Re: Smoke [5.11.0] Intel(R) FAIL(F) MSWin32 WinXP/.Net SP3 (x86/4 cpu)
Abe Timmerman
> Hello Abe,
Hi Max,
>>> Locally applied patches:
>>> SMOKE?????
>>
>> Your reports seem to miss the rather important information about
>> commit id. This seems to point at the fact that you do not have the
>> '.patch' file.
>>
>> Can you tell me how you sync the source-tree so we can find a way to
>> fake the .patch file.
> Ooops - yes, that was the one issue I wanted to remember before starting
> to send reports outside :(
>
> I'm copying a synced git tree with the following commands. The git repo
> doesn't have the .patch and hence, things get interesting. I guess we
> could (ab)use the code in make_patchnum.pl to find out the git
> patchnumber for the current branch.
...
> I haven't yet looked at what .patch should contain. I guess running
> make_patchnum.pl with an installed perl instead of the just-built
> miniperl should simply work, because all make_patchnum.pl does is run
> git several times to extract the current state.
Nope, make_patchnum.pl provides de full sha1 for patchlevel.h
.patch is generated as part of the hole post-commit stuff for the rsync-archives.
you can either rsync it from rsync://perl5.git.perl.org/perl-current/.patch or ftp it with wget from
ftp://public.activestate.com/pub/apc/perl-current/.patch
The .patch file contains 4 fields (branch, timestamp, sha1, git-describe). My choice to use the timestamp field, to identify the
commit in the subject of the reports, might not have been the smartest as the timestamp is generated by the post-commit handler and
not reproducible from the git repository (it's not the same stamp as the commit message).
I think I will be changing this, so we can reproduce these fields from the git repository.
thanks again for your interest in perl-core smoking +
good luck,
Abe.
--
_ | "Pinky, are you pondering what I'm pondering?"
( ) | I think so, Brain, but if the plural of mouse is mice, wouldn't the
X | plural of spouse be spice? (27 Sep 97 - Brain Acres)
/ \ |