fdo-mirrors/mesa
1
0
Fork 0
mirror of https://gitlab.freedesktop.org/mesa/mesa.git synced 2026-01-03 13:40:11 +01:00
Code Issues Projects Releases Packages Wiki Activity Actions 1

venus: fix to ensure sfb cmds can get recycled

Browse source 
The prior refactor has missed a case that timeline sempahore can be
ping-pong'ed between device signal and host wait.

Fixes: d63432012d ("venus: refactor semaphore feedback")
Signed-off-by: Yiwei Zhang <zzyiwei@chromium.org>
Reported-by: Juston Li <justonli@google.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27807>
This commit is contained in:
Yiwei Zhang 2024-02-26 17:34:54 -08:00 • committed by Marge Bot
parent 90e38bbb3b
commit e9290ec0bb
1 changed files with 22 additions and 31 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

53
src/virtio/vulkan/vn_queue.c
View file

@ -760,10 +760,6 @@ vn_queue_submission_setup_batches(struct vn_queue_submission *submit)
return VK_SUCCESS;
}
static void
vn_semaphore_recycle_feedback_cmd(VkDevice dev_handle,
VkSemaphore sem_handle);
static void
vn_queue_submission_cleanup_semaphore_feedback(
struct vn_queue_submission *submit)
@ -775,7 +771,13 @@ vn_queue_submission_cleanup_semaphore_feedback(
const uint32_t wait_count = vn_get_wait_semaphore_count(submit, i);
for (uint32_t j = 0; j < wait_count; j++) {
VkSemaphore sem_handle = vn_get_wait_semaphore(submit, i, j);
vn_semaphore_recycle_feedback_cmd(dev_handle, sem_handle);
struct vn_semaphore *sem = vn_semaphore_from_handle(sem_handle);
if (!sem->feedback.slot)
continue;
/* sfb pending cmds are recycled when signaled counter is updated */
uint64_t counter = 0;
vn_GetSemaphoreCounterValue(dev_handle, sem_handle, &counter);
}
}
}
@ -1791,26 +1793,6 @@ vn_semaphore_get_feedback_cmd(struct vn_device *dev, struct vn_semaphore *sem)
return sfb_cmd;
}
static void
vn_semaphore_recycle_feedback_cmd(VkDevice dev_handle, VkSemaphore sem_handle)
{
struct vn_semaphore *sem = vn_semaphore_from_handle(sem_handle);
if (!sem->feedback.slot)
return;
uint64_t curr_counter = 0;
vn_GetSemaphoreCounterValue(dev_handle, sem_handle, &curr_counter);
/* search pending cmds for already signaled values */
simple_mtx_lock(&sem->feedback.cmd_mtx);
list_for_each_entry_safe(struct vn_semaphore_feedback_cmd, sfb_cmd,
&sem->feedback.pending_cmds, head) {
if (curr_counter >= vn_feedback_get_counter(sfb_cmd->src_slot))
list_move_to(&sfb_cmd->head, &sem->feedback.free_cmds);
}
simple_mtx_unlock(&sem->feedback.cmd_mtx);
}
static VkResult
vn_semaphore_feedback_init(struct vn_device *dev,
struct vn_semaphore *sem,
@ -1965,10 +1947,8 @@ vn_GetSemaphoreCounterValue(VkDevice device,
if (sem->feedback.slot) {
simple_mtx_lock(&sem->feedback.async_wait_mtx);
*pValue = vn_feedback_get_counter(sem->feedback.slot);
if (sem->feedback.signaled_counter < *pValue) {
const uint64_t counter = vn_feedback_get_counter(sem->feedback.slot);
if (sem->feedback.signaled_counter < counter) {
/* When the timeline semaphore feedback slot gets signaled, the real
* semaphore signal operation follows after but the signaling isr can
* be deferred or preempted. To avoid racing, we let the renderer
@ -1990,15 +1970,26 @@ vn_GetSemaphoreCounterValue(VkDevice device,
.flags = 0,
.semaphoreCount = 1,
.pSemaphores = &semaphore,
.pValues = pValue,
.pValues = &counter,
};
vn_async_vkWaitSemaphores(dev->primary_ring, device, &wait_info,
UINT64_MAX);
sem->feedback.signaled_counter = *pValue;
/* search pending cmds for already signaled values */
simple_mtx_lock(&sem->feedback.cmd_mtx);
list_for_each_entry_safe(struct vn_semaphore_feedback_cmd, sfb_cmd,
&sem->feedback.pending_cmds, head) {
if (counter >= vn_feedback_get_counter(sfb_cmd->src_slot))
list_move_to(&sfb_cmd->head, &sem->feedback.free_cmds);
}
simple_mtx_unlock(&sem->feedback.cmd_mtx);
sem->feedback.signaled_counter = counter;
}
simple_mtx_unlock(&sem->feedback.async_wait_mtx);
*pValue = counter;
return VK_SUCCESS;
} else {
return vn_call_vkGetSemaphoreCounterValue(dev->primary_ring, device,