check #46301
ydshieh
wants to merge
30 commits into
main
from
debug_fix_more_read_only_cache_issues
check
9f10f944
check
bef84bb9
check
95e78121
check
bd3709c5
tests: skip SIGABRT after destroy_process_group:done as gloo shutdown…
30767aae
tests: richer SIGABRT diagnostics — thread snapshot + ctypes C backtrace
fd846a28
tests: restore faulthandler for SIGABRT, add glibc/OS/torch env to br…
6e3e9c4d
try again
6618391d
tests: add torch.utils.collect_env to SIGABRT diagnostics
3d8e1576
try again
0573c2db
check
36febf9f
check
361441c7
check both
a557c120
try again
6bfed662
try with shared cache
a255d1a7
check pytest -n 1
c3b5ce7e
try with n1
48174d9a
check pytest -n 1
48d4884c
try with n1
358ca156
check wait
28bfbff1
check without wait
816fea25
check without torch job
23013af9
check run-slow
73e4d6a0
check torch.set_num_threads(1)
43a3abe7
check torch.set_num_threads(1)
0e2bd32c
check torch.set_num_threads(1)
0b4fe240
add richer diagnostics to TP/EP distributed test crash handler
0f562ff2
Revert "add richer diagnostics to TP/EP distributed test crash handler"
97712101
fix SIGABRT in TP/EP tests by forcing gc.collect() after destroy_proc…
ec79b167
Revert "fix SIGABRT in TP/EP tests by forcing gc.collect() after dest…
f2462bbc
Reviewers
No reviews
Assignees
No one assigned
Labels
None yet
Milestone
No milestone
Login to write a write a comment.
Login via GitHub