -o sunos5gc.so $(OBJS) dyn_load_sunos53.o -ldl
|
||||
ln sunos5gc.so libgc.so
|
||||
|
||||
# Alpha/OSF shared library version of the collector
|
||||
libalphagc.so: $(OBJS)
|
||||
ld -shared -o libalphagc.so $(OBJS) dyn_load.o -lc
|
||||
ln libalphagc.so libgc.so
|
||||
|
||||
# IRIX shared library version of the collector
|
||||
libirixgc.so: $(OBJS) dyn_load.o
|
||||
ld -shared $(ABI_FLAG) -o libirixgc.so $(OBJS) dyn_load.o -lc
|
||||
ln libirixgc.so libgc.so
|
||||
|
||||
# Linux shared library version of the collector
|
||||
liblinuxgc.so: $(OBJS) dyn_load.o
|
||||
gcc -shared -o liblinuxgc.so $(OBJS) dyn_load.o
|
||||
ln liblinuxgc.so libgc.so
|
||||
|
||||
# Alternative Linux rule. This is preferable, but is likely to break the
|
||||
# Makefile for some non-linux platforms.
|
||||
# LIBOBJS= $(patsubst %.o, %.lo, $(OBJS))
|
||||
#
|
||||
#.SUFFIXES: .lo $(SUFFIXES)
|
||||
#
|
||||
#.c.lo:
|
||||
# $(CC) $(CFLAGS) $(CPPFLAGS) -fPIC -c $< -o $@
|
||||
#
|
||||
# liblinuxgc.so: $(LIBOBJS) dyn_load.lo
|
||||
# gcc -shared -Wl,-soname=libgc.so.0 -o libgc.so.0 $(LIBOBJS) dyn_load.lo
|
||||
# touch liblinuxgc.so
|
||||
|
||||
mach_dep.o: $(srcdir)/mach_dep.c $(srcdir)/mips_sgi_mach_dep.s \
|
||||
$(srcdir)/mips_ultrix_mach_dep.s \
|
||||
$(srcdir)/rs6000_mach_dep.s $(srcdir)/powerpc_darwin_mach_dep.s \
|
||||
$(srcdir)/sparc_mach_dep.S $(srcdir)/sparc_sunos4_mach_dep.s \
|
||||
$(srcdir)/ia64_save_regs_in_stack.s \
|
||||
$(srcdir)/sparc_netbsd_mach_dep.s $(UTILS)
|
||||
rm -f mach_dep.o
|
||||
./if_mach MIPS IRIX5 $(CC) -c -o mach_dep.o $(srcdir)/mips_sgi_mach_dep.s
|
||||
./if_mach MIPS RISCOS $(AS) -o mach_dep.o $(srcdir)/mips_ultrix_mach_dep.s
|
||||
./if_mach MIPS ULTRIX $(AS) -o mach_dep.o $(srcdir)/mips_ultrix_mach_dep.s
|
||||
./if_mach POWERPC DARWIN $(AS) -o mach_dep.o $(srcdir)/powerpc_darwin_mach_dep.s
|
||||
./if_mach ALPHA LINUX $(CC) -c -o mach_dep.o $(srcdir)/alpha_mach_dep.S
|
||||
./if_mach SPARC SUNOS5 $(CC) -c -o mach_dep.o $(srcdir)/sparc_mach_dep.S
|
||||
./if_mach SPARC SUNOS4 $(AS) -o mach_dep.o $(srcdir)/sparc_sunos4_mach_dep.s
|
||||
./if_mach SPARC OPENBSD $(AS) -o mach_dep.o $(srcdir)/sparc_sunos4_mach_dep.s
|
||||
./if_mach SPARC NETBSD $(AS) -o mach_dep.o $(srcdir)/sparc_netbsd_mach_dep.s
|
||||
./if_mach IA64 "" as $(AS_ABI_FLAG) -o ia64_save_regs_in_stack.o $(srcdir)/ia64_save_regs_in_stack.s
|
||||
./if_mach IA64 "" $(CC) -c -o mach_dep1.o $(SPECIALCFLAGS) $(srcdir)/mach_dep.c
|
||||
./if_mach IA64 "" ld -r -o mach_dep.o mach_dep1.o ia64_save_regs_in_stack.o
|
||||
./if_not_there mach_dep.o $(CC) -c $(SPECIALCFLAGS) $(srcdir)/mach_dep.c
|
||||
|
||||
mark_rts.o: $(srcdir)/mark_rts.c $(UTILS)
