Mesa (master): freedreno/a6xx: fix hang with large render target

[GitLab Mirror]

Module: Mesa
Branch: master
Commit: 39d00722b22a0059bbc58f0158a22f384519cd39
URL:    http://cgit.freedesktop.org/mesa/mesa/commit/?id=39d00722b22a0059bbc58f0158a22f384519cd39

Author: Rob Clark <robdclark at chromium.org>
Date:   Tue Aug 25 13:02:34 2020 -0700

freedreno/a6xx: fix hang with large render target

It seems we do have some limits.  Similar to older gens, # of tiles per
pipe cannot be more than 32.  But I could not trigger any hangs with 16
or more tiles per pipe in either X or Y direction, so that limit does
not seem to apply.

Signed-off-by: Rob Clark <robdclark at chromium.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6461>

---

 src/gallium/drivers/freedreno/a6xx/fd6_gmem.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/src/gallium/drivers/freedreno/a6xx/fd6_gmem.c b/src/gallium/drivers/freedreno/a6xx/fd6_gmem.c
index 236151e01a7..eb4cbbf52df 100644
--- a/src/gallium/drivers/freedreno/a6xx/fd6_gmem.c
+++ b/src/gallium/drivers/freedreno/a6xx/fd6_gmem.c
@@ -254,7 +254,8 @@ use_hw_binning(struct fd_batch *batch)
 {
 	const struct fd_gmem_stateobj *gmem = batch->gmem_state;
 
-	// TODO figure out hw limits for binning
+	if ((gmem->maxpw * gmem->maxph) > 32)
+		return false;
 
 	return fd_binning_enabled && ((gmem->nbins_x * gmem->nbins_y) >= 2) &&
 			(batch->num_draws > 0);