anv: Add support for the PMA fix on Broadwell

This helps Dota 2 on Broadwell by 8-9%.  I also hacked up the driver and
used the Sascha "shadowmapping" demo to get some results.  Setting
uses_kill to true dropped the framerate on the demo by 25-30%.  Enabling
the PMA fix brought it back up to around 90% of the original framerate.

Reviewed-by: Lionel Landwerlin <[email protected]>
Reviewed-by: Nanley Chery <[email protected]>

1 files changed, 14 insertions, 1 deletions

diff --git a/src/intel/vulkan/genX_cmd_buffer.c b/src/intel/vulkan/genX_cmd_buffer.c
index 14338b22ece..40a72f4d141 100644
--- a/src/intel/vulkan/genX_cmd_buffer.c
+++ b/src/intel/vulkan/genX_cmd_buffer.c
@@ -637,6 +637,11 @@ genX(EndCommandBuffer)(
 {
    ANV_FROM_HANDLE(anv_cmd_buffer, cmd_buffer, commandBuffer);
 
+   /* We want every command buffer to start with the PMA fix in a known state,
+    * so we disable it at the end of the command buffer.
+    */
+   genX(cmd_buffer_enable_pma_fix)(cmd_buffer, false);
+
    genX(cmd_buffer_apply_pipe_flushes)(cmd_buffer);
 
    anv_cmd_buffer_end_batch_buffer(cmd_buffer);
@@ -654,6 +659,11 @@ genX(CmdExecuteCommands)(
 
    assert(primary->level == VK_COMMAND_BUFFER_LEVEL_PRIMARY);
 
+   /* The secondary command buffers will assume that the PMA fix is disabled
+    * when they begin executing.  Make sure this is true.
+    */
+   genX(cmd_buffer_enable_pma_fix)(primary, false);
+
    for (uint32_t i = 0; i < commandBufferCount; i++) {
       ANV_FROM_HANDLE(anv_cmd_buffer, secondary, pCmdBuffers[i]);
 
@@ -2227,7 +2237,8 @@ cmd_buffer_emit_depth_stencil(struct anv_cmd_buffer *cmd_buffer)
    const bool has_stencil =
       image && (image->aspects & VK_IMAGE_ASPECT_STENCIL_BIT);
 
-   /* FIXME: Implement the PMA stall W/A */
+   cmd_buffer->state.hiz_enabled = has_hiz;
+
    /* FIXME: Width and Height are wrong */
 
    genX(cmd_buffer_emit_gen7_depth_flush)(cmd_buffer);
@@ -2465,6 +2476,8 @@ void genX(CmdEndRenderPass)(
 
    anv_cmd_buffer_resolve_subpass(cmd_buffer);
 
+   cmd_buffer->state.hiz_enabled = false;
+
 #ifndef NDEBUG
    anv_dump_add_framebuffer(cmd_buffer, cmd_buffer->state.framebuffer);
 #endif