Commit Graph

829 Commits

Author SHA1 Message Date
Wes Kocher
baf6cddc4c Merge inbound to central, a=merge
MozReview-Commit-ID: 9NFjSEt96iT
2017-08-02 17:11:51 -07:00
Nicholas Nethercote
406a9ed175 Bug 1384819 (part 3) - Remove the return value from the stack walker functions. r=glandium.
Just one caller (in DMD) actually looks at it, and that's in an unimportant way
-- if the return value was false, mLength would be zero anyway.

--HG--
extra : rebase_source : 0463ab3765744742a9e854964342d631095fa55f
2017-07-27 16:46:55 +10:00
Nicholas Nethercote
8a20eb82d2 Bug 1384819 (part 2) - Tweak FramePointerStackWalk() arguments. r=glandium.
This patch does he following.

- Avoids some unnecessary casting.

- Renames the |bp| parameter as |aBp|.

- Makes the no-op FramePointerStackWalk() signature match the real one.
  (Clearly it's dead code in all built configurations!)

--HG--
extra : rebase_source : 3fe606d1ff9b063294f4028ff884c20661ed9e0a
2017-07-27 16:46:33 +10:00
Nicholas Nethercote
08e54b7c13 Bug 1384819 (part 1) - Split MozStackWalk(). r=glandium.
MozStackWalk() is different on Windows to the other platforms. It has two extra
arguments, which can be used to walk the stack of a different thread.

This patch makes those differences clearer. Instead of having a single function
and forbidding those two arguments on non-Windows, it removes those arguments
from MozStackWalk, and splits off MozStackWalkThread() which retains them. This
also allows those arguments to have more appropriate types (HANDLE instead of
uintptr_t; CONTEXT* instead of than void*) and names (aContext instead of
aPlatformData).

The patch also removes unnecessary reinterpret_casts for the aClosure argument
at a couple of MozStackWalk() callsites.

--HG--
extra : rebase_source : 111ab7d6426d7be921facc2264f6db86c501d127
2017-07-27 12:46:47 +10:00
Botond Ballo
b93629ee1d Bug 1383919 - Use |noexcept| instead of an exception-specification in mozalloc.h. r=glandium
MozReview-Commit-ID: DwrE8cmZOQN

--HG--
extra : rebase_source : e45f35dd009a3a8a94f50afc55177976c532f16a
2017-07-24 19:19:07 -04:00
Nicholas Nethercote
f7a538edd4 Bug 1382076 - Use FramePointerStackWalk() in DMD on Win32. r=erahm.
The patch also uses a better value for skipFrames on Win64.

--HG--
extra : rebase_source : ef3d6d68fbea45e7fd571db2ba78904d97f5e21e
2017-07-20 11:33:31 +10:00
Nicholas Nethercote
7d699f9fa4 Bug 1382099 - Remove MOZ_WIDGET_GONK from memory/. r=glandium.
--HG--
extra : rebase_source : 641de85945cb2f0ccdb9d6f5207bdd22b597b318
2017-07-19 18:39:00 +10:00
Mike Hommey
4956a3cafa Bug 1379897 - Remove opt_abort. r=njn
It causes abort() on errors from munmap/VirtualFree (which in practice
don't happen except maybe in case of memory corruption), and was only
enabled on debug builds.

--HG--
extra : rebase_source : fb0c8c82dc1e2d1f14a588a82144cf82e4f4f773
2017-07-11 14:24:35 +09:00
Carsten "Tomcat" Book
31311070d9 merge mozilla-inbound to mozilla-central a=merge 2017-07-11 12:51:59 +02:00
Mike Hommey
1246710041 Bug 1379890 - Remove unused mozjemalloc stats. r=njn
Since bug 1378258 remove malloc_print_stats, there are a bunch of
allocator stats that are now unused, reducing the memory footprint of
allocator metadata.

--HG--
extra : rebase_source : 337ef3b647c20119334b6576d591006f6bb3dd16
2017-07-11 14:42:57 +09:00
Mike Hommey
1ecc9df3fc Bug 1379878 - Use C++ templates to make _malloc_message calls cleaner. r=njn
--HG--
extra : rebase_source : 695fd9ce6b1f216a8bda29b849e1372601a43fe4
2017-07-11 14:15:04 +09:00
Mike Hommey
26fb8eb7b0 Bug 1379878 - Remove umax2s, its last use was removed in bug 1378258. r=njn
--HG--
extra : rebase_source : d9cb0631662ec0015e9c6d268884adb31c5a38bb
2017-07-11 14:09:09 +09:00
Mike Hommey
860374cbca Bug 1360772 - Allow to initialize non-zeroed arena chunks. r=njn
When initializing a new chunk for use as an arena, we started by zeroing
out the chunk (if that wasn't the case) and then initializing a new
arena chunk in there. It turns out this can have a noticeable overhead,
especially when e.g. the new arena chunk is used for a large allocation
filled out by something that is realloc()ated.

