OpenHarmony/third_party_mesa3d
2
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

freedreno/a6xx: Avoid stalling for occlusion queries

Browse Source 
If we postpone computing the counter delta until after each tile (or
sysmem pass), we don't have to stall in the middle of the draw stream.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5064>
This commit is contained in:
Kristian H. Kristensen
2020-05-15 12:23:18 -07:00
committed by Marge Bot
parent 1c21577246
commit 13fc03f4c0
4 changed files with 35 additions and 12 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
src/gallium/drivers/freedreno/a6xx/fd6_gmem.c
View File

@@ -1377,6 +1377,9 @@ fd6_emit_tile_fini(struct fd_batch *batch)
{
struct fd_ringbuffer *ring = batch->gmem;
if (batch->epilogue)
fd6_emit_ib(batch->gmem, batch->epilogue);
OUT_PKT4(ring, REG_A6XX_GRAS_LRZ_CNTL, 1);
OUT_RING(ring, A6XX_GRAS_LRZ_CNTL_ENABLE | A6XX_GRAS_LRZ_CNTL_UNK3);
@@ -1527,6 +1530,9 @@ fd6_emit_sysmem_fini(struct fd_batch *batch)
{
struct fd_ringbuffer *ring = batch->gmem;
if (batch->epilogue)
fd6_emit_ib(batch->gmem, batch->epilogue);
OUT_PKT7(ring, CP_SKIP_IB2_ENABLE_GLOBAL, 1);
OUT_RING(ring, 0x0);

23
src/gallium/drivers/freedreno/a6xx/fd6_query.c
View File

@@ -95,21 +95,20 @@ occlusion_pause(struct fd_acc_query *aq, struct fd_batch *batch)
fd6_event_write(batch, ring, ZPASS_DONE, false);
OUT_PKT7(ring, CP_WAIT_REG_MEM, 6);
OUT_RING(ring, 0x00000014); // XXX
OUT_RELOC(ring, query_sample(aq, stop));
OUT_RING(ring, 0xffffffff);
OUT_RING(ring, 0xffffffff);
OUT_RING(ring, 0x00000010); // XXX
/* To avoid stalling in the draw buffer, emit code the code to compute the
* counter delta in the epilogue ring.
*/
struct fd_ringbuffer *epilogue = fd_batch_get_epilogue(batch);
fd_wfi(batch, epilogue);
/* result += stop - start: */
OUT_PKT7(ring, CP_MEM_TO_MEM, 9);
OUT_RING(ring, CP_MEM_TO_MEM_0_DOUBLE |
OUT_PKT7(epilogue, CP_MEM_TO_MEM, 9);
OUT_RING(epilogue, CP_MEM_TO_MEM_0_DOUBLE |
CP_MEM_TO_MEM_0_NEG_C);
OUT_RELOC(ring, query_sample(aq, result)); /* dst */
OUT_RELOC(ring, query_sample(aq, result)); /* srcA */
OUT_RELOC(ring, query_sample(aq, stop)); /* srcB */
OUT_RELOC(ring, query_sample(aq, start)); /* srcC */
OUT_RELOC(epilogue, query_sample(aq, result)); /* dst */
OUT_RELOC(epilogue, query_sample(aq, result)); /* srcA */
OUT_RELOC(epilogue, query_sample(aq, stop)); /* srcB */
OUT_RELOC(epilogue, query_sample(aq, start)); /* srcC */
fd6_context(batch->ctx)->samples_passed_queries--;
}

5
src/gallium/drivers/freedreno/freedreno_batch.c
View File

@@ -161,6 +161,11 @@ batch_fini(struct fd_batch *batch)
batch->lrz_clear = NULL;
}
if (batch->epilogue) {
fd_ringbuffer_del(batch->epilogue);
batch->epilogue = NULL;
}
if (batch->tile_setup) {
fd_ringbuffer_del(batch->tile_setup);
batch->tile_setup = NULL;

13
src/gallium/drivers/freedreno/freedreno_batch.h
View File

@@ -184,6 +184,9 @@ struct fd_batch {
/** tiling/gmem (IB0) cmdstream: */
struct fd_ringbuffer *gmem;
/** epilogue cmdstream: */
struct fd_ringbuffer *epilogue;
// TODO maybe more generically split out clear and clear_binning rings?
struct fd_ringbuffer *lrz_clear;
struct fd_ringbuffer *tile_setup;
@@ -336,4 +339,14 @@ fd_event_write(struct fd_batch *batch, struct fd_ringbuffer *ring,
fd_reset_wfi(batch);
}
static inline struct fd_ringbuffer *
fd_batch_get_epilogue(struct fd_batch *batch)
{
if (batch->epilogue == NULL)
batch->epilogue = fd_submit_new_ringbuffer(batch->submit, 0x1000, 0);
return batch->epilogue;
}
#endif /* FREEDRENO_BATCH_H_ */