You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I somehow interrupted a precompilation process, resulting in a stale pidlock somewhere in my depot. This causes precompilation to hang, which for all intents and purposes looks like a plain hang (IIUC JuliaLang/Pkg.jl#3519 wouldn't help here, because this is triggered by using Package in code, not by Pkg.precompile()).
Setting JULIA_DEBUG shows that its stuck waiting for a pidlock, as does the backtrace when interrupting:
┌ Debug: Waiting for another process (pid: 5719) to finish precompiling CUDA [052768ef-5323-5732-b1bb-66c8b64840ba]
└ @ Base loading.jl:2831
^C
[7169] signal (2): Interrupt
in expression starting at /home/tim/Julia/pkg/CUDA/wip.jl:1
epoll_wait at /usr/lib/libc.so.6 (unknown line)
uv__io_poll at /workspace/srcdir/libuv/src/unix/epoll.c:236
uv_run at /workspace/srcdir/libuv/src/unix/core.c:400
ijl_task_get_next at /home/tim/Julia/src/julia/src/partr.c:436
unknown function (ip: 0x7f59f774c93e)
unknown function (ip: 0x7f59f75d646c)
unknown function (ip: 0x7f59f77a27c5)
unknown function (ip: 0x7f59f704af20)
unknown function (ip: 0x7f59f6fa0c03)
watch_file at /home/tim/Julia/src/julia/build/dev/usr/share/julia/stdlib/v1.10/FileWatching/src/FileWatching.jl:778 [inlined]
JuliaLang/Pkg.jl#13 at /home/tim/Julia/src/julia/build/dev/usr/share/julia/stdlib/v1.10/FileWatching/src/pidfile.jl:257
unknown function (ip: 0x7f5a0d07daf9)
_jl_invoke at /home/tim/Julia/src/julia/src/gf.c:2870 [inlined]
ijl_apply_generic at /home/tim/Julia/src/julia/src/gf.c:3071
jl_apply at /home/tim/Julia/src/julia/src/julia.h:1963 [inlined]
start_task at /home/tim/Julia/src/julia/src/task.c:1238
unknown function (ip: (nil))
Allocations: 2836 (Pool: 2827; Big: 9); GC: 0
Crucially, process 5719 just doesn't exist anymore. Instead of waiting for the pidlock to go stale, which seems like ages (it's been almost 10 minutes), shouldn't we first check if the process even exists and if not break the lock?
The text was updated successfully, but these errors were encountered:
I somehow interrupted a precompilation process, resulting in a stale pidlock somewhere in my depot. This causes precompilation to hang, which for all intents and purposes looks like a plain hang (IIUC JuliaLang/Pkg.jl#3519 wouldn't help here, because this is triggered by
using Package
in code, not byPkg.precompile()
).Setting JULIA_DEBUG shows that its stuck waiting for a pidlock, as does the backtrace when interrupting:
Crucially, process 5719 just doesn't exist anymore. Instead of waiting for the pidlock to go stale, which seems like ages (it's been almost 10 minutes), shouldn't we first check if the process even exists and if not break the lock?
The text was updated successfully, but these errors were encountered: