v3d: Add support for using the TFU to do some blits.

This will be useful in particular for blits from raster to UIF for X11.
This commit is contained in:
Eric Anholt 2018-11-28 17:59:51 -08:00
parent e5b4d1f55f
commit 976ea90bdc
1 changed files with 152 additions and 65 deletions

View File

@ -330,6 +330,8 @@ v3d_stencil_blit(struct pipe_context *ctx, const struct pipe_blit_info *info)
#define V3D_TFU_ICFG_NUMMM_SHIFT 5
#define V3D_TFU_ICFG_TTYPE_SHIFT 9
#define V3D_TFU_ICFG_OPAD_SHIFT 22
#define V3D_TFU_ICFG_FORMAT_SHIFT 18
#define V3D_TFU_ICFG_FORMAT_RASTER 0
#define V3D_TFU_ICFG_FORMAT_SAND_128 1
@ -340,6 +342,121 @@ v3d_stencil_blit(struct pipe_context *ctx, const struct pipe_blit_info *info)
#define V3D_TFU_ICFG_FORMAT_UIF_NO_XOR 14
#define V3D_TFU_ICFG_FORMAT_UIF_XOR 15
static bool
v3d_tfu(struct pipe_context *pctx,
struct pipe_resource *pdst,
struct pipe_resource *psrc,
unsigned int src_level,
unsigned int base_level,
unsigned int last_level,
unsigned int src_layer,
unsigned int dst_layer)
{
struct v3d_context *v3d = v3d_context(pctx);
struct v3d_screen *screen = v3d->screen;
struct v3d_resource *src = v3d_resource(psrc);
struct v3d_resource *dst = v3d_resource(pdst);
struct v3d_resource_slice *src_base_slice = &src->slices[src_level];
struct v3d_resource_slice *dst_base_slice = &dst->slices[base_level];
int msaa_scale = pdst->nr_samples > 1 ? 2 : 1;
int width = u_minify(pdst->width0, base_level) * msaa_scale;
int height = u_minify(pdst->height0, base_level) * msaa_scale;
if (psrc->format != pdst->format)
return false;
if (psrc->nr_samples != pdst->nr_samples)
return false;
uint32_t tex_format = v3d_get_tex_format(&screen->devinfo,
pdst->format);
if (!v3d_tfu_supports_tex_format(&screen->devinfo, tex_format))
return false;
if (pdst->target != PIPE_TEXTURE_2D || psrc->target != PIPE_TEXTURE_2D)
return false;
/* Can't write to raster. */
if (dst_base_slice->tiling == VC5_TILING_RASTER)
return false;
v3d_flush_jobs_writing_resource(v3d, psrc);
v3d_flush_jobs_reading_resource(v3d, pdst);
struct drm_v3d_submit_tfu tfu = {
.ios = (height << 16) | width,
.bo_handles = {
src->bo->handle,
src != dst ? dst->bo->handle : 0
},
.in_sync = v3d->out_sync,
.out_sync = v3d->out_sync,
};
uint32_t src_offset = (src->bo->offset +
v3d_layer_offset(psrc, src_level, src_layer));
tfu.iia |= src_offset;
if (src_base_slice->tiling == VC5_TILING_RASTER) {
tfu.icfg |= (V3D_TFU_ICFG_FORMAT_RASTER <<
V3D_TFU_ICFG_FORMAT_SHIFT);
} else {
tfu.icfg |= ((V3D_TFU_ICFG_FORMAT_LINEARTILE +
(src_base_slice->tiling - VC5_TILING_LINEARTILE)) <<
V3D_TFU_ICFG_FORMAT_SHIFT);
}
uint32_t dst_offset = (dst->bo->offset +
v3d_layer_offset(pdst, src_level, dst_layer));
tfu.ioa |= dst_offset;
if (last_level != base_level)
tfu.ioa |= V3D_TFU_IOA_DIMTW;
tfu.ioa |= ((V3D_TFU_IOA_FORMAT_LINEARTILE +
(dst_base_slice->tiling - VC5_TILING_LINEARTILE)) <<
V3D_TFU_IOA_FORMAT_SHIFT);
tfu.icfg |= tex_format << V3D_TFU_ICFG_TTYPE_SHIFT;
tfu.icfg |= (last_level - base_level) << V3D_TFU_ICFG_NUMMM_SHIFT;
switch (src_base_slice->tiling) {
case VC5_TILING_UIF_NO_XOR:
case VC5_TILING_UIF_XOR:
tfu.iis |= (src_base_slice->padded_height /
(2 * v3d_utile_height(src->cpp)));
break;
case VC5_TILING_RASTER:
tfu.iis |= src_base_slice->stride / src->cpp;
break;
case VC5_TILING_LINEARTILE:
case VC5_TILING_UBLINEAR_1_COLUMN:
case VC5_TILING_UBLINEAR_2_COLUMN:
break;
}
/* If we're writing level 0 (!IOA_DIMTW), then we need to supply the
* OPAD field for the destination (how many extra UIF blocks beyond
* those necessary to cover the height). When filling mipmaps, the
* miplevel 1+ tiling state is inferred.
*/
if (dst_base_slice->tiling == VC5_TILING_UIF_NO_XOR ||
dst_base_slice->tiling == VC5_TILING_UIF_XOR) {
int uif_block_h = 2 * v3d_utile_height(dst->cpp);
int implicit_padded_height = align(height, uif_block_h);
tfu.icfg |= (((dst_base_slice->padded_height -
