[Mesa-dev] [PATCH 2/2] radeonsi/compute: Clamp COMPUTE_TMPRING_SIZE.WAVES to: num_cu * 32

Tom Stellard thomas.stellard at amd.com
Fri Dec 5 15:59:11 PST 2014


This is the maximum value allowed for this field.
---
 src/gallium/drivers/radeonsi/si_compute.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/src/gallium/drivers/radeonsi/si_compute.c b/src/gallium/drivers/radeonsi/si_compute.c
index 6ddb478..bf935dc 100644
--- a/src/gallium/drivers/radeonsi/si_compute.c
+++ b/src/gallium/drivers/radeonsi/si_compute.c
@@ -371,6 +371,9 @@ static void si_launch_grid(
 		| S_00B85C_SH1_CU_EN(0xffff /* Default value */))
 		;
 
+	num_waves_for_scratch =
+		MIN2(num_waves_for_scratch,
+		     32 * sctx->screen->b.info.max_compute_units);
 	si_pm4_set_reg(pm4, R_00B860_COMPUTE_TMPRING_SIZE,
 		/* The maximum value for WAVES is 32 * num CU.
 		 * If you program this value incorrectly, the GPU will hang if
-- 
2.0.4



More information about the mesa-dev mailing list