|
||||
rm -f mark_rts.o
|
||||
-./if_mach ALPHA OSF1 $(CC) -c $(CFLAGS) -Wo,-notail $(srcdir)/mark_rts.c
|
||||
./if_not_there mark_rts.o $(CC) -c $(CFLAGS) $(srcdir)/mark_rts.c
|
||||
# Work-around for DEC optimizer tail recursion elimination bug.
|
||||
# The ALPHA-specific line should be removed if gcc is used.
|
||||
|
||||
alloc.o: version.h
|
||||
|
||||
cord:
|
||||
mkdir cord
|
||||
|
||||
cord/cordbscs.o: cord $(srcdir)/cord/cordbscs.c $(CORD_INCLUDE_FILES)
|
||||
$(CC) $(CFLAGS) -c -I$(srcdir) $(srcdir)/cord/cordbscs.c
|
||||
mv cordbscs.o cord/cordbscs.o
|
||||
# not all compilers understand -o filename
|
||||
|
||||
cord/cordxtra.o: cord $(srcdir)/cord/cordxtra.c $(CORD_INCLUDE_FILES)
|
||||
$(CC) $(CFLAGS) -c -I$(srcdir) $(srcdir)/cord/cordxtra.c
|
||||
mv cordxtra.o cord/cordxtra.o
|
||||
|
||||
cord/cordprnt.o: cord $(srcdir)/cord/cordprnt.c $(CORD_INCLUDE_FILES)
|
||||
$(CC) $(CFLAGS) -c -I$(srcdir) $(srcdir)/cord/cordprnt.c
|
||||
mv cordprnt.o cord/cordprnt.o
|
||||
|
||||
cord/cordtest: $(srcdir)/cord/cordtest.c $(CORD_OBJS) gc.a $(UTILS)
|
||||
rm -f cord/cordtest
|
||||
./if_mach SPARC DRSNX $(CC) $(CFLAGS) -o cord/cordtest $(srcdir)/cord/cordtest.c $(CORD_OBJS) gc.a -lucb
|
||||
./if_mach HP_PA HPUX $(CC) $(CFLAGS) -o cord/cordtest $(srcdir)/cord/cordtest.c $(CORD_OBJS) gc.a -ldld `./threadlibs`
|
||||
./if_mach M68K AMIGA $(CC) $(CFLAGS) -UGC_AMIGA_MAKINGLIB -o cord/cordtest $(srcdir)/cord/cordtest.c $(CORD_OBJS) gc.a `./threadlibs`
|
||||
./if_not_there cord/cordtest $(CC) $(CFLAGS) -o cord/cordtest $(srcdir)/cord/cordtest.c $(CORD_OBJS) gc.a `./threadlibs`
|
||||
|
||||
cord/de: $(srcdir)/cord/de.c cord/cordbscs.o cord/cordxtra.o gc.a $(UTILS)
|
||||
rm -f cord/de
|
||||
./if_mach SPARC DRSNX $(CC) $(CFLAGS) -o cord/de $(srcdir)/cord/de.c cord/cordbscs.o cord/cordxtra.o gc.a $(CURSES) -lucb `./threadlibs`
|
||||
./if_mach HP_PA HPUX $(CC) $(CFLAGS) -o cord/de $(srcdir)/cord/de.c cord/cordbscs.o cord/cordxtra.o gc.a $(CURSES) -ldld `./threadlibs`
|
||||
./if_mach RS6000 "" $(CC) $(CFLAGS) -o cord/de $(srcdir)/cord/de.c cord/cordbscs.o cord/cordxtra.o gc.a -lcurses
|
||||
./if_mach POWERPC DARWIN $(CC) $(CFLAGS) -o cord/de $(srcdir)/cord/de.c cord/cordbscs.o cord/cordxtra.o gc.a
|
||||
./if_mach I386 LINUX $(CC) $(CFLAGS) -o cord/de $(srcdir)/cord/de.c cord/cordbscs.o cord/cordxtra.o gc.a -lcurses `./threadlibs`
|
||||
./if_mach ALPHA LINUX $(CC) $(CFLAGS) -o cord/de $(srcdir)/cord/de.c cord/cordbscs.o cord/cordxtra.o gc.a -lcurses `./threadlibs`
|
||||
./if_mach IA64 LINUX $(CC) $(CFLAGS) -o cord/de $(srcdir)/cord/de.c cord/cordbscs.o cord/cordxtra.o gc.a -lcurses `./threadlibs`
|
||||
./if_mach M68K AMIGA $(CC) $(CFLAGS) -UGC_AMIGA_MAKINGLIB -o cord/de $(srcdir)/cord/de.c cord/cordbscs.o cord/cordxtra.o gc.a -lcurses
|
||||
./if_not_there cord/de $(CC) $(CFLAGS) -o cord/de $(srcdir)/cord/de.c cord/cordbscs.o cord/cordxtra.o gc.a $(CURSES) `./threadlibs`
|
||||
|
||||
if_mach: $(srcdir)/if_mach.c $(srcdir)/include/private/gcconfig.h
|
||||
$(HOSTCC) $(HOSTCFLAGS) -o if_mach $(srcdir)/if_mach.c
|
||||
|
||||
threadlibs: $(srcdir)/threadlibs.c $(srcdir)/include/private/gcconfig.h Makefile
|
||||
$(HOSTCC) $(HOSTCFLAGS) -o threadlibs $(srcdir)/threadlibs.c
|
||||
|
||||
if_not_there: $(srcdir)/if_not_there.c
|
||||
$(HOSTCC) $(HOSTCFLAGS) -o if_not_there $(srcdir)/if_not_there.c
|
||||
|
||||
clean:
|
||||
rm -f gc.a *.o *.exe tests/*.o gctest gctest_dyn_link test_cpp \
|
||||
setjmp_test mon.out gmon.out a.out core if_not_there if_mach \
|
||||
threadlibs $(CORD_OBJS) cord/cordtest cord/de
|
||||
-rm -f *~
|
||||
|
||||
gctest: tests/test.o gc.a $(UTILS)
|
||||
rm -f gctest
|
||||
./if_mach SPARC DRSNX $(CC) $(CFLAGS) -o gctest tests/test.o gc.a -lucb
|
||||
./if_mach HP_PA HPUX $(CC) $(CFLAGS) -o gctest tests/test.o gc.a -ldld `./threadlibs`
|
||||
./if_mach M68K AMIGA $(CC) $(CFLAGS) -UGC_AMIGA_MAKINGLIB -o gctest tests/test.o gc.a `./threadlibs`
