Commit message (Collapse) | Author | Age | Files | Lines | ||
---|---|---|---|---|---|---|
... | ||||||
* | aarch64: fix start code for static pie | Szabolcs Nagy | 2017-12-18 | 1 | -1/+10 | |
| | | | | | | | | | | | | | | | | | | | | | | | | There are three flavors of the crt startup code: 1) crt1.o used for non-pie, 2) Scrt1.o used for dynamic linked pie (dynamic linker relocates), 3) rcrt1.o used for static linked pie (self relocation is needed) In the --enable-static-pie case crt1.o is built with -DPIC and in case of static linking it interposes _dl_relocate_static_pie in libc to avoid self relocation. Scrt1.o is built with -DPIC -DSHARED and it relies on GOT entries that the static linker cannot relax and thus need relocation before the start code is executed, so rcrt1.o needs separate implementation. This implementation does not work for .text > 4G position independent executables, which is fine since the toolchain does not support -mcmodel=large with -fPIE. Tests pass with ld/22269 and ld/22263 binutils bugs fixed. * sysdeps/aarch64/start.S (_start): Handle PIC && !SHARED case. | |||||
* | aarch64: Improve strcmp unaligned performance | Siddhesh Poyarekar | 2017-12-13 | 1 | -2/+29 | |
| | | | | | | | | | | | | | | Replace the simple byte-wise compare in the misaligned case with a dword compare with page boundary checks in place. For simplicity I've chosen a 4K page boundary so that we don't have to query the actual page size on the system. This results in up to 3x improvement in performance in the unaligned case on falkor and about 2.5x improvement on mustang as measured using bench-strcmp. * sysdeps/aarch64/strcmp.S (misaligned8): Compare dword at a time whenever possible. | |||||
* | aarch64: Avoid hidden symbols for memcpy/memmove into static binaries | Siddhesh Poyarekar | 2017-12-04 | 1 | -0/+2 | |
| | | | | | | | | | | | The __GI_* symbol aliases for __memcpy_generic are unnecessary since they're never used. Add them only for libc.so to avoid PLT. Maybe some time in future we need to evaluate the relative cost of PLT vs gains from multiarch memcpy implementations and take a call on whether to drop this completely. * sysdeps/aarch64/multiarch/memcpy_generic.S (__GI_memcpy): Define only for libc.so. | |||||
* | Use libm_alias_float for aarch64. | Joseph Myers | 2017-11-28 | 13 | -13/+26 | |
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Continuing the preparation for additional _FloatN / _FloatNx function aliases, this patch makes aarch64 libm function implementations use libm_alias_float to define function aliases. Tested with build-many-glibcs.py for aarch64-linux-gnu that installed stripped shared libraries are unchanged by the patch. * sysdeps/aarch64/fpu/s_ceilf.c: Include <libm-alias-float.h>. (ceilf): Define using libm_alias_float. * sysdeps/aarch64/fpu/s_floorf.c: Include <libm-alias-float.h>. (floorf): Define using libm_alias_float. * sysdeps/aarch64/fpu/s_fmaf.c: Include <libm-alias-float.h>. (fmaf): Define using libm_alias_float. * sysdeps/aarch64/fpu/s_fmaxf.c: Include <libm-alias-float.h>. (fmaxf): Define using libm_alias_float. * sysdeps/aarch64/fpu/s_fminf.c: Include <libm-alias-float.h>. (fminf): Define using libm_alias_float. * sysdeps/aarch64/fpu/s_llrintf.c: Include <libm-alias-float.h>. (llrintf): Define using libm_alias_float. * sysdeps/aarch64/fpu/s_llroundf.c: Include <libm-alias-float.h>. (llroundf): Define using libm_alias_float. * sysdeps/aarch64/fpu/s_lrintf.c: Include <libm-alias-float.h>. (lrintf): Define using libm_alias_float. * sysdeps/aarch64/fpu/s_lroundf.c: Include <libm-alias-float.h>. (lroundf): Define using libm_alias_float. * sysdeps/aarch64/fpu/s_nearbyintf.c: Include <libm-alias-float.h>. (nearbyintf): Define using libm_alias_float. * sysdeps/aarch64/fpu/s_rintf.c: Include <libm-alias-float.h>. (rintf): Define using libm_alias_float. * sysdeps/aarch64/fpu/s_roundf.c: Include <libm-alias-float.h>. (roundf): Define using libm_alias_float. * sysdeps/aarch64/fpu/s_truncf.c: Include <libm-alias-float.h>. (truncf): Define using libm_alias_float. | |||||
* | Use libm_alias_double for aarch64. | Joseph Myers | 2017-11-27 | 13 | -13/+26 | |
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Continuing the preparation for additional _FloatN / _FloatNx function aliases, this patch makes aarch64 libm function implementations use libm_alias_double to define function aliases. Tested with build-many-glibcs.py for aarch64-linux-gnu that installed stripped shared libraries are unchanged by the patch. * sysdeps/aarch64/fpu/s_ceil.c: Include <libm-alias-double.h>. (ceil): Define using libm_alias_double. * sysdeps/aarch64/fpu/s_floor.c: Include <libm-alias-double.h>. (floor): Define using libm_alias_double. * sysdeps/aarch64/fpu/s_fma.c: Include <libm-alias-double.h>. (fma): Define using libm_alias_double. * sysdeps/aarch64/fpu/s_fmax.c: Include <libm-alias-double.h>. (fmax): Define using libm_alias_double. * sysdeps/aarch64/fpu/s_fmin.c: Include <libm-alias-double.h>. (fmin): Define using libm_alias_double. * sysdeps/aarch64/fpu/s_llrint.c: Include <libm-alias-double.h>. (llrint): Define using libm_alias_double. * sysdeps/aarch64/fpu/s_llround.c: Include <libm-alias-double.h>. (llround): Define using libm_alias_double. * sysdeps/aarch64/fpu/s_lrint.c: Include <libm-alias-double.h>. (lrint): Define using libm_alias_double. * sysdeps/aarch64/fpu/s_lround.c: Include <libm-alias-double.h>. (lround): Define using libm_alias_double. * sysdeps/aarch64/fpu/s_nearbyint.c: Include <libm-alias-double.h>. (nearbyint): Define using libm_alias_double. * sysdeps/aarch64/fpu/s_rint.c: Include <libm-alias-double.h>. (rint): Define using libm_alias_double. * sysdeps/aarch64/fpu/s_round.c: Include <libm-alias-double.h>. (round): Define using libm_alias_double. * sysdeps/aarch64/fpu/s_trunc.c: Include <libm-alias-double.h>. (trunc): Define using libm_alias_double. | |||||
* | aarch64: Optimized memset for falkor | Siddhesh Poyarekar | 2017-11-20 | 9 | -22/+193 | |
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The generic memset reads dczid_el0 on every memset. This has a significant impact on falkor for a range of sizes because reading dczid_el0 is slow. The DZP bit in the dczid_el0 register does not change dynamically, so it is safe to read once during program startup. With this patch dczid_el0 is read once during startup and zva_size is cached. This is used to invoke the falkor-specific memset; the generic memset routine remains unchanged. The gains due to this are significant for falkor, with run time reductions as high as 48%. Here's a sample from the falkor tests: Function: memset Variant: walk simple_memset __memset_falkor __memset_generic ===================================================================== length=256, char=0: 139.96 (-698.28%) 9.07 ( 48.26%) 17.53 length=257, char=0: 140.50 (-699.03%) 9.53 ( 45.80%) 17.58 length=258, char=0: 140.96 (-703.95%) 9.58 ( 45.36%) 17.53 length=259, char=0: 141.56 (-705.16%) 9.53 ( 45.79%) 17.58 length=260, char=0: 142.15 (-710.76%) 9.57 ( 45.39%) 17.53 length=261, char=0: 142.50 (-710.39%) 9.53 ( 45.78%) 17.58 length=262, char=0: 142.97 (-715.09%) 9.57 ( 45.42%) 17.54 length=263, char=0: 143.51 (-716.18%) 9.53 ( 45.80%) 17.58 length=264, char=0: 143.93 (-720.55%) 9.58 ( 45.39%) 17.54 length=265, char=0: 144.56 (-722.07%) 9.53 ( 45.80%) 17.59 length=266, char=0: 144.98 (-726.42%) 9.58 ( 45.42%) 17.54 length=267, char=0: 145.53 (-727.53%) 9.53 ( 45.80%) 17.59 length=268, char=0: 146.25 (-731.81%) 9.53 ( 45.79%) 17.58 length=269, char=0: 146.52 (-735.39%) 9.53 ( 45.66%) 17.54 length=270, char=0: 146.97 (-735.81%) 9.53 ( 45.80%) 17.58 length=271, char=0: 147.54 (-741.08%) 9.58 ( 45.38%) 17.54 length=512, char=0: 268.26 (-1307.85%) 12.06 ( 36.71%) 19.05 length=513, char=0: 268.73 (-1273.89%) 13.56 ( 30.68%) 19.56 length=514, char=0: 269.31 (-1276.89%) 13.56 ( 30.68%) 19.56 length=515, char=0: 269.73 (-1279.05%) 13.56 ( 30.68%) 19.56 length=516, char=0: 270.34 (-1282.24%) 13.56 ( 30.67%) 19.56 length=517, char=0: 270.83 (-1284.71%) 13.56 ( 30.66%) 19.56 length=518, char=0: 271.20 (-1286.54%) 13.56 ( 30.67%) 19.56 length=519, char=0: 271.67 (-1288.67%) 13.65 ( 30.24%) 19.56 length=520, char=0: 272.14 (-1291.04%) 13.65 ( 30.22%) 19.56 length=521, char=0: 272.66 (-1293.69%) 13.65 ( 30.23%) 19.56 length=522, char=0: 273.14 (-1296.13%) 13.65 ( 30.20%) 19.56 length=523, char=0: 273.64 (-1298.75%) 13.65 ( 30.23%) 19.56 length=524, char=0: 274.34 (-1302.16%) 13.66 ( 30.20%) 19.57 length=525, char=0: 274.64 (-1297.78%) 13.56 ( 30.99%) 19.65 length=526, char=0: 275.20 (-1300.04%) 13.56 ( 31.01%) 19.66 length=527, char=0: 275.66 (-1302.86%) 13.56 ( 30.99%) 19.65 length=1024, char=0: 524.46 (-2169.75%) 20.12 ( 12.92%) 23.11 length=1025, char=0: 525.14 (-2124.63%) 21.62 ( 8.40%) 23.61 length=1026, char=0: 525.59 (-2125.36%) 21.88 ( 7.37%) 23.62 length=1027, char=0: 525.98 (-2127.14%) 21.62 ( 8.46%) 23.62 length=1028, char=0: 526.68 (-2131.10%) 21.62 ( 8.42%) 23.61 length=1029, char=0: 527.10 (-2131.70%) 21.79 ( 7.73%) 23.62 length=1030, char=0: 527.54 (-2118.51%) 21.62 ( 9.10%) 23.78 length=1031, char=0: 527.98 (-2136.37%) 21.62 ( 8.43%) 23.61 length=1032, char=0: 528.70 (-2139.38%) 21.62 ( 8.43%) 23.61 length=1033, char=0: 529.25 (-2124.37%) 21.62 ( 9.11%) 23.79 length=1034, char=0: 529.48 (-2142.95%) 21.62 ( 8.43%) 23.61 length=1035, char=0: 530.11 (-2145.13%) 21.62 ( 8.44%) 23.61 length=1036, char=0: 530.76 (-2147.10%) 21.79 ( 7.73%) 23.62 length=1037, char=0: 531.03 (-2149.45%) 21.62 ( 8.42%) 23.61 length=1038, char=0: 531.64 (-2151.87%) 21.62 ( 8.42%) 23.61 length=1039, char=0: 531.99 (-2151.63%) 21.80 ( 7.75%) 23.63 * sysdeps/aarch64/memset-reg.h: New file. * sysdeps/aarch64/memset.S: Use it. (__memset): Rename to MEMSET macro. [ZVA_MACRO]: Use zva_macro. * sysdeps/aarch64/multiarch/Makefile (sysdep_routines): Add memset_generic and memset_falkor. * sysdeps/aarch64/multiarch/ifunc-impl-list.c (__libc_ifunc_impl_list): Add memset ifuncs. * sysdeps/aarch64/multiarch/init-arch.h (INIT_ARCH): New local variable zva_size. * sysdeps/aarch64/multiarch/memset.c: New file. * sysdeps/aarch64/multiarch/memset_generic.S: New file. * sysdeps/aarch64/multiarch/memset_falkor.S: New file. * sysdeps/aarch64/multiarch/rtld-memset.S: New file. * sysdeps/unix/sysv/linux/aarch64/cpu-features.c (DCZID_DZP_MASK): New macro. (DCZID_BS_MASK): Likewise. (init_cpu_features): Read and set zva_size. * sysdeps/unix/sysv/linux/aarch64/cpu-features.h (struct cpu_features): New member zva_size. | |||||
* | aarch64: Fix f{max,min}{f} build for GCC 4.9 and 5 | Adhemerval Zanella | 2017-11-17 | 1 | -0/+6 | |
| | | | | | | | | | | | | | | | | | | | | | | | GCC 4.9 and 5 do not generate a correct f{max,min}nm instruction for __builtin_{fmax,fmin}{f} without -ffinite-math-only. It is clear a compiler issue since the instruction can handle NaN and Inf correctly and GCC6+ does not show this issue. We can backport a fix to GCC 5, raise the minimum required GCC version for aarch64 (since GCC 4.9 branch is now closed [1]) and/or add configure check to check for this issue. However I think -ffinite-math-only should be safe for these specific implementations and it is a simpler solution. Checked on aarch64-linux-gnu with GCC 5.3.1. * sysdeps/aarch64/fpu/Makefile (CFLAGS-s_fmax.c, CFLAGS-s_fmaxf.c, CFLAGS-s_fmin.c, CFLAGS-s_fminf.c): New rule: add -ffinite-math-only. [1] https://gcc.gnu.org/ml/gcc/2016-08/msg00010.html Signed-off-by: Adhemerval Zanella <adhemerval.zanella@linaro.org> Reviewed-by: Szabolcs Nagy <szabolcs.nagy@arm.com> | |||||
* | nptl: Define __PTHREAD_MUTEX_{NUSERS_AFTER_KIND,USE_UNION} | Adhemerval Zanella | 2017-11-07 | 1 | -0/+2 | |
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This patch adds two new internal defines to set the internal pthread_mutex_t layout required by the supported ABIS: 1. __PTHREAD_MUTEX_NUSERS_AFTER_KIND which control whether to define __nusers fields before or after __kind. The preferred value for is 0 for new ports and it sets __nusers before __kind. 2. __PTHREAD_MUTEX_USE_UNION which control whether internal __spins and __list members will be place inside an union for linuxthreads compatibility. The preferred value is 0 for ports and it sets to not use an union to define both fields. It fixes the wrong offsets value for __kind value on x86_64-linux-gnu-x32. Checked with a make check run-built-tests=no on all afected ABIs. [BZ #22298] * nptl/allocatestack.c (allocate_stack): Check if __PTHREAD_MUTEX_HAVE_PREV is non-zero, instead if __PTHREAD_MUTEX_HAVE_PREV is defined. * nptl/descr.h (pthread): Likewise. * nptl/nptl-init.c (__pthread_initialize_minimal_internal): Likewise. * nptl/pthread_create.c (START_THREAD_DEFN): Likewise. * sysdeps/nptl/fork.c (__libc_fork): Likewise. * sysdeps/nptl/pthread.h (PTHREAD_MUTEX_INITIALIZER): Likewise. * sysdeps/nptl/bits/thread-shared-types.h (__PTHREAD_MUTEX_NUSERS_AFTER_KIND, __PTHREAD_MUTEX_USE_UNION): New defines. (__pthread_internal_list): Check __PTHREAD_MUTEX_USE_UNION instead of __WORDSIZE for internal layout. (__pthread_mutex_s): Check __PTHREAD_MUTEX_NUSERS_AFTER_KIND instead of __WORDSIZE for internal __nusers layout and __PTHREAD_MUTEX_USE_UNION instead of __WORDSIZE whether to use an union for __spins and __list fields. (__PTHREAD_MUTEX_HAVE_PREV): Define also for __PTHREAD_MUTEX_USE_UNION case. * sysdeps/aarch64/nptl/bits/pthreadtypes-arch.h (__PTHREAD_MUTEX_NUSERS_AFTER_KIND, __PTHREAD_MUTEX_USE_UNION): New defines. * sysdeps/alpha/nptl/bits/pthreadtypes-arch.h (__PTHREAD_MUTEX_NUSERS_AFTER_KIND, __PTHREAD_MUTEX_USE_UNION): Likewise. * sysdeps/arm/nptl/bits/pthreadtypes-arch.h (__PTHREAD_MUTEX_NUSERS_AFTER_KIND, __PTHREAD_MUTEX_USE_UNION): Likewise. * sysdeps/hppa/nptl/bits/pthreadtypes-arch.h (__PTHREAD_MUTEX_NUSERS_AFTER_KIND, __PTHREAD_MUTEX_USE_UNION): Likewise. * sysdeps/ia64/nptl/bits/pthreadtypes-arch.h (__PTHREAD_MUTEX_NUSERS_AFTER_KIND, __PTHREAD_MUTEX_USE_UNION): Likewise. * sysdeps/m68k/nptl/bits/pthreadtypes-arch.h (__PTHREAD_MUTEX_NUSERS_AFTER_KIND, __PTHREAD_MUTEX_USE_UNION): Likewise. * sysdeps/microblaze/nptl/bits/pthreadtypes-arch.h (__PTHREAD_MUTEX_NUSERS_AFTER_KIND, __PTHREAD_MUTEX_USE_UNION): Likewise. * sysdeps/mips/nptl/bits/pthreadtypes-arch.h (__PTHREAD_MUTEX_NUSERS_AFTER_KIND, __PTHREAD_MUTEX_USE_UNION): Likewise. * sysdeps/nios2/nptl/bits/pthreadtypes-arch.h (__PTHREAD_MUTEX_NUSERS_AFTER_KIND, __PTHREAD_MUTEX_USE_UNION): Likewise. * sysdeps/powerpc/nptl/bits/pthreadtypes-arch.h (__PTHREAD_MUTEX_NUSERS_AFTER_KIND, __PTHREAD_MUTEX_USE_UNION): Likewise. * sysdeps/s390/nptl/bits/pthreadtypes-arch.h (__PTHREAD_MUTEX_NUSERS_AFTER_KIND, __PTHREAD_MUTEX_USE_UNION): Likewise. * sysdeps/sh/nptl/bits/pthreadtypes-arch.h (__PTHREAD_MUTEX_NUSERS_AFTER_KIND, __PTHREAD_MUTEX_USE_UNION): Likewise. * sysdeps/sparc/nptl/bits/pthreadtypes-arch.h (__PTHREAD_MUTEX_NUSERS_AFTER_KIND, __PTHREAD_MUTEX_USE_UNION): Likewise. * sysdeps/tile/nptl/bits/pthreadtypes-arch.h (__PTHREAD_MUTEX_NUSERS_AFTER_KIND, __PTHREAD_MUTEX_USE_UNION): Likewise. * sysdeps/x86/nptl/bits/pthreadtypes-arch.h (__PTHREAD_MUTEX_NUSERS_AFTER_KIND, __PTHREAD_MUTEX_USE_UNION): Likewise. Signed-off-by: Adhemerval Zanella <adhemerval.zanella@linaro.org> | |||||
* | nptl: Add tests for internal pthread_mutex_t offsets | Adhemerval Zanella | 2017-11-07 | 1 | -0/+5 | |
