aboutsummaryrefslogtreecommitdiffstats
path: root/Tools/portbuild
Commit message (Expand)AuthorAgeFilesLines
* Explicitly mount the NFS images ro, instead of relying on the server tokris2006-04-161-1/+1
* Don't bother archiving the ports tree, we don't use it for anything so itkris2006-04-161-1/+1
* Mount linprocfs on amd64 for the convenience of jdk builds.kris2006-04-031-1/+1
* The squid rc.d script no longer has the .sh suffixkris2006-04-032-4/+4
* Do not preserve the previous package sets in bak/. This is taking upkris2006-04-031-2/+2
* Clean up linprocfs on amd64 as well.kris2006-04-031-1/+1
* jdk is buildable on amd64, so we need to also mount linprocfs there.kris2006-04-031-2/+2
* This script preserves packages that will be removed by the restricted.shkris2006-04-031-0/+11
* Temporarily back out the pre.log change, it doesn't work properly.kris2006-03-131-6/+6
* Mount the md image async here tookris2006-03-131-1/+1
* Add some more directories temporarily until these ports can be made tokris2006-03-131-4/+8
* Add a mechanism for reporting client machine error conditions back to thekris2006-02-151-1/+14
* Trim some unused cruftkris2006-02-151-69/+12
* Save the output of portbuild run on the client to a temporary log file.kris2006-02-151-1/+11
* Add some sanity checking of the build environment:kris2006-02-151-1/+30
* * Don't try to pkg_delete packages that are not installed (the usualkris2006-02-151-6/+9
* Add support for 6-exp builds (FreeBSD 6.x is our new reference platform)kris2006-02-154-4/+14
* * Switch back to using pkg_delete -f; we are not yet ready to dealkris2006-02-141-25/+26
* * In kill_procs(), pass in the directory and mountpoint as separatekris2006-02-111-2/+8
* Support for building ports in a jail instead of a chroot. If use_jail=1kris2006-02-111-9/+14
* * Use ${mailto} portbuild.conf variable instead of hard-coded list [1]kris2006-02-111-4/+7
* Use ${mailto} portbuild.conf variable instead of hard-coded listkris2006-02-111-1/+1
* Checkpoint a major round of changes:kris2006-02-111-64/+201
* Flip from noclean to cleankris2006-02-041-3/+3
* Don't bother chowning the client /var/portbuild/, it just takes too longkris2006-02-041-3/+0
* Use clean-chroot with clean=2 to remove the chroot instead of duplicatingkris2006-02-041-5/+1
* Flip "noclean" to "clean"kris2006-02-041-3/+3
* Instead of passing in a "noclean" boolean, pass in a "clean" level:kris2006-02-041-4/+11
* Also look for chroots that are marked 'dirty' (previous build left chrootkris2006-02-041-1/+5
* Don't try to remove the lock file now that we use lockf to manage itkris2006-02-041-2/+0
* Correctly set SRCBASE; the version inherited from buildenv is wrong forkris2006-01-281-0/+1
* Don't try to rsync docs, we no longer need to distribute themkris2006-01-281-1/+0
* Update the list of quickportskris2006-01-281-1/+1
* Don't bother mounting doc tree, it's not needed by any portskris2006-01-221-12/+8
* Remove -nodoccvs option, it turns out mounting the doc tree is notkris2006-01-221-14/+1
* Also clean up linprocfs mountskris2006-01-221-0/+1
* Pull in buildenv for LOCALBASE/X11BASE locationskris2006-01-221-2/+4
* Use LOCALBASE and X11BASE from buildenvkris2006-01-221-1/+4
* Allow for the possibility of overriding LOCALBASE and X11BASEkris2006-01-221-0/+2
* Exit with the correct error code when a make target failskris2006-01-151-3/+3
* Back out last commit, it was the wrong filekris2006-01-151-17/+15
* Exit with the correct error code when a make target failskris2006-01-151-15/+17
* This script may be run on a client machine to force it to flush its squidkris2005-12-111-0/+15
* Bump the file size limit to 300MB now that openoffice distfiles exceedkris2005-12-111-5/+5
* Shuffle the order of a cdkris2005-12-111-2/+1
* Remove 0-sized entries to avoid misbehaviourkris2005-12-111-0/+4
* Don't use a lockfile, the script will be called with lockf -t0 instead tokris2005-12-112-16/+0
* Don't use a lockfile, the script will be called with lockf -t0 insteadkris2005-12-111-6/+0
* Update the list of quickportskris2005-12-111-1/+1
