From 2abcc512a3ce81bc11ff2b45a2208d3400a2385d Mon Sep 17 00:00:00 2001 From: Eric Anholt Date: Sat, 2 Feb 2008 23:27:19 -0800 Subject: [965] Convert brw_draw_upload to managing dri_bos, not gl_buffer_objects. This helps us avoid a bunch of mess with gl_client_arrays that we filled with unused data and confused readers. --- src/mesa/drivers/dri/i965/brw_context.h | 13 +- src/mesa/drivers/dri/i965/brw_draw.c | 52 +----- src/mesa/drivers/dri/i965/brw_draw_upload.c | 264 +++++++++++----------------- src/mesa/drivers/dri/i965/brw_vtbl.c | 6 +- 4 files changed, 126 insertions(+), 209 deletions(-) (limited to 'src') diff --git a/src/mesa/drivers/dri/i965/brw_context.h b/src/mesa/drivers/dri/i965/brw_context.h index 7fdaeaee17c..6bdb8342aa3 100644 --- a/src/mesa/drivers/dri/i965/brw_context.h +++ b/src/mesa/drivers/dri/i965/brw_context.h @@ -372,8 +372,16 @@ struct brw_cached_batch_item { struct brw_vertex_element { const struct gl_client_array *glarray; + /** Size of a complete element */ GLuint element_size; + /** Number of uploaded elements for this input. */ GLuint count; + /** Byte stride between elements in the uploaded array */ + GLuint stride; + /** Offset of the first element within the buffer object */ + unsigned int offset; + /** Buffer object containing the uploaded vertex data */ + dri_bo *bo; }; @@ -433,11 +441,8 @@ struct brw_context #define BRW_UPLOAD_INIT_SIZE (128*1024) struct { - struct gl_buffer_object *vbo[BRW_NR_UPLOAD_BUFS]; - GLuint buf; + dri_bo *bo; GLuint offset; - GLuint size; - GLuint wrap; } upload; /* Summary of size and varying of active arrays, so we can check diff --git a/src/mesa/drivers/dri/i965/brw_draw.c b/src/mesa/drivers/dri/i965/brw_draw.c index 72c6221ef76..0990dcfac42 100644 --- a/src/mesa/drivers/dri/i965/brw_draw.c +++ b/src/mesa/drivers/dri/i965/brw_draw.c @@ -327,20 +327,6 @@ static GLboolean brw_try_draw_prims( GLcontext *ctx, brw->no_batch_wrap = GL_FALSE; - /* Free any completed data so it doesn't clog up texture memory - we - * won't be referencing it again. - */ - while (brw->vb.upload.wrap != brw->vb.upload.buf) { - ctx->Driver.BufferData(ctx, - GL_ARRAY_BUFFER_ARB, - BRW_UPLOAD_INIT_SIZE, - NULL, - GL_DYNAMIC_DRAW_ARB, - brw->vb.upload.vbo[brw->vb.upload.wrap]); - brw->vb.upload.wrap++; - brw->vb.upload.wrap %= BRW_NR_UPLOAD_BUFS; - } - UNLOCK_HARDWARE(intel); if (!retval) @@ -418,44 +404,16 @@ void brw_draw_init( struct brw_context *brw ) { GLcontext *ctx = &brw->intel.ctx; struct vbo_context *vbo = vbo_context(ctx); - GLuint i; - + /* Register our drawing function: */ vbo->draw_prims = brw_draw_prims; - - brw->vb.upload.size = BRW_UPLOAD_INIT_SIZE; - - for (i = 0; i < BRW_NR_UPLOAD_BUFS; i++) { - brw->vb.upload.vbo[i] = ctx->Driver.NewBufferObject(ctx, 1, GL_ARRAY_BUFFER_ARB); - - ctx->Driver.BufferData(ctx, - GL_ARRAY_BUFFER_ARB, - BRW_UPLOAD_INIT_SIZE, - NULL, - GL_DYNAMIC_DRAW_ARB, - brw->vb.upload.vbo[i]); - - /* Set the internal VBOs to no-backing-store. We only use them as a - * temporary within a brw_try_draw_prims while the lock is held. - */ - if (!brw->intel.ttm) { - struct intel_buffer_object *intel_bo = - intel_buffer_object(brw->vb.upload.vbo[i]); - - dri_bo_fake_disable_backing_store(intel_bufferobj_buffer(&brw->intel, - intel_bo, - INTEL_READ), - NULL, NULL); - } - } } void brw_draw_destroy( struct brw_context *brw ) { - GLcontext *ctx = &brw->intel.ctx; - GLuint i; - - for (i = 0; i < BRW_NR_UPLOAD_BUFS; i++) - ctx->Driver.DeleteBuffer(ctx, brw->vb.upload.vbo[i]); + if (brw->vb.upload.bo != NULL) { + dri_bo_unreference(brw->vb.upload.bo); + brw->vb.upload.bo = NULL; + } } diff --git a/src/mesa/drivers/dri/i965/brw_draw_upload.c b/src/mesa/drivers/dri/i965/brw_draw_upload.c index 19fb6be6e27..839735daec8 100644 --- a/src/mesa/drivers/dri/i965/brw_draw_upload.c +++ b/src/mesa/drivers/dri/i965/brw_draw_upload.c @@ -44,13 +44,6 @@ #include "intel_buffer_objects.h" #include "intel_tex.h" -static dri_bo *array_buffer( struct intel_context *intel, - const struct gl_client_array *array ) -{ - return intel_bufferobj_buffer(intel, intel_buffer_object(array->BufferObj), - INTEL_WRITE_PART); -} - static GLuint double_types[5] = { 0, BRW_SURFACEFORMAT_R64_FLOAT, @@ -246,34 +239,40 @@ static void copy_strided_array( GLubyte *dest, static void wrap_buffers( struct brw_context *brw, GLuint size ) { - GLcontext *ctx = &brw->intel.ctx; - if (size < BRW_UPLOAD_INIT_SIZE) size = BRW_UPLOAD_INIT_SIZE; - brw->vb.upload.buf++; - brw->vb.upload.buf %= BRW_NR_UPLOAD_BUFS; brw->vb.upload.offset = 0; - ctx->Driver.BufferData(ctx, - GL_ARRAY_BUFFER_ARB, - size, - NULL, - GL_DYNAMIC_DRAW_ARB, - brw->vb.upload.vbo[brw->vb.upload.buf]); + if (brw->vb.upload.bo != NULL) + dri_bo_unreference(brw->vb.upload.bo); + brw->vb.upload.bo = dri_bo_alloc(brw->intel.bufmgr, "temporary VBO", + size, 1, + DRM_BO_FLAG_MEM_LOCAL | + DRM_BO_FLAG_CACHED | + DRM_BO_FLAG_CACHED_MAPPED); + + /* Set the internal VBO\ to no-backing-store. We only use them as a + * temporary within a brw_try_draw_prims while the lock is held. + */ + if (!brw->intel.ttm) + dri_bo_fake_disable_backing_store(brw->vb.upload.bo, NULL, NULL); } static void get_space( struct brw_context *brw, GLuint size, - struct gl_buffer_object **vbo_return, + dri_bo **bo_return, GLuint *offset_return ) { size = ALIGN(size, 64); - - if (brw->vb.upload.offset + size > BRW_UPLOAD_INIT_SIZE) + + if (brw->vb.upload.bo == NULL || + brw->vb.upload.offset + size > brw->vb.upload.bo->size) { wrap_buffers(brw, size); + } - *vbo_return = brw->vb.upload.vbo[brw->vb.upload.buf]; + dri_bo_reference(brw->vb.upload.bo); + *bo_return = brw->vb.upload.bo; *offset_return = brw->vb.upload.offset; brw->vb.upload.offset += size; @@ -281,88 +280,29 @@ static void get_space( struct brw_context *brw, static void copy_array_to_vbo_array( struct brw_context *brw, - struct gl_client_array *vbo_array, - const struct gl_client_array *array, - GLuint element_size, - GLuint count) + struct brw_vertex_element *element, + GLuint dst_stride) { - GLcontext *ctx = &brw->intel.ctx; - GLuint size = count * element_size; - struct gl_buffer_object *vbo; - GLuint offset; - GLuint new_stride; - - get_space(brw, size, &vbo, &offset); - - if (array->StrideB == 0) { - assert(count == 1); - new_stride = 0; - } - else - new_stride = element_size; - - vbo_array->Size = array->Size; - vbo_array->Type = array->Type; - vbo_array->Stride = new_stride; - vbo_array->StrideB = new_stride; - vbo_array->Ptr = (const void *)offset; - vbo_array->Enabled = 1; - vbo_array->Normalized = array->Normalized; - vbo_array->_MaxElement = array->_MaxElement; /* ? */ - vbo_array->BufferObj = vbo; - - { - GLubyte *map = ctx->Driver.MapBuffer(ctx, - GL_ARRAY_BUFFER_ARB, - GL_DYNAMIC_DRAW_ARB, - vbo); - - map += offset; + GLuint size = element->count * dst_stride; - copy_strided_array( map, - array->Ptr, - element_size, - array->StrideB, - count); + get_space(brw, size, &element->bo, &element->offset); - ctx->Driver.UnmapBuffer(ctx, GL_ARRAY_BUFFER_ARB, vbo_array->BufferObj); + if (element->glarray->StrideB == 0) { + assert(element->count == 1); + element->stride = 0; + } else { + element->stride = dst_stride; } -} -/** - * Just a wrapper to highlight which cause of copy_array_to_vbo_array - * is happening in the profile. - */ -static void -interleaved_copy_array_to_vbo_array(struct brw_context *brw, - struct gl_client_array *vbo_array, - const struct gl_client_array *array, - GLuint element_size, - GLuint count) -{ - copy_array_to_vbo_array(brw, vbo_array, array, element_size, count); + dri_bo_map(element->bo, GL_TRUE); + copy_strided_array((unsigned char *)element->bo->virtual + element->offset, + element->glarray->Ptr, + dst_stride, + element->glarray->StrideB, + element->count); + dri_bo_unmap(element->bo); } -static void -interleaved_vbo_array( struct brw_context *brw, - struct gl_client_array *vbo_array, - const struct gl_client_array *uploaded_array, - const struct gl_client_array *array, - const char *ptr) -{ - vbo_array->Size = array->Size; - vbo_array->Type = array->Type; - vbo_array->Stride = array->Stride; - vbo_array->StrideB = array->StrideB; - vbo_array->Ptr = (const void *)((const char *)uploaded_array->Ptr + - ((const char *)array->Ptr - ptr)); - vbo_array->Enabled = 1; - vbo_array->Normalized = array->Normalized; - vbo_array->_MaxElement = array->_MaxElement; - vbo_array->BufferObj = uploaded_array->BufferObj; -} - - GLboolean brw_upload_vertices( struct brw_context *brw, GLuint min_index, GLuint max_index ) @@ -371,9 +311,8 @@ GLboolean brw_upload_vertices( struct brw_context *brw, struct intel_context *intel = intel_context(ctx); GLuint tmp = brw->vs.prog_data->inputs_read; GLuint i; - const void *ptr = NULL; + const unsigned char *ptr = NULL; GLuint interleave = 0; - struct gl_client_array vbo_array_temp[VERT_ATTRIB_MAX]; struct brw_vertex_element *enabled[VERT_ATTRIB_MAX]; GLuint nr_enabled = 0; @@ -385,18 +324,45 @@ GLboolean brw_upload_vertices( struct brw_context *brw, */ if (0) _mesa_printf("%s %d..%d\n", __FUNCTION__, min_index, max_index); - + + /* Accumulate the list of enabled arrays. */ while (tmp) { GLuint i = _mesa_ffsll(tmp)-1; struct brw_vertex_element *input = &brw->vb.inputs[i]; tmp &= ~(1< 17 vertex attributes enabled, so it probably + * isn't an issue at this point. + */ + if (nr_enabled >= BRW_VEP_MAX) + return GL_FALSE; + + for (i = 0; i < nr_enabled; i++) { + struct brw_vertex_element *input = enabled[i]; input->element_size = get_size(input->glarray->Type) * input->glarray->Size; input->count = input->glarray->StrideB ? max_index + 1 - min_index : 1; - if (!input->glarray->BufferObj->Name) { + if (input->glarray->BufferObj->Name != 0) { + struct intel_buffer_object *intel_buffer = + intel_buffer_object(input->glarray->BufferObj); + + /* Named buffer object: Just reference its contents directly. */ + input->bo = intel_bufferobj_buffer(intel, intel_buffer, + INTEL_READ); + dri_bo_reference(input->bo); + input->offset = (unsigned long)input->glarray->Ptr; + input->stride = input->glarray->StrideB; + } else { + /* Queue the buffer object up to be uploaded in the next pass, + * when we've decided if we're doing interleaved or not. + */ if (i == 0) { /* Position array not properly enabled: */ @@ -407,8 +373,9 @@ GLboolean brw_upload_vertices( struct brw_context *brw, ptr = input->glarray->Ptr; } else if (interleave != input->glarray->StrideB || - (const char *)input->glarray->Ptr - (const char *)ptr < 0 || - (const char *)input->glarray->Ptr - (const char *)ptr > interleave) { + (const unsigned char *)input->glarray->Ptr - ptr < 0 || + (const unsigned char *)input->glarray->Ptr - ptr > interleave) + { interleave = 0; } @@ -425,43 +392,29 @@ GLboolean brw_upload_vertices( struct brw_context *brw, } } - /* Upload interleaved arrays if all uploads are interleaved - */ + /* Handle any arrays to be uploaded. */ if (nr_uploads > 1 && interleave && interleave <= 256) { - interleaved_copy_array_to_vbo_array(brw, &vbo_array_temp[0], - upload[0]->glarray, - interleave, - upload[0]->count); - upload[0]->glarray = &vbo_array_temp[0]; + /* All uploads are interleaved, so upload the arrays together as + * interleaved. First, upload the contents and set up upload[0]. + */ + copy_array_to_vbo_array(brw, upload[0], interleave); for (i = 1; i < nr_uploads; i++) { - interleaved_vbo_array(brw, - &vbo_array_temp[i], - upload[0]->glarray, - upload[i]->glarray, - ptr); - upload[i]->glarray = &vbo_array_temp[i]; + /* Then, just point upload[i] at upload[0]'s buffer. */ + upload[i]->stride = interleave; + upload[i]->offset = upload[0]->offset + + ((const unsigned char *)upload[i]->glarray->Ptr - ptr); + upload[i]->bo = upload[0]->bo; + dri_bo_reference(upload[i]->bo); } } else { + /* Upload non-interleaved arrays */ for (i = 0; i < nr_uploads; i++) { - copy_array_to_vbo_array(brw, &vbo_array_temp[i], - upload[i]->glarray, - upload[i]->element_size, - upload[i]->count); - upload[i]->glarray = &vbo_array_temp[i]; + copy_array_to_vbo_array(brw, upload[i], upload[i]->element_size); } } - /* XXX: In the rare cases where this happens we fallback all - * the way to software rasterization, although a tnl fallback - * would be sufficient. I don't know of *any* real world - * cases with > 17 vertex attributes enabled, so it probably - * isn't an issue at this point. - */ - if (nr_enabled >= BRW_VEP_MAX) - return GL_FALSE; - /* Now emit VB and VEP state packets. * * This still defines a hardware VB for each input, even if they @@ -477,12 +430,18 @@ GLboolean brw_upload_vertices( struct brw_context *brw, OUT_BATCH((i << BRW_VB0_INDEX_SHIFT) | BRW_VB0_ACCESS_VERTEXDATA | - (input->glarray->StrideB << BRW_VB0_PITCH_SHIFT)); - OUT_RELOC(array_buffer(intel, input->glarray), + (input->stride << BRW_VB0_PITCH_SHIFT)); + OUT_RELOC(input->bo, DRM_BO_FLAG_MEM_TT | DRM_BO_FLAG_READ, - (GLuint)input->glarray->Ptr); + input->offset); OUT_BATCH(max_index); OUT_BATCH(0); /* Instance