|
||||
./if_not_there gctest $(CC) $(CFLAGS) -o gctest tests/test.o gc.a `./threadlibs`
|
||||
|
||||
# If an optimized setjmp_test generates a segmentation fault,
|
||||
# odds are your compiler is broken. Gctest may still work.
|
||||
# Try compiling setjmp_t.c unoptimized.
|
||||
setjmp_test: $(srcdir)/setjmp_t.c $(srcdir)/include/gc.h $(UTILS)
|
||||
$(CC) $(CFLAGS) -o setjmp_test $(srcdir)/setjmp_t.c
|
||||
|
||||
test: KandRtest cord/cordtest
|
||||
cord/cordtest
|
||||
|
||||
# Those tests that work even with a K&R C compiler:
|
||||
KandRtest: setjmp_test gctest
|
||||
./setjmp_test
|
||||
./gctest
|
||||
|
||||
add_gc_prefix: $(srcdir)/add_gc_prefix.c $(srcdir)/version.h
|
||||
$(CC) -o add_gc_prefix $(srcdir)/add_gc_prefix.c
|
||||
|
||||
gcname: $(srcdir)/gcname.c $(srcdir)/version.h
|
||||
$(CC) -o gcname $(srcdir)/gcname.c
|
||||
|
||||
gc.tar: $(SRCS) $(DOC_FILES) $(OTHER_FILES) add_gc_prefix gcname
|
||||
cp Makefile Makefile.old
|
||||
cp Makefile.direct Makefile
|
||||
rm -f `./gcname`
|
||||
ln -s . `./gcname`
|
||||
./add_gc_prefix $(SRCS) $(DOC_FILES) $(OTHER_FILES) > /tmp/gc.tar-files
|
||||
tar cvfh gc.tar `cat /tmp/gc.tar-files`
|
||||
cp gc.tar `./gcname`.tar
|
||||
gzip `./gcname`.tar
|
||||
rm `./gcname`
|
||||
|
||||
pc_gc.tar: $(SRCS) $(OTHER_FILES)
|
||||
tar cvfX pc_gc.tar pc_excludes $(SRCS) $(OTHER_FILES)
|
||||
|
||||
floppy: pc_gc.tar
|
||||
-mmd a:/cord
|
||||
-mmd a:/cord/private
|
||||
-mmd a:/include
|
||||
-mmd a:/include/private
|
||||
mkdir /tmp/pc_gc
|
||||
cat pc_gc.tar | (cd /tmp/pc_gc; tar xvf -)
|
||||
-mcopy -tmn /tmp/pc_gc/* a:
|
||||
-mcopy -tmn /tmp/pc_gc/cord/* a:/cord
|
||||
-mcopy -mn /tmp/pc_gc/cord/de_win.ICO a:/cord
|
||||
-mcopy -tmn /tmp/pc_gc/cord/private/* a:/cord/private
|
||||
-mcopy -tmn /tmp/pc_gc/include/* a:/include
|
||||
-mcopy -tmn /tmp/pc_gc/include/private/* a:/include/private
|
||||
rm -r /tmp/pc_gc
|
||||
|
||||
gc.tar.Z: gc.tar
|
||||
compress gc.tar
|
||||
|
||||
gc.tar.gz: gc.tar
|
||||
gzip gc.tar
|
||||
|
||||
lint: $(CSRCS) tests/test.c
|
||||
lint -DLINT $(CSRCS) tests/test.c | egrep -v "possible pointer alignment problem|abort|exit|sbrk|mprotect|syscall|change in ANSI|improper alignment"
|
||||
|
||||
# BTL: added to test shared library version of collector.
|
||||
# Currently works only under SunOS5. Requires GC_INIT call from statically
|
||||
# loaded client code.
|
||||
ABSDIR = `pwd`
|
||||
gctest_dyn_link: tests/test.o libgc.so
|
||||
$(CC) -L$(ABSDIR) -R$(ABSDIR) -o gctest_dyn_link tests/test.o -lgc -ldl -lthread
|
||||
|
||||
gctest_irix_dyn_link: tests/test.o libirixgc.so
|
||||
$(CC) -L$(ABSDIR) -o gctest_irix_dyn_link tests/test.o -lirixgc
|
||||
|
||||
# The following appear to be dead, especially since libgc_globals.h
|
||||
# is apparently lost.
|
||||
test_dll.o: tests/test.c libgc_globals.h
|
||||
$(CC) $(CFLAGS) -DGC_USE_DLL -c tests/test.c -o test_dll.o
|
||||
|
||||
test_dll: test_dll.o libgc_dll.a libgc.dll
|
||||
$(CC) test_dll.o -L$(ABSDIR) -lgc_dll -o test_dll
|
||||
|
||||
SYM_PREFIX-libgc=GC
|
||||
|
||||
# Uncomment the following line to build a GNU win32 DLL
|
||||
# include Makefile.DLLs
|
||||
|
||||
reserved_namespace: $(SRCS)
|
||||
for file in $(SRCS) tests/test.c tests/test_cpp.cc; do \
|
||||
sed s/GC_/_GC_/g < $$file > tmp; \
|
||||
cp tmp $$file; \
|
||||
done
|
||||
|
||||
user_namespace: $(SRCS)
|
||||
for file in $(SRCS) tests/test.c tests/test_cpp.cc; do \
|
||||
sed s/_GC_/GC_/g < $$file > tmp; \
|
||||
cp tmp $$file; \
|
||||
done
|
||||
@@ -1,430 +0,0 @@
|
||||
# This Makefile is intended only for DJGPP use.
|
||||
# It is mainly a copy of the main Makefile, but tends to get out of sync
|
||||
# with it. A merge would probably be appropriate.
|
||||
|
||||
# Primary targets:
|
||||
# gc.a - builds basic library
|
||||
# libgc.a - builds library for use with g++ "-fgc-keyword" extension
|
||||
# -fgc-keyword was never really available. Historical
|
||||
# interest only.
|
||||
# c++ - adds C++ interface to library
|
||||
# cords - adds cords (heavyweight strings) to library
|
||||
# test - prints porting information, then builds basic version of gc.a,
|
||||
# and runs some tests of collector and cords. Does not add cords or
|
||||
# c++ interface to gc.a
|
||||
# cord/de$(EXE_SUFFIX) - builds dumb editor based on cords.
|
||||
ABI_FLAG=
|
||||
CC=gcc $(ABI_FLAG)
|
||||
CXX=gxx $(ABI_FLAG)
|
||||
AS=gcc -c -x assembler-with-cpp $(ABI_FLAG)
|
||||
# The above doesn't work with gas, which doesn't run cpp.
|
||||
# Define AS as `gcc -c -x assembler-with-cpp' instead.
|
||||
# Under Irix 6, you will have to specify the ABI (-o32, -n32, or -64)
|
||||
# if you use something other than the default ABI on your machine.
|
||||
|
||||
# special defines for DJGPP
|
||||
CXXLD=gxx $(ABI_FLAG)
|
||||
EXE_SUFFIX=.exe
|
||||
|
||||
srcdir= .
