depot - monorepo for the virus lounge

Age	Commit message (Collapse)	Author	Files	Lines
2014-07-24	tests/remote-builds.nix: Test failing build	Eelco Dolstra	1	-1/+4

2014-07-24	nix-store --serve: Only monitor stdin during builds	Eelco Dolstra	1	-2/+2
	Other operations cannot hang indefinitely (except when we're reading from stdin, in which case we'll notice a client disconnect). But monitoring works badly during compressed imports, since there the client can close the connection before we've sent an ack. http://hydra.nixos.org/build/12711638
2014-07-24	tests/remote-builds.nix: Don't try cache.nixos.org	Eelco Dolstra	1	-0/+1

2014-07-24	Use pthread_cancel instead of a signal	Eelco Dolstra	1	-5/+2
	Signal handlers are process-wide, so sending SIGINT to the monitor thread will cause the normal SIGINT handler to run. This sets the isInterrupted flag, which is not what we want. So use pthread_cancel instead.
2014-07-24	Fix bogus pass by reference	Eelco Dolstra	1	-6/+1
	http://hydra.nixos.org/build/12711659
2014-07-24	More debugging	Eelco Dolstra	1	-0/+1

2014-07-24	Add some assertions	Eelco Dolstra	1	-0/+6

2014-07-24	Remove some dead code	Eelco Dolstra	2	-10/+0

2014-07-23	Remove some obsolete files	Eelco Dolstra	2	-310/+0

2014-07-23	Pass -pthread only for programs that need it	Eelco Dolstra	3	-2/+3

2014-07-23	nix-daemon: Less verbosity	Eelco Dolstra	1	-1/+1

2014-07-23	nix-daemon: Simplify stderr handling	Eelco Dolstra	2	-7/+2

2014-07-23	nix-store --serve: Monitor for client disconnects	Eelco Dolstra	1	-0/+3
	This is necessary because build-remote.pl now builds via ‘nix-store --serve’. So if a build hangs without writing to stdout/stderr, and the client disconnects, then we need to detect that.
2014-07-23	nix-daemon: Use a thread instead of SIGPOLL to catch client disconnects	Eelco Dolstra	3	-146/+51
	The thread calls poll() to wait until a HUP (or other error event) happens on the client connection. If so, it sends SIGINT to the main thread, which is then cleaned up normally. This is much nicer than messing around with SIGPOLL.
2014-07-23	startProcess: Make writing error messages from the child more robust	Eelco Dolstra	1	-2/+4

2014-07-23	Remove dead code	Eelco Dolstra	2	-3/+0

2014-07-19	Revert old useBuildHook behaviour	Eelco Dolstra	3	-3/+3

2014-07-18	Better fix for strcasecmp on Darwin	Eelco Dolstra	1	-3/+1

2014-07-18	Bump	Eelco Dolstra	1	-0/+9

2014-07-17	Ugly hack to fix building on old Darwin	Eelco Dolstra	1	-0/+3
	http://hydra.nixos.org/build/12580878
2014-07-17	nix-daemon: Add trusted-users and allowed-users options	Eelco Dolstra	4	-3/+90
	‘trusted-users’ is a list of users and groups that have elevated rights, such as the ability to specify binary caches. It defaults to ‘root’. A typical value would be ‘@wheel’ to specify all users in the wheel group. ‘allowed-users’ is a list of users and groups that are allowed to connect to the daemon. It defaults to ‘*’. A typical value would be ‘@users’ to specify the ‘users’ group.
2014-07-17	nix-daemon: Show name of connecting user	Eelco Dolstra	1	-6/+7

2014-07-17	nix-daemon: Only print connection info if we have SO_PEERCRED	Eelco Dolstra	1	-9/+12

