/**
* Implements a rectangular block transfer (blit) of pixels between two
* miptrees.
*
* Our blitter can operate on 1, 2, or 4-byte-per-pixel data, with generous,
* but limited, pitches and sizes allowed.
*
* The src/dst coordinates are relative to the given level/slice of the
* miptree.
*
* If @src_flip or @dst_flip is set, then the rectangle within that miptree
* will be inverted (including scanline order) when copying. This is common
* in GL when copying between window system and user-created
* renderbuffers/textures.
*/
bool
intel_miptree_blit(struct brw_context *brw,
struct intel_mipmap_tree *src_mt,
int src_level, int src_slice,
uint32_t src_x, uint32_t src_y, bool src_flip,
struct intel_mipmap_tree *dst_mt,
int dst_level, int dst_slice,
uint32_t dst_x, uint32_t dst_y, bool dst_flip,
uint32_t width, uint32_t height,
GLenum logicop)
{
/* The blitter doesn't understand multisampling at all. */
if (src_mt->num_samples > 0 || dst_mt->num_samples > 0)
return false;
/* No sRGB decode or encode is done by the hardware blitter, which is
* consistent with what we want in the callers (glCopyTexSubImage(),
* glBlitFramebuffer(), texture validation, etc.).
*/
mesa_format src_format = _mesa_get_srgb_format_linear(src_mt->format);
mesa_format dst_format = _mesa_get_srgb_format_linear(dst_mt->format);
/* The blitter doesn't support doing any format conversions. We do also
* support blitting ARGB8888 to XRGB8888 (trivial, the values dropped into
* the X channel don't matter), and XRGB8888 to ARGB8888 by setting the A
* channel to 1.0 at the end.
*/
if (src_format != dst_format &&
((src_format != MESA_FORMAT_B8G8R8A8_UNORM &&
src_format != MESA_FORMAT_B8G8R8X8_UNORM) ||
(dst_format != MESA_FORMAT_B8G8R8A8_UNORM &&
dst_format != MESA_FORMAT_B8G8R8X8_UNORM))) {
perf_debug("%s: Can't use hardware blitter from %s to %s, "
"falling back.\n", __FUNCTION__,
_mesa_get_format_name(src_format),
_mesa_get_format_name(dst_format));
return false;
}
/* According to the Ivy Bridge PRM, Vol1 Part4, section 1.2.1.2 (Graphics
* Data Size Limitations):
*
* The BLT engine is capable of transferring very large quantities of
* graphics data. Any graphics data read from and written to the
* destination is permitted to represent a number of pixels that
* occupies up to 65,536 scan lines and up to 32,768 bytes per scan line
* at the destination. The maximum number of pixels that may be
* represented per scan line’s worth of graphics data depends on the
* color depth.
*
* Furthermore, intelEmitCopyBlit (which is called below) uses a signed
* 16-bit integer to represent buffer pitch, so it can only handle buffer
* pitches < 32k.
*
* As a result of these two limitations, we can only use the blitter to do
* this copy when the miptree's pitch is less than 32k.
*/
if (src_mt->pitch >= 32768 ||
dst_mt->pitch >= 32768) {
perf_debug("Falling back due to >=32k pitch\n");
return false;
}
/* The blitter has no idea about HiZ or fast color clears, so we need to
* resolve the miptrees before we do anything.
*/
intel_miptree_slice_resolve_depth(brw, src_mt, src_level, src_slice);
intel_miptree_slice_resolve_depth(brw, dst_mt, dst_level, dst_slice);
intel_miptree_resolve_color(brw, src_mt);
intel_miptree_resolve_color(brw, dst_mt);
if (src_flip)
src_y = minify(src_mt->physical_height0, src_level - src_mt->first_level) - src_y - height;
if (dst_flip)
dst_y = minify(dst_mt->physical_height0, dst_level - dst_mt->first_level) - dst_y - height;
int src_pitch = src_mt->pitch;
if (src_flip != dst_flip)
src_pitch = -src_pitch;
uint32_t src_image_x, src_image_y;
intel_miptree_get_image_offset(src_mt, src_level, src_slice,
&src_image_x, &src_image_y);
src_x += src_image_x;
src_y += src_image_y;
/* The blitter interprets the 16-bit src x/y as a signed 16-bit value,
* where negative values are invalid. The values we're working with are
* unsigned, so make sure we don't overflow.
*/
if (src_x >= 32768 || src_y >= 32768) {
perf_debug("Falling back due to >=32k src offset (%d, %d)\n",
src_x, src_y);
return false;
}
uint32_t dst_image_x, dst_image_y;
intel_miptree_get_image_offset(dst_mt, dst_level, dst_slice,
&dst_image_x, &dst_image_y);
dst_x += dst_image_x;
dst_y += dst_image_y;
/* The blitter interprets the 16-bit destination x/y as a signed 16-bit
* value. The values we're working with are unsigned, so make sure we
* don't overflow.
