builtin memcmp() could be optimised
>Submitter-Id: net
>Originator: herbert@gondor.apana.org.au
>Organization: The Debian project
>Confidential: no
>Synopsis: builtin memcmp() could be optimised
>Severity: non-critical
>Priority: low
>Category: optimization
>Class: sw-bug
>Release: 3.0 (Debian GNU/Linux) and HEAD 20010701
>Environment:
System: Debian GNU/Linux (testing/unstable)
Architecture: i686
host: i386-linux
build: i386-linux
target: i386-linux
configured with: ../src/configure -v --enable-languages=c,c++,java,f77,proto,objc --prefix=/usr --infodir=/share/info --mandir=/share/man --enable-shared --with-gnu-as --with-gnu-ld --with-system-zlib --enable-long-long --enable-nls --without-included-gettext --disable-checking --enable-threads=posix --enable-java-gc=boehm --with-cpp-install-dir=bin --enable-objc-gc i386-linux
>Description:
[ Reported to the Debian BTS as report #85535.
Please CC 85535-quiet@bugs.debian.org on replies.
Log of report can be found at http://bugs.debian.org/85535 ]
int main() {
return !memcmp("abcd", "efgh", 4);
}
produces the assembly (-S -g -O2),
movl $.LC1, %edi
movl $.LC0, %esi
movl $4, %ecx
cld
repz
cmpsb
seta %cl
setb %dl
popl %esi
xorl %eax, %eax
popl %edi
cmpb %dl, %cl
sete %al
which is a waste because it could've simply loaded the two objects into
registers and compared them. The same is true on other architectures as
well.
>How-To-Repeat:
>Fix:
Reply to: