Linux from 3.9 through 4.2 does not abort HTM transaction on syscalls,
instead it suspend and resume it when leaving the kernel. The
side-effects of the syscall will always remain visible, even if the
transaction is aborted. This is an issue when transaction is used along
with futex syscall, on pthread_cond_wait for instance, where the futex
call might succeed but the transaction is rolled back leading the
pthread_cond object in an inconsistent state.
Glibc used to prevent it by always aborting a transaction before issuing
a syscall. Linux 4.2 also decided to abort active transaction in
syscalls which makes the glibc workaround superfluous. Worse, glibc
transaction abortion leads to a performance issue on recent kernels
where the HTM state is saved/restore lazily (v4.9). By aborting a
transaction on every syscalls, regardless whether a transaction has being
initiated before, GLIBS makes the kernel always save/restore HTM state
(it can not even lazily disable it after a certain number of syscall
iterations).
Because of this shortcoming, Transactional Lock Elision is just enabled
when it has been explicitly set (either by tunables of by a configure
switch) and if kernel aborts HTM transactions on syscalls
(PPC_FEATURE2_HTM_NOSC). It is reported that using simple benchmark [1],
the context-switch is about 5% faster by not issuing a tabort in every
syscall in newer kernels.
Checked on powerpc64le-linux-gnu with 4.4.0 kernel (Ubuntu 16.04).
* NEWS: Add note about new TLE support on powerpc64le.
* sysdeps/powerpc/nptl/tcb-offsets.sym (TM_CAPABLE): Remove.
* sysdeps/powerpc/nptl/tls.h (tcbhead_t): Rename tm_capable to
__ununsed1.
(TLS_INIT_TP, TLS_DEFINE_INIT_TP): Remove tm_capable setup.
(THREAD_GET_TM_CAPABLE, THREAD_SET_TM_CAPABLE): Remove macros.
* sysdeps/powerpc/powerpc32/sysdep.h,
sysdeps/powerpc/powerpc64/sysdep.h (ABORT_TRANSACTION_IMPL,
ABORT_TRANSACTION): Remove macros.
* sysdeps/powerpc/sysdep.h (ABORT_TRANSACTION): Likewise.
* sysdeps/unix/sysv/linux/powerpc/elision-conf.c (elision_init): Set
__pthread_force_elision iff PPC_FEATURE2_HTM_NOSC is set.
* sysdeps/unix/sysv/linux/powerpc/powerpc32/sysdep.h,
sysdeps/unix/sysv/linux/powerpc/powerpc64/sysdep.h
sysdeps/unix/sysv/linux/powerpc/syscall.S (ABORT_TRANSACTION): Remove
usage.
* sysdeps/unix/sysv/linux/powerpc/not-errno.h: Remove file.
Reported-by: Breno Leitão <leitao@debian.org>
Some SPE opcodes clashes with some recent PowerISA opcodes and
until recently gas did not complain about it. However binutils
recently changed it and now VLE configured gas does not support to
assembler some instruction that might class with VLE (HTM for
instance). It also does not help that glibc build hardware lock
elision support as default (regardless of assembler support).
Although runtime will not actually enables TLE on SPE hardware
(since kernel will not advertise it), I see little advantage on
adding HTM support on SPE built glibc. SPE uses an incompatible
ABI which does not allow share the same build with default
powerpc and HTM code slows down SPE without any benefict.
This patch fixes it by only building HTM when SPE configuration
is not used.
Checked with a powerpc-linux-gnuspe build. I also did some sniff
tests on a e500 hardware without any issue.
[BZ #22926]
* sysdeps/powerpc/powerpc32/sysdep.h (ABORT_TRANSACTION_IMPL): Define
empty for __SPE__.
* sysdeps/powerpc/sysdep.h (ABORT_TRANSACTION): Likewise.
* sysdeps/unix/sysv/linux/powerpc/elision-lock.c (__lll_lock_elision):
Do not build hardware transactional code for __SPE__.
* sysdeps/unix/sysv/linux/powerpc/elision-trylock.c
(__lll_trylock_elision): Likewise.