*/
if (dst_x >= 32768 || dst_y >= 32768) {
perf_debug("Falling back due to >=32k dst offset (%d, %d)\n",
dst_x, dst_y);
return false;
}
if (!intelEmitCopyBlit(brw,
src_mt->cpp,
src_pitch,
src_mt->bo, src_mt->offset,
src_mt->tiling,
dst_mt->pitch,
dst_mt->bo, dst_mt->offset,
dst_mt->tiling,
src_x, src_y,
dst_x, dst_y,
width, height,
logicop)) {
return false;
}
if (src_mt->format == MESA_FORMAT_B8G8R8X8_UNORM &&
dst_mt->format == MESA_FORMAT_B8G8R8A8_UNORM) {
intel_miptree_set_alpha_to_one(brw, dst_mt,
dst_x, dst_y,
width, height);
}
return true;
}
/**
* Implements a rectangular block transfer (blit) of pixels between two
* miptrees.
*
* Our blitter can operate on 1, 2, or 4-byte-per-pixel data, with generous,
* but limited, pitches and sizes allowed.
*
* The src/dst coordinates are relative to the given level/slice of the
* miptree.
*
* If @src_flip or @dst_flip is set, then the rectangle within that miptree
* will be inverted (including scanline order) when copying. This is common
* in GL when copying between window system and user-created
* renderbuffers/textures.
*/
bool
intel_miptree_blit(struct intel_context *intel,
struct intel_mipmap_tree *src_mt,
int src_level, int src_slice,
uint32_t src_x, uint32_t src_y, bool src_flip,
struct intel_mipmap_tree *dst_mt,
int dst_level, int dst_slice,
uint32_t dst_x, uint32_t dst_y, bool dst_flip,
uint32_t width, uint32_t height,
GLenum logicop)
{
/* No sRGB decode or encode is done by the hardware blitter, which is
* consistent with what we want in the callers (glCopyTexSubImage(),
* glBlitFramebuffer(), texture validation, etc.).
*/
gl_format src_format = _mesa_get_srgb_format_linear(src_mt->format);
gl_format dst_format = _mesa_get_srgb_format_linear(dst_mt->format);
/* The blitter doesn't support doing any format conversions. We do also
* support blitting ARGB8888 to XRGB8888 (trivial, the values dropped into
* the X channel don't matter), and XRGB8888 to ARGB8888 by setting the A
* channel to 1.0 at the end.
*/
if (src_format != dst_format &&
((src_format != MESA_FORMAT_ARGB8888 &&
src_format != MESA_FORMAT_XRGB8888) ||
(dst_format != MESA_FORMAT_ARGB8888 &&
dst_format != MESA_FORMAT_XRGB8888))) {
perf_debug("%s: Can't use hardware blitter from %s to %s, "
"falling back.\n", __FUNCTION__,
_mesa_get_format_name(src_format),
_mesa_get_format_name(dst_format));
return false;
}
/* According to the Ivy Bridge PRM, Vol1 Part4, section 1.2.1.2 (Graphics
* Data Size Limitations):
*
* The BLT engine is capable of transferring very large quantities of
* graphics data. Any graphics data read from and written to the
* destination is permitted to represent a number of pixels that
* occupies up to 65,536 scan lines and up to 32,768 bytes per scan line
* at the destination. The maximum number of pixels that may be
* represented per scan line’s worth of graphics data depends on the
* color depth.
*
* Furthermore, intelEmitCopyBlit (which is called below) uses a signed
* 16-bit integer to represent buffer pitch, so it can only handle buffer
* pitches < 32k.
*
* As a result of these two limitations, we can only use the blitter to do
* this copy when the region's pitch is less than 32k.
