aboutsummaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* check for all mandatory archs if there is at least one arched packageOlivier Blin2010-08-251-2/+3
| | | | | | (and not only if there is no noarch package, since now rpm can build both noarch and arched from the same src package)
* do not guess built modules from media with different arch (i.e. do not use ↵Olivier Blin2010-05-261-0/+1
| | | | i586 media from x86_64 media.cfg)
* use release sub-media by default (has been the default repository layout for ↵Olivier Blin2010-05-191-2/+2
| | | | years now)
* allow armv5tejl to build armv5tl packagesOlivier Blin2010-05-191-0/+1
|
* Missing buildrequires is for the maintainer tooPascal Terjan2010-01-191-1/+1
|
* Use maintainers.mandriva.com instead of rpmmon to get maintainerPascal Terjan2010-01-191-1/+1
|
* - Fix --log optionPascal Terjan2010-01-181-3/+4
|
* Fix missing }Pascal Terjan2010-01-051-13/+15
|
* Fail when deps would not be installed for a src.rpmPascal Terjan2010-01-051-2/+3
|
* Differentiate missing src.rpm from failure to rebuild itPascal Terjan2010-01-041-0/+8
|
* Don't break the case when new and old src.rpm have same namePascal Terjan2009-12-301-1/+1
|
* Try harder to decide which one is re-generated rpmPascal Terjan2009-12-301-8/+5
|
* Commit the working version of build_allPascal Terjan2009-12-071-1/+1
|
* Allow rebuilding all the mediaPascal Terjan2009-12-021-4/+8
|
* perl_checker cleanupsThierry Vignaud2009-12-024-30/+30
|
* supplement blacklistThierry Vignaud2009-12-021-0/+2
|
* Add a --delete-on-success for the cases where we only want failure logsPascal Terjan2009-12-011-2/+10
|
* Oops they were differentPascal Terjan2009-12-011-0/+3
|
* Remove duplicate linesPascal Terjan2009-12-011-3/+0
|
* Don't lie, use the same plog level for the group of messages, so they always ↵Gustavo De Nardin2009-08-071-2/+2
| | | | show up complete.
* - don't output control code for back-to-normal color when not using colorsGustavo De Nardin2009-08-061-5/+4
| | | | | - add an 'ERROR' plog level, as many places use that instead of 'ERR'
* Workaround failure to create the log file.Gustavo De Nardin2009-08-061-0/+2
|
* - support passing the place of iurt logs to the command, so the iurt2.shGustavo De Nardin2009-08-061-1/+2
| | | | | | | | wrapper can save iurt output there - revert a change from r258923: don't try using tee in the SSH command, the single quotes break the command line, probably due to them already being used somewhere else
* Adding --iurtlogdir option, to be told where to save output from iurt.Gustavo De Nardin2009-08-061-3/+7
|
* Adding iurt2 wrapper, which outputs the PID=.. line as required by ulri, as ↵Gustavo De Nardin2009-08-051-0/+6
| | | | it exists in the cluster currently.
* Small change for preparing get_pid_from_file() to work on a complete Iurt ↵Gustavo De Nardin2009-08-051-1/+1
| | | | output log.
* Save the remote bot command output to botcmd.log in the remote log dir. SoGustavo De Nardin2009-08-041-2/+3
| | | | | it will be fetched with the rest of the logs in case of build failure.
* do not test presence of /usr/lib/rpm/rpmb since rpm 4.6.0 do not have it anymorePascal Rigaux2008-12-101-1/+1
|
* do not try to uselessly detect latest modulesOlivier Blin2008-08-211-1/+1
|
* use the defined $sudo instead of whole path to sudoGustavo De Nardin2008-07-101-1/+1
|
* remove hack for urpmi 6.0 (now unneded since 6.1 is out and nodes have been ↵Pascal Rigaux2008-07-101-1/+0
| | | | updated)
* temporary hack to workaround bug in urpmi 6.0: create tmp_chroot/var/lib/urpmiPascal Rigaux2008-07-101-0/+1
|
* really keep .youri files for debuggingOlivier Blin2008-06-241-1/+1
|
* use better algorithm to prefer devel packages over source packagesOlivier Blin2008-04-021-11/+16
| | | | | and to handle contrib kernels correctly
* do not force incorrect symlink in /lib/modules if already presentOlivier Blin2008-04-021-4/+8
|
* do not explicitely specify kernel source dir, dkms will use /lib/modules by ↵Olivier Blin2008-04-021-1/+1
| | | | itself
* add package in the cache earlier, not to mail multiple times about incorrect ↵Olivier Blin2008-04-021-1/+3
| | | | dkms module
* prefer match with exact same name first (for devel packages to be preferred)Olivier Blin2008-04-021-1/+1
|
* print source package name instead of internal nameOlivier Blin2008-04-021-1/+1
|
* factorize (and avoid printing harmless error messages)Olivier Blin2008-04-021-15/+12
|
* use better regexp and rpm output formatOlivier Blin2008-04-021-3/+3
|
* do not run make prepare, dkms handles itOlivier Blin2008-04-021-2/+0
|
* remove old commentOlivier Blin2008-04-021-2/+0
|
* do not uses hacks if the dkms module is built for the wrong kernel, a ↵Olivier Blin2008-04-021-23/+5
| | | | warning mail is sent to maintainers instead
* handle rpms being installed in /var/lib/dkms (dkms-2.0.19)Olivier Blin2008-04-021-1/+1
|
* factorize (for next commit)Olivier Blin2008-04-021-2/+14
|
* send mail to admins when a DKMS module wrongly build for current kernelOlivier Blin2008-04-021-0/+8
|
* do not try to move modules built for current kernel if it was the targetOlivier Blin2008-04-021-1/+1
|
* make sure dkms commands are not run in rpm post scriptsOlivier Blin2008-04-021-0/+6
|
* fix removing /lib/modules/`uname -r` symlink hackOlivier Blin2008-04-021-1/+1
|