lethliel/glibc
1
0
Fork 0
forked from rpms/glibc
Code Issues Pull Requests Releases Activity

import glibc-2.28-168.el8

Browse Source
This commit is contained in:
CentOS Sources 2021-11-10 04:19:51 +00:00 committed by Stepan Oksanichenko
parent 6e9cfdd108
commit b6400f3e09
3 changed files with 590 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

306
SOURCES/glibc-rh1983203-1.patch Normal file
View File

@ -0,0 +1,306 @@
commit a55e2da2702e235fa0ae66a116d304d1bffc060a
Author: Lucas A. M. Magalhaes <lamm@linux.ibm.com>
Date: Thu May 6 17:01:52 2021 -0300
powerpc: Optimized memcmp for power10
This patch was based on the __memcmp_power8 and the recent
__strlen_power10.
Improvements from __memcmp_power8:
1. Don't need alignment code.
On POWER10 lxvp and lxvl do not generate alignment interrupts, so
they are safe for use on caching-inhibited memory. Notice that the
comparison on the main loop will wait for both VSR to be ready.
Therefore aligning one of the input address does not improve
performance. In order to align both registers a vperm is necessary
which add too much overhead.
2. Uses new POWER10 instructions
This code uses lxvp to decrease contention on load by loading 32 bytes
per instruction.
The vextractbm is used to have a smaller tail code for calculating the
return value.
3. Performance improvement
This version has around 35% better performance on average. I saw no
performance regressions for any length or alignment.
Thanks Matheus for helping me out with some details.
Co-authored-by: Matheus Castanho <msc@linux.ibm.com>
Reviewed-by: Raphael M Zinsly <rzinsly@linux.ibm.com>
diff --git a/sysdeps/powerpc/powerpc64/le/power10/memcmp.S b/sysdeps/powerpc/powerpc64/le/power10/memcmp.S
new file mode 100644
index 0000000000000000..52f244e7e77cbdf9
--- /dev/null
+++ b/sysdeps/powerpc/powerpc64/le/power10/memcmp.S
@@ -0,0 +1,179 @@
+/* Optimized memcmp implementation for POWER10.
+ Copyright (C) 2021 Free Software Foundation, Inc.
+ This file is part of the GNU C Library.
+
+ The GNU C Library is free software; you can redistribute it and/or
+ modify it under the terms of the GNU Lesser General Public
+ License as published by the Free Software Foundation; either
+ version 2.1 of the License, or (at your option) any later version.
+
+ The GNU C Library is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ Lesser General Public License for more details.
+
+ You should have received a copy of the GNU Lesser General Public
+ License along with the GNU C Library; if not, see
+ <https://www.gnu.org/licenses/>. */
+
+#include <sysdep.h>
+
+/* TODO: Replace macros by the actual instructions when minimum binutils becomes