* sysdeps/unix/sysv/linux/powerpc/elision-unlock.c
(__lll_unlock_elision): Likewise.
This patch adds several new tunables to control the behavior of
elision on supported platforms[1]. Since elision now depends
on tunables, we should always *compile* with elision enabled,
and leave the code disabled, but available for runtime
selection. This gives us *much* better compile-time testing of
the existing code to avoid bit-rot[2].
Tested on ppc, ppc64, ppc64le, s390x and x86_64.
[1] This part of the patch was initially proposed by
Paul Murphy but was "staled" because the framework have changed
since the patch was originally proposed:
https://patchwork.sourceware.org/patch/10342/
[2] This part of the patch was inititally proposed as a RFC by
Carlos O'Donnell. Make sense to me integrate this on the patch:
https://sourceware.org/ml/libc-alpha/2017-05/msg00335.html
* elf/dl-tunables.list: Add elision parameters.
* manual/tunables.texi: Add entries about elision tunable.
* sysdeps/unix/sysv/linux/powerpc/elision-conf.c:
Add callback functions to dynamically enable/disable elision.
Add multiple callbacks functions to set elision parameters.
Deleted __libc_enable_secure check.
* sysdeps/unix/sysv/linux/s390/elision-conf.c: Likewise.
* sysdeps/unix/sysv/linux/x86/elision-conf.c: Likewise.
* configure: Regenerated.
* configure.ac: Option enable_lock_elision was deleted.
* config.h.in: ENABLE_LOCK_ELISION flag was deleted.
* config.make.in: Remove references to enable_lock_elision.
* manual/install.texi: Elision configure option was removed.
* INSTALL: Regenerated to remove enable_lock_elision.
* nptl/Makefile:
Disable elision so it can verify error case for destroying a mutex.
* sysdeps/powerpc/nptl/elide.h:
Cleanup ENABLE_LOCK_ELISION check.
Deleted macros for the case when ENABLE_LOCK_ELISION was not defined.
* sysdeps/s390/configure: Regenerated.
* sysdeps/s390/configure.ac: Remove references to enable_lock_elision..
* nptl/tst-mutex8.c:
Deleted all #ifndef ENABLE_LOCK_ELISION from the test.
* sysdeps/powerpc/powerpc32/sysdep.h:
Deleted all ENABLE_LOCK_ELISION checks.
* sysdeps/powerpc/powerpc64/sysdep.h: Likewise.
* sysdeps/powerpc/sysdep.h: Likewise.
* sysdeps/s390/nptl/bits/pthreadtypes-arch.h: Likewise.
* sysdeps/unix/sysv/linux/powerpc/force-elision.h: Likewise.
* sysdeps/unix/sysv/linux/s390/elision-conf.h: Likewise.
* sysdeps/unix/sysv/linux/s390/force-elision.h: Likewise.
* sysdeps/unix/sysv/linux/s390/lowlevellock.h: Likewise.
* sysdeps/unix/sysv/linux/s390/Makefile: Remove references to
enable-lock-elision.
Reviewed-by: Tulio Magno Quites Machado Filho <tuliom@linux.vnet.ibm.com>
Work around a GCC behavior with hardware transactional memory built-ins.
GCC doesn't treat the PowerPC transactional built-ins as compiler
barriers, moving instructions past the transaction boundaries and
altering their atomicity.
Linux kernel powerpc documentation states issuing a syscall inside a
transaction is not recommended and may lead to undefined behavior. It
also states syscalls does not abort transactoin neither they run in
transactional state.
To avoid side-effects being visible outside transactions, GLIBC with
lock elision enabled will issue a transaction abort instruction just
before all syscalls if hardware supports hardware transactions.
elf/tst-auxv.c includes misc/sys/auxv.h, which ends up not actually
being included due to the guard overlap, and getauxval becomes an
implicit declaration and implicit pointer conversion which means, at
best, the test isn't actually testing what it thinks it is and, at
worst, it'll crash and burn on platforms where implict pointer
conversion is a Very Bad Thing.
