This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: A truely ugly hack...
- To: bkorb at cruzio dot com
- Subject: Re: A truely ugly hack...
- From: Mark Mitchell <mark at codesourcery dot com>
- Date: Sat, 26 Aug 2000 09:40:16 -0700
- Cc: gcc at gcc dot gnu dot org, info-cvs at gnu dot org
- Organization: CodeSourcery, LLC
- References: <39A7F0E9.2AD8323E@cruzio.com>
>>>>> "Bruce" == Bruce Korb <bkorb@cruzio.com> writes:
Bruce> *that* special? That renders #3 (and #2) completely
Bruce> impossible. Unless you implement a Truely Ugly Hack.
I believe the standard hack is something like this, which we use on
x86-64.org in `loginfo':
^x86-64-ABI (/home/cvs/bin/x8664-update-abi.sh &) >> $CVSROOT/CVSROOT/commitlog
Then, that script looks like:
# Sleep a few seconds so that we're not doing this at the same time
# that the commit is occurring.
sleep 5
# Update the ABI document.
cd ~anoncvs/x86-64-ABI || exit 1
cvs -q update -d
make abi.pdf
You could replace the last bits with something that did more what you
want.
This script is perhaps not robust in general -- but it works OK in
practice. The key thing is that it runs in the background, which lets
the commit finish. You then do whatever other action you want to do
afterwards; either via the crude `sleep' hack, or via a proper loop of
some kind.
Of course, this still leaves a window where the new files are checked
in, but the generated files are not updated -- but that window is
pretty small.
--
Mark Mitchell mark@codesourcery.com
CodeSourcery, LLC http://www.codesourcery.com