ac macros and ac releases

2006-02-03 Thread Thomas Porschberg
Hi, how can I find out when a certain autoconf macro was introduced by autoconf ? E.g. when I run autoconf with the --warnings=all option I get good advices how to improve my macros. However I guess when I upgrade the macros I'm indeed on top with the current autoconf release but possibly break so

Re: Checking for GCC 4

2006-02-03 Thread Ralf Corsepius
On Fri, 2006-02-03 at 06:56 +, Harlan Stenn wrote: > > On Fri, 2006-02-03 at 06:07 +, Harlan Stenn wrote: > > > Ralf, > > > > > > How about something that would use AC_RUN_IFELSE()? > > > > We are talking about distinguishing different compiler > > versions ! > > No, we are talki

Re: Checking for GCC 4

2006-02-03 Thread Ralf Corsepius
On Fri, 2006-02-03 at 18:06 +1100, Erik de Castro Lopo wrote: > Ralf Corsepius wrote: > > > Which compiler or language feature requires you to run code on a target? > > I have an M4 macro which checks the overflow behaviour of float > to int conversions. Are you running it on the target? Ralf

Re: Checking for GCC 4

2006-02-03 Thread Harlan Stenn
> > > Why on earth would you want to run something conditionally, depending on > > > which compilers it had been compiled with? > > I don't know, but that doesn't mean there isn't a good reason do to it. > My point is: There can't be the slightest reason. Why could there never be a reason? The

Re: Computed values in AC_INIT, AC_CONFIG_SRCDIR, AC_CONFIG_HEADERS, etc.

2006-02-03 Thread Ralf Wildenhues
* Harlan Stenn wrote on Thu, Feb 02, 2006 at 11:37:04PM CET: > > > I am always surprized, as well, to see so many warnings due to the > > > latest libtool, still using obsolete macros ! libtool 2.0 will do away with most of these warnings. > > I assume that you prefer that anyone using the latest

Re: ac macros and ac releases

2006-02-03 Thread Ralf Wildenhues
Hi Thomas, * Thomas Porschberg wrote on Fri, Feb 03, 2006 at 08:35:13AM CET: > > how can I find out when a certain autoconf macro was > introduced by autoconf ? Here's what I do to find out, in decreasing order of preference: Go to the top source dir of autoconf (either from a tarball, or CVS).

Re: Checking for GCC 4

2006-02-03 Thread Ralf Wildenhues
* E. Rosten wrote on Thu, Feb 02, 2006 at 02:52:14PM CET: > > How can I make sure user has at least gcc4 installed? > > (after establishing that the compiler is gcc), and doing AC_PROG_AWK: > > version=`$CXX -v 2>&1 | $AWK -F'[ .]' '/version/{print $3}'` > > should do the job. There are some rea

Re: ac macros and ac releases

2006-02-03 Thread Thomas Porschberg
Hi Ralf, thanks for your instructions. Just a thought: How about collect this information in a separate file and extend autoconf with an option --min-required-version which outputs the minimal required version for putting in AC_PREREQ ? Thomas On Fri, Feb 03, 2006 at 09:27:27AM +0100, Ralf Wild

Re: Checking for GCC 4

2006-02-03 Thread Erik de Castro Lopo
Ralf Corsepius wrote: > On Fri, 2006-02-03 at 18:06 +1100, Erik de Castro Lopo wrote: > > > I have an M4 macro which checks the overflow behaviour of float > > to int conversions. > > Are you running it on the target? Only if the host and target are the same. If they aren't it defaults to a less

Re: Checking for GCC 4

2006-02-03 Thread Ralf Corsepius
On Fri, 2006-02-03 at 07:51 +, Harlan Stenn wrote: > > > > Why on earth would you want to run something conditionally, depending on > > > > which compilers it had been compiled with? > > In case of cross-compilation, they are always user-specified input > > (--host=mytarget), and not automatic

Re: Checking for GCC 4

2006-02-03 Thread Harlan Stenn
> > And I'm not sure which is more fragile, user input or coded checks. > Best would be functional and safe automatic checks ;) Ideally, I agree with you. But in reality, the ideal pretty much never happens. > The worst are semi-thought out automatic checks silently failing and > producing bogu

Re: Checking for GCC 4

2006-02-03 Thread Ralf Corsepius
On Fri, 2006-02-03 at 09:56 +, Harlan Stenn wrote: > > > And I'm not sure which is more fragile, user input or coded checks. > I don't see a significant difference between something that (apparently) > works that is based on a version number check, an OS version check, or a > feature test. Gru

Re: ac macros and ac releases

2006-02-03 Thread Thomas Dickey
On Fri, 3 Feb 2006, Thomas Porschberg wrote: how can I find out when a certain autoconf macro was introduced by autoconf ? Since the autoconf maintainers choose not to provide any relevant information, your only recourse is to read the CVS for the corresponding files. E.g. when I run autoc

Re: ac macros and ac releases

2006-02-03 Thread Ralf Wildenhues
* Thomas Porschberg wrote on Fri, Feb 03, 2006 at 09:40:04AM CET: > > * Thomas Porschberg wrote on Fri, Feb 03, 2006 at 08:35:13AM CET: > > > > > > how can I find out when a certain autoconf macro was > > > introduced by autoconf ? > Just a thought: > How about collect this information in a separ

Re: ac macros and ac releases

2006-02-03 Thread Thomas Dickey
On Fri, 3 Feb 2006, Thomas Porschberg wrote: Hi Ralf, thanks for your instructions. Just a thought: How about collect this information in a separate file and extend autoconf with an option --min-required-version which outputs the minimal required version for putting in AC_PREREQ ? That would