X-Git-Url: https://git.libre-soc.org/?a=blobdiff_plain;ds=sidebyside;f=src%2Fmesa%2Fstate_tracker%2Fst_cb_drawpixels.c;h=311ba25c17fc00668d2172bb1235e3b92bd98705;hb=b5dd8e0cf877db2adc249db8e132beeea79ba1cf;hp=5e078a85edfa35b28aba155ae566038bb54f85ce;hpb=6c1fbe912f1623cbcf474a0b985841e44122a911;p=mesa.git diff --git a/src/mesa/state_tracker/st_cb_drawpixels.c b/src/mesa/state_tracker/st_cb_drawpixels.c index 5e078a85edf..311ba25c17f 100644 --- a/src/mesa/state_tracker/st_cb_drawpixels.c +++ b/src/mesa/state_tracker/st_cb_drawpixels.c @@ -1,6 +1,6 @@ /************************************************************************** * - * Copyright 2007 Tungsten Graphics, Inc., Cedar Park, Texas. + * Copyright 2007 VMware, Inc. * All Rights Reserved. * * Permission is hereby granted, free of charge, to any person obtaining a @@ -18,7 +18,7 @@ * 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 + * IN NO EVENT SHALL VMWARE 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. @@ -33,9 +33,10 @@ #include "main/imports.h" #include "main/image.h" #include "main/bufferobj.h" +#include "main/blit.h" #include "main/format_pack.h" +#include "main/framebuffer.h" #include "main/macros.h" -#include "main/mfeatures.h" #include "main/mtypes.h" #include "main/pack.h" #include "main/pbo.h" @@ -43,17 +44,20 @@ #include "main/texformat.h" #include "main/teximage.h" #include "main/texstore.h" +#include "main/glformats.h" #include "program/program.h" #include "program/prog_print.h" #include "program/prog_instruction.h" #include "st_atom.h" #include "st_atom_constbuf.h" +#include "st_cb_bitmap.h" #include "st_cb_drawpixels.h" #include "st_cb_readpixels.h" #include "st_cb_fbo.h" #include "st_context.h" #include "st_debug.h" +#include "st_draw.h" #include "st_format.h" #include "st_program.h" #include "st_texture.h" @@ -61,128 +65,42 @@ #include "pipe/p_context.h" #include "pipe/p_defines.h" #include "tgsi/tgsi_ureg.h" -#include "util/u_draw_quad.h" #include "util/u_format.h" #include "util/u_inlines.h" #include "util/u_math.h" #include "util/u_tile.h" -#include "util/u_upload_mgr.h" #include "cso_cache/cso_context.h" -#if FEATURE_drawpix - -/** - * Check if the given program is: - * 0: MOVE result.color, fragment.color; - * 1: END; - */ -static GLboolean -is_passthrough_program(const struct gl_fragment_program *prog) -{ - if (prog->Base.NumInstructions == 2) { - const struct prog_instruction *inst = prog->Base.Instructions; - if (inst[0].Opcode == OPCODE_MOV && - inst[1].Opcode == OPCODE_END && - inst[0].DstReg.File == PROGRAM_OUTPUT && - inst[0].DstReg.Index == FRAG_RESULT_COLOR && - inst[0].DstReg.WriteMask == WRITEMASK_XYZW && - inst[0].SrcReg[0].File == PROGRAM_INPUT && - inst[0].SrcReg[0].Index == FRAG_ATTRIB_COL0 && - inst[0].SrcReg[0].Swizzle == SWIZZLE_XYZW) { - return GL_TRUE; - } - } - return GL_FALSE; -} - - -/** - * Returns a fragment program which implements the current pixel transfer ops. - */ -static struct gl_fragment_program * -get_glsl_pixel_transfer_program(struct st_context *st, - struct st_fragment_program *orig) -{ - int pixelMaps = 0, scaleAndBias = 0; - struct gl_context *ctx = st->ctx; - struct st_fragment_program *fp = (struct st_fragment_program *) - ctx->Driver.NewProgram(ctx, GL_FRAGMENT_PROGRAM_ARB, 0); - - if (!fp) - return NULL; - - if (ctx->Pixel.RedBias != 0.0 || ctx->Pixel.RedScale != 1.0 || - ctx->Pixel.GreenBias != 0.0 || ctx->Pixel.GreenScale != 1.0 || - ctx->Pixel.BlueBias != 0.0 || ctx->Pixel.BlueScale != 1.0 || - ctx->Pixel.AlphaBias != 0.0 || ctx->Pixel.AlphaScale != 1.0) { - scaleAndBias = 1; - } - - pixelMaps = ctx->Pixel.MapColorFlag; - - if (pixelMaps) { - /* create the colormap/texture now if not already done */ - if (!st->pixel_xfer.pixelmap_texture) { - st->pixel_xfer.pixelmap_texture = st_create_color_map_texture(ctx); - st->pixel_xfer.pixelmap_sampler_view = - st_create_texture_sampler_view(st->pipe, - st->pixel_xfer.pixelmap_texture); - } - } - - get_pixel_transfer_visitor(fp, orig->glsl_to_tgsi, - scaleAndBias, pixelMaps); - - return &fp->Base; -} - - /** - * Make fragment shader for glDraw/CopyPixels. This shader is made - * by combining the pixel transfer shader with the user-defined shader. - * \param fpIn the current/incoming fragment program - * \param fpOut returns the combined fragment program + * We have a simple glDrawPixels cache to try to optimize the case where the + * same image is drawn over and over again. It basically works as follows: + * + * 1. After we construct a texture map with the image and draw it, we do + * not discard the texture. We keep it around, plus we note the + * glDrawPixels width, height, format, etc. parameters and keep a copy + * of the image in a malloc'd buffer. + * + * 2. On the next glDrawPixels we check if the parameters match the previous + * call. If those match, we check if the image matches the previous image + * via a memcmp() call. If everything matches, we re-use the previous + * texture, thereby avoiding the cost creating a new texture and copying + * the image to it. + * + * The effectiveness of this cache depends upon: + * 1. If the memcmp() finds a difference, it happens relatively quickly. + Hopefully, not just the last pixels differ! + * 2. If the memcmp() finds no difference, doing that check is faster than + * creating and loading a texture. + * + * Notes: + * 1. We don't support any pixel unpacking parameters. + * 2. We don't try to cache images in Pixel Buffer Objects. + * 3. Instead of saving the whole image, perhaps some sort of reliable + * checksum function could be used instead. */ -void -st_make_drawpix_fragment_program(struct st_context *st, - struct gl_fragment_program *fpIn, - struct gl_fragment_program **fpOut) -{ - struct gl_program *newProg; - struct st_fragment_program *stfp = (struct st_fragment_program *) fpIn; +#define USE_DRAWPIXELS_CACHE 1 - if (is_passthrough_program(fpIn)) { - newProg = (struct gl_program *) _mesa_clone_fragment_program(st->ctx, - &st->pixel_xfer.program->Base); - } - else if (stfp->glsl_to_tgsi != NULL) { - newProg = (struct gl_program *) get_glsl_pixel_transfer_program(st, stfp); - } - else { -#if 0 - /* debug */ - printf("Base program:\n"); - _mesa_print_program(&fpIn->Base); - printf("DrawPix program:\n"); - _mesa_print_program(&st->pixel_xfer.program->Base.Base); -#endif - newProg = _mesa_combine_programs(st->ctx, - &st->pixel_xfer.program->Base.Base, - &fpIn->Base); - } - -#if 0 - /* debug */ - printf("Combined DrawPixels program:\n"); - _mesa_print_program(newProg); - printf("InputsRead: 