about summary refs log tree commit diff
AgeCommit message (Collapse)AuthorFilesLines
2009-04-21 * Use foreach in a lot of places.Eelco Dolstra10-96/+59
2009-04-21 * nix-store --verify: don't bail out if a referenced path is missing.Eelco Dolstra1-14/+13
(It can't fix it though.)
2009-04-16 * Fix a few "comparison is always false/true due to limited range ofEelco Dolstra3-5/+5
data type" warnings on 64-bit platforms. The one in parser.y is likely to be a real bug.
2009-04-16 * Cleanup.Eelco Dolstra1-15/+6
2009-04-15 * Don't cache transient build hook problems.Eelco Dolstra1-6/+11
2009-04-14 * Remove references to __XXX primops.Eelco Dolstra2-14/+4
2009-04-14 * Release notes.Eelco Dolstra1-0/+60
2009-04-02 * Debian packages: use /etc, not /usr/etc.Eelco Dolstra1-0/+1
2009-03-31 * Do a substitution even if --max-jobs == 0.Eelco Dolstra2-18/+23
2009-03-31 * Doh!Eelco Dolstra1-1/+1
2009-03-30 (no commit message)Eelco Dolstra1-0/+1
2009-03-30 * Make the poll interval configurable.Eelco Dolstra2-3/+3
2009-03-29 * Update the mtime on the lock file to make it easy to see when a slotEelco Dolstra1-4/+8
was last used.
2009-03-29 * Undocument the "system" option. No sane person would use it :-)Eelco Dolstra1-20/+0
2009-03-29 * With `--max-jobs 0', print a nicer error message than "AssertionEelco Dolstra1-1/+5
`!awake.empty()' failed."
2009-03-29 * Use polling to wait for a remote build slot when using a build hookEelco Dolstra2-36/+3
(that is, call the build hook with a certain interval until it accepts the build). * build-remote.pl was totally broken: for all system types other than the local system type, it would send all builds to the *first* machine of the appropriate type.
2009-03-28 * nix-copy-closure: compute the closure in one call to nix-store,Eelco Dolstra2-26/+10
which is much faster.
2009-03-28 * Scan for references and compute the SHA-256 hash of the output inEelco Dolstra3-89/+50
one pass. This halves the amount of I/O.
2009-03-28 * Don't use the non-standard __gnu_cxx::stdio_filebuf class.Eelco Dolstra2-33/+17
2009-03-28 * Simplify communication with the hook a bit (don't use fileEelco Dolstra6-117/+68
descriptors 3/4, just use stdin/stderr).
2009-03-28 * Clean up the output a bit.Eelco Dolstra1-5/+5
2009-03-28 * The `fixedOutput' variable didn't get initialised when using a buildEelco Dolstra1-5/+6
hook, causing negative caching to fail randomly.
2009-03-27 * Argh, stupid timing sensitive tests...Eelco Dolstra1-4/+4
2009-03-27 * Debian 5.0, Ubuntu 8.10 builds.Eelco Dolstra1-4/+8
2009-03-27 * Create /nix/store if it doesn't exist.Eelco Dolstra1-0/+2
2009-03-26 * A job for creating a fully statically linked binary tarball.Eelco Dolstra1-1/+22
2009-03-26 * Support doing a fully static build. Statically linked Nix binariesEelco Dolstra1-11/+24
will be useful as a fallback on platforms for which there are no RPMs/Debs available.
2009-03-26 * Don't use ULLONG_MAX in maxFreed - use 0 to mean "no limit".Eelco Dolstra4-10/+8
18446744073709551615ULL breaks on GCC 3.3.6 (`integer constant is too large for "long" type').
2009-03-25 * Negative caching, i.e. caching of build failures. Disabled byEelco Dolstra7-7/+140
default. This is mostly useful for Hydra.
2009-03-25 * Removed the locking.sh test; it's redundant because of the extendedEelco Dolstra5-45/+11
parallel.sh test. Also, don't call multiple nix-builds in parallel, since they can race creating .nix-build-tmp-derivation.
2009-03-25 * Use bash in the tests.Eelco Dolstra2-1/+2
2009-03-24 * Doh.Eelco Dolstra1-1/+1
2009-03-24 (no commit message)Eelco Dolstra1-1/+1
2009-03-23 * Make this test a bit more robust. It's still timing dependentEelco Dolstra3-10/+14
though.
2009-03-23 * Shut up a GCC warning.Eelco Dolstra1-2/+2
2009-03-23 * No longer block while waiting for a lock on a store path. InsteadEelco Dolstra5-28/+85
poll for it (i.e. if we can't acquire the lock, then let the main select() loop wait for at most a few seconds and then try again). This improves parallelism: if two nix-store processes are both trying to build a path at the same time, the second one shouldn't block; it should first see if it can build other goals. Also, it prevents the deadlocks that have been occuring in Hydra lately, where a process waits for a lock held by another process that's waiting for a lock held by the first. The downside is that polling isn't really elegant, but POSIX doesn't provide a way to wait for locks in a select() loop. The only solution would be to spawn a thread for each lock to do a blocking fcntl() and then signal the main thread, but that would require pthreads.
2009-03-22 * Refactoring.Eelco Dolstra1-136/+85
2009-03-22 * Test case (currently fails): multiple Nix builds shouldn't blockEelco Dolstra1-2/+34
waiting on the same lock when there are other builds that can be done.
2009-03-22 * Better error message when nix-store --import is applied to garbageEelco Dolstra1-1/+1
(previously it would likely say "implementation cannot deal with > 32-bit integers").
2009-03-22 * NAR archives: handle files larger than 2^32 bytes. Previously itEelco Dolstra7-17/+48
would just silently store only (fileSize % 2^32) bytes. * Use posix_fallocate if available when unpacking archives. * Provide a better error message when trying to unpack something that isn't a NAR archive.
2009-03-19 * Future proofing: assume we can read manifests up to version 10Eelco Dolstra1-1/+6
(which should therefore be backwards compatible).
2009-03-19 * Disregard the Hash field in manifests.Eelco Dolstra1-22/+5
2009-03-18 * Unify exportReferencesGraph and exportBuildReferencesGraph, and makeEelco Dolstra4-67/+40
sure that it works as expected when you pass it a derivation. That is, we have to make sure that all build-time dependencies are built, and that they are all in the input closure (otherwise remote builds might fail, for example). This is ensured at instantiation time by adding all derivations and their sources to inputDrvs and inputSrcs.
2009-03-18 * Improve the test.Eelco Dolstra1-2/+6
2009-03-18 * Better cleanup after tests.Eelco Dolstra1-5/+8
2009-03-18 * Missing file.Eelco Dolstra1-0/+1
2009-03-18 * Acquire the locks on the output paths before trying to run the buildEelco Dolstra4-150/+78
hook. This fixes a problem with log files being partially or completely filled with 0's because another nix-store process truncates the log file. It should also be more efficient.
2009-03-18 * Clean up some tests (use nix-build where appropriate).Eelco Dolstra7-34/+25
2009-03-17 * Regression test for exportBuildReferencesGraph. It currently fails.Eelco Dolstra3-6/+34
2009-03-17 * Refactoring: renamed *.nix.in to *.nix.Eelco Dolstra22-264/+209