fdo-mirrors/mesa
1
0
Fork 0
mirror of https://gitlab.freedesktop.org/mesa/mesa.git synced 2025-12-27 10:30:08 +01:00
Code Issues Projects Releases Packages Wiki Activity Actions 1

isl: fix a number of errors on storage format support on Gfx9/12.5

Browse source 
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Cc: mesa-stable
Reviewed-by: Ivan Briano <ivan.briano@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/22302>
This commit is contained in:
Lionel Landwerlin 2023-04-04 13:32:53 +03:00 • committed by Marge Bot
parent c221bfbd85
commit d4f498a583
7 changed files with 64 additions and 31 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

3
src/intel/isl/isl.h
View file

@ -2078,7 +2078,8 @@ void isl_color_value_unpack(union isl_color_value *value,
enum isl_format format,
const uint32_t *data_in);
bool isl_is_storage_image_format(enum isl_format fmt);
bool isl_is_storage_image_format(const struct intel_device_info *devinfo,
enum isl_format fmt);
enum isl_format
isl_lower_storage_image_format(const struct intel_device_info *devinfo,

12
src/intel/isl/isl_format.c
View file

@ -147,12 +147,12 @@ static const struct surface_format_info format_info[] = {
SF( x, x, x, x, x, x, Y, x, x, x, x, x, x, R32G32_USCALED)
SF( x, x, x, x, x, x, 75, x, x, x, x, x, x, R32G32_SFIXED)
SF( x, x, x, x, x, x, 80, x, x, x, x, x, 90, R64_PASSTHRU)
SF( Y, Y, x, Y, Y, Y, Y, x, 60, 70, x, 90, x, B8G8R8A8_UNORM)
SF( Y, Y, x, Y, Y, Y, Y, x, 60, 70, 125, 90, x, B8G8R8A8_UNORM)
SF( Y, Y, x, x, Y, Y, x, x, x, x, x, 110, x, B8G8R8A8_UNORM_SRGB)
/* smpl filt shad CK RT AB VB SO color TW TR ccs_e TA */
SF( Y, Y, x, x, Y, Y, Y, x, 60, 70, x, 110, x, R10G10B10A2_UNORM)
SF( Y, Y, x, x, Y, Y, Y, x, 60, 70, 125, 110, x, R10G10B10A2_UNORM)
SF( Y, Y, x, x, x, x, x, x, 60, x, x, 120, x, R10G10B10A2_UNORM_SRGB)
SF( Y, x, x, x, Y, x, Y, x, x, 70, x, 110, x, R10G10B10A2_UINT)
SF( Y, x, x, x, Y, x, Y, x, x, 70, 125, 110, x, R10G10B10A2_UINT)
SF( Y, Y, x, x, x, x, Y, x, x, x, x, x, x, R10G10B10_SNORM_A2_UNORM)
SF( Y, Y, x, x, Y, Y, Y, x, 60, 70, 110, 90, x, R8G8B8A8_UNORM)
SF( Y, Y, x, x, Y, Y, x, x, 60, x, x, 110, x, R8G8B8A8_UNORM_SRGB)
@ -164,10 +164,10 @@ static const struct surface_format_info format_info[] = {
SF( Y, x, x, x, Y, x, Y, x, x, 70, 90, 90, x, R16G16_SINT)
SF( Y, x, x, x, Y, x, Y, x, x, 70, 75, 90, x, R16G16_UINT)
SF( Y, Y, x, x, Y, Y, Y, x, x, 70, 90, 90, x, R16G16_FLOAT)
SF( Y, Y, x, x, Y, Y, 75, x, 60, 70, x, 110, x, B10G10R10A2_UNORM)
SF( Y, Y, x, x, Y, Y, 75, x, 60, 70, 125, 110, x, B10G10R10A2_UNORM)
SF( Y, Y, x, x, Y, Y, x, x, 60, x, x, 110, x, B10G10R10A2_UNORM_SRGB)
SF( Y, Y, x, x, Y, Y, Y, x, x, 70, x, 110, x, R11G11B10_FLOAT)
SF(120, 120, x, x, 120, 120, x, x, x, x, x, 120, x, R10G10B10_FLOAT_A2_UNORM)
SF( Y, Y, x, x, Y, Y, Y, x, x, 70, 125, 110, x, R11G11B10_FLOAT)
SF(120, 120, x, x, 120, 120, x, x, x, 125, 125, 120, x, R10G10B10_FLOAT_A2_UNORM)
SF( Y, x, x, x, Y, x, Y, Y, x, 70, 70, 90, 70, R32_SINT)
SF( Y, x, x, x, Y, x, Y, Y, x, 70, 70, 90, 70, R32_UINT)
SF( Y, 50, Y, x, Y, Y, Y, Y, x, 70, 70, 90, 110, R32_FLOAT)

