aboutsummaryrefslogtreecommitdiff
path: root/ports/sysdeps
AgeCommit message (Collapse)Author
2012-11-17Fix "set but not used" warnings from MIPS bits/atomic.h.Joseph Myers
2012-11-14Fix thumb2 build breakage in libcryptMarcus Shawcroft
2012-11-09AArch64 PortMarcus Shawcroft
2012-11-06tile: use memcpy in memmove when possibleChris Metcalf
2012-11-06Use memcpy in memmove when possibleMaxim Kuvyrkov
2012-11-06tile: use atomic op to unlock pthread_spinlock_tChris Metcalf
Atomic ops are issued directly from the core, rather than potentially sitting in the write buffer, so can improve the performance of other waiters. In addition, if we didn't end up pulling a copy of the cache line where the lock is into cache, by using an atomic op we don't have to acquire the cache line before we can unlock.
2012-11-06Make ieee754 fma tolerate architectures without exception support.Chris Metcalf
2012-11-06tile: support very large shared objectsChris Metcalf
With gcc 4.8 tilegx has support for -mcmodel=large, to tolerate very large shared objects. This option changes the compiler output to not include direct jump instructions, which have a range of only 2^30, i.e +/- 512MB. Instead the compiler marshalls the target PCs into registers and then uses jump- or call-to-register instructions. For glibc, the upshot is that we need to arrange for a few functions to tolerate the possibility of a large range between the PC and the target. In particular, the crti.S and start.S code needs to be able to reach from .init to the PLT, as does gmon-start.c. The elf-init.c code has the reverse problem, needing to call from libc_nonshared.a (linked at the end of shared objects) back to the _init section at the beginning. No other functions in *_nonshared.a need to be built this way, as they only call the PLT (or potentially each other), but all of that code is linked at the very end of the shared object. We don't build the standard -static archives with this option as the performance cost is high enough and the use case is rare enough that it doesn't seem worthwhile. Instead, we would encourage developers who need the -static model with huge executables to build a private copy of glibc and configure it with -mcmodel=large. Note that libc.so et al don't need any changes; the only changes are for code that is statically linked into user code built with -mcmodel=large. For the assembly code, I just rewrote it so that it unconditionally uses the large model. To be able to pass -mcmodel=large to csu/elf-init.c and csu/gmon-start.c, I need to check to see if the compiler supports that flag, since gcc 4.7 doesn't; I added the support by creating a small Makefile fragment that just runs the compiler to check.
2012-11-06tile: improve simulator notification for relative paths in dlopenChris Metcalf
Normally, the simulator is notified of absolute pathnames by the _dl_load_hook hook. However, when a relative pathname is used, the simulator may not know that the relative path matches a path that it could figure out in the file system that it has access to. Instead we provide a simplified version of the realpath function so we can pass a plausible absolute pathname to the simulator. Since we're now doing more work at object load time, we also add a guard so we do no work at all if we're not running on the simulator.
2012-11-06Optimize tile (mostly tilegx) memcpy and memmove performance.Chris Metcalf
- Override <memcopy.h> so we use full 8-byte word copies on tilegx32 for memmove, then use op_t in memcpy instead of the previous locally-defined word_t just to avoid proliferating identical types. - Fix bug in memcpy prefetch that caused us to never prefetch past the first cache line. - Optimize misaligned memcpy by inlining _wordcopy_fwd_dest_aligned instead of just doing a dumb word-at-a-time copy. - Make memcpy safe for forward copies by doing all the loads from a given cache line prior to doing a wh64 (cache line zero-fill) on the destination. Remove now-redundant src == dst check. - Copy and optimize the generic wordcopy.c routines to use the tile "double align" instruction instead of the MERGE macro; to avoid offset addressing mode (which tile doesn't have) by rewriting the pointer math to load and store with a zero index; and to use post-increment addresses in the inner loops to improve scheduling.
2012-11-05Define lll_futex_timed_wait_bitset for MIPS.Joseph Myers
2012-11-05Define lll_futex_timed_wait_bitset for ARM.Joseph Myers
2012-11-05Fix types of FE_DFL_ENV and FE_NOMASK_ENV (bug 14805).Joseph Myers
2012-11-04[BZ #157] Remove include/stub-tag.h for good.Thomas Schwinge
2012-11-03Make fma use of Dekker and Knuth algorithms use round-to-nearest (bug 14796).Joseph Myers
2012-11-03Make fenv.h FE_* macros usable in #if (bug 3439).Joseph Myers
2012-11-03tile: Account for new tests (missed test-double changes)Chris Metcalf
2012-11-01Regenerate all configure scripts.Thomas Schwinge
2012-11-01Convert alpha to use <bits/fcntl-linux.h>Andreas Jaeger
2012-11-01BZ#14743: Move clock_* symbols from librt to libc.Thomas Schwinge
Change missing from commit 6e6249d0b461b952d0f544792372663feb6d792a.
2012-10-31Fix powerpc abilist sort orderAndreas Schwab
2012-10-31Remove 'Contributed' line, glibc is not using them anymore.Steve Ellcey
2012-10-31Remove trailing white spaces.Steve Ellcey
2012-10-312012-10-31 Steve Ellcey <sellcey@mips.com>Steve Ellcey
* sysdeps/mips/memcpy.S: Add prefetching and more unrolling, make it work in 32 or 64 bit modes. * sysdeps/mips/mips64/memcpy.S: Remove.
2012-10-30Correct cacheline size to 32-bytes for ppc405 memset.S (bug 14595).Ryan S. Arnold
This patch also creates a version of memset.S for the ppc476 processor which uses a 128-byte cacheline size for dcbz insns.
2012-10-30Invoke DL_AFTER_LOAD if definedChris Metcalf
This hook is useful for any arch-specific functionality that should be done on loaded objects. For the tile architecture, the hook is already provided (though we switch to using the new macro name with this commit) and implements a simulator notifier so that the simulator can load Elf symbols to match the object and generate better error messages for PC's. Also, remove a spurious definition of DL_UNMAP in dl-runtime.c
2012-10-30tile: improve ABI and #include use in gettimeofday vdso supportChris Metcalf
2012-10-30tile: update libm-test-ulps for new testsChris Metcalf
2012-10-30Fix strtod handling of underflow (bug 14047).Joseph Myers
2012-10-29hppa: Add cfi direcvtives.Carlos O'Donell
We add cfi directives to the syscall assembly.
2012-10-29hppa: Fix r19 save and restore.Carlos O'Donell
We must save and restore r19 in both PIC and non-PIC situations since the kernel paths that clobber r19 are independent of that PIC-ness of userspace. In addition we choose r4 as the temporary register over r3 which is being used by recent gcc's as the frame pointer.
2012-10-29hppa: Optimize syscall sequence.Carlos O'Donell
We remove an extraneous nop from the syscall sequence. No DO_CALL* may end in an instruction with a delay slot.
2012-10-29Fix trailing whitespace.Carlos O'Donell
2012-10-292012-10-29 Steve Ellcey <sellcey@mips.com>Steve Ellcey
* sysdeps/unix/sysv/linux/mips/mips32/Makefile: Remove. * sysdeps/unix/sysv/linux/mips/mips64/n32/Makefile: Remove. * sysdeps/unix/sysv/linux/mips/mips64/n64/Makefile: Remove. * sysdeps/unix/sysv/linux/mips/configure.in: Create default_abi.make. * sysdeps/unix/sysv/linux/mips/configure: Regenerate. * sysdeps/unix/sysv/linux/mips/Makefile: Include default_abi.make. (abi-variants): Add hard and soft float versions. (abi-o32-options): Remove. (abi-o32-condition): Remove. (abi-n32-options): Remove. (abi-n32-condition): Remove. (abi-n64-options): Remove. (abi-n64-condition): Remove. (abi-o32_soft-options): New. (abi-o32_soft-condition): New. (abi-o32_hard-options): New. (abi-o32_hard-condition): New. (abi-n32_soft-options): New. (abi-n32_soft-condition): New. (abi-n32_hard-options): New. (abi-n32_hard-condition): New. (abi-n64_soft-options): New. (abi-n64_soft-condition): New. (abi-n64_hard-options): New. (abi-n64_hard-condition): New.