OTOH, the chunk recycle code only ever keeps zeroed or arena chunks
around (there is a "recycled" type too, but in practice, at the moment,
this means they were arena chunks before). Arena chunks that were
recycled were totally emptied, so all the runs they may contain will
contain zeroed-out or poisoned data. They also contain a header, that is
overwritten by the new arena chunk initialization.

This means we can get away with reusing non-zeroed recycled chunks
without zeroing them, as long as the arena chunk header marks the runs
as madvised instead of zeroed.

Code-wise, this would benefit from getting a ChunkType out of
chunk_alloc, but this would require more refactoring than I'm willing to
do at the moment.
2017-07-11 13:55:26 +09:00
Mike Hommey
e6c1516607 Bug 1360772 - Indicate to chunk_recycle callers that pages_commit has zeroed the chunk. r=njn
Before returning a chunk, chunk_recycle calls pages_commit (when
MALLOC_DECOMMIT is enabled), which is guaranteed to zero the chunk.

The code further zeroing the chunk afterwards, which is now moved out to
chunk_alloc callers, never took advantage of that fact, duplicating the
effort of zeroing the chunk on Windows.

By indicating to the callers that the chunk has already been zeroed, we
allow callers to skip zeroing on their own.
2017-07-11 13:44:40 +09:00
Mike Hommey
78a584db8a Bug 1360772 - Make chunk_alloc callers handle zeroing when they need it. r=njn
The current code only allows chunk_calloc() callers to tell whether they
want zeroed memory or not, but some might be okay either way, assuming
they act accordingly afterwards. So move the zeroing out of chunk_alloc.
2017-07-11 13:44:39 +09:00
Mike Hommey
f26f174a0e Bug 1360772 - Store the chunk type in chunk recycling queue. r=njn 2017-07-11 13:44:21 +09:00
Mike Hommey
0380bc7b32 Bug 1360772 - Make pages_purge return whether the pages were purged. r=njn
Many functions in the mozjemalloc codebase like to return the opposite
boolean one would tend to expect. Pages_purge is one of them, and this
reverses the logic to match expectations.

Also make it static.
2017-07-11 13:11:55 +09:00
Mike Hommey
bf7a9e79f6 Bug 1379139 - Instead of not recycling huge chunks, zero them. r=njn
It turns out that not recycling some kinds of chunk can lead to the
recycle queue being starved in some scenarios. When that happens, we end
up mmap()ing new memory, but that turns out to be significantly slower.

So instead of not recycling huge chunks, we force-clean them, before
madvising so that the pages can still be reclaimed in case of memory
pressure.

--HG--
extra : rebase_source : 2dbd028daca92c9cd7c8079eb3dc5a0cfa06495b
2017-07-10 06:41:17 +09:00
Nicholas Nethercote
4325ae218a Bug 1371397 - Use FramePointerStackWalk() in DMD on Mac. r=erahm.
This avoids MozStackWalk(), which has become unusably slow on Mac due to
changes in libunwind, and gets us back to decent speed.

The code for getting the frame pointer and stack end was copied from the Gecko
Profiler, which also uses FramePointerStackWalk() on Mac.

--HG--
extra : rebase_source : 58c32c2df8716c7c8123a4a8fb692182d066caca
2017-07-07 07:08:08 +10:00
Carsten "Tomcat" Book
b5c809a8bc merge mozilla-inbound to mozilla-central a=merge 2017-07-07 10:35:44 +02:00
Mike Hommey
19448fe709 Bug 1356701 - Export unprefixed malloc and duplication functions on OSX. r=njn
Going through the system zone allocator for every call to realloc/free
on OSX is costly, because the zone allocator needs to first verify that
the allocations do belong to the allocator it invokes (which ends up
calling jemalloc's malloc_usable_size), which is unnecessary when we
expect the allocations to belong to jemalloc.

So, we export the malloc/realloc/free/etc. symbols from
libmozglue.dylib, such that libraries and programs linked against it
call directly into jemalloc instead of going through the system zone
allocator, effectively shortcutting the allocator verification.

The risk is that some things in Gecko try to realloc/free pointers it
got from system libraries, if those were allocated with a system zone
that is not jemalloc.