+ >= 2.35. This is used to keep compatibility with older versions. */
+#define VEXTRACTBM(rt,vrb) \
+ .long(((4)<<(32-6)) \
+ | ((rt)<<(32-11)) \
+ | ((8)<<(32-16)) \
+ | ((vrb)<<(32-21)) \
+ | 1602)
+
+#define LXVP(xtp,dq,ra) \
+ .long(((6)<<(32-6)) \
+ | ((((xtp)-32)>>1)<<(32-10)) \
+ | ((1)<<(32-11)) \
+ | ((ra)<<(32-16)) \
+ | dq)
+
+/* Compare 32 bytes. */
+#define COMPARE_32(vr1,vr2,offset,tail_1,tail_2)\
+ LXVP(32+vr1,offset,r3); \
+ LXVP(32+vr2,offset,r4); \
+ vcmpneb. v5,vr1+1,vr2+1; \
+ bne cr6,L(tail_2); \
+ vcmpneb. v4,vr1,vr2; \
+ bne cr6,L(tail_1); \
+
+#define TAIL(v_res,s1,s2) \
+ vctzlsbb r7,v_res; \
+ vextubrx r8,r7,s1; \
+ vextubrx r9,r7,s2; \
+ subf r3,r9,r8; \
+ blr; \
+
+/* int [r3] memcmp (const char *s1 [r3], const char *s2 [r4],
+ size_t size [r5]) */
+
+#ifndef MEMCMP
+# define MEMCMP memcmp
+#endif
+ .machine power9
+ENTRY_TOCLESS (MEMCMP, 4)
+ CALL_MCOUNT 3
+
+ cmpldi cr6,r5,64
+ bgt cr6,L(loop_head)
+
+/* Compare 64 bytes. This section is used for lengths <= 64 and for the last
+ bytes for larger lengths. */
+L(last_compare):
+ li r8,16
+
+ sldi r9,r5,56
+ sldi r8,r8,56
+ addi r6,r3,16
+ addi r7,r4,16
+
+ /* Align up to 16 bytes. */
+ lxvl 32+v0,r3,r9
+ lxvl 32+v2,r4,r9
+
+ /* The sub. and vcmpneb. results are concatenated by the crnand in order
+ to do a single branch. It's doing a NOT(CR0.GT AND CR6.EQ) then
+ loading to CR0.LT. That means r9 is not bigger than 0 and v4 is not
+ all equal to 0. */
+ sub. r9,r9,r8
+ vcmpneb. v4,v0,v2
+ crnand 4*cr0+lt,4*cr0+gt,4*cr6+eq
+ bt 4*cr0+lt,L(tail1)
+
+ addi r3,r3,32
+ addi r4,r4,32
+
+ lxvl 32+v1,r6,r9
+ lxvl 32+v3,r7,r9
+ sub. r9,r9,r8
+ vcmpneb. v5,v1,v3
+ crnand 4*cr0+lt,4*cr0+gt,4*cr6+eq
+ bt 4*cr0+lt,L(tail2)
+
+ addi r6,r3,16
+ addi r7,r4,16
+
+ lxvl 32+v6,r3,r9
+ lxvl 32+v8,r4,r9
+ sub. r9,r9,r8
+ vcmpneb. v4,v6,v8
+ crnand 4*cr0+lt,4*cr0+gt,4*cr6+eq
+ bt 4*cr0+lt,L(tail3)
+
+ lxvl 32+v7,r6,r9
+ lxvl 32+v9,r7,r9
+ vcmpneb. v5,v7,v9
+ bne cr6,L(tail4)
+
+L(finish):
+ /* The contents are equal. */
+ li r3,0
+ blr
+
+L(loop_head):
+ /* Calculate how many loops to run. */
+ srdi. r8,r5,7
+ beq L(loop_tail)
+ mtctr r8
+
+/* Main loop. Compares 128 bytes each loop. */
+ .p2align 5
+L(loop_128):
+ COMPARE_32(v0,v2,0,tail1,tail2)
+ COMPARE_32(v6,v8,32,tail3,tail4)
+ COMPARE_32(v10,v12,64,tail5,tail6)
+ COMPARE_32(v14,v16,96,tail7,tail8)
+
+ addi r3,r3,128
+ addi r4,r4,128
+ bdnz L(loop_128)
+
+ /* Account loop comparisons. */
+ clrldi. r5,r5,57
+ beq L(finish)
+
+/* Compares 64 bytes if length is still bigger than 64 bytes. */
+ .p2align 5
+L(loop_tail):
+ cmpldi r5,64
+ ble L(last_compare)
+ COMPARE_32(v0,v2,0,tail1,tail2)
+ COMPARE_32(v6,v8,32,tail3,tail4)
+ addi r3,r3,64