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This patch adds a new build test to check for internal fields offsets for user visible internal field. Although currently the only field which is statically initialized to a non zero value is pthread_mutex_t.__data.__kind value, the tests also check the offset of __kind, __spins, __elision (if supported), and __list internal member. A internal header (pthread-offset.h) is added to each major ABI with the reference value. Checked on x86_64-linux-gnu and with a build check for all affected ABIs (aarch64-linux-gnu, alpha-linux-gnu, arm-linux-gnueabihf, hppa-linux-gnu, i686-linux-gnu, ia64-linux-gnu, m68k-linux-gnu, microblaze-linux-gnu, mips64-linux-gnu, mips64-n32-linux-gnu, mips-linux-gnu, powerpc64le-linux-gnu, powerpc-linux-gnu, s390-linux-gnu, s390x-linux-gnu, sh4-linux-gnu, sparc64-linux-gnu, sparcv9-linux-gnu, tilegx-linux-gnu, tilegx-linux-gnu-x32, tilepro-linux-gnu, x86_64-linux-gnu, and x86_64-linux-x32). * nptl/pthreadP.h (ASSERT_PTHREAD_STRING, ASSERT_PTHREAD_INTERNAL_OFFSET): New macro. * nptl/pthread_mutex_init.c (__pthread_mutex_init): Add build time checks for internal pthread_mutex_t offsets. * sysdeps/aarch64/nptl/pthread-offsets.h (__PTHREAD_MUTEX_NUSERS_OFFSET, __PTHREAD_MUTEX_KIND_OFFSET, __PTHREAD_MUTEX_SPINS_OFFSET, __PTHREAD_MUTEX_ELISION_OFFSET, __PTHREAD_MUTEX_LIST_OFFSET): New macro. * sysdeps/alpha/nptl/pthread-offsets.h: Likewise. * sysdeps/arm/nptl/pthread-offsets.h: Likewise. * sysdeps/hppa/nptl/pthread-offsets.h: Likewise. * sysdeps/i386/nptl/pthread-offsets.h: Likewise. * sysdeps/ia64/nptl/pthread-offsets.h: Likewise. * sysdeps/m68k/nptl/pthread-offsets.h: Likewise. * sysdeps/microblaze/nptl/pthread-offsets.h: Likewise. * sysdeps/mips/nptl/pthread-offsets.h: Likewise. * sysdeps/nios2/nptl/pthread-offsets.h: Likewise. * sysdeps/powerpc/nptl/pthread-offsets.h: Likewise. * sysdeps/s390/nptl/pthread-offsets.h: Likewise. * sysdeps/sh/nptl/pthread-offsets.h: Likewise. * sysdeps/sparc/nptl/pthread-offsets.h: Likewise. * sysdeps/tile/nptl/pthread-offsets.h: Likewise. * sysdeps/x86_64/nptl/pthread-offsets.h: Likewise. Signed-off-by: Adhemerval Zanella <adhemerval.zanella@linaro.org> | |||||
* | aarch64: optimize _dl_tlsdesc_dynamic fast path | Szabolcs Nagy | 2017-11-03 | 1 | -54/+51 | |
| | | | | | | | | Remove some load/store instructions from the dynamic tlsdesc resolver fast path. This gives around 20% faster tls access in dlopened shared libraries (assuming glibc ran out of static tls space). * sysdeps/aarch64/dl-tlsdesc.S (_dl_tlsdesc_dynamic): Optimize. | |||||
* | aarch64: Remove barriers from TLS descriptor functions | Szabolcs Nagy | 2017-11-03 | 4 | -342/+1 | |
| | | | | | | | | | | | | | | | | | | | | Remove ldar synchronization and most lazy TLSDESC initialization related code. * sysdeps/aarch64/dl-machine.h (elf_machine_runtime_setup): Remove DT_TLSDESC_GOT initialization. * sysdeps/aarch64/dl-tlsdesc.S (_dl_tlsdesc_return_lazy): Remove. (_dl_tlsdesc_resolve_rela): Likewise. (_dl_tlsdesc_resolve_hold): Likewise. (_dl_tlsdesc_undefweak): Remove ldar. (_dl_tlsdesc_dynamic): Likewise. * sysdeps/aarch64/dl-tlsdesc.h (_dl_tlsdesc_return_lazy): Remove. (_dl_tlsdesc_resolve_rela): Likewise. (_dl_tlsdesc_resolve_hold): Likewise. * sysdeps/aarch64/tlsdesc.c (_dl_tlsdesc_resolve_rela_fixup): Remove. (_dl_tlsdesc_resolve_hold_fixup): Likewise. (_dl_tlsdesc_resolve_rela): Likewise. (_dl_tlsdesc_resolve_hold): Likewise. | |||||
* | aarch64: Disable lazy symbol binding of TLSDESC | Szabolcs Nagy | 2017-11-03 | 1 | -5/+14 | |
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Always do TLS descriptor initialization at load time during relocation processing to avoid barriers at every TLS access. In non-dlopened shared libraries the overhead of tls access vs static global access is > 3x bigger when lazy initialization is used (_dl_tlsdesc_return_lazy) compared to bind-now (_dl_tlsdesc_return) so the barriers dominate tls access performance. TLSDESC relocs are in DT_JMPREL which are processed at load time using elf_machine_lazy_rel which is only supposed to do lightweight initialization using the DT_TLSDESC_PLT trampoline (the trampoline code jumps to the entry point in DT_TLSDESC_GOT which does the lazy tlsdesc initialization at runtime). This patch changes elf_machine_lazy_rel in aarch64 to do the symbol binding and initialization as if DF_BIND_NOW was set, so the non-lazy code path of elf/do-rel.h was replicated. The static linker could be changed to emit TLSDESC relocs in DT_REL*, which are processed non-lazily, but the goal of this patch is to always guarantee bind-now semantics, even if the binary was produced with an old linker, so the barriers can be dropped in tls descriptor functions. After this change the synchronizing ldar instructions can be dropped as well as the lazy initialization machinery including the DT_TLSDESC_GOT setup. I believe this should be done on all targets, including ones where no barrier is needed for lazy initialization. There is very little gain in optimizing for large number of symbolic tlsdesc relocations which is an extremely uncommon case. And currently the tlsdesc entries are only readonly protected with -z now and some hardennings against writable JUMPSLOT relocs don't work for TLSDESC so they are a security hazard. (But to fix that the static linker has to be changed.) * sysdeps/aarch64/dl-machine.h (elf_machine_lazy_rel): Do symbol binding and initialization non-lazily for R_AARCH64_TLSDESC. | |||||
* | aarch64: Add missing math Makefile for recent commit | Szabolcs Nagy | 2017-10-23 | 1 | -0/+8 | |
| | | | | | Without -fno-math-errno, the builtins just do a call instead of inlining a single instruction. | |||||
* | aarch64: Implement math acceleration via builtins | Michael Collison | 2017-10-23 | 30 | -288/+250 | |
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This patch converts asm statements into builtins for AArch64. As an example for the file sysdeps/aarch64/fpu/s_ceil.c, we convert the function from double __ceil (double x) { double result; asm ("frintp\t%d0, %d1" : "=w" (result) : "w" (x) ); return result; } into double __ceil (double x) { return __builtin_ceil (x); } Tested on aarch64-linux-gnu with gcc-4.9.4 and gcc-6. * sysdeps/aarch64/fpu/e_sqrt.c (ieee754_sqrt): Replace asm statements with __builtin_sqrt. * sysdeps/aarch64/fpu/e_sqrtf.c (ieee754_sqrtf): Replace asm statements with __builtin_sqrtf. * sysdeps/aarch64/fpu/s_ceil.c (__ceil): Replace asm statements with __builtin_ceil. * sysdeps/aarch64/fpu/s_ceilf.c (__ceilf): Replace asm statements with __builtin_ceilf. * sysdeps/aarch64/fpu/s_floor.c (__floor): Replace asm statements with __builtin_floor. * sysdeps/aarch64/fpu/s_floorf.c (__floorf): Replace asm statements with __builtin_floorf. * sysdeps/aarch64/fpu/s_fma.c (__fma): Replace asm statements with __builtin_fma. * sysdeps/aarch64/fpu/s_fmaf.c (__fmaf): Replace asm statements with __builtin_fmaf. * sysdeps/aarch64/fpu/s_fmax.c (__fmax): Replace asm statements with __builtin_fmax. * sysdeps/aarch64/fpu/s_fmaxf.c (__fmaxf): Replace asm statements with __builtin_fmaxf. * sysdeps/aarch64/fpu/s_fmin.c (__fmin): Replace asm statements with __builtin_fmin. * sysdeps/aarch64/fpu/s_fminf.c (__fminf): Replace asm statements with __builtin_fminf. * sysdeps/aarch64/fpu/s_frint.c: Delete file. * sysdeps/aarch64/fpu/s_frintf.c: Delete file. * sysdeps/aarch64/fpu/s_llrint.c (__llrint): Replace asm statements with builtin_rint and conversion to int. * sysdeps/aarch64/fpu/s_llrintf.c (__llrintf): Likewise. * sysdeps/aarch64/fpu/s_llround.c (__llround): Replace asm statements with builtin_llround. * sysdeps/aarch64/fpu/s_llroundf.c (__llroundf): Likewise. * sysdeps/aarch64/fpu/s_lrint.c (__lrint): Replace asm statements with builtin_rint and conversion to long int. * sysdeps/aarch64/fpu/s_lrintf.c (__lrintf): Likewise. * sysdeps/aarch64/fpu/s_lround.c (__lround): Replace asm statements with builtin_lround. * sysdeps/aarch64/fpu/s_lroundf.c (__lroundf): Replace asm statements with builtin_lroundf. * sysdeps/aarch64/fpu/s_nearbyint.c (__nearbyint): Replace asm statements with __builtin_nearbyint. * sysdeps/aarch64/fpu/s_nearbyintf.c (__nearbyintf): Replace asm statements with __builtin_nearbyintf. * sysdeps/aarch64/fpu/s_rint.c (__rint): Replace asm statements with __builtin_rint. * sysdeps/aarch64/fpu/s_rintf.c (__rintf): Replace asm statements with __builtin_rintf. * sysdeps/aarch64/fpu/s_round.c (__round): Replace asm statements with __builtin_round. * sysdeps/aarch64/fpu/s_roundf.c (__roundf): Replace asm statements with __builtin_roundf. * sysdeps/aarch64/fpu/s_trunc.c (__trunc): Replace asm statements with __builtin_trunc. * sysdeps/aarch64/fpu/s_truncf.c (__truncf): Replace asm statements with __builtin_truncf. * sysdeps/aarch64/fpu/Makefile: Build e_sqrt[f].c with -fno-math-errno. | |||||
* | [AARCH64] Rewrite elf_machine_load_address using _DYNAMIC symbol | Szabolcs Nagy | 2017-10-18 | 1 | -34/+5 | |
| | | | | | | | | | | | | | | | | | | | | This patch rewrites aarch64 elf_machine_load_address to use special _DYNAMIC symbol instead of _dl_start. The static address of _DYNAMIC symbol is stored in the first GOT entry. Here is the change which makes this solution work (part of binutils 2.24): https://sourceware.org/ml/binutils/2013-06/msg00248.html i386, x86_64 targets use the same method to do this as well. The original implementation relies on a trick that R_AARCH64_ABS32 relocation being resolved at link time and the static address fits in the 32bits. However, in LP64, normally, the address is defined to be 64 bit. Here is the C version one which should be portable in all cases. * sysdeps/aarch64/dl-machine.h (elf_machine_load_address): Use _DYNAMIC symbol to calculate load address. | |||||
* | aarch64: Optimized implementation of memmove for Qualcomm Falkor | Siddhesh Poyarekar | 2017-10-05 | 4 | -2/+241 | |
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This is an optimized memmove implementation for the Qualcomm Falkor processor core. Due to the way the falkor memcpy needs to be written, code cannot be easily shared between memmove and memcpy like in case of other aarch64 memcpy implementations due to which this routine is separate. The underlying principle is the same as that of memcpy where it tries to use registers with the same lower 4 bits for fetching the same stream, thus optimizing hardware prefetcher performance. The memcpy copy loop copies 64 bytes at a time using the same register pair since that's the way to train the hardware prefetcher on the falkor core. memmove cannot quite do that since it needs to avoid overlaps, so it does the next best thing, i.e. has a 32 byte loop with a 32 byte end (prefetch a loop ahead to account for overlapping locations) with register pairs that alias so that they hit the same prefetcher. Due to this difference in loop size, they have to currently be separate implementations but efforts are on to try and get memmove to fall back into memcpy whenever it can without simply duplicating all of the code. Performance: The routine fares around 20-25% better than the generic memmove for most medium to large sizes (i.e. > 128 bytes) for the new walking memmove benchmark (memmove-walk) with an unexplained regression between 1K and 2K. The minor regression is something worth looking into for us, but the remaining gains are significant enough that we would like this included upstream as we looking into the cause for the regression. Here is a snippet of the numbers as generated from the microbenchmark by the compare_strings script. Comparisons are against __memmove_generic: Function: memmove Variant: walk __memmove_thunderx __memmove_falkor __memmove_generic ======================================================================================================================== <snip> length=16384: 12508800.00 ( 6.09%) 11486800.00 ( 13.76%) 13319600.00 length=16400: 13614200.00 ( -0.67%) 11585000.00 ( 14.33%) 13523600.00 length=16385: 13448400.00 ( 0.10%) 11732700.00 ( 12.84%) 13461200.00 length=16399: 13594100.00 ( -0.22%) 11859600.00 ( 12.57%) 13564400.00 length=16386: 13211600.00 ( 1.13%) 11503800.00 ( 13.91%) 13362400.00 length=16398: 13218600.00 ( 2.12%) 11573200.00 ( 14.30%) 13504700.00 length=16387: 13510900.00 ( -0.37%) 11744200.00 ( 12.76%) 13461300.00 length=16397: 13603700.00 ( -0.15%) 11878200.00 ( 12.55%) 13583200.00 length=16388: 13461700.00 ( -0.13%) 11558000.00 ( 14.03%) 13444100.00 length=16396: 13517500.00 ( -0.03%) 11561300.00 ( 14.45%) 13513900.00 length=16389: 13534100.00 ( 0.17%) 11756800.00 ( 13.28%) 13556900.00 length=16395: 13585600.00 ( 0.11%) 11791800.00 ( 13.30%) 13601200.00 length=16390: 13480100.00 ( -0.13%) 11685500.00 ( 13.20%) 13462100.00 length=16394: 13529900.00 ( -0.23%) 11549800.00 ( 14.43%) 13498200.00 length=16391: 13595400.00 ( -0.26%) 11768200.00 ( 13.22%) 13560600.00 length=16393: 13567000.00 ( 0.20%) 11779700.00 ( 13.35%) 13594700.00 length=32768: 71308800.00 ( -6.53%) 50220800.00 ( 24.98%) 66939200.00 length=32784: 72100800.00 (-11.55%) 50114100.00 ( 22.47%) 64636300.00 length=32769: 71767000.00 ( -7.10%) 51238400.00 ( 23.54%) 67010000.00 length=32783: 70113700.00 (-40.95%) 51129000.00 ( -2.78%) 49744400.00 length=32770: 71367600.00 ( -6.52%) 50244700.00 ( 25.01%) 67000900.00 length=32782: 64366700.00 ( 4.71%) 50101400.00 ( 25.83%) 67545600.00 length=32771: 71440100.00 ( -6.51%) 51263900.00 ( 23.57%) 67074900.00 length=32781: 66993000.00 ( 0.34%) 51108300.00 ( 23.97%) 67220300.00 length=32772: 71443900.00 (-60.50%) 50062100.00 (-12.47%) 44512600.00 length=32780: 71759100.00 ( -6.58%) 50263200.00 ( 25.35%) 67328600.00 length=32773: 71714900.00 (-33.21%) 51076600.00 ( 5.12%) 53835400.00 length=32779: 71756900.00 ( -6.56%) 51290800.00 ( 23.83%) 67337800.00 length=32774: 59689300.00 (-34.55%) 50068400.00 (-12.86%) 44363300.00 length=32778: 71847500.00 (-18.20%) 50084100.00 ( 17.61%) 60786500.00 length=32775: 71599300.00 ( -6.54%) 51278200.00 ( 23.70%) 67204800.00 length=32777: 71862900.00 (-60.85%) 51094000.00 (-14.36%) 44677900.00 length=65536: 282848000.00 ( -6.60%) 199187000.00 ( 24.93%) 265325000.00 length=65552: 243285000.00 (-41.61%) 198512000.00 (-15.54%) 171805000.00 length=65537: 255415000.00 (-23.47%) 202499000.00 ( 2.11%) 206858000.00 length=65551: 280122000.00 (-62.95%) 203349000.00 (-18.29%) 171911000.00 length=65538: 283676000.00 (-14.46%) 198368000.00 ( 19.96%) 247848000.00 length=65550: 275566000.00 (-51.76%) 198494000.00 ( -9.31%) 181581000.00 length=65539: 283699000.00 ( -6.58%) 203453000.00 ( 23.57%) 266195000.00 length=65549: 286572000.00 ( -6.65%) 202607000.00 ( 24.60%) 268712000.00 length=65540: 283710000.00 ( -6.59%) 199161000.00 ( 25.17%) 266160000.00 length=65548: 237573000.00 ( 11.48%) 198462000.00 ( 26.06%) 268395000.00 length=65541: 284150000.00 ( -6.58%) 203273000.00 ( 23.75%) 266600000.00 length=65547: 286250000.00 ( -6.70%) 202594000.00 ( 24.48%) 268263000.00 length=65542: 284167000.00 ( -6.60%) 199122000.00 ( 25.31%) 266584000.00 length=65546: 285656000.00 ( -6.59%) 198443000.00 ( 25.95%) 268002000.00 length=65543: 284600000.00 ( -6.58%) 203247000.00 ( 23.89%) 267030000.00 length=65545: 285665000.00 ( -6.40%) 202575000.00 ( 24.55%) 268472000.00 <snip> * sysdeps/aarch64/multiarch/Makefile (sysdep_routines): Add memmove_falkor. * sysdeps/aarch64/multiarch/ifunc-impl-list.c (__libc_ifunc_impl_list): Likewise. * sysdeps/aarch64/multiarch/memmove.c: Likewise. * sysdeps/aarch64/multiarch/memmove_falkor.S: New file. | |||||
* | aarch64: don't use MIN in dl-machine.h | Szabolcs Nagy | 2017-10-04 | 1 | -1/+2 | |
| | | | | | | | MIN is used, but param.h may not be included, so expand its single use inline. * sysdeps/aarch64/dl-machine.h (elf_machine_rela): Expand MIN. | |||||
* | AArch64: update libm-test-ulps | Szabolcs Nagy | 2017-09-28 | 1 | -2/+8 | |
| | | | | | | Update for new expf and logf. * sysdeps/aarch64/libm-test-ulps: Update. | |||||
* | Optimized generic expf and exp2f with wrappers | Szabolcs Nagy | 2017-09-25 | 1 | -0/+20 | |
