This is the mail archive of the
mailing list for the GCC project.
Re: [PATCH] Adjust dg-extract-results.sh for lame awk implementations (PR testsuite/39807, take 2)
On Sun, 2009-04-26 at 22:53 +0200, Jakub Jelinek wrote:
> On Mon, Apr 20, 2009 at 07:53:54PM +0200, Jakub Jelinek wrote:
> > Apparently as reported in the PR some awk implementations can't handle more
> > than a few concurrently opened files (or is it extremely low system limit on
> > number of open files?).
> > I've bootstrapped/regtested this on x86_64-linux and i686-linux (where I
> > can't reproduce it though) and Andrew said in the PR that this fixes it on
> > powerpc-darwin.
> Apparently Solaris awk has a lot of other issues, this patch fixes other 2
> issues that Eric reported in bugzilla.
> Bootstrapped/regtested on x86_64-linux and Eric said he tested on Solaris.
> Ok for trunk/4.4.1/4.3?
> 2009-04-26 Jakub Jelinek <email@example.com>
> PR testsuite/39807
> * dg-extract-results.sh: Close open files and use >> instead of >
> to decrease number of concurrently open files from awk. Avoid
> = at the beginning of a regexp and redirect to a file determined
> by curfile variable rather than concatenated strings to workaround
> Solaris nawk bug.