Add optimizations for classification functions (32-bit and 64-bit) and
string functions (32-bit and 64-bit).
powerpc: Re-work the Implies structure
This patch tries to organize the implies files for ppc, since there are
a number of processors and most of them are compatible with each other
(backwards compatible).
Having in mind that we start the search for processor-specific files in
the sysdeps/unix/sysv/linux tree
(sysdeps/unix/sysv/linux/powerpc/powerpc[32|64]/[processor]/fpu to be
exact), we would like to grab any linux-specific code from that tree
prior to going through the other tree (sysdeps/powerpc/...).
For that, i removed the Implies files that were originally inside the
fpu directories and placed then in the non-fpu directories (still inside
the unix/sysv/linux tree). If no processor-specific/linux-specific files
could be found, we "imply" the other tree's (sysdeps/powerpc/...) fpu
directory for that specific processor AND also the non-fpu directory for
that same tree.
If, again, no processor-specific code is found, we read another Implies
file that will point to the most compatible processor that we should
grab code from, and so on, until we reach the power4 processor.
So, in summary, the Implies files will live inside these directories
now:
* sysdeps/unix/sysv/linux/powerpc/powerpc[32|64]/[processor]
* sysdeps/powerpc/powerpc[32|64]/[processor]
Practical example of the order we will use to pick power6-specific code
with the new structure.
sysdeps/unix/sysv/linux/powerpc/powerpc[32|64]/power6/fpu ->
sysdeps/unix/sysv/linux/powerpc/powerpc[32|64]/power6 ->
sysdeps/powerpc/powerpc[32|64]/power6/fpu ->
sysdeps/powerpc/powerpc[32|64]/power6 ->
sysdeps/powerpc/powerpc[32|64]/power5+/fpu ->
sysdeps/powerpc/powerpc[32|64]/power5+ ->
sysdeps/powerpc/powerpc[32|64]/power5/fpu ->
sysdeps/powerpc/powerpc[32|64]/power5 ->
sysdeps/powerpc/powerpc[32|64]/power4/fpu ->
sysdeps/powerpc/powerpc[32|64]/power4 (from here, it'll go to the
generic path as usual)
to be abstracted to remove the guards from common files and put the stack initialization into a header file that can be overridden. 2010-06-11 Ryan S. Arnold <rsa@us.ibm.com> * elf/dl-load.c (_dl_map_object_from_fd): Set default stack_flags to no-exec for ARCH_PPC64. * elf/dl-support.c [_dl_stack_flags]: Set default stack_flags to no-exec for ARCH_PPC64. * elf/rtld.c (_dl_starting_up): Set default stack_flags to no-exec for ARCH_PPC64.
Fix use of 64-bit insn in 32-bit memcpy for POWER6. 2009-08-31 Joshua W. Boyer <jwboyer@linux.vnet.ibm.com> * sysdeps/powerpc/powerpc32/power6/memcpy.S: Change srdi instruction to srwi in 32-bit memcpy for power6.
Fix incorrect use of cmpldi in 32-bit PPC code. The 32-bit PowerPC POWER6 memcpy uses the cmpldi insn when it should use a cmplwi. 2009-04-22 Ryan S. Arnold <rsa@us.ibm.com> [BZ #10107] * sysdeps/powerpc/powerpc32/power6/memcpy.S (memcpy): Replace cmpldi with cmplwi. * sysdeps/powerpc/powerpc32/power6/memset.S (memset): Likewise.
2009-09-05 H.J. Lu <hongjiu.lu@intel.com> * configure.in: Support binutils 2.20. * configure: regenerate.
2008-12-09 Jakub Jelinek <jakub@redhat.com> * pthread_mutex_lock.c (__pthread_mutex_lock): Handle only the fast path here, for robust/PI/PP mutexes call __pthread_mutex_lock_full. Don't use switch, instead use a series of ifs according to their probability. (__pthread_mutex_lock_full): New function. * pthread_mutex_unlock.c: Include assert.h. (__pthread_mutex_unlock_usercnt): Handle only the fast path here, for robust/PI/PP mutexes call __pthread_mutex_unlock_full. Don't use switch, instead use a series of ifs according to their probability. (__pthread_mutex_unlock_full): New function. * sysdeps/unix/sysv/linux/pthread_mutex_cond_lock.c (__pthread_mutex_lock_full): Define.
[BZ #6955] 2008-11-01 Ulrich Drepper <drepper@redhat.com> [BZ #6955] * pthread_mutex_lock.c: Add support for private PI mutexes. * pthread_mutex_timedlock.c: Likewise. * pthread_mutex_trylock.c: Likewise. * pthread_mutex_unlock.c: Likewise. Patch mostly by Ben Jackson <ben@ben.com>.
* ChangeLog: Updated to reflect following patch. Selective Cherry Pick of Roland McGrath's head commit 7967983fd42380b090a63b53fe57a05e1d1c0b4f. The configure.in and Makefile parts are the same as Roland's patch but since this is a backport to a GLIBC 2.8 branch the individual configure files are different so those were regenerated locally by Ryan. Roland's ChangeLog is still relevant: 2009-09-15 Roland McGrath <roland@redhat.com> * aclocal.m4 (GLIBC_PROVIDES): Add _AS_ECHO_PREPARE. * nptl/sysdeps/pthread/configure: Regenerated. * nptl/sysdeps/x86_64/configure: Regenerated. * sysdeps/unix/sysv/linux/configure: Regenerated. * sysdeps/x86_64/elf/configure: Regenerated. * Makefile (autoconf-it): Set execute bit only on top-level configure. * configure.in: Let add-ons set $libc_add_on_config_subdirs to have the effect of AC_CONFIG_SUBDIRS relative to the add-on directory. * aclocal.m4 (GLIBC_PROVIDES): Add AC_CONFIG_SUBDIRS. * configure: Regenerated. This fix makes configure recurse into an add-on's standalone directories where the directories in libc_add_on_config_subdirs (relative path to the add-on) have their own pre-configured 'configure' and 'Makefile.in' that is maintained 'out-of-project'.
than __iovector which is reserved in GCC 4.3. 2009-08-03 Ryan S. Arnold <rsa@us.ibm.com> * include/sys/uio.h (__readv, __libc_readv, __writev, __libc_writev): Change prototypes to use __iovec for struct iovec * parameters rather than __iovector which is a reserved word in GCC 4.3.