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Based on new expf and exp2f code from https://github.com/ARM-software/optimized-routines/ with wrapper on aarch64: expf reciprocal-throughput: 2.3x faster expf latency: 1.7x faster without wrapper on aarch64: expf reciprocal-throughput: 3.3x faster expf latency: 1.7x faster without wrapper on aarch64: exp2f reciprocal-throughput: 2.8x faster exp2f latency: 1.3x faster libm.so size on aarch64: .text size: -152 bytes .rodata size: -1740 bytes expf/exp2f worst case nearest rounding error: 0.502 ulp worst case non-nearest rounding error: 1 ulp Error checks are inline and errno setting is in separate tail called functions, but the wrappers are kept in this patch to handle the _LIB_VERSION==_SVID_ case. (So e.g. errno is set twice for expf calls and once for __expf_finite calls on targets where the new code is used.) Double precision arithmetics is used which is expected to be faster on most targets (including soft-float) than using single precision and it is easier to get good precision result with it. Const data is kept in a separate translation unit which complicates maintenance a bit, but is expected to give good code for literal loads on most targets and allows sharing data across expf, exp2f and powf. (This data is disabled on i386, m68k and ia64 which have their own expf, exp2f and powf code.) Some details may need target specific tweaks: - best convert and round to int operation in the arg reduction may be different across targets. - code was optimized on fma target, optimal polynomial eval may be different without fma. - gcc does not always generate good code for fp bit representation access via unions or it may be inherently slow on some targets. The libm-test-ulps will need adjustment because.. - The argument reduction ideally uses nearest rounded rint, but that is not efficient on most targets, so the polynomial can get evaluated on a wider interval in non-nearest rounding mode making 1 ulp errors common in that case. - The polynomial is evaluated such that it may have 1 ulp error on negative tiny inputs with upward rounding. * math/Makefile (type-float-routines): Add math_errf and e_exp2f_data. * sysdeps/aarch64/fpu/math_private.h (TOINT_INTRINSICS): Define. (roundtoint, converttoint): Likewise. * sysdeps/ieee754/flt-32/e_expf.c: New implementation. * sysdeps/ieee754/flt-32/e_exp2f.c: New implementation. * sysdeps/ieee754/flt-32/e_exp2f_data.c: New file. * sysdeps/ieee754/flt-32/math_config.h: New file. * sysdeps/ieee754/flt-32/math_errf.c: New file. * sysdeps/ieee754/flt-32/t_exp2f.h: Remove. * sysdeps/i386/fpu/e_exp2f_data.c: New file. * sysdeps/i386/fpu/math_errf.c: New file. * sysdeps/ia64/fpu/e_exp2f_data.c: New file. * sysdeps/ia64/fpu/math_errf.c: New file. * sysdeps/m68k/m680x0/fpu/e_exp2f_data.c: New file. * sysdeps/m68k/m680x0/fpu/math_errf.c: New file. | |||||
* | Enable unwind info in libc-start.c and backtrace.c | Wilco Dijkstra | 2017-09-19 | 1 | -4/+0 | |
| | | | | | | | | | | | | | | | | | Add unwind info to __libc_start_main so that unwinding continues one extra level to _start. Similarly add unwind info to backtrace. Given many targets require this, do this in a general way. * csu/Makefile: Add -funwind-tables to libc-start.c. * debug/Makefile: Add -funwind-tables to backtrace.c. * sysdeps/aarch64/Makefile: Remove CFLAGS-backtrace.c. * sysdeps/arm/Makefile: Likewise. * sysdeps/i386/Makefile: Likewise. * sysdeps/m68k/Makefile: Likewise. * sysdeps/mips/Makefile: Likewise. * sysdeps/nios2/Makefile: Likewise. * sysdeps/sh/Makefile: Likewise. * sysdeps/sparc/Makefile: Likewise. | |||||
* | AArch64: use movz/movk instead of literal pools in start.S | Wang Boshi | 2017-09-18 | 2 | -3/+17 | |
| | | | | | | | | | | | | | | | eXecute-Only Memory (XOM) is a protection mechanism against some ROP attacks. XOM sets the code as executable and unreadable, so the access to any data, like literal pools, in the code section causes the fault with XOM. The compiler can disable literal pools for C source files, but not for assembly files, so I use movz/movk instead of literal pools in start.S for XOM. I add MOVL macro with movz/movk instructions like movl pseudo-instruction in armasm, and use the macro instead of literal pools. * sysdeps/aarch64/start.S: Use MOVL instead of literal pools. * sysdeps/aarch64/sysdep.h (MOVL): Add MOVL macro. | |||||
* | Obsolete pow10 functions. | Joseph Myers | 2017-09-01 | 1 | -30/+0 | |
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This patch obsoletes the pow10, pow10f and pow10l functions (makes them into compat symbols, not available for new ports or static linking). The exp10 names for these functions are standardized (in TS 18661-4) and were added in the same glibc version (2.1) as pow10 so source code can change to use them without any loss of portability. Since pow10 is deliberately not provided for _Float128, only exp10, this slightly simplifies moving to the new wrapper templates in the !LIBM_SVID_COMPAT case, by avoiding needing to arrange for pow10, pow10f and pow10l to be defined by those templates. Tested for x86_64, and with build-many-glibcs.py. * manual/math.texi (pow10): Do not document. (pow10f): Likewise. (pow10l): Likewise. * math/bits/mathcalls.h [__USE_GNU] (pow10): Do not declare. * math/bits/math-finite.h [__USE_GNU] (pow10): Likewise. * math/libm-test-exp10.inc (pow10_test): Remove. (do_test): Do not call pow10. * math/w_exp10_compat.c (pow10): Make into compat symbol. [NO_LONG_DOUBLE] (pow10l): Likewise. * math/w_exp10f_compat.c (pow10f): Likewise. * math/w_exp10l_compat.c (pow10l): Likewise. * sysdeps/ia64/fpu/e_exp10.S: Include <shlib-compat.h>. (pow10): Make into compat symbol. * sysdeps/ia64/fpu/e_exp10f.S: Include <shlib-compat.h>. (pow10f): Make into compat symbol. * sysdeps/ia64/fpu/e_exp10l.S: Include <shlib-compat.h>. (pow10l): Make into compat symbol. * sysdeps/ieee754/ldbl-opt/Makefile (libnldbl-calls): Remove pow10. (CFLAGS-nldbl-pow10.c): Remove variable.. * sysdeps/ieee754/ldbl-opt/nldbl-pow10.c: Remove file. * sysdeps/ieee754/ldbl-opt/w_exp10_compat.c (pow10l): Condition on [SHLIB_COMPAT (libm, GLIBC_2_1, GLIBC_2_27)]. * sysdeps/ieee754/ldbl-opt/w_exp10l_compat.c (compat_symbol): Undefine and redefine. (pow10l): Make into compat symbol. * sysdeps/aarch64/libm-test-ulps: Remove pow10 ulps. * sysdeps/alpha/fpu/libm-test-ulps: Likewise. * sysdeps/arm/libm-test-ulps: Likewise. * sysdeps/hppa/fpu/libm-test-ulps: Likewise. * sysdeps/i386/fpu/libm-test-ulps: Likewise. * sysdeps/i386/i686/fpu/multiarch/libm-test-ulps: Likewise. * sysdeps/microblaze/libm-test-ulps: Likewise. * sysdeps/mips/mips32/libm-test-ulps: Likewise. * sysdeps/mips/mips64/libm-test-ulps: Likewise. * sysdeps/nios2/libm-test-ulps: Likewise. * sysdeps/powerpc/fpu/libm-test-ulps: Likewise. * sysdeps/powerpc/nofpu/libm-test-ulps: Likewise. * sysdeps/s390/fpu/libm-test-ulps: Likewise. * sysdeps/sh/libm-test-ulps: Likewise. * sysdeps/sparc/fpu/libm-test-ulps: Likewise. * sysdeps/tile/libm-test-ulps: Likewise. * sysdeps/x86_64/fpu/libm-test-ulps: Likewise. | |||||
* | ILP32 math changes | Steve Ellcey | 2017-08-31 | 8 | -12/+87 | |
| | | | | | | | | | | | | | | | | | | | | | | * sysdeps/aarch64/fpu/s_llrint.c (OREG_SIZE): New macro. * sysdeps/aarch64/fpu/s_llround.c (OREG_SIZE): Likewise. * sysdeps/aarch64/fpu/s_llrintf.c (OREGS, IREGS): Remove. (IREG_SIZE, OREG_SIZE): New macros. * sysdeps/aarch64/fpu/s_llroundf.c: (OREGS, IREGS): Remove. (IREG_SIZE, OREG_SIZE): New macros. * sysdeps/aarch64/fpu/s_lrintf.c (IREGS): Remove. (IREG_SIZE): New macro. * sysdeps/aarch64/fpu/s_lroundf.c (IREGS): Remove. (IREG_SIZE): New macro. * sysdeps/aarch64/fpu/s_lrint.c (get-rounding-mode.h, stdint.h): New includes. (IREG_SIZE, OREG_SIZE): Initialize if not already set. (OREGS, IREGS): Set based on IREG_SIZE and OREG_SIZE. (__CONCATX): Handle exceptions correctly on large values that may set FE_INVALID. * sysdeps/aarch64/fpu/s_lround.c (IREG_SIZE, OREG_SIZE): Initialize if not already set. (OREGS, IREGS): Set based on IREG_SIZE and OREG_SIZE. | |||||
* | Change argument type passed to ifunc resolvers | Steve Ellcey | 2017-08-31 | 1 | -1/+1 | |
| | | | | | * sysdeps/aarch64/dl-irel.h: (elf_ifunc_invoke): Change argument type in resolver call. | |||||
* | elf: Remove internal_function attribute | Florian Weimer | 2017-08-31 | 2 | -3/+1 | |
| | ||||||
* | aarch64: Use PTR_REG macro to fix ILP32 bug and make code consistent | Steve Ellcey | 2017-08-22 | 1 | -1/+1 | |
| | | | | | * sysdeps/aarch64/dl-tlsdesc.S (_dl_tlsdesc_dynamic): Use PTR_REG macro in cmp instruction. | |||||
* | [AArch64] Optimized memcmp. | Wilco Dijkstra | 2017-08-10 | 1 | -105/+71 | |
| | | | | | | | | | | | | | | | | | | | | This is an optimized memcmp for AArch64. This is a complete rewrite using a different algorithm. The previous version split into cases where both inputs were aligned, the inputs were mutually aligned and unaligned using a byte loop. The new version combines all these cases, while small inputs of less than 8 bytes are handled separately. This allows the main code to be sped up using unaligned loads since there are now at least 8 bytes to be compared. After the first 8 bytes, align the first input. This ensures each iteration does at most one unaligned access and mutually aligned inputs behave as aligned. After the main loop, process the last 8 bytes using unaligned accesses. This improves performance of (mutually) aligned cases by 25% and unaligned by >500% (yes >6 times faster) on large inputs. * sysdeps/aarch64/memcmp.S (memcmp): Rewrite of optimized memcmp. | |||||
* | memcpy_falkor: Fix code style in comments | Siddhesh Poyarekar | 2017-08-09 | 1 | -7/+4 | |
| | ||||||
* | aarch64: Optimized memcpy for Qualcomm Falkor processor | Siddhesh Poyarekar | 2017-08-09 | 4 | -3/+196 | |
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This is an optimized implementation of the memcpy routine that gives a significant gain in performance for all sizes of copies on the Qualcomm Falkor processor. A detailed rationale of the implementation is written in a comment in the patch. This implementation improves time for copies up to 128 bytes by up to 15% and for larger copies by up to 35% in the glibc microbenchmark. The memcpy-random benchmark sees improvements in all sizes in the range of 13%-18%. Here are the full numbers extracted from the glibc microbenchmark using the commands: ../benchtests/scripts/compare_strings.py benchtests/bench-memcpy.out \ ../benchtests/scripts/benchout_strings.schema.json \ -base=__memcpy_generic length align1 align2 ../benchtests/scripts/compare_strings.py benchtests/bench-memcpy-large.out \ ../benchtests/scripts/benchout_strings.schema.json \ -base=__memcpy_generic length align1 align2 ../benchtests/scripts/compare_strings.py benchtests/bench-memcpy-random.out \ ../benchtests/scripts/benchout_strings.schema.json \ -base=__memcpy_generic max-size Function: memcpy __memcpy_thunderx __memcpy_falkor __memcpy_generic Variant: default ================================================================================ length=1,align1=0,align2=0: 33.59 (-115.00%) 15.62 (0.00%) 15.62 length=1,align1=0,align2=0: 16.41 (-10.53%) 14.06 (5.26%) 14.84 length=1,align1=0,align2=0: 14.84 (0.00%) 14.84 (0.00%) 14.84 length=1,align1=0,align2=0: 15.62 (-5.26%) 14.06 (5.26%) 14.84 length=2,align1=0,align2=0: 15.62 (-5.26%) 14.06 (5.26%) 14.84 length=2,align1=1,align2=0: 15.62 (-5.26%) 14.06 (5.26%) 14.84 length=2,align1=0,align2=1: 14.84 (0.00%) 14.06 (5.26%) 14.84 length=2,align1=1,align2=1: 14.84 (-5.56%) 14.06 (0.00%) 14.06 length=4,align1=0,align2=0: 14.06 (0.00%) 14.06 (0.00%) 14.06 length=4,align1=2,align2=0: 14.06 (-5.88%) 14.06 (-5.88%) 13.28 length=4,align1=0,align2=2: 14.06 (0.00%) 14.06 (0.00%) 14.06 length=4,align1=2,align2=2: 14.06 (-5.88%) 14.06 (-5.88%) 13.28 length=8,align1=0,align2=0: 14.84 (-5.56%) 13.28 (5.56%) 14.06 length=8,align1=3,align2=0: 14.06 (0.00%) 13.28 (5.56%) 14.06 length=8,align1=0,align2=3: 13.28 (0.00%) 13.28 (0.00%) 13.28 length=8,align1=3,align2=3: 13.28 (-6.25%) 13.28 (-6.25%) 12.50 length=16,align1=0,align2=0: 13.28 (0.00%) 13.28 (0.00%) 13.28 length=16,align1=4,align2=0: 13.28 (0.00%) 12.50 (5.88%) 13.28 length=16,align1=0,align2=4: 13.28 (0.00%) 13.28 (0.00%) 13.28 length=16,align1=4,align2=4: 13.28 (-6.25%) 12.50 (0.00%) 12.50 length=32,align1=0,align2=0: 14.06 (0.00%) 12.50 (11.11%) 14.06 length=32,align1=5,align2=0: 13.28 (0.00%) 12.50 (5.88%) 13.28 length=32,align1=0,align2=5: 14.06 (-5.88%) 12.50 (5.88%) 13.28 length=32,align1=5,align2=5: 14.06 (-5.88%) 12.50 (5.88%) 13.28 length=64,align1=0,align2=0: 14.06 (-5.88%) 13.28 (0.00%) 13.28 length=64,align1=6,align2=0: 13.28 (0.00%) 13.28 (0.00%) 13.28 length=64,align1=0,align2=6: 14.06 (5.26%) 14.06 (5.26%) 14.84 length=64,align1=6,align2=6: 14.84 (-11.77%) 14.06 (-5.88%) 13.28 length=128,align1=0,align2=0: 17.19 (-4.76%) 14.84 (9.52%) 16.41 length=128,align1=7,align2=0: 16.41 (4.55%) 15.62 (9.09%) 17.19 length=128,align1=0,align2=7: 16.41 (0.00%) 14.06 (14.29%) 16.41 length=128,align1=7,align2=7: 16.41 (4.55%) 15.62 (9.09%) 17.19 length=256,align1=0,align2=0: 21.88 (-3.70%) 21.09 (0.00%) 21.09 length=256,align1=8,align2=0: 21.09 (-3.85%) 21.09 (-3.85%) 20.31 length=256,align1=0,align2=8: 20.31 (-4.00%) 20.31 (-4.00%) 19.53 length=256,align1=8,align2=8: 21.88 (-7.69%) 20.31 (0.00%) 20.31 length=512,align1=0,align2=0: 28.91 (-2.78%) 28.91 (-2.78%) 28.12 length=512,align1=9,align2=0: 30.47 (-2.63%) 30.47 (-2.63%) 29.69 length=512,align1=0,align2=9: 29.69 (0.00%) 29.69 (0.00%) 29.69 length=512,align1=9,align2=9: 28.12 (-2.86%) 28.12 (-2.86%) 27.34 length=1024,align1=0,align2=0: 44.53 (0.00%) 44.53 (0.00%) 44.53 length=1024,align1=10,align2=0: 50.00 (0.00%) 50.00 (0.00%) 50.00 length=1024,align1=0,align2=10: 49.22 (1.56%) 50.78 (-1.56%) 50.00 length=1024,align1=10,align2=10: 44.53 (-1.79%) 43.75 (0.00%) 43.75 length=2048,align1=0,align2=0: 77.34 (-1.02%) 76.56 (0.00%) 76.56 length=2048,align1=11,align2=0: 89.84 (0.00%) 89.84 (0.00%) 89.84 length=2048,align1=0,align2=11: 89.84 (0.00%) 89.84 (0.00%) 89.84 length=2048,align1=11,align2=11: 75.78 (0.00%) 75.78 (0.00%) 75.78 length=4096,align1=0,align2=0: 141.41 (-0.56%) 140.62 (0.00%) 140.62 length=4096,align1=12,align2=0: 171.09 (-0.46%) 170.31 (0.00%) 170.31 length=4096,align1=0,align2=12: 170.31 (0.00%) 170.31 (0.00%) 170.31 length=4096,align1=12,align2=12: 140.62 (0.00%) 140.62 (0.00%) 140.62 length=8192,align1=0,align2=0: 278.91 (-0.28%) 275.78 (0.84%) 278.12 length=8192,align1=13,align2=0: 338.28 (0.23%) 335.94 (0.92%) 339.06 length=8192,align1=0,align2=13: 338.28 (0.00%) 455.47 (-34.64%) 338.28 length=8192,align1=13,align2=13: 278.12 (-0.28%) 275.78 (0.56%) 277.34 length=16384,align1=0,align2=0: 535.94 (-0.15%) 531.25 (0.73%) 535.16 length=16384,align1=14,align2=0: 659.38 (0.12%) 659.38 (0.12%) 660.16 length=16384,align1=0,align2=14: 659.38 (0.00%) 657.03 (0.36%) 659.38 length=16384,align1=14,align2=14: 535.16 (0.44%) 532.81 (0.87%) 537.50 length=32768,align1=0,align2=0: 1260.94 (10.68%) 1121.88 (20.53%) 1411.72 length=32768,align1=15,align2=0: 1368.75 (10.02%) 1376.56 (9.50%) 1521.09 length=32768,align1=0,align2=15: 1333.59 (10.91%) 1373.44 (8.25%) 1496.88 length=32768,align1=15,align2=15: 1256.25 (13.96%) 1125.78 (22.90%) 1460.16 length=65536,align1=0,align2=0: 2853.91 (30.11%) 2589.06 (36.60%) 4083.59 length=65536,align1=16,align2=0: 2850.00 (30.14%) 2589.84 (36.52%) 4079.69 length=65536,align1=0,align2=16: 2853.12 (30.60%) 2589.84 (37.00%) 4110.94 length=65536,align1=16,align2=16: 2850.78 (30.07%) 2589.06 (36.49%) 4076.56 length=0,align1=0,align2=0: 15.62 (-5.26%) 16.41 (-10.53%) 14.84 length=0,align1=0,align2=0: 14.84 (-5.56%) 14.84 (-5.56%) 14.06 length=0,align1=0,align2=0: 14.84 (0.00%) 14.84 (0.00%) 14.84 length=0,align1=0,align2=0: 16.41 (-16.67%) 14.84 (-5.56%) 14.06 length=1,align1=0,align2=0: 15.62 (4.76%) 15.62 (4.76%) 16.41 length=1,align1=1,align2=0: 15.62 (0.00%) 14.84 (5.00%) 15.62 length=1,align1=0,align2=1: 14.84 (0.00%) 14.84 (0.00%) 14.84 length=1,align1=1,align2=1: 14.84 (0.00%) 14.06 (5.26%) 14.84 length=2,align1=0,align2=0: 14.84 (0.00%) 14.06 (5.26%) 14.84 length=2,align1=2,align2=0: 14.84 (0.00%) 14.06 (5.26%) 14.84 