diff options
Diffstat (limited to 'src/gallium/auxiliary')
-rw-r--r-- | src/gallium/auxiliary/Makefile | 1 | ||||
-rw-r--r-- | src/gallium/auxiliary/SConscript | 1 | ||||
-rw-r--r-- | src/gallium/auxiliary/draw/draw_pipe_aapoint.c | 2 | ||||
-rw-r--r-- | src/gallium/auxiliary/tgsi/tgsi_exec.c | 47 | ||||
-rw-r--r-- | src/gallium/auxiliary/tgsi/tgsi_exec.h | 6 | ||||
-rw-r--r-- | src/gallium/auxiliary/util/u_format_pack.py | 2 | ||||
-rw-r--r-- | src/gallium/auxiliary/util/u_format_rgtc.c | 452 | ||||
-rw-r--r-- | src/gallium/auxiliary/util/u_format_rgtc.h | 108 | ||||
-rwxr-xr-x | src/gallium/auxiliary/util/u_format_table.py | 3 | ||||
-rw-r--r-- | src/gallium/auxiliary/util/u_math.h | 11 | ||||
-rw-r--r-- | src/gallium/auxiliary/util/u_vbuf_mgr.c | 10 |
11 files changed, 631 insertions, 12 deletions
diff --git a/src/gallium/auxiliary/Makefile b/src/gallium/auxiliary/Makefile index 7d7d700eacd..e48063166ff 100644 --- a/src/gallium/auxiliary/Makefile +++ b/src/gallium/auxiliary/Makefile @@ -111,6 +111,7 @@ C_SOURCES = \ util/u_format.c \ util/u_format_other.c \ util/u_format_s3tc.c \ + util/u_format_rgtc.c \ util/u_format_srgb.c \ util/u_format_table.c \ util/u_format_tests.c \ diff --git a/src/gallium/auxiliary/SConscript b/src/gallium/auxiliary/SConscript index 0ec63071615..ef0b15f06ce 100644 --- a/src/gallium/auxiliary/SConscript +++ b/src/gallium/auxiliary/SConscript @@ -158,6 +158,7 @@ source = [ 'util/u_format.c', 'util/u_format_other.c', 'util/u_format_s3tc.c', + 'util/u_format_rgtc.c', 'util/u_format_srgb.c', 'util/u_format_table.c', 'util/u_format_tests.c', diff --git a/src/gallium/auxiliary/draw/draw_pipe_aapoint.c b/src/gallium/auxiliary/draw/draw_pipe_aapoint.c index 5ea552f51c1..60f6380c503 100644 --- a/src/gallium/auxiliary/draw/draw_pipe_aapoint.c +++ b/src/gallium/auxiliary/draw/draw_pipe_aapoint.c @@ -874,6 +874,8 @@ draw_install_aapoint_stage(struct draw_context *draw, { struct aapoint_stage *aapoint; + pipe->draw = (void *) draw; + /* * Create / install AA point drawing / prim stage */ diff --git a/src/gallium/auxiliary/tgsi/tgsi_exec.c b/src/gallium/auxiliary/tgsi/tgsi_exec.c index cbb090b2803..9cf74a838fe 100644 --- a/src/gallium/auxiliary/tgsi/tgsi_exec.c +++ b/src/gallium/auxiliary/tgsi/tgsi_exec.c @@ -672,6 +672,31 @@ tgsi_exec_machine_bind_shader( mach->Processor = parse.FullHeader.Processor.Processor; mach->ImmLimit = 0; + if (mach->Processor == TGSI_PROCESSOR_GEOMETRY && + !mach->UsedGeometryShader) { + struct tgsi_exec_vector *inputs = + align_malloc(sizeof(struct tgsi_exec_vector) * + TGSI_MAX_PRIM_VERTICES * PIPE_MAX_ATTRIBS, + 16); + struct tgsi_exec_vector *outputs = + align_malloc(sizeof(struct tgsi_exec_vector) * + TGSI_MAX_TOTAL_VERTICES, 16); + + if (!inputs) + return; + if (!outputs) { + align_free(inputs); + return; + } + + align_free(mach->Inputs); + align_free(mach->Outputs); + + mach->Inputs = inputs; + mach->Outputs = outputs; + mach->UsedGeometryShader = TRUE; + } + declarations = (struct tgsi_full_declaration *) MALLOC( maxDeclarations * sizeof(struct tgsi_full_declaration) ); @@ -801,6 +826,11 @@ tgsi_exec_machine_create( void ) mach->MaxGeometryShaderOutputs = TGSI_MAX_TOTAL_VERTICES; mach->Predicates = &mach->Temps[TGSI_EXEC_TEMP_P0]; + mach->Inputs = align_malloc(sizeof(struct tgsi_exec_vector) * PIPE_MAX_ATTRIBS, 16); + mach->Outputs = align_malloc(sizeof(struct