Re: feature request: detect CPU instruction set

Sam, > > AH_TOP([/* CPU and C ABI indicator */ > > #ifndef __i386__ > > #undef __i386__ > > #endif > > ... > > #ifndef __sparc64__ > > #undef __sparc64__ > > #endif > > ]) > > what is this? It's a set of preprocessor commands that get copied into config.h.in. So that config.h will define the a

Re: feature request: detect CPU instruction set

Bruno Haible wrote: AH_TOP([/* CPU and C ABI indicator */ #ifndef __i386__ #undef __i386__ #endif #ifndef __x86_64__ #undef __x86_64__ #endif #ifndef __alpha__ #undef __alpha__ #endif #ifndef __arm__ #undef __arm__ #endif #ifndef __armel__ #undef __armel__ #endif #ifndef __hppa__ #undef __hppa_

Re: feature request: detect CPU instruction set

Sam Steingold wrote: > > So, except for code style, you are fine with that proposed module? > yes. OK. In the absence of other comments, I committed and pushed the new module. Bruno

Re: feature request: detect CPU instruction set

Hi Bruno, On 8/31/10, Bruno Haible wrote: > > It apparently depends for whom the code should be readable. since you will be maintaining the code, it is your choice. > So, except for code style, you are fine with that proposed module? yes. -- Sam Steingold

Re: feature request: detect CPU instruction set

Hi Sam, > you say you want to cache just once. > fine, remove AC_CACHE_CHECK. Yes, I removed the AC_CACHE_CHECK invocations and then found that there was no good reason any more to keep the CL_CACHE_EGREP_CPP and CL_SET_CPU_ABI macros. > I have > dnl CL_CACHE_EGREP_CPP(what,variable,condition) >

Re: feature request: detect CPU instruction set

Bruno Haible wrote: Compared to the code that you proposed: - I changed the name, of course. - I chose to use AC_CACHE_CHECK once for the result of the entire macro, not for each AC_EGREP_CPP inside it. - I dropped Convex CPUs; these are dead for more than 10 years. - I added an AH_TO

Re: feature request: detect CPU instruction set

Hi Sam, > PS. I do not see the new module under git. It is not yet committed, because I'm waiting for comments and reviews from you and others. > I think I have already contributed and thus you should already have > all the papers. I'm following up to this in private mail. Bruno

Re: feature request: detect CPU instruction set

Hi Bruno, On 8/28/10, Bruno Haible wrote: > Sam, you contributed a significant part to the > FFCALL_CANONICAL_HOST_CPU code that you proposed, from which this one here > is derived. Would you mind starting the paperwork process for assigning > your gnulib contributions to the FSF? Please refe

Re: feature request: detect CPU instruction set

Hi Sam, all, Replying to : Ralf Wildenhues wrote: > > The already existing AC_CANONICAL_HOST is insufficient in that it > > does not detect the actual cpu instruction set to which the C > > compiler compiles. > > Is there someth

Re: feature request: detect CPU instruction set

so, what's the verdict? Sam Steingold wrote: The already existing AC_CANONICAL_HOST is insufficient in that it does not detect the actual cpu instruction set to which the C compiler compiles. E.g., it is quite common that "uname -m" prints sparc64 while gcc compile 32-bit binaries by default. I

Re: feature request: detect CPU instruction set

Hello Ralf, On 7/21/10, Ralf Wildenhues wrote: > * Sam Steingold wrote on Wed, Jul 21, 2010 at 05:08:02PM CEST: > > The already existing AC_CANONICAL_HOST is insufficient in that it > > does not detect the actual cpu instruction set to which the C > > compiler compiles. > > Is there something

Re: feature request: detect CPU instruction set

Hello Sam, * Sam Steingold wrote on Wed, Jul 21, 2010 at 05:08:02PM CEST: > The already existing AC_CANONICAL_HOST is insufficient in that it > does not detect the actual cpu instruction set to which the C > compiler compiles. Is there something in GCC and/or binutils to bite off of? I mean, idea

feature request: detect CPU instruction set

The already existing AC_CANONICAL_HOST is insufficient in that it does not detect the actual cpu instruction set to which the C compiler compiles. E.g., it is quite common that "uname -m" prints sparc64 while gcc compile 32-bit binaries by default. It would be nice if gnulib offered a module cont