diff options
| author | Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com> | 2022-02-03 09:35:07 +0100 |
|---|---|---|
| committer | Dylan Baker <dylan.c.baker@intel.com> | 2022-02-03 10:32:02 -0800 |
| commit | b6e296f82333f8a1290f3aefbf554d793bd8003a (patch) | |
| tree | 9fdc5de7f780e44671cef34ebfc145fe187c5da1 | |
| parent | fabb6b5c5e7785214fefdfa89354146f3c50eaf0 (diff) | |
radeonsi: limit loop unrolling for LLVM < 13
Without this change LLVM 12 hits this error:
"""
LLVM ERROR: Error while trying to spill SGPR0_SGPR1 from class SReg_64:
Cannot scavenge register without an emergency spill slot!
"""
when running glcts KHR-GL46.arrays_of_arrays_gl.AtomicUsage test.
Fixes: 9ff086052ab ("radeonsi: unroll loops of up to 128 iterations")
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/14848>
(cherry picked from commit eaa87b1a4659c1c06b0ac9e5659943c888dff6d3)
| -rw-r--r-- | .pick_status.json | 2 | ||||
| -rw-r--r-- | src/gallium/drivers/radeonsi/si_get.c | 4 |
2 files changed, 3 insertions, 3 deletions
diff --git a/.pick_status.json b/.pick_status.json index 20718c7f8bf..b97ba0008ec 100644 --- a/.pick_status.json +++ b/.pick_status.json @@ -139,7 +139,7 @@ "description": "radeonsi: limit loop unrolling for LLVM < 13", "nominated": true, "nomination_type": 1, - "resolution": 0, + "resolution": 1, "main_sha": null, "because_sha": "9ff086052ab7bff3cb55c06365543190a3afe188" }, diff --git a/src/gallium/drivers/radeonsi/si_get.c b/src/gallium/drivers/radeonsi/si_get.c index 795461a1c90..985daa1c776 100644 --- a/src/gallium/drivers/radeonsi/si_get.c +++ b/src/gallium/drivers/radeonsi/si_get.c @@ -1072,8 +1072,8 @@ void si_init_screen_get_functions(struct si_screen *sscreen) .has_udot_4x8 = sscreen->info.has_accelerated_dot_product, .has_dot_2x16 = sscreen->info.has_accelerated_dot_product, .optimize_sample_mask_in = true, - .max_unroll_iterations = 128, - .max_unroll_iterations_aggressive = 128, + .max_unroll_iterations = LLVM_VERSION_MAJOR >= 13 ? 128 : 32, + .max_unroll_iterations_aggressive = LLVM_VERSION_MAJOR >= 13 ? 128 : 32, .use_interpolated_input_intrinsics = true, .lower_uniforms_to_ubo = true, .support_16bit_alu = sscreen->options.fp16, |