diff options
author | Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl> | 2020-01-30 17:58:55 +0100 |
---|---|---|
committer | Marge Bot <eric+marge@anholt.net> | 2020-02-04 21:22:30 +0000 |
commit | 65a6dc5139fddd5e01eaedcc57fc67e0a6a28c94 (patch) | |
tree | 6b30a846477028120cff627e272dc34255da28e8 /src/amd | |
parent | 17303c9851e32194550f899399859944fa5b3009 (diff) |
radv: Do not set SX DISABLE bits for RB+ with unused surfaces.
The extra bits in CB_SHADER_MASK break dual source blending in
SkQP on a Stoney device. However:
- As far as I can tell, some other dual source blend tests are passing
before and after the change.
- A hacked around skqp passes on my Vega desktop and Raven laptop
- Getting Skqp to give any useful info or to run it outside of Android
on ChromeOS is proving difficult.
I have confirmed 3 strategies that seem to work:
- The old radv behavior of setting CB_SHADER_MASK to 0xF
- AMDVLK: CB_SHADER_MASK = 0xFF, and the 3 RB+ regs
are 0.
- radeonsi: CB_SHADER_MASK = 0xFF, but does not set DISABLE
bits in SX_BLEND_OPT_CONTROL for CB 1-7.
Let us use the radeonsi solution as that solution also seems like the correct
thing to do for holes. I have tested on my Raven laptop that setting the high
surfaces to not disabled and downconvert to 32_R does not imply a performance
penalty.
Fixes: e9316fdfd48 "radv: fix setting CB_SHADER_MASK for dual source blending"
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Tested-by: Marge Bot <https://gitlab.freedesktop.org/mesa/mesa/merge_requests/3670>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/merge_requests/3670>
Diffstat (limited to 'src/amd')
-rw-r--r-- | src/amd/vulkan/radv_cmd_buffer.c | 13 |
1 files changed, 7 insertions, 6 deletions
diff --git a/src/amd/vulkan/radv_cmd_buffer.c b/src/amd/vulkan/radv_cmd_buffer.c index 21ef5caa8e5..549b7e8843d 100644 --- a/src/amd/vulkan/radv_cmd_buffer.c +++ b/src/amd/vulkan/radv_cmd_buffer.c @@ -996,8 +996,9 @@ radv_emit_rbplus_state(struct radv_cmd_buffer *cmd_buffer) for (unsigned i = 0; i < subpass->color_count; ++i) { if (subpass->color_attachments[i].attachment == VK_ATTACHMENT_UNUSED) { - sx_blend_opt_control |= S_02875C_MRT0_COLOR_OPT_DISABLE(1) << (i * 4); - sx_blend_opt_control |= S_02875C_MRT0_ALPHA_OPT_DISABLE(1) << (i * 4); + /* We don't set the DISABLE bits, because the HW can't have holes, + * so the SPI color format is set to 32-bit 1-component. */ + sx_ps_downconvert |= V_028754_SX_RT_EXPORT_32_R << (i * 4); continue; } @@ -1113,10 +1114,10 @@ radv_emit_rbplus_state(struct radv_cmd_buffer *cmd_buffer) } } - for (unsigned i = subpass->color_count; i < 8; ++i) { - sx_blend_opt_control |= S_02875C_MRT0_COLOR_OPT_DISABLE(1) << (i * 4); - sx_blend_opt_control |= S_02875C_MRT0_ALPHA_OPT_DISABLE(1) << (i * 4); - } + /* Do not set the DISABLE bits for the unused attachments, as that + * breaks dual source blending in SkQP and does not seem to improve + * performance. */ + /* TODO: avoid redundantly setting context registers */ radeon_set_context_reg_seq(cmd_buffer->cs, R_028754_SX_PS_DOWNCONVERT, 3); radeon_emit(cmd_buffer->cs, sx_ps_downconvert); |