|x86 code generation firstname.lastname@example.org (Nick Shaffner) (2000-06-14)|
|Re: x86 code generation email@example.com (2000-06-20)|
|Re: x86 code generation firstname.lastname@example.org (Juergen Kahrs) (2000-06-20)|
|Re: x86 code generation email@example.com (2000-06-20)|
|From:||"Nick Shaffner" <firstname.lastname@example.org>|
|Date:||14 Jun 2000 12:47:08 -0400|
|Organization:||Concentric Internet Services|
|Keywords:||386, code, question|
Anyone know where I can get some good information on code
generation stratagies for x86 (thinking Pentium+) specific processors?
It seems that nearly all the compiler texts and papers I've read
seem only to dedicate a few pages to dealing with the x86's major
idosyncracies, and they usually just show you a couple of ways to hack
'x86' support into a general purpose back end.
I am interested x86 specific code generation, including
intermediete representations that are friendly to the architecture,
strategies for managing the floating point stack, dealing with the
lack of registers - but making use of register renaming, etc.
Could anyone offer any guidance as to where I could find such
[I've seen thick manuals from Intel full of advice on x86 code
generation. You can probably find PDFs on their developer web
Return to the
Search the comp.compilers archives again.