+ addi r4,r4,64
+ subi r5,r5,64
+ b L(last_compare)
+
+L(tail1):
+ TAIL(v4,v0,v2)
+
+L(tail2):
+ TAIL(v5,v1,v3)
+
+L(tail3):
+ TAIL(v4,v6,v8)
+
+L(tail4):
+ TAIL(v5,v7,v9)
+
+L(tail5):
+ TAIL(v4,v10,v12)
+
+L(tail6):
+ TAIL(v5,v11,v13)
+
+L(tail7):
+ TAIL(v4,v14,v16)
+
+L(tail8):
+ TAIL(v5,v15,v17)
+
+END (MEMCMP)
+libc_hidden_builtin_def (memcmp)
+weak_alias (memcmp, bcmp)
diff --git a/sysdeps/powerpc/powerpc64/multiarch/Makefile b/sysdeps/powerpc/powerpc64/multiarch/Makefile
index ac2446aca62cc4ab..ee98417f4a383356 100644
--- a/sysdeps/powerpc/powerpc64/multiarch/Makefile
+++ b/sysdeps/powerpc/powerpc64/multiarch/Makefile
@@ -32,7 +32,7 @@ sysdep_routines += memcpy-power8-cached memcpy-power7 memcpy-a2 memcpy-power6 \
strncase-power8
ifneq (,$(filter %le,$(config-machine)))
-sysdep_routines += memcpy-power10 memmove-power10 memset-power10 \
+sysdep_routines += memcmp-power10 memcpy-power10 memmove-power10 memset-power10 \
rawmemchr-power9 rawmemchr-power10 \
strcmp-power9 strncmp-power9 strcpy-power9 stpcpy-power9 \
strlen-power9 strncpy-power9 stpncpy-power9 strlen-power10
diff --git a/sysdeps/powerpc/powerpc64/multiarch/ifunc-impl-list.c b/sysdeps/powerpc/powerpc64/multiarch/ifunc-impl-list.c
index 127af84b32a8196f..5213abdf87c79c88 100644
--- a/sysdeps/powerpc/powerpc64/multiarch/ifunc-impl-list.c
+++ b/sysdeps/powerpc/powerpc64/multiarch/ifunc-impl-list.c
@@ -184,6 +184,12 @@ __libc_ifunc_impl_list (const char *name, struct libc_ifunc_impl *array,
/* Support sysdeps/powerpc/powerpc64/multiarch/memcmp.c. */
IFUNC_IMPL (i, name, memcmp,
+#ifdef __LITTLE_ENDIAN__
+ IFUNC_IMPL_ADD (array, i, memcmp,
+ hwcap2 & PPC_FEATURE2_ARCH_3_1
+ && hwcap & PPC_FEATURE_HAS_VSX,
+ __memcmp_power10)
+#endif
IFUNC_IMPL_ADD (array, i, memcmp, hwcap2 & PPC_FEATURE2_ARCH_2_07,
__memcmp_power8)
IFUNC_IMPL_ADD (array, i, memcmp, hwcap & PPC_FEATURE_HAS_VSX,
diff --git a/sysdeps/powerpc/powerpc64/multiarch/memcmp-power10.S b/sysdeps/powerpc/powerpc64/multiarch/memcmp-power10.S
new file mode 100644
index 0000000000000000..73a0debd4a811d8e
--- /dev/null
+++ b/sysdeps/powerpc/powerpc64/multiarch/memcmp-power10.S
@@ -0,0 +1,26 @@
+/* Optimized memcmp implementation for POWER10.
+ Copyright (C) 2017-2021 Free Software Foundation, Inc.
+ This file is part of the GNU C Library.
+
+ The GNU C Library is free software; you can redistribute it and/or
+ modify it under the terms of the GNU Lesser General Public
+ License as published by the Free Software Foundation; either
+ version 2.1 of the License, or (at your option) any later version.
+
+ The GNU C Library is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ Lesser General Public License for more details.
