aboutsummaryrefslogtreecommitdiff
path: root/src/libstore
AgeCommit message (Collapse)Author
2015-02-17Use $<attr>Path instead of $<attr> for passAsFileEelco Dolstra
2015-02-17Allow passing attributes via files instead of environment variablesEelco Dolstra
Closes #473.
2015-02-16Use pivot_root in addition to chroot when possibleHarald van Dijk
chroot only changes the process root directory, not the mount namespace root directory, and it is well-known that any process with chroot capability can break out of a chroot "jail". By using pivot_root as well, and unmounting the original mount namespace root directory, breaking out becomes impossible. Non-root processes typically have no ability to use chroot() anyway, but they can gain that capability through the use of clone() or unshare(). For security reasons, these syscalls are limited in functionality when used inside a normal chroot environment. Using pivot_root() this way does allow those syscalls to be put to their full use.
2015-02-04Require linux 3.13 or later for chrootShea Levy
Fixes #453
2015-01-18Make inputs writeable in the sandbox (builds still can’t actually write ↵Daniel Peebles
due to user permissions)
2015-01-13Allow using /bin and /usr/bin as impure prefixes on non-darwin by defaultShea Levy
These directories are generally world-readable anyway, and give us the two most common linux impurities (env and sh)
2015-01-13SysError -> ErrorEelco Dolstra
2015-01-13Don't resolve symlinks while checking __impureHostDepsEelco Dolstra
Since these come from untrusted users, we shouldn't do any I/O on them before we've checked that they're in an allowed prefix.
2015-01-12Add basic Apple sandbox supportDaniel Peebles
2015-01-08Doh^2Eelco Dolstra
2015-01-08DohEelco Dolstra
2015-01-08Set /nix/store permission to 1737Eelco Dolstra
I.e., not readable to the nixbld group. This improves purity a bit for non-chroot builds, because it prevents a builder from enumerating store paths (i.e. it can only access paths it knows about).
2015-01-06Fix building on DarwinEelco Dolstra
Fixes #433.
2014-12-29LocalStore initialization: Don't die if build-users-group doesn't existShea Levy
See NixOS/nixpkgs@9245516
2014-12-23Revive running builds in a PID namespaceEelco Dolstra
2014-12-14PedantryEelco Dolstra
2014-12-14Merge branch 'cygwin-master' of https://github.com/ternaris/nixEelco Dolstra
2014-12-13Better error messageEelco Dolstra
2014-12-12Silence some warnings on GCC 4.9Eelco Dolstra
2014-12-12Shut up a Valgrind warningEelco Dolstra
2014-12-12Fix some memory leaksEelco Dolstra
2014-12-12Ensure we're writing to stderr in the builderEelco Dolstra
http://hydra.nixos.org/build/17862041
2014-12-12Remove chatty messageEelco Dolstra
This broke building with "-vv", because the builder is not allowed to write to stderr at this point.
2014-12-12DohEelco Dolstra
2014-12-10Don't do vfork in conjunction with setuidEelco Dolstra
2014-12-10Rename functionEelco Dolstra
2014-12-10Don't wait for PID -1Eelco Dolstra
The pid field can be -1 if forking the substituter process failed.
2014-12-09Explicitly include required C headersMarko Durkovic
2014-11-24Build derivations in a more predictable orderEelco Dolstra
Derivations are now built in order of derivation name, so a package named "aardvark" is built before "baboon". Fixes #399.
2014-11-24Don't create unnecessary substitution goals for derivationsEelco Dolstra
2014-11-19Disable vacuuming the DB after garbage collectionEelco Dolstra
Especially in WAL mode on a highly loaded machine, this is not a good idea because it results in a WAL file of approximately the same size ad the database, which apparently cannot be deleted while anybody is accessing it.
2014-11-19Clean up temp roots in a more C++ wayEelco Dolstra
2014-11-17Fix messageEelco Dolstra
2014-11-14Don't use ADDR_LIMIT_3GBEelco Dolstra
This gives 32-bit builds on x86_64-linux more memory.
2014-11-12Make ~DerivationGoal more reliableEelco Dolstra
2014-11-04nix-store --gc: Don't warn about missing manifests directoryEelco Dolstra
2014-10-31Improve error message if the daemon worker fails to startEelco Dolstra
2014-10-29Remove comments claiming we use a private PID namespaceEelco Dolstra
This is no longer the case since 524f89f1399724e596f61faba2c6861b1bb7b9c5.
2014-10-14Improved error message when encountering unsupported file typesEelco Dolstra
Fixes #269.
2014-10-03Remove some duplicate codeEelco Dolstra
2014-09-18Update spec fileEelco Dolstra
http://hydra.nixos.org/build/14344391
2014-09-18Install some pkgconfig filesEelco Dolstra
2014-09-17On Linux, disable address space randomizationEelco Dolstra
2014-09-17Settings: Add bool get()Eelco Dolstra
2014-09-01Add an 'optimiseStore' remote procedure call.Ludovic Courtès
2014-08-28Add disallowedReferences / disallowedRequisitesEelco Dolstra
For the "stdenv accidentally referring to bootstrap-tools", it seems easier to specify the path that we don't want to depend on, e.g. disallowedRequisites = [ bootstrapTools ];
2014-08-28Introduce allowedRequisites featureGergely Risko
2014-08-21fix disappearing bash argumentsJoel Taylor
2014-08-21Use unshare() instead of clone()Eelco Dolstra
It turns out that using clone() to start a child process is unsafe in a multithreaded program. It can cause the initialisation of a build child process to hang in setgroups(), as seen several times in the build farm: The reason is that Glibc thinks that the other threads of the parent exist in the child, so in setxid_mark_thread() it tries to get a futex that has been acquired by another thread just before the clone(). With fork(), Glibc runs pthread_atfork() handlers that take care of this (in particular, __reclaim_stacks()). But clone() doesn't do that. Fortunately, we can use fork()+unshare() instead of clone() to set up private namespaces. See also https://www.mail-archive.com/lxc-devel@lists.linuxcontainers.org/msg03434.html.
2014-08-20Use proper quotes everywhereEelco Dolstra