Demi Marie Obenour <demiobenour@gmail.com> writes:
On 10/26/25 05:09, Alyssa Ross wrote:
Yureka Lilian <yureka@cyberchaos.dev> writes:
Signed-off-by: Yureka Lilian <yureka@cyberchaos.dev> --- lib/nixpkgs.default.nix | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
I'm seeing a timeout in the networking test in release/checks/integration with this patch applied. I'll try bisecting, but I get the impression you might have the resources to be able to do that faster, if you want to try to beat me. I wonder if this is not a new bug, but rather a pre-existing race being exposed. I have run into this before on x86.
I think it would be better to actually figure out what is happening. One option would be allow the person running the tests to break into a shell.
Interesting — I haven't. It should be possible to get a shell by entering the nix-shell for release/checks/integration, running the test with meson test -v networking, and then running vm-console in the graphical QEMU that spawns. I'd be a bit surprised if this turns out to be the same issue, since this happens consistently, but I guess not /that/ surprised.