tgsi_exec_vector) * PIPE_MAX_ATTRIBS, 16); + if (!mach->Inputs || !mach->Outputs) + goto fail; + /* Setup constants needed by the SSE2 executor. */ for( i = 0; i < 4; i++ ) { mach->Temps[TGSI_EXEC_TEMP_00000000_I].xyzw[TGSI_EXEC_TEMP_00000000_C].u[i] = 0x00000000; @@ -824,7 +854,11 @@ tgsi_exec_machine_create( void ) return mach; fail: - align_free(mach); + if (mach) { + align_free(mach->Inputs); + align_free(mach->Outputs); + align_free(mach); + } return NULL; } @@ -836,10 +870,13 @@ tgsi_exec_machine_destroy(struct tgsi_exec_machine *mach) if (mach->Instructions) FREE(mach->Instructions); if (mach->Declarations) - FREE(mach->Declarations); - } + FREE(mach->Declarations); + + align_free(mach->Inputs); + align_free(mach->Outputs); - align_free(mach); + align_free(mach); + } } static void @@ -1047,7 +1084,7 @@ fetch_src_file_channel(const struct tgsi_exec_machine *mach, }*/ int pos = index2D->i[i] * TGSI_EXEC_MAX_INPUT_ATTRIBS + index->i[i]; assert(pos >= 0); - assert(pos < Elements(mach->Inputs)); + assert(pos < TGSI_MAX_PRIM_VERTICES * PIPE_MAX_ATTRIBS); chan->u[i] = mach->Inputs[pos].xyzw[swizzle].u[i]; } break; diff --git a/src/gallium/auxiliary/tgsi/tgsi_exec.h b/src/gallium/auxiliary/tgsi/tgsi_exec.h index 4a423b5bb4e..33f33aa82c7 100644 --- a/src/gallium/auxiliary/tgsi/tgsi_exec.h +++ b/src/gallium/auxiliary/tgsi/tgsi_exec.h @@ -228,8 +228,8 @@ struct tgsi_exec_machine float ImmArray[TGSI_EXEC_NUM_IMMEDIATES][4]; - struct tgsi_exec_vector Inputs[TGSI_MAX_PRIM_VERTICES * PIPE_MAX_ATTRIBS]; - struct tgsi_exec_vector Outputs[TGSI_MAX_TOTAL_VERTICES]; + struct tgsi_exec_vector *Inputs; + struct tgsi_exec_vector *Outputs; /* System values */ unsigned SysSemanticToIndex[TGSI_SEMANTIC_COUNT]; @@ -309,6 +309,8 @@ struct tgsi_exec_machine uint NumDeclarations; struct tgsi_declaration_resource Resources[PIPE_MAX_SHADER_RESOURCES]; + + boolean UsedGeometryShader; }; struct tgsi_exec_machine * diff --git a/src/gallium/auxiliary/util/u_format_pack.py b/src/gallium/auxiliary/util/u_format_pack.py index 6d0016c0ad8..cc173f808ae 100644 --- a/src/gallium/auxiliary/util/u_format_pack.py +++ b/src/gallium/auxiliary/util/u_format_pack.py @@ -632,7 +632,7 @@ def generate_format_fetch(format, dst_channel, dst_native_type, dst_suffix): def is_format_hand_written(format): - return format.layout in ('s3tc', 'subsampled', 'other') or format.colorspace == ZS + return format.layout in ('s3tc', 'rgtc', 'subsampled', 'other') or format.colorspace == ZS def generate(formats): diff --git a/src/gallium/auxiliary/util/u_format_rgtc.c b/src/gallium/auxiliary/util/u_format_rgtc.c new file mode 100644 index 00000000000..6ffcd7e99ef --- /dev/null +++ b/src/gallium/auxiliary/util/u_format_rgtc.c @@ -0,0 +1,452 @@ +/************************************************************************** + * + * Copyright (C) 2011 Red Hat Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included + * in all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS + * OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * BRIAN PAUL BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN + * AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN + * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE. + * + **************************************************************************/ + +#include <stdio.h> +#include "u_math.h" +#include "u_format.h" +#include "u_format_rgtc.h" + +static void u_format_unsigned_encode_rgtc_chan(uint8_t *blkaddr, uint8_t srccolors[4][4], + int numxpixels, int numypixels); + +static void u_format_unsigned_fetch_texel_rgtc(unsigned srcRowStride, const uint8_t *pixdata, + unsigned i, unsigned j, uint8_t *value, unsigned comps); + +static void u_format_signed_encode_rgtc_chan(int8_t *blkaddr, int8_t srccolors[4][4], + int numxpixels, int numypixels); + +static void u_format_signed_fetch_texel_rgtc(unsigned srcRowStride, const int8_t *pixdata, + unsigned i, unsigned j, int8_t *value, unsigned comps); + +void +util_format_rgtc1_unorm_fetch_rgba_8unorm(uint8_t *dst, const uint8_t *src, unsigned i, unsigned j) +{ + u_format_unsigned_fetch_texel_rgtc(0, src, i, j, dst, 1); +} + +void +util_format_rgtc1_unorm_unpack_rgba_8unorm(uint8_t *dst_row, unsigned dst_stride, const uint8_t *src_row, unsigned src_stride, unsigned width, unsigned height) +{ + const unsigned bw = 4, bh = 4, comps = 4; + unsigned x, y, i, j; + unsigned block_size = 8; + + for(y = 0; y < height; y += bh) { + const uint8_t *src = src_row; + for(x = 0; x < width; x += bw) { + for(j = 0; j < bh; ++j) { + for(i = 0; i < bw; ++i) { + uint8_t *dst = dst_row + (y + j)*dst_stride/sizeof(*dst_row) + (x + i)*comps; + u_format_unsigned_fetch_texel_rgtc(0, src, i, j, dst, 1); + } + } + src += block_size; + } + src_row += src_stride; + } +} + +void +util_format_rgtc1_unorm_pack_rgba_8unorm(uint8_t *dst_row, unsigned dst_stride, const uint8_t *src_row, + unsigned src_stride, unsigned width, unsigned height) +{ + const unsigned bw = 4, bh = 4, bytes_per_block = 8; + unsigned x, y, i, j; + + for(y = 0; y < height; y += bh) { + uint8_t *dst = dst_row; + for(x = 0; x < width; x += bw) { + uint8_t tmp[4][4]; /* [bh][bw][comps] */ + for(j = 0; j < bh; ++j) { + for(i = 0; i < bw; ++i) { + tmp[j][i] = src_row[(y + j)*src_stride/sizeof(*src_row) + (x + i)*4]; + } + } + u_format_unsigned_encode_rgtc_chan(dst, tmp, 4, 4); + dst += bytes_per_block; + } + dst_row += dst_stride / sizeof(*dst_row); + } +} + +void +util_format_rgtc1_unorm_unpack_rgba_float(float *dst_row, unsigned dst_stride, const uint8_t *src_row, unsigned src_stride, unsigned width, unsigned height) +{ + unsigned x, y, i, j; + int block_size = 8; + for(y = 0; y < height; y += 4) { + const uint8_t *src = src_row; + for(x = 0; x < width; x += 4) { + for(j = 0; j < 4; ++j) { + for(i = 0; i < 4; ++i) { + float *dst = dst_row + (y + j)*dst_stride/sizeof(*dst_row) + (x + i)*4; + uint8_t tmp_r; + u_format_unsigned_fetch_texel_rgtc(0, src, i, j, &tmp_r, 1); + dst[0] = ubyte_to_float(tmp_r); + dst[1] = 0.0; + dst[2] = 0.0; + dst[3] = 1.0; + } + } + src += block_size; + } + src_row += src_stride; + } +} + +void +util_format_rgtc1_unorm_pack_rgba_float(uint8_t *dst_row, unsigned dst_stride, const float *src_row, unsigned src_stride, unsigned width, unsigned height) +{ + const unsigned bw = 4, bh = 4, bytes_per_block = 8; + unsigned x, y, i, j; + + for(y = 0; y < height; y += bh) { + uint8_t *dst = dst_row; + for(x = 0; x < width; x += bw) { + uint8_t tmp[4][4]; /* [bh][bw][comps] */ + for(j = 0; j < bh; ++j) { + for(i = 0; i < bw; ++i) { + tmp[j][i] = float_to_ubyte(src_row[(y + j)*src_stride/sizeof(*src_row) + (x + i)*4]); + } + } + u_format_unsigned_encode_rgtc_chan(dst, tmp, 4, 4); + dst += bytes_per_block; + } + dst_row += dst_stride / sizeof(*dst_row); + } +} + +void +util_format_rgtc1_unorm_fetch_rgba_float(float *dst, const uint8_t *src, unsigned i, unsigned j) +{ + uint8_t tmp_r; + u_format_unsigned_fetch_texel_rgtc(0, src, i, j, &tmp_r, 1); + dst[0] = ubyte_to_float(tmp_r); + dst[1] = 0.0; + dst[2] = 0.0; + dst[3] = 1.0; +} + +void +util_format_rgtc1_snorm_fetch_rgba_8unorm(uint8_t *dst, const uint8_t *src, unsigned i, unsigned j) +{ + fprintf(stderr,"%s\n", __func__); +} + +void +util_format_rgtc1_snorm_unpack_rgba_8unorm(uint8_t *dst_row, unsigned dst_stride, const uint8_t *src_row, unsigned src_stride, unsigned width, unsigned height) +{ + fprintf(stderr,"%s\n", __func__); +} + +void +util_format_rgtc1_snorm_pack_rgba_8unorm(uint8_t *dst_row, unsigned dst_stride, const uint8_t *src_row, unsigned src_stride, unsigned width, unsigned height) +{ + fprintf(stderr,"%s\n", __func__); +} + +void +util_format_rgtc1_snorm_pack_rgba_float(uint8_t *dst_row, unsigned dst_stride, const float *src_row, unsigned src_stride, unsigned width, unsigned height) +{ + const unsigned bw = 4, bh = 4, bytes_per_block = 8; + unsigned x, y, i, j; + + for(y = 0; y < height; y += bh) { + int8_t *dst = (int8_t *)dst_row; + for(x = 0; x < width; x += bw) { + int8_t tmp[4][4]; /* [bh][bw][comps] */ + for(j = 0; j < bh; ++j) { + for(i = 0; i < bw; ++i) { + tmp[j][i] = float_to_byte_tex(src_row[(y + j)*src_stride/sizeof(*src_row) + (x + i)*4]); + } + } + u_format_signed_encode_rgtc_chan(dst, tmp, 4, 4); + dst += bytes_per_block; + } + dst_row += dst_stride / sizeof(*dst_row); + } +} + +void +util_format_rgtc1_snorm_unpack_rgba_float(float *dst_row, unsigned dst_stride, const uint8_t *src_row, unsigned src_stride, unsigned width, unsigned height) +{ + unsigned x, y, i, j; + int block_size = 8; + for(y = 0; y < height; y += 4) { + const int8_t *src = (int8_t *)src_row; + for(x = 0; x < width; x += 4) { + for(j = 0; j < 4; ++j) { + for(i = 0; i < 4; ++i) { + float *dst = dst_row + (y + j)*dst_stride/sizeof(*dst_row) + (x + i)*4; + int8_t tmp_r; + u_format_signed_fetch_texel_rgtc(0, src, i, j, &tmp_r, 1); + dst[0] = byte_to_float_tex(tmp_r); + dst[1] = 0.0; + dst[2] = 0.0; + dst[3] = 1.0; + } + } + src += block_size; + } + src_row += src_stride; + } +} + +void +util_format_rgtc1_snorm_fetch_rgba_float(float *dst, const uint8_t *src, unsigned i, unsigned j) +{ + int8_t tmp_r; + u_format_signed_fetch_texel_rgtc(0, (int8_t *)src, i, j, &tmp_r, 1); + dst[0] = byte_to_float_tex(tmp_r); + dst[1] = 0.0; + dst[2] = 0.0; + dst[3] = 1.0; +} + + +void +util_format_rgtc2_unorm_fetch_rgba_8unorm(uint8_t *dst, const uint8_t *src, unsigned i, unsigned j) +{ + u_format_unsigned_fetch_texel_rgtc(0, src, i, j, dst, 2); + u_format_unsigned_fetch_texel_rgtc(0, src + 8, i, j, dst + 1, 2); +} + +void +util_format_rgtc2_unorm_unpack_rgba_8unorm(uint8_t *dst_row, unsigned dst_stride, const uint8_t *src_row, unsigned src_stride, unsigned width, unsigned height) +{ + const unsigned bw = 4, bh = 4, comps = 4; + unsigned x, y, i, j; + unsigned block_size = 16; + + for(y = 0; y < height; y += bh) { + const uint8_t *src = src_row; + for(x = 0; x < width; x += bw) { + for(j = 0; j < bh; ++j) { + for(i = 0; i < bw; ++i) { + uint8_t *dst = dst_row + (y + j)*dst_stride/sizeof(*dst_row) + (x + i)*comps; + u_format_unsigned_fetch_texel_rgtc(0, src, i, j, dst, 2); + u_format_unsigned_fetch_texel_rgtc(0, src + 8, i, j, dst + 1, 2); + + } + } + src += block_size; + } + src_row += src_stride; + } +} + +void +util_format_rgtc2_unorm_pack_rgba_8unorm(uint8_t *dst_row, unsigned dst_stride, const uint8_t *src_row, unsigned src_stride, unsigned width, unsigned height) +{ + const unsigned bw = 4, bh = 4, bytes_per_block = 16; + unsigned x, y, i, j; + + for(y = 0; y < height; y += bh) { + uint8_t *dst = dst_row; + for(x = 0; x < width; x += bw) { + uint8_t tmp_r[4][4]; /* [bh][bw] */ + uint8_t tmp_g[4][4]; /* [bh][bw] */ + for(j = 0; j < bh; ++j) { + for(i = 0; i < bw; ++i) { + tmp_r[j][i] = src_row[(y + j)*src_stride/sizeof(*src_row) + (x + i)*4]; + tmp_g[j][i] = src_row[((y + j)*src_stride/sizeof(*src_row) + (x + i)*4) + 1]; + } + } + u_format_unsigned_encode_rgtc_chan(dst, tmp_r, 4, 4); + u_format_unsigned_encode_rgtc_chan(dst + 8, tmp_g, 4, 4); + dst += bytes_per_block; + } + dst_row += dst_stride / sizeof(*dst_row); + } +} + +void +util_format_rgtc2_unorm_pack_rgba_float(uint8_t *dst_row, unsigned dst_stride, const float *src_row, unsigned src_stride, unsigned width, unsigned height) +{ + const unsigned bw = 4, bh = 4, bytes_per_block = 16; + unsigned x, y, i, j; + + for(y = 0; y < height; y += bh) { + uint8_t *dst = dst_row; + for(x = 0; x < width; x += bw) { + uint8_t tmp_r[4][4]; /* [bh][bw][comps] */ + uint8_t tmp_g[4][4]; /* [bh][bw][comps] */ + for(j = 0; j < bh; ++j) { + for(i = 0; i < bw; ++i) { + tmp_r[j][i] = float_to_ubyte(src_row[(y + j)*src_stride/sizeof(*src_row) + (x + i)*4]); + tmp_g[j][i] = float_to_ubyte(src_row[(y + j)*src_stride/sizeof(*src_row) + (x + i)*4 + 1]); + } + } + u_format_unsigned_encode_rgtc_chan(dst, tmp_r, 4, 4); + u_format_unsigned_encode_rgtc_chan(dst + 8, tmp_g, 4, 4); + dst += bytes_per_block; + } + dst_row += dst_stride / sizeof(*dst_row); + } +} + +void +util_format_rgtc2_unorm_unpack_rgba_float(float *dst_row, unsigned dst_stride, const uint8_t *src_row, unsigned src_stride, unsigned width, unsigned height) +{ + unsigned x, y, i, j; + int block_size = 16; + for(y = 0; y < height; y += 4) { + const uint8_t *src = src_row; + for(x = 0; x < width; x += 4) { + for(j = 0; j < 4; ++j) { + for(i = 0; i < 4; ++i) { + float *dst = dst_row + (y + j)*dst_stride/sizeof(*dst_row) + (x + i)*4; + uint8_t tmp_r, tmp_g; + u_format_unsigned_fetch_texel_rgtc(0, src, i, j, &tmp_r, 2); + u_format_unsigned_fetch_texel_rgtc(0, src + 8, i, j, &tmp_g, 2); + dst[0] = ubyte_to_float(tmp_r); + dst[1] = ubyte_to_float(tmp_g); + dst[2] = 0.0; + dst[3] = 1.0; + } + } + src += block_size; + } + src_row += src_stride; + } +} + +void +util_format_rgtc2_unorm_fetch_rgba_float(float *dst, const uint8_t *src, unsigned i, unsigned j) +{ + uint8_t tmp_r, tmp_g; + u_format_unsigned_fetch_texel_rgtc(0, src, i, j, &tmp_r, 2); + u_format_unsigned_fetch_texel_rgtc(0, src + 8, i, j, &tmp_g, 2); + dst[0] = ubyte_to_float(tmp_r); + dst[1] = ubyte_to_float(tmp_g); + dst[2] = 0.0; + dst[3] = 1.0; +} + + +void +util_format_rgtc2_snorm_fetch_rgba_8unorm(uint8_t *dst, const uint8_t *src, unsigned i, unsigned j) +{ + fprintf(stderr,"%s\n", __func__); +} + +void +util_format_rgtc2_snorm_unpack_rgba_8unorm(uint8_t *dst_row, unsigned dst_stride, const uint8_t *src_row, unsigned src_stride, unsigned width, unsigned height) +{ + fprintf(stderr,"%s\n", __func__); +} + +void +util_format_rgtc2_snorm_pack_rgba_8unorm(uint8_t *dst_row, unsigned dst_stride, const uint8_t *src_row, unsigned src_stride, unsigned width, unsigned height) +{ + fprintf(stderr,"%s\n", __func__); +} + +void +util_format_rgtc2_snorm_unpack_rgba_float(float *dst_row, unsigned dst_stride, const uint8_t *src_row, unsigned src_stride, unsigned width, unsigned height) +{ + unsigned x, y, i, j; + int block_size = 16; + for(y = 0; y < height; y += 4) { + const int8_t *src = (int8_t *)src_row; + for(x = 0; x < width; x += 4) { + for(j = 0; j < 4; ++j) { + for(i = 0; i < 4; ++i) { + float *dst = dst_row + (y + j)*dst_stride/sizeof(*dst_row) + (x + i)*4; + int8_t tmp_r, tmp_g; + u_format_signed_fetch_texel_rgtc(0, src, i, j, &tmp_r, 2); + u_format_signed_fetch_texel_rgtc(0, src + 8, i, j, &tmp_g, 2); + dst[0] = byte_to_float_tex(tmp_r); + dst[1] = byte_to_float_tex(tmp_g); + dst[2] = 0.0; + dst[3] = 1.0; + } + } + src += block_size; + } + src_row += src_stride; + } +} + +void +util_format_rgtc2_snorm_pack_rgba_float(uint8_t *dst_row, unsigned dst_stride, const float *src_row, unsigned src_stride, unsigned width, unsigned height) +{ + const unsigned bw = 4, bh = 4, bytes_per_block = 16; + unsigned x, y, i, j; + + for(y = 0; y < height; y += bh) { + int8_t *dst = (int8_t *)dst_row; + for(x = 0; x < width; x += bw) { + int8_t tmp_r[4][4]; /* [bh][bw][comps] */ + int8_t tmp_g[4][4]; /* [bh][bw][comps] */ + for(j = 0; j < bh; ++j) { + for(i = 0; i < bw; ++i) { + tmp_r[j][i] = float_to_byte_tex(src_row[(y + j)*src_stride/sizeof(*src_row) + (x + i)*4]); + tmp_g[j][i] = float_to_byte_tex(src_row[(y + j)*src_stride/sizeof(*src_row) + (x + i)*4 + 1]); + } + } + u_format_signed_encode_rgtc_chan(dst, tmp_r, 4, 4); + u_format_signed_encode_rgtc_chan(dst + 8, tmp_g, 4, 4); + dst += bytes_per_block; + } + dst_row += dst_stride / sizeof(*dst_row); + } +} + +void +util_format_rgtc2_snorm_fetch_rgba_float(float *dst, const uint8_t *src, unsigned i, unsigned j) +{ + int8_t tmp_r, tmp_g; + u_format_signed_fetch_texel_rgtc(0, (int8_t *)src, i, j, &tmp_r, 2); + u_format_signed_fetch_texel_rgtc(0, (int8_t *)src + 8, i, j, &tmp_g, 2); + dst[0] = byte_to_float_tex(tmp_r); + dst[1] = byte_to_float_tex(tmp_g); + dst[2] = 0.0; + dst[3] = 1.0; +} + + +#define TAG(x) u_format_unsigned_##x +#define TYPE uint8_t +#define T_MIN 0 +#define T_MAX 255 + +#include "../../../mesa/main/texcompress_rgtc_tmp.h" + +#undef TYPE +#undef TAG +#undef T_MIN +#undef T_MAX + + +#define TAG(x) u_format_signed_##x +#define TYPE int8_t +#define T_MIN (int8_t)-128 +#define T_MAX (int8_t)127 + +#include "../../../mesa/main/texcompress_rgtc_tmp.h" + +#undef TYPE +#undef TAG +#undef T_MIN +#undef T_MAX diff --git a/src/gallium/auxiliary/util/u_format_rgtc.h b/src/gallium/auxiliary/util/u_format_rgtc.h new file mode 100644 index 00000000000..3e8636d110c --- /dev/null +++ b/src/gallium/auxiliary/util/u_format_rgtc.h @@ -0,0 +1,108 @@ +/************************************************************************** + * + * Copyright 2011 Red Hat Inc. + * All Rights Reserved. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the + * "Software"), to deal in the Software without restriction, including + * without limitation the rights to use, copy, modify, merge, publish, + * distribute, sub license, and/or sell copies of the Software, and to + * permit persons to whom the Software is furnished to do so, subject to + * the following conditions: + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NON-INFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDERS, AUTHORS AND/OR ITS SUPPLIERS BE LIABLE FOR ANY CLAIM, + * DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR + * OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE + * USE OR OTHER DEALINGS IN THE SOFTWARE. + * + * The above copyright notice and this permission notice (including the + * next paragraph) shall be included in all copies or substantial portions + * of the Software. + * + **************************************************************************/ + +#ifndef U_FORMAT_RGTC_H_ +#define U_FORMAT_RGTC_H_ + +void +util_format_rgtc1_unorm_fetch_rgba_8unorm(uint8_t *dst, const uint8_t *src, unsigned i, unsigned j); + +void +util_format_rgtc1_unorm_unpack_rgba_8unorm(uint8_t *dst_row, unsigned dst_stride, const uint8_t *src_row, unsigned src_stride, unsigned width, unsigned height); + +void +util_format_rgtc1_unorm_pack_rgba_8unorm(uint8_t *dst_row, unsigned dst_stride, const uint8_t *src_row, unsigned src_stride, unsigned width, unsigned height); + +void +util_format_rgtc1_unorm_unpack_rgba_float(float *dst_row, unsigned dst_stride, const uint8_t *src_row, unsigned src_stride, unsigned width, unsigned height); + +void +util_format_rgtc1_unorm_pack_rgba_float(uint8_t *dst_row, unsigned dst_stride, const float *src_row, unsigned src_stride, unsigned width, unsigned height); + +void +util_format_rgtc1_unorm_fetch_rgba_float(float *dst, const uint8_t *src, unsigned i, unsigned j); + + + +void +util_format_rgtc1_snorm_fetch_rgba_8unorm(uint8_t *dst, const uint8_t *src, unsigned i, unsigned j); + +void +util_format_rgtc1_snorm_unpack_rgba_8unorm(uint8_t *dst_row, unsigned dst_stride, const uint8_t *src_row, unsigned src_stride, unsigned width, unsigned height); + +void +util_format_rgtc1_snorm_pack_rgba_8unorm(uint8_t *dst_row, unsigned dst_stride, const uint8_t *src_row, unsigned src_stride, unsigned width, unsigned height); + +void +util_format_rgtc1_snorm_pack_rgba_float(uint8_t *dst_row, unsigned dst_stride, const float *src_row, unsigned src_stride, unsigned width, unsigned height); + +void +util_format_rgtc1_snorm_unpack_rgba_float(float *dst_row, unsigned dst_stride, const uint8_t *src_row, unsigned src_stride, unsigned width, unsigned height); + +void +util_format_rgtc1_snorm_fetch_rgba_float(float *dst, const uint8_t *src, unsigned i, unsigned j); + + +void +util_format_rgtc2_unorm_fetch_rgba_8unorm(uint8_t *dst, const uint8_t *src, unsigned i, unsigned j); + +void +util_format_rgtc2_unorm_unpack_rgba_8unorm(uint8_t *dst_row, unsigned dst_stride, const uint8_t *src_row, unsigned src_stride, unsigned width, unsigned height); + +void +util_format_rgtc2_unorm_pack_rgba_8unorm(uint8_t *dst_row, unsigned dst_stride, const uint8_t *src_row, unsigned src_stride, unsigned width, unsigned height); + +void +util_format_rgtc2_unorm_pack_rgba_float(uint8_t *dst_row, unsigned dst_stride, const float *src_row, unsigned src_stride, unsigned width, unsigned height); + +void +util_format_rgtc2_unorm_unpack_rgba_float(float *dst_row, unsigned dst_stride, const uint8_t *src_row, unsigned src_stride, unsigned width, unsigned height); + +void +util_format_rgtc2_unorm_fetch_rgba_float(float *dst, const uint8_t *src, unsigned i, unsigned j); + + +void +util_format_rgtc2_snorm_fetch_rgba_8unorm(uint8_t *dst, const uint8_t *src, unsigned i, unsigned