While the comment says:
* To prevent rpciod from hanging, this allocator never sleeps,
* returning NULL if the request cannot be serviced immediately.
The function does not actually check for NULL pointers being returned.
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Make it more useful for debugging purposes.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
The maximum size of an NFSv4 SETATTR compound reply should include the
GETATTR operation that we send.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Use a cleaner method to find the size of an rpc_buffer. This actually
works on x86-64!
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
The check for nfs_attribute_timeout(dir) in nfs_check_verifier is
redundant: nfs_lookup_revalidate() will already call nfs_revalidate_inode()
on the parent dir when necessary.
The only case where this is not done is the case of a negative dentry. Fix
this case by moving up the revalidation code.
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
dentry verifiers are always set to the parent directory's
cache_change_attribute. There is no reason to be testing for anything other
than equality when we're trying to find out if the dentry has been checked
since the last time the directory was modified.
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
* git://git.kernel.org/pub/scm/linux/kernel/git/aegl/linux-2.6:
[IA64] wire up pselect, ppoll
[IA64] Add TIF_RESTORE_SIGMASK
[IA64] unwind did not work for processes born with CLONE_STOPPED
[IA64] Optional method to purge the TLB on SN systems
[IA64] SPIN_LOCK_UNLOCKED macro cleanup in arch/ia64
[IA64-SN2][KJ] mmtimer.c-kzalloc
[IA64] fix stack alignment for ia32 signal handlers
[IA64] - Altix: hotplug after intr redirect can crash system
[IA64] save and restore cpus_allowed in cpu_idle_wait
[IA64] Removal of percpu TR cleanup in kexec code
[IA64] Fix some section mismatch errors
* git://git.infradead.org/mtd-2.6: (21 commits)
[MTD] [CHIPS] Remove MTD_OBSOLETE_CHIPS (jedec, amd_flash, sharp)
[MTD] Delete allegedly obsolete "bank_size" field of mtd_info.
[MTD] Remove unnecessary user space check from mtd.h.
[MTD] [MAPS] Remove flash maps for no longer supported 405LP boards
[MTD] [MAPS] Fix missing printk() parameter in physmap_of.c MTD driver
[MTD] [NAND] platform NAND driver: add driver
[MTD] [NAND] platform NAND driver: update header
[JFFS2] Simplify and clean up jffs2_add_tn_to_tree() some more.
[JFFS2] Remove another bogus optimisation in jffs2_add_tn_to_tree()
[JFFS2] Remove broken insert_point optimisation in jffs2_add_tn_to_tree()
[JFFS2] Remember to calculate overlap on nodes which replace older nodes
[JFFS2] Don't advance c->wbuf_ofs to next eraseblock after wbuf flush
[MTD] [NAND] at91_nand.c: CMDLINE_PARTS support
[MTD] [NAND] Tidy up handling of page number in nand_block_bad()
[MTD] block2mtd_paramline[] mustn't be __initdata
[MTD] [NAND] Support multiple chips in CAFÉ driver
[MTD] [NAND] Rename cafe.c to cafe_nand.c and remove the multi-obj magic
[MTD] [NAND] Use rslib for CAFÉ ECC
[RSLIB] Support non-canonical GF representations
[JFFS2] Remove dead file histo_mips.h
...
Commit 6fe6900e1e introduced a nasty bug
in read_cache_page_async().
It added a "mark_page_accessed(page)" at the final return path in
read_cache_page_async(). But in error cases, 'page' holds the error
code, and you can't mark it accessed.
[ and Glauber de Oliveira Costa points out that we can use a return
instead of adding more goto's ]
Signed-off-by: David Howells <dhowells@redhat.com>
Acked-by: Nick Piggin <npiggin@suse.de>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* 'master' of git://git.kernel.org/pub/scm/linux/kernel/git/paulus/powerpc:
[POWERPC] Further fixes for the removal of 4level-fixup hack from ppc32
[POWERPC] EEH: log all PCI-X and PCI-E AER registers
[POWERPC] EEH: capture and log pci state on error
[POWERPC] EEH: Split up long error msg
[POWERPC] EEH: log error only after driver notification.
[POWERPC] fsl_soc: Make mac_addr const in fs_enet_of_init().
[POWERPC] Don't use SLAB/SLUB for PTE pages
[POWERPC] Spufs support for 64K LS mappings on 4K kernels
[POWERPC] Add ability to 4K kernel to hash in 64K pages
[POWERPC] Introduce address space "slices"
[POWERPC] Small fixes & cleanups in segment page size demotion
[POWERPC] iSeries: Make HVC_ISERIES the default
[POWERPC] iSeries: suppress build warning in lparmap.c
[POWERPC] Mark pages that don't exist as nosave
[POWERPC] swsusp: Introduce register_nosave_region_late
* git://git.kernel.org/pub/scm/linux/kernel/git/bunk/trivial: (25 commits)
sound: convert "sound" subdirectory to UTF-8
MAINTAINERS: Add cxacru website/mailing list
include files: convert "include" subdirectory to UTF-8
general: convert "kernel" subdirectory to UTF-8
documentation: convert the Documentation directory to UTF-8
Convert the toplevel files CREDITS and MAINTAINERS to UTF-8.
remove broken URLs from net drivers' output
Magic number prefix consistency change to Documentation/magic-number.txt
trivial: s/i_sem /i_mutex/
fix file specification in comments
drivers/base/platform.c: fix small typo in doc
misc doc and kconfig typos
Remove obsolete fat_cvf help text
Fix occurrences of "the the "
Fix minor typoes in kernel/module.c
Kconfig: Remove reference to external mqueue library
Kconfig: A couple of grammatical fixes in arch/i386/Kconfig
Correct comments in genrtc.c to refer to correct /proc file.
Fix more "deprecated" spellos.
Fix "deprecated" typoes.
...
Fix trivial comment conflict in kernel/relay.c.
When implementing things as macros, make sure we use typecasts and
parentheses where needed. The macros as defined were vulnerable to
surreptitious promotion causing problems.
Avoid macros where practical; e.g. wrmsr() can be an inline instead.
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
All users to the two-part rdtsc() macro have already switched to using
rdtscl() or rdtscll(). Remove the now-obsolete macro.
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The definition of USER686 is supposed to be a mask of feature bits,
not an OR of feature numbers! It happened to work anyway on the only
processor affected, simply by pure coincidence. Fix.
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
md currently uses ->media_changed to make sure rescan_partitions
is call on md array after they are assembled.
However that doesn't happen until the array is opened, which is later
than some people would like.
So use blkdev_ioctl to do the rescan immediately that the
array has been assembled.
This means we can remove all the ->change infrastructure as it was only used
to trigger a partition rescan.
Signed-off-by: Neil Brown <neilb@suse.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
"reshape_position" records how much progress has been made on a "reshape"
(adding drives, changing layout or chunksize).
When it is set, the number of drives, layout and chunksize can have
two possible values, an old an a new.
So allow these different values to be visible, and allow both old and new to
be set: Set the old ones first, then the reshape_position, then the new
values.
Signed-off-by: Neil Brown <neilb@suse.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
SLUB doesn't like slashes as it wants to use the cache name as the name of a
directory (or symlink) in sysfs.
Signed-off-by: Neil Brown <neilb@suse.de>
Acked-by: Christoph Lameter <clameter@sgi.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
If CONFIG_NET is not selected, csum_partial is not exported, so md.ko cannot
use it. We shouldn't really be using csum_partial anyway as it is an
internal-to-networking interface.
So replace it with C code to do the same thing. Speed is not crucial here, so
something simple and correct is best.
Signed-off-by: Neil Brown <neilb@suse.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
We need to check for internal-consistency of superblock in load_super.
validate_super is for inter-device consistency.
With the test in the wrong place, a badly created array will confuse md rather
an produce sensible errors.
Signed-off-by: Neil Brown <neilb@suse.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
We can save some lines of code by using seq_release_private().
Signed-off-by: Martin Peschke <mp3@de.ibm.com>
Acked-by: Neil Brown <neilb@suse.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Use ARRAY_SIZE macro already defined in kernel.h
Signed-off-by: Ahmed S. Darwish <darwish.07@gmail.com>
Acked-by: Neil Brown <neilb@suse.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Provide framebuffer page protection flags and definitions of
fb_readl/fb_writel for AVR32.
Signed-off-by: Haavard Skinnemoen <hskinnemoen@atmel.com>
Cc: "Antonino A. Daplas" <adaplas@pol.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Move fb_get_caps() method to svgalib.c as svga_get_caps() so it can be used by
s3fb, arkfb and vt8623fb.
Signed-off-by: Antonino Daplas <adaplas@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
This patch adds fbdev driver for graphics core in VIA VT8623
[adaplas@gmail.com: build fixes]
Signed-off-by: Ondrej Zajicek <santiago@crfreenet.org>
Signed-off-by: Antonino Daplas <adaplas@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Replace automatic variable instances of __attribute__ ((unused)) with
__maybe_unused.
Cc: Andy Whitcroft <apw@shadowen.org>
Signed-off-by: David Rientjes <rientjes@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Replace automatic variable instances of __attribute__((unused)) with
__maybe_unused in mca_nmi_hook().
Cc: James Bottomley <James.Bottomley@HansenPartnership.com>
Signed-off-by: David Rientjes <rientjes@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
There is no such thing as labeling a variable as __attribute__((used)). Since
ts_shift is not referenced in inline assembly, we assume that we're simply
suppressing a warning here if the variable is declared but unreferenced.
