Re: count2.asm
- From: "wolfgang kern" <nowhere@xxxxxxxxxxx>
- Date: Wed, 27 Jul 2005 18:51:08 +0200
"¬a\/b" asked:
| >[The compiler can never win. If the compiler produces faster code, we
| >use that. If the human produces faster code, we use that. The best the
| >compiler can do is break even.]
| Is it possible that one can do some C code that assembly people can
| not do faster? (>0.1%) (i think yes, because i have seen that code)
A few C-coders cheat their tool with ASM-inline and _intrisinc code,
here the timing may be almost identical (depends on programmers skills).
But in general, a standard written C-source compiled program will never
ever beat or become equal to 'well written' ASM-code.
Using ASM doesn't automatically mean someone knows how to program :)
We had this comparision already here, my code was 50 % shorter and
2 times faster until the HLL-coder starts with above mentioned tricks,
then the timing became more close.
| LF CR are 10 13?
Yes.
BS 08h
TAB 09h
LF 0Ah
CR 0Dh
BTAB 0Fh
other standard text-format controls exist (<20h),
but are usually not used in normal text files.
__
wolfgang
.
- References:
- count2.asm
- From: Frank Kotler
- Re: count2.asm
- From: ¬a\\/b
- Re: count2.asm
- From: Frank Kotler
- Re: count2.asm
- From: ¬a\\/b
- count2.asm
- Prev by Date: Re: HAY Betov, still no support for GPL or LINUX, WHY ????
- Next by Date: Re: HAY Betov, still no support for GPL or LINUX, WHY ????
- Previous by thread: Re: count2.asm
- Next by thread: Re: count2.asm
- Index(es):
Relevant Pages
|