length=2,align1=0,align2=2: 14.84 (-5.56%) 14.06 (0.00%) 14.06 length=2,align1=2,align2=2: 14.84 (0.00%) 14.06 (5.26%) 14.84 length=3,align1=0,align2=0: 14.84 (0.00%) 14.84 (0.00%) 14.84 length=3,align1=3,align2=0: 14.84 (-5.56%) 14.06 (0.00%) 14.06 length=3,align1=0,align2=3: 15.62 (-11.11%) 14.06 (0.00%) 14.06 length=3,align1=3,align2=3: 14.84 (0.00%) 14.06 (5.26%) 14.84 length=4,align1=0,align2=0: 17.97 (-27.78%) 14.06 (0.00%) 14.06 length=4,align1=4,align2=0: 13.28 (5.56%) 14.06 (0.00%) 14.06 length=4,align1=0,align2=4: 14.06 (0.00%) 13.28 (5.56%) 14.06 length=4,align1=4,align2=4: 13.28 (5.56%) 13.28 (5.56%) 14.06 length=5,align1=0,align2=0: 13.28 (5.56%) 13.28 (5.56%) 14.06 length=5,align1=5,align2=0: 14.06 (0.00%) 14.06 (0.00%) 14.06 length=5,align1=0,align2=5: 14.06 (0.00%) 13.28 (5.56%) 14.06 length=5,align1=5,align2=5: 14.06 (-5.88%) 14.06 (-5.88%) 13.28 length=6,align1=0,align2=0: 14.06 (-5.88%) 14.06 (-5.88%) 13.28 length=6,align1=6,align2=0: 14.06 (0.00%) 14.06 (0.00%) 14.06 length=6,align1=0,align2=6: 14.06 (0.00%) 13.28 (5.56%) 14.06 length=6,align1=6,align2=6: 14.06 (0.00%) 13.28 (5.56%) 14.06 length=7,align1=0,align2=0: 14.84 (-11.77%) 14.06 (-5.88%) 13.28 length=7,align1=7,align2=0: 13.28 (0.00%) 14.06 (-5.88%) 13.28 length=7,align1=0,align2=7: 14.06 (0.00%) 14.06 (0.00%) 14.06 length=7,align1=7,align2=7: 14.06 (0.00%) 14.06 (0.00%) 14.06 length=8,align1=0,align2=0: 14.06 (-5.88%) 13.28 (0.00%) 13.28 length=8,align1=8,align2=0: 14.06 (0.00%) 13.28 (5.56%) 14.06 length=8,align1=0,align2=8: 13.28 (0.00%) 13.28 (0.00%) 13.28 length=8,align1=8,align2=8: 14.06 (-5.88%) 13.28 (0.00%) 13.28 length=9,align1=0,align2=0: 13.28 (0.00%) 13.28 (0.00%) 13.28 length=9,align1=9,align2=0: 13.28 (0.00%) 13.28 (0.00%) 13.28 length=9,align1=0,align2=9: 13.28 (0.00%) 14.06 (-5.88%) 13.28 length=9,align1=9,align2=9: 14.06 (-5.88%) 13.28 (0.00%) 13.28 length=10,align1=0,align2=0: 14.06 (0.00%) 13.28 (5.56%) 14.06 length=10,align1=10,align2=0: 14.06 (-5.88%) 14.06 (-5.88%) 13.28 length=10,align1=0,align2=10: 14.06 (-5.88%) 13.28 (0.00%) 13.28 length=10,align1=10,align2=10: 14.06 (0.00%) 13.28 (5.56%) 14.06 length=11,align1=0,align2=0: 14.06 (-5.88%) 13.28 (0.00%) 13.28 length=11,align1=11,align2=0: 14.06 (-5.88%) 13.28 (0.00%) 13.28 length=11,align1=0,align2=11: 13.28 (0.00%) 13.28 (0.00%) 13.28 length=11,align1=11,align2=11: 13.28 (0.00%) 13.28 (0.00%) 13.28 length=12,align1=0,align2=0: 14.06 (-5.88%) 13.28 (0.00%) 13.28 length=12,align1=12,align2=0: 14.06 (-5.88%) 13.28 (0.00%) 13.28 length=12,align1=0,align2=12: 14.06 (-5.88%) 13.28 (0.00%) 13.28 length=12,align1=12,align2=12: 14.06 (0.00%) 13.28 (5.56%) 14.06 length=13,align1=0,align2=0: 14.06 (-5.88%) 13.28 (0.00%) 13.28 length=13,align1=13,align2=0: 14.06 (-5.88%) 13.28 (0.00%) 13.28 length=13,align1=0,align2=13: 14.06 (-5.88%) 13.28 (0.00%) 13.28 length=13,align1=13,align2=13: 13.28 (0.00%) 13.28 (0.00%) 13.28 length=14,align1=0,align2=0: 13.28 (0.00%) 13.28 (0.00%) 13.28 length=14,align1=14,align2=0: 13.28 (5.56%) 13.28 (5.56%) 14.06 length=14,align1=0,align2=14: 14.06 (-5.88%) 13.28 (0.00%) 13.28 length=14,align1=14,align2=14: 14.06 (-5.88%) 13.28 (0.00%) 13.28 length=15,align1=0,align2=0: 14.06 (-5.88%) 13.28 (0.00%) 13.28 length=15,align1=15,align2=0: 14.06 (-5.88%) 14.06 (-5.88%) 13.28 length=15,align1=0,align2=15: 13.28 (0.00%) 13.28 (0.00%) 13.28 length=15,align1=15,align2=15: 13.28 (0.00%) 14.06 (-5.88%) 13.28 length=16,align1=0,align2=0: 14.06 (-5.88%) 13.28 (0.00%) 13.28 length=16,align1=16,align2=0: 13.28 (5.56%) 14.06 (0.00%) 14.06 length=16,align1=0,align2=16: 14.84 (-11.77%) 13.28 (0.00%) 13.28 length=16,align1=16,align2=16: 13.28 (-6.25%) 12.50 (0.00%) 12.50 length=17,align1=0,align2=0: 14.06 (-5.88%) 12.50 (5.88%) 13.28 length=17,align1=17,align2=0: 14.84 (-11.77%) 12.50 (5.88%) 13.28 length=17,align1=0,align2=17: 14.84 (-5.56%) 12.50 (11.11%) 14.06 length=17,align1=17,align2=17: 14.84 (-11.77%) 12.50 (5.88%) 13.28 length=18,align1=0,align2=0: 14.06 (0.00%) 12.50 (11.11%) 14.06 length=18,align1=18,align2=0: 13.28 (5.56%) 12.50 (11.11%) 14.06 length=18,align1=0,align2=18: 14.06 (-5.88%) 12.50 (5.88%) 13.28 length=18,align1=18,align2=18: 14.06 (0.00%) 12.50 (11.11%) 14.06 length=19,align1=0,align2=0: 14.06 (-5.88%) 13.28 (0.00%) 13.28 length=19,align1=19,align2=0: 14.06 (-5.88%) 13.28 (0.00%) 13.28 length=19,align1=0,align2=19: 14.84 (-5.56%) 12.50 (11.11%) 14.06 length=19,align1=19,align2=19: 14.06 (-5.88%) 12.50 (5.88%) 13.28 length=20,align1=0,align2=0: 14.84 (-11.77%) 12.50 (5.88%) 13.28 length=20,align1=20,align2=0: 14.06 (0.00%) 12.50 (11.11%) 14.06 length=20,align1=0,align2=20: 14.06 (-5.88%) 12.50 (5.88%) 13.28 length=20,align1=20,align2=20: 14.06 (0.00%) 13.28 (5.56%) 14.06 length=21,align1=0,align2=0: 14.84 (-5.56%) 12.50 (11.11%) 14.06 length=21,align1=21,align2=0: 14.06 (-5.88%) 13.28 (0.00%) 13.28 length=21,align1=0,align2=21: 14.84 (-11.77%) 12.50 (5.88%) 13.28 length=21,align1=21,align2=21: 13.28 (5.56%) 13.28 (5.56%) 14.06 length=22,align1=0,align2=0: 14.06 (-5.88%) 12.50 (5.88%) 13.28 length=22,align1=22,align2=0: 14.06 (-5.88%) 13.28 (0.00%) 13.28 length=22,align1=0,align2=22: 14.06 (0.00%) 12.50 (11.11%) 14.06 length=22,align1=22,align2=22: 14.06 (0.00%) 12.50 (11.11%) 14.06 length=23,align1=0,align2=0: 14.06 (-5.88%) 12.50 (5.88%) 13.28 length=23,align1=23,align2=0: 14.06 (-5.88%) 13.28 (0.00%) 13.28 length=23,align1=0,align2=23: 14.06 (-5.88%) 12.50 (5.88%) 13.28 length=23,align1=23,align2=23: 14.06 (-5.88%) 13.28 (0.00%) 13.28 length=24,align1=0,align2=0: 14.06 (-5.88%) 12.50 (5.88%) 13.28 length=24,align1=24,align2=0: 14.06 (0.00%) 13.28 (5.56%) 14.06 length=24,align1=0,align2=24: 14.84 (-11.77%) 12.50 (5.88%) 13.28 length=24,align1=24,align2=24: 14.06 (-5.88%) 13.28 (0.00%) 13.28 length=25,align1=0,align2=0: 14.06 (0.00%) 12.50 (11.11%) 14.06 length=25,align1=25,align2=0: 14.06 (0.00%) 13.28 (5.56%) 14.06 length=25,align1=0,align2=25: 14.06 (0.00%) 12.50 (11.11%) 14.06 length=25,align1=25,align2=25: 13.28 (0.00%) 13.28 (0.00%) 13.28 length=26,align1=0,align2=0: 14.06 (-5.88%) 12.50 (5.88%) 13.28 length=26,align1=26,align2=0: 14.06 (0.00%) 13.28 (5.56%) 14.06 length=26,align1=0,align2=26: 14.06 (-5.88%) 12.50 (5.88%) 13.28 length=26,align1=26,align2=26: 14.06 (0.00%) 13.28 (5.56%) 14.06 length=27,align1=0,align2=0: 14.06 (-5.88%) 12.50 (5.88%) 13.28 length=27,align1=27,align2=0: 14.06 (-5.88%) 12.50 (5.88%) 13.28 length=27,align1=0,align2=27: 14.06 (-5.88%) 12.50 (5.88%) 13.28 length=27,align1=27,align2=27: 14.06 (0.00%) 12.50 (11.11%) 14.06 length=28,align1=0,align2=0: 14.06 (-5.88%) 12.50 (5.88%) 13.28 length=28,align1=28,align2=0: 14.06 (0.00%) 12.50 (11.11%) 14.06 length=28,align1=0,align2=28: 14.06 (0.00%) 12.50 (11.11%) 14.06 length=28,align1=28,align2=28: 14.84 (-11.77%) 13.28 (0.00%) 13.28 length=29,align1=0,align2=0: 14.06 (-5.88%) 12.50 (5.88%) 13.28 length=29,align1=29,align2=0: 13.28 (0.00%) 12.50 (5.88%) 13.28 length=29,align1=0,align2=29: 14.06 (0.00%) 12.50 (11.11%) 14.06 length=29,align1=29,align2=29: 13.28 (5.56%) 12.50 (11.11%) 14.06 length=30,align1=0,align2=0: 14.06 (-5.88%) 12.50 (5.88%) 13.28 length=30,align1=30,align2=0: 13.28 (5.56%) 12.50 (11.11%) 14.06 length=30,align1=0,align2=30: 14.06 (-5.88%) 12.50 (5.88%) 13.28 length=30,align1=30,align2=30: 13.28 (0.00%) 12.50 (5.88%) 13.28 length=31,align1=0,align2=0: 13.28 (0.00%) 12.50 (5.88%) 13.28 length=31,align1=31,align2=0: 14.06 (0.00%) 12.50 (11.11%) 14.06 length=31,align1=0,align2=31: 13.28 (0.00%) 12.50 (5.88%) 13.28 length=31,align1=31,align2=31: 14.06 (0.00%) 12.50 (11.11%) 14.06 length=48,align1=0,align2=0: 14.06 (0.00%) 14.06 (0.00%) 14.06 length=48,align1=3,align2=0: 14.06 (0.00%) 14.06 (0.00%) 14.06 length=48,align1=0,align2=3: 14.06 (-5.88%) 14.06 (-5.88%) 13.28 length=48,align1=3,align2=3: 13.28 (5.56%) 14.06 (0.00%) 14.06 length=80,align1=0,align2=0: 15.62 (-11.11%) 14.84 (-5.56%) 14.06 length=80,align1=5,align2=0: 15.62 (-11.11%) 16.41 (-16.67%) 14.06 length=80,align1=0,align2=5: 14.06 (0.00%) 15.62 (-11.11%) 14.06 length=80,align1=5,align2=5: 15.62 (-5.26%) 17.19 (-15.79%) 14.84 length=96,align1=0,align2=0: 14.06 (0.00%) 14.84 (-5.56%) 14.06 length=96,align1=6,align2=0: 14.84 (-5.56%) 16.41 (-16.67%) 14.06 length=96,align1=0,align2=6: 14.06 (0.00%) 14.84 (-5.56%) 14.06 length=96,align1=6,align2=6: 14.84 (-5.56%) 17.19 (-22.22%) 14.06 length=112,align1=0,align2=0: 17.19 (-4.76%) 14.06 (14.29%) 16.41 length=112,align1=7,align2=0: 17.19 (0.00%) 16.41 (4.55%) 17.19 length=112,align1=0,align2=7: 16.41 (0.00%) 14.84 (9.52%) 16.41 length=112,align1=7,align2=7: 17.19 (0.00%) 17.19 (0.00%) 17.19 length=144,align1=0,align2=0: 17.19 (-10.00%) 17.97 (-15.00%) 15.62 length=144,align1=9,align2=0: 17.19 (-4.76%) 18.75 (-14.29%) 16.41 length=144,align1=0,align2=9: 20.31 (-8.33%) 18.75 (0.00%) 18.75 length=144,align1=9,align2=9: 18.75 (-4.35%) 18.75 (-4.35%) 17.97 length=160,align1=0,align2=0: 18.75 (-4.35%) 17.97 (0.00%) 17.97 length=160,align1=10,align2=0: 18.75 (4.00%) 18.75 (4.00%) 19.53 length=160,align1=0,align2=10: 19.53 (-4.17%) 17.97 (4.17%) 18.75 length=160,align1=10,align2=10: 18.75 (-4.35%) 18.75 (-4.35%) 17.97 length=176,align1=0,align2=0: 18.75 (-4.35%) 17.19 (4.35%) 17.97 length=176,align1=11,align2=0: 19.53 (0.00%) 19.53 (0.00%) 19.53 length=176,align1=0,align2=11: 19.53 (-4.17%) 18.75 (0.00%) 18.75 length=176,align1=11,align2=11: 18.75 (0.00%) 17.97 (4.17%) 18.75 length=192,align1=0,align2=0: 18.75 (0.00%) 17.97 (4.17%) 18.75 length=192,align1=12,align2=0: 21.09 (-8.00%) 18.75 (4.00%) 19.53 length=192,align1=0,align2=12: 18.75 (0.00%) 18.75 (0.00%) 18.75 length=192,align1=12,align2=12: 18.75 (0.00%) 17.97 (4.17%) 18.75 length=208,align1=0,align2=0: 17.97 (0.00%) 20.31 (-13.04%) 17.97 length=208,align1=13,align2=0: 19.53 (7.41%) 21.09 (0.00%) 21.09 length=208,align1=0,align2=13: 23.44 (-11.11%) 21.09 (0.00%) 21.09 length=208,align1=13,align2=13: 21.09 (-3.85%) 21.09 (-3.85%) 20.31 length=224,align1=0,align2=0: 21.09 (-8.00%) 20.31 (-4.00%) 19.53 length=224,align1=14,align2=0: 23.44 (-11.11%) 20.31 (3.70%) 21.09 length=224,align1=0,align2=14: 21.09 (3.57%) 20.31 (7.14%) 21.88 length=224,align1=14,align2=14: 20.31 (0.00%) 19.53 (3.85%) 20.31 length=240,align1=0,align2=0: 20.31 (-4.00%) 19.53 (0.00%) 19.53 length=240,align1=15,align2=0: 22.66 (0.00%) 20.31 (10.34%) 22.66 length=240,align1=0,align2=15: 20.31 (-4.00%) 20.31 (-4.00%) 19.53 length=240,align1=15,align2=15: 21.88 (0.00%) 21.09 (3.57%) 21.88 length=272,align1=0,align2=0: 20.31 (0.00%) 28.12 (-38.46%) 20.31 length=272,align1=17,align2=0: 22.66 (0.00%) 27.34 (-20.69%) 22.66 length=272,align1=0,align2=17: 25.78 (-10.00%) 28.12 (-20.00%) 23.44 length=272,align1=17,align2=17: 22.66 (-3.57%) 27.34 (-25.00%) 21.88 length=288,align1=0,align2=0: 23.44 (-7.14%) 27.34 (-25.00%) 21.88 length=288,align1=18,align2=0: 22.66 (0.00%) 27.34 (-20.69%) 22.66 length=288,align1=0,align2=18: 23.44 (-3.45%) 25.00 (-10.35%) 22.66 length=288,align1=18,align2=18: 22.66 (-3.57%) 21.88 (0.00%) 21.88 length=304,align1=0,align2=0: 21.88 (0.00%) 21.88 (0.00%) 21.88 length=304,align1=19,align2=0: 23.44 (-3.45%) 22.66 (0.00%) 22.66 length=304,align1=0,align2=19: 22.66 (0.00%) 22.66 (0.00%) 22.66 length=304,align1=19,align2=19: 22.66 (-3.57%) 21.88 (0.00%) 21.88 length=320,align1=0,align2=0: 22.66 (-3.57%) 21.88 (0.00%) 21.88 length=320,align1=20,align2=0: 22.66 (0.00%) 22.66 (0.00%) 22.66 length=320,align1=0,align2=20: 22.66 (0.00%) 22.66 (0.00%) 22.66 length=320,align1=20,align2=20: 22.66 (-3.57%) 21.88 (0.00%) 21.88 length=336,align1=0,align2=0: 21.88 (0.00%) 24.22 (-10.71%) 21.88 length=336,align1=21,align2=0: 22.66 (0.00%) 25.00 (-10.35%) 22.66 length=336,align1=0,align2=21: 25.78 (0.00%) 25.00 (3.03%) 25.78 length=336,align1=21,align2=21: 25.00 (0.00%) 23.44 (6.25%) 25.00 length=352,align1=0,align2=0: 24.22 (0.00%) 24.22 (0.00%) 24.22 length=352,align1=22,align2=0: 25.00 (0.00%) 25.00 (0.00%) 25.00 length=352,align1=0,align2=22: 25.00 (-3.23%) 25.00 (-3.23%) 24.22 length=352,align1=22,align2=22: 25.00 (-3.23%) 24.22 (0.00%) 24.22 length=368,align1=0,align2=0: 25.00 (-3.23%) 23.44 (3.23%) 24.22 length=368,align1=23,align2=0: 25.00 (0.00%) 24.22 (3.12%) 25.00 length=368,align1=0,align2=23: 25.00 (-3.23%) 25.00 (-3.23%) 24.22 length=368,align1=23,align2=23: 25.00 (-6.67%) 23.44 (0.00%) 23.44 length=384,align1=0,align2=0: 24.22 (0.00%) 24.22 (0.00%) 24.22 length=384,align1=24,align2=0: 25.00 (0.00%) 24.22 (3.12%) 25.00 length=384,align1=0,align2=24: 25.00 (0.00%) 25.78 (-3.12%) 25.00 length=384,align1=24,align2=24: 24.22 (-3.33%) 23.44 (0.00%) 23.44 length=400,align1=0,align2=0: 25.00 (-3.23%) 26.56 (-9.68%) 24.22 length=400,align1=25,align2=0: 25.78 (-3.12%) 27.34 (-9.38%) 25.00 length=400,align1=0,align2=25: 27.34 (0.00%) 27.34 (0.00%) 27.34 length=400,align1=25,align2=25: 26.56 (0.00%) 25.78 (2.94%) 26.56 length=416,align1=0,align2=0: 26.56 (-3.03%) 25.78 (0.00%) 25.78 length=416,align1=26,align2=0: 28.12 (-2.86%) 27.34 (0.00%) 27.34 length=416,align1=0,align2=26: 27.34 (-2.94%) 28.12 (-5.88%) 26.56 length=416,align1=26,align2=26: 25.78 (0.00%) 26.56 (-3.03%) 25.78 length=432,align1=0,align2=0: 27.34 (-2.94%) 25.78 (2.94%) 26.56 length=432,align1=27,align2=0: 28.12 (-2.86%) 27.34 (0.00%) 27.34 length=432,align1=0,align2=27: 27.34 (0.00%) 28.12 (-2.86%) 27.34 length=432,align1=27,align2=27: 25.78 (0.00%) 25.78 (0.00%) 25.78 length=448,align1=0,align2=0: 26.56 (-3.03%) 25.78 (0.00%) 25.78 length=448,align1=28,align2=0: 27.34 (0.00%) 27.34 (0.00%) 27.34 length=448,align1=0,align2=28: 27.34 (0.00%) 28.12 (-2.86%) 27.34 length=448,align1=28,align2=28: 25.78 (0.00%) 25.78 (0.00%) 25.78 length=464,align1=0,align2=0: 25.78 (0.00%) 28.12 (-9.09%) 25.78 length=464,align1=29,align2=0: 28.12 (-2.86%) 29.69 (-8.57%) 27.34 length=464,align1=0,align2=29: 30.47 (0.00%) 30.47 (0.00%) 30.47 length=464,align1=29,align2=29: 28.12 (0.00%) 27.34 (2.78%) 28.12 length=480,align1=0,align2=0: 29.69 (-5.56%) 28.12 (0.00%) 28.12 length=480,align1=30,align2=0: 31.25 (-2.56%) 29.69 (2.56%) 30.47 length=480,align1=0,align2=30: 29.69 (0.00%) 30.47 (-2.63%) 29.69 length=480,align1=30,align2=30: 28.12 (0.00%) 28.12 (0.00%) 28.12 length=496,align1=0,align2=0: 28.12 (0.00%) 27.34 (2.78%) 28.12 length=496,align1=31,align2=0: 30.47 (-2.63%) 29.69 (0.00%) 29.69 length=496,align1=0,align2=31: 29.69 (0.00%) 30.47 (-2.63%) 29.69 length=496,align1=31,align2=31: 28.12 (-2.86%) 28.12 (-2.86%) 27.34 length=1024,align1=0,align2=0: 44.53 (0.00%) 44.53 (0.00%) 44.53 length=1024,align1=32,align2=0: 44.53 (-1.79%) 44.53 (-1.79%) 43.75 length=1024,align1=0,align2=32: 44.53 (-1.79%) 43.75 (0.00%) 43.75 length=1024,align1=32,align2=32: 43.75 (1.75%) 43.75 (1.75%) 44.53 length=1056,align1=0,align2=0: 46.88 (-1.69%) 46.88 (-1.69%) 46.09 length=1056,align1=33,align2=0: 53.12 (0.00%) 52.34 (1.47%) 53.12 length=1056,align1=0,align2=33: 52.34 (0.00%) 53.12 (-1.49%) 52.34 length=1056,align1=33,align2=33: 46.09 (0.00%) 46.88 (-1.69%) 46.09 length=1088,align1=0,align2=0: 46.88 (-1.69%) 46.09 (0.00%) 46.09 length=1088,align1=34,align2=0: 52.34 (0.00%) 52.34 (0.00%) 52.34 length=1088,align1=0,align2=34: 53.12 (-3.03%) 53.12 (-3.03%) 51.56 length=1088,align1=34,align2=34: 46.09 (0.00%) 46.88 (-1.69%) 46.09 length=1120,align1=0,align2=0: 49.22 (-1.61%) 48.44 (0.00%) 48.44 length=1120,align1=35,align2=0: 54.69 (1.41%) 55.47 (0.00%) 55.47 length=1120,align1=0,align2=35: 57.03 (0.00%) 55.47 (2.74%) 57.03 length=1120,align1=35,align2=35: 48.44 (0.00%) 49.22 (-1.61%) 48.44 length=1152,align1=0,align2=0: 47.66 (1.61%) 48.44 (0.00%) 48.44 length=1152,align1=36,align2=0: 55.47 (-1.43%) 55.47 (-1.43%) 54.69 length=1152,align1=0,align2=36: 58.59 (-1.35%) 55.47 (4.05%) 57.81 length=1152,align1=36,align2=36: 48.44 (0.00%) 49.22 (-1.61%) 48.44 length=1184,align1=0,align2=0: 53.12 (-3.03%) 50.78 (1.52%) 51.56 length=1184,align1=37,align2=0: 61.72 (-2.60%) 57.03 (5.19%) 60.16 