!lymvtcwDJ7ZA9Npq:lix.systems

Lix Development

415 Members
(Technical) development of Lix, the package manager, a Nix implementation. Please be mindful of ongoing technical conversations in this channel.140 Servers

You have reached the beginning of time (for this room).


SenderMessageTime
12 Nov 2025
@winter:catgirl.cloudWinter...'t die on sigpipe and groaning20:45:23
@raitobezarius:matrix.orgraitobezarius i think there's a bunch of CL improvements from aloisw on n-e-j 21:56:52
13 Nov 2025
@aloisw:julia0815.dealoisw SIGPIPE is not in it though. Also it seems like the entirety of Lix just ignores it completely. I think the best way forward would not be to special-case it in nix-eval-jobs, but to treat as an interrupt like SIGTERM and SIGHUP. 05:59:10
@winter:catgirl.cloudWinteri believe i only looked at nix but it appears that they’re not explicitly ignoring it at least, instead just saying to use the OS’ default signal handler for it?06:15:06
@winter:catgirl.cloudWinterand yeah, agreed06:15:10
@winter:catgirl.cloudWinterguess i could write a patch for that pretty easily06:15:40
@winter:catgirl.cloudWinteronce i wrap my head around the current state of affairs described above06:15:49
@aloisw:julia0815.dealoisw
In reply to @winter:catgirl.cloud
i believe i only looked at nix but it appears that they’re not explicitly ignoring it at least, instead just saying to use the OS’ default signal handler for it?
https://git.lix.systems/lix-project/lix/src/commit/78bdc60c11031d2293b2e77a0e44140ff7b88f12/lix/libutil/signals.cc#L189-L197
07:25:58
@aloisw:julia0815.dealoisw The solution I imagined above was just adding SIGPIPE here: https://git.lix.systems/lix-project/lix/src/commit/78bdc60c11031d2293b2e77a0e44140ff7b88f12/lix/libutil/signals.cc#L119 07:26:54
@aloisw:julia0815.dealoisw Of course that's not going to work because SIGPIPE is only sent to the writing thread. 11:45:10
@alicia:catgirl.cloudAlicia joined the room.22:44:56
15 Nov 2025
@347online:matrix.orgKatie joined the room.18:22:52
16 Nov 2025
@grimmauld:grapevine.grimmauld.deGrimmauld (any/all) Quick heads-up:
I believe https://git.lix.systems/lix-project/lix/src/commit/e101400359558ec89f0a3afa29631ca1b6546fd3/lix/libutil/processes.cc#L20-L22 should be unconditional. On glibc gcc this header is included implicitly. On musl llvm, that include is needed or otherwise SYS_capset (https://git.lix.systems/lix-project/lix/src/commit/243836ed0a0b963eebd033d109e6cb817d67fde4/lix/libutil/processes.cc#L414) is undefined.
08:49:40
@xokdvium:matrix.orgSergei Zimmerman (xokdvium)

On my dev machine I'm getting:

+(tarball.sh:31) nix-build -o /tmp/nix-test/tarball/result -E 'import (fetchTarball { url = "file:///does-not-exist/must-remain-unused//tmp/nix-test/tarball/tarball.tar"; sha256 = "sha256-PRgXPIzueCAIFHesj5pSLlDYbJsrGmUD8B9gpbs9p9Q="; })'
++(tarball.sh:36) mktemp -d
+(tarball.sh:36) export HOME=/tmp/tmp.0S4oIlI3Xo
+(tarball.sh:36) HOME=/tmp/tmp.0S4oIlI3Xo
tarball.sh: line 37: TMPDIR: unbound variable

Seems like function tests need to provide a sensible default for TMPDIR if it's not set

23:49:08
@xokdvium:matrix.orgSergei Zimmerman (xokdvium) *

On my dev machine I'm getting:

+(tarball.sh:31) nix-build -o /tmp/nix-test/tarball/result -E 'import (fetchTarball { url = "file:///does-not-exist/must-remain-unused//tmp/nix-test/tarball/tarball.tar"; sha256 = "sha256-PRgXPIzueCAIFHesj5pSLlDYbJsrGmUD8B9gpbs9p9Q="; })'
++(tarball.sh:36) mktemp -d
+(tarball.sh:36) export HOME=/tmp/tmp.0S4oIlI3Xo
+(tarball.sh:36) HOME=/tmp/tmp.0S4oIlI3Xo
tarball.sh: line 37: TMPDIR: unbound variable

Seems like functional tests need to provide a sensible default for TMPDIR if it's not set

23:49:27
17 Nov 2025
@raitobezarius:matrix.orgraitobezariusCan you open an issue with this trace so I don't forget about it? Thanks!00:03:55
@xokdvium:matrix.orgSergei Zimmerman (xokdvium)https://git.lix.systems/lix-project/lix/issues/103500:09:21

Show newer messages


Back to Room ListRoom Version: 10