+
+ You should have received a copy of the GNU Lesser General Public
+ License along with the GNU C Library; if not, see
+ <https://www.gnu.org/licenses/>. */
+
+#define MEMCMP __memcmp_power10
+
+#undef libc_hidden_builtin_def
+#define libc_hidden_builtin_def(name)
+#undef weak_alias
+#define weak_alias(name,alias)
+
+#include <sysdeps/powerpc/powerpc64/le/power10/memcmp.S>
diff --git a/sysdeps/powerpc/powerpc64/multiarch/memcmp.c b/sysdeps/powerpc/powerpc64/multiarch/memcmp.c
index 2c7a083a6560f920..0b8c0c1d8aa3f90a 100644
--- a/sysdeps/powerpc/powerpc64/multiarch/memcmp.c
+++ b/sysdeps/powerpc/powerpc64/multiarch/memcmp.c
@@ -27,11 +27,17 @@ extern __typeof (memcmp) __memcmp_ppc attribute_hidden;
extern __typeof (memcmp) __memcmp_power4 attribute_hidden;
extern __typeof (memcmp) __memcmp_power7 attribute_hidden;
extern __typeof (memcmp) __memcmp_power8 attribute_hidden;
+extern __typeof (memcmp) __memcmp_power10 attribute_hidden;
# undef memcmp
/* Avoid DWARF definition DIE on ifunc symbol so that GDB can handle
ifunc symbol properly. */
libc_ifunc_redirected (__redirect_memcmp, memcmp,
+#ifdef __LITTLE_ENDIAN__
+ (hwcap2 & PPC_FEATURE2_ARCH_3_1
+ && hwcap & PPC_FEATURE_HAS_VSX)
+ ? __memcmp_power10 :
+#endif
(hwcap2 & PPC_FEATURE2_ARCH_2_07)
? __memcmp_power8 :
(hwcap & PPC_FEATURE_HAS_VSX)

278
SOURCES/glibc-rh1983203-2.patch Normal file
View File

@ -0,0 +1,278 @@
commit 813c6ec808556553be9d39e900a3fc97ceb32330
Author: Pedro Franco de Carvalho <pedromfc@linux.ibm.com>
Date: Wed Jun 30 12:36:07 2021 -0300
powerpc: optimize strcpy/stpcpy for POWER9/10
This patch modifies the current POWER9 implementation of strcpy and
stpcpy to optimize it for POWER9/10.
Since no new POWER10 instructions are used, the original POWER9 strcpy is
modified instead of creating a new implementation for POWER10. This
implementation is based on both the original POWER9 implementation of
strcpy and the preamble of the new POWER10 implementation of strlen.
The changes also affect stpcpy, which uses the same implementation with
some additional code before returning.
On POWER9, averaging improvements across the benchmark
inputs (length/source alignment/destination alignment), for an
experiment that ran the benchmark five times, bench-strcpy showed an
improvement of 5.23%, and bench-stpcpy showed an improvement of 6.59%.
On POWER10, bench-strcpy showed 13.16%, and bench-stpcpy showed 13.59%.
The changes are:
1. Removed the null string optimization.
Although this results in a few extra cycles for the null string, in
combination with the second change, this resulted in improvements for
for other cases.
2. Adapted the preamble from strlen for POWER10.
This is the part of the function that handles up to the first 16 bytes
of the string.
3. Increased number of unrolled iterations in the main loop to 6.
Reviewed-by: Matheus Castanho <msc@linux.ibm.com>
Tested-by: Matheus Castanho <msc@linux.ibm.com>
diff --git a/sysdeps/powerpc/powerpc64/le/power9/strcpy.S b/sysdeps/powerpc/powerpc64/le/power9/strcpy.S
index ce8f50329177fd06..9845a1d4cf0e1e5d 100644
--- a/sysdeps/powerpc/powerpc64/le/power9/strcpy.S