implicit_padded_height) / uif_block_h) <<
V3D_TFU_ICFG_OPAD_SHIFT);
}
int ret = v3d_ioctl(screen->fd, DRM_IOCTL_V3D_SUBMIT_TFU, &tfu);
if (ret != 0) {
fprintf(stderr, "Failed to submit TFU job: %d\n", ret);
return false;
}
dst->writes++;
return true;
}
boolean
v3d_generate_mipmap(struct pipe_context *pctx,
struct pipe_resource *prsc,
@ -349,78 +466,49 @@ v3d_generate_mipmap(struct pipe_context *pctx,
unsigned int first_layer,
unsigned int last_layer)
{
struct v3d_context *v3d = v3d_context(pctx);
struct v3d_screen *screen = v3d->screen;
struct v3d_resource *rsc = v3d_resource(prsc);
struct v3d_resource_slice *base_slice = &rsc->slices[base_level];
int width = u_minify(prsc->width0, base_level);
int height = u_minify(prsc->height0, base_level);
uint32_t tex_format = v3d_get_tex_format(&screen->devinfo,
prsc->format);
if (!v3d_tfu_supports_tex_format(&screen->devinfo, tex_format))
if (format != prsc->format)
return false;
if (prsc->target != PIPE_TEXTURE_2D)
return false;
/* Since we don't support array or 3D textures, there should be only
* one layer.
/* We could maybe support looping over layers for array textures, but
* we definitely don't support 3D.
*/
int layer = first_layer;
assert(first_layer == last_layer);
/* Can't write to raster. */
if (base_slice->tiling == VC5_TILING_RASTER)
if (first_layer != last_layer)
return false;
v3d_flush_jobs_reading_resource(v3d, prsc);
return v3d_tfu(pctx,
prsc, prsc,
base_level,
base_level, last_level,
first_layer, first_layer);
}
struct drm_v3d_submit_tfu tfu = {
.ios = (height << 16) | width,
.bo_handles = { rsc->bo->handle },
.in_sync = v3d->out_sync,
.out_sync = v3d->out_sync,
};
uint32_t offset = (rsc->bo->offset +
v3d_layer_offset(prsc, base_level, layer));
tfu.iia |= offset;
tfu.icfg |= ((V3D_TFU_ICFG_FORMAT_LINEARTILE +
(base_slice->tiling - VC5_TILING_LINEARTILE)) <<
V3D_TFU_ICFG_FORMAT_SHIFT);
static bool
v3d_tfu_blit(struct pipe_context *pctx, const struct pipe_blit_info *info)
{
int dst_width = u_minify(info->dst.resource->width0, info->dst.level);
int dst_height = u_minify(info->dst.resource->height0, info->dst.level);
tfu.ioa |= offset;
tfu.ioa |= V3D_TFU_IOA_DIMTW;
tfu.ioa |= ((V3D_TFU_IOA_FORMAT_LINEARTILE +
(base_slice->tiling - VC5_TILING_LINEARTILE)) <<
V3D_TFU_IOA_FORMAT_SHIFT);
if ((info->mask & PIPE_MASK_RGBA) == 0)
return false;
tfu.icfg |= tex_format << V3D_TFU_ICFG_TTYPE_SHIFT;
tfu.icfg |= (last_level - base_level) << V3D_TFU_ICFG_NUMMM_SHIFT;
switch (base_slice->tiling) {
case VC5_TILING_UIF_NO_XOR:
case VC5_TILING_UIF_XOR:
tfu.iis |= (base_slice->padded_height /
(2 * v3d_utile_height(rsc->cpp)));
break;
case VC5_TILING_RASTER:
tfu.iis |= base_slice->stride / rsc->cpp;
break;
case VC5_TILING_LINEARTILE:
case VC5_TILING_UBLINEAR_1_COLUMN:
case VC5_TILING_UBLINEAR_2_COLUMN:
break;
}
int ret = v3d_ioctl(screen->fd, DRM_IOCTL_V3D_SUBMIT_TFU, &tfu);
if (ret != 0) {
fprintf(stderr, "Failed to submit TFU job: %d\n", ret);
if (info->dst.box.x != 0 ||
info->dst.box.y != 0 ||
info->dst.box.width != dst_width ||
info->dst.box.height != dst_height ||
info->src.box.x != 0 ||
info->src.box.y != 0 ||
info->src.box.width != info->dst.box.width ||
info->src.box.height != info->dst.box.height) {
return false;
}
rsc->writes++;
if (info->dst.format != info->src.format)
return false;
return true;
return v3d_tfu(pctx, info->dst.resource, info->src.resource,
info->src.level,
info->dst.level, info->dst.level,
info->src.box.z, info->dst.box.z);
}
/* Optimal hardware path for blitting pixels.
@ -436,10 +524,9 @@ v3d_blit(struct pipe_context *pctx, const struct pipe_blit_info *blit_info)
info.mask &= ~PIPE_MASK_S;
}
#if 0
if (v3d_tile_blit(pctx, blit_info))
return;
#endif
if (v3d_tfu_blit(pctx, blit_info))
info.mask &= ~PIPE_MASK_RGBA;
v3d_render_blit(pctx, &info);
if (info.mask)
v3d_render_blit(pctx, &info);
}