Cc: Paul Mundt <lethal@linux-sh.org>
Signed-off-by: David Rientjes <rientjes@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Use the new macro here
Cc: Andi Kleen <ak@suse.de>
Signed-off-by: David Rientjes <rientjes@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
__used is defined to be __attribute__((unused)) for all pre-3.3 gcc
compilers to suppress warnings for unused functions because perhaps they
are referenced only in inline assembly. It is defined to be
__attribute__((used)) for gcc 3.3 and later so that the code is still
emitted for such functions.
__maybe_unused is defined to be __attribute__((unused)) for both function
and variable use if it could possibly be unreferenced due to the evaluation
of preprocessor macros. Function prototypes shall be marked with
__maybe_unused if the actual definition of the function is dependant on
preprocessor macros.
No update to compiler-intel.h is necessary because ICC supports both
__attribute__((used)) and __attribute__((unused)) as specified by the gcc
manual.
__attribute_used__ is deprecated and will be removed once all current
code is converted to using __used.
Cc: Rusty Russell <rusty@rustcorp.com.au>
Cc: Adrian Bunk <bunk@stusta.de>
Signed-off-by: David Rientjes <rientjes@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
This finally renames the thread_info field in task structure to stack, so that
the assumptions about this field are gone and archs have more freedom about
placing the thread_info structure.
Nonbroken archs which have a proper thread pointer can do the access to both
current thread and task structure via a single pointer.
It'll allow for a few more cleanups of the fork code, from which e.g. ia64
could benefit.
Signed-off-by: Roman Zippel <zippel@linux-m68k.org>
[akpm@linux-foundation.org: build fix]
Cc: Richard Henderson <rth@twiddle.net>
Cc: Ivan Kokshaysky <ink@jurassic.park.msu.ru>
Cc: Russell King <rmk@arm.linux.org.uk>
Cc: Ian Molton <spyro@f2s.com>
Cc: Haavard Skinnemoen <hskinnemoen@atmel.com>
Cc: Mikael Starvik <starvik@axis.com>
Cc: David Howells <dhowells@redhat.com>
Cc: Yoshinori Sato <ysato@users.sourceforge.jp>
Cc: "Luck, Tony" <tony.luck@intel.com>
Cc: Hirokazu Takata <takata@linux-m32r.org>
Cc: Geert Uytterhoeven <geert@linux-m68k.org>
Cc: Roman Zippel <zippel@linux-m68k.org>
Cc: Greg Ungerer <gerg@uclinux.org>
Cc: Ralf Baechle <ralf@linux-mips.org>
Cc: Ralf Baechle <ralf@linux-mips.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Paul Mundt <lethal@linux-sh.org>
Cc: Kazumoto Kojima <kkojima@rr.iij4u.or.jp>
Cc: Richard Curnow <rc@rc0.org.uk>
Cc: William Lee Irwin III <wli@holomorphy.com>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Jeff Dike <jdike@addtoit.com>
Cc: Paolo 'Blaisorblade' Giarrusso <blaisorblade@yahoo.it>
Cc: Miles Bader <uclinux-v850@lsi.nec.co.jp>
Cc: Andi Kleen <ak@muc.de>
Cc: Chris Zankel <chris@zankel.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Recently a few direct accesses to the thread_info in the task structure snuck
back, so this wraps them with the appropriate wrapper.
Signed-off-by: Roman Zippel <zippel@linux-m68k.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
This will later allow an arch to add module specific information via linker
generated tables instead of poking directly in the module object structure.
Signed-off-by: Roman Zippel <zippel@linux-m68k.org>
Signed-off-by: Geert Uytterhoeven <geert@linux-m68k.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
We need to make sure that the clocksources are resumed, when timekeeping is
resumed. The current resume logic does not guarantee this.
Add a resume function pointer to the clocksource struct, so clocksource
drivers which need to reinitialize the clocksource can provide a resume
function.
Add a resume function, which calls the maybe available clocksource resume
functions and resets the watchdog function, so a stable TSC can be used
accross suspend/resume.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: john stultz <johnstul@us.ibm.com>
Cc: Andi Kleen <ak@suse.de>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: <stable@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Currently the slab allocators contain callbacks into the page allocator to
perform the draining of pagesets on remote nodes. This requires SLUB to have
a whole subsystem in order to be compatible with SLAB. Moving node draining
out of the slab allocators avoids a section of code in SLUB.
