Oh, update_version_svn can be apply to
trunk/gcc-4.5-branch/gcc-4.4-branch/gcc-4.3-branch, not only trunk.

On Mon, Jan 31, 2011 at 10:45, Dongsheng Song <dongsheng.s...@gmail.com> wrote:
> It's very simple (only for trunk, although it maybe more useful for branches):
>
> Index: update_version_svn
> ===================================================================
> --- update_version_svn  (revision 169428)
> +++ update_version_svn  (working copy)
> @@ -42,6 +42,12 @@
>     SVNROOT2=${SVNROOT}/branches/${BRANCH}
>   fi
>
> +  LAST_COMMITER=`${SVN} log -q -l 1 ${SVNROOT2} | awk '{if (NR==2)
> {print $3; exit}}'`
> +  if test "${LAST_COMMITER}" = "gccadmin"; then
> +    echo "The last commiter is gccadmin, bump DATASTAMP skipped."
> +    continue
> +  fi
> +
>   for i in $datestamp_FILES; do
>     ${SVN} -q co -N ${SVNROOT2}/`dirname $i` `basename $i`
>   done
>
> On Mon, Jan 31, 2011 at 06:23, Gerald Pfeifer <ger...@pfeifer.com> wrote:
>>
>> On Sat, 29 Jan 2011, Dongsheng Song wrote:
>> > Just for curiousness, why we bump the DATESTAMP when the last commit
>> > is DATESTAMP changes on the branch ?
>>
>> As far as I am concerned, that's a bug (or a missing feature).  The
>> script in question is maintainer-scripts/update_version_svn in the GCC
>> source repository.  I already made some simplifications today and am
>> planning to do a bit more tonight.  Are you interested in improving
>> the script to address this issue?
>>
>> Gerald
>

Reply via email to