2014-07-17	nix-daemon: Fix compat with older clients	Eelco Dolstra	1	-1/+1

2014-07-16	Get rid of a compiler warning	Eelco Dolstra	1	-1/+2

2014-07-16	Be more strict about file names in NARs	Eelco Dolstra	1	-1/+6

2014-07-16	Handle case collisions on case-insensitive systems	Eelco Dolstra	6	-76/+122
	When running NixOps under Mac OS X, we need to be able to import store paths built on Linux into the local Nix store. However, HFS+ is usually case-insensitive, so if there are directories with file names that differ only in case, then importing will fail. The solution is to add a suffix ("~nix~case~hack~<integer>") to colliding files. For instance, if we have a directory containing xt_CONNMARK.h and xt_connmark.h, then the latter will be renamed to "xt_connmark.h~nix~case~hack~1". If a store path is dumped as a NAR, the suffixes are removed. Thus, importing and exporting via a case-insensitive Nix store is round-tripping. So when NixOps calls nix-copy-closure to copy the path to a Linux machine, you get the original file names back. Closes #119.
2014-07-16	Make dev-shell script work on Darwin	Eelco Dolstra	1	-2/+3

2014-07-16	Install systemd and Upstart stuff only on Linux	Eelco Dolstra	2	-2/+10

2014-07-16	Pass *_proxy vars to bootstrap fetchurl	Shea Levy	1	-0/+8

2014-07-16	Manual: Typo	Eelco Dolstra	1	-1/+1

2014-07-14	Remove cruft	Eelco Dolstra	2	-127/+0

2014-07-14	build-remote.pl: Fix building multiple output derivations	Eelco Dolstra	3	-4/+6
	We were importing paths without sorting them topologically, leading to "path is not valid" errors. See e.g. http://hydra.nixos.org/build/12451761
2014-07-12	build-remote.pl: Don't keep a shell process around	Eelco Dolstra	1	-1/+2

2014-07-12	build-remote.pl: Fix build log	Eelco Dolstra	2	-3/+5

2014-07-11	Fix test	Eelco Dolstra	1	-1/+1

2014-07-11	build-remote.pl: Use ‘nix-store --serve’ on the remote side	Eelco Dolstra	9	-115/+185
	This makes things more efficient (we don't need to use an SSH master connection, and we only start a single remote process) and gets rid of locking issues (the remote nix-store process will keep inputs and outputs locked as long as they're needed). It also makes it more or less secure to connect directly to the root account on the build machine, using a forced command (e.g. ‘command="nix-store --serve --write"’). This bypasses the Nix daemon and is therefore more efficient. Also, don't call nix-store to import the output paths.
2014-07-11	Fix closure size display	Eelco Dolstra	1	-1/+1

2014-07-11	Allow $NIX_BUILD_HOOK to be relative to Nix libexec directory	Eelco Dolstra	2	-3/+5

2014-07-10	Fix broken Pid constructor	Eelco Dolstra	1	-5/+2

2014-07-10	Replace message "importing path <...>" with "exporting path <...>"	Eelco Dolstra	1	-2/+2
	This causes nix-copy-closure to show what it's doing before rather than after.
2014-07-10	Fix use of sysread	Eelco Dolstra	1	-6/+18

2014-07-10	nix-copy-closure -s: Do substitutions via ‘nix-store --serve’	Eelco Dolstra	4	-11/+43
	This means we no longer need an SSH master connection, since we only execute a single command on the remote host.
2014-07-10	nix-copy-closure: Fix --dry-run	Eelco Dolstra	1	-8/+8

2014-07-10	Remove tabs	Eelco Dolstra	1	-3/+3

2014-07-10	Refactoring: Move all fork handling into a higher-order function	Eelco Dolstra	7	-206/+128
	C++11 lambdas ftw.
2014-07-10	nix-copy-closure: Restore compression and the progress viewer	Eelco Dolstra	3	-16/+89

2014-07-10	Remove maybeVfork	Eelco Dolstra	5	-18/+4

2014-07-10	nix-copy-closure: Fix race condition	Eelco Dolstra	4	-60/+136
	There is a long-standing race condition when copying a closure to a remote machine, particularly affecting build-remote.pl: the client first asks the remote machine which paths it already has, then copies over the missing paths. If the garbage collector kicks in on the remote machine between the first and second step, the already-present paths may be deleted. The missing paths may then refer to deleted paths, causing nix-copy-closure to fail. The client now performs both steps using a single remote Nix call (using ‘nix-store --serve’), locking all paths in the closure while querying. I changed the --serve protocol a bit (getting rid of QueryCommand), so this breaks the SSH substituter from older versions. But it was marked experimental anyway. Fixes #141.
2014-07-10	Fix security hole in ‘nix-store --serve’	Eelco Dolstra	1	-1/+1
	Since it didn't check that the path received from the client is a store path, the client could dump any path in the file system.