68
src/intel/isl/isl_storage_image.c
View file

@ -25,7 +25,8 @@
#include "compiler/brw_compiler.h"
bool
isl_is_storage_image_format(enum isl_format format)
isl_is_storage_image_format(const struct intel_device_info *devinfo,
enum isl_format format)
{
/* XXX: Maybe we should put this in the CSV? */
@ -94,9 +95,24 @@ isl_lower_storage_image_format(const struct intel_device_info *devinfo,
*
* "The surface format for the typed atomic integer operations must
* be R32_UINT or R32_SINT."
*
* But checking the BSpec 1706, you find a different restriction. There the
* wording is :
*
* "The surface format must be one of R32_UINT, R32_SINT or R32_FLOAT"
*
* The confusion is probably related to atomic integer messages. For
* example an IADD instruction would require a R32_UINT/R32_SINT surface.
* But a CMPXCHG instruction does not really care about the type, it just
* does bit to bit comparison and swap.
*
* The confusion seems to have propagated to the simulation environment.
* Gfx12 has the same restrictions as Gfx11 regarding doing a CMPXCHG on a
* R32_FLOAT surface, but the Gfx11 environment will report an error while
* Gfx12 passes fine. More importantly HW doesn't seem to mind.
*/
case ISL_FORMAT_R32_FLOAT:
return ISL_FORMAT_R32_UINT;
return format;
/* From HSW to BDW the only 64bpp format supported for typed access is
* RGBA_UINT16. IVB falls back to untyped.
@ -150,39 +166,55 @@ isl_lower_storage_image_format(const struct intel_device_info *devinfo,
case ISL_FORMAT_R8_SINT:
return (devinfo->ver >= 9 ? format : ISL_FORMAT_R8_UINT);
/* Neither the 2/10/10/10 nor the 11/11/10 packed formats are supported
/* Here the PRMs are a bit out of date. But according to BSpec 47635
* (Gfx12.5), the 2/10/10/10 and the 11/11/10 packed formats are supported
* by the hardware.
*/
case ISL_FORMAT_R10G10B10A2_UINT:
case ISL_FORMAT_R10G10B10A2_UNORM:
case ISL_FORMAT_R11G11B10_FLOAT:
return ISL_FORMAT_R32_UINT;
return devinfo->verx10 >= 125 ? format : ISL_FORMAT_R32_UINT;
/* No normalized fixed-point formats are supported by the hardware. */
/* No normalized fixed-point formats are supported by the hardware until Gfx11. */
case ISL_FORMAT_R16G16B16A16_UNORM:
case ISL_FORMAT_R16G16B16A16_SNORM:
return (devinfo->ver >= 11 ? format :
devinfo->verx10 >= 75 ?
ISL_FORMAT_R16G16B16A16_UINT :
ISL_FORMAT_R32G32_UINT);
if (devinfo->ver >= 11)
return format;
if (devinfo->ver >= 9)
return ISL_FORMAT_R32G32_UINT;
if (devinfo->verx10 >= 75)
return ISL_FORMAT_R16G16B16A16_UINT;
return ISL_FORMAT_R32G32_UINT;
case ISL_FORMAT_R8G8B8A8_UNORM:
case ISL_FORMAT_R8G8B8A8_SNORM:
return (devinfo->ver >= 11 ? format :
devinfo->verx10 >= 75 ?
ISL_FORMAT_R8G8B8A8_UINT : ISL_FORMAT_R32_UINT);
if (devinfo->ver >= 11)
return format;
if (devinfo->ver >= 9)
return ISL_FORMAT_R32_UINT;
if (devinfo->verx10 >= 75)
return ISL_FORMAT_R8G8B8A8_UINT;
return ISL_FORMAT_R32_UINT;
case ISL_FORMAT_R16G16_UNORM:
case ISL_FORMAT_R16G16_SNORM:
return (devinfo->ver >= 11 ? format :
devinfo->verx10 >= 75 ?
ISL_FORMAT_R16G16_UINT : ISL_FORMAT_R32_UINT);
if (devinfo->ver >= 11)
return format;
if (devinfo->ver >= 9)
return ISL_FORMAT_R32_UINT;
if (devinfo->verx10 >= 75)
return ISL_FORMAT_R16G16_UINT;
return ISL_FORMAT_R32_UINT;
case ISL_FORMAT_R8G8_UNORM:
case ISL_FORMAT_R8G8_SNORM:
return (devinfo->ver >= 11 ? format :
devinfo->verx10 >= 75 ?
ISL_FORMAT_R8G8_UINT : ISL_FORMAT_R16_UINT);
if (devinfo->ver >= 11)
return format;
if (devinfo->ver >= 9)
return ISL_FORMAT_R16_UINT;
if (devinfo->verx10 >= 75)
return ISL_FORMAT_R8G8_UINT;
return ISL_FORMAT_R16_UINT;
case ISL_FORMAT_R16_UNORM:
case ISL_FORMAT_R16_SNORM:

