diff options
Diffstat (limited to 'src/gallium/drivers/i965/brw_draw_upload.c')
| -rw-r--r-- | src/gallium/drivers/i965/brw_draw_upload.c | 542 |
1 files changed, 542 insertions, 0 deletions
diff --git a/src/gallium/drivers/i965/brw_draw_upload.c b/src/gallium/drivers/i965/brw_draw_upload.c new file mode 100644 index 0000000000..a27da5f1c1 --- /dev/null +++ b/src/gallium/drivers/i965/brw_draw_upload.c @@ -0,0 +1,542 @@ +/************************************************************************** + * + * Copyright 2003 Tungsten Graphics, Inc., Cedar Park, Texas. + * 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 above copyright notice and this permission notice (including the + * next paragraph) 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 NON-INFRINGEMENT. + * IN NO EVENT SHALL TUNGSTEN GRAPHICS 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. + * + **************************************************************************/ + +#include "pipe/p_context.h" + +#include "util/u_upload_mgr.h" +#include "util/u_math.h" + +#include "brw_draw.h" +#include "brw_defines.h" +#include "brw_context.h" +#include "brw_state.h" +#include "brw_screen.h" +#include "brw_batchbuffer.h" +#include "brw_debug.h" + + + + +static unsigned brw_translate_surface_format( unsigned id ) +{ + switch (id) { + case PIPE_FORMAT_R64_FLOAT: + return BRW_SURFACEFORMAT_R64_FLOAT; + case PIPE_FORMAT_R64G64_FLOAT: + return BRW_SURFACEFORMAT_R64G64_FLOAT; + case PIPE_FORMAT_R64G64B64_FLOAT: + return BRW_SURFACEFORMAT_R64G64B64_FLOAT; + case PIPE_FORMAT_R64G64B64A64_FLOAT: + return BRW_SURFACEFORMAT_R64G64B64A64_FLOAT; + + case PIPE_FORMAT_R32_FLOAT: + return BRW_SURFACEFORMAT_R32_FLOAT; + case PIPE_FORMAT_R32G32_FLOAT: + return BRW_SURFACEFORMAT_R32G32_FLOAT; + case PIPE_FORMAT_R32G32B32_FLOAT: + return BRW_SURFACEFORMAT_R32G32B32_FLOAT; + case PIPE_FORMAT_R32G32B32A32_FLOAT: + return BRW_SURFACEFORMAT_R32G32B32A32_FLOAT; + + case PIPE_FORMAT_R32_UNORM: + return BRW_SURFACEFORMAT_R32_UNORM; + case PIPE_FORMAT_R32G32_UNORM: + return BRW_SURFACEFORMAT_R32G32_UNORM; + case PIPE_FORMAT_R32G32B32_UNORM: + return BRW_SURFACEFORMAT_R32G32B32_UNORM; + case PIPE_FORMAT_R32G32B32A32_UNORM: + return BRW_SURFACEFORMAT_R32G32B32A32_UNORM; + + case PIPE_FORMAT_R32_USCALED: + return BRW_SURFACEFORMAT_R32_USCALED; + case PIPE_FORMAT_R32G32_USCALED: + return BRW_SURFACEFORMAT_R32G32_USCALED; + case PIPE_FORMAT_R32G32B32_USCALED: + return BRW_SURFACEFORMAT_R32G32B32_USCALED; + case PIPE_FORMAT_R32G32B32A32_USCALED: + return BRW_SURFACEFORMAT_R32G32B32A32_USCALED; + + case PIPE_FORMAT_R32_SNORM: + return BRW_SURFACEFORMAT_R32_SNORM; + case PIPE_FORMAT_R32G32_SNORM: + return BRW_SURFACEFORMAT_R32G32_SNORM; + case PIPE_FORMAT_R32G32B32_SNORM: + return BRW_SURFACEFORMAT_R32G32B32_SNORM; + case PIPE_FORMAT_R32G32B32A32_SNORM: + return BRW_SURFACEFORMAT_R32G32B32A32_SNORM; + + case PIPE_FORMAT_R32_SSCALED: + return BRW_SURFACEFORMAT_R32_SSCALED; + case PIPE_FORMAT_R32G32_SSCALED: + return BRW_SURFACEFORMAT_R32G32_SSCALED; + case PIPE_FORMAT_R32G32B32_SSCALED: + return BRW_SURFACEFORMAT_R32G32B32_SSCALED; + case PIPE_FORMAT_R32G32B32A32_SSCALED: + return BRW_SURFACEFORMAT_R32G32B32A32_SSCALED; + + case PIPE_FORMAT_R16_UNORM: + return BRW_SURFACEFORMAT_R16_UNORM; + case PIPE_FORMAT_R16G16_UNORM: + return BRW_SURFACEFORMAT_R16G16_UNORM; + case PIPE_FORMAT_R16G16B16_UNORM: + return BRW_SURFACEFORMAT_R16G16B16_UNORM; + case PIPE_FORMAT_R16G16B16A16_UNORM: + return BRW_SURFACEFORMAT_R16G16B16A16_UNORM; + + case PIPE_FORMAT_R16_USCALED: + return BRW_SURFACEFORMAT_R16_USCALED; + case PIPE_FORMAT_R16G16_USCALED: + return BRW_SURFACEFORMAT_R16G16_USCALED; + case PIPE_FORMAT_R16G16B16_USCALED: + return BRW_SURFACEFORMAT_R16G16B16_USCALED; + case PIPE_FORMAT_R16G16B16A16_USCALED: + return BRW_SURFACEFORMAT_R16G16B16A16_USCALED; + + case PIPE_FORMAT_R16_SNORM: + return BRW_SURFACEFORMAT_R16_SNORM; + case PIPE_FORMAT_R16G16_SNORM: + return BRW_SURFACEFORMAT_R16G16_SNORM; + case PIPE_FORMAT_R16G16B16_SNORM: + return BRW_SURFACEFORMAT_R16G16B16_SNORM; + case PIPE_FORMAT_R16G16B16A16_SNORM: + return BRW_SURFACEFORMAT_R16G16B16A16_SNORM; + + case PIPE_FORMAT_R16_SSCALED: + return BRW_SURFACEFORMAT_R16_SSCALED; + case PIPE_FORMAT_R16G16_SSCALED: + return BRW_SURFACEFORMAT_R16G16_SSCALED; + case PIPE_FORMAT_R16G16B16_SSCALED: + return BRW_SURFACEFORMAT_R16G16B16_SSCALED; + case PIPE_FORMAT_R16G16B16A16_SSCALED: + return BRW_SURFACEFORMAT_R16G16B16A16_SSCALED; + + case PIPE_FORMAT_R8_UNORM: + return BRW_SURFACEFORMAT_R8_UNORM; + case PIPE_FORMAT_R8G8_UNORM: + return BRW_SURFACEFORMAT_R8G8_UNORM; + case PIPE_FORMAT_R8G8B8_UNORM: + return BRW_SURFACEFORMAT_R8G8B8_UNORM; + case PIPE_FORMAT_R8G8B8A8_UNORM: + return BRW_SURFACEFORMAT_R8G8B8A8_UNORM; + + case PIPE_FORMAT_R8_USCALED: + return BRW_SURFACEFORMAT_R8_USCALED; + case PIPE_FORMAT_R8G8_USCALED: + return BRW_SURFACEFORMAT_R8G8_USCALED; + case PIPE_FORMAT_R8G8B8_USCALED: + return BRW_SURFACEFORMAT_R8G8B8_USCALED; + case PIPE_FORMAT_R8G8B8A8_USCALED: + return BRW_SURFACEFORMAT_R8G8B8A8_USCALED; + + case PIPE_FORMAT_R8_SNORM: + return BRW_SURFACEFORMAT_R8_SNORM; + case PIPE_FORMAT_R8G8_SNORM: + return BRW_SURFACEFORMAT_R8G8_SNORM; + case PIPE_FORMAT_R8G8B8_SNORM: + return BRW_SURFACEFORMAT_R8G8B8_SNORM; + case PIPE_FORMAT_R8G8B8A8_SNORM: + return BRW_SURFACEFORMAT_R8G8B8A8_SNORM; + + case PIPE_FORMAT_R8_SSCALED: + return BRW_SURFACEFORMAT_R8_SSCALED; + case PIPE_FORMAT_R8G8_SSCALED: + return BRW_SURFACEFORMAT_R8G8_SSCALED; + case PIPE_FORMAT_R8G8B8_SSCALED: + return BRW_SURFACEFORMAT_R8G8B8_SSCALED; + case PIPE_FORMAT_R8G8B8A8_SSCALED: + return BRW_SURFACEFORMAT_R8G8B8A8_SSCALED; + + default: + assert(0); + return 0; + } +} + +static unsigned get_index_type(int type) +{ + switch (type) { + case 1: return BRW_INDEX_BYTE; + case 2: return BRW_INDEX_WORD; + case 4: return BRW_INDEX_DWORD; + default: assert(0); return 0; + } +} + + +static int brw_prepare_vertices(struct brw_context *brw) +{ + unsigned int min_index = brw->curr.min_index; + unsigned int max_index = brw->curr.max_index; + GLuint i; + int ret; + + if (BRW_DEBUG & DEBUG_VERTS) + debug_printf("%s %d..