[Mesa-dev] [PATCH 2/3] radeonsi: CP DMA clear supports unaligned destination addresses

[Nicolai Hähnle]

From: Nicolai Hähnle <nicolai.haehnle at amd.com>

---
 src/gallium/drivers/radeonsi/si_cp_dma.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/src/gallium/drivers/radeonsi/si_cp_dma.c b/src/gallium/drivers/radeonsi/si_cp_dma.c
index b40f5cc..0cf7b3b 100644
--- a/src/gallium/drivers/radeonsi/si_cp_dma.c
+++ b/src/gallium/drivers/radeonsi/si_cp_dma.c
@@ -190,34 +190,35 @@ static void si_clear_buffer(struct pipe_context *ctx, struct pipe_resource *dst,
 	if (!size)
 		return;
 
 	/* Mark the buffer range of destination as valid (initialized),
 	 * so that transfer_map knows it should wait for the GPU when mapping
 	 * that range. */
 	util_range_add(&rdst->valid_buffer_range, offset,
 		       offset + size);
 
 	/* Fallback for unaligned clears. */
-	if (offset % 4 != 0 || size % 4 != 0) {
+	if (size % 4 != 0) {
 		uint8_t *map = r600_buffer_map_sync_with_rings(&sctx->b, rdst,
 							       PIPE_TRANSFER_WRITE);
 		map += offset;
 		for (uint64_t i = 0; i < size; i++) {
 			unsigned byte_within_dword = (offset + i) % 4;
 			*map++ = (value >> (byte_within_dword * 8)) & 0xff;
 		}
 		return;
 	}
 
 	/* dma_clear_buffer can use clear_buffer on failure. Make sure that
 	 * doesn't happen. We don't want an infinite recursion: */
 	if (sctx->b.dma.cs &&
+	    (offset % 4 == 0) &&
 	    /* CP DMA is very slow. Always use SDMA for big clears. This
 	     * alone improves DeusEx:MD performance by 70%. */
 	    (size > 128 * 1024 ||
 	     /* Buffers not used by the GFX IB yet will be cleared by SDMA.
 	      * This happens to move most buffer clears to SDMA, including
 	      * DCC and CMASK clears, because pipe->clear clears them before
 	      * si_emit_framebuffer_state (in a draw call) adds them.
 	      * For example, DeusEx:MD has 21 buffer clears per frame and all
 	      * of them are moved to SDMA thanks to this. */
 	     !ws->cs_is_buffer_referenced(sctx->b.gfx.cs, rdst->buf,
-- 
2.9.3