Re: [patch] remove config.status race condition

On Fri, Feb 23, 2001 at 02:54:06PM -0500, Pavel Roskin wrote: > > > Are you sure that $$ is always preserved across subshells? What it I > > > create conf$$ in a subshell but want to access it in the top level shell? > > > > Sorry, but I don't understand why you are referring to subshells here. >

AC_CONFIGURE_HEADER from subdirectory

Hi, I'm trying to set up a project that has most of the configuration related files (including configure.ac and acconfig.h) in $SRCDIR/config, but it's causing me some grief, so I'm hoping someone can help me. First off, if I run 'autoconf' from $PROJECT I have to tell it where to find 'configu

Re: 2.13 issue with AC_PROG_CXX

On Feb 22, 2001, Peter Eisentraut <[EMAIL PROTECTED]> wrote: > Alexandre Oliva writes: >> I wish all of AC_PROG_CC, AC_PROG_CXX, etc, would have variants or >> extra arguments that would cause them to be non-fatal in case a >> compiler isn't found. > How about AC_PROG_CC([list], [ACTION-IF-NO]),

Re: [patch] remove config.status race condition

> > Are you sure that $$ is always preserved across subshells? What it I > > create conf$$ in a subshell but want to access it in the top level shell? > > Sorry, but I don't understand why you are referring to subshells here. > If anyway we were actually playing with subshells, we'd be in trouble

RE: [patch] remove config.status race condition

> First of all, it is not obvious at all that config.status should run > this, since Tim still did not convince me that AS_LN_S was a good > thing. As is, AS_LN_S is just *useless*. We were happier before. Hmm: CONFIG_LINKS code BEFORE: ln -s foo bar || ln foo bar fine, but breaks DJGPP, whe

Re: [patch] remove config.status race condition

Pavel Roskin <[EMAIL PROTECTED]> writes: > Hello, Akim! > > > Second means consists in just using $$. We have room for this: > > s/conftest/conf$$/, I really can't understand why we'd need something > > more complex. > > Are you sure that $$ is always preserved across subshells? What it I > cr

Re: [patch] remove config.status race condition

Hello, Akim! > Second means consists in just using $$. We have room for this: > s/conftest/conf$$/, I really can't understand why we'd need something > more complex. Are you sure that $$ is always preserved across subshells? What it I create conf$$ in a subshell but want to access it in the top

Re: [patch] remove config.status race condition

"Lars J. Aas" <[EMAIL PROTECTED]> writes: > Hey, I sent is a patch that fixes a *bug*! Why do I have to justify > myself to you on this one? Agreed. Nonetheless I don't agree with patch :) First of all, it is not obvious at all that config.status should run this, since Tim still did not convi

Re: [patch] don't discard previous config.log before necessary

I agree with Lars we want to keep the log parts, but my suggestion was to just move the setting up of config.log *after* arguments have been parsed. AFAIR, it is possible to do so and keeping AC_MSG_ERROR uses because *now* there is some magic in AC_MSG_ERROR (well the AS stuff). I dislike the

Re: suppressing the newline from echo

Phil Edwards <[EMAIL PROTECTED]> writes: > I've seen logic in generated configure files to determine how to keep echo > from suppressing the newline (-n, \c, etc). Some ac_ variables get set > and then used in echo invocations. > > Is this available in a macro? I don't see anything in the 2.13

Re: autoconf 2.49c fails if '.' is in PATH

Well, it's all too messy, and because of this show stopper we are delaying Autoconf 2.50 way too much. My personal opinion is back up to test -f and period. Sorry for the exotic guys, but no answer seems to have arose, and yet trying to address this issue screwed the last snapshot.

Re: Autotest

"Derek R. Price" <[EMAIL PROTECTED]> writes: > What's the reason autotest uses a while/shift loop is used to read > arguments? 'getopts' is specified by XPG4, Posix.2, and SUS2. Is that > not portable enough? Nope. At least it is very suspicious. Prove it's portable, then we'll use it :) I

Autotest

What's the reason autotest uses a while/shift loop is used to read arguments? 'getopts' is specified by XPG4, Posix.2, and SUS2. Is that not portable enough? Derek -- Derek Price CVS Solutions Architect ( http://CVSHome.org ) mailto:[EMAIL PROTECTED] OpenAvenue ( http:

Re: CVS Autoconf broken - no acceptable cc found

Pavel Roskin <[EMAIL PROTECTED]> wrote: | Hello! | | Autoconf is badly broken today. Don't use it. It cannot find any compilers | on my RedHat 6.2. Most tests fail. Running "sh -x configure" shows e.g. | this: Argh. What was I thinking? I've just reverted that. Sorry. Next time, I'll run `make

CVS Autoconf broken - no acceptable cc found

Hello! Autoconf is badly broken today. Don't use it. It cannot find any compilers on my RedHat 6.2. Most tests fail. Running "sh -x configure" shows e.g. this: + test -x +/home/proski/bin:/usr/local/bin:/usr/kerberos/bin:/bin:/usr/bin:/usr/X11R6/bin:/usr/sbin:/usr/local/sbin:/sbin/cc + continue