length=1184,align1=0,align2=37: 62.50 (-1.27%) 57.03 (7.60%) 61.72 length=1184,align1=37,align2=37: 53.12 (-1.49%) 50.78 (2.99%) 52.34 length=1216,align1=0,align2=0: 53.91 (-4.55%) 50.78 (1.52%) 51.56 length=1216,align1=38,align2=0: 60.94 (0.00%) 57.03 (6.41%) 60.94 length=1216,align1=0,align2=38: 60.16 (0.00%) 57.81 (3.90%) 60.16 length=1216,align1=38,align2=38: 52.34 (-1.52%) 50.00 (3.03%) 51.56 length=1248,align1=0,align2=0: 54.69 (-2.94%) 53.12 (0.00%) 53.12 length=1248,align1=39,align2=0: 64.06 (-1.23%) 60.16 (4.94%) 63.28 length=1248,align1=0,align2=39: 60.94 (-2.63%) 60.16 (-1.32%) 59.38 length=1248,align1=39,align2=39: 53.12 (0.00%) 52.34 (1.47%) 53.12 length=1280,align1=0,align2=0: 52.34 (-1.52%) 52.34 (-1.52%) 51.56 length=1280,align1=40,align2=0: 61.72 (3.66%) 59.38 (7.32%) 64.06 length=1280,align1=0,align2=40: 60.94 (-2.63%) 60.16 (-1.32%) 59.38 length=1280,align1=40,align2=40: 52.34 (-1.52%) 52.34 (-1.52%) 51.56 length=1312,align1=0,align2=0: 54.69 (-1.45%) 55.47 (-2.90%) 53.91 length=1312,align1=41,align2=0: 63.28 (0.00%) 62.50 (1.23%) 63.28 length=1312,align1=0,align2=41: 62.50 (0.00%) 62.50 (0.00%) 62.50 length=1312,align1=41,align2=41: 53.91 (0.00%) 54.69 (-1.45%) 53.91 length=1344,align1=0,align2=0: 54.69 (0.00%) 54.69 (0.00%) 54.69 length=1344,align1=42,align2=0: 62.50 (0.00%) 62.50 (0.00%) 62.50 length=1344,align1=0,align2=42: 62.50 (-1.27%) 62.50 (-1.27%) 61.72 length=1344,align1=42,align2=42: 53.91 (0.00%) 53.91 (0.00%) 53.91 length=1376,align1=0,align2=0: 65.62 (-16.67%) 68.75 (-22.22%) 56.25 length=1376,align1=43,align2=0: 71.88 (-9.52%) 73.44 (-11.90%) 65.62 length=1376,align1=0,align2=43: 72.66 (-12.05%) 74.22 (-14.46%) 64.84 length=1376,align1=43,align2=43: 64.06 (-13.89%) 67.97 (-20.83%) 56.25 length=1408,align1=0,align2=0: 57.03 (-1.39%) 68.75 (-22.22%) 56.25 length=1408,align1=44,align2=0: 65.62 (-1.20%) 73.44 (-13.25%) 64.84 length=1408,align1=0,align2=44: 64.84 (0.00%) 74.22 (-14.46%) 64.84 length=1408,align1=44,align2=44: 56.25 (-1.41%) 68.75 (-23.94%) 55.47 length=1440,align1=0,align2=0: 67.97 (-14.47%) 64.84 (-9.21%) 59.38 length=1440,align1=45,align2=0: 74.22 (-10.47%) 68.75 (-2.33%) 67.19 length=1440,align1=0,align2=45: 72.66 (-6.90%) 69.53 (-2.30%) 67.97 length=1440,align1=45,align2=45: 65.62 (-13.51%) 58.59 (-1.35%) 57.81 length=1472,align1=0,align2=0: 66.41 (-14.86%) 58.59 (-1.35%) 57.81 length=1472,align1=46,align2=0: 73.44 (-9.30%) 67.19 (0.00%) 67.19 length=1472,align1=0,align2=46: 70.31 (-4.65%) 67.97 (-1.16%) 67.19 length=1472,align1=46,align2=46: 57.81 (0.00%) 58.59 (-1.35%) 57.81 length=1504,align1=0,align2=0: 60.94 (0.00%) 60.94 (0.00%) 60.94 length=1504,align1=47,align2=0: 71.09 (-1.11%) 70.31 (0.00%) 70.31 length=1504,align1=0,align2=47: 70.31 (-1.12%) 70.31 (-1.12%) 69.53 length=1504,align1=47,align2=47: 60.94 (-1.30%) 60.16 (0.00%) 60.16 length=1536,align1=0,align2=0: 62.50 (-3.90%) 60.16 (0.00%) 60.16 length=1536,align1=48,align2=0: 60.94 (-1.30%) 60.16 (0.00%) 60.16 length=1536,align1=0,align2=48: 61.72 (-3.95%) 60.16 (-1.32%) 59.38 length=1536,align1=48,align2=48: 60.94 (-1.30%) 60.16 (0.00%) 60.16 length=1568,align1=0,align2=0: 80.47 (-27.16%) 63.28 (0.00%) 63.28 length=1568,align1=49,align2=0: 86.72 (-18.09%) 72.66 (1.06%) 73.44 length=1568,align1=0,align2=49: 74.22 (-3.26%) 74.22 (-3.26%) 71.88 length=1568,align1=49,align2=49: 62.50 (0.00%) 61.72 (1.25%) 62.50 length=1600,align1=0,align2=0: 62.50 (-1.27%) 62.50 (-1.27%) 61.72 length=1600,align1=50,align2=0: 73.44 (0.00%) 71.88 (2.13%) 73.44 length=1600,align1=0,align2=50: 72.66 (0.00%) 73.44 (-1.08%) 72.66 length=1600,align1=50,align2=50: 62.50 (-1.27%) 62.50 (-1.27%) 61.72 length=1632,align1=0,align2=0: 64.84 (0.00%) 64.84 (0.00%) 64.84 length=1632,align1=51,align2=0: 75.78 (0.00%) 75.00 (1.03%) 75.78 length=1632,align1=0,align2=51: 78.91 (0.00%) 75.78 (3.96%) 78.91 length=1632,align1=51,align2=51: 64.84 (-2.47%) 64.84 (-2.47%) 63.28 length=1664,align1=0,align2=0: 64.84 (-1.22%) 64.84 (-1.22%) 64.06 length=1664,align1=52,align2=0: 75.78 (0.00%) 75.00 (1.03%) 75.78 length=1664,align1=0,align2=52: 80.47 (-0.98%) 75.78 (4.90%) 79.69 length=1664,align1=52,align2=52: 64.06 (-1.23%) 65.62 (-3.70%) 63.28 length=1696,align1=0,align2=0: 69.53 (-3.49%) 72.66 (-8.14%) 67.19 length=1696,align1=53,align2=0: 80.47 (-0.98%) 82.03 (-2.94%) 79.69 length=1696,align1=0,align2=53: 80.47 (0.96%) 82.03 (-0.96%) 81.25 length=1696,align1=53,align2=53: 68.75 (-2.33%) 72.66 (-8.14%) 67.19 length=1728,align1=0,align2=0: 67.97 (0.00%) 72.66 (-6.90%) 67.97 length=1728,align1=54,align2=0: 80.47 (-0.98%) 82.81 (-3.92%) 79.69 length=1728,align1=0,align2=54: 78.91 (-1.00%) 82.03 (-5.00%) 78.12 length=1728,align1=54,align2=54: 68.75 (0.00%) 72.66 (-5.68%) 68.75 length=1760,align1=0,align2=0: 77.34 (-12.50%) 68.75 (0.00%) 68.75 length=1760,align1=55,align2=0: 91.41 (-8.33%) 79.69 (5.56%) 84.38 length=1760,align1=0,align2=55: 88.28 (-10.78%) 80.47 (-0.98%) 79.69 length=1760,align1=55,align2=55: 77.34 (-11.24%) 68.75 (1.12%) 69.53 length=1792,align1=0,align2=0: 78.12 (-14.94%) 68.75 (-1.15%) 67.97 length=1792,align1=56,align2=0: 88.28 (-4.63%) 79.69 (5.56%) 84.38 length=1792,align1=0,align2=56: 88.28 (-9.71%) 80.47 (0.00%) 80.47 length=1792,align1=56,align2=56: 77.34 (-11.24%) 68.75 (1.12%) 69.53 length=1824,align1=0,align2=0: 72.66 (7.92%) 70.31 (10.89%) 78.91 length=1824,align1=57,align2=0: 85.94 (5.17%) 82.03 (9.48%) 90.62 length=1824,align1=0,align2=57: 82.03 (3.67%) 82.81 (2.75%) 85.16 length=1824,align1=57,align2=57: 70.31 (-1.12%) 70.31 (-1.12%) 69.53 length=1856,align1=0,align2=0: 70.31 (-1.12%) 70.31 (-1.12%) 69.53 length=1856,align1=58,align2=0: 83.59 (-0.94%) 82.03 (0.94%) 82.81 length=1856,align1=0,align2=58: 178.12 (-115.09%) 82.81 (0.00%) 82.81 length=1856,align1=58,align2=58: 70.31 (-1.12%) 70.31 (-1.12%) 69.53 length=1888,align1=0,align2=0: 73.44 (-1.08%) 78.91 (-8.60%) 72.66 length=1888,align1=59,align2=0: 85.94 (0.00%) 89.84 (-4.55%) 85.94 length=1888,align1=0,align2=59: 84.38 (0.00%) 89.06 (-5.56%) 84.38 length=1888,align1=59,align2=59: 72.66 (-1.09%) 78.12 (-8.70%) 71.88 length=1920,align1=0,align2=0: 72.66 (-1.09%) 78.12 (-8.70%) 71.88 length=1920,align1=60,align2=0: 85.94 (0.00%) 89.84 (-4.55%) 85.94 length=1920,align1=0,align2=60: 85.16 (0.00%) 89.06 (-4.59%) 85.16 length=1920,align1=60,align2=60: 72.66 (-1.09%) 78.91 (-9.78%) 71.88 length=1952,align1=0,align2=0: 75.00 (-1.05%) 75.00 (-1.05%) 74.22 length=1952,align1=61,align2=0: 88.28 (0.00%) 87.50 (0.88%) 88.28 length=1952,align1=0,align2=61: 87.50 (0.00%) 88.28 (-0.89%) 87.50 length=1952,align1=61,align2=61: 74.22 (0.00%) 74.22 (0.00%) 74.22 length=1984,align1=0,align2=0: 75.00 (-1.05%) 73.44 (1.05%) 74.22 length=1984,align1=62,align2=0: 89.06 (-0.89%) 87.50 (0.88%) 88.28 length=1984,align1=0,align2=62: 87.50 (0.00%) 88.28 (-0.89%) 87.50 length=1984,align1=62,align2=62: 74.22 (0.00%) 74.22 (0.00%) 74.22 length=2016,align1=0,align2=0: 77.34 (-1.02%) 76.56 (0.00%) 76.56 length=2016,align1=63,align2=0: 91.41 (-0.86%) 90.62 (0.00%) 90.62 length=2016,align1=0,align2=63: 89.84 (0.00%) 90.62 (-0.87%) 89.84 length=2016,align1=63,align2=63: 77.34 (-1.02%) 76.56 (0.00%) 76.56 length=4096,align1=0,align2=0: 141.41 (-0.56%) 146.88 (-4.44%) 140.62 Function: memcpy __memcpy_thunderx __memcpy_falkor __memcpy_generic Variant: large ================================================================================ length=65543,align1=0,align2=0: 4018.75 (3.09%) 2634.38 (36.47%) 4146.88 length=65551,align1=0,align2=3: 4425.00 (-6.47%) 3134.38 (24.59%) 4156.25 length=65567,align1=3,align2=0: 2909.38 (29.95%) 3134.38 (24.53%) 4153.12 length=65599,align1=3,align2=5: 4415.62 (-6.16%) 3134.38 (24.64%) 4159.38 length=131079,align1=0,align2=0: 5765.62 (30.38%) 5240.62 (36.72%) 8281.25 length=131087,align1=0,align2=3: 8831.25 (-6.56%) 6271.88 (24.32%) 8287.50 length=131103,align1=3,align2=0: 5793.75 (29.05%) 6268.75 (23.23%) 8165.62 length=131135,align1=3,align2=5: 5806.25 (29.97%) 6259.38 (24.50%) 8290.62 length=262151,align1=0,align2=0: 11850.00 (28.91%) 10762.50 (35.43%) 16668.80 length=262159,align1=0,align2=3: 12043.80 (27.72%) 12700.00 (23.78%) 16662.50 length=262175,align1=3,align2=0: 12046.90 (27.90%) 12687.50 (24.07%) 16709.40 length=262207,align1=3,align2=5: 11984.40 (28.08%) 12678.10 (23.91%) 16662.50 length=524295,align1=0,align2=0: 24825.00 (25.00%) 24268.80 (27.34%) 33400.00 length=524303,align1=0,align2=3: 35731.20 (-6.53%) 25678.10 (23.44%) 33540.60 length=524319,align1=3,align2=0: 25893.80 (22.71%) 25725.00 (23.22%) 33503.10 length=524351,align1=3,align2=5: 25887.50 (22.86%) 25690.60 (23.45%) 33559.40 length=1048583,align1=0,align2=0: 50621.90 (0.30%) 50600.00 (0.34%) 50771.90 length=1048591,align1=0,align2=3: 53206.20 (0.54%) 51081.20 (4.51%) 53493.80 length=1048607,align1=3,align2=0: 53221.90 (0.32%) 51975.00 (2.66%) 53393.80 length=1048639,align1=3,align2=5: 53240.60 (0.36%) 51953.10 (2.77%) 53431.20 length=2097159,align1=0,align2=0: 103744.00 (-2.00%) 102447.00 (-1.00%) 102425.00 length=2097167,align1=0,align2=3: 108588.00 (-1.00%) 105159.00 (2.00%) 107606.00 length=2097183,align1=3,align2=0: 107678.00 (0.00%) 105250.00 (2.00%) 108125.00 length=2097215,align1=3,align2=5: 107906.00 (1.00%) 105841.00 (3.00%) 109475.00 length=4194311,align1=0,align2=0: 202994.00 (0.00%) 202500.00 (1.00%) 204809.00 length=4194319,align1=0,align2=3: 213350.00 (0.00%) 205997.00 (3.00%) 213384.00 length=4194335,align1=3,align2=0: 212653.00 (0.00%) 206444.00 (3.00%) 212900.00 length=4194367,align1=3,align2=5: 213044.00 (0.00%) 206084.00 (3.00%) 213847.00 length=8388615,align1=0,align2=0: 401294.00 (0.00%) 401231.00 (0.00%) 401944.00 length=8388623,align1=0,align2=3: 480872.00 (-14.00%) 406444.00 (3.00%) 422900.00 length=8388639,align1=3,align2=0: 422147.00 (0.00%) 407750.00 (3.00%) 422803.00 length=8388671,align1=3,align2=5: 442003.00 (-5.00%) 407125.00 (3.00%) 423509.00 length=16777223,align1=0,align2=0: 799809.00 (0.00%) 800000.00 (0.00%) 801756.00 length=16777231,align1=0,align2=3: 841184.00 (0.00%) 808525.00 (4.00%) 843775.00 length=16777247,align1=3,align2=0: 841166.00 (0.00%) 810147.00 (3.00%) 843147.00 length=16777279,align1=3,align2=5: 972569.00 (-16.00%) 808588.00 (4.00%) 843731.00 length=33554439,align1=0,align2=0: 1842240.00 (-0.01%) 1863590.00 (-1.17%) 1841990.00 length=33554447,align1=0,align2=3: 2103470.00 (-2.74%) 1919460.00 (6.25%) 2047440.00 length=33554463,align1=3,align2=0: 2075690.00 (-1.07%) 1930040.00 (6.02%) 2053720.00 length=33554495,align1=3,align2=5: 2110590.00 (-2.82%) 1924440.00 (6.25%) 2052650.00 Function: memcpy __memcpy_thunderx __memcpy_falkor __memcpy_generic Variant: random ================================================================================ max-size=4096: 44061.90 (5.85%) 38568.20 (17.59%) 46799.90 max-size=8192: 42790.90 (5.27%) 38158.90 (15.52%) 45171.50 max-size=16384: 44912.10 (2.25%) 38710.40 (15.75%) 45945.00 max-size=32768: 43577.90 (1.23%) 37975.10 (13.93%) 44120.00 max-size=65536: 44375.50 (1.04%) 38474.20 (14.20%) 44840.60 * manual/tunables.texi (Tunable glibc.tune.cpu): Add falkor. * sysdeps/aarch64/multiarch/Makefile (sysdep_routines): Add memcpy_falkor. * sysdeps/aarch64/multiarch/ifunc-impl-list.c (MAX_IFUNC): Bump. (__libc_ifunc_impl_list): Add __memcpy_falkor. * sysdeps/aarch64/multiarch/memcpy.c: Likewise. * sysdeps/aarch64/multiarch/memcpy_falkor.S: New file. * sysdeps/unix/sysv/linux/aarch64/cpu-features.c (cpu_list): Add falkor. * sysdeps/unix/sysv/linux/aarch64/cpu-features.h (IS_FALKOR): New macro. | |||||
* | tunables, aarch64: New tunable to override cpu | Siddhesh Poyarekar | 2017-06-30 | 1 | -0/+25 | |
| | | | | | | | | | | | | | | | | | | | | Add a new tunable (glibc.tune.cpu) to override CPU identification on aarch64. This is useful in two cases: one where it is desirable to pretend to be another CPU for purposes of testing or because routines written for that CPU are beneficial for specific workloads and second where the underlying kernel does not support emulation of MRS to get the MIDR of the CPU. * elf/dl-tunables.h (tunable_is_name): Move from... * elf/dl-tunables.c (is_name): ... here. (parse_tunables, __tunables_init): Adjust. * manual/tunables.texi: Document glibc.tune.cpu. * sysdeps/aarch64/dl-tunables.list: New file. * sysdeps/unix/sysv/linux/aarch64/cpu-features.c (struct cpu_list): New type. (cpu_list): New list of CPU names and their MIDR. (get_midr_from_mcpu): New function. (init_cpu_features): Override MIDR if necessary. | |||||
* | aarch64: Call all string function implementations in tests | Siddhesh Poyarekar | 2017-06-30 | 1 | -4/+2 | |
| | | | | | | | | | | | The string function implementations implemented so far do not use any instructions that may deviate from standard aarch64, so it is possible for all routines to run on all armv8 hardware. Select all implementations in the benchmarks and tests. * sysdeps/aarch64/multiarch/ifunc-impl-list.c (__libc_ifunc_impl_list): Unconditionally select thunderx routine for testing. | |||||
* | [AArch64] Add more cfi annotations to tlsdesc entry points | Szabolcs Nagy | 2017-06-21 | 1 | -0/+70 | |
| | | | | | | | | Backtrace through _dl_tlsdesc_resolve_rela was broken because the offset of x30 from cfa was not in the debug info. Add enough annotation so backtracing from the dynamic linker through tlsdesc entry points works and the debugger shows registers correctly. | |||||
* | [AArch64] Use hidden __GI__dl_argv in rtld startup code | Szabolcs Nagy | 2017-06-21 | 1 | -2/+2 | |
| | | | | | We rely on the symbol being locally defined so using extern symbol is not correct and the linker may complain about the relocations. | |||||
* | Remove bits/string.h. | Zack Weinberg | 2017-06-20 | 1 | -24/+0 | |
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | These machine-dependent inline string functions have never been on by default, and even if they were a good idea at the time they were introduced, they haven't really been touched in ten to fifteen years and probably aren't a good idea on current-gen processors. Current thinking is that this class of optimization is best left to the compiler. * bits/string.h, string/bits/string.h * sysdeps/aarch64/bits/string.h * sysdeps/m68k/m680x0/m68020/bits/string.h * sysdeps/s390/bits/string.h, sysdeps/sparc/bits/string.h * sysdeps/x86/bits/string.h: Delete file. * string/string.h: Don't include bits/string.h. * string/bits/string3.h: Rename to bits/string_fortified.h. No need to undef various symbols that the removed headers might have defined as macros. * string/Makefile (headers): Remove bits/string.h, change bits/string3.h to bits/string_fortified.h. * string/string-inlines.c: Update commentary. Remove definitions of various macros that nothing looks at anymore. Don't directly include bits/string.h. Set _STRING_INLINE_unaligned here, based on compiler-predefined macros. * string/strncat.c: If STRNCAT is not defined, or STRNCAT_PRIMARY _is_ defined, provide internal hidden alias __strncat. * include/string.h: Declare internal hidden alias __strncat. Only forward __stpcpy to __builtin_stpcpy if __NO_STRING_INLINES is not defined. * include/bits/string3.h: Rename to bits/string_fortified.h, update to match above. * sysdeps/i386/string-inlines.c: Define compat symbols for everything formerly defined by sysdeps/x86/bits/string.h. Make existing definitions into compat symbols as well. Remove some no-longer-necessary messing around with macros. * sysdeps/powerpc/powerpc32/power4/multiarch/mempcpy.c * sysdeps/powerpc/powerpc64/multiarch/mempcpy.c * sysdeps/powerpc/powerpc64/multiarch/stpcpy.c * sysdeps/s390/multiarch/mempcpy.c No need to define _HAVE_STRING_ARCH_mempcpy. Do define __NO_STRING_INLINES and NO_MEMPCPY_STPCPY_REDIRECT. * sysdeps/i386/i686/multiarch/strncat-c.c * sysdeps/s390/multiarch/strncat-c.c * sysdeps/x86_64/multiarch/strncat-c.c Define STRNCAT_PRIMARY. Don't change definition of libc_hidden_def. | |||||