Move the node draining so that is is done when the vm statistics are updated.
At that point we are already touching all the cachelines with the pagesets of
a processor.
Add a expire counter there. If we have to update per zone or global vm
statistics then assume that the pageset will require subsequent draining.
The expire counter will be decremented on each vm stats update pass until it
reaches zero. Then we will drain one batch from the pageset. The draining
will cause vm counter updates which will then cause another expiration until
the pcp is empty. So we will drain a batch every 3 seconds.
Note that remote node draining is a somewhat esoteric feature that is required
on large NUMA systems because otherwise significant portions of system memory
can become trapped in pcp queues. The number of pcp is determined by the
number of processors and nodes in a system. A system with 4 processors and 2
nodes has 8 pcps which is okay. But a system with 1024 processors and 512
nodes has 512k pcps with a high potential for large amount of memory being
caught in them.
Signed-off-by: Christoph Lameter <clameter@sgi.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Make it configurable. Code in mm makes the vm statistics intervals
independent from the cache reaper use that opportunity to make it
configurable.
Signed-off-by: Christoph Lameter <clameter@sgi.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
vmstat is currently using the cache reaper to periodically bring the
statistics up to date. The cache reaper does only exists in SLUB as a way to
provide compatibility with SLAB. This patch removes the vmstat calls from the
slab allocators and provides its own handling.
The advantage is also that we can use a different frequency for the updates.
Refreshing vm stats is a pretty fast job so we can run this every second and
stagger this by only one tick. This will lead to some overlap in large
systems. F.e a system running at 250 HZ with 1024 processors will have 4 vm
updates occurring at once.
However, the vm stats update only accesses per node information. It is only
necessary to stagger the vm statistics updates per processor in each node. Vm
counter updates occurring on distant nodes will not cause cacheline
contention.
We could implement an alternate approach that runs the first processor on each
node at the second and then each of the other processor on a node on a
subsequent tick. That may be useful to keep a large amount of the second free
of timer activity. Maybe the timer folks will have some feedback on this one?
[jirislaby@gmail.com: add missing break]
Cc: Arjan van de Ven <arjan@linux.intel.com>
Signed-off-by: Christoph Lameter <clameter@sgi.com>
Signed-off-by: Jiri Slaby <jirislaby@gmail.com>
Cc: Oleg Nesterov <oleg@tv-sign.ru>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Make the microcode driver use the suspend-related CPU hotplug notifications
to handle the CPU hotplug events occuring during system-wide suspend and
resume transitions. Remove the global variable suspend_cpu_hotplug
previously used for this purpose.
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
Cc: Gautham R Shenoy <ego@in.ibm.com>
Cc: Pavel Machek <pavel@ucw.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Since nonboot CPUs are now disabled after tasks and devices have been
frozen and the CPU hotplug infrastructure is used for this purpose, we need
special CPU hotplug notifications that will help the CPU-hotplug-aware
subsystems distinguish normal CPU hotplug events from CPU hotplug events
related to a system-wide suspend or resume operation in progress. This
patch introduces such notifications and causes them to be used during
suspend and resume transitions. It also changes all of the
CPU-hotplug-aware subsystems to take these notifications into consideration
(for now they are handled in the same way as the corresponding "normal"
ones).
[oleg@tv-sign.ru: cleanups]
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
Cc: Gautham R Shenoy <ego@in.ibm.com>
Cc: Pavel Machek <pavel@ucw.cz>
Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Now that all the in-tree users are converted over to zero_user_page(),
deprecate the old memclear_highpage_flush() call.
Signed-off-by: Nate Diller <nate.diller@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Use zero_user_page() instead of open-coding it.
Signed-off-by: Nate Diller <nate.diller@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Use zero_user_page() instead of open-coding it.
Signed-off-by: Nate Diller <nate.diller@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Use zero_user_page() instead of open-coding it.
Signed-off-by: Nate Diller <nate.diller@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
It's very common for file systems to need to zero part or all of a page,
the simplist way is just to use kmap_atomic() and memset(). There's
actually a library function in include/linux/highmem.h that does exactly
that, but it's confusingly named memclear_highpage_flush(), which is
descriptive of *how* it does the work rather than what the *purpose* is.
So this patchset renames the function to zero_user_page(), and calls it
from the various places that currently open code it.
This first patch introduces the new function call, and converts all the
core kernel callsites, both the open-coded ones and the old
memclear_highpage_flush() ones. Following this patch is a series of
conversions for each file system individually, per AKPM, and finally a
patch deprecating the old call. The diffstat below shows the entire
patchset.