|
||||
VPATH= $(srcdir)
|
||||
|
||||
CFLAGS= -gstabs+ -O2 -I$(srcdir)/include -DATOMIC_UNCOLLECTABLE -DNO_SIGNALS -DALL_INTERIOR_POINTERS -DNO_EXECUTE_PERMISSION -DSILENT
|
||||
|
||||
# Setjmp_test may yield overly optimistic results when compiled
|
||||
# without optimization.
|
||||
# -DSILENT disables statistics printing, and improves performance.
|
||||
# -DFIND_LEAK causes GC_find_leak to be initially set.
|
||||
# This causes the collector to assume that all inaccessible
|
||||
# objects should have been explicitly deallocated, and reports exceptions.
|
||||
# Finalization and the test program are not usable in this mode.
|
||||
# -DALL_INTERIOR_POINTERS allows all pointers to the interior
|
||||
# of objects to be recognized. (See gc_priv.h for consequences.)
|
||||
# -DSMALL_CONFIG tries to tune the collector for small heap sizes,
|
||||
# usually causing it to use less space in such situations.
|
||||
# Incremental collection no longer works in this case.
|
||||
# -DLARGE_CONFIG tunes the collector for unusually large heaps.
|
||||
# Necessary for heaps larger than about 500 MB on most machines.
|
||||
# Recommended for heaps larger than about 64 MB.
|
||||
# -DDONT_ADD_BYTE_AT_END is meaningful only with
|
||||
# -DALL_INTERIOR_POINTERS. Normally -DALL_INTERIOR_POINTERS
|
||||
# causes all objects to be padded so that pointers just past the end of
|
||||
# an object can be recognized. This can be expensive. (The padding
|
||||
# is normally more than one byte due to alignment constraints.)
|
||||
# -DDONT_ADD_BYTE_AT_END disables the padding.
|
||||
# -DNO_SIGNALS does not disable signals during critical parts of
|
||||
# the GC process. This is no less correct than many malloc
|
||||
# implementations, and it sometimes has a significant performance
|
||||
# impact. However, it is dangerous for many not-quite-ANSI C
|
||||
# programs that call things like printf in asynchronous signal handlers.
|
||||
# This is on by default. Turning it off has not been extensively tested with
|
||||
# compilers that reorder stores. It should have been.
|
||||
# -DNO_EXECUTE_PERMISSION may cause some or all of the heap to not
|
||||
# have execute permission, i.e. it may be impossible to execute
|
||||
# code from the heap. Currently this only affects the incremental
|
||||
# collector on UNIX machines. It may greatly improve its performance,
|
||||
# since this may avoid some expensive cache synchronization.
|
||||
# -DGC_NO_OPERATOR_NEW_ARRAY declares that the C++ compiler does not support
|
||||
# the new syntax "operator new[]" for allocating and deleting arrays.
|
||||
# See gc_cpp.h for details. No effect on the C part of the collector.
|
||||
# This is defined implicitly in a few environments. Must also be defined
|
||||
# by clients that use gc_cpp.h.
|
||||
# -DREDIRECT_MALLOC=X causes malloc, realloc, and free to be defined
|
||||
# as aliases for X, GC_realloc, and GC_free, respectively.
|
||||
# Calloc is redefined in terms of the new malloc. X should
|
||||
# be either GC_malloc or GC_malloc_uncollectable.
|
||||
# The former is occasionally useful for working around leaks in code
|
||||
# you don't want to (or can't) look at. It may not work for
|
||||
# existing code, but it often does. Neither works on all platforms,
|
||||
# since some ports use malloc or calloc to obtain system memory.
|
||||
# (Probably works for UNIX, and win32.)
|
||||
# -DIGNORE_FREE turns calls to free into a noop. Only useful with
|
||||
# -DREDIRECT_MALLOC.
|
||||
# -DNO_DEBUGGING removes GC_dump and the debugging routines it calls.
|
||||
# Reduces code size slightly at the expense of debuggability.
|
||||
# -DJAVA_FINALIZATION makes it somewhat safer to finalize objects out of
|
||||
# order by specifying a nonstandard finalization mark procedure (see
|
||||
# finalize.c). Objects reachable from finalizable objects will be marked
|
||||
# in a sepearte postpass, and hence their memory won't be reclaimed.
|
||||
# Not recommended unless you are implementing a language that specifies
|
||||
# these semantics. Since 5.0, determines only only the initial value
|
||||
# of GC_java_finalization variable.
|
||||
# -DFINALIZE_ON_DEMAND causes finalizers to be run only in response
|
||||
# to explicit GC_invoke_finalizers() calls.
|
||||
# In 5.0 this became runtime adjustable, and this only determines the
|
||||
# initial value of GC_finalize_on_demand.
|
||||
# -DATOMIC_UNCOLLECTABLE includes code for GC_malloc_atomic_uncollectable.
|
||||
# This is useful if either the vendor malloc implementation is poor,
|
||||
# or if REDIRECT_MALLOC is used.
|
||||
# -DHBLKSIZE=ddd, where ddd is a power of 2 between 512 and 16384, explicitly
|
||||
# sets the heap block size. Each heap block is devoted to a single size and
|
||||
# kind of object. For the incremental collector it makes sense to match
|
||||
# the most likely page size. Otherwise large values result in more
|
||||
# fragmentation, but generally better performance for large heaps.
|
||||
# -DPRINT_BLACK_LIST Whenever a black list entry is added, i.e. whenever
|
||||
# the garbage collector detects a value that looks almost, but not quite,
|
||||
# like a pointer, print both the address containing the value, and the
|
||||
# value of the near-bogus-pointer. Can be used to identifiy regions of
|
||||
# memory that are likely to contribute misidentified pointers.
|
||||
# -DKEEP_BACK_PTRS Add code to save back pointers in debugging headers
|
||||
# for objects allocated with the debugging allocator. If all objects
|
||||
# through GC_MALLOC with GC_DEBUG defined, this allows the client
|
||||
# to determine how particular or randomly chosen objects are reachable
|