* Don't set -nocopy when -norsync is set, otherwise the setupnode scriptkris2005-12-111-1/+1
* Don't use a lock file, the script will be called with lockf -t 0 instead.kris2005-12-111-6/+0
* Mount the md image async since this gives much better performancekris2005-12-111-1/+1
* Remove, don't comment out, unused variableskris2005-10-111-2/+0
* Remove unused variables. Don't bother protecting against negativekris2005-10-111-4/+2
* Update lockf to lockf -k in example usagekris2005-10-111-1/+1
* Update the lockf command to add -k in example usagekris2005-10-111-1/+1
* When -norsync is set, don't bother to run md5 over the bindist.tarkris2005-10-111-3/+9
* Change lockf to lockf -k in a commented-out command, just in case it'skris2005-10-111-1/+1
* Add support for use_md_swapkris2005-10-111-0/+5
* Add support for use_md_swap. When this variable is set in akris2005-10-111-4/+18
* No need to hold the queue lock while polling client loadkris2005-10-111-1/+1
* * Use lockf -k instead of lockf. If there is ever contention from >1kris2005-10-111-5/+5
* * Add support for use_md_swapkris2005-10-111-12/+22
* * Unmount linprocfs around pkg_adding dependencies, to try and workkris2005-10-111-3/+41
* * Finish flipping the switch on -noplistcheck - this is activated bykris2005-10-111-11/+6
* * Remove support for the unused -nodummy functionkris2005-10-111-65/+66
* * Remove unused -nodummy argumentkris2005-10-111-34/+6
* Silence some possible errors from chflags -Rkris2005-10-111-1/+1
* Silence possible errors from some recursive commandskris2005-10-111-2/+2
* Check for extra files on !i386 alsokris2005-07-251-6/+4
* Support FreeBSD 7.xkris2005-07-255-5/+15
* Drop the removed openoffice 1.0 ports from the quickports listkris2005-07-251-1/+1
* Correct typokris2005-07-251-1/+1
* Use WRKDIRPREFIX instead of hardcoded value.kris2005-04-121-1/+3
* Change rsync -l to -L for the tarballs/ directory (so that danglingkris2005-04-031-1/+1
* Build packages with WRKDIRPREFIX=/work (which does not exist bykris2005-04-031-1/+9
* Increase the nc poll timeout to better deal with network latencies.kris2005-04-031-1/+1
* Flip over the -plistcheck switch to -noplistcheck since we now want tokris2005-04-031-5/+5
* Update the list of files and directories to filter from the plist checking.kris2005-04-031-1/+3
* * Instead of using umount -f to unmount things, first use fstat tokris2005-02-121-6/+44
* Only try to chown the portbuild directory if we are a disconnected clientkris2005-02-121-1/+1
* Instead of using umount -f to unmount things, first use fstat to look forkris2005-02-121-1/+35
* * Instead of using umount -f to unmount things, first use fstat tokris2005-02-121-7/+44
* Update the plist checking to look for a subset of files and weed out somekris2005-02-121-1/+6
* Remove traling spaces.krion2005-01-0912-29/+29
* Always set up the bindist-${hostname}.tar file since it's usually small.kris2005-01-071-1/+1
* Unset the __MAKE_CONF variable after setting up buildenv, because wekris2005-01-071-0/+3
* Allow machines to use an alternate address for polling for job load (e.g.kris2005-01-071-1/+3
* * Include portbuild.conf before portbuild.${hostname} to reset the defaultkris2005-01-071-4/+6
* * The mlist file now only has one field per line.kris2005-01-071-4/+9
* krion wants to drink from the pointyhat email firehose as well.kris2005-01-071-1/+1
* Now that I have so many remote build machines, don't force trying tokris2005-01-071-7/+0
* * Support the -norsync argument to only remove the existing build chrootskris2004-12-281-2/+8
* * It does in fact make sense to specify -incremental and -noindex at thekris2004-12-281-7/+7
* Fix the Xvfb code to run Xvfb before the build phase if detectedkris2004-12-281-15/+25
* Overhaul of the job scheduler. The new scheduler runs buildskris2004-12-284-29/+119
* Overhaul of the job scheduler. The new scheduler runs buildskris2004-12-281-26/+33
* * Catch up to change in formatting of 'extra files' section in logskris2004-12-281-4/+3
* /rescue/mount_linprocfs isn't functional when called bykris2004-12-281-1/+2
