anv: Implement VK_KHR_performance_query

This has the same kernel requirements are VK_INTEL_performance_query

v2: Fix empty queue submit (Lionel)

v3: Fix autotool build issue (Piotr Byszewski)

v4: Fix Reset & Begin/End in same command buffer, using soft-pin &
    relocation on the same buffer won't work currently. This version
    uses a somewhat dirty trick in anv_execbuf_add_bo (Piotr Byszewski)

v5: Fix enumeration with null pointers for either pCounters or
    pCounterDescriptions (Piotr)
    Fix return condition on enumeration (Lionel)
    Set counter uuid using sha1 hashes (Lionel)

v6: Fix counters scope, should be COMMAND_KHR not COMMAND_BUFFER_KHR (Lionel)

v7: Rebase (Lionel)

v8: Rework checking for loaded queries (Lionel)

v9: Use new i915-perf interface

v10: Use anv_multialloc (Jason)

v11: Implement perf query passes using self modifying batches (Lionel)
     Limit support to softpin/gen8

v12: Remove spurious changes (Jason)

v13: Drop relocs (Jason)

v14: Avoid overwritting .sType in
     VkPerformanceCounterKHR/VkPerformanceCounterDescriptionKHR (Lionel)

v15: Don't copy the entire
     VkPerformanceCounterKHR/VkPerformanceCounterDescriptionKHR (Jason)
     Reuse anv_batch rather than custom packing (Jason)

v16: Fix missing MI_BB_END in reconfiguration batch
     Only report the extension with kernel support (perf_version >= 3)

v17: Some cleanup of unused stuff

Signed-off-by: Lionel Landwerlin <[email protected]>
Reviewed-by: Jason Ekstrand <[email protected]>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/2775>

1 files changed, 14 insertions, 7 deletions

diff --git a/src/intel/vulkan/anv_queue.c b/src/intel/vulkan/anv_queue.c
index 009675e23ea..f6e3fdd6177 100644
--- a/src/intel/vulkan/anv_queue.c
+++ b/src/intel/vulkan/anv_queue.c
@@ -544,7 +544,7 @@ anv_queue_submit_add_timeline_signal(struct anv_queue_submit* submit,
 }
 
 static struct anv_queue_submit *
-anv_queue_submit_alloc(struct anv_device *device)
+anv_queue_submit_alloc(struct anv_device *device, int perf_query_pass)
 {
    const VkAllocationCallbacks *alloc = &device->vk.alloc;
    VkSystemAllocationScope alloc_scope = VK_SYSTEM_ALLOCATION_SCOPE_DEVICE;
@@ -557,6 +557,7 @@ anv_queue_submit_alloc(struct anv_device *device)
    submit->alloc_scope = alloc_scope;
    submit->in_fence = -1;
    submit->out_fence = -1;
+   submit->perf_query_pass = perf_query_pass;
 
    return submit;
 }
@@ -569,7 +570,7 @@ anv_queue_submit_simple_batch(struct anv_queue *queue,
       return VK_SUCCESS;
 
    struct anv_device *device = queue->device;
-   struct anv_queue_submit *submit = anv_queue_submit_alloc(device);
+   struct anv_queue_submit *submit = anv_queue_submit_alloc(device, -1);
    if (!submit)
       return vk_error(VK_ERROR_OUT_OF_HOST_MEMORY);
 
@@ -720,12 +721,13 @@ anv_queue_submit(struct anv_queue *queue,
                  const uint64_t *out_values,
                  uint32_t num_out_semaphores,
                  struct anv_bo *wsi_signal_bo,
-                 VkFence _fence)
+                 VkFence _fence,
+                 int perf_query_pass)
 {
    ANV_FROM_HANDLE(anv_fence, fence, _fence);
    struct anv_device *device = queue->device;
    UNUSED struct anv_physical_device *pdevice = device->physical;
-   struct anv_queue_submit *submit = anv_queue_submit_alloc(device);
+   struct anv_queue_submit *submit = anv_queue_submit_alloc(device, perf_query_pass);
    if (!submit)
       return vk_error(VK_ERROR_OUT_OF_HOST_MEMORY);
 
@@ -972,7 +974,7 @@ VkResult anv_QueueSubmit(
        * common case.
        */
       result = anv_queue_submit(queue, NULL, NULL, NULL, 0, NULL, NULL, 0,
-                                NULL, fence);
+                                NULL, fence, -1);
       goto out;
    }
 
@@ -990,6 +992,9 @@ VkResult anv_QueueSubmit(
       const VkTimelineSemaphoreSubmitInfoKHR *timeline_info =
          vk_find_struct_const(pSubmits[i].pNext,
                               TIMELINE_SEMAPHORE_SUBMIT_INFO_KHR);
+      const VkPerformanceQuerySubmitInfoKHR *perf_info =
+         vk_find_struct_const(pSubmits[i].pNext,
+                              PERFORMANCE_QUERY_SUBMIT_INFO_KHR);
       const uint64_t *wait_values =
          timeline_info && timeline_info->waitSemaphoreValueCount ?
          timeline_info->pWaitSemaphoreValues : NULL;
@@ -1011,7 +1016,8 @@ VkResult anv_QueueSubmit(
                                    signal_values,
                                    pSubmits[i].signalSemaphoreCount,
                                    wsi_signal_bo,
-                                   submit_fence);
+                                   submit_fence,
+                                   -1);
          if (result != VK_SUCCESS)
             goto out;
 
@@ -1049,7 +1055,8 @@ VkResult anv_QueueSubmit(
          result = anv_queue_submit(queue, cmd_buffer,
                                    in_semaphores, in_values, num_in_semaphores,
                                    out_semaphores, out_values, num_out_semaphores,
-                                   wsi_signal_bo, execbuf_fence);
+                                   wsi_signal_bo, execbuf_fence,
+                                   perf_info ? perf_info->counterPassIndex : 0);
          if (result != VK_SUCCESS)
             goto out;
       }