j); + +void +util_format_rgtc2_snorm_unpack_rgba_8unorm(uint8_t *dst_row, unsigned dst_stride, const uint8_t *src_row, unsigned src_stride, unsigned width, unsigned height); + +void +util_format_rgtc2_snorm_pack_rgba_8unorm(uint8_t *dst_row, unsigned dst_stride, const uint8_t *src_row, unsigned src_stride, unsigned width, unsigned height); + +void +util_format_rgtc2_snorm_unpack_rgba_float(float *dst_row, unsigned dst_stride, const uint8_t *src_row, unsigned src_stride, unsigned width, unsigned height); + +void +util_format_rgtc2_snorm_pack_rgba_float(uint8_t *dst_row, unsigned dst_stride, const float *src_row, unsigned src_stride, unsigned width, unsigned height); + +void +util_format_rgtc2_snorm_fetch_rgba_float(float *dst, const uint8_t *src, unsigned i, unsigned j); + + +#endif diff --git a/src/gallium/auxiliary/util/u_format_table.py b/src/gallium/auxiliary/util/u_format_table.py index 8cc22a56371..7468bc38b32 100755 --- a/src/gallium/auxiliary/util/u_format_table.py +++ b/src/gallium/auxiliary/util/u_format_table.py @@ -87,6 +87,7 @@ def write_format_table(formats): print print '#include "u_format.h"' print '#include "u_format_s3tc.h"' + print '#include "u_format_rgtc.h"' print u_format_pack.generate(formats) @@ -132,7 +133,7 @@ def write_format_table(formats): if format.colorspace != ZS: print " &util_format_%s_unpack_rgba_8unorm," % format.short_name() print " &util_format_%s_pack_rgba_8unorm," % format.short_name() - if format.layout == 's3tc': + if format.layout == 's3tc' or format.layout == 'rgtc': print " &util_format_%s_fetch_rgba_8unorm," % format.short_name() else: print " NULL, /* fetch_rgba_8unorm */" diff --git a/src/gallium/auxiliary/util/u_math.h b/src/gallium/auxiliary/util/u_math.h index 30555f92a6d..e3d4c06b6f9 100644 --- a/src/gallium/auxiliary/util/u_math.h +++ b/src/gallium/auxiliary/util/u_math.h @@ -475,6 +475,17 @@ float_to_ubyte(float f) } } +static INLINE float +byte_to_float_tex(int8_t b) +{ + return (b == -128) ? -1.0F : b * 1.0F / 127.0F; +} + +static INLINE int8_t +float_to_byte_tex(float f) +{ + return (int8_t) (127.0F * f); +} /** * Calc log base 2 diff --git a/src/gallium/auxiliary/util/u_vbuf_mgr.c b/src/gallium/auxiliary/util/u_vbuf_mgr.c index 3cf8ee0831d..7d157c99ccc 100644 --- a/src/gallium/auxiliary/util/u_vbuf_mgr.c +++ b/src/gallium/auxiliary/util/u_vbuf_mgr.c @@ -515,6 +515,7 @@ void u_vbuf_mgr_set_vertex_buffers(struct u_vbuf_mgr *mgrb, static void u_vbuf_upload_buffers(struct u_vbuf_mgr_priv *mgr, int min_index, int max_index, + unsigned instance_count, boolean *upload_flushed) { int i, nr = mgr->ve->count; @@ -530,10 +531,12 @@ static void u_vbuf_upload_buffers(struct u_vbuf_mgr_priv *mgr, !uploaded[index]) { unsigned first, size; boolean flushed; + unsigned instance_div = mgr->ve->ve[i].instance_divisor; - if (mgr->ve->ve[i].instance_divisor) { + if (instance_div) { first = 0; - size = vb->buffer->width0; + size = vb->stride * + ((instance_count + instance_div - 1) / instance_div); } else if (vb->stride) { first = vb->stride * min_index; size = vb->stride * count; @@ -581,7 +584,8 @@ void u_vbuf_mgr_draw_begin(struct u_vbuf_mgr *mgrb, /* Upload user buffers. */ if (mgr->any_user_vbs) { - u_vbuf_upload_buffers(mgr, min_index, max_index, &upload_flushed); + u_vbuf_upload_buffers(mgr, min_index, max_index, info->instance_count, + &upload_flushed); bufs_updated = TRUE; } |