* | PowerPC64 ELFv2 PPC64_OPT_LOCALENTRY | Alan Modra | 2017-06-14 | 1 | -0/+1 | |
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | ELFv2 functions with localentry:0 are those with a single entry point, ie. global entry == local entry, that have no requirement on r2 or r12 and guarantee r2 is unchanged on return. Such an external function can be called via the PLT without saving r2 or restoring it on return, avoiding a common load-hit-store for small functions. This patch implements the ld.so changes necessary for this optimization. ld.so needs to check that an optimized plt call sequence is in fact calling a function implemented with localentry:0, end emit a fatal error otherwise. The elf/testobj6.c change is to stop "error while loading shared libraries: expected localentry:0 `preload'" when running elf/preloadtest, which we'd get otherwise. * elf/elf.h (PPC64_OPT_LOCALENTRY): Define. * sysdeps/alpha/dl-machine.h (elf_machine_fixup_plt): Add refsym and sym parameters. Adjust callers. * sysdeps/aarch64/dl-machine.h (elf_machine_fixup_plt): Likewise. * sysdeps/arm/dl-machine.h (elf_machine_fixup_plt): Likewise. * sysdeps/generic/dl-machine.h (elf_machine_fixup_plt): Likewise. * sysdeps/hppa/dl-machine.h (elf_machine_fixup_plt): Likewise. * sysdeps/i386/dl-machine.h (elf_machine_fixup_plt): Likewise. * sysdeps/ia64/dl-machine.h (elf_machine_fixup_plt): Likewise. * sysdeps/m68k/dl-machine.h (elf_machine_fixup_plt): Likewise. * sysdeps/microblaze/dl-machine.h (elf_machine_fixup_plt): Likewise. * sysdeps/mips/dl-machine.h (elf_machine_fixup_plt): Likewise. * sysdeps/nios2/dl-machine.h (elf_machine_fixup_plt): Likewise. * sysdeps/powerpc/powerpc32/dl-machine.h (elf_machine_fixup_plt): Likewise. * sysdeps/s390/s390-32/dl-machine.h (elf_machine_fixup_plt): Likewise. * sysdeps/s390/s390-64/dl-machine.h (elf_machine_fixup_plt): Likewise. * sysdeps/sh/dl-machine.h (elf_machine_fixup_plt): Likewise. * sysdeps/sparc/sparc32/dl-machine.h (elf_machine_fixup_plt): Likewise. * sysdeps/sparc/sparc64/dl-machine.h (elf_machine_fixup_plt): Likewise. * sysdeps/tile/dl-machine.h (elf_machine_fixup_plt): Likewise. * sysdeps/x86_64/dl-machine.h (elf_machine_fixup_plt): Likewise. * sysdeps/powerpc/powerpc64/dl-machine.c (_dl_error_localentry): New. (_dl_reloc_overflow): Increase buffser size. Formatting. * sysdeps/powerpc/powerpc64/dl-machine.h (ppc64_local_entry_offset): Delete reloc param, add refsym and sym. Check optimized plt call stubs for localentry:0 functions. Adjust callers. (elf_machine_fixup_plt, elf_machine_plt_conflict): Add refsym and sym parameters. Adjust callers. (_dl_reloc_overflow): Move attribute. (_dl_error_localentry): Declare. * elf/dl-runtime.c (_dl_fixup): Save original sym. Pass refsym and sym to elf_machine_fixup_plt. * elf/testobj6.c (preload): Call printf. | |||||
* | Optimize generic spinlock code and use C11 like atomic macros. | Stefan Liebler | 2017-06-06 | 2 | -24/+1 | |
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This patch optimizes the generic spinlock code. The type pthread_spinlock_t is a typedef to volatile int on all archs. Passing a volatile pointer to the atomic macros which are not mapped to the C11 atomic builtins can lead to extra stores and loads to stack if such a macro creates a temporary variable by using "__typeof (*(mem)) tmp;". Thus, those macros which are used by spinlock code - atomic_exchange_acquire, atomic_load_relaxed, atomic_compare_exchange_weak - have to be adjusted. According to the comment from Szabolcs Nagy, the type of a cast expression is unqualified (see http://www.open-std.org/jtc1/sc22/wg14/www/docs/dr_423.htm): __typeof ((__typeof (*(mem)) *(mem)) tmp; Thus from spinlock perspective the variable tmp is of type int instead of type volatile int. This patch adjusts those macros in include/atomic.h. With this construct GCC >= 5 omits the extra stores and loads. The atomic macros are replaced by the C11 like atomic macros and thus the code is aligned to it. The pthread_spin_unlock implementation is now using release memory order instead of sequentially consistent memory order. The issue with passed volatile int pointers applies to the C11 like atomic macros as well as the ones used before. I've added a glibc_likely hint to the first atomic exchange in pthread_spin_lock in order to return immediately to the caller if the lock is free. Without the hint, there is an additional jump if the lock is free. I've added the atomic_spin_nop macro within the loop of plain reads. The plain reads are also realized by C11 like atomic_load_relaxed macro. The new define ATOMIC_EXCHANGE_USES_CAS determines if the first try to acquire the spinlock in pthread_spin_lock or pthread_spin_trylock is an exchange or a CAS. This is defined in atomic-machine.h for all architectures. The define SPIN_LOCK_READS_BETWEEN_CMPXCHG is now removed. There is no technical reason for throwing in a CAS every now and then, and so far we have no evidence that it can improve performance. If that would be the case, we have to adjust other spin-waiting loops elsewhere, too! Using a CAS loop without plain reads is not a good idea on many targets and wasn't used by one. Thus there is now no option to do so. Architectures are now using the generic spinlock automatically if they do not provide an own implementation. Thus the pthread_spin_lock.c files in sysdeps folder are deleted. ChangeLog: * NEWS: Mention new spinlock implementation. * include/atomic.h: (__atomic_val_bysize): Cast type to omit volatile qualifier. (atomic_exchange_acq): Likewise. (atomic_load_relaxed): Likewise. (ATOMIC_EXCHANGE_USES_CAS): Check definition. * nptl/pthread_spin_init.c (pthread_spin_init): Use atomic_store_relaxed. * nptl/pthread_spin_lock.c (pthread_spin_lock): Use C11-like atomic macros. * nptl/pthread_spin_trylock.c (pthread_spin_trylock): Likewise. * nptl/pthread_spin_unlock.c (pthread_spin_unlock): Use atomic_store_release. * sysdeps/aarch64/nptl/pthread_spin_lock.c: Delete File. * sysdeps/arm/nptl/pthread_spin_lock.c: Likewise. * sysdeps/hppa/nptl/pthread_spin_lock.c: Likewise. * sysdeps/m68k/nptl/pthread_spin_lock.c: Likewise. * sysdeps/microblaze/nptl/pthread_spin_lock.c: Likewise. * sysdeps/mips/nptl/pthread_spin_lock.c: Likewise. * sysdeps/nios2/nptl/pthread_spin_lock.c: Likewise. * sysdeps/aarch64/atomic-machine.h (ATOMIC_EXCHANGE_USES_CAS): Define. * sysdeps/alpha/atomic-machine.h: Likewise. * sysdeps/arm/atomic-machine.h: Likewise. * sysdeps/i386/atomic-machine.h: Likewise. * sysdeps/ia64/atomic-machine.h: Likewise. * sysdeps/m68k/coldfire/atomic-machine.h: Likewise. * sysdeps/m68k/m680x0/m68020/atomic-machine.h: Likewise. * sysdeps/microblaze/atomic-machine.h: Likewise. * sysdeps/mips/atomic-machine.h: Likewise. * sysdeps/powerpc/powerpc32/atomic-machine.h: Likewise. * sysdeps/powerpc/powerpc64/atomic-machine.h: Likewise. * sysdeps/s390/atomic-machine.h: Likewise. * sysdeps/sparc/sparc32/atomic-machine.h: Likewise. * sysdeps/sparc/sparc32/sparcv9/atomic-machine.h: Likewise. * sysdeps/sparc/sparc64/atomic-machine.h: Likewise. * sysdeps/tile/tilegx/atomic-machine.h: Likewise. * sysdeps/tile/tilepro/atomic-machine.h: Likewise. * sysdeps/unix/sysv/linux/hppa/atomic-machine.h: Likewise. * sysdeps/unix/sysv/linux/m68k/coldfire/atomic-machine.h: Likewise. * sysdeps/unix/sysv/linux/nios2/atomic-machine.h: Likewise. * sysdeps/unix/sysv/linux/sh/atomic-machine.h: Likewise. * sysdeps/x86_64/atomic-machine.h: Likewise. | |||||
* | aarch64: Thunderx specific memcpy and memmove | Steve Ellcey | 2017-05-24 | 8 | -10/+540 | |
| | | | | | | | | | | | | | | * sysdeps/aarch64/memcpy.S (MEMMOVE, MEMCPY): New macros. (memmove): Use MEMMOVE for name. (memcpy): Use MEMCPY for name. Change internal labels to external labels. * sysdeps/aarch64/multiarch/Makefile: New file. * sysdeps/aarch64/multiarch/ifunc-impl-list.c: Likewise. * sysdeps/aarch64/multiarch/init-arch.h: Likewise. * sysdeps/aarch64/multiarch/memcpy.c: Likewise. * sysdeps/aarch64/multiarch/memcpy_generic.S: Likewise. * sysdeps/aarch64/multiarch/memcpy_thunderx.S: Likewise. * sysdeps/aarch64/multiarch/memmove.c: Likewise. | |||||
* | Move shared pthread definitions to common headers | Adhemerval Zanella | 2017-05-09 | 2 | -203/+69 | |
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This patch removes all the replicated pthread definition accross the architectures and consolidates it on shared headers. The new organization is as follow: * Architecture specific definition (such as pthread types sizes) are place in the new pthreadtypes-arch.h header in arch specific path. * All shared structure definition are moved to a common NPTL header at sysdeps/nptl/bits/pthreadtypes.h (with now includes the arch specific one for internal definitions). * Also, for C11 future thread support, both mutex and condition definition are placed in a common header at sysdeps/nptl/bits/thread-shared-types.h. It is also a refactor patch without expected functional changes. Checked with a build for all major ABI (aarch64-linux-gnu, alpha-linux-gnu, arm-linux-gnueabi, i386-linux-gnu, ia64-linux-gnu, m68k-linux-gnu, microblaze-linux-gnu, mips{64}-linux-gnu, nios2-linux-gnu, powerpc{64le}-linux-gnu, s390{x}-linux-gnu, sparc{64}-linux-gnu, tile{pro,gx}-linux-gnu, and x86_64-linux-gnu). * posix/Makefile (headers): Add pthreadtypes-arch.h and thread-shared-types.h. * sysdeps/aarch64/nptl/bits/pthreadtypes-arch.h: New file: arch specific thread definition. * sysdeps/alpha/nptl/bits/pthreadtypes-arch.h: Likewise. * sysdeps/arm/nptl/bits/pthreadtypes-arch.h: Likewise. * sysdeps/hppa/nptl/bits/pthreadtypes-arch.h: Likewise. * sysdeps/ia64/nptl/bits/pthreadtypes-arch.h: Likewise. * sysdeps/m68k/nptl/bits/pthreadtypes-arch.h: Likewise. * sysdeps/microblaze/nptl/bits/pthreadtypes-arch.h: Likewise. * sysdeps/mips/nptl/bits/pthreadtypes-arch.h: Likewise. * sysdeps/nios2/nptl/bits/pthreadtypes-arch.h: Likewise. * sysdeps/powerpc/nptl/bits/pthreadtypes-arch.h: Likewise. * sysdeps/s390/nptl/bits/pthreadtypes-arch.h: Likewise. * sysdeps/sh/nptl/bits/pthreadtypes-arch.h: Likewise. * sysdeps/sparc/nptl/bits/pthreadtypes-arch.h: Likewise. * sysdeps/tile/nptl/bits/pthreadtypes-arch.h: Likewise. * sysdeps/x86/nptl/bits/pthreadtypes-arch.h: Likewise. * sysdeps/nptl/bits/thread-shared-types.h: New file: shared thread definition between POSIX and C11. * sysdeps/aarch64/nptl/bits/pthreadtypes.h.: Remove file. * sysdeps/alpha/nptl/bits/pthreadtypes.h: Likewise. * sysdeps/arm/nptl/bits/pthreadtypes.h: Likewise. * sysdeps/hppa/nptl/bits/pthreadtypes.h: Likewise. * sysdeps/m68k/nptl/bits/pthreadtypes.h: Likewise. * sysdeps/microblaze/nptl/bits/pthreadtypes.h: Likewise. * sysdeps/mips/nptl/bits/pthreadtypes.h: Likewise. * sysdeps/nios2/nptl/bits/pthreadtypes.h: Likewise. * sysdeps/ia64/nptl/bits/pthreadtypes.h: Likewise. * sysdeps/powerpc/nptl/bits/pthreadtypes.h: Likewise. * sysdeps/s390/nptl/bits/pthreadtypes.h: Likewise. * sysdeps/sh/nptl/bits/pthreadtypes.h: Likewise. * sysdeps/sparc/nptl/bits/pthreadtypes.h: Likewise. * sysdeps/tile/nptl/bits/pthreadtypes.h: Likewise. * sysdeps/x86/nptl/bits/pthreadtypes.h: Likewise. * sysdeps/nptl/bits/pthreadtypes.h: New file: common thread definitions shared across all architectures. | |||||
* | [AArch64] Update libm-test-ulps | Szabolcs Nagy | 2017-03-27 | 1 | -80/+88 | |
| | | | | * sysdeps/aarch64/libm-test-ulps: Update. | |||||
* | Add ifunc support for aarch64. | Steve Ellcey | 2017-03-15 | 2 | -0/+19 | |
| | | | | | | | | | | | * sysdeps/aarch64/dl-machine.h: Include cpu-features.c. (DL_PLATFORM_INIT): New define. (dl_platform_init): New function. * sysdeps/aarch64/ldsodefs.h: Include cpu-features.h. * sysdeps/unix/sysv/linux/aarch64/cpu-features.c: New file. * sysdeps/unix/sysv/linux/aarch64/cpu-features.h: Likewise. * sysdeps/unix/sysv/linux/aarch64/dl-procinfo.c: Likewise. * sysdeps/unix/sysv/linux/aarch64/libc-start.c: Likewise. | |||||
* | New pthread rwlock that is more scalable. | Torvald Riegel | 2017-01-10 | 1 | -7/+7 | |
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This replaces the pthread rwlock with a new implementation that uses a more scalable algorithm (primarily through not using a critical section anymore to make state changes). The fast path for rdlock acquisition and release is now basically a single atomic read-modify write or CAS and a few branches. See nptl/pthread_rwlock_common.c for details. * nptl/DESIGN-rwlock.txt: Remove. * nptl/lowlevelrwlock.sym: Remove. * nptl/Makefile: Add new tests. * nptl/pthread_rwlock_common.c: New file. Contains the new rwlock. * nptl/pthreadP.h (PTHREAD_RWLOCK_PREFER_READER_P): Remove. (PTHREAD_RWLOCK_WRPHASE, PTHREAD_RWLOCK_WRLOCKED, PTHREAD_RWLOCK_RWAITING, PTHREAD_RWLOCK_READER_SHIFT, PTHREAD_RWLOCK_READER_OVERFLOW, PTHREAD_RWLOCK_WRHANDOVER, PTHREAD_RWLOCK_FUTEX_USED): New. * nptl/pthread_rwlock_init.c (__pthread_rwlock_init): Adapt to new implementation. * nptl/pthread_rwlock_rdlock.c (__pthread_rwlock_rdlock_slow): Remove. (__pthread_rwlock_rdlock): Adapt. * nptl/pthread_rwlock_timedrdlock.c (pthread_rwlock_timedrdlock): Adapt. * nptl/pthread_rwlock_timedwrlock.c (pthread_rwlock_timedwrlock): Adapt. * nptl/pthread_rwlock_trywrlock.c (pthread_rwlock_trywrlock): Adapt. * nptl/pthread_rwlock_tryrdlock.c (pthread_rwlock_tryrdlock): Adapt. * nptl/pthread_rwlock_unlock.c (pthread_rwlock_unlock): Adapt. * nptl/pthread_rwlock_wrlock.c (__pthread_rwlock_wrlock_slow): Remove. (__pthread_rwlock_wrlock): Adapt. * nptl/tst-rwlock10.c: Adapt. * nptl/tst-rwlock11.c: Adapt. * nptl/tst-rwlock17.c: New file. * nptl/tst-rwlock18.c: New file. * nptl/tst-rwlock19.c: New file. * nptl/tst-rwlock2b.c: New file. * nptl/tst-rwlock8.c: Adapt. * nptl/tst-rwlock9.c: Adapt. * sysdeps/aarch64/nptl/bits/pthreadtypes.h (pthread_rwlock_t): Adapt. * sysdeps/arm/nptl/bits/pthreadtypes.h (pthread_rwlock_t): Adapt. * sysdeps/hppa/nptl/bits/pthreadtypes.h (pthread_rwlock_t): Adapt. * sysdeps/ia64/nptl/bits/pthreadtypes.h (pthread_rwlock_t): Adapt. * sysdeps/m68k/nptl/bits/pthreadtypes.h (pthread_rwlock_t): Adapt. * sysdeps/microblaze/nptl/bits/pthreadtypes.h (pthread_rwlock_t): Adapt. * sysdeps/mips/nptl/bits/pthreadtypes.h (pthread_rwlock_t): Adapt. * sysdeps/nios2/nptl/bits/pthreadtypes.h (pthread_rwlock_t): Adapt. * sysdeps/s390/nptl/bits/pthreadtypes.h (pthread_rwlock_t): Adapt. * sysdeps/sh/nptl/bits/pthreadtypes.h (pthread_rwlock_t): Adapt. * sysdeps/sparc/nptl/bits/pthreadtypes.h (pthread_rwlock_t): Adapt. * sysdeps/tile/nptl/bits/pthreadtypes.h (pthread_rwlock_t): Adapt. * sysdeps/unix/sysv/linux/alpha/bits/pthreadtypes.h (pthread_rwlock_t): Adapt. * sysdeps/unix/sysv/linux/powerpc/bits/pthreadtypes.h (pthread_rwlock_t): Adapt. * sysdeps/x86/bits/pthreadtypes.h (pthread_rwlock_t): Adapt. * nptl/nptl-printers.py (): Adapt. * nptl/nptl_lock_constants.pysym: Adapt. * nptl/test-rwlock-printers.py: Adapt. * nptl/test-rwlockattr-printers.c: Adapt. * nptl/test-rwlockattr-printers.py: Adapt. | |||||
* | Update copyright dates with scripts/update-copyrights. | Joseph Myers | 2017-01-01 | 107 | -107/+107 | |
| | ||||||
* | New condvar implementation that provides stronger ordering guarantees. | Torvald Riegel | 2016-12-31 | 1 | -9/+22 | |
