summaryrefslogtreecommitdiff
path: root/usr.sbin/syspatch
AgeCommit message (Expand)Author
2016-12-02Few consistency fixes (no intended change in behaviour).Antoine Jacoutot
2016-12-01Fix kernel matching installation:Antoine Jacoutot
2016-11-30Simplify:Antoine Jacoutot
2016-11-30During early testing, hardcode the syspatch repo to 'syspatch.openbsd.org'Antoine Jacoutot
2016-11-30Change the hierarchy under /var/syspatch/ so that the output of installed orAntoine Jacoutot
2016-11-29Document the fact that *for now* syspatch only works on official releases andAntoine Jacoutot
2016-11-27Remove the backup kernel in rollback_patch() instead of the cleanupAntoine Jacoutot
2016-11-27Explicitely set umask.Antoine Jacoutot
2016-11-25Check args.Antoine Jacoutot
2016-11-22Make sure /var/syspatch/$OSREV is not on a read-only mount-point when doingAntoine Jacoutot
2016-11-22Allow passing an explicit return code to sp_err() in case we don't want to exitAntoine Jacoutot
2016-11-21Use the same basename for the rollback tarball as the original patch.sig.Antoine Jacoutot
2016-11-21Tweak comment. Reverse cmp logic while here.Antoine Jacoutot
2016-11-17Re-add the root user checks in the quirks; that part is to be removed soon.Antoine Jacoutot
2016-11-17Drop priv. to the _pkgfetch (for now) user when running ftp(1) and signify(1).Antoine Jacoutot
2016-11-16Merge comments altogether, it's easier to understand this way.Antoine Jacoutot
2016-11-16If we install a new kernel, add /bsd twice (for size checking) when:Antoine Jacoutot
2016-11-16/bsd.rollbackXX -> bsd.syspatchXX to make it more obvious where it comes fromAntoine Jacoutot
2016-11-15Small wording tweak and a couple of fixes from jmc@Antoine Jacoutot
2016-11-15checkfs: ignore nonexistent files and extend comment. There's actually no needAntoine Jacoutot
2016-11-15Double the required size to make sure we have enough space for install(1) safeAntoine Jacoutot
2016-11-14Check for available space before installing a patch.Antoine Jacoutot
2016-11-14uniq -> sort -uAntoine Jacoutot
2016-11-12Improve the cleanup error handling to make sure the exit code isAlexander Hall
2016-11-10Extend documentation a bit. It'll probably need some adjustments soon.Antoine Jacoutot
2016-11-10Simplify checkfs() and fix read-only/remote fs detection.Antoine Jacoutot
2016-11-10Extend mtree(8) comment.Antoine Jacoutot
2016-11-10Remove redundant check.Antoine Jacoutot
2016-11-10Create the bsd rollback kernel in create_rollback() (it is contained inAntoine Jacoutot
2016-11-08syspatch-60-001_foobar -> syspatch60-001_foobar to match base system sets name.Antoine Jacoutot
2016-11-08Simplify.Antoine Jacoutot
2016-11-08Few fixes and consistency.Antoine Jacoutot
2016-11-06Rework the cleanup trap handling using the EXIT trap;Alexander Hall
2016-11-04Be verbose when PATCH_PATH is not set (that is temporary until we agree onAntoine Jacoutot
2016-11-04One more XXX.Antoine Jacoutot
2016-11-04Zap extra space.Antoine Jacoutot
2016-11-04Make sure our filesystems are local and not read-only.Antoine Jacoutot
2016-11-04Use 'rm -f' to remove the rollback tarball if we have an errir; it mayAntoine Jacoutot
2016-11-03Missing local.Antoine Jacoutot
2016-11-03rollback -> revert where it makes sense.Antoine Jacoutot
2016-11-03XXX match with installed sets (comp, x...)?Antoine Jacoutot
2016-11-03Hardlinks are properly handled; for the rest, we'll see if we need to careAntoine Jacoutot
2016-11-03Make sure PATCH_PATH is an URL that ftp(1) can cope with.Antoine Jacoutot
2016-11-03Merge ls_avail() into ls_missing(), it's only used once.Antoine Jacoutot
2016-11-03Use hw.ncpufound.Antoine Jacoutot
2016-11-03Simplify for loops; prompted by a comment from rpe@Antoine Jacoutot
2016-11-02Be verbose when reverting a patch.Antoine Jacoutot
2016-11-01Simplify.Antoine Jacoutot
2016-11-01Tap SIGINT while install(1)ing so that we can properly rollback and notAntoine Jacoutot
2016-11-01Add an XXX.Antoine Jacoutot