Fix SIGUSR1/SIGUSR2 handler clobbering breaking JVM processes#160
Open
charford wants to merge 1 commit intoProject-HAMi:mainfrom
Open
Fix SIGUSR1/SIGUSR2 handler clobbering breaking JVM processes#160charford wants to merge 1 commit intoProject-HAMi:mainfrom
charford wants to merge 1 commit intoProject-HAMi:mainfrom
Conversation
Contributor
|
[APPROVALNOTIFIER] This PR is NOT APPROVED This pull-request has been approved by: charford The full list of commands accepted by this bot can be found here. DetailsNeeds approval from an approver in each of these files:Approvers can indicate their approval by writing |
Contributor
|
Welcome @charford! It looks like this is your first PR to Project-HAMi/HAMi-core 🎉 |
signal() overwrote SIGUSR1 and SIGUSR2 without saving previous handlers, causing JVM crashes (SIGSEGV in Monitor::wait) since the JVM uses these signals internally for GC safepoints and thread management. Replace signal() with sigaction() to save the old handlers, and chain to them in the stubs so other runtimes (JVM, etc.) can still process the signals correctly. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com> Signed-off-by: Casey Harford <casey@caseyharford.com>
charford
force-pushed
the
fix-sigusr-signal-handler-chaining
branch
from
3413af0 to
bf1e307
Compare
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Problem
libvgpu.soregisters signal handlers forSIGUSR1andSIGUSR2usingsignal(), which overwrites any previously installed handlers without saving them. This causes JVM processes to crash withSIGSEGVinMonitor::wait()because the JVM usesSIGUSR1/SIGUSR2internally for GC safepoints and thread management.Observed on HAMi volcano-vgpu nodes (hami-core mode) when running PyTorch jobs with a JVM component — the crash occurs at startup before CUDA initializes.
Additional risks from the current implementation:
libvgpu.sointerceptsdlsym(), which the JVM also uses for native library loadingENSURE_RUNNING()spin loop can cause a deadlock if a Java thread holding a JVM monitor gets suspendedFix
Replace
signal()calls ininit_proc_slot_withlock()withsigaction(), saving the previous handlers into staticstruct sigactionvariables and chaining to them insig_swap_stub/sig_restore_stub. This ensures other runtimes (JVM, Python, etc.) can still processSIGUSR1/SIGUSR2correctly.Changes
src/multiprocess/multiprocess_memory_limit.c: Replacesignal()withsigaction(), chain to previous handlersTesting
SIGSEGVinMonitor::wait()when running underlibvgpu.soinjection on HAMi volcano-vgpu nodes (prod-ltx1-k8s-1, prod-ltx1-k8s-2)Closes #161