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This is a new implementation for condition variables, required after http://austingroupbugs.net/view.php?id=609 to fix bug 13165. In essence, we need to be stricter in which waiters a signal or broadcast is required to wake up; this couldn't be solved using the old algorithm. ISO C++ made a similar clarification, so this also fixes a bug in current libstdc++, for example. We can't use the old algorithm anymore because futexes do not guarantee to wake in FIFO order. Thus, when we wake, we can't simply let any waiter grab a signal, but we need to ensure that one of the waiters happening before the signal is woken up. This is something the previous algorithm violated (see bug 13165). There's another issue specific to condvars: ABA issues on the underlying futexes. Unlike mutexes that have just three states, or semaphores that have no tokens or a limited number of them, the state of a condvar is the *order* of the waiters. A waiter on a semaphore can grab a token whenever one is available; a condvar waiter must only consume a signal if it is eligible to do so as determined by the relative order of the waiter and the signal. Therefore, this new algorithm maintains two groups of waiters: Those eligible to consume signals (G1), and those that have to wait until previous waiters have consumed signals (G2). Once G1 is empty, G2 becomes the new G1. 64b counters are used to avoid ABA issues. This condvar doesn't yet use a requeue optimization (ie, on a broadcast, waking just one thread and requeueing all others on the futex of the mutex supplied by the program). I don't think doing the requeue is necessarily the right approach (but I haven't done real measurements yet): * If a program expects to wake many threads at the same time and make that scalable, a condvar isn't great anyway because of how it requires waiters to operate mutually exclusive (due to the mutex usage). Thus, a thundering herd problem is a scalability problem with or without the optimization. Using something like a semaphore might be more appropriate in such a case. * The scalability problem is actually at the mutex side; the condvar could help (and it tries to with the requeue optimization), but it should be the mutex who decides how that is done, and whether it is done at all. * Forcing all but one waiter into the kernel-side wait queue of the mutex prevents/avoids the use of lock elision on the mutex. Thus, it prevents the only cure against the underlying scalability problem inherent to condvars. * If condvars use short critical sections (ie, hold the mutex just to check a binary flag or such), which they should do ideally, then forcing all those waiter to proceed serially with kernel-based hand-off (ie, futex ops in the mutex' contended state, via the futex wait queues) will be less efficient than just letting a scalable mutex implementation take care of it. Our current mutex impl doesn't employ spinning at all, but if critical sections are short, spinning can be much better. * Doing the requeue stuff requires all waiters to always drive the mutex into the contended state. This leads to each waiter having to call futex_wake after lock release, even if this wouldn't be necessary. [BZ #13165] * nptl/pthread_cond_broadcast.c (__pthread_cond_broadcast): Rewrite to use new algorithm. * nptl/pthread_cond_destroy.c (__pthread_cond_destroy): Likewise. * nptl/pthread_cond_init.c (__pthread_cond_init): Likewise. * nptl/pthread_cond_signal.c (__pthread_cond_signal): Likewise. * nptl/pthread_cond_wait.c (__pthread_cond_wait): Likewise. (__pthread_cond_timedwait): Move here from pthread_cond_timedwait.c. (__condvar_confirm_wakeup, __condvar_cancel_waiting, __condvar_cleanup_waiting, __condvar_dec_grefs, __pthread_cond_wait_common): New. (__condvar_cleanup): Remove. * npt/pthread_condattr_getclock.c (pthread_condattr_getclock): Adapt. * npt/pthread_condattr_setclock.c (pthread_condattr_setclock): Likewise. * npt/pthread_condattr_getpshared.c (pthread_condattr_getpshared): Likewise. * npt/pthread_condattr_init.c (pthread_condattr_init): Likewise. * nptl/tst-cond1.c: Add comment. * nptl/tst-cond20.c (do_test): Adapt. * nptl/tst-cond22.c (do_test): Likewise. * sysdeps/aarch64/nptl/bits/pthreadtypes.h (pthread_cond_t): Adapt structure. * sysdeps/arm/nptl/bits/pthreadtypes.h (pthread_cond_t): Likewise. * sysdeps/ia64/nptl/bits/pthreadtypes.h (pthread_cond_t): Likewise. * sysdeps/m68k/nptl/bits/pthreadtypes.h (pthread_cond_t): Likewise. * sysdeps/microblaze/nptl/bits/pthreadtypes.h (pthread_cond_t): Likewise. * sysdeps/mips/nptl/bits/pthreadtypes.h (pthread_cond_t): Likewise. * sysdeps/nios2/nptl/bits/pthreadtypes.h (pthread_cond_t): Likewise. * sysdeps/s390/nptl/bits/pthreadtypes.h (pthread_cond_t): Likewise. * sysdeps/sh/nptl/bits/pthreadtypes.h (pthread_cond_t): Likewise. * sysdeps/tile/nptl/bits/pthreadtypes.h (pthread_cond_t): Likewise. * sysdeps/unix/sysv/linux/alpha/bits/pthreadtypes.h (pthread_cond_t): Likewise. * sysdeps/unix/sysv/linux/powerpc/bits/pthreadtypes.h (pthread_cond_t): Likewise. * sysdeps/x86/bits/pthreadtypes.h (pthread_cond_t): Likewise. * sysdeps/nptl/internaltypes.h (COND_NWAITERS_SHIFT): Remove. (COND_CLOCK_BITS): Adapt. * sysdeps/nptl/pthread.h (PTHREAD_COND_INITIALIZER): Adapt. * nptl/pthreadP.h (__PTHREAD_COND_CLOCK_MONOTONIC_MASK, __PTHREAD_COND_SHARED_MASK): New. * nptl/nptl-printers.py (CLOCK_IDS): Remove. (ConditionVariablePrinter, ConditionVariableAttributesPrinter): Adapt. * nptl/nptl_lock_constants.pysym: Adapt. * nptl/test-cond-printers.py: Adapt. * sysdeps/unix/sysv/linux/hppa/internaltypes.h (cond_compat_clear, cond_compat_check_and_clear): Adapt. * sysdeps/unix/sysv/linux/hppa/pthread_cond_timedwait.c: Remove file ... * sysdeps/unix/sysv/linux/hppa/pthread_cond_wait.c (__pthread_cond_timedwait): ... and move here. * nptl/DESIGN-condvar.txt: Remove file. * nptl/lowlevelcond.sym: Likewise. * nptl/pthread_cond_timedwait.c: Likewise. * sysdeps/unix/sysv/linux/i386/i486/pthread_cond_broadcast.S: Likewise. * sysdeps/unix/sysv/linux/i386/i486/pthread_cond_signal.S: Likewise. * sysdeps/unix/sysv/linux/i386/i486/pthread_cond_timedwait.S: Likewise. * sysdeps/unix/sysv/linux/i386/i486/pthread_cond_wait.S: Likewise. * sysdeps/unix/sysv/linux/i386/i586/pthread_cond_broadcast.S: Likewise. * sysdeps/unix/sysv/linux/i386/i586/pthread_cond_signal.S: Likewise. * sysdeps/unix/sysv/linux/i386/i586/pthread_cond_timedwait.S: Likewise. * sysdeps/unix/sysv/linux/i386/i586/pthread_cond_wait.S: Likewise. * sysdeps/unix/sysv/linux/i386/i686/pthread_cond_broadcast.S: Likewise. * sysdeps/unix/sysv/linux/i386/i686/pthread_cond_signal.S: Likewise. * sysdeps/unix/sysv/linux/i386/i686/pthread_cond_timedwait.S: Likewise. * sysdeps/unix/sysv/linux/i386/i686/pthread_cond_wait.S: Likewise. * sysdeps/unix/sysv/linux/x86_64/pthread_cond_broadcast.S: Likewise. * sysdeps/unix/sysv/linux/x86_64/pthread_cond_signal.S: Likewise. * sysdeps/unix/sysv/linux/x86_64/pthread_cond_timedwait.S: Likewise. * sysdeps/unix/sysv/linux/x86_64/pthread_cond_wait.S: Likewise. | |||||
* | Refactor long double information into bits/long-double.h. | Joseph Myers | 2016-12-14 | 1 | -21/+0 | |
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Information about whether the ABI of long double is the same as that of double is split between bits/mathdef.h and bits/wordsize.h. When the ABIs are the same, bits/mathdef.h defines __NO_LONG_DOUBLE_MATH. In addition, in the case where the same glibc binary supports both -mlong-double-64 and -mlong-double-128, bits/wordsize.h defines __LONG_DOUBLE_MATH_OPTIONAL, along with __NO_LONG_DOUBLE_MATH if this particular compilation is with -mlong-double-64. As part of the refactoring I proposed in <https://sourceware.org/ml/libc-alpha/2016-11/msg00745.html>, this patch puts all that information in a single header, bits/long-double.h. It is included from sys/cdefs.h alongside the include of bits/wordsize.h, so other headers generally do not need to include bits/long-double.h directly. Previously, various bits/mathdef.h headers and bits/wordsize.h headers had this long double information (including implicitly in some bits/mathdef.h headers through not having the defines present in the default version). After the patch, it's all in six bits/long-double.h headers. Furthermore, most of those new headers are not architecture-specific. Architectures with optional long double all use the ldbl-opt sysdeps directory, either in the order (ldbl-64-128, ldbl-opt, ldbl-128) or (ldbl-128ibm, ldbl-opt). Thus a generic header for the case where long double = double, and headers in ldbl-128, ldbl-96 and ldbl-opt, suffices to cover every architecture except for cases where long double properties vary between different ABIs sharing a set of installed headers; fortunately all the ldbl-opt cases share a single compiler-predefined macro __LONG_DOUBLE_128__ that can be used to tell whether this compilation is -mlong-double-64 or -mlong-double-128. The two cases where a set of headers is shared between ABIs with different long double properties, MIPS (o32 has long double = double, other ABIs use ldbl-128) and SPARC (32-bit has optional long double, 64-bit has required long double), need their own bits/long-double.h headers. As with bits/wordsize.h, multiple-include protection for this header is generally implicit through the include guards on sys/cdefs.h, and multiple inclusion is harmless in any case. There is one subtlety: the header must not define __LONG_DOUBLE_MATH_OPTIONAL if __NO_LONG_DOUBLE_MATH was defined before its inclusion, because doing so breaks how sysdeps/ieee754/ldbl-opt/nldbl-compat.h defines __NO_LONG_DOUBLE_MATH itself before including system headers. Subject to keeping that working, it would be reasonable to move these macros from defined/undefined #ifdef to always-defined 1/0 #if semantics, but this patch does not attempt to do so, just rearranges where the macros are defined. After this patch, the only use of bits/mathdef.h is the alpha one for modifying complex function ABIs for old GCC. Thus, all versions of the header other than the default and alpha versions are removed, as is the include from math.h. Tested for x86_64 and x86. Also did compilation-only testing with build-many-glibcs.py. * bits/long-double.h: New file. * sysdeps/ieee754/ldbl-128/bits/long-double.h: Likewise. * sysdeps/ieee754/ldbl-96/bits/long-double.h: Likewise. * sysdeps/ieee754/ldbl-opt/bits/long-double.h: Likewise. * sysdeps/mips/bits/long-double.h: Likewise. * sysdeps/unix/sysv/linux/sparc/bits/long-double.h: Likewise. * math/Makefile (headers): Add bits/long-double.h. * misc/sys/cdefs.h: Include <bits/long-double.h>. * stdlib/strtold.c: Include <bits/long-double.h> instead of <bits/wordsize.h>. * bits/mathdef.h [!_COMPLEX_H]: Do not allow inclusion. [!__NO_LONG_DOUBLE_MATH]: Remove conditional code. * math/math.h: Do not include <bits/mathdef.h>. * sysdeps/aarch64/bits/mathdef.h: Remove file. * sysdeps/alpha/bits/mathdef.h [!_COMPLEX_H]: Do not allow inclusion. * sysdeps/ia64/bits/mathdef.h: Remove file. * sysdeps/m68k/m680x0/bits/mathdef.h: Likewise. * sysdeps/mips/bits/mathdef.h: Likewise. * sysdeps/powerpc/bits/mathdef.h: Likewise. * sysdeps/s390/bits/mathdef.h: Likewise. * sysdeps/sparc/bits/mathdef.h: Likewise. * sysdeps/x86/bits/mathdef.h: Likewise. * sysdeps/s390/s390-32/bits/wordsize.h [!__NO_LONG_DOUBLE_MATH && !__LONG_DOUBLE_MATH_OPTIONAL]: Remove conditional code. * sysdeps/s390/s390-64/bits/wordsize.h [!__NO_LONG_DOUBLE_MATH && !__LONG_DOUBLE_MATH_OPTIONAL]: Likewise. * sysdeps/unix/sysv/linux/alpha/bits/wordsize.h [!__NO_LONG_DOUBLE_MATH && !__LONG_DOUBLE_MATH_OPTIONAL]: Likewise. * sysdeps/unix/sysv/linux/powerpc/bits/wordsize.h [!__NO_LONG_DOUBLE_MATH && !__LONG_DOUBLE_MATH_OPTIONAL]: Likewise. * sysdeps/unix/sysv/linux/sparc/bits/wordsize.h [!__NO_LONG_DOUBLE_MATH && !__LONG_DOUBLE_MATH_OPTIONAL]: Likewise. | |||||
* | aarch64: Use explicit offsets in _dl_tlsdesc_dynamic | Florian Weimer | 2016-12-02 | 2 | -9/+12 | |
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Commit 389d1f1b232b3d6b9d73ee2c50e543ace6675621 (“Partial ILP32 support for aarch64”) broke dynamic TLS support because a load offset changed: 0000000000000030 <_dl_tlsdesc_dynamic>: 30: a9bc7bfd stp x29, x30, [sp,#-64]! 34: 910003fd mov x29, sp 38: a9020be1 stp x1, x2, [sp,#32] 3c: a90313e3 stp x3, x4, [sp,#48] 40: d53bd044 mrs x4, tpidr_el0 44: c8dffc1f ldar xzr, [x0] 48: f9400401 ldr x1, [x0,#8] 4c: f9400080 ldr x0, [x4] 50: f9400823 ldr x3, [x1,#16] 54: f9400002 ldr x2, [x0] 58: eb02007f cmp x3, x2 5c: 540001a8 b.hi 90 <_dl_tlsdesc_dynamic+0x60> 60: f9400022 ldr x2, [x1] 64: 8b021000 add x0, x0, x2, lsl #4 68: f9400000 ldr x0, [x0] 6c: b100041f cmn x0, #0x1 70: 54000100 b.eq 90 <_dl_tlsdesc_dynamic+0x60> - 74: f9400421 ldr x1, [x1,#8] + 74: f9400821 ldr x1, [x1,#16] 78: 8b010000 add x0, x0, x1 … This commit introduces explicit struct offsets, generated from the C headers, fixing the regression. | |||||
* | Refactor FP_ILOGB* out of bits/mathdef.h. | Joseph Myers | 2016-12-01 | 1 | -9/+0 | |
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Continuing the refactoring of bits/mathdef.h, this patch stops it defining FP_ILOGB0 and FP_ILOGBNAN, moving the required information to a new header bits/fp-logb.h. There are only two possible values of each of those macros permitted by ISO C. TS 18661-1 adds corresponding macros for llogb, and their values are required to correspond to those of the ilogb macros in the obvious way. Thus two boolean values - for which the same choices are correct for most architectures - suffice to determine the value of all these macros, and by defining macros for those boolean values in bits/fp-logb.h we can then define the public FP_* macros in math.h and avoid the present duplication of the associated feature test macro logic. This patch duly moves to bits/fp-logb.h defining __FP_LOGB0_IS_MIN and __FP_LOGBNAN_IS_MIN. Default definitions of those to 0 are correct for both architectures, while ia64, m68k and x86 get their own versions of bits/fp-logb.h to reflect their use of values different from the defaults. The patch renders many copies of bits/mathdef.h trivial (needed only to avoid the default __NO_LONG_DOUBLE_MATH). I'll revise <https://sourceware.org/ml/libc-alpha/2016-11/msg00865.html> accordingly so that it removes all bits/mathdef.h headers except the default one and the alpha one, and arranges for the header to be included only by complex.h as the only remaining use at that point will be for the alpha ABI issues there. Tested for x86_64 and x86. Also did compile-only testing with build-many-glibcs.py (using glibc sources from before the commit that introduced many build failures with undefined __GI___sigsetjmp). * bits/fp-logb.h: New file. * sysdeps/ia64/bits/fp-logb.h: Likewise. * sysdeps/m68k/m680x0/bits/fp-logb.h: Likewise. * sysdeps/x86/bits/fp-logb.h: Likewise. * math/Makefile (headers): Add bits/fp-logb.h. * math/math.h: Include <bits/fp-logb.h>. [__USE_ISOC99] (FP_ILOGB0): Define based on __FP_LOGB0_IS_MIN. [__USE_ISOC99] (FP_ILOGBNAN): Define based on __FP_LOGBNAN_IS_MIN. * bits/mathdef.h (FP_ILOGB0): Remove. (FP_ILOGBNAN): Likewise. * sysdeps/aarch64/bits/mathdef.h (FP_ILOGB0): Likewise. (FP_ILOGBNAN): Likewise. * sysdeps/alpha/bits/mathdef.h (FP_ILOGB0): Likewise. (FP_ILOGBNAN): Likewise. * sysdeps/ia64/bits/mathdef.h (FP_ILOGB0): Likewise. (FP_ILOGBNAN): Likewise. * sysdeps/m68k/m680x0/bits/mathdef.h (FP_ILOGB0): Likewise. (FP_ILOGBNAN): Likewise. * sysdeps/mips/bits/mathdef.h (FP_ILOGB0): Likewise. (FP_ILOGBNAN): Likewise. * sysdeps/powerpc/bits/mathdef.h (FP_ILOGB0): Likewise. (FP_ILOGBNAN): Likewise. * sysdeps/s390/bits/mathdef.h (FP_ILOGB0): Likewise. (FP_ILOGBNAN): Likewise. * sysdeps/sparc/bits/mathdef.h (FP_ILOGB0): Likewise. (FP_ILOGBNAN): Likewise. * sysdeps/x86/bits/mathdef.h (FP_ILOGB0): Likewise. (FP_ILOGBNAN): Likewise. | |||||
* | Refactor FP_FAST_* into bits/fp-fast.h. | Joseph Myers | 2016-11-29 | 2 | -3/+34 | |
