Sigh, Wrong mail... Resending...
Luiz Fernando Capitulino wrote:
Hi Thomas,
On Sat, 11 Mar 2006 15:01:01 +0200
Thomas Backlund <[EMAIL PROTECTED]> wrote:
| o Thomas Backlund <[EMAIL PROTECTED]>
| - Patch ZZA1: fix missing BITS_PER_BYTE definition in x86_64 /
NUMA
| introduced in CVE-2006-0557
Ok, I've moved the patch into another series, because ZZXX is really for
security issues. But the final result is exactly the same.
Yeah, forgot about that...
I didn't add the changelog that way, either, but something like:
o Luiz Capitulino <[EMAIL PROTECTED]>
- Define BITS_PER_BYTE (patch from Thomas Backlund <[EMAIL PROTECTED]>)
instead. The reason is that (IMO) the 'o' line is for the person who
added the commit (and that way is the responsible for it, and who should
be blamed when things goes wrong).
Fine by me ;-)
Also, would be good if we might use the same directory organization in
the servers. I'm using RPM's build default organization, does it bother
you? :)
Well If I go by RPM it will be:to change:
/Kernels/2006.0/x86_64/
to
/Kernels/2006.0/RPMS/i586/
/Kernels/2006.0/RPMS/x86_64/
/Kernels/2006.0/SRPMS/
Or did you mean like you, a separate directory for every version? Like:
/Kernels/2006.0/x86_64/2.6.12.18.uc1mdk/RPMS/
/Kernels/2006.0/x86_64/2.6.12.18.uc1mdk/SRPMS/
/Kernels/2006.0/x86_64/2.6.12.18.uc2mdk/RPMS/
/Kernels/2006.0/x86_64/2.6.12.18.uc2mdk/SRPMS/
/Kernels/2006.0/x86_64/2.6.12.18.uc3mdk/RPMS/
/Kernels/2006.0/x86_64/2.6.12.18.uc3mdk/SRPMS/
/Kernels/2006.0/x86_64/2.6.12.18.uc4mdk/RPMS/
/Kernels/2006.0/x86_64/2.6.12.18.uc4mdk/SRPMS/
/Kernels/2006.0/i586/2.6.12.18.uc1mdk/RPMS/
/Kernels/2006.0/i586/2.6.12.18.uc1mdk/SRPMS/
/Kernels/2006.0/i586/2.6.12.18.uc2mdk/RPMS/
/Kernels/2006.0/i586/2.6.12.18.uc2mdk/SRPMS/
/Kernels/2006.0/i586/2.6.12.18.uc3mdk/RPMS/
/Kernels/2006.0/i586/2.6.12.18.uc3mdk/SRPMS/
/Kernels/2006.0/i586/2.6.12.18.uc4mdk/RPMS/
/Kernels/2006.0/i586/2.6.12.18.uc4mdk/SRPMS/
It will be a lot of subdirectories, since I keep every version to be
able to track regressions...
and it wont help me making it an "urpmi"-able source...
--
Regards
Thomas
