src/lib/libc/amd64/string/memcmp.S, branch releng/13.1 FreeBSD source tree amd64: bring back asm bcmp, shared with memcmp 2022-03-29T12:40:32+00:00 Mateusz Guzik mjg@FreeBSD.org 2022-03-25T15:04:04+00:00 1544cd332b9635a4a3c1821d895c3ad40bc630f9 Turns out clang converts "memcmp(foo, bar, len) == 0" and similar to bcmp calls. Reviewed by: emaste (previous version), jhb (previous version) Differential Revision: https://reviews.freebsd.org/D34673 Approved by: re (gjb) (cherry picked from commit fbc002cb72d2d9bb435cce99630d5d7da9f59390) (cherry picked from commit c5890784f6d315c501586b39cf8c0c80b609be5b) 
Turns out clang converts "memcmp(foo, bar, len) == 0" and similar to
bcmp calls.

Reviewed by:	emaste (previous version), jhb (previous version)
Differential Revision:	https://reviews.freebsd.org/D34673
Approved by:	re (gjb)

(cherry picked from commit fbc002cb72d2d9bb435cce99630d5d7da9f59390)
(cherry picked from commit c5890784f6d315c501586b39cf8c0c80b609be5b)
amd64: move memcmp checks upfront 2021-02-04T17:59:10+00:00 Mateusz Guzik mjg@FreeBSD.org 2021-01-31T15:46:18+00:00 88a580ebeea10e7bf9eb823dbbf48349eb8da540 This is a tradeoff which saves jumps for smaller sizes while making the 8-16 range slower (roughly in line with the other cases). Tested with glibc test suite. For example size 3 (most common with vfs namecache) (ops/s): before: 407086026 after: 461391995 The regressed range of 8-16 (with 8 as example): before: 540850489 after: 461671032 (cherry picked from commit f1be262ec11c1c35e6485f432415b5b52adb505d) 
This is a tradeoff which saves jumps for smaller sizes while making
the 8-16 range slower (roughly in line with the other cases).

Tested with glibc test suite.

For example size 3 (most common with vfs namecache) (ops/s):
before:	407086026
after:	461391995

The regressed range of 8-16 (with 8 as example):
before:	540850489
after:	461671032

(cherry picked from commit f1be262ec11c1c35e6485f432415b5b52adb505d)
amd64: add a note about simd to libc memset, memmove and memcmp 2021-02-04T17:59:05+00:00 Mateusz Guzik mjg@FreeBSD.org 2021-01-31T15:50:34+00:00 068f2402d28bf2ddee884c83be1dff3a7631569b (cherry picked from commit 0db6aef407f30c138982b8cde43189aad098b337) 
(cherry picked from commit 0db6aef407f30c138982b8cde43189aad098b337)
amd64: sync up libc memcmp with the kernel version (r357309) 2020-01-30T19:57:05+00:00 Mateusz Guzik mjg@FreeBSD.org 2020-01-30T19:57:05+00:00 8291e8874845b258acc349e9ac9c37d0b14046d3 






amd64: sync up libc memcmp with the kernel version (r357208)
2020-01-29T01:57:07+00:00

Mateusz Guzik
mjg@FreeBSD.org

2020-01-29T01:57:07+00:00

4846152a081cb3f0643287dab84c5e2778d9b54e












amd64: reimplement libc memcmp and bcmp with kernel memcmp
2018-09-27T17:08:29+00:00

Mateusz Guzik
mjg@FreeBSD.org

2018-09-27T17:08:29+00:00

5bbde333cdbd9054fd9b4257a2a3e51d4a819ab4

Both are significantly slower than hand-coded loops. See r338963 for
kernel commit.

bcmp differs from memcmp by always returning 1 when a difference is
found, as opposed to going for a value bigger or lower than 0
depending on what it is. This means it can do less work. For now the
code is duplicated and modified. This will get deduplicated after
another round of optimization when memcmp will get a longer-term form.

Both tested with the glibc suite. While the suite does not have a test
for bcmp, I created a wrapper routine which verified that values match
(0 vs 0, 1 vs non-zero).

Reviewed by:	kib
Approved by:	re (gjb)
Sponsored by:	The FreeBSD Foundation
Differential Revision:	https://reviews.freebsd.org/D17336





Both are significantly slower than hand-coded loops. See r338963 for
kernel commit.

bcmp differs from memcmp by always returning 1 when a difference is
found, as opposed to going for a value bigger or lower than 0
depending on what it is. This means it can do less work. For now the
code is duplicated and modified. This will get deduplicated after
another round of optimization when memcmp will get a longer-term form.

Both tested with the glibc suite. While the suite does not have a test
for bcmp, I created a wrapper routine which verified that values match
(0 vs 0, 1 vs non-zero).

Reviewed by:	kib
Approved by:	re (gjb)
Sponsored by:	The FreeBSD Foundation
Differential Revision:	https://reviews.freebsd.org/D17336







Add section .note.GNU-stack for assembly files used by 386 and amd64.
2011-01-07T16:08:40+00:00

Konstantin Belousov
kib@FreeBSD.org

2011-01-07T16:08:40+00:00

93ab75867017bed8892f9f3b1e1bbd6120d49fcd












We've been lax about matching END() macros in asm code for some time.  This
2008-11-02T01:10:54+00:00

Peter Wemm
peter@FreeBSD.org

2008-11-02T01:10:54+00:00

5d053f461caeb73f6de165aa1e07b2003101605c

is used to set the ELF size attribute for functions.  It isn't normally
critical but some things can make use of it (gdb for stack traces).
Valgrind needs it so I'm adding it in.  The problem is present on all
branches and on both i386 and amd64.





is used to set the ELF size attribute for functions.  It isn't normally
critical but some things can make use of it (gdb for stack traces).
Valgrind needs it so I'm adding it in.  The problem is present on all
branches and on both i386 and amd64.







Add machine-specific, optimized implementations of bcmp and memcmp.
2005-04-08T05:15:55+00:00

Alan Cox
alc@FreeBSD.org

2005-04-08T05:15:55+00:00

26f6218be92649295dbb54aa1a18fbd9110631fd

PR: 73111
Submitted by: Ville-Pertti Keinonen <will@iki.fi> (taken from NetBSD)
MFC after: 3 weeks





PR: 73111
Submitted by: Ville-Pertti Keinonen <will@iki.fi> (taken from NetBSD)
MFC after: 3 weeks