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Continuing the refactoring of bits/mathdef.h, this patch moves the FP_FAST_* definitions into a new bits/fp-fast.h header. Currently this is only for FP_FAST_FMA*, but in future it would be the appropriate place for the FP_FAST_* macros from TS 18661-1 as well. The generic bits/mathdef.h header defines these macros based on whether the compiler defines __FP_FAST_*. Most architecture-specific headers, however, fail to do so, meaning that if the architecture (or some particular processors) does in fact have fused operations, and GCC knows to use them inline, the FP_FAST_* macros will still not be defined. By refactoring, this patch causes the generic version (based on __FP_FAST_*) to be used in more cases, and so the macro definitions to be more accurate. Architectures that already defined some or all of these macros other than based on the predefines have their own versions of fp-fast.h, which are arranged so they define FP_FAST_* if either the architecture-specific conditions are true or __FP_FAST_* are defined. After this refactoring, various bits/mathdef.h headers for architectures with long double = double are semantically identical to the generic version. The patch removes those headers that are redundant. (In fact two of the four removed were already redundant before this patch because they did use __FP_FAST_*.) Tested for x86_64 and x86, and compilation-only with build-many-glibcs.py. * bits/fp-fast.h: New file. * sysdeps/aarch64/bits/fp-fast.h: Likewise. * sysdeps/powerpc/bits/fp-fast.h: Likewise. * math/Makefile (headers): Add bits/fp-fast.h. * math/math.h: Include <bits/fp-fast.h>. * bits/mathdef.h (FP_FAST_FMA): Remove. (FP_FAST_FMAF): Likewise. (FP_FAST_FMAL): Likewise. * sysdeps/aarch64/bits/mathdef.h (FP_FAST_FMA): Likewise. (FP_FAST_FMAF): Likewise. * sysdeps/powerpc/bits/mathdef.h (FP_FAST_FMA): Likewise. (FP_FAST_FMAF): Likewise. * sysdeps/x86/bits/mathdef.h (FP_FAST_FMA): Likewise. (FP_FAST_FMAF): Likewise. (FP_FAST_FMAL): Likewise. * sysdeps/arm/bits/mathdef.h: Remove file. * sysdeps/hppa/fpu/bits/mathdef.h: Likewise. * sysdeps/sh/sh4/bits/mathdef.h: Likewise. * sysdeps/tile/bits/mathdef.h: Likewise. | |||||
* | Partial ILP32 support for aarch64. | Steve Ellcey | 2016-11-28 | 23 | -156/+266 | |
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | * sysdeps/aarch64/crti.S: Add include of sysdep.h. (call_weak_fn): Use PTR_REG to get correct reg name in ILP32. * sysdeps/aarch64/dl-irel.h: Add include of sysdep.h. (elf_irela): Use AARCH64_R macro to get correct relocation in ILP32. * sysdeps/aarch64/dl-machine.h: Add include of sysdep.h. (elf_machine_load_address, RTLD_START, RTLD_START_1, RTLD_START, elf_machine_type_class, ELF_MACHINE_JMP_SLOT, elf_machine_rela, elf_machine_lazy_rel): Add ifdef's for ILP32 support. * sysdeps/aarch64/dl-tlsdesc.S (_dl_tlsdesc_return, _dl_tlsdesc_return_lazy, _dl_tlsdesc_dynamic, _dl_tlsdesc_resolve_hold): Extend pointers in ILP32, use PTR_REG to get correct reg name for ILP32. * sysdeps/aarch64/dl-trampoline.S (ip01): New Macro. (RELA_SIZE): New Macro. (_dl_runtime_resolve, _dl_runtime_profile): Use new macros and PTR_REG to support ILP32. * sysdeps/aarch64/jmpbuf-unwind.h (_JMPBUF_CFA_UNWINDS_ADJ): Add cast for ILP32 mode. * sysdeps/aarch64/memcmp.S (memcmp): Extend arg pointers for ILP32 mode. * sysdeps/aarch64/memcpy.S (memmove, memcpy): Ditto. * sysdeps/aarch64/memset.S (__memset): Ditto. * sysdeps/aarch64/strchr.S (strchr): Ditto. * sysdeps/aarch64/strchrnul.S (__strchrnul): Ditto. * sysdeps/aarch64/strcmp.S (strcmp): Ditto. * sysdeps/aarch64/strcpy.S (strcpy): Ditto. * sysdeps/aarch64/strlen.S (__strlen): Ditto. * sysdeps/aarch64/strncmp.S (strncmp): Ditto. * sysdeps/aarch64/strnlen.S (strnlen): Ditto. * sysdeps/aarch64/strrchr.S (strrchr): Ditto. * sysdeps/unix/sysv/linux/aarch64/clone.S: Ditto. * sysdeps/unix/sysv/linux/aarch64/setcontext.S (__setcontext): Ditto. * sysdeps/unix/sysv/linux/aarch64/swapcontext.S (__swapcontext): Ditto. * sysdeps/aarch64/__longjmp.S (__longjmp): Extend pointers in ILP32, change PTR_MANGLE call to use register numbers instead of names. * sysdeps/unix/sysv/linux/aarch64/getcontext.S (__getcontext): Ditto. * sysdeps/aarch64/setjmp.S (__sigsetjmp): Extend arg pointers for ILP32 mode, change PTR_MANGLE calls to use register numbers. * sysdeps/aarch64/start.S (_start): Ditto. * sysdeps/aarch64/nptl/bits/pthreadtypes.h (__PTHREAD_RWLOCK_INT_FLAGS_SHARED): New define. (__SIZEOF_PTHREAD_ATTR_T, __SIZEOF_PTHREAD_MUTEX_T, __SIZEOF_PTHREAD_MUTEXATTR_T, __SIZEOF_PTHREAD_COND_T, __SIZEOF_PTHREAD_COND_COMPAT_T, __SIZEOF_PTHREAD_CONDATTR_T, __SIZEOF_PTHREAD_RWLOCK_T, __SIZEOF_PTHREAD_RWLOCKATTR_T, __SIZEOF_PTHREAD_BARRIER_T, __SIZEOF_PTHREAD_BARRIERATTR_T): Make defined values dependent on __ILP32__. * sysdeps/aarch64/nptl/bits/semaphore.h (__SIZEOF_SEM_T): Change define. (sem_t): Change __align type. * sysdeps/aarch64/sysdep.h (AARCH64_R, PTR_REG, PTR_LOG_SIZE, DELOUSE, PTR_SIZE): New Macros. (LDST_PCREL, LDST_GLOBAL) Update to use PTR_REG. * sysdeps/unix/sysv/linux/aarch64/bits/fcntl.h (O_LARGEFILE): Set when in ILP32 mode. (F_GETLK64, F_SETLK64, F_SETLKW64): Only set in LP64 mode. * sysdeps/unix/sysv/linux/aarch64/dl-cache.h (DL_CACHE_DEFAULT_ID): Set elf flags for ILP32. (add_system_dir): Set ILP32 library directories. * sysdeps/unix/sysv/linux/aarch64/init-first.c (_libc_vdso_platform_setup): Set minimum kernel version for ILP32. * sysdeps/unix/sysv/linux/aarch64/ldconfig.h (SYSDEP_KNOWN_INTERPRETER_NAMES): Add ILP32 names. * sysdeps/unix/sysv/linux/aarch64/sigcontextinfo.h (GET_PC, SET_PC): New Macros. * sysdeps/unix/sysv/linux/aarch64/sysdep.h: Handle ILP32 pointers. | |||||
* | Remove cached PID/TID in clone | Adhemerval Zanella | 2016-11-24 | 1 | -1/+0 | |
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This patch remove the PID cache and usage in current GLIBC code. Current usage is mainly used a performance optimization to avoid the syscall, however it adds some issues: - The exposed clone syscall will try to set pid/tid to make the new thread somewhat compatible with current GLIBC assumptions. This cause a set of issue with new workloads and usecases (such as BZ#17214 and [1]) as well for new internal usage of clone to optimize other algorithms (such as clone plus CLONE_VM for posix_spawn, BZ#19957). - The caching complexity also added some bugs in the past [2] [3] and requires more effort of each port to handle such requirements (for both clone and vfork implementation). - Caching performance gain in mainly on getpid and some specific code paths. The getpid performance leverage is questionable [4], either by the idea of getpid being a hotspot as for the getpid implementation itself (if it is indeed a justifiable hotspot a vDSO symbol could let to a much more simpler solution). Other usage is mainly for non usual code paths, such as pthread cancellation signal and handling. For thread creation (on stack allocation) the code simplification in fact adds some performance gain due the no need of transverse the stack cache and invalidate each element pid. Other thread usages will require a direct getpid syscall, such as cancellation/setxid signal, thread cancellation, thread fail path (at create_thread), and thread signal (pthread_kill and pthread_sigqueue). However these are hardly usual hotspots and I think adding a syscall is justifiable. It also simplifies both the clone and vfork arch-specific implementation. And by review each fork implementation there are some discrepancies that this patch also solves: - microblaze clone/vfork does not set/reset the pid/tid field - hppa uses the default vfork implementation that fallback to fork. Since vfork is deprecated I do not think we should bother with it. The patch also removes the TID caching in clone. My understanding for such semantic is try provide some pthread usage after a user program issue clone directly (as done by thread creation with CLONE_PARENT_SETTID and pthread tid member). However, as stated before in multiple discussions threads, GLIBC provides clone syscalls without further supporting all this semantics. I ran a full make check on x86_64, x32, i686, armhf, aarch64, and powerpc64le. For sparc32, sparc64, and mips I ran the basic fork and vfork tests from posix/ folder (on a qemu system). So it would require further testing on alpha, hppa, ia64, m68k, nios2, s390, sh, and tile (I excluded microblaze because it is already implementing the patch semantic regarding clone/vfork). [1] https://codereview.chromium.org/800183004/ [2] https://sourceware.org/ml/libc-alpha/2006-07/msg00123.html [3] https://sourceware.org/bugzilla/show_bug.cgi?id=15368 [4] http://yarchive.net/comp/linux/getpid_caching.html * sysdeps/nptl/fork.c (__libc_fork): Remove pid cache setting. * nptl/allocatestack.c (allocate_stack): Likewise. (__reclaim_stacks): Likewise. (setxid_signal_thread): Obtain pid through syscall. * nptl/nptl-init.c (sigcancel_handler): Likewise. (sighandle_setxid): Likewise. * nptl/pthread_cancel.c (pthread_cancel): Likewise. * sysdeps/unix/sysv/linux/pthread_kill.c (__pthread_kill): Likewise. * sysdeps/unix/sysv/linux/pthread_sigqueue.c (pthread_sigqueue): Likewise. * sysdeps/unix/sysv/linux/createthread.c (create_thread): Likewise. * sysdeps/unix/sysv/linux/getpid.c: Remove file. * nptl/descr.h (struct pthread): Change comment about pid value. * nptl/pthread_getattr_np.c (pthread_getattr_np): Remove thread pid assert. * sysdeps/unix/sysv/linux/pthread-pids.h (__pthread_initialize_pids): Do not set pid value. * nptl_db/td_ta_thr_iter.c (iterate_thread_list): Remove thread pid cache check. * nptl_db/td_thr_validate.c (td_thr_validate): Likewise. * sysdeps/aarch64/nptl/tcb-offsets.sym: Remove pid offset. * sysdeps/alpha/nptl/tcb-offsets.sym: Likewise. * sysdeps/arm/nptl/tcb-offsets.sym: Likewise. * sysdeps/hppa/nptl/tcb-offsets.sym: Likewise. * sysdeps/i386/nptl/tcb-offsets.sym: Likewise. * sysdeps/ia64/nptl/tcb-offsets.sym: Likewise. * sysdeps/m68k/nptl/tcb-offsets.sym: Likewise. * sysdeps/microblaze/nptl/tcb-offsets.sym: Likewise. * sysdeps/mips/nptl/tcb-offsets.sym: Likewise. * sysdeps/nios2/nptl/tcb-offsets.sym: Likewise. * sysdeps/powerpc/nptl/tcb-offsets.sym: Likewise. * sysdeps/s390/nptl/tcb-offsets.sym: Likewise. * sysdeps/sh/nptl/tcb-offsets.sym: Likewise. * sysdeps/sparc/nptl/tcb-offsets.sym: Likewise. * sysdeps/tile/nptl/tcb-offsets.sym: Likewise. * sysdeps/x86_64/nptl/tcb-offsets.sym: Likewise. * sysdeps/unix/sysv/linux/aarch64/clone.S: Remove pid and tid caching. * sysdeps/unix/sysv/linux/alpha/clone.S: Likewise. * sysdeps/unix/sysv/linux/arm/clone.S: Likewise. * sysdeps/unix/sysv/linux/hppa/clone.S: Likewise. * sysdeps/unix/sysv/linux/i386/clone.S: Likewise. * sysdeps/unix/sysv/linux/ia64/clone2.S: Likewise. * sysdeps/unix/sysv/linux/mips/clone.S: Likewise. * sysdeps/unix/sysv/linux/nios2/clone.S: Likewise. * sysdeps/unix/sysv/linux/powerpc/powerpc32/clone.S: Likewise. * sysdeps/unix/sysv/linux/powerpc/powerpc64/clone.S: Likewise. * sysdeps/unix/sysv/linux/s390/s390-32/clone.S: Likewise. * sysdeps/unix/sysv/linux/s390/s390-64/clone.S: Likewise. * sysdeps/unix/sysv/linux/sh/clone.S: Likewise. * sysdeps/unix/sysv/linux/sparc/sparc32/clone.S: Likewise. * sysdeps/unix/sysv/linux/sparc/sparc64/clone.S: Likewise. * sysdeps/unix/sysv/linux/tile/clone.S: Likewise. * sysdeps/unix/sysv/linux/x86_64/clone.S: Likewise. * sysdeps/unix/sysv/linux/aarch64/vfork.S: Remove pid set and reset. * sysdeps/unix/sysv/linux/alpha/vfork.S: Likewise. * sysdeps/unix/sysv/linux/arm/vfork.S: Likewise. * sysdeps/unix/sysv/linux/i386/vfork.S: Likewise. * sysdeps/unix/sysv/linux/ia64/vfork.S: Likewise. * sysdeps/unix/sysv/linux/m68k/clone.S: Likewise. * sysdeps/unix/sysv/linux/m68k/vfork.S: Likewise. * sysdeps/unix/sysv/linux/mips/vfork.S: Likewise. * sysdeps/unix/sysv/linux/nios2/vfork.S: Likewise. * sysdeps/unix/sysv/linux/powerpc/powerpc32/vfork.S: Likewise. * sysdeps/unix/sysv/linux/powerpc/powerpc64/vfork.S: Likewise. * sysdeps/unix/sysv/linux/s390/s390-32/vfork.S: Likewise. * sysdeps/unix/sysv/linux/s390/s390-64/vfork.S: Likewise. * sysdeps/unix/sysv/linux/sh/vfork.S: Likewise. * sysdeps/unix/sysv/linux/sparc/sparc32/vfork.S: Likewise. * sysdeps/unix/sysv/linux/sparc/sparc64/vfork.S: Likewise. * sysdeps/unix/sysv/linux/tile/vfork.S: Likewise. * sysdeps/unix/sysv/linux/x86_64/vfork.S: Likewise. * sysdeps/unix/sysv/linux/tst-clone2.c (f): Remove direct pthread struct access. (clone_test): Remove function. (do_test): Rewrite to take in consideration pid is not cached anymore. | |||||
* | Refactor float_t, double_t information into bits/flt-eval-method.h. | Joseph Myers | 2016-11-24 | 1 | -6/+0 | |
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | At present, definitions of float_t and double_t are split among many bits/mathdef.h headers. For all but three architectures, these types are float and double. Furthermore, if you assume __FLT_EVAL_METHOD__ to be defined, that provides a more generic way of determining the correct values of these typedefs. Defining these typedefs more generally based on __FLT_EVAL_METHOD__ was previously proposed by Paul Eggert in <https://sourceware.org/ml/libc-alpha/2012-02/msg00002.html>. This patch refactors things in the way I proposed in <https://sourceware.org/ml/libc-alpha/2016-11/msg00745.html>. A new header bits/flt-eval-method.h defines a single macro, __GLIBC_FLT_EVAL_METHOD, which is then used by math.h to define float_t and double_t. The default is based on __FLT_EVAL_METHOD__ (although actually a default to 0 would have the same effect for current ports, because ports where values other than 0 or 16 are possible all have their own headers). To avoid changing the existing semantics in any case, including for compilers not defining __FLT_EVAL_METHOD__, architecture-specific files are then added for m68k, s390, x86 which replicate the existing semantics. At least with __FLT_EVAL_METHOD__ values possible with GCC, there should be no change to the choices of float_t and double_t for any supported configuration. Architecture maintainer notes: * m68k: sysdeps/m68k/m680x0/bits/flt-eval-method.h always defines __GLIBC_FLT_EVAL_METHOD to 2 to replicate the existing logic. But actually GCC defines __FLT_EVAL_METHOD__ to 0 if TARGET_68040. It might make sense to make the header prefer to base things on __FLT_EVAL_METHOD__ if defined, like the x86 version, and so make the choices of these types more accurate (with a NEWS entry as for the other changes to these types on particular architectures). * s390: sysdeps/s390/bits/flt-eval-method.h always defines __GLIBC_FLT_EVAL_METHOD to 1 to replicate the existing logic. As previously discussed, it might make sense in coordination with GCC to eliminate the historic mistake, avoid excess precision in the -fexcess-precision=standard case and make the typedefs match (with a NEWS entry, again). Tested for x86-64 and x86. Also did compilation-only testing with build-many-glibcs.py. * bits/flt-eval-method.h: New file. * sysdeps/m68k/m680x0/bits/flt-eval-method.h: Likewise. * sysdeps/s390/bits/flt-eval-method.h: Likewise. * sysdeps/x86/bits/flt-eval-method.h: Likewise. * math/Makefile (headers): Add bits/flt-eval-method.h. * math/math.h: Include <bits/flt-eval-method.h>. [__USE_ISOC99] (float_t): Define based on __GLIBC_FLT_EVAL_METHOD. [__USE_ISOC99] (double_t): Likewise. * bits/mathdef.h (float_t): Remove. (double_t): Likewise. * sysdeps/aarch64/bits/mathdef.h (float_t): Likewise. (double_t): Likewise. * sysdeps/alpha/bits/mathdef.h (float_t): Likewise. (double_t): Likewise. * sysdeps/arm/bits/mathdef.h (float_t): Likewise. (double_t): Likewise. * sysdeps/hppa/fpu/bits/mathdef.h (float_t): Likewise. (double_t): Likewise. * sysdeps/ia64/bits/mathdef.h (float_t): Likewise. (double_t): Likewise. * sysdeps/m68k/m680x0/bits/mathdef.h (float_t): Likewise. (double_t): Likewise. * sysdeps/mips/bits/mathdef.h (float_t): Likewise. (double_t): Likewise. * sysdeps/powerpc/bits/mathdef.h (float_t): Likewise. (double_t): Likewise. * sysdeps/s390/bits/mathdef.h (float_t): Likewise. (double_t): Likewise. * sysdeps/sh/sh4/bits/mathdef.h (float_t): Likewise. (double_t): Likewise. * sysdeps/sparc/bits/mathdef.h (float_t): Likewise. (double_t): Likewise. * sysdeps/tile/bits/mathdef.h (float_t): Likewise. (double_t): Likewise. * sysdeps/x86/bits/mathdef.h (float_t): Likewise. (double_t): Likewise. |