[Mesa-dev] [PATCH 4/4] i965: Use intel_bufferobj_buffer() wrapper in image surface state setup.

[Nanley Chery]

On Mon, Mar 19, 2018 at 11:26:59AM -0700, Francisco Jerez wrote:
> Instead of directly using intel_obj->buffer.  Among other things
> intel_bufferobj_buffer() will update intel_buffer_object::
> gpu_active_start/end, which are used by glBufferSubData() to decide
> which path to take.  Fixes a failure in the Piglit
> ARB_shader_image_load_store-host-mem-barrier Buffer Update/WaW tests,
> which could be reproduced with a non-standard glGetTexSubImage
> implementation (see bug report).
> 
> Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=105351

What do you think about leaving the bug open until we create a piglit
test to reproduce it? We could create a new bug otherwise.

> Reported-by: Nanley Chery <nanleychery at gmail.com>
> Cc: mesa-stable at lists.freedesktop.org
> ---
>  src/mesa/drivers/dri/i965/brw_wm_surface_state.c | 8 +++++---
>  1 file changed, 5 insertions(+), 3 deletions(-)
> 
> diff --git a/src/mesa/drivers/dri/i965/brw_wm_surface_state.c b/src/mesa/drivers/dri/i965/brw_wm_surface_state.c
> index 2ab15af793a..5d4c84bb55a 100644
> --- a/src/mesa/drivers/dri/i965/brw_wm_surface_state.c
> +++ b/src/mesa/drivers/dri/i965/brw_wm_surface_state.c
> @@ -1510,14 +1510,16 @@ update_image_surface(struct brw_context *brw,
>        const unsigned format = get_image_format(brw, u->_ActualFormat, access);
>  
>        if (obj->Target == GL_TEXTURE_BUFFER) {
> -         struct intel_buffer_object *intel_obj =
> -            intel_buffer_object(obj->BufferObject);
>           const unsigned texel_size = (format == ISL_FORMAT_RAW ? 1 :
>                                        _mesa_get_format_bytes(u->_ActualFormat));
>           const unsigned buffer_size = buffer_texture_range_size(brw, obj);
> +         struct brw_bo *const bo = !obj->BufferObject ? NULL :

Interesting. Did the old code wrongly assume that
obj->BufferObject != NULL? Maybe we need a test that tries to do image
load/store on a texture buffer without a backing buffer object?

> +            intel_bufferobj_buffer(brw, intel_buffer_object(obj->BufferObject),
> +                                   obj->BufferOffset, buffer_size,
> +                                   access != GL_READ_ONLY);
>  

Looks right to me. The comment above the helper function says:
   [...]
   * Anywhere that uses buffer objects in the pipeline should be using this to
   * mark the range of the buffer that is being accessed by the pipeline.
   */
  
...and that is what we're doing here. This patch is
Reviewed-by: Nanley Chery <nanley.g.chery at intel.com>

>           brw_emit_buffer_surface_state(
> -            brw, surf_offset, intel_obj->buffer, obj->BufferOffset,
> +            brw, surf_offset, bo, obj->BufferOffset,
>              format, buffer_size, texel_size,
>              access != GL_READ_ONLY ? RELOC_WRITE : 0);
>  
> -- 
> 2.16.1
>