*/
if (src_mt->region->pitch > 32768 ||
dst_mt->region->pitch > 32768) {
perf_debug("Falling back due to >32k pitch\n");
return false;
}
if (src_flip)
src_y = src_mt->level[src_level].height - src_y - height;
if (dst_flip)
dst_y = dst_mt->level[dst_level].height - dst_y - height;
int src_pitch = src_mt->region->pitch;
if (src_flip != dst_flip)
src_pitch = -src_pitch;
uint32_t src_image_x, src_image_y;
intel_miptree_get_image_offset(src_mt, src_level, src_slice,
&src_image_x, &src_image_y);
src_x += src_image_x;
src_y += src_image_y;
uint32_t dst_image_x, dst_image_y;
intel_miptree_get_image_offset(dst_mt, dst_level, dst_slice,
&dst_image_x, &dst_image_y);
dst_x += dst_image_x;
dst_y += dst_image_y;
if (!intelEmitCopyBlit(intel,
src_mt->cpp,
src_pitch,
src_mt->region->bo, src_mt->offset,
src_mt->region->tiling,
dst_mt->region->pitch,
dst_mt->region->bo, dst_mt->offset,
dst_mt->region->tiling,
src_x, src_y,
dst_x, dst_y,
width, height,
logicop)) {
return false;
}
if (src_mt->format == MESA_FORMAT_XRGB8888 &&
dst_mt->format == MESA_FORMAT_ARGB8888) {
intel_miptree_set_alpha_to_one(intel, dst_mt,
dst_x, dst_y,
width, height);
}
return true;
}
/**
* Implements a rectangular block transfer (blit) of pixels between two
* miptrees.
*
* Our blitter can operate on 1, 2, or 4-byte-per-pixel data, with generous,
* but limited, pitches and sizes allowed.
*
* The src/dst coordinates are relative to the given level/slice of the
* miptree.
*
* If @src_flip or @dst_flip is set, then the rectangle within that miptree
* will be inverted (including scanline order) when copying. This is common
* in GL when copying between window system and user-created
* renderbuffers/textures.
*/
bool
intel_miptree_blit(struct brw_context *brw,
struct intel_mipmap_tree *src_mt,
int src_level, int src_slice,
uint32_t src_x, uint32_t src_y, bool src_flip,
struct intel_mipmap_tree *dst_mt,
int dst_level, int dst_slice,
uint32_t dst_x, uint32_t dst_y, bool dst_flip,
uint32_t width, uint32_t height,
enum gl_logicop_mode logicop)
{
/* The blitter doesn't understand multisampling at all. */
if (src_mt->surf.samples > 1 || dst_mt->surf.samples > 1)
return false;
/* No sRGB decode or encode is done by the hardware blitter, which is
* consistent with what we want in many callers (glCopyTexSubImage(),
* texture validation, etc.).
*/
mesa_format src_format = _mesa_get_srgb_format_linear(src_mt->format);
mesa_format dst_format = _mesa_get_srgb_format_linear(dst_mt->format);
/* The blitter doesn't support doing any format conversions. We do also
* support blitting ARGB8888 to XRGB8888 (trivial, the values dropped into
* the X channel don't matter), and XRGB8888 to ARGB8888 by setting the A
* channel to 1.0 at the end. Also trivially ARGB2101010 to XRGB2101010,
* but not XRGB2101010 to ARGB2101010 yet.
*/
if (!intel_miptree_blit_compatible_formats(src_format, dst_format)) {
perf_debug("%s: Can't use hardware blitter from %s to %s, "
"falling back.\n", __func__,
_mesa_get_format_name(src_format),
_mesa_get_format_name(dst_format));
return false;
}
/* The blitter has no idea about HiZ or fast color clears, so we need to
* resolve the miptrees before we do anything.
*/
intel_miptree_access_raw(brw, src_mt, src_level, src_slice, false);
intel_miptree_access_raw(brw, dst_mt, dst_level, dst_slice, true);
if (src_flip) {
const unsigned h0 = src_mt->surf.phys_level0_sa.height;
src_y = minify(h0, src_level - src_mt->first_level) - src_y - height;
}
if (dst_flip) {
const unsigned h0 = dst_mt->surf.phys_level0_sa.height;
dst_y = minify(h0, dst_level - dst_mt->first_level) - dst_y - height;
}
uint32_t src_image_x, src_image_y, dst_image_x, dst_image_y;
intel_miptree_get_image_offset(src_mt, src_level, src_slice,
&src_image_x, &src_image_y);
intel_miptree_get_image_offset(dst_mt, dst_level, dst_slice,
&dst_image_x, &dst_image_y);
src_x += src_image_x;
src_y += src_image_y;
dst_x += dst_image_x;
dst_y += dst_image_y;
if (!emit_miptree_blit(brw, src_mt, src_x, src_y,
dst_mt, dst_x, dst_y, width, height,
src_flip != dst_flip, logicop)) {
return false;
}
/* XXX This could be done in a single pass using XY_FULL_MONO_PATTERN_BLT */
if (_mesa_get_format_bits(src_format, GL_ALPHA_BITS) == 0 &&
_mesa_get_format_bits(dst_format, GL_ALPHA_BITS) > 0) {
intel_miptree_set_alpha_to_one(brw, dst_mt,
dst_x, dst_y,
width, height);
}
return true;
}