0x%x\n", newProg->InputsRead); - printf("OutputsWritten: 0x%x\n", newProg->OutputsWritten); - _mesa_print_parameter_list(newProg->Parameters); -#endif - - *fpOut = (struct gl_fragment_program *) newProg; -} /** @@ -190,91 +108,80 @@ st_make_drawpix_fragment_program(struct st_context *st, * stencil value value, then writes to FRAG_RESULT_DEPTH/FRAG_RESULT_STENCIL. * Used for glDrawPixels(GL_DEPTH_COMPONENT / GL_STENCIL_INDEX). * Pass fragment color through as-is. - * \return pointer to the gl_fragment program + * + * \return CSO of the fragment shader. */ -struct gl_fragment_program * -st_make_drawpix_z_stencil_program(struct st_context *st, - GLboolean write_depth, - GLboolean write_stencil) +static void * +get_drawpix_z_stencil_program(struct st_context *st, + GLboolean write_depth, + GLboolean write_stencil) { - struct gl_context *ctx = st->ctx; - struct gl_program *p; - struct gl_fragment_program *fp; - GLuint ic = 0; + struct ureg_program *ureg; + struct ureg_src depth_sampler, stencil_sampler; + struct ureg_src texcoord, color; + struct ureg_dst out_color, out_depth, out_stencil; const GLuint shaderIndex = write_depth * 2 + write_stencil; + void *cso; - assert(shaderIndex < Elements(st->drawpix.shaders)); + assert(shaderIndex < ARRAY_SIZE(st->drawpix.zs_shaders)); - if (st->drawpix.shaders[shaderIndex]) { + if (st->drawpix.zs_shaders[shaderIndex]) { /* already have the proper shader */ - return st->drawpix.shaders[shaderIndex]; + return st->drawpix.zs_shaders[shaderIndex]; } - /* - * Create shader now - */ - p = ctx->Driver.NewProgram(ctx, GL_FRAGMENT_PROGRAM_ARB, 0); - if (!p) + ureg = ureg_create(PIPE_SHADER_FRAGMENT); + if (ureg == NULL) return NULL; - p->NumInstructions = write_depth ? 2 : 1; - p->NumInstructions += write_stencil ? 1 : 0; - - p->Instructions = _mesa_alloc_instructions(p->NumInstructions); - if (!p->Instructions) { - ctx->Driver.DeleteProgram(ctx, p); - return NULL; - } - _mesa_init_instructions(p->Instructions, p->NumInstructions); + ureg_property(ureg, TGSI_PROPERTY_FS_COLOR0_WRITES_ALL_CBUFS, TRUE); if (write_depth) { - /* TEX result.depth, fragment.texcoord[0], texture[0], 2D; */ - p->Instructions[ic].Opcode = OPCODE_TEX; - p->Instructions[ic].DstReg.File = PROGRAM_OUTPUT; - p->Instructions[ic].DstReg.Index = FRAG_RESULT_DEPTH; - p->Instructions[ic].DstReg.WriteMask = WRITEMASK_Z; - p->Instructions[ic].SrcReg[0].File = PROGRAM_INPUT; - p->Instructions[ic].SrcReg[0].Index = FRAG_ATTRIB_TEX0; - p->Instructions[ic].TexSrcUnit = 0; - p->Instructions[ic].TexSrcTarget = TEXTURE_2D_INDEX; - ic++; + color = ureg_DECL_fs_input(ureg, TGSI_SEMANTIC_COLOR, 0, + TGSI_INTERPOLATE_COLOR); + out_color = ureg_DECL_output(ureg, TGSI_SEMANTIC_COLOR, 0); + + depth_sampler = ureg_DECL_sampler(ureg, 0); + ureg_DECL_sampler_view(ureg, 0, TGSI_TEXTURE_2D, + TGSI_RETURN_TYPE_FLOAT, + TGSI_RETURN_TYPE_FLOAT, + TGSI_RETURN_TYPE_FLOAT, + TGSI_RETURN_TYPE_FLOAT); + out_depth = ureg_DECL_output(ureg, TGSI_SEMANTIC_POSITION, 0); } if (write_stencil) { - /* TEX result.stencil, fragment.texcoord[0], texture[0], 2D; */ - p->Instructions[ic].Opcode = OPCODE_TEX; - p->Instructions[ic].DstReg.File = PROGRAM_OUTPUT; - p->Instructions[ic].DstReg.Index = FRAG_RESULT_STENCIL; - p->Instructions[ic].DstReg.WriteMask = WRITEMASK_Y; - p->Instructions[ic].SrcReg[0].File = PROGRAM_INPUT; - p->Instructions[ic].SrcReg[0].Index = FRAG_ATTRIB_TEX0; - p->Instructions[ic].TexSrcUnit = 1; - p->Instructions[ic].TexSrcTarget = TEXTURE_2D_INDEX; - ic++; + stencil_sampler = ureg_DECL_sampler(ureg, 1); + ureg_DECL_sampler_view(ureg, 1, TGSI_TEXTURE_2D, + TGSI_RETURN_TYPE_UINT, + TGSI_RETURN_TYPE_UINT, + TGSI_RETURN_TYPE_UINT, + TGSI_RETURN_TYPE_UINT); + out_stencil = ureg_DECL_output(ureg, TGSI_SEMANTIC_STENCIL, 0); } - /* END; */ - p->Instructions[ic++].Opcode = OPCODE_END; - - assert(ic == p->NumInstructions); + texcoord = ureg_DECL_fs_input(ureg, + st->needs_texcoord_semantic ? + TGSI_SEMANTIC_TEXCOORD : + TGSI_SEMANTIC_GENERIC, + 0, TGSI_INTERPOLATE_LINEAR); - p->InputsRead = FRAG_BIT_TEX0 | FRAG_BIT_COL0; - p->OutputsWritten = 0; - if (write_depth) - p->OutputsWritten |= BITFIELD64_BIT(FRAG_RESULT_DEPTH); - if (write_stencil) - p->OutputsWritten |= BITFIELD64_BIT(FRAG_RESULT_STENCIL); + if (write_depth) { + ureg_TEX(ureg, ureg_writemask(out_depth, TGSI_WRITEMASK_Z), + TGSI_TEXTURE_2D, texcoord, depth_sampler); + ureg_MOV(ureg, out_color, color); + } - p->SamplersUsed = 0x1; /* sampler 0 (bit 0) is used */ if (write_stencil) - p->SamplersUsed |= 1 << 1; + ureg_TEX(ureg, ureg_writemask(out_stencil, TGSI_WRITEMASK_Y), + TGSI_TEXTURE_2D, texcoord, stencil_sampler); - fp = (struct gl_fragment_program *) p; + ureg_END(ureg); + cso = ureg_create_shader_and_destroy(ureg, st->pipe); /* save the new shader */ - st->drawpix.shaders[shaderIndex] = fp; - - return fp; + st->drawpix.zs_shaders[shaderIndex] = cso; + return cso; } @@ -286,29 +193,32 @@ static void * make_passthrough_vertex_shader(struct st_context *st, GLboolean passColor) { + const unsigned texcoord_semantic = st->needs_texcoord_semantic ? + TGSI_SEMANTIC_TEXCOORD : TGSI_SEMANTIC_GENERIC; + if (!st->drawpix.vert_shaders[passColor]) { - struct ureg_program *ureg = ureg_create( TGSI_PROCESSOR_VERTEX ); + struct ureg_program *ureg = ureg_create( PIPE_SHADER_VERTEX ); if (ureg == NULL) return NULL; /* MOV result.pos, vertex.pos; */ - ureg_MOV(ureg, + ureg_MOV(ureg, ureg_DECL_output( ureg, TGSI_SEMANTIC_POSITION, 0 ), ureg_DECL_vs_input( ureg, 0 )); - - /* MOV result.texcoord0, vertex.attr[1]; */ - ureg_MOV(ureg, - ureg_DECL_output( ureg, TGSI_SEMANTIC_GENERIC, 0 ), - ureg_DECL_vs_input( ureg, 1 )); - + if (passColor) { - /* MOV result.color0, vertex.attr[2]; */ - ureg_MOV(ureg, + /* MOV result.color0, vertex.attr[1]; */ + ureg_MOV(ureg, ureg_DECL_output( ureg, TGSI_SEMANTIC_COLOR, 0 ), - ureg_DECL_vs_input( ureg, 2 )); + ureg_DECL_vs_input( ureg, 1 )); } + /* MOV result.texcoord0, vertex.attr[2]; */ + ureg_MOV(ureg, + ureg_DECL_output( ureg, texcoord_semantic, 0 ), + ureg_DECL_vs_input( ureg, 2 )); + ureg_END( ureg ); st->drawpix.vert_shaders[passColor] = @@ -359,7 +269,7 @@ internal_format(struct gl_context *ctx, GLenum format, GLenum type) return GL_STENCIL_INDEX; default: - if (_mesa_is_integer_format(format)) { + if (_mesa_is_enum_format_integer(format)) { switch (type) { case GL_BYTE: return GL_RGBA8I; @@ -388,12 +298,16 @@ internal_format(struct gl_context *ctx, GLenum format, GLenum type) case GL_UNSIGNED_BYTE_3_3_2: case GL_UNSIGNED_BYTE_2_3_3_REV: + return GL_R3_G3_B2; + case