||||
# for debugging/profiling purposes. The gc_backptr.h interface is
|
||||
# implemented only if this is defined.
|
||||
# -DGC_ASSERTIONS Enable some internal GC assertion checking. Currently
|
||||
# this facility is only used in a few places. It is intended primarily
|
||||
# for debugging of the garbage collector itself, but could also
|
||||
# -DDBG_HDRS_ALL Make sure that all objects have debug headers. Increases
|
||||
# the reliability (from 99.9999% to 100%) of some of the debugging
|
||||
# code (especially KEEP_BACK_PTRS). Makes -DSHORT_DBG_HDRS possible.
|
||||
# Assumes that all client allocation is done through debugging
|
||||
# allocators.
|
||||
# -DSHORT_DBG_HDRS Assume that all objects have debug headers. Shorten
|
||||
# the headers to minimize object size, at the expense of checking for
|
||||
# writes past the end of an object. This is intended for environments
|
||||
# in which most client code is written in a "safe" language, such as
|
||||
# Scheme or Java. Assumes that all client allocation is done using
|
||||
# the GC_debug_ functions (or through the macros that expand to these.
|
||||
# (Also eliminates the field for the requested object size.)
|
||||
# occasionally be useful for debugging of client code. Slows down the
|
||||
# collector somewhat, but not drastically.
|
||||
# -DCHECKSUMS reports on erroneously clear dirty bits, and unexpectedly
|
||||
# altered stubborn objects, at substantial performance cost.
|
||||
# Use only for debugging of the incremental collector.
|
||||
# -DGC_GCJ_SUPPORT includes support for gcj (and possibly other systems
|
||||
# that include a pointer to a type descriptor in each allocated object).
|
||||
# Building this way requires an ANSI C compiler.
|
||||
# -DUSE_I686_PREFETCH causes the collector to issue Pentium III style
|
||||
# prefetch instructions. No effect except on X86 Linux platforms.
|
||||
# Assumes a very recent gcc-compatible compiler and assembler.
|
||||
# (Gas prefetcht0 support was added around May 1999.)
|
||||
# Empirically the code appears to still run correctly on Pentium II
|
||||
# processors, though with no performance benefit. May not run on other
|
||||
# X86 processors? In some cases this improves performance by
|
||||
# 15% or so.
|
||||
# -DUSE_3DNOW_PREFETCH causes the collector to issue AMD 3DNow style
|
||||
# prefetch instructions. Same restrictions as USE_I686_PREFETCH.
|
||||
# UNTESTED!!
|
||||
# -DGC_USE_LD_WRAP in combination with the gld flags listed in README.linux
|
||||
# causes the collector some system and pthread calls in a more transparent
|
||||
# fashion than the usual macro-based approach. Requires GNU ld, and
|
||||
# currently probably works only with Linux.
|
||||
|
||||
|
||||
CXXFLAGS= $(CFLAGS) -DGC_OPERATOR_NEW_ARRAY
|
||||
AR= ar
|
||||
RANLIB= ranlib
|
||||
|
||||
|
||||
OBJS= alloc.o reclaim.o allchblk.o misc.o mach_dep.o os_dep.o mark_rts.o headers.o mark.o obj_map.o blacklst.o finalize.o new_hblk.o dbg_mlc.o malloc.o stubborn.o checksums.o solaris_threads.o typd_mlc.o ptr_chck.o mallocx.o solaris_pthreads.o gcj_mlc.o specific.o
|
||||
|
||||
CSRCS= reclaim.c allchblk.c misc.c alloc.c mach_dep.c os_dep.c mark_rts.c headers.c mark.c obj_map.c pcr_interface.c blacklst.c finalize.c new_hblk.c real_malloc.c dyn_load.c dbg_mlc.c malloc.c stubborn.c checksums.c solaris_threads.c typd_mlc.c ptr_chck.c mallocx.c solaris_pthreads.c gcj_mlc.c specific.c
|
||||
|
||||
CORD_SRCS= cord/cordbscs.c cord/cordxtra.c cord/cordprnt.c cord/de.c cord/cordtest.c include/cord.h include/ec.h include/private/cord_pos.h cord/de_win.c cord/de_win.h cord/de_cmds.h cord/de_win.ICO cord/de_win.RC cord/SCOPTIONS.amiga cord/SMakefile.amiga
|
||||
|
||||
CORD_OBJS= cord/cordbscs.o cord/cordxtra.o cord/cordprnt.o
|
||||
|
||||
SRCS= $(CSRCS) mips_sgi_mach_dep.S rs6000_mach_dep.s alpha_mach_dep.S \
|
||||
sparc_mach_dep.S include/gc.h include/gc_typed.h \
|
||||
include/private/gc_hdrs.h include/private/gc_priv.h \
|
||||
include/private/gcconfig.h include/private/gc_mark.h \
|
||||
include/gc_inl.h include/gc_inline.h gc.man \
|
||||
threadlibs.c if_mach.c if_not_there.c gc_cpp.cc include/gc_cpp.h \
|
||||
include/weakpointer.h include/private/gc_locks.h \
|
||||
gcc_support.c mips_ultrix_mach_dep.s include/gc_alloc.h \
|
||||
include/new_gc_alloc.h include/javaxfc.h sparc_sunos4_mach_dep.s \
|
||||
include/private/solaris_threads.h include/gc_backptr.h \
|
||||
hpux_test_and_clear.s include/gc_gcj.h \
|
||||
include/gc_local_alloc.h include/private/dbg_mlc.h \
|
||||
include/private/specific.h powerpc_darwin_mach_dep.s \
|
||||
include/leak_detector.h $(CORD_SRCS)
|
||||
|
||||
OTHER_FILES= Makefile PCR-Makefile OS2_MAKEFILE NT_MAKEFILE BCC_MAKEFILE \
|
||||
README tests/test.c test_cpp.cc setjmp_t.c SMakefile.amiga \
|
||||
SCoptions.amiga README.amiga README.win32 cord/README \
|
||||
README.rs6000 README.QUICK callprocs pc_excludes \
|
||||
barrett_diagram README.OS2 README.Mac MacProjects.sit.hqx \
|
||||
MacOS.c EMX_MAKEFILE README.debugging \
|
||||
Mac_files/datastart.c Mac_files/dataend.c \
|
||||
Mac_files/MacOS_config.h Mac_files/MacOS_Test_config.h \
|
||||
add_gc_prefix.c README.solaris2 README.sgi README.hp README.uts \
|
||||
win32_threads.c NT_THREADS_MAKEFILE gc.mak README.dj Makefile.dj \
|
||||
README.alpha README.linux README.MacOSX version.h Makefile.DLLs \
|
||||
WCC_MAKEFILE nursery.c include/gc_nursery.h include/gc_copy_descr.h
|
||||
|
||||
CORD_INCLUDE_FILES= $(srcdir)/include/gc.h $(srcdir)/include/cord.h \
|
||||
$(srcdir)/include/ec.h $(srcdir)/include/private/cord_pos.h
|
||||
|
||||
UTILS= if_mach$(EXE_SUFFIX) if_not_there$(EXE_SUFFIX)