* Support 5-exp buildskris2004-12-284-4/+14
* Use X_WINDOW_SYSTEM for 4.x builds too.kris2004-12-101-2/+2
* Belatedly catch up to XFREE86_VERSION -> X_WINDOW_SYSTEM change; applykris2004-12-101-3/+3
* The explanations of each build error no longer live on bento, they livelinimon2004-12-092-7/+7
* Construct permanent links to the error logs here too.kris2004-12-051-4/+4
* Copy the statically-linked binaries from /rescue on the host, insteadkris2004-12-051-4/+11
* Be silent about attempts to unmount filesystems from within the chrootkris2004-12-051-4/+4
* Clean up the symlink in ${pb}/${arch}/${branch}/latest/${dir} when we findkris2004-11-251-0/+5
* If sudo_cmd is set, assume we need to chown the files we created onkris2004-11-251-0/+3
* Rework the -incremental support slightly; this should now be robust enoughkris2004-11-251-3/+5
* Catch up to the changed format of the logs.kris2004-11-251-4/+4
* * Recursively list all files detected as having been left behind afterkris2004-11-251-7/+22
* Fix a couple of bugs in the extraction of this from dopackages, andkris2004-10-251-33/+20
* Redefine the -norestr option to mean 'don't build the list ofkris2004-10-251-78/+73
* If the symlink under latest/ exists, then use that to find the most recentkris2004-10-251-1/+6
* We now also do 6.x builds in the '6' branch.kris2004-10-251-3/+3
* Use the absolute path to nc(1) to deal with it not being in the PATH.kris2004-10-251-3/+1
* Clean up the 'latest error log' symlink when the build is successful.kris2004-10-251-0/+4
* When a port build fails, create a symlink to the error log underkris2004-10-251-3/+4
* Support ${client_user} and ${sudo_cmd} for running commands as root onkris2004-09-271-1/+5
* * Break out setupnode into the dosetupnode script so it may be calledkris2004-09-271-37/+3
* Break out the setupnode function of dopackages into a separate script sokris2004-09-271-0/+47
* * Bail out if we can't claim a chroot directory in the client (e.g. ifkris2004-09-271-13/+17
* Refuse to "clean" / if something goes wrong upstream and we are passedkris2004-09-271-0/+5
* Pull in portbuild.$(hostname) for per-client customizationkris2004-09-271-1/+3
* Support 6.x buildskris2004-09-272-1/+5
* * Support 6.x buildskris2004-09-271-2/+12
* Use a consistent PATH in both portbuild and dopackages.marcus2004-08-281-1/+1
* Clean up the specified chroot (umount filesystems, and scrub thekris2004-07-141-0/+36
* Atomically claim an unused chroot directory. The previous algorithmkris2004-07-141-0/+52
* Bump up the cleanup timeout to directories that have been in use forkris2004-07-141-3/+3
* Back out the attempt to list the error type in the failure summary;kris2004-07-141-14/+1
* Generate permanent URLs in the error reports. The directories likekris2004-07-141-1/+10
* * Add support for ${TRYBROKEN}.kris2004-07-141-6/+67
* * Add a cleanup() function and use it so that we archive the ${WRKDIR}kris2004-07-141-68/+122
* * Remove the quickports stuff, it didn't (ever?) work as designed andkris2004-07-141-112/+156
* Parse the slightly-different format of the build logs.kris2004-07-141-1/+1
* Parse the slightly-different format of the build logs.kris2004-07-141-3/+3
* * Accept another argument, the value of the disconnected variable thatkris2004-07-141-25/+25
* Allow local customization of the node using the portbuild.$(hostname) file.kris2004-07-141-0/+1
* Send me mail when a previously-broken port builds successfully.kris2004-07-141-2/+5
* * Initial support for disconnected nodes (not on same LAN as buildkris2004-07-141-83/+88
* Hack around the sparc64 64-bit time_t test in 'installworld' sincekris2004-07-141-0/+1
* * Break out the 'parallel' target from /usr/ports/Makefile. Onekris2004-07-141-6/+40
* Set INDEX_JOBS to control index build concurrencykris2004-07-141-0/+3
* Drop the ignorelist concurrency back to 1 because I'm occasionally seeingkris2004-07-141-1/+1
* * If we time out connecting to a machine, give it a weight of 999 to makekris2004-07-141-2/+2
* Chase bento to pointyhatkris2004-07-141-1/+1
* Add dopackages.wrapper, a wrapper script around dopackages that sets upmarcus2004-06-031-0/+53