+++ b/sysdeps/powerpc/powerpc64/le/power9/strcpy.S
@@ -45,91 +45,78 @@
The implementation can load bytes past a null terminator, but only
up to the next 16B boundary, so it never crosses a page. */
+/* Load quadword at addr+offset to vreg, check for null bytes,
+ and branch to label if any are found. */
+#define CHECK16(vreg,offset,addr,label) \
+ lxv vreg+32,offset(addr); \
+ vcmpequb. v6,vreg,v18; \
+ bne cr6,L(label);
+
.machine power9
ENTRY_TOCLESS (FUNC_NAME, 4)
CALL_MCOUNT 2
- /* NULL string optimisation */
- lbz r0,0(r4)
- stb r0,0(r3)
- cmpwi r0,0
- beqlr
-
- addi r4,r4,1
- addi r11,r3,1
-
vspltisb v18,0 /* Zeroes in v18 */
+ vspltisb v19,-1 /* 0xFF bytes in v19 */
- neg r5,r4
- rldicl r9,r5,0,60 /* How many bytes to get source 16B aligned? */
+ /* Next 16B-aligned address. Prepare address for L(loop). */
+ addi r5,r4,16
+ clrrdi r5,r5,4
+ subf r8,r4,r5
+ add r11,r3,r8
- /* Get source 16B aligned */
+ /* Align data and fill bytes not loaded with non matching char. */
lvx v0,0,r4
lvsr v1,0,r4
- vperm v0,v18,v0,v1
-
- vcmpequb v6,v0,v18 /* 0xff if byte is NULL, 0x00 otherwise */
- vctzlsbb r7,v6 /* Number of trailing zeroes */
- addi r8,r7,1 /* Add null terminator */
+ vperm v0,v19,v0,v1
- /* r8 = bytes including null
- r9 = bytes to get source 16B aligned
- if r8 > r9
- no null, copy r9 bytes
- else
- there is a null, copy r8 bytes and return. */
- cmpd r8,r9
- bgt L(no_null)
+ vcmpequb. v6,v0,v18 /* 0xff if byte is NULL, 0x00 otherwise */
+ beq cr6,L(no_null)
- sldi r10,r8,56 /* stxvl wants size in top 8 bits */
- stxvl 32+v0,r11,r10 /* Partial store */
+ /* There's a null byte. */
+ vctzlsbb r8,v6 /* Number of trailing zeroes */
+ addi r9,r8,1 /* Add null byte. */
+ sldi r10,r9,56 /* stxvl wants size in top 8 bits. */
+ stxvl 32+v0,r3,r10 /* Partial store */
#ifdef USE_AS_STPCPY
/* stpcpy returns the dest address plus the size not counting the
final '\0'. */
- add r3,r11,r7
+ add r3,r3,r8
#endif
blr
L(no_null):
- sldi r10,r9,56 /* stxvl wants size in top 8 bits */
- stxvl 32+v0,r11,r10 /* Partial store */
-
- add r4,r4,r9
- add r11,r11,r9
+ sldi r10,r8,56 /* stxvl wants size in top 8 bits */
+ stxvl 32+v0,r3,r10 /* Partial store */
+ .p2align 4
L(loop):
- lxv 32+v0,0(r4)
- vcmpequb. v6,v0,v18 /* Any zero bytes? */
- bne cr6,L(tail1)
-
- lxv 32+v1,16(r4)
- vcmpequb. v6,v1,v18 /* Any zero bytes? */
- bne cr6,L(tail2)
-
- lxv 32+v2,32(r4)
- vcmpequb. v6,v2,v18 /* Any zero bytes? */
- bne cr6,L(tail3)
-
- lxv 32+v3,48(r4)
- vcmpequb. v6,v3,v18 /* Any zero bytes? */
- bne cr6,L(tail4)
+ CHECK16(v0,0,r5,tail1)
+ CHECK16(v1,16,r5,tail2)
+ CHECK16(v2,32,r5,tail3)
+ CHECK16(v3,48,r5,tail4)
+ CHECK16(v4,64,r5,tail5)
+ CHECK16(v5,80,r5,tail6)
stxv 32+v0,0(r11)
stxv 32+v1,16(r11)
stxv 32+v2,32(r11)
stxv 32+v3,48(r11)
+ stxv 32+v4,64(r11)
+ stxv 32+v5,80(r11)
- addi r4,r4,64
- addi r11,r11,64
+ addi r5,r5,96
+ addi r11,r11,96
b L(loop)
+ .p2align 4
L(tail1):
- vctzlsbb r8,v6
- addi r9,r8,1
+ vctzlsbb r8,v6 /* Number of trailing zeroes */
+ addi r9,r8,1 /* Add null terminator */
sldi r9,r9,56 /* stxvl wants size in top 8 bits */
- stxvl 32+v0,r11,r9
+ stxvl 32+v0,r11,r9 /* Partial store */