2
src/intel/vulkan/anv_formats.c
View file

@ -830,7 +830,7 @@ get_buffer_format_features2(const struct intel_device_info *devinfo,
if (isl_format_supports_vertex_fetch(devinfo, isl_format))
flags |= VK_FORMAT_FEATURE_2_VERTEX_BUFFER_BIT;
if (isl_is_storage_image_format(isl_format))
if (isl_is_storage_image_format(devinfo, isl_format))
flags |= VK_FORMAT_FEATURE_2_STORAGE_TEXEL_BUFFER_BIT;
if (isl_format == ISL_FORMAT_R32_SINT || isl_format == ISL_FORMAT_R32_UINT)

4
src/intel/vulkan/anv_image.c
View file

@ -430,7 +430,7 @@ anv_get_isl_format_with_usage(const struct intel_device_info *devinfo,
vk_tiling);
if ((vk_usage == VK_IMAGE_USAGE_STORAGE_BIT) &&
isl_is_storage_image_format(format.isl_format)) {
isl_is_storage_image_format(devinfo, format.isl_format)) {
enum isl_format lowered_format =
isl_lower_storage_image_format(devinfo, format.isl_format);
@ -2689,7 +2689,7 @@ anv_CreateImageView(VkDevice _device,
iview->planes[vplane].lowered_storage_surface_state.state =
alloc_bindless_surface_state(device);
if (isl_is_storage_image_format(format.isl_format)) {
if (isl_is_storage_image_format(device->info, format.isl_format)) {
anv_image_fill_surface_state(device, image, 1ULL << iaspect_bit,
&storage_view,
ISL_SURF_USAGE_STORAGE_BIT,

2
src/intel/vulkan_hasvk/anv_formats.c
View file

@ -843,7 +843,7 @@ get_buffer_format_features2(const struct intel_device_info *devinfo,
if (isl_format_supports_vertex_fetch(devinfo, isl_format))
flags |= VK_FORMAT_FEATURE_2_VERTEX_BUFFER_BIT;
if (isl_is_storage_image_format(isl_format))
if (isl_is_storage_image_format(devinfo, isl_format))
flags |= VK_FORMAT_FEATURE_2_STORAGE_TEXEL_BUFFER_BIT;
if (isl_format == ISL_FORMAT_R32_SINT || isl_format == ISL_FORMAT_R32_UINT)

4
src/intel/vulkan_hasvk/anv_image.c
View file

@ -448,7 +448,7 @@ anv_get_isl_format_with_usage(const struct intel_device_info *devinfo,
vk_tiling);
if ((vk_usage == VK_IMAGE_USAGE_STORAGE_BIT) &&
isl_is_storage_image_format(format.isl_format)) {
isl_is_storage_image_format(devinfo, format.isl_format)) {
enum isl_format lowered_format =
isl_lower_storage_image_format(devinfo, format.isl_format);
@ -2523,7 +2523,7 @@ anv_CreateImageView(VkDevice _device,
&iview->planes[vplane].storage_surface_state,
NULL);
if (isl_is_storage_image_format(format.isl_format)) {
if (isl_is_storage_image_format(device->info, format.isl_format)) {
iview->planes[vplane].lowered_storage_surface_state.state =
alloc_surface_state(device);