data step rate */ + + /* Unreference the buffer so it can get freed, now that we won't + * touch it any more. + */ + dri_bo_unreference(input->bo); + input->bo = NULL; } ADVANCE_BATCH(); @@ -527,6 +486,7 @@ void brw_upload_indices( struct brw_context *brw, GLcontext *ctx = &brw->intel.ctx; struct intel_context *intel = &brw->intel; GLuint ib_size = get_size(index_buffer->type) * index_buffer->count; + dri_bo *bo; struct gl_buffer_object *bufferobj = index_buffer->obj; GLuint offset = (GLuint)index_buffer->ptr; @@ -536,40 +496,31 @@ void brw_upload_indices( struct brw_context *brw, /* Get new bufferobj, offset: */ - get_space(brw, ib_size, &bufferobj, &offset); + get_space(brw, ib_size, &bo, &offset); /* Straight upload */ - ctx->Driver.BufferSubData( ctx, - GL_ELEMENT_ARRAY_BUFFER_ARB, - offset, - ib_size, - index_buffer->ptr, - bufferobj); + dri_bo_subdata(bo, offset, ib_size, index_buffer->ptr); } else { /* If the index buffer isn't aligned to its element size, we have to * rebase it into a temporary. */ if ((get_size(index_buffer->type) - 1) & offset) { - struct gl_buffer_object *vbo; - GLuint voffset; GLubyte *map = ctx->Driver.MapBuffer(ctx, GL_ELEMENT_ARRAY_BUFFER_ARB, GL_DYNAMIC_DRAW_ARB, bufferobj); map += offset; - get_space(brw, ib_size, &vbo, &voffset); - - ctx->Driver.BufferSubData(ctx, - GL_ELEMENT_ARRAY_BUFFER_ARB, - voffset, - ib_size, - map, - vbo); - ctx->Driver.UnmapBuffer(ctx, GL_ELEMENT_ARRAY_BUFFER_ARB, bufferobj); - bufferobj = vbo; - offset = voffset; + get_space(brw, ib_size, &bo, &offset); + + dri_bo_subdata(bo, offset, ib_size, map); + + ctx->Driver.UnmapBuffer(ctx, GL_ELEMENT_ARRAY_BUFFER_ARB, bufferobj); + } else { + bo = intel_bufferobj_buffer(intel, intel_buffer_object(bufferobj), + INTEL_READ); + dri_bo_reference(bo); } } @@ -577,9 +528,6 @@ void brw_upload_indices( struct brw_context *brw, */ { struct brw_indexbuffer ib; - dri_bo *buffer = intel_bufferobj_buffer(intel, - intel_buffer_object(bufferobj), - INTEL_READ); memset(&ib, 0, sizeof(ib)); @@ -591,10 +539,12 @@ void brw_upload_indices( struct brw_context *brw, BEGIN_BATCH(4, IGNORE_CLIPRECTS); OUT_BATCH( ib.header.dword ); - OUT_RELOC( buffer, DRM_BO_FLAG_MEM_TT | DRM_BO_FLAG_READ, offset); - OUT_RELOC( buffer, DRM_BO_FLAG_MEM_TT | DRM_BO_FLAG_READ, + OUT_RELOC( bo, DRM_BO_FLAG_MEM_TT | DRM_BO_FLAG_READ, offset); + OUT_RELOC( bo, DRM_BO_FLAG_MEM_TT | DRM_BO_FLAG_READ, offset + ib_size); OUT_BATCH( 0 ); ADVANCE_BATCH(); + + dri_bo_unreference(bo); } } diff --git a/src/mesa/drivers/dri/i965/brw_vtbl.c b/src/mesa/drivers/dri/i965/brw_vtbl.c index 0b70888071d..cdbbe7b6994 100644 --- a/src/mesa/drivers/dri/i965/brw_vtbl.c +++ b/src/mesa/drivers/dri/i965/brw_vtbl.c @@ -110,7 +110,11 @@ static void brw_new_batch( struct intel_context *intel ) /* Move to the end of the current upload buffer so that we'll force choosing * a new buffer next time. */ - brw->vb.upload.offset = brw->vb.upload.vbo[brw->vb.upload.buf]->Size; + if (brw->vb.upload.bo != NULL) { + dri_bo_unreference(brw->vb.upload.bo); + brw->vb.upload.bo = NULL; + brw->vb.upload.offset = 0; + } } static void brw_note_fence( struct intel_context *intel, -- cgit v1.2.3