|
||||
|
||||
# Libraries needed for curses applications. Only needed for de.
|
||||
CURSES= -lcurses -ltermlib
|
||||
|
||||
# The following is irrelevant on most systems. But a few
|
||||
# versions of make otherwise fork the shell specified in
|
||||
# the SHELL environment variable.
|
||||
SHELL= /bin/sh
|
||||
|
||||
SPECIALCFLAGS = -I$(srcdir)/include
|
||||
# Alternative flags to the C compiler for mach_dep.c.
|
||||
# Mach_dep.c often doesn't like optimization, and it's
|
||||
# not time-critical anyway.
|
||||
# Set SPECIALCFLAGS to -q nodirect_code on Encore.
|
||||
|
||||
all: gc.a gctest$(EXE_SUFFIX)
|
||||
|
||||
$(OBJS) test.o dyn_load.o dyn_load_sunos53.o: \
|
||||
$(srcdir)/include/private/gc_priv.h \
|
||||
$(srcdir)/include/private/gc_hdrs.h $(srcdir)/include/private/gc_locks.h \
|
||||
$(srcdir)/include/gc.h \
|
||||
$(srcdir)/include/private/gcconfig.h $(srcdir)/include/gc_typed.h \
|
||||
Makefile
|
||||
# The dependency on Makefile is needed. Changing
|
||||
# options such as -DSILENT affects the size of GC_arrays,
|
||||
# invalidating all .o files that rely on gc_priv.h
|
||||
|
||||
mark.o typd_mlc.o finalize.o: $(srcdir)/include/gc_mark.h
|
||||
|
||||
base_lib gc.a: $(OBJS) dyn_load.o $(UTILS)
|
||||
echo > base_lib
|
||||
rm -f on_sparc_sunos5_1
|
||||
./if_mach SPARC SUNOS5 touch on_sparc_sunos5_1
|
||||
./if_mach SPARC SUNOS5 $(AR) rus gc.a $(OBJS) dyn_load.o
|
||||
./if_not_there on_sparc_sunos5_1 $(AR) ru gc.a $(OBJS) dyn_load.o
|
||||
-./if_not_there on_sparc_sunos5_1 $(RANLIB) gc.a
|
||||
# ignore ranlib failure; that usually means it doesn't exist, and isn't needed
|
||||
|
||||
cords: $(CORD_OBJS) cord/cordtest$(EXE_SUFFIX) $(UTILS)
|
||||
rm -f on_sparc_sunos5_3
|
||||
./if_mach SPARC SUNOS5 touch on_sparc_sunos5_3
|
||||
./if_mach SPARC SUNOS5 $(AR) rus gc.a $(CORD_OBJS)
|
||||
./if_not_there on_sparc_sunos5_3 $(AR) ru gc.a $(CORD_OBJS)
|
||||
-./if_not_there on_sparc_sunos5_3 $(RANLIB) gc.a
|
||||
|
||||
gc_cpp.o: $(srcdir)/gc_cpp.cc $(srcdir)/include/gc_cpp.h $(srcdir)/include/gc.h Makefile
|
||||
$(CXX) -c $(CXXFLAGS) $(srcdir)/gc_cpp.cc
|
||||
|
||||
test_cpp$(EXE_SUFFIX): $(srcdir)/test_cpp.cc $(srcdir)/include/gc_cpp.h gc_cpp.o $(srcdir)/include/gc.h \
|
||||
base_lib $(UTILS)
|
||||
rm -f test_cpp test_cpp$(EXE_SUFFIX)
|
||||
./if_mach HP_PA "" $(CXX) $(CXXFLAGS) -o test_cpp $(srcdir)/test_cpp.cc gc_cpp.o gc.a -ldld
|
||||
./if_not_there test_cpp$(EXE_SUFFIX) $(CXXLD) $(CXXFLAGS) -o test_cpp$(EXE_SUFFIX) $(srcdir)/test_cpp.cc gc_cpp.o gc.a
|
||||
rm -f test_cpp
|
||||
|
||||
c++: gc_cpp.o $(srcdir)/include/gc_cpp.h test_cpp$(EXE_SUFFIX)
|
||||
rm -f on_sparc_sunos5_4
|
||||
./if_mach SPARC SUNOS5 touch on_sparc_sunos5_4
|
||||
./if_mach SPARC SUNOS5 $(AR) rus gc.a gc_cpp.o
|
||||
./if_not_there on_sparc_sunos5_4 $(AR) ru gc.a gc_cpp.o
|
||||
-./if_not_there on_sparc_sunos5_4 $(RANLIB) gc.a
|
||||
./test_cpp$(EXE_SUFFIX) 1
|
||||
echo > c++
|
||||
|
||||
dyn_load_sunos53.o: dyn_load.c
|
||||
$(CC) $(CFLAGS) -DSUNOS53_SHARED_LIB -c $(srcdir)/dyn_load.c -o $@
|
||||
|
||||
# SunOS5 shared library version of the collector
|
||||
sunos5gc.so: $(OBJS) dyn_load_sunos53.o
|
||||
$(CC) -G -o sunos5gc.so $(OBJS) dyn_load_sunos53.o -ldl
|
||||
ln sunos5gc.so libgc.so
|
||||
|
||||
# Alpha/OSF shared library version of the collector
|
||||
libalphagc.so: $(OBJS)
|
||||
ld -shared -o libalphagc.so $(OBJS) dyn_load.o -lc
|
||||
ln libalphagc.so libgc.so
|
||||
|
||||
# IRIX shared library version of the collector
|
||||
libirixgc.so: $(OBJS) dyn_load.o
|
||||
ld -shared $(ABI_FLAG) -o libirixgc.so $(OBJS) dyn_load.o -lc
|
||||
ln libirixgc.so libgc.so
|
||||
|
||||
# Linux shared library version of the collector
|
||||
liblinuxgc.so: $(OBJS) dyn_load.o
|
||||
gcc -shared -o liblinuxgc.so $(OBJS) dyn_load.o -lo
|
||||
ln liblinuxgc.so libgc.so
|
||||
|
||||
mach_dep.o: $(srcdir)/mach_dep.c $(srcdir)/mips_sgi_mach_dep.S $(srcdir)/mips_ultrix_mach_dep.s \