--HG--
extra : rebase_source : ee0b29e1275176f52e64f4648dfa7ce25d61292e
2017-07-04 15:01:50 +09:00
Mike Hommey
ed863255b4 Bug 1378592 - Remove MOZ_REPLACE_MALLOC_LINKAGE. r=froydnj
MOZ_REPLACE_MALLOC_LINKAGE was added back when there were problems with
getting weak references working properly for replace-malloc.

Versions of OSX < 10.6 needed flat namespace, but aren't supported
anymore.

Versions of Xcode < 4.5 required flat namespace + a dummy library in
order to produce proper weak references. There is virtually nobody still
building with such an ancient toolchain.

Keeping those around doesn't /really/ hurt, except recent versions of
Xcode don't expose dyldinfo in /usr/bin, used for the configure test.
Consequently, MOZ_REPLACE_MALLOC_LINKAGE ended up being set to use the
dummy library setup, which, by using flat namespace, now causes harm in
bug 1356701, causing bug 1378332.

--HG--
extra : rebase_source : e3edc1f2cf905943c33fafeb631f2f88fc87167e
2017-07-06 10:26:04 +09:00
Mike Hommey
a02826e8ee Bug 1378278 - Use nullptr in mozjemalloc. r=njn
At the same time:
- replace "(nullptr)" with "nullptr",
- replace "x == nullptr" with "!x",
- replace "x != nullptr" with "x".

--HG--
extra : rebase_source : 5bc5577d0de53244afae9ebadb84962f38306368
2017-07-05 14:02:50 +09:00
Ryan VanderMeulen
5e1e8d2f24 Backed out changeset 261d61f4eeec (bug 1356701) for causing bug 1378339. 2017-07-05 12:58:39 -04:00
Mike Hommey
230dccc3d5 Bug 1378258 - Remove malloc_print_stats. r=njn
malloc_print_stats is designed as a end-of-process dump of malloc stats,
but it has limited value in Firefox itself. It has some usefulness as
something to invoke from a debugger while Firefox is running (after
manually setting opt_print_stats to true), but it also lacks information
about live allocations, which would make it more useful.

All in all, if we want some stats about jemalloc allocations in a more
useful way, we'd be better off augmenting the jemalloc_stats() API and
make the data available to e.g. about:memory, than keeping
malloc_print_stats.

--HG--
extra : rebase_source : f6fbc97dfddd52c8620e37e2e189ae0087a845b0
2017-07-05 10:07:43 +09:00
Mike Hommey
c65bedc1c8 Bug 1356701 - Export unprefixed malloc and duplication functions on OSX. r=njn
Going through the system zone allocator for every call to realloc/free
on OSX is costly, because the zone allocator needs to first verify that
the allocations do belong to the allocator it invokes (which ends up
calling jemalloc's malloc_usable_size), which is unnecessary when we
expect the allocations to belong to jemalloc.

So, we export the malloc/realloc/free/etc. symbols from
libmozglue.dylib, such that libraries and programs linked against it
calls directly into jemalloc instead of going through the system zone
allocator, effectively shortcutting the allocator verification.

The risk is that some things in Gecko try to realloc/free pointers it
got from system libraries, if those were allocated with a system zone
that is not jemalloc.

--HG--
extra : rebase_source : 45b9b98499760a7f946878d41d2fdaadb6dff4d6
2017-07-04 15:01:50 +09:00
Petr Sumbera
a393f2a6cf Bug 1375467 - Redefining abort() in C++ requires extern "C". r=glandium 2017-06-22 05:09:05 -07:00
Jeff Gilbert
58ae26fbc7 Bug 1376057 - Replace MSVC wrappers with std::exception::_Set_raise_handler. - r=froydnj
MozReview-Commit-ID: MG5c4bzDlI
2017-06-27 11:51:34 -07:00
Mike Hommey
7ccc8afe84 Bug 1378658 - When page_purges doesn't zero pages, only record arena chunks or previously recycled chunks for future recycling. r=njn 2017-07-07 09:18:03 +09:00
Mike Hommey
29524a9451 Bug 1369626 - Remove MALLOC_XMALLOC. r=njn
Setting MALLOC_XMALLOC enables a runtime toggle that allows to make
allocations abort() on OOM instead of returning NULL. In other words,
it enables a toggle that allows to turn all allocations into infallible
allocations.

The toggle however still defaults to being disabled, which means even
when MALLOC_XMALLOC is defined when building mozjemalloc, there is no
change in behavior, unless a MALLOC_OPTIONS is set.

Even if this were useful to anyone (MALLOC_XMALLOC is only defined on
debug builds, limiting the usefulness), this is something
replace-malloc, in Firefox, is meant to be used for.

So let's remove this feature, and possibly add an equivalent
replace-malloc later if deemed necessary.