[akpm@linux-foundation.org: fix a few things]
Signed-off-by: Nate Diller <nate.diller@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Analysis of current linux futex code :
--------------------------------------
A central hash table futex_queues[] holds all contexts (futex_q) of waiting
threads.
Each futex_wait()/futex_wait() has to obtain a spinlock on a hash slot to
perform lookups or insert/deletion of a futex_q.
When a futex_wait() is done, calling thread has to :
1) - Obtain a read lock on mmap_sem to be able to validate the user pointer
(calling find_vma()). This validation tells us if the futex uses
an inode based store (mapped file), or mm based store (anonymous mem)
2) - compute a hash key
3) - Atomic increment of reference counter on an inode or a mm_struct
4) - lock part of futex_queues[] hash table
5) - perform the test on value of futex.
(rollback is value != expected_value, returns EWOULDBLOCK)
(various loops if test triggers mm faults)
6) queue the context into hash table, release the lock got in 4)
7) - release the read_lock on mmap_sem
<block>
8) Eventually unqueue the context (but rarely, as this part may be done
by the futex_wake())
Futexes were designed to improve scalability but current implementation has
various problems :
- Central hashtable :
This means scalability problems if many processes/threads want to use
futexes at the same time.
This means NUMA unbalance because this hashtable is located on one node.
- Using mmap_sem on every futex() syscall :
Even if mmap_sem is a rw_semaphore, up_read()/down_read() are doing atomic
ops on mmap_sem, dirtying cache line :
- lot of cache line ping pongs on SMP configurations.
mmap_sem is also extensively used by mm code (page faults, mmap()/munmap())
Highly threaded processes might suffer from mmap_sem contention.
mmap_sem is also used by oprofile code. Enabling oprofile hurts threaded
programs because of contention on the mmap_sem cache line.
- Using an atomic_inc()/atomic_dec() on inode ref counter or mm ref counter:
It's also a cache line ping pong on SMP. It also increases mmap_sem hold time
because of cache misses.
Most of these scalability problems come from the fact that futexes are in
one global namespace. As we use a central hash table, we must make sure
they are all using the same reference (given by the mm subsystem). We
chose to force all futexes be 'shared'. This has a cost.
But fact is POSIX defined PRIVATE and SHARED, allowing clear separation,
and optimal performance if carefuly implemented. Time has come for linux
to have better threading performance.
The goal is to permit new futex commands to avoid :
- Taking the mmap_sem semaphore, conflicting with other subsystems.
- Modifying a ref_count on mm or an inode, still conflicting with mm or fs.
This is possible because, for one process using PTHREAD_PROCESS_PRIVATE
futexes, we only need to distinguish futexes by their virtual address, no
matter the underlying mm storage is.
If glibc wants to exploit this new infrastructure, it should use new
_PRIVATE futex subcommands for PTHREAD_PROCESS_PRIVATE futexes. And be
prepared to fallback on old subcommands for old kernels. Using one global
variable with the FUTEX_PRIVATE_FLAG or 0 value should be OK.
PTHREAD_PROCESS_SHARED futexes should still use the old subcommands.
Compatibility with old applications is preserved, they still hit the
scalability problems, but new applications can fly :)
Note : the same SHARED futex (mapped on a file) can be used by old binaries
*and* new binaries, because both binaries will use the old subcommands.
Note : Vast majority of futexes should be using PROCESS_PRIVATE semantic,
as this is the default semantic. Almost all applications should benefit
of this changes (new kernel and updated libc)
Some bench results on a Pentium M 1.6 GHz (SMP kernel on a UP machine)
/* calling futex_wait(addr, value) with value != *addr */
433 cycles per futex(FUTEX_WAIT) call (mixing 2 futexes)
424 cycles per futex(FUTEX_WAIT) call (using one futex)
334 cycles per futex(FUTEX_WAIT_PRIVATE) call (mixing 2 futexes)
334 cycles per futex(FUTEX_WAIT_PRIVATE) call (using one futex)
For reference :
187 cycles per getppid() call
188 cycles per umask() call
181 cycles per ni_syscall() call
Signed-off-by: Eric Dumazet <dada1@cosmosbay.com>
Pierre Peiffer <pierre.peiffer@bull.net>
Cc: "Ulrich Drepper" <drepper@gmail.com>
Cc: "Nick Piggin" <nickpiggin@yahoo.com.au>
Cc: "Ingo Molnar" <mingo@elte.hu>
Cc: Rusty Russell <rusty@rustcorp.com.au>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
