From a683012a80a3408b3b71f22b2a97d9eaaac11a46 Mon Sep 17 00:00:00 2001 From: Paul Berry Date: Thu, 26 Apr 2012 06:35:56 -0700 Subject: [PATCH] i965/Gen6: Work around GPU hangs due to misaligned depth coordinate offsets. In i965 Gen6, Mesa has for a long time used the "depth coordinate offset X/Y" settings (in 3DSTATE_DEPTH_BUFFER) to cause the GPU to render to miplevels other than 0. Unfortunately, this doesn't work, because these offsets must be aligned to multiples of 8, and miplevels in the depth buffer are only guaranteed to be aligned to multiples of 4. When the offsets aren't aligned to a multiple of 8, the GPU sometimes hangs. As a temporary measure, to avoid GPU hangs, this patch smashes the 3 LSB's of "depth coordinate offset X/Y" to 0. This results in incorrect rendering to mipmapped depth textures, but that seems like a reasonable stopgap while we figure out a better solution. (Note that we have only ever observed this GPU hang on Gen6 when HiZ is enabled, so another possible stopgap would be to disable HiZ). Avoids GPU hangs in piglit test "depthstencil-render-miplevels" at texture sizes that are not powers of 2. Reviewed-by: Chad Verace --- src/mesa/drivers/dri/i965/brw_misc_state.c | 36 ++++++++++++++++++++++++++++++ src/mesa/drivers/dri/i965/gen6_hiz.c | 18 +++++++++++++++ 2 files changed, 54 insertions(+) diff --git a/src/mesa/drivers/dri/i965/brw_misc_state.c b/src/mesa/drivers/dri/i965/brw_misc_state.c index 7a0f5a3..0c0389f 100644 --- a/src/mesa/drivers/dri/i965/brw_misc_state.c +++ b/src/mesa/drivers/dri/i965/brw_misc_state.c @@ -398,6 +398,24 @@ static void emit_depthbuffer(struct brw_context *brw) tile_x = draw_x & tile_mask_x; tile_y = draw_y & tile_mask_y; + /* According to the Sandy Bridge PRM, volume 2 part 1, pp326-327 + * (3DSTATE_DEPTH_BUFFER dw5), in the documentation for "Depth + * Coordinate Offset X/Y": + * + * "The 3 LSBs of both offsets must be zero to ensure correct + * alignment" + * + * We have no guarantee that tile_x and tile_y are correctly aligned, + * since they are determined by the mipmap layout, which is only aligned + * to multiples of 4. + * + * So, to avoid hanging the GPU, just smash the low order 3 bits of + * tile_x and tile_y to 0. This is a temporary workaround until we come + * up with a better solution. + */ + tile_x &= ~7; + tile_y &= ~7; + BEGIN_BATCH(len); OUT_BATCH(_3DSTATE_DEPTH_BUFFER << 16 | (len - 2)); OUT_BATCH((BRW_DEPTHFORMAT_D32_FLOAT << 18) | @@ -436,6 +454,24 @@ static void emit_depthbuffer(struct brw_context *brw) tile_x = draw_x & tile_mask_x; tile_y = draw_y & tile_mask_y; + /* According to the Sandy Bridge PRM, volume 2 part 1, pp326-327 + * (3DSTATE_DEPTH_BUFFER dw5), in the documentation for "Depth + * Coordinate Offset X/Y": + * + * "The 3 LSBs of both offsets must be zero to ensure correct + * alignment" + * + * We have no guarantee that tile_x and tile_y are correctly aligned, + * since they are determined by the mipmap layout, which is only aligned + * to multiples of 4. + * + * So, to avoid hanging the GPU, just smash the low order 3 bits of + * tile_x and tile_y to 0. This is a temporary workaround until we come + * up with a better solution. + */ + tile_x &= ~7; + tile_y &= ~7; + offset = intel_region_get_aligned_offset(region, draw_x & ~tile_mask_x, draw_y & ~tile_mask_y); diff --git a/src/mesa/drivers/dri/i965/gen6_hiz.c b/src/mesa/drivers/dri/i965/gen6_hiz.c index d9b547c..67d6bdb 100644 --- a/src/mesa/drivers/dri/i965/gen6_hiz.c +++ b/src/mesa/drivers/dri/i965/gen6_hiz.c @@ -487,6 +487,24 @@ gen6_hiz_exec(struct intel_context *intel, draw_x & ~tile_mask_x, draw_y & ~tile_mask_y); + /* According to the Sandy Bridge PRM, volume 2 part 1, pp326-327 + * (3DSTATE_DEPTH_BUFFER dw5), in the documentation for "Depth + * Coordinate Offset X/Y": + * + * "The 3 LSBs of both offsets must be zero to ensure correct + * alignment" + * + * We have no guarantee that tile_x and tile_y are correctly aligned, + * since they are determined by the mipmap layout, which is only aligned + * to multiples of 4. + * + * So, to avoid hanging the GPU, just smash the low order 3 bits of + * tile_x and tile_y to 0. This is a temporary workaround until we come + * up with a better solution. + */ + tile_x &= ~7; + tile_y &= ~7; + uint32_t format; switch (mt->format) { case MESA_FORMAT_Z16: format = BRW_DEPTHFORMAT_D16_UNORM; break; -- 2.7.4