SSE2 generation bug with 4.1.2 and -O3

to behave correctly. :-( Do you need anything else? Cheers, -- (°= =°) //\ Prakash Punnoor /\\ V_/ \_V pgpqqR78NXdXu.pgp Description: PGP signature

Re: SSE2 generation bug with 4.1.2 and -O3

Now filed under http://gcc.gnu.org/bugzilla/show_bug.cgi?id=31245 at David's request. -- (°= =°) //\ Prakash Punnoor /\\ V_/ \_V pgpmJBx7yF8Fe.pgp Description: PGP signature

Re: SSE2 generation bug with 4.1.2 and -O3

Am Samstag 17 März 2007 schrieb Prakash Punnoor: > Now filed under > > http://gcc.gnu.org/bugzilla/show_bug.cgi?id=31245 > > at David's request. I noticed another SSE2 related bug, filed under http://gcc.gnu.org/bugzilla/show_bug.cgi?id=31361 -- (°= =°)

gcc 4.0.x: MMX built-ins regression

return; } 3a1: c9 leave 3a2: c3 ret -- (°= =°) //\ Prakash Punnoor /\\ V_/ \_V signature.asc Description: OpenPGP digital signature

Re: gcc 4.0.x: MMX built-ins regression

h > 'gcc version 4.1.0 20050716 (experimental)' the code looks OK to me: Yes, looks OK to me, as well. -- (°= =°) //\ Prakash Punnoor /\\ V_/ \_V signature.asc Description: OpenPGP digital signature

Re: mmx register moves through memory

m), %mmxreg' I haven't read entirely through your mail, but I have reported above issue already and it seems to be fixed in cvs: http://gcc.gnu.org/bugzilla/show_bug.cgi?id=23630 Cheers, -- (°= =°) //\ Prakash Punnoor /\\ V_/ \_V signature.asc Description: OpenPGP digital signature

MMX built-ins performance oddities

LOOP_MIX; } __builtin_ia32_emms(); return; } I attached the objdumps: old.dump - using unions -> fast n3.dump - using vectors directly -> 10% slower on my athlon-xp, even when generated asm seems to be shorter BTW, the buffers were 16-byte aligned. -- Prakash Punno

Re: MMX built-ins performance oddities

Andrew Pinski schrieb: On Feb 19, 2005, at 8:21 AM, Prakash Punnoor wrote: Is this a known issue with gcc-3.4.3? I compiled the code using -O2 -march=athlon-xp -g3. If you want a smaller test case, I could try to do so. Right now I just didn't want to waste my time in case this is a know iss

Re: MMX built-ins performance oddities

Andrew Pinski schrieb: On Feb 19, 2005, at 8:21 AM, Prakash Punnoor wrote: Is this a known issue with gcc-3.4.3? I compiled the code using -O2 -march=athlon-xp -g3. If you want a smaller test case, I could try to do so. Right now I just didn't want to waste my time in case this is a know iss

Re: MMX built-ins performance oddities

stable state. :-) Sorry for my mistake. -- Prakash Punnoor formerly known as Prakash K. Cheemplavam signature.asc Description: OpenPGP digital signature

gcc 4.2.3 and MMX to mem move oddity

ntoo) x86_64 As you see in the intrinsic version gcc moves to mmx register to the stack, reloads from the stack and writes to the destination. Why? I don't know whether earlier gcc 4.2 versions produced such stupid code. Compiling as 32 does similar stupidity, though gcc reload

Re: gcc 4.2.3 and MMX to mem move oddity

jne60 9e: f3 c3 repz retq Why is movaps (SSE, floating point data) instead of movdqa (SSE2. integer data) used as store? Bug or feature? Even with -O0 compiled it is used. Regards, -- (°= =°) //\ Prakash Punnoor /\\ V_/ \_V signature.asc Description: This is a digitally signed message part.

Re: gcc 4.2.3 and MMX to mem move oddity

On the day of Saturday 23 February 2008 Prakash Punnoor hast written: > On the day of Saturday 23 February 2008 Uros Bizjak hast written: > > Hello! > > > > > f7: 0f 7f 5c 24 f0 movq %mm3,-0x10(%rsp) > > > fc: 0f 7f 54 24 f8 movq %mm2,