601650f878
* Tue May 31 2022 Arjun Shankar <arjun@redhat.com> - 2.34-35
- Sync with upstream branch release/2.34/master,
commit ff450cdbdee0b8cb6b9d653d6d2fa892de29be31:
- Fix deadlock when pthread_atfork handler calls pthread_atfork or dlclose
- x86: Fallback {str|wcs}cmp RTM in the ncmp overflow case [BZ #29127]
- string.h: fix __fortified_attr_access macro call [BZ #29162]
- linux: Add a getauxval test [BZ #23293]
- rtld: Use generic argv adjustment in ld.so [BZ #23293]
- S390: Enable static PIE
* Thu May 19 2022 Florian Weimer <fweimer@redhat.com> - 2.34-34
- Sync with upstream branch release/2.34/master,
commit ede8d94d154157d269b18f3601440ac576c1f96a:
- csu: Implement and use _dl_early_allocate during static startup
- Linux: Introduce __brk_call for invoking the brk system call
- Linux: Implement a useful version of _startup_fatal
- ia64: Always define IA64_USE_NEW_STUB as a flag macro
- Linux: Define MMAP_CALL_INTERNAL
- i386: Honor I386_USE_SYSENTER for 6-argument Linux system calls
- i386: Remove OPTIMIZE_FOR_GCC_5 from Linux libc-do-syscall.S
- elf: Remove __libc_init_secure
- Linux: Consolidate auxiliary vector parsing (redo)
- Linux: Include <dl-auxv.h> in dl-sysdep.c only for SHARED
- Revert "Linux: Consolidate auxiliary vector parsing"
- Linux: Consolidate auxiliary vector parsing
- Linux: Assume that NEED_DL_SYSINFO_DSO is always defined
- Linux: Remove DL_FIND_ARG_COMPONENTS
- Linux: Remove HAVE_AUX_SECURE, HAVE_AUX_XID, HAVE_AUX_PAGESIZE
- elf: Merge dl-sysdep.c into the Linux version
- elf: Remove unused NEED_DL_BASE_ADDR and _dl_base_addr
- x86: Optimize {str|wcs}rchr-evex
- x86: Optimize {str|wcs}rchr-avx2
- x86: Optimize {str|wcs}rchr-sse2
- x86: Cleanup page cross code in memcmp-avx2-movbe.S
- x86: Remove memcmp-sse4.S
- x86: Small improvements for wcslen
- x86: Remove AVX str{n}casecmp
- x86: Add EVEX optimized str{n}casecmp
- x86: Add AVX2 optimized str{n}casecmp
- x86: Optimize str{n}casecmp TOLOWER logic in strcmp-sse42.S
- x86: Optimize str{n}casecmp TOLOWER logic in strcmp.S
- x86: Remove strspn-sse2.S and use the generic implementation
- x86: Remove strpbrk-sse2.S and use the generic implementation
- x87: Remove strcspn-sse2.S and use the generic implementation
- x86: Optimize strspn in strspn-c.c
- x86: Optimize strcspn and strpbrk in strcspn-c.c
- x86: Code cleanup in strchr-evex and comment justifying branch
- x86: Code cleanup in strchr-avx2 and comment justifying branch
- x86_64: Remove bcopy optimizations
- x86-64: Remove bzero weak alias in SS2 memset
- x86_64/multiarch: Sort sysdep_routines and put one entry per line
- x86: Improve L to support L(XXX_SYMBOL (YYY, ZZZ))
- fortify: Ensure that __glibc_fortify condition is a constant [BZ #29141]
* Thu May 12 2022 Florian Weimer <fweimer@redhat.com> - 2.34-33
- Sync with upstream branch release/2.34/master,
commit 91c2e6c3db44297bf4cb3a2e3c40236c5b6a0b23:
- dlfcn: Implement the RTLD_DI_PHDR request type for dlinfo
- manual: Document the dlinfo function
- x86: Fix fallback for wcsncmp_avx2 in strcmp-avx2.S [BZ #28896]
- x86: Fix bug in strncmp-evex and strncmp-avx2 [BZ #28895]
- x86: Set .text section in memset-vec-unaligned-erms
- x86-64: Optimize bzero
- x86: Remove SSSE3 instruction for broadcast in memset.S (SSE2 Only)
- x86: Improve vec generation in memset-vec-unaligned-erms.S
- x86-64: Fix strcmp-evex.S
- x86-64: Fix strcmp-avx2.S
- x86: Optimize strcmp-evex.S
- x86: Optimize strcmp-avx2.S
- manual: Clarify that abbreviations of long options are allowed
- Add HWCAP2_AFP, HWCAP2_RPRES from Linux 5.17 to AArch64 bits/hwcap.h
- aarch64: Add HWCAP2_ECV from Linux 5.16
- Add SOL_MPTCP, SOL_MCTP from Linux 5.16 to bits/socket.h
- Update kernel version to 5.17 in tst-mman-consts.py
- Update kernel version to 5.16 in tst-mman-consts.py
- Update syscall lists for Linux 5.17
- Add ARPHRD_CAN, ARPHRD_MCTP to net/if_arp.h
- Update kernel version to 5.15 in tst-mman-consts.py
- Add PF_MCTP, AF_MCTP from Linux 5.15 to bits/socket.h
Resolves: #2091541
36 lines
1.1 KiB
Diff
36 lines
1.1 KiB
Diff
commit 190ea5f7e4e7e98b9b6e3f29835ae8b1f6a5442e
|
|
Author: Noah Goldstein <goldstein.w.n@gmail.com>
|
|
Date: Mon Feb 7 00:32:23 2022 -0600
|
|
|
|
x86: Remove SSSE3 instruction for broadcast in memset.S (SSE2 Only)
|
|
|
|
commit b62ace2740a106222e124cc86956448fa07abf4d
|
|
Author: Noah Goldstein <goldstein.w.n@gmail.com>
|
|
Date: Sun Feb 6 00:54:18 2022 -0600
|
|
|
|
x86: Improve vec generation in memset-vec-unaligned-erms.S
|
|
|
|
Revert usage of 'pshufb' in broadcast logic as it is an SSSE3
|
|
instruction and memset.S is restricted to only SSE2 instructions.
|
|
|
|
(cherry picked from commit 1b0c60f95bbe2eded80b2bb5be75c0e45b11cde1)
|
|
|
|
diff --git a/sysdeps/x86_64/memset.S b/sysdeps/x86_64/memset.S
|
|
index 34ee0bfdcb81fb39..954471e5a5bf225b 100644
|
|
--- a/sysdeps/x86_64/memset.S
|
|
+++ b/sysdeps/x86_64/memset.S
|
|
@@ -30,9 +30,10 @@
|
|
|
|
# define MEMSET_SET_VEC0_AND_SET_RETURN(d, r) \
|
|
movd d, %xmm0; \
|
|
- pxor %xmm1, %xmm1; \
|
|
- pshufb %xmm1, %xmm0; \
|
|
- movq r, %rax
|
|
+ movq r, %rax; \
|
|
+ punpcklbw %xmm0, %xmm0; \
|
|
+ punpcklwd %xmm0, %xmm0; \
|
|
+ pshufd $0, %xmm0, %xmm0
|
|
|
|
# define WMEMSET_SET_VEC0_AND_SET_RETURN(d, r) \
|
|
movd d, %xmm0; \
|