--HG--
extra : rebase_source : 1ccc893e9a8e842c3fa90e91f076a528df2f7dfe
2017-06-02 15:19:20 +09:00
Mike Hommey
1b12c351f2 Bug 1368932 - Add a testcase for a replace-malloc library that doesn't implement all functions. r=njn
--HG--
rename : memory/replace/logalloc/moz.build => memory/replace/logalloc/logalloc.mozbuild
rename : memory/replace/logalloc/replay/replay.log => memory/replace/logalloc/replay/expected_output_minimal.log
extra : rebase_source : 94f8fb4444e2e0ecda61e5ffd283ab560d1e9dfe
2017-05-31 15:04:32 +09:00
Mike Hommey
4ae1509807 Bug 1368932 - Handle missing replace_posix_memalign at the replace-malloc level. r=njn
Replace-malloc libraries, such as DMD, don't really need to care about
the details of implementing all the variants of aligned memory
allocation functions. Currently, by defining MOZ_REPLACE_ONLY_MEMALIGN
before including replace_malloc.h, they get predefined functions.

Instead of making that an opt-in at build time, we make the
replace-malloc initialization just fill the replace-malloc
malloc_table_t with implementations that rely on the replace_memalign
the library provides.

--HG--
extra : rebase_source : 0842a67d9bc27a9a86c33d14d98b9c25f39982fb
2017-05-31 13:47:17 +09:00
Mike Hommey
7fd2c94e79 Bug 1368932 - Move the replace_malloc_init_funcs function around. r=njn
--HG--
extra : rebase_source : 4e3b3bbd18a37a3ee01d6ec30449249be5481b77
2017-05-31 13:47:11 +09:00
Mike Hommey
565f74102e Bug 1368932 - Fill the replace-malloc malloc_table_t with the real allocator as a fallback. r=njn
Until now, the malloc implementation functions would call the
replace-malloc functions if they exist, and fallback to the real
allocator in no such function exists. Instead of doing this, we now
fill the empty slots in the malloc_table_t with the real allocator
functions.

--HG--
extra : rebase_source : b54634f23188906939e4dc01fc5a3007de0f3f2c
2017-05-30 15:57:28 +09:00
Mike Hommey
1a6ac2f8fb Bug 1368932 - Use a malloc_table_t for most replace-malloc function pointers, on all platforms. r=njn
We make replace_malloc_init_funcs called on all platforms and fill out a
malloc_table_t for the replace-malloc functions with what comes from
dlsym/GetProcAddress on Android/Windows, and from the dynamically linked
weak symbols replace_* on other platforms.

replace_malloc.h contains definitions of *_impl_t types for each of the
functions in the malloc_table_t, which is redundant with the
replace_*_impl_t types we were creating, so we remove those typedefs,
except for the two functions (init and get_bridge) that don't have such
a typedef. Those functions don't appear in malloc_table_t.

--HG--
extra : rebase_source : 3705a99ee07f63dbaa66973eef19ddab224e0911
2017-05-30 15:57:28 +09:00
Mike Hommey
409495f283 Bug 1368932 - Refactor such that there is only one definition of replace_malloc_init_funcs. r=njn
We want, in a subsequent patch, to have replace_malloc_init_funcs be
called on all platforms (including those relying on the replace-malloc
library being loaded already) and perform more initialization.

To prepare for that, we move the non-platform-specific pieces out.

--HG--
extra : rebase_source : 239ed363ee168bf4f8a96e0a1ca52981cb941b71
2017-05-30 15:57:28 +09:00
Mike Hommey
37b66db19d Bug 1368932 - Generate all the _impl functions with macros in replace-malloc. r=njn
All the _impl functions in replace-malloc.c are largely identical. This
replaces all of them with macro expansions.

--HG--
extra : rebase_source : 67a1809b0b0fc4645ea5041154fa3a6dcb6cce6b
2017-05-25 16:47:57 +09:00
Mike Hommey
e4fddc104f Bug 1368932 - Add argument names to malloc implementation declarations in replace-malloc. r=njn
This transforms the declarations from e.g.:
    void *realloc(void *, size_t);
into:
    void *realloc(void *arg1, size_t arg2);

--HG--
extra : rebase_source : 627eb833a98d11f044c776f9d29715ccad31174d
2017-05-25 16:04:46 +09:00
Mike Hommey
2357f45d41 Bug 1368932 - Don't rely on the default MALLOC_DECL_VOID for malloc function declarations in replace-malloc. r=njn
In practice, this induces no change in what the expanded code looks
like.