* sysdeps/powerpc/bits/hwcap.h: Allow _SYSDEPS_SYSDEP_H guard as a
synonym for _SYS_AUXV_H to allow direct inclusion.
* sysdeps/sparc/bits/hwcap.h: Likewise.
* sysdeps/powerpc/sysdep.h: Define _SYSDEPS_SYSDEP_H instead of
_SYS_AUXV_H so we can include sysdep.h and sys/auxv.h together.
* sysdeps/sparc/sysdep.h: Likewise.
2008-08-01 Steven Munroe <sjmunroe@us.ibm.com>
Carlos Eduardo Seo <cseo@linux.vnet.ibm.com>
[BZ #6817]
* sysdeps/powerpc/dl-procinfo.c (_dl_powerpc_cap_flags):
Added the members 'vsx' and 'arch_2_06'.
(_dl_powerpc_platforms): Add the member 'power7'.
* sysdeps/powerpc/dl-procinfo.h: Modify _DL_HWCAP_FIRST
to reflect the changes required by VSX and ISA 2.06.
Modify _DL_PLATFORMS_COUNT to reflect the addition of
'power7'.
Defined PPC_PLATFORM_POWER7.
(_dl_string_platform): Add support for POWER7.
* sysdeps/powerpc/sysdep.h: Define bit masks for VSX
capability and ISA 2.06.
Steven Munroe <sjmunroe@us.ibm.com>
* sysdeps/powerpc/dl-procinfo.c (_dl_powerpc_cap_flags): Add 4 new cap
names to the beginning. Rename "cell" to "cellbe".
(_dl_powerpc_platforms): New.
* sysdeps/powerpc/dl-procinfo.h (_DL_HWCAP_FIRST): Decrease.
(HWCAP_IMPORTANT): Remove power{4,5,5+} and cell.
(_DL_PLATFORMS_COUNT, _DL_FIRST_PLATFORM): Define.
(_DL_HWCAP_PLATFORM): Define to new mask.
(_dl_platform_string, _dl_string_platform): New functions.
* sysdeps/powerpc/sysdep.h (PPC_FEATURE_BOOKE, PPC_FEATURE_SMT,
PPC_FEATURE_ICACHE_SNOOP, PPC_FEATURE_ARCH_2_05): Define.
2004-05-25 Steven Munroe <sjmunroe@us.ibm.com>
* sysdeps/powerpc/fpu/Makefile: Make ld.so a dependency of libm.so.
* sysdeps/powerpc/fpu/bits/mathinline.h [__LIBC_INERNAL_MATH_INLINES]
(__ieee754_sqrt): Define as __MATH_INLINE using fsqrt instruction.
(__ieee754_sqrtf): Define as __MATH_INLINE using fsqrts instruction.
* sysdeps/powerpc/fpu/e_sqrt.c (__slow_ieee754_sqrt): Moved
implementation from w_sqrt.c.
* sysdeps/powerpc/fpu/e_sqrtf.c (__slow_ieee754_sqrtf): Moved
implementation from w_sqrtf.c.
* sysdeps/powerpc/fpu/w_sqrt.c (__sqrt): Wrapper implementation
using inline __ieee754_sqrt().
* sysdeps/powerpc/fpu/w_sqrtf.c (__sqrtf): Wrapper implementation
using inline __ieee754_sqrtf().
* sysdeps/powerpc/powerpc32/sysdep.h [__ASSEMBLER__]: Include
<sysdeps/powerpc/sysdep.h> independent of __ASSEMBLER__.
* sysdeps/powerpc/sysdep.h [__ASSEMBLER__] (PPC_FEATURE_*): Define
PPC_FEATURE_* independent of __ASSEMBLER__.
2004-05-25 Jakub Jelinek <jakub@redhat.com>
* sysdeps/pthread/aio_notify.c: Use <> instead of "" for aio_misc.h
include.
(aio_start_notify_thread): Define if not defined.
(notify_func_wrapper): Use it.
* sysdeps/pthread/aio_misc.c: Use <> instead of "" for aio_misc.h
include.