* Set INDEX_PRISTINE and INDEX_QUIET to catch up to new index buildkris2004-04-021-3/+5
* Run a command on all client machines of a certain architecture, or allkris2004-03-081-0/+50
* Walk through a package set and prune files that are not listed in INDEXkris2004-03-081-0/+66
* Verify the archive integrity of a package set, and move aside any thatkris2004-03-081-0/+45
* To be run on the client, this script looks for chroot directories that have n...kris2004-03-081-0/+48
* Break out a script from dopackages for building the restricted.sh filekris2004-03-081-0/+34
* Replace gnome with gnome2 in quickports. I'm still not sure this actuallykris2004-03-081-1/+1
* Remove the integrity testing code now that the corruption bug has beenkris2004-03-081-4/+0
* * Fix 4.x builds with 5.x host systems by copying some additional neededkris2004-03-081-0/+6
* Expand out some shell globbing syntaxkris2004-03-081-1/+1
* Prune out packages that will not be built because a dependency is markedkris2004-03-081-1/+1
* Create a duds.full file that lists not only files that are marked IGNOREkris2004-03-081-0/+4
* Increase timeout on slow build machines (alpha, sparc64) to 16 hours.kris2004-03-081-2/+8
* * Set the PACKAGE_BUILDING variable in the build environment so ports cankris2004-03-081-1/+4
* Set up the DISTDIR and PACKAGES variables for use by client scriptskris2004-03-081-0/+3
* Fix up the ordering and remove some duplicates so the resulting summary ismarcus2004-02-041-22/+26
* Comment out the munmap reason since it is quite unlikely this will bemarcus2004-01-221-4/+4
* * Refactor processlogs so that new error types can be added easily to a newmarcus2004-01-203-312/+441
* * Add support for linprocfs (needed for JDK ports) [1]marcus2004-01-201-2/+23
* Use editors/openoffice-1.1 instead of editors/openoffice since the latter ismarcus2004-01-161-1/+1
* Condense the table a bitkris2003-10-131-3/+3
* The newfailure file is no longer updated by buildskris2003-10-131-3/+0
* Script for migrating from old-style to new-style failure log file.kris2003-10-131-0/+74
* * newfailure is no longer updated, no need to prune itkris2003-10-131-42/+6
* * Generate newfailure file (sorted list by original failure date) from failurekris2003-10-131-10/+13
* * Try to be more careful when writing the failure file in case the diskkris2003-10-131-13/+16
* Add a script that prunes the contents of the failure and newfailure fileskris2003-10-131-0/+114
* Teach this script about different architectures.kris2003-09-132-61/+54
* Add amd64 to the list of architectures to poll.kris2003-09-131-1/+1
* Add a couple of amd64 failure messages.kris2003-09-131-0/+6
* Remove some ldconfig -R's that were apparently confusing things, coalescekris2003-09-131-6/+3
* Set PACKAGE_BUILDING before making the duds file to prevent listing ofkris2003-09-131-0/+1
* Export MASTER_SITE_OVERRIDE with DIST_SUBDIR correctly escapedkris2003-09-131-2/+1
* Catch up to makefile changes: build in etc/ before trying to installkris2003-09-131-0/+2
* Add a -fetch-original switch to dopackages that fetches the distfiles fromkris2003-05-183-1/+16
* Only transfer the build distfiles to bento if requested.kris2003-05-171-7/+14
* Pass down the request to keep distfiles if appropriate.kris2003-05-171-0/+3
* Add a -distfiles flag to specify that we should retain distfiles for laterkris2003-05-171-4/+12
* Switch to polling the clients with netcat to collect load stats, insteadkris2003-05-171-13/+10
* Update sed pattern used to process logskris2003-04-061-1/+1
* Zap some old/non-useful codekris2003-04-061-108/+106
* - Define __MAKE_CONF to allow per-arch make.conf files in buildworldkris2003-03-231-5/+19
* Instead of pushing the load data to the master with scp every 10 seconds,kris2003-02-231-13/+3
* * Allow port builds to be started with a certain niceness ('nice' parameterkris2003-02-231-1/+15
* If ${buildworld}=1 in mkbindist.conf, then use the makeworld script tokris2003-02-231-48/+44
* Simplify the collection of load statistics for the cluster. Loadskris2003-02-231-33/+35