--HG--
extra : rebase_source : 0df6a5dd3ca5161c0969c814e4222eeb51f9f9a2
2017-05-25 15:58:56 +09:00
Mike Hommey
7cd6d8b45d Bug 1368932 - Factor out function declarations for malloc implementation. r=njn
--HG--
extra : rebase_source : a8df78c7ade473eb5d7c71228f0939cce4554f57
2017-05-25 15:54:05 +09:00
Mike Hommey
16215c85f1 Bug 1368932 - Remove void argument from declarations in malloc_decls.h. r=njn
This makes no significant difference in practice in the macro
expansions, but will help down the line.

--HG--
extra : rebase_source : 6d61c1f28c558321478d7e5f26390d27ae8ae3ac
2017-05-25 13:56:40 +09:00
Mike Hommey
f9a4ab0e55 Bug 1367695 - Remove leftovers from jemalloc4. r=njn
MOZ_REPLACE_JEMALLOC was only defined when building jemalloc4 as a
replace-malloc library.

--HG--
extra : rebase_source : fa5c402da07fa96448c170b6db99629469691efe
2017-05-25 16:24:13 +09:00
Mike Hommey
fade48cbe6 Bug 1365206 - Remove Sun Studio compiler optimization flags for mozjemalloc. r=gps
We don't support that compiler anymore.
2017-05-19 08:29:05 +09:00
Mike Hommey
fbbe2d361d Bug 1365460 - Remove function prototypes but keep necessary forward declarations. r=njn
--HG--
extra : rebase_source : 670aecc65cf709f2312a29c80e3c3171ccb2c8c2
2017-05-18 16:27:06 +09:00
Mike Hommey
1dfc7df618 Bug 1365460 - Make umax2s support only base 10. r=njn
Other bases are never used.

--HG--
extra : rebase_source : 6c8edcbde86480995b2d96ac513724e3212457cf
2017-05-18 16:22:10 +09:00
Mike Hommey
eaa9da1203 Bug 1365460 - Avoid unused variables. r=njn
--HG--
extra : rebase_source : 18cd68afc595aca1177e453f6c60a65de1d0dece
2017-05-18 12:24:57 +09:00
Mike Hommey
be0c2ffbc5 Bug 1365460 - Remove warning exceptions for MSVC and work around them. r=njn
--HG--
extra : rebase_source : 182c1e5583950f52a5eadc9fcf5d0b72ba4c56a9
2017-05-18 12:13:34 +09:00
Mike Hommey
61860e7f63 Bug 1365460 - Don't define integer types that MSVC >= 2015 know about. r=njn
--HG--
extra : rebase_source : 6a13970a477fda647eaed1dc381ecb993297922f
2017-05-18 12:09:38 +09:00
Mike Hommey
d3f17931b7 Bug 1365460 - Remove JEMALLOC_USES_MAP_ALIGN. r=njn
It's a Solaris-only optimization that was used as a workaround for some
infinite loop in pages_map (bug 457189). In the meanwhile, the way
pages_map works has changed in such a way the infinite loop can't happen
anymore.

Specifically, the original problem was that pages_map would try to
allocate something larger than a chunk, then deallocate it, and
reallocate at a hinted aligned address near the address of the now
deallocated mapping, hopefully getting an aligned chunk. But Solaris
would ignore the hint and the chunk would never be aligned, causing the
infinite loop.

What the code does now it over-allocate, find an aligned chunk in there,
and deallocate what ends up not being needed. Which leaves no room for
the original infinite loop.

We thus remove the workaround and put Solaris in par with other Unix
platforms.

--HG--
extra : rebase_source : 9ce4509d9c5ac6123cedabf87c5738672af35d1b
2017-05-18 11:30:52 +09:00
Mike Hommey
3dce495fff Bug 1365460 - Use MOZ_DIAGNOSTIC_ASSERT where mozjemalloc uses RELEASE_ASSERT. r=njn
In Gecko code, MOZ_RELEASE_ASSERT means assertions that happen on all
types of builds.

In mozjemalloc, RELEASE_ASSERT means assertions that happen when
MOZ_JEMALLOC_HARD_ASSERTS is set, otherwise normal assertions. Which is
confusing. On the other hand, it's very similar to
MOZ_DIAGNOSTIC_ASSERT, and we may just want to use that instead.

Moreover, with release promotion, the check setting
MOZ_JEMALLOC_HARD_ASSERTS means releases (promoted from beta) would end
up with those asserts while they're not meant to, so
MOZ_DIAGNOSTIC_ASSERT is actually closer to the intent. It however means
we'd lose the beta population running those assertions.

--HG--
extra : rebase_source : 606ab47af8d9ee793b13b806acadb9781c99a078
2017-05-18 10:57:43 +09:00