(aio_create_helper_thread): Define if not defined.
(__aio_create_helper_thread): New function.
(__aio_enqueue_request): Use aio_create_helper_thread.
* nis/ypclnt.c (ypall_data, ypall_foreach): Remove.
(struct ypresp_all_data): New type.
(__xdr_ypresp_all): Change second argument to
struct ypresp_all_data *. Replace ypall_foreach and
ypall_data with objp->foreach and objp->data.
(yp_all): Remove status variable, add data. Replace
all uses of status with data.status. Initialize data.foreach
and data.data instead of ypall_foreach and ypall_data.
2004-05-24 Jakub Jelinek <jakub@redhat.com>
* elf/dl-lookup.c (add_dependency): Set DF_1_NODELETE bit
in l_flags_1, not in l_flags.
2004-01-14 Steven Munroe <sjmunroe@us.ibm.com>
* include/libc-symbols.h [HAVE_ASM_GLOBAL_DOT_NAME]
(_symbol_version): Use C_SYMBOL_DOT_NAME to create '.'ed symbols.
(_default_symbol_version): Use C_SYMBOL_DOT_NAME to create '.'ed
symbols.
* sysdeps/powerpc/Makefile: Add rtld-global-offsets.sym to
gen-as-const-headers.
* sysdeps/powerpc/elf/rtld-global-offsets.sym: New file.
* sysdeps/powerpc/sysdep.h: Define v# symbols for vector registers.
Define PPC_FEATURE_* masks for Aux Vector AT_HWCAP.
* sysdeps/unix/sysv/linux/kernel-features.h
(__ASSUME_SWAPCONTEXT_SYSCALL): Define for PPC and 2.6.0 kernels.
* sysdeps/unix/sysv/linux/powerpc/sys/ucontext.h [!__WORDSIZE == 32]:
Declare mcontext_t inline and include altivec state for 64-bit.
* sysdeps/unix/sysv/linux/powerpc/powerpc64/Versions: Add GLIBC_2.3.4
versions for setcontext, getcontext, and swapcontext.
* sysdeps/unix/sysv/linux/powerpc/powerpc64/Versions: Add GLIBC_2.3.4
versions for setcontext, getcontext, swapcontext, and makecontext.
* sysdeps/unix/sysv/linux/powerpc/powerpc64/getcontext.S
(__getcontext): Upgrade to save Altivec regs and version GLIBC_2_3_4.
[SHLIB_COMPAT (libc, GLIBC_2_3, GLIBC_2_3_4)](__novec_getcontext):
Compatible with GLIBC_2.3.3 release.
* sysdeps/unix/sysv/linux/powerpc/powerpc64/makecontext.S
(__makecontext): Use parm save area instead of compiler_dw to hold
context pointer.
* sysdeps/unix/sysv/linux/powerpc/powerpc64/setcontext.S
(__setcontext): Upgrade to restore Altivec regs and version
GLIBC_2_3_4.
[SHLIB_COMPAT (libc, GLIBC_2_3, GLIBC_2_3_4)](__novec_setcontext):
Compatible with GLIBC_2.3.3 release.
* sysdeps/unix/sysv/linux/powerpc/powerpc64/swapcontext.S
(__swapcontext): Upgrade to swap Altivec regs and version GLIBC_2_3_4.
[SHLIB_COMPAT (libc, GLIBC_2_3, GLIBC_2_3_4)](__novec_swapcontext):
Compatible with GLIBC_2.3.3 release.
* sysdeps/unix/sysv/linux/powerpc/powerpc64/ucontext_i.h
(SIGCONTEXT_V_REGS_PTR, SIGCONTEXT_V_RESERVE): Defined.
2004-01-12 Steven Munroe <sjmunroe@us.ibm.com>
* sysdeps/unix/sysv/linux/powerpc/powerpc32/Versions: Add GLIBC_2.3.4
versions for setcontext, getcontext, swapcontext, and makecontext.
* sysdeps/unix/sysv/linux/powerpc/powerpc32/getcontext.S
(__getcontext): Upgrade to save Altivec regs and version GLIBC_2_3_4.