GL_UNSIGNED_SHORT_4_4_4_4: case GL_UNSIGNED_SHORT_4_4_4_4_REV: return GL_RGBA4; case GL_UNSIGNED_SHORT_5_6_5: case GL_UNSIGNED_SHORT_5_6_5_REV: + return GL_RGB565; + case GL_UNSIGNED_SHORT_5_5_5_1: case GL_UNSIGNED_SHORT_1_5_5_5_REV: return GL_RGB5_A1; @@ -446,12 +360,12 @@ internal_format(struct gl_context *ctx, GLenum format, GLenum type) */ static struct pipe_resource * alloc_texture(struct st_context *st, GLsizei width, GLsizei height, - enum pipe_format texFormat) + enum pipe_format texFormat, unsigned bind) { struct pipe_resource *pt; pt = st_texture_create(st, st->internal_target, texFormat, 0, - width, height, 1, 1, PIPE_BIND_SAMPLER_VIEW); + width, height, 1, 1, 0, bind); return pt; } @@ -465,31 +379,81 @@ static struct pipe_resource * make_texture(struct st_context *st, GLsizei width, GLsizei height, GLenum format, GLenum type, const struct gl_pixelstore_attrib *unpack, - const GLvoid *pixels) + const void *pixels) { struct gl_context *ctx = st->ctx; struct pipe_context *pipe = st->pipe; - gl_format mformat; - struct pipe_resource *pt; + mesa_format mformat; + struct pipe_resource *pt = NULL; enum pipe_format pipeFormat; - GLenum baseInternalFormat, intFormat; + GLenum baseInternalFormat; + +#if USE_DRAWPIXELS_CACHE + const GLint bpp = _mesa_bytes_per_pixel(format, type); + + /* Check if the glDrawPixels() parameters and state matches the cache */ + if (width == st->drawpix_cache.width && + height == st->drawpix_cache.height && + format == st->drawpix_cache.format && + type == st->drawpix_cache.type && + pixels == st->drawpix_cache.user_pointer && + !_mesa_is_bufferobj(unpack->BufferObj) && + (unpack->RowLength == 0 || unpack->RowLength == width) && + unpack->SkipPixels == 0 && + unpack->SkipRows == 0 && + unpack->SwapBytes == GL_FALSE && + st->drawpix_cache.image) { + assert(st->drawpix_cache.texture); + + /* check if the pixel data is the same */ + if (memcmp(pixels, st->drawpix_cache.image, width * height * bpp) == 0) { + /* OK, re-use the cached texture */ + pipe_resource_reference(&pt, st->drawpix_cache.texture); + /* refcount of returned texture should be at least two here. One + * reference for the cache to hold on to, one for the caller (which + * it will release), and possibly more held by the driver. + */ + assert(pt->reference.count >= 2); + return pt; + } + } - intFormat = internal_format(ctx, format, type); - baseInternalFormat = _mesa_base_tex_format(ctx, intFormat); + /* discard the cached image and texture (if there is one) */ + st->drawpix_cache.width = 0; + st->drawpix_cache.height = 0; + st->drawpix_cache.user_pointer = NULL; + if (st->drawpix_cache.image) { + free(st->drawpix_cache.image); + st->drawpix_cache.image = NULL; + } + pipe_resource_reference(&st->drawpix_cache.texture, NULL); +#endif + + /* Choose a pixel format for the temp texture which will hold the + * image to draw. + */ + pipeFormat = st_choose_matching_format(st, PIPE_BIND_SAMPLER_VIEW, + format, type, unpack->SwapBytes); - mformat = st_ChooseTextureFormat_renderable(ctx, intFormat, - format, type, GL_FALSE); - assert(mformat); + if (pipeFormat == PIPE_FORMAT_NONE) { + /* Use the generic approach. */ + GLenum intFormat = internal_format(ctx, format, type); - pipeFormat = st_mesa_format_to_pipe_format(mformat); - assert(pipeFormat); + pipeFormat = st_choose_format(st, intFormat, format, type, + st->internal_target, 0, + PIPE_BIND_SAMPLER_VIEW, FALSE); + assert(pipeFormat != PIPE_FORMAT_NONE); + } + + mformat = st_pipe_format_to_mesa_format(pipeFormat); + baseInternalFormat = _mesa_get_format_base_format(mformat); pixels = _mesa_map_pbo_source(ctx, unpack, pixels); if (!pixels) return NULL; /* alloc temporary texture */ - pt = alloc_texture(st, width, height, pipeFormat); + pt = alloc_texture(st, width, height, pipeFormat, PIPE_BIND_SAMPLER_VIEW); if (!pt) { _mesa_unmap_pbo_source(ctx, unpack); return NULL; @@ -504,31 +468,48 @@ make_texture(struct st_context *st, /* we'll do pixel transfer in a fragment shader */ ctx->_ImageTransferState = 0x0; - transfer = pipe_get_transfer(st->pipe, pt, 0, 0, - PIPE_TRANSFER_WRITE, 0, 0, - width, height); - /* map texture transfer */ - dest = pipe_transfer_map(pipe, transfer); + dest = pipe_transfer_map(pipe, pt, 0, 0, + PIPE_TRANSFER_WRITE, 0, 0, + width, height, &transfer); /* Put image into texture transfer. * Note that the image is actually going to be upside down in * the texture. We deal with that with texcoords. */ - success = _mesa_texstore(ctx, 2, /* dims */ - baseInternalFormat, /* baseInternalFormat */ - mformat, /* gl_format */ - transfer->stride, /* dstRowStride, bytes */ - &dest, /* destSlices */ - width, height, 1, /* size */ - format, type, /* src format/type */ - pixels, /* data source */ - unpack); + if ((format == GL_RGBA || format == GL_BGRA) + && type == GL_UNSIGNED_BYTE) { + /* Use a memcpy-based texstore to avoid software pixel swizzling. + * We'll do the necessary swizzling with the pipe_sampler_view to + * give much better performance. + * XXX in the future, expand this to accomodate more format and + * type combinations. + */ + _mesa_memcpy_texture(ctx, 2, + mformat, /* mesa_format */ + transfer->stride, /* dstRowStride, bytes */ + &dest, /* destSlices */ + width, height, 1, /* size */ + format, type, /* src format/type */ + pixels, /* data source */ + unpack); + success = GL_TRUE; + } + else { + success = _mesa_texstore(ctx, 2, /* dims */ + baseInternalFormat, /* baseInternalFormat */ + mformat, /* mesa_format */ + transfer->stride, /* dstRowStride, bytes */ + &dest, /* destSlices */ + width, height, 1, /* size */ + format, type, /* src format/type */ + pixels, /* data source */ + unpack); + } /* unmap */ pipe_transfer_unmap(pipe, transfer); - pipe->transfer_destroy(pipe, transfer); assert(success); @@ -538,105 +519,35 @@ make_texture(struct st_context *st, _mesa_unmap_pbo_source(ctx, unpack); - return pt; -} - - -/** - * Draw quad with texcoords and optional color. - * Coords are gallium window coords with y=0=top. - * \param color may be null - * \param invertTex if true, flip texcoords vertically - */ -static void -draw_quad(struct gl_context *ctx, GLfloat x0, GLfloat y0, GLfloat z, - GLfloat x1, GLfloat y1, const GLfloat *color, - GLboolean invertTex, GLfloat maxXcoord, GLfloat maxYcoord) -{ - struct st_context *st = st_context(ctx); - struct pipe_context *pipe = st->pipe; - GLfloat (*verts)[3][4]; /* four verts, three attribs, XYZW */ - struct pipe_resource *buf = NULL; - unsigned offset; - - u_upload_alloc(st->uploader, 0, 4 * sizeof(verts[0]), &offset, &buf, - (void**)&verts); - if (!buf) { - return; - } - - /* setup vertex data */ - { - const struct gl_framebuffer *fb = st->ctx->DrawBuffer; - const GLfloat fb_width = (GLfloat) fb->Width; - const GLfloat fb_height = (GLfloat) fb->Height; - const GLfloat clip_x0 = x0 / fb_width * 2.0f - 1.0f; - const GLfloat clip_y0 = y0 / fb_height * 2.0f - 1.0f; - const GLfloat clip_x1 = x1 / fb_width * 2.0f - 1.0f; - const GLfloat clip_y1 = y1 / fb_height * 2.0f - 1.0f; - const GLfloat sLeft = 0.0f, sRight = maxXcoord; - const GLfloat tTop = invertTex ? maxYcoord : 0.0f; - const GLfloat tBot = invertTex ? 0.0f : maxYcoord; - GLuint i; - - /* upper-left */ - verts[0][0][0] = clip_x0; /* v[0].attr[0].x */ - verts[0][0][1] = clip_y0; /* v[0].attr[0].y */ - - /* upper-right */ - verts[1][0][0] = clip_x1; - verts[1][0][1] = clip_y0; - - /* lower-right */ - verts[2][0][0] = clip_x1; - verts[2][0][1] = clip_y1; - - /* lower-left */ - verts[3][0][0] = clip_x0; - verts[3][0][1] = clip_y1; - - verts[0][1][0] = sLeft; /* v[0].attr[1].S */ - verts[0][1][1] = tTop; /* v[0].attr[1].T */ - verts[1][1][0] = sRight; - verts[1][1][1] = tTop; - verts[2][1][0] = sRight; - verts[2][1][1] = tBot; - verts[3][1][0] = sLeft; - verts[3][1][1] = tBot; - - /* same for all verts: */ - if (color) { - for (i = 0; i < 4; i++) { - verts[i][0][2] = z; /* v[i].attr[0].z */ - verts[i][0][3] = 1.0f; /* v[i].attr[0].w */ - verts[i][2][0] = color[0]; /* v[i].attr[2].r */ - verts[i][2][1] = color[1]; /* v[i].attr[2].g */ - verts[i][2][2] = color[2]; /* v[i].attr[2].b */ - verts[i][2][3] = color[3]; /* v[i].attr[2].a */ - verts[i][1][2] = 0.0f; /* v[i].attr[1].R */ - verts[i][1][3] = 1.0f; /* v[i].attr[1].Q */ - } +#if USE_DRAWPIXELS_CACHE + /* Save the glDrawPixels parameter and image in the cache */ + if ((unpack->RowLength == 0 || unpack->RowLength == width) && + unpack->SkipPixels == 0 && + unpack->SkipRows == 0) { + st->drawpix_cache.width = width; + st->drawpix_cache.height = height; + st->drawpix_cache.format = format; + st->drawpix_cache.type = type; + st->drawpix_cache.user_pointer = pixels; + assert(!st->drawpix_cache.image); + st->drawpix_cache.image = malloc(width * height * bpp); + if (st->drawpix_cache.image) { + memcpy(st->drawpix_cache.image, pixels, width * height * bpp); + pipe_resource_reference(&st->drawpix_cache.texture, pt); } else { - for (i = 0; i < 4; i++) { - verts[i][0][2] = z; /*Z*/ - verts[i][0][3] = 1.0f; /*W*/ - verts[i][1][2] = 0.0f; /*R*/ - verts[i][1][3] = 1.0f; /*Q*/ - } + /* out of memory, free/disable cached texture */ + st->drawpix_cache.width = 0; + st->drawpix_cache.height = 0; + pipe_resource_reference(&st->drawpix_cache.texture, NULL); } } +#endif - u_upload_unmap(st->uploader); - util_draw_vertex_buffer(pipe, st->cso_context, buf, offset, - PIPE_PRIM_QUADS, - 4, /* verts */ - 3); /* attribs/vert */ - pipe_resource_reference(&buf, NULL); + return pt; } - static void draw_textured_quad(struct gl_context *ctx, GLint x, GLint y, GLfloat z, GLsizei width, GLsizei height, @@ -645,6 +556,7 @@ draw_textured_quad(struct gl_context *ctx, GLint x, GLint y, GLfloat z, int num_sampler_view, void *driver_vp, void *driver_fp, + struct st_fp_variant *fpv, const GLfloat *color, GLboolean invertTex, GLboolean write_depth, GLboolean write_stencil) @@ -652