2012-10-292012-10-29 Steve Ellcey <sellcey@mips.com>Steve Ellcey
* sysdeps/unix/sysv/linux/mips/mips32/Makefile: Remove. * sysdeps/unix/sysv/linux/mips/mips64/n32/Makefile: Remove. * sysdeps/unix/sysv/linux/mips/mips64/n64/Makefile: Remove. * sysdeps/unix/sysv/linux/mips/configure.in: Create default_abi.make. * sysdeps/unix/sysv/linux/mips/configure: Regenerate. * sysdeps/unix/sysv/linux/mips/Makefile: Include default_abi.make. (abi-variants): Add hard and soft float versions. (abi-o32-options): Remove. (abi-o32-condition): Remove. (abi-n32-options): Remove. (abi-n32-condition): Remove. (abi-n64-options): Remove. (abi-n64-condition): Remove. (abi-o32_soft-options): New. (abi-o32_soft-condition): New. (abi-o32_hard-options): New. (abi-o32_hard-condition): New. (abi-n32_soft-options): New. (abi-n32_soft-condition): New. (abi-n32_hard-options): New. (abi-n32_hard-condition): New. (abi-n64_soft-options): New. (abi-n64_soft-condition): New. (abi-n64_hard-options): New. (abi-n64_hard-condition): New.
2012-10-29tile: add vDSO support for gettimeofday()Chris Metcalf
2012-10-27Remove redundant defines of __O_LARGEFILEAndreas Jaeger
* sysdeps/unix/sysv/linux/x86/bits/fcntl.h (__O_LARGEFILE) [!__x86_64]: Do not define, take value from <bits/fcntl-linux.h>. * sysdeps/unix/sysv/linux/s390/bits/fcntl.h (__O_LARGEFILE): [__WORDSIZE != 64]: Likewise. * sysdeps/unix/sysv/linux/generic/bits/fcntl.h: (__O_LARGEFILE) [__WORDSIZE != 64]: Do not define, take value from <bits/fcntl-linux.h>.
2012-10-26HPPA: Use <bits/fcntl-linux.h>Andreas Jaeger
* sysdeps/unix/sysv/linux/hppa/bits/fcntl.h: Remove all definitions and declarations that are provided by <bits/fcntl-linux.h> and include <bits/fcntl-linux.h>. (__O_PATH): Define.
2012-10-25Convert ia64, linux-generic, m68k <bits/fcntl.h> to use <bits/fcntl-linux.h>Andreas Jaeger
* sysdeps/unix/sysv/linux/m68k/bits/fcntl.h: Remove all definitions and declarations that are provided by <bits/fcntl-linux.h> and include <bits/fcntl-linux.h>. * sysdeps/unix/sysv/linux/generic/bits/fcntl.h: Remove all definitions and declarations that are provided by <bits/fcntl-linux.h> and include <bits/fcntl-linux.h>. * sysdeps/unix/sysv/linux/ia64/bits/fcntl.h: Remove all definitions and declarations that are provided by <bits/fcntl-linux.h> and include <bits/fcntl-linux.h>.
2012-10-24BZ#14743: Move clock_* symbols from librt to libc.Roland McGrath
2012-10-24Convert MIPS to use <bits/fcntl-linux.h>Andreas Jaeger
* sysdeps/unix/sysv/linux/mips/bits/fcntl.h: Remove all definitions and declarations that are provided by <bits/fcntl-linux.h> and include <bits/fcntl-linux.h>.
2012-10-24Don't set resource limits in tst-oddstacklimit-ENV.Joseph Myers
2012-10-22Always define __O_LARGEFILEAndreas Jaeger
2012-10-22ARM: Use <bits/fcntl-linux.h>Andreas Jaeger
* sysdeps/unix/sysv/linux/arm/bits/fcntl.h: Remove all definitions and declarations that are provided by <bits/fcntl-linux.h> and include <bits/fcntl-linux.h>.
2012-10-12tilegx: add optimized sched_getcpu() using TILE_COORD SPRChris Metcalf
We can discover our x,y coordinate in the core mesh with an mfspr instruction, multiply y by the core mesh width, and have the core number without needing to ask the kernel.
2012-10-10hppa: Update pthread.hCarlos O'Donell
Updates the hppa-specific pthread.h from the generic version. After this update the only difference between the generic version and the hppa version is the footer protected by the _PTHREAD_H_HPPA_ guard.
2012-10-10hppa: Remove pt-initfini.cCarlos O'Donell
Rmoves the unused pt-initfini.c
2012-10-10hppa: Restore nptl/shlib-versions.Carlos O'Donell
On 2012-01-07 the nptl/shlib-versions file removed the hppa entry. This restores that entry in the hppa ports tree.
2012-10-09Add missing magic to GLIBC_PROVIDES.Roland McGrath
2012-10-05Clean up R_ARM_PC24 handling.Roland McGrath
generated by cgit v1.2.3 (git 2.25.1) at 2024-11-05 05:39:30 +0000