|
||||
$(srcdir)/rs6000_mach_dep.s $(srcdir)/powerpc_darwin_mach_dep.s $(UTILS)
|
||||
rm -f mach_dep.o
|
||||
./if_mach MIPS IRIX5 $(AS) -o mach_dep.o $(srcdir)/mips_sgi_mach_dep.S
|
||||
./if_mach MIPS RISCOS $(AS) -o mach_dep.o $(srcdir)/mips_ultrix_mach_dep.s
|
||||
./if_mach MIPS ULTRIX $(AS) -o mach_dep.o $(srcdir)/mips_ultrix_mach_dep.s
|
||||
./if_mach RS6000 "" $(AS) -o mach_dep.o $(srcdir)/rs6000_mach_dep.s
|
||||
./if_mach POWERPC MACOSX $(AS) -o mach_dep.o $(srcdir)/powerpc_darwin_mach_dep.s
|
||||
./if_mach ALPHA "" $(AS) -o mach_dep.o $(srcdir)/alpha_mach_dep.S
|
||||
./if_mach SPARC SUNOS5 $(AS) -o mach_dep.o $(srcdir)/sparc_mach_dep.S
|
||||
./if_mach SPARC SUNOS4 $(AS) -o mach_dep.o $(srcdir)/sparc_sunos4_mach_dep.s
|
||||
./if_not_there mach_dep.o $(CC) -c $(SPECIALCFLAGS) $(srcdir)/mach_dep.c
|
||||
|
||||
mark_rts.o: $(srcdir)/mark_rts.c if_mach if_not_there $(UTILS)
|
||||
rm -f mark_rts.o
|
||||
-./if_mach ALPHA OSF1 $(CC) -c $(CFLAGS) -Wo,-notail $(srcdir)/mark_rts.c
|
||||
./if_not_there mark_rts.o $(CC) -c $(CFLAGS) $(srcdir)/mark_rts.c
|
||||
# Work-around for DEC optimizer tail recursion elimination bug.
|
||||
# The ALPHA-specific line should be removed if gcc is used.
|
||||
|
||||
alloc.o: version.h
|
||||
|
||||
cord/cordbscs.o: $(srcdir)/cord/cordbscs.c $(CORD_INCLUDE_FILES)
|
||||
$(CC) $(CFLAGS) -c -I$(srcdir) $(srcdir)/cord/cordbscs.c
|
||||
mv cordbscs.o cord/cordbscs.o
|
||||
# not all compilers understand -o filename
|
||||
|
||||
cord/cordxtra.o: $(srcdir)/cord/cordxtra.c $(CORD_INCLUDE_FILES)
|
||||
$(CC) $(CFLAGS) -c -I$(srcdir) $(srcdir)/cord/cordxtra.c
|
||||
mv cordxtra.o cord/cordxtra.o
|
||||
|
||||
cord/cordprnt.o: $(srcdir)/cord/cordprnt.c $(CORD_INCLUDE_FILES)
|
||||
$(CC) $(CFLAGS) -c -I$(srcdir) $(srcdir)/cord/cordprnt.c
|
||||
mv cordprnt.o cord/cordprnt.o
|
||||
|
||||
cord/cordtest$(EXE_SUFFIX): $(srcdir)/cord/cordtest.c $(CORD_OBJS) gc.a $(UTILS) /tmp
|
||||
rm -f cord/cordtest$(EXE_SUFFIX)
|
||||
./if_mach SPARC DRSNX $(CC) $(CFLAGS) -o cord/cordtest$(EXE_SUFFIX) $(srcdir)/cord/cordtest.c $(CORD_OBJS) gc.a -lucb
|
||||
./if_mach HP_PA "" $(CC) $(CFLAGS) -o cord/cordtest$(EXE_SUFFIX) $(srcdir)/cord/cordtest.c $(CORD_OBJS) gc.a -ldld
|
||||
./if_not_there cord/cordtest$(EXE_SUFFIX) $(CC) $(CFLAGS) -o cord/cordtest $(srcdir)/cord/cordtest.c $(CORD_OBJS) gc.a
|
||||
rm -f cord/cordtest cordtest
|
||||
-mv cordtest$(EXE_SUFFIX) cord/
|
||||
|
||||
/tmp: $(UTILS)
|
||||
./if_not_there /tmp mkdir /tmp
|
||||
|
||||
cord/de$(EXE_SUFFIX): $(srcdir)/cord/de.c cord/cordbscs.o cord/cordxtra.o gc.a $(UTILS)
|
||||
rm -f cord/de cord/de$(EXE_SUFFIX)
|
||||
./if_mach SPARC DRSNX $(CC) $(CFLAGS) -o cord/de $(srcdir)/cord/de.c cord/cordbscs.o cord/cordxtra.o gc.a $(CURSES) -lucb `./threadlibs`
|
||||
./if_mach HP_PA "" $(CC) $(CFLAGS) -o cord/de $(srcdir)/cord/de.c cord/cordbscs.o cord/cordxtra.o gc.a $(CURSES) -ldld
|
||||
./if_mach RS6000 "" $(CC) $(CFLAGS) -o cord/de $(srcdir)/cord/de.c cord/cordbscs.o cord/cordxtra.o gc.a -lcurses
|
||||
./if_mach I386 LINUX $(CC) $(CFLAGS) -o cord/de $(srcdir)/cord/de.c cord/cordbscs.o cord/cordxtra.o gc.a -lcurses `./threadlibs`
|
||||
./if_mach ALPHA LINUX $(CC) $(CFLAGS) -o cord/de $(srcdir)/cord/de.c cord/cordbscs.o cord/cordxtra.o gc.a -lcurses
|
||||
./if_not_there cord/de$(EXE_SUFFIX) $(CC) $(CFLAGS) -o cord/de$(EXE_SUFFIX) $(srcdir)/cord/de.c cord/cordbscs.o cord/cordxtra.o gc.a $(CURSES)
|
||||
|
||||
if_mach$(EXE_SUFFIX): $(srcdir)/if_mach.c $(srcdir)/include/private/gcconfig.h
|
||||
rm -f if_mach if_mach$(EXE_SUFFIX)
|
||||
$(CC) $(CFLAGS) -o if_mach $(srcdir)/if_mach.c
|
||||
|
||||
threadlibs$(EXE_SUFFIX): $(srcdir)/threadlibs.c $(srcdir)include/private/gcconfig.h Makefile
|
||||
rm -f threadlibs threadlibs$(EXE_SUFFIX)
|
||||
$(CC) $(CFLAGS) -o threadlibs $(srcdir)/threadlibs.c
|
||||
|
||||
if_not_there$(EXE_SUFFIX): $(srcdir)/if_not_there.c
|
||||
rm -f if_not_there if_not_there$(EXE_SUFFIX)