* Store dates in +%s formatkris2003-02-231-3/+3
* Search for a lot of new failure strings to reduce the number of buildkris2003-02-141-4/+177
* Work around broken rand() in -current by changing the way we pick akris2003-02-141-3/+3
* - If ${ftp} = 0 in mkbindist.conf, then use the makeworld script to buildkris2003-02-141-77/+56
* - Correct usage messagekris2003-02-141-7/+2
* If the port directory contains a file called .keep, then tar up thekris2003-01-242-0/+10
* These scripts have been replaced by .c versions that do not require perl.kris2003-01-242-111/+0
* - Use ${arch} and rename html files accordinglykris2003-01-241-9/+11
* First cut at a script to process the logfile maintained by 'buildsuccess'kris2003-01-241-0/+68
* These scripts are run when a build completes (successfully or not). Theykris2003-01-242-0/+80
* Simple script to check the number of packages built on the variouskris2003-01-241-0/+17
* Script to automate the process of cvs updating the build tree, performingkris2003-01-241-0/+91
* - Use ${arch}kris2003-01-241-5/+6
* Determine client architecture using /etc/archkris2003-01-241-4/+2
* - Use ${arch}kris2003-01-241-2/+3
* Check status of multiple architecture client machineskris2003-01-241-35/+40
* - Use ${arch}kris2003-01-241-20/+27
* - Use ${arch}kris2003-01-241-11/+11
* - Use ${arch}kris2003-01-241-6/+7
* - Use ${arch} and ${INDEXFILE}kris2003-01-241-6/+11
* Add the ${arch} argumentkris2003-01-241-4/+6
* - Use the ${arch} subdirectory and pass ${arch} around to other scriptskris2003-01-241-155/+161
* Add makeparallel script, an improved version of the 'parallel' targetkris2003-01-241-0/+30
* Use pkg_add instead of pkg_add -fkris2003-01-241-3/+3
* Sync with bento:kris2003-01-241-13/+13
* Remove these scripts, which are no longer useful.kris2003-01-243-47/+0
* Convert this to buildenv to make it actually work properly.kris2002-11-181-38/+7
* Fix the html output of this script. There was neither a closingkris2002-11-101-15/+21
* Catch up to lots of changes from bento/beta:kris2002-11-091-144/+180
* Revamp this script:kris2002-11-041-64/+243
* Add C versions of the pnohang and ptimeout scripts (previously written inkris2002-11-032-0/+192
* Use the correct filename for the FreeBSD version string used by thekris2002-11-021-14/+2
* Read FTP url from ${ftpurl}. Remove hack to fetch perl package into thekris2002-11-021-8/+2
* Support either .tgz or .tbz packageskris2002-11-021-2/+2
* Add FTP_PASSIVE_MODE. Build tgz packages on 4 and 4-exp, and tbz on 5.kris2002-11-021-2/+8
* Correct a messagekris2002-09-161-1/+1
* Conditionalize one more thing on i386kris2002-09-161-1/+1
* Optimize the html output by putting multiple missing files in the samekris2002-09-161-7/+9
* Add a bunch of new failure modes (mostly to do with ports failing on 5.0)kris2002-09-161-2/+14
* Use PKGSUFFIX from buildenvkris2002-09-161-1/+5
* Add PKG_PATH so packages can be found automatically (not sure what causedkris2002-09-161-1/+3
* .tbz packages are now standardkris2002-09-161-2/+2
* * XFREE86_VERSION=4 is now standardkris2002-09-161-3/+4
* * Support multiple arches and conditionalize i386-only stuffkris2002-09-161-2/+12
* * Catch up to .tbz packageskris2002-09-161-6/+11
* Allow packages to be installed in the build chroot; this is neededkris2002-06-071-1/+9
* Only run ldconfig -aout commands on i386kris2002-06-071-6/+16
* - Catch up to buildenv changes.kris2002-05-131-12/+21
* Comment out the kernel building and just link to kernel.GENERIC.kris2002-05-131-5/+6
* Zap some unused code. Catch up to buildenv changes.kris2002-05-131-7/+1
* - Switch default to XFree86-4.x for all brancheskris2002-05-131-3/+6
* * Clean up this script a bitkris2002-03-251-36/+49
* * Clean up this script a bitkris2002-03-251-19/+27
* Detect a couple more failure modeskris2002-03-251-0/+6
* Add MASTER_SITE_OVERRIDE to the list of variables propagated, so thatkris2002-03-251-0/+2
* Update OSVERSIONS. This needs to be switched to buildenv.kris2002-03-161-3/+3
generated by cgit (git 2.34.1) at 2024-11-09 04:53:45 +0800