#ifdef USE_AS_STPCPY
/* stpcpy returns the dest address plus the size not counting the
final '\0'. */
@@ -137,50 +124,81 @@ L(tail1):
#endif
blr
+ .p2align 4
L(tail2):
stxv 32+v0,0(r11)
- vctzlsbb r8,v6 /* Number of trailing zeroes */
- addi r9,r8,1 /* Add null terminator */
- sldi r10,r9,56 /* stxvl wants size in top 8 bits */
+ vctzlsbb r8,v6
+ addi r9,r8,1
+ sldi r9,r9,56
addi r11,r11,16
- stxvl 32+v1,r11,r10 /* Partial store */
+ stxvl 32+v1,r11,r9
#ifdef USE_AS_STPCPY
- /* stpcpy returns the dest address plus the size not counting the
- final '\0'. */
add r3,r11,r8
#endif
blr
+ .p2align 4
L(tail3):
stxv 32+v0,0(r11)
stxv 32+v1,16(r11)
- vctzlsbb r8,v6 /* Number of trailing zeroes */
- addi r9,r8,1 /* Add null terminator */
- sldi r10,r9,56 /* stxvl wants size in top 8 bits */
+ vctzlsbb r8,v6
+ addi r9,r8,1
+ sldi r9,r9,56
addi r11,r11,32
- stxvl 32+v2,r11,r10 /* Partial store */
+ stxvl 32+v2,r11,r9
#ifdef USE_AS_STPCPY
- /* stpcpy returns the dest address plus the size not counting the
- final '\0'. */
add r3,r11,r8
#endif
blr
+ .p2align 4
L(tail4):
stxv 32+v0,0(r11)
stxv 32+v1,16(r11)
stxv 32+v2,32(r11)
- vctzlsbb r8,v6 /* Number of trailing zeroes */
- addi r9,r8,1 /* Add null terminator */
- sldi r10,r9,56 /* stxvl wants size in top 8 bits */
+ vctzlsbb r8,v6
+ addi r9,r8,1
+ sldi r9,r9,56
addi r11,r11,48
- stxvl 32+v3,r11,r10 /* Partial store */
+ stxvl 32+v3,r11,r9
#ifdef USE_AS_STPCPY
- /* stpcpy returns the dest address plus the size not counting the
- final '\0'. */
add r3,r11,r8
#endif
blr
+
+ .p2align 4
+L(tail5):
+ stxv 32+v0,0(r11)
+ stxv 32+v1,16(r11)
+ stxv 32+v2,32(r11)
+ stxv 32+v3,48(r11)
+ vctzlsbb r8,v6
+ addi r9,r8,1
+ sldi r9,r9,56
+ addi r11,r11,64
+ stxvl 32+v4,r11,r9
+#ifdef USE_AS_STPCPY
+ add r3,r11,r8
+#endif
+ blr
+
+ .p2align 4
+L(tail6):
+ stxv 32+v0,0(r11)
+ stxv 32+v1,16(r11)
+ stxv 32+v2,32(r11)
+ stxv 32+v3,48(r11)
+ stxv 32+v4,64(r11)
+ vctzlsbb r8,v6
+ addi r9,r8,1
+ sldi r9,r9,56
+ addi r11,r11,80
+ stxvl 32+v5,r11,r9
+#ifdef USE_AS_STPCPY
+ add r3,r11,r8
+#endif
+ blr
+
END (FUNC_NAME)
#ifndef USE_AS_STPCPY
libc_hidden_builtin_def (strcpy)

7
SPECS/glibc.spec
View File

@ -1,6 +1,6 @@
%define glibcsrcdir glibc-2.28
%define glibcversion 2.28
%define glibcrelease 167%{?dist}
%define glibcrelease 168%{?dist}
# Pre-release tarballs are pulled in from git using a command that is
# effectively:
#
@ -776,6 +776,8 @@ Patch598: glibc-rh1971664-13.patch
Patch599: glibc-rh1971664-14.patch
Patch600: glibc-rh1971664-15.patch
Patch601: glibc-rh1977614.patch
Patch602: glibc-rh1983203-1.patch
Patch603: glibc-rh1983203-2.patch
##############################################################################
# Continued list of core "glibc" package information:
@ -2726,6 +2728,9 @@ fi
%files -f compat-libpthread-nonshared.filelist -n compat-libpthread-nonshared
%changelog
* Fri Oct 29 2021 Arjun Shankar <arjun@redhat.com> - 2.28-168
- Optimize memcmp, strcpy, and stpcpy for IBM POWER10 (#1983203)
* Wed Oct 13 2021 Arjun Shankar <arjun@redhat.com> - 2.28-167
- malloc: Initiate tcache shutdown even without allocations (#1977614)