|
||||
$(CC) $(CFLAGS) -o if_not_there $(srcdir)/if_not_there.c
|
||||
|
||||
# Clean removes *.o several times,
|
||||
# because as the first one doesn't seem to get them all!
|
||||
clean:
|
||||
rm -f gc.a *.o
|
||||
rm -f *.o
|
||||
rm -f *.o
|
||||
rm -f cord/*.o
|
||||
rm -f gctest gctest_dyn_link test_cpp
|
||||
rm -f setjmp_test mon.out gmon.out a.out core if_not_there if_mach
|
||||
rm -f threadlibs $(CORD_OBJS) cordtest cord/cordtest de cord/de
|
||||
rm -f gctest$(EXE_SUFFIX) gctest_dyn_link$(EXE_SUFFIX) test_cpp$(EXE_SUFFIX)
|
||||
rm -f setjmp_test$(EXE_SUFFIX) if_not_there$(EXE_SUFFIX) if_mach$(EXE_SUFFIX)
|
||||
rm -f threadlibs$(EXE_SUFFIX) cord/cordtest$(EXE_SUFFIX)
|
||||
-rm -f *~
|
||||
|
||||
gctest$(EXE_SUFFIX): tests/test.o gc.a if_mach$(EXE_SUFFIX) if_not_there$(EXE_SUFFIX)
|
||||
rm -f gctest gctest$(EXE_SUFFIX)
|
||||
./if_mach SPARC DRSNX $(CC) $(CFLAGS) -o gctest tests/test.o gc.a -lucb
|
||||
./if_mach HP_PA "" $(CC) $(CFLAGS) -o gctest tests/test.o gc.a -ldld
|
||||
./if_not_there gctest$(EXE_SUFFIX) $(CC) $(CFLAGS) -o gctest$(EXE_SUFFIX) tests/test.o gc.a
|
||||
rm -f gctest
|
||||
|
||||
# If an optimized setjmp_test generates a segmentation fault,
|
||||
# odds are your compiler is broken. Gctest may still work.
|
||||
# Try compiling setjmp_t.c unoptimized.
|
||||
setjmp_test$(EXE_SUFFIX): $(srcdir)/setjmp_t.c $(srcdir)/include/gc.h \
|
||||
if_mach$(EXE_SUFFIX) if_not_there$(EXE_SUFFIX)
|
||||
rm -f setjmp_test$(EXE_SUFFIX)
|
||||
$(CC) $(CFLAGS) -o setjmp_test $(srcdir)/setjmp_t.c
|
||||
rm -f setjmp_test
|
||||
|
||||
test: KandRtest cord/cordtest$(EXE_SUFFIX)
|
||||
./cord/cordtest$(EXE_SUFFIX)
|
||||
|
||||
# Those tests that work even with a K&R C compiler:
|
||||
KandRtest: setjmp_test$(EXE_SUFFIX) gctest$(EXE_SUFFIX)
|
||||
./setjmp_test$(EXE_SUFFIX)
|
||||
./gctest$(EXE_SUFFIX)
|
||||
|
||||
add_gc_prefix$(EXE_SUFFIX): add_gc_prefix.c
|
||||
$(CC) -o add_gc_prefix$(EXE_SUFFIX) $(srcdir)/add_gc_prefix.c
|
||||
rm -f add_gc_prefix
|
||||
|
||||
gc.tar: $(SRCS) $(OTHER_FILES) add_gc_prefix
|
||||
./add_gc_prefix$(EXE_SUFFIX) $(SRCS) $(OTHER_FILES) > /tmp/gc.tar-files
|
||||
(cd $(srcdir)/.. ; tar cvfh - `cat /tmp/gc.tar-files`) > gc.tar
|
||||
|
||||
pc_gc.tar: $(SRCS) $(OTHER_FILES)
|
||||
tar cvfX pc_gc.tar pc_excludes $(SRCS) $(OTHER_FILES)
|
||||
|
||||
gc.tar.Z: gc.tar
|
||||
compress gc.tar
|
||||
|
||||
gc.tar.gz: gc.tar
|
||||
gzip gc.tar
|
||||
|
||||
lint: $(CSRCS) tests/test.c
|
||||
lint -DLINT $(CSRCS) tests/test.c | egrep -v "possible pointer alignment problem|abort|exit|sbrk|mprotect|syscall"
|
||||
|
||||
# BTL: added to test shared library version of collector.
|
||||
# Currently works only under SunOS5. Requires GC_INIT call from statically
|
||||
# loaded client code.
|
||||
ABSDIR = `pwd`
|
||||
gctest_dyn_link: test.o libgc.so
|
||||
$(CC) -L$(ABSDIR) -R$(ABSDIR) -o gctest_dyn_link test.o -lgc -ldl -lthread
|
||||
|
||||
test_dll.o: tests/test.c libgc_globals.h
|
||||
$(CC) $(CFLAGS) -DGC_USE_DLL -c tests/test.c -o test_dll.o
|
||||
|
||||
test_dll: test_dll.o libgc_dll.a libgc.dll
|
||||
$(CC) test_dll.o -L$(ABSDIR) -lgc_dll -o test_dll
|
||||
|
||||
SYM_PREFIX-libgc=GC
|
||||
|
||||
# Uncomment the following line to build a GNU win32 DLL
|
||||
# include Makefile.DLLs
|
||||
|
||||