Skip to content

Fix hang issue in logging submodule #625

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
Binyang2014 merged 1 commit into from
Sep 5, 2025
Merged

Fix hang issue in logging submodule #625

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
13 changes: 5 additions & 8 deletions src/debug.cc
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -12,6 +12,7 @@
#include <mscclpp/env.hpp>
#include <mscclpp/gpu_utils.hpp>
#include <mscclpp/utils.hpp>
#include <mutex>
#include <string>

int mscclppDebugLevel = -1;
Expand All @@ -22,17 +23,16 @@ char mscclppLastError[1024] = ""; // Global string for the last error i
uint64_t mscclppDebugMask = MSCCLPP_INIT; // Default debug sub-system mask is INIT
FILE* mscclppDebugFile = stdout;
mscclppLogHandler_t mscclppDebugLogHandler = NULL;
pthread_mutex_t mscclppDebugLock = PTHREAD_MUTEX_INITIALIZER;
std::recursive_mutex mscclppDebugMutex;
std::chrono::steady_clock::time_point mscclppEpoch;

static __thread int tid = -1;

void mscclppDebugDefaultLogHandler(const char* msg) { fwrite(msg, 1, strlen(msg), mscclppDebugFile); }

void mscclppDebugInit() {
pthread_mutex_lock(&mscclppDebugLock);
std::lock_guard<std::recursive_mutex> lock(mscclppDebugMutex);
if (mscclppDebugLevel != -1) {
pthread_mutex_unlock(&mscclppDebugLock);
return;
}
const char* mscclpp_debug = mscclpp::env()->debug.c_str();
Expand Down Expand Up @@ -152,7 +152,6 @@ void mscclppDebugInit() {

mscclppEpoch = std::chrono::steady_clock::now();
__atomic_store_n(&mscclppDebugLevel, tempNcclDebugLevel, __ATOMIC_RELEASE);
pthread_mutex_unlock(&mscclppDebugLock);
}

/* Common logging function used by the INFO, WARN and TRACE macros
Expand All @@ -168,12 +167,11 @@ void mscclppDebugLog(mscclppDebugLogLevel level, unsigned long flags, const char
}
// Save the last error (WARN) as a human readable string
if (level == MSCCLPP_LOG_WARN) {
pthread_mutex_lock(&mscclppDebugLock);
std::lock_guard<std::recursive_mutex> lock(mscclppDebugMutex);
va_list vargs;
va_start(vargs, fmt);
(void)vsnprintf(mscclppLastError, sizeof(mscclppLastError), fmt, vargs);
va_end(vargs);
pthread_mutex_unlock(&mscclppDebugLock);
}
if (mscclppDebugLevel < level || ((flags & mscclppDebugMask) == 0)) return;

Expand Down Expand Up @@ -219,9 +217,8 @@ void mscclppDebugLog(mscclppDebugLogLevel level, unsigned long flags, const char
mscclppResult_t mscclppDebugSetLogHandler(mscclppLogHandler_t handler) {
if (__atomic_load_n(&mscclppDebugLevel, __ATOMIC_ACQUIRE) == -1) mscclppDebugInit();
if (handler == NULL) return mscclppInvalidArgument;
pthread_mutex_lock(&mscclppDebugLock);
std::lock_guard<std::recursive_mutex> lock(mscclppDebugMutex);
mscclppDebugLogHandler = handler;
pthread_mutex_unlock(&mscclppDebugLock);
return mscclppSuccess;
}

Expand Down
Loading