sdgoij/llama.cpp
1
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2026-05-10 11:04:06 +00:00
Code Issues Packages Projects Releases Wiki Activity

state

Browse Source
This commit is contained in:
Ruben Ortlam
2026-04-09 12:51:43 +02:00
parent 88335c0490
commit 9f001cae27
1 changed files with 21 additions and 4 deletions
Download Patch File Download Diff File Expand all files Collapse all files

25
ggml/src/ggml-vulkan/ggml-vulkan.cpp
View File

@@ -13404,14 +13404,31 @@ static void ggml_vk_graph_cleanup(ggml_backend_vk_context * ctx) {
ggml_vk_queue_command_pools_cleanup(ctx->device);
}
GGML_LOG_DEBUG("ggml_vulkan: graph_cleanup(%s) binary_semaphore_idx=%zu\n",
ctx->name.c_str(), ctx->binary_semaphore_idx);
// Destroy binary semaphores rather than pooling — drivers may hold
// imported sync_fd handles until the semaphore is destroyed.
GGML_LOG_DEBUG("ggml_vulkan: graph_cleanup(%s) binary_semaphore_idx=%zu gc.semaphores=%zu\n",
ctx->name.c_str(), ctx->binary_semaphore_idx, ctx->gc.semaphores.size());
#if !defined(_WIN32)
if (!ctx->gc.semaphores.empty()) {
auto count_fds = []() -> int {
int n = 0;
DIR * d = opendir("/proc/self/fd");
if (d) { while (readdir(d)) n++; closedir(d); n -= 3; }
return n;
};
int fds_before = count_fds();
for (size_t i = 0; i < ctx->gc.semaphores.size(); i++) {
ctx->device->device.destroySemaphore(ctx->gc.semaphores[i].s);
}
int fds_after = count_fds();
GGML_LOG_DEBUG("ggml_vulkan: graph_cleanup destroyed %zu binary sems: fds %d -> %d (delta=%d)\n",
ctx->gc.semaphores.size(), fds_before, fds_after, fds_after - fds_before);
ctx->gc.semaphores.clear();
}
#else
for (size_t i = 0; i < ctx->gc.semaphores.size(); i++) {
ctx->device->device.destroySemaphore(ctx->gc.semaphores[i].s);
}
ctx->gc.semaphores.clear();
#endif
ctx->binary_semaphore_idx = 0;
for (auto& [peer, staging] : ctx->device->peer_staging) {