Lumpiasty/llama.cpp
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

vulkan: when using transfer queue for async copies, sync on event_wait to avoid race (#25229)

Browse Source
This commit is contained in:
Ruben Ortlam
2026-07-16 15:34:24 +02:00
committed by GitHub
parent 3278e921b1
commit b15ca938ad
1 changed files with 21 additions and 14 deletions
Download Patch File Download Diff File Expand all files Collapse all files
ggml/src/ggml-vulkan/ggml-vulkan.cpp
+21 -14
View File
@@ -7829,6 +7829,20 @@ static vk_context ggml_vk_get_compute_ctx(ggml_backend_vk_context * ctx) {
return result; return result;
} }
static vk_context ggml_vk_get_transfer_ctx(ggml_backend_vk_context * ctx) {
vk_context result;
if (!ctx->transfer_ctx.expired()) {
result = ctx->transfer_ctx.lock();
} else {
result = ggml_vk_create_context(ctx, ctx->transfer_cmd_pool);
ctx->transfer_ctx = result;
ggml_vk_ctx_begin(ctx->device, result);
}
return result;
}
// Submit any pending transfer queue work and signal the transfer semaphore. // Submit any pending transfer queue work and signal the transfer semaphore.
// The next compute context created via ggml_vk_get_compute_ctx will wait on this semaphore. // The next compute context created via ggml_vk_get_compute_ctx will wait on this semaphore.
// Returns true if work was submitted. // Returns true if work was submitted.
@@ -15633,13 +15647,7 @@ static void ggml_backend_vk_set_tensor_2d_async(ggml_backend_t backend, ggml_ten
vk_context cpy_ctx; vk_context cpy_ctx;
if (ctx->device->async_use_transfer_queue) { if (ctx->device->async_use_transfer_queue) {
if (ctx->transfer_ctx.expired()) { cpy_ctx = ggml_vk_get_transfer_ctx(ctx);
cpy_ctx = ggml_vk_create_context(ctx, ctx->transfer_cmd_pool);
ctx->transfer_ctx = cpy_ctx;
ggml_vk_ctx_begin(ctx->device, cpy_ctx);
} else {
cpy_ctx = ctx->transfer_ctx.lock();
}
} else { } else {
cpy_ctx = ggml_vk_get_compute_ctx(ctx); cpy_ctx = ggml_vk_get_compute_ctx(ctx);
} }
@@ -15785,13 +15793,7 @@ static bool ggml_backend_vk_cpy_tensor_async(ggml_backend_t backend_src, ggml_ba
vk_context cpy_ctx; vk_context cpy_ctx;
if (ctx->device->async_use_transfer_queue) { if (ctx->device->async_use_transfer_queue) {
if (ctx->transfer_ctx.expired()) { cpy_ctx = ggml_vk_get_transfer_ctx(ctx);
cpy_ctx = ggml_vk_create_context(ctx, ctx->transfer_cmd_pool);
ctx->transfer_ctx = cpy_ctx;
ggml_vk_ctx_begin(ctx->device, cpy_ctx);
} else {
cpy_ctx = ctx->transfer_ctx.lock();
}
} else { } else {
cpy_ctx = ggml_vk_get_compute_ctx(ctx); cpy_ctx = ggml_vk_get_compute_ctx(ctx);
} }
@@ -17123,6 +17125,11 @@ static void ggml_backend_vk_event_wait(ggml_backend_t backend, ggml_backend_even
if (vkev->has_event) { if (vkev->has_event) {
// Wait for latest event // Wait for latest event
ggml_vk_wait_events(compute_ctx, { vkev->event }); ggml_vk_wait_events(compute_ctx, { vkev->event });
if (ctx->device->async_use_transfer_queue) {
vk_context transfer_ctx = ggml_vk_get_transfer_ctx(ctx);
transfer_ctx->s->wait_semaphores.push_back(vkev->tl_semaphore);
}
} }
} }