[SHLIB_COMPAT (libc, GLIBC_2_3_3, GLIBC_2_3_4)](__novec_getcontext):
Compatible with GLIBC_2.3.3 release.
* sysdeps/unix/sysv/linux/powerpc/powerpc32/makecontext.S
(__makecontext): Upgrade to align for Altivec regs and version
GLIBC_2_3_4.
[SHLIB_COMPAT (libc, GLIBC_2_3_3, GLIBC_2_3_4)](__novec_makecontext):
Compatible with GLIBC_2.3.3 release.
* sysdeps/unix/sysv/linux/powerpc/powerpc32/setcontext.S
(__setcontext): Upgrade to restore Altivec regs and version
GLIBC_2_3_4.
[SHLIB_COMPAT (libc, GLIBC_2_3_3, GLIBC_2_3_4)](__novec_setcontext):
Compatible with GLIBC_2.3.3 release.
* sysdeps/unix/sysv/linux/powerpc/powerpc32/swapcontext.S
(__swapcontext): Upgrade to swap Altivec regs and version GLIBC_2_3_4.
[SHLIB_COMPAT (libc, GLIBC_2_3_3, GLIBC_2_3_4)]
(__novec_swapcontext): Compatible with GLIBC_2.3.3 release.
* sysdeps/unix/sysv/linux/powerpc/powerpc32/ucontext_i.h:
(_UC_VSCR, _UC_VRSAVE): Define.
(_FRAME_BACKCHAIN, _FRAME_LR_SAVE,_FRAME_PARM_SAVE1,_FRAME_PARM_SAVE2,
_FRAME_PARM_SAVE3, _FRAME_PARM_SAVE4): Defined.
* sysdeps/powerpc/powerpc32/sysdep.h: New file. Define powerpc32
specific macros copied from sysdeps/powerpc/sysdep.h.
* sysdeps/powerpc/powerpc64/sysdep.h: New file. Define powerpc64
specific macros.
* sysdeps/unix/powerpc/sysdep.h: Include <bits/wordsize.h>.
[__WORDSIZE==64]: Include sysdeps/powerpc/powerpc64/sysdep.h.
[!__WORDSIZE==64]: Include sysdeps/powerpc/powerpc32/sysdep.h.
* sysdeps/powerpc/elf/sysdep.h: Remove this file, because it overrides
the more-specific unix/sysv/linux/powerpc file and breaks everything.
* sysdeps/powerpc/sysdep.h (ALIGNARG, ASM_TYPE_DIRECTIVE,
ASM_SIZE_DIRECTIVE, CALL_MCOUNT, ENTRY, EALIGN, DO_CALL, JUMPTARGET,
PSEUDO, PSEUDO_RET, PSEUDO_END, L, C_TEXT): These macros moved here
here, put under [__ELF__].
2001-07-06 Paul Eggert <eggert@twinsun.com>
* manual/argp.texi: Remove ignored LGPL copyright notice; it's
not appropriate for documentation anyway.
* manual/libc-texinfo.sh: "Library General Public License" ->
"Lesser General Public License".
2001-07-06 Andreas Jaeger <aj@suse.de>
* All files under GPL/LGPL version 2: Place under LGPL version
2.1.
1999-10-10 Ulrich Drepper <drepper@cygnus.com>
* sysdeps/powerpc/sysdep.h: New file.
* sysdeps/unix/powerpc/sysdep.h: New file.
* sysdeps/unix/sysv/linux/powerpc/sysdep.h: Include
sysdeps/unix/powerpc/sysdep.h, not sysdeps/unix/sysdep.h.
* sysdeps/powerpc/fpu/s_copysign.S: Use symbolic names for registers
instead of bare numbers.
* sysdeps/powerpc/fpu/s_fabs.S: Likewise.
* sysdeps/powerpc/fpu/s_fmax.S: Likewise.
* sysdeps/powerpc/fpu/s_fmin.S: Likewise.
* sysdeps/hppa/sysdep.h: New file.