%d\n", __FUNCTION__, min_index, max_index); + + + for (i = 0; i < brw->curr.num_vertex_buffers; i++) { + struct pipe_vertex_buffer *vb = &brw->curr.vertex_buffer[i]; + struct brw_winsys_buffer *bo; + struct pipe_buffer *upload_buf = NULL; + unsigned offset; + + if (BRW_DEBUG & DEBUG_VERTS) + debug_printf("%s vb[%d] user:%d offset:0x%x sz:0x%x stride:0x%x\n", + __FUNCTION__, i, + brw_buffer_is_user_buffer(vb->buffer), + vb->buffer_offset, + vb->buffer->size, + vb->stride); + + if (brw_buffer_is_user_buffer(vb->buffer)) { + + /* XXX: simplify this. Stop the state trackers from generating + * zero-stride buffers & have them use additional constants (or + * add support for >1 constant buffer) instead. + */ + unsigned size = (vb->stride == 0 ? + vb->buffer->size - vb->buffer_offset : + MAX2(vb->buffer->size - vb->buffer_offset, + vb->stride * (max_index + 1 - min_index))); + + ret = u_upload_buffer( brw->vb.upload_vertex, + vb->buffer_offset + min_index * vb->stride, + size, + vb->buffer, + &offset, + &upload_buf ); + if (ret) + return ret; + + bo = brw_buffer(upload_buf)->bo; + + assert(offset + size <= bo->size); + } + else + { + offset = vb->buffer_offset; + bo = brw_buffer(vb->buffer)->bo; + } + + assert(offset < bo->size); + + /* Set up post-upload info about this vertex buffer: + */ + brw->vb.vb[i].offset = offset; + brw->vb.vb[i].stride = vb->stride; + brw->vb.vb[i].vertex_count = (vb->stride == 0 ? + 1 : + (bo->size - offset) / vb->stride); + + bo_reference( &brw->vb.vb[i].bo, bo ); + + /* Don't need to retain this reference. We have a reference on + * the underlying winsys buffer: + */ + pipe_buffer_reference( &upload_buf, NULL ); + } + + brw->vb.nr_vb = i; + brw_prepare_query_begin(brw); + + for (i = 0; i < brw->vb.nr_vb; i++) { + brw_add_validated_bo(brw, brw->vb.vb[i].bo); + } + + return 0; +} + +static int brw_emit_vertex_buffers( struct brw_context *brw ) +{ + int i; + + /* If the VS doesn't read any inputs (calculating vertex position from + * a state variable for some reason, for example), just bail. + * + * The stale VB state stays in place, but they don't do anything unless + * a VE loads from them. + */ + if (brw->vb.nr_vb == 0) { + if (BRW_DEBUG & DEBUG_VERTS) + debug_printf("%s: no active vertex buffers\n", __FUNCTION__); + + return 0; + } + + /* Emit VB state packets. + */ + BEGIN_BATCH(1 + brw->vb.nr_vb * 4, IGNORE_CLIPRECTS); + OUT_BATCH((CMD_VERTEX_BUFFER << 16) | + ((1 + brw->vb.nr_vb * 4) - 2)); + + for (i = 0; i < brw->vb.nr_vb; i++) { + OUT_BATCH((i << BRW_VB0_INDEX_SHIFT) | + BRW_VB0_ACCESS_VERTEXDATA | + (brw->vb.vb[i].stride << BRW_VB0_PITCH_SHIFT)); + OUT_RELOC(brw->vb.vb[i].bo, + BRW_USAGE_VERTEX, + brw->vb.vb[i].offset); + if (BRW_IS_IGDNG(brw)) { + OUT_RELOC(brw->vb.vb[i].bo, + BRW_USAGE_VERTEX, + brw->vb.vb[i].bo->size - 1); + } else + OUT_BATCH(brw->vb.vb[i].stride ? brw->vb.vb[i].vertex_count : 0); + OUT_BATCH(0); /* Instance data step rate */ + } + ADVANCE_BATCH(); + return 0; +} + + + + +static int brw_emit_vertex_elements(struct brw_context *brw) +{ + GLuint nr = brw->curr.num_vertex_elements; + GLuint i; + + brw_emit_query_begin(brw); + + /* If the VS doesn't read any inputs (calculating vertex position from + * a state variable for some reason, for example), emit a single pad + * VERTEX_ELEMENT struct and bail. + * + * The stale VB state stays in place, but they don't do anything unless + * a VE loads from them. + */ + if (nr == 0) { + BEGIN_BATCH(3, IGNORE_CLIPRECTS); + OUT_BATCH((CMD_VERTEX_ELEMENT << 16) | 1); + OUT_BATCH((0 << BRW_VE0_INDEX_SHIFT) | + BRW_VE0_VALID | + (BRW_SURFACEFORMAT_R32G32B32A32_FLOAT << BRW_VE0_FORMAT_SHIFT) | + (0 << BRW_VE0_SRC_OFFSET_SHIFT)); + OUT_BATCH((BRW_VE1_COMPONENT_STORE_0 << BRW_VE1_COMPONENT_0_SHIFT) | + (BRW_VE1_COMPONENT_STORE_0 << BRW_VE1_COMPONENT_1_SHIFT) | + (BRW_VE1_COMPONENT_STORE_0 << BRW_VE1_COMPONENT_2_SHIFT) | + (BRW_VE1_COMPONENT_STORE_1_FLT << BRW_VE1_COMPONENT_3_SHIFT)); + ADVANCE_BATCH(); + return 0; + } + + /* Now emit vertex element (VEP) state packets. + * + */ + BEGIN_BATCH(1 + nr * 2, IGNORE_CLIPRECTS); + OUT_BATCH((CMD_VERTEX_ELEMENT << 16) | ((1 + nr * 2) - 2)); + for (i = 0; i < nr; i++) { + const struct pipe_vertex_element *input = &brw->curr.vertex_element[i]; + uint32_t format = brw_translate_surface_format( input->src_format ); + uint32_t comp0 = BRW_VE1_COMPONENT_STORE_SRC; + uint32_t comp1 = BRW_VE1_COMPONENT_STORE_SRC; + uint32_t comp2 = BRW_VE1_COMPONENT_STORE_SRC; + uint32_t comp3 = BRW_VE1_COMPONENT_STORE_SRC; + + switch (input->nr_components) { + case 0: comp0 = BRW_VE1_COMPONENT_STORE_0; + case 1: comp1 = BRW_VE1_COMPONENT_STORE_0; + case 2: comp2 = BRW_VE1_COMPONENT_STORE_0; + case 3: comp3 = BRW_VE1_COMPONENT_STORE_1_FLT; + break; + } + + OUT_BATCH((input->vertex_buffer_index << BRW_VE0_INDEX_SHIFT) | + BRW_VE0_VALID | + (format << BRW_VE0_FORMAT_SHIFT) | + (input->src_offset << BRW_VE0_SRC_OFFSET_SHIFT)); + + if (BRW_IS_IGDNG(brw)) + OUT_BATCH((comp0 << BRW_VE1_COMPONENT_0_SHIFT) | + (comp1 << BRW_VE1_COMPONENT_1_SHIFT) | + (comp2 << BRW_VE1_COMPONENT_2_SHIFT) | + (comp3 << BRW_VE1_COMPONENT_3_SHIFT)); + else + OUT_BATCH((comp0 << BRW_VE1_COMPONENT_0_SHIFT) | + (comp1 << BRW_VE1_COMPONENT_1_SHIFT) | + (comp2 << BRW_VE1_COMPONENT_2_SHIFT) | + (comp3 << BRW_VE1_COMPONENT_3_SHIFT) | + ((i * 4) << BRW_VE1_DST_OFFSET_SHIFT)); + } + ADVANCE_BATCH(); + return 0; +} + + +static int brw_emit_vertices( struct brw_context *brw ) +{ + int ret; + + ret = brw_emit_vertex_buffers( brw ); + if (ret) + return ret; + + ret = brw_emit_vertex_elements( brw ); + if (ret) + return ret; + + return 0; +} + + +const struct brw_tracked_state brw_vertices = { + .dirty = { + .mesa = (PIPE_NEW_INDEX_RANGE | + PIPE_NEW_VERTEX_BUFFER), + .brw = BRW_NEW_BATCH, + .cache = 0, + }, + .prepare = brw_prepare_vertices, + .emit = brw_emit_vertices, +}; + + +static int brw_prepare_indices(struct brw_context *brw) +{ + struct pipe_buffer *index_buffer = brw->curr.index_buffer; + struct pipe_buffer *upload_buf = NULL; + struct brw_winsys_buffer *bo = NULL; + GLuint offset; + GLuint index_size; + GLuint ib_size; + int ret; + + if (index_buffer == NULL) + return 0; + + if (BRW_DEBUG & DEBUG_VERTS) + debug_printf("%s: index_size:%d index_buffer->size:%d\n", + __FUNCTION__, + brw->curr.index_size, + brw->curr.index_buffer->size); + + ib_size = index_buffer->size; + index_size = brw->curr.index_size; + + /* Turn userbuffer into a proper hardware buffer? + */ + if (brw_buffer_is_user_buffer(index_buffer)) { + + ret = u_upload_buffer( brw->vb.upload_index, + 0, + ib_size, + index_buffer, + &offset, + &upload_buf ); + if (ret) + return ret; + + bo = brw_buffer(upload_buf)->bo; + + /* XXX: annotate the userbuffer with the upload information so + * that successive calls don't get re-uploaded. + */ + } + else { + bo = brw_buffer(index_buffer)->bo; + ib_size = bo->size; + offset = 0; + } + + /* Use CMD_3D_PRIM's start_vertex_offset to avoid re-uploading the + * index buffer state when we're just moving the start index of our + * drawing. + * + * In gallium this will happen in the case where successive draw + * calls are made with (distinct?) userbuffers, but the upload_mgr + * places the data into a single winsys buffer. + * + * This statechange doesn't raise any state flags and is always + * just merged into the final draw packet: + */ + if (1) { + assert((offset & (index_size - 1)) == 0); + brw->ib.start_vertex_offset = offset / index_size; + } + + /* These statechanges trigger a new CMD_INDEX_BUFFER packet: + */ + if (brw->ib.bo != bo || + brw->ib.size != ib_size) + { + bo_reference(&brw->ib.bo, bo); + brw->ib.size = ib_size; + brw->state.dirty.brw |= BRW_NEW_INDEX_BUFFER; + } + + pipe_buffer_reference( &upload_buf, NULL ); + brw_add_validated_bo(brw, brw->ib.bo); + return 0; +} + +const struct brw_tracked_state brw_indices = { + .dirty = { + .mesa = PIPE_NEW_INDEX_BUFFER, + .brw = 0, + .cache = 0, + }, + .prepare = brw_prepare_indices, +}; + +static int brw_emit_index_buffer(struct brw_context *brw) +{ + /* Emit the indexbuffer packet: + */ + if (brw->ib.bo) + { + struct brw_indexbuffer ib; + + memset(&ib, 0, sizeof(ib)); + + ib.header.bits.opcode = CMD_INDEX_BUFFER; + ib.header.bits.length = sizeof(ib)/4 - 2; + ib.header.bits.index_format = get_index_type(brw->ib.size); + ib.header.bits.cut_index_enable = 0; + + BEGIN_BATCH(4, IGNORE_CLIPRECTS); + OUT_BATCH( ib.header.dword ); + OUT_RELOC(brw->ib.bo, + BRW_USAGE_VERTEX, + brw->ib.offset); + OUT_RELOC(brw->ib.bo, + BRW_USAGE_VERTEX, + brw->ib.offset + brw->ib.size - 1); + OUT_BATCH( 0 ); + ADVANCE_BATCH(); + } + + return 0; +} + +const struct brw_tracked_state brw_index_buffer = { + .dirty = { + .mesa = 0, + .brw = BRW_NEW_BATCH | BRW_NEW_INDEX_BUFFER, + .cache = 0, + }, + .emit = brw_emit_index_buffer, +}; |