about summary refs log tree commit diff
AgeCommit message (Collapse)AuthorFilesLines
2011-04-11 * Create $manifestDir if it doesn't exist.Eelco Dolstra1-0/+3
2011-04-10 * Cache the manifests in /nix/var/nix/manifests in a SQLite database.Eelco Dolstra2-30/+164
This significantly speeds up the download-using-manifests substituter, especially if manifests are very large. For instance, one "nix-build -A geeqie" operation that updated four packages using binary patches went from 18.5s to 1.6s. It also significantly reduces memory use. The cache is kept in /nix/var/nix/manifests/cache.sqlite. It's updated automatically when manifests are added to or removed from /nix/var/nix/manifests. It might be interesting to have nix-pull store manifests directly in the DB, rather than storing them as separate flat files, but then we would need a command line interface to delete manifests from the DB.
2011-04-06 * Remove the localPaths feature in manifests since it's no longer usedEelco Dolstra7-47/+10
and redundant anyway.
2011-03-16 * Print a better error message.Eelco Dolstra1-1/+1
2011-03-15 * Fix a bug in the documentation (reported by Olexiy Buyanskyy,Eelco Dolstra1-1/+1
Nix/215).
2011-02-17 (no commit message)Eelco Dolstra1-1/+1
2011-02-17 * nix-push: handle the case where the hash is not set in the DB.Eelco Dolstra1-0/+10
2011-02-14 * Build for Ubuntu 10.04.Eelco Dolstra1-0/+2
2011-02-14 make nix-mode provide 'nix-modeFlorian Friesdorf1-0/+2
this enables (require 'nix-mode)
2011-02-10 * Don't allocate a big initial GC address space on machines withEelco Dolstra2-12/+25
little RAM. Even if the memory isn't actually used, it can cause problems with the overcommit heuristics in the kernel. So use a VM space of 25% of RAM, up to 384 MB.
2011-02-09 * Don't call GC_expand_hp unless we're actually using the garbageEelco Dolstra2-12/+18
collector.
2011-02-09 Use $BDW_GC_LIBS instead of a custom variable.Ludovic Courtès3-4/+2
2011-02-09 * A better fix. $boehmgc isn't set anywhere, we should use the flagsEelco Dolstra2-2/+2
returned by pkg-config.
2011-02-09 * The GC library can't be found on SolarisEelco Dolstra1-1/+1
(http://hydra.nixos.org/build/890714), so don't build with GC support for now.
2011-02-09 * Urgh, FreeBSD doesn't have a "seq" command.Eelco Dolstra1-2/+2
2011-02-09 * Obsolete.Eelco Dolstra22-353/+0
2011-02-09 * Remove obsolete directory.Eelco Dolstra1-252/+0
2011-02-09 * Remove obsolete file.Eelco Dolstra1-149/+0
2011-02-09 * Merged the SQLite branch.Eelco Dolstra82-1890/+2676
2011-02-09 * Sync with the trunk.Eelco Dolstra1-2/+0
2011-02-08 (no commit message)Eelco Dolstra1-1/+1
2011-02-05 * Propagate the CC setting.Eelco Dolstra1-1/+1
2011-02-05 * Clang compatibility fix. Clang actually checks template definitionsEelco Dolstra1-1/+1
when they are defined
2011-01-14 Remove useless <config.h> inclusion from public header.Ludovic Courtès1-2/+0
2011-01-14 * builtins.substring: if "start" is beyond the end of the string,Eelco Dolstra3-2/+4
return the empty string.
2010-12-22 * In Hydra manifests the Size field is missing, so don't rely on it.Eelco Dolstra1-2/+2
This caused a lot of "Use of uninitialized value" warnings from Perl.
2010-12-17 * Do a short sleep after SQLITE_BUSY.Eelco Dolstra2-2/+12
2010-12-17 * Propagate the "100" exit status for failed builds through the NixEelco Dolstra3-5/+9
daemon.
2010-12-17 * Use the right `make'.Eelco Dolstra1-1/+1
2010-12-15 * Disable X11 forwarding, it's not needed.Eelco Dolstra1-0/+2
2010-12-15 * nix-copy-closure: make sure that the shell doesn't do globbing ofEelco Dolstra1-6/+6
`+' and `?' in filenames. This is very slow if /nix/store is very large. (This is a quick hack - a cleaner solution would be to bypass the shell entirely.)
2010-12-14 * I forgot to catch SQLiteBusy in registerValidPaths(). SoEelco Dolstra1-38/+31
registerValidPaths() now handles busy errors and registerValidPath() is simply a wrapper around it.
2010-12-13 * nix-instantiate: return exit status 100 to denote a permanent buildEelco Dolstra5-14/+29
failure. The build hook can use this to distinguish between transient and permanent failures on the remote side.
2010-12-13 * Create /nix/var/nix/db if it's missing.Eelco Dolstra1-0/+1
2010-12-13 * createDirs(path): if path already exists, make sure it's aEelco Dolstra2-20/+26
directory. * Provide a C++ wrapper around lstat().
2010-12-13 * Update some comments.Eelco Dolstra2-4/+3
2010-12-13 * When doing a query (e.g. `nix-store -r --dry-run'), don't make a lotEelco Dolstra1-4/+12
of expensive calls to `nix-store --check-validity'.
2010-12-10 * Use SQLite 3.7.4.Eelco Dolstra2-6/+6
2010-12-08 * Ignore the result of sqlite3_reset().Eelco Dolstra1-3/+4
2010-12-07 * Bad things happen when a filehandle has the same name as a module.Eelco Dolstra1-2/+2
2010-12-06 * `nix-store --verify --check-contents': don't hold the global GC lockEelco Dolstra2-11/+49
while checking the contents, since this operation can take a very long time to finish. Also, fill in missing narSize fields in the DB while doing this.
2010-12-05 * Retry a transaction if SQLite returns SQLITE_BUSY. This can happenEelco Dolstra1-61/+83
even with a very long busy timeout, because SQLITE_BUSY is also returned to resolve deadlocks. This should get rid of random "database is locked" errors. This is kind of hard to test though. * Fix a horrible bug in deleteFromStore(): deletePathWrapped() should be called after committing the transaction, not before, because the commit might not succeed.
2010-12-05 * Use CamelCase for the Perl modules.Eelco Dolstra12-19/+18
2010-12-05 * Add a script `nix-generate-patches'.Eelco Dolstra4-7/+52
* Fix the binary patching test.
2010-11-29 * Quick hack around one of the pathological cases in bsdiffEelco Dolstra1-0/+1
(namely, when there is a long region of 0s). On one 86 MiB case, this cut patch generation time from 44m to 2m.
2010-11-29 * Move the patch generator into a module.Eelco Dolstra4-437/+345
2010-11-29 * Don't generate patches if the system type differs.Eelco Dolstra2-2/+13
2010-11-29 * Impose a configurable time limit on patch generation. This isEelco Dolstra1-6/+12
necessary because bsdiff has some pathological worst-case behaviour, e.g. O(n^2) on files consisting only of 0s.
2010-11-25 * Sync with the trunk.Eelco Dolstra1-0/+4
2010-11-25 * DrvInfo contains pointers to expressions, so DrvInfos should beEelco Dolstra1-0/+4
traced by the garbage collector. Otherwise "nix-env -u" can crash randomly.