Commit Graph

294 Commits

Author SHA1 Message Date
Roland McGrath ec136444ca x86: Consolidate NPTL fork. 2014-05-14 11:31:37 -07:00
Roland McGrath 2fbdf5339a Consolidate not-cancel.h files. 2014-05-14 11:06:36 -07:00
Roland McGrath 941d7dfd24 x86: Consolidate NPTL/non versions of clone 2014-05-14 10:16:27 -07:00
Roland McGrath e27871374e x86: Consolidate NPTL/non versions of vfork 2014-05-14 09:44:39 -07:00
Andi Kleen df5b85da90 Fix dwarf2 unwinding through futex functions.
When profiling programs with lock problems with perf record -g dwarf,
libunwind can currently not backtrace through the futex and unlock
functions in pthread. This is because they use out of line sections,
and those are not correctly described in dwarf2 (I believe needs
dwarf3 or 4).

This patch first removes the out of line sections. They only save a
single jump, but cause a lot of pain. Then it converts the now inline
lock code to use the now standard gas .cfi_* commands.

With these changes libunwind/perf can backtrace through the futex
functions now.

Longer term it would be likely better to just use C futex() functions
on x86 like all the other architectures. This would clean the code up
even more.
2014-03-26 23:18:15 -04:00
Allan McRae d4697bc93d Update copyright notices with scripts/update-copyrights 2014-01-01 22:00:23 +10:00
Andreas Schwab caecf2a0ce Fix CFI annotations in pthread_cond_timedwait for i486+ 2013-12-09 13:53:41 +01:00
Andi Kleen 1cdbe57948 Add the low level infrastructure for pthreads lock elision with TSX
Lock elision using TSX is a technique to optimize lock scaling
It allows to run locks in parallel using hardware support for
a transactional execution mode in 4th generation Intel Core CPUs.
See http://www.intel.com/software/tsx for more Information.

This patch implements a simple adaptive lock elision algorithm based
on RTM. It enables elision for the pthread mutexes and rwlocks.
The algorithm keeps track whether a mutex successfully elides or not,
and stops eliding for some time when it is not.

When the CPU supports RTM the elision path is automatically tried,
otherwise any elision is disabled.

The adaptation algorithm and its tuning is currently preliminary.

The code adds some checks to the lock fast paths. Micro-benchmarks
show little to no difference without RTM.

This patch implements the low level "lll_" code for lock elision.
Followon patches hook this into the pthread implementation

Changes with the RTM mutexes:
-----------------------------
Lock elision in pthreads is generally compatible with existing programs.
There are some obscure exceptions, which are expected to be uncommon.
See the manual for more details.

- A broken program that unlocks a free lock will crash.
  There are ways around this with some tradeoffs (more code in hot paths)
  I'm still undecided on what approach to take here; have to wait for testing reports.
- pthread_mutex_destroy of a lock mutex will not return EBUSY but 0.
- There's also a similar situation with trylock outside the mutex,
  "knowing" that the mutex must be held due to some other condition.
  In this case an assert failure cannot be recovered. This situation is
  usually an existing bug in the program.
- Same applies to the rwlocks. Some of the return values changes
  (for example there is no EDEADLK for an elided lock, unless it aborts.
   However when elided it will also never deadlock of course)
- Timing changes, so broken programs that make assumptions about specific timing
  may expose already existing latent problems.  Note that these broken programs will
  break in other situations too (loaded system, new faster hardware, compiler
  optimizations etc.)
- Programs with non recursive mutexes that take them recursively in a thread and
  which would always deadlock without elision may not always see a deadlock.
  The deadlock will only happen on an early or delayed abort (which typically
  happens at some point)
  This only happens for mutexes not explicitely set to PTHREAD_MUTEX_NORMAL
  or PTHREAD_MUTEX_ADAPTIVE_NP.  PTHREAD_MUTEX_NORMAL mutexes do not elide.

The elision default can be set at configure time.

This patch implements the basic infrastructure for elision.
2013-07-02 08:46:54 -07:00
Carlos O'Donell be11d71394 x86*: Return syscall error for lll_futex_wake.
It is very very possible that the futex syscall returns an
error and that the caller of lll_futex_wake may want to
look at that error and propagate the failure.

This patch allows a caller to see the syscall error.

There are no users of the syscall error at present, but
future cleanups are now be able to check for the error.

--

nplt/

2013-06-10  Carlos O'Donell  <carlos@redhat.com>

	* sysdeps/unix/sysv/linux/i386/lowlevellock.h
	(lll_futex_wake): Return syscall error.
	* sysdeps/unix/sysv/linux/x86_64/lowlevellock.h
	(lll_futex_wake): Return syscall error.
2013-06-10 12:05:11 -04:00
Joseph Myers c7afae94ca Remove trailing whitespace in nptl. 2013-06-06 12:06:15 +00:00
Andreas Schwab 4f682b2ae9 Extend i486 pthread_cond_timedwait to use futex syscall with absolute timeout 2013-04-11 10:40:39 +02:00
Roland McGrath b43769a3f5 Rejigger i386 dl-sysdep.h files. 2013-03-04 09:40:25 -08:00
Andreas Schwab e3f45e2bbe Revert "Extend i486 pthread_cond_timedwait to use futex syscall with absolute timeout"
This reverts commit 1bd57044e9.
2013-01-10 10:44:05 +01:00
Andreas Schwab 1bd57044e9 Extend i486 pthread_cond_timedwait to use futex syscall with absolute timeout
* sysdeps/unix/sysv/linux/i386/i486/pthread_cond_timedwait.S
	(__pthread_cond_timedwait): If possible use FUTEX_WAIT_BITSET to
	directly use absolute timeout.
2013-01-10 09:59:58 +01:00
Joseph Myers 568035b787 Update copyright notices with scripts/update-copyrights. 2013-01-02 19:05:09 +00:00
Siddhesh Poyarekar 9485a40444 Adjust mutex lock in condvar_cleanup if we got it from requeue_pi
This completes the fix to bz #14652.
2012-10-16 14:23:35 +05:30
Siddhesh Poyarekar 0e3b5d6a68 Take lock in pthread_cond_wait cleanup handler only when needed
[BZ #14652]
When a thread waiting in pthread_cond_wait with a PI mutex is
cancelled after it has returned successfully from the futex syscall
but just before async cancellation is disabled, it enters its
cancellation handler with the mutex held and simply calling a
mutex_lock again will result in a deadlock.  Hence, it is necessary to
see if the thread owns the lock and try to lock it only if it doesn't.
2012-10-10 12:52:56 +05:30
Siddhesh Poyarekar c30e8edf7c Unlock mutex before going back to waiting for PI mutexes
[BZ #14417]
A futex call with FUTEX_WAIT_REQUEUE_PI returns with the mutex locked
on success.  If such a successful thread is pipped to the cond_lock by
another spuriously woken waiter, it could be sent back to wait on the
futex with the mutex lock held, thus causing a deadlock.  So it is
necessary that the thread relinquishes the mutex before going back to
sleep.
2012-10-05 18:52:36 +05:30
Siddhesh Poyarekar adcdc775e1 Fix clone flag name in comment to CLONE_CHILD_CLEARTID. 2012-10-02 08:52:55 +05:30
Siddhesh Poyarekar 55a051c985 Fix exception table for i386 pthread_cond_wait
[BZ #14477]
Add an additional entry in the exception table to jump to
__condvar_w_cleanup2 instead of __condvar_w_cleanup for PI mutexes
when %ebx contains the address of the futex instead of the condition
variable.
2012-10-01 23:21:39 +05:30
Joseph Myers 93a78ac437 Remove __ASSUME_POSIX_TIMERS. 2012-08-16 14:03:43 +00:00
Andreas Schwab 842a39cd1a Remove unused pseudo_end label 2012-07-25 21:58:17 +02:00
H.J. Lu 0e20515a17 Use x86-64 bits/pthreadtypes.h/semaphore.h for i386/x86-64 2012-05-30 17:29:22 -07:00
Andreas Schwab 4d17e68350 Remove use of INTDEF/INTUSE in nptl 2012-05-30 00:45:53 +02:00
Rayson Ho 1755728208 i386 port of the pthread SystemTap probes 2012-05-25 13:41:04 -07:00
Andreas Jaeger de7f5ce7c5 Fix warnings on Linux/i686
Fixes:
../sysdeps/i386/dl-machine.h:336:30: warning: unused variable ‘refsym’ [-Wunused-variable]
rtld.c:1435:3: warning: implicit declaration of function ‘_dl_discover_osversion’ [-Wimplicit-function-declaration]
2012-05-17 20:17:53 +02:00
Thomas Schwinge 48aff7765b Fix 9554ebf2d4.
| Invalid timeouts in i386 sem_timedwait.
|
| We adjusted nwaiters even though this isn't necessary.
2012-03-08 09:33:12 +01:00
David S. Miller 22f9d9df92 Fix stray references to __pthread_attr
* sysdeps/unix/sysv/linux/i386/bits/pthreadtypes.h: Don't refer to non-existing
	__pthread_attr.
	* sysdeps/unix/sysv/linux/powerpc/bits/pthreadtypes.h: Likewise.
	* sysdeps/unix/sysv/linux/s390/bits/pthreadtypes.h: Likewise.
	* sysdeps/unix/sysv/linux/sh/bits/pthreadtypes.h: Likewise.
	* sysdeps/unix/sysv/linux/sparc/bits/pthreadtypes.h: Likewise.
	* sysdeps/unix/sysv/linux/x86_64/bits/pthreadtypes.h: Likewise.
2012-02-27 12:59:16 -08:00
Ulrich Drepper 8da0464f6f Fix name mangling of pthread_attr_t after change 2012-02-26 21:42:53 -05:00
Ulrich Drepper 4a3dded527 Work around problem of pthread_attr_t definition with old compilers 2012-02-26 17:41:17 -05:00
Ulrich Drepper 4efeffc1d5 Fix up POSIX testing in conformtest 2012-02-26 13:17:27 -05:00
Paul Eggert 59ba27a63a Replace FSF snail mail address with URLs. 2012-02-09 23:18:22 +00:00
Joseph Myers 3add8e1353 Support crti.S and crtn.S provided directly by architectures. 2012-02-08 01:45:26 +00:00
Joseph Myers 9a1d92541f Consistently use macros for x86 PIC thunks. 2012-02-03 23:22:53 +00:00
Andreas Schwab c5a0802a68 Handle EAGAIN from FUTEX_WAIT_REQUEUE_PI 2011-11-30 11:03:19 +01:00
Ulrich Drepper d063d16433 Remove support for !USE___THREAD 2011-09-10 16:50:28 -04:00
Dinakar Guniguntala 022f6b8920 Fix x86 pthread_cond_signal() FUTEX_WAKE_OP fallback 2010-08-19 00:46:19 -07:00
H.J. Lu f8392f4004 Add -fno-asynchronous-unwind-tables to initfini.s for i386 2010-08-15 20:09:23 -07:00
Andreas Schwab 2983d85ee5 Work around kernel rejecting valid absolute timestamps 2010-07-01 19:07:14 -07:00
Michal Schmidt 893549c5a0 Fix pthread_cond_*wait with requeue-PI on i386. 2010-01-15 18:09:14 -08:00
Ulrich Drepper 23b8575625 More whitespace fixes. 2009-12-13 11:52:28 -08:00
Ulrich Drepper 5a9e37df22 Fix whitespaces. 2009-12-13 11:51:54 -08:00
Dinakar Guniguntala 75956694f3 Add Requeue-PI support for x86 arch. 2009-12-13 11:50:16 -08:00
Ulrich Drepper 9554ebf2d4 Invalid timeouts in i386 sem_timedwait.
We adjusted nwaiters even though this isn't necessary.
2009-12-12 11:06:23 -08:00
Roland McGrath 57a299fe47 Use #include instead of duplication for i386 vs i686 dl-sysdep.h in NPTL. 2009-11-22 17:49:21 -08:00
Ulrich Drepper dfedb126d6 Minor optimizations and cleanups of x86 cond_broadcast. 2009-11-18 06:40:25 -08:00
Dinakar Guniguntala dd7106b3c6 Update cfi offsets for pthread_cond_broadcast and signal fir x86. 2009-11-18 06:33:23 -08:00
Ulrich Drepper f8c10bb45b Use cfi directives in x86 pthread_cond_{,timed}wait. 2009-11-17 17:59:33 -08:00
Bryan Kadzban 1828530f09 Fix futex syscall parameter for x86 absolute timeout waits.
This affects only installations compiled for kernels older than 2.6.18.
2009-06-11 11:32:51 -07:00
Ulrich Drepper 45f7760695 (__lll_timedlock_wait): Use FUTEX_WAIT_BITSET|FUTEX_CLOCK_READTIME instead of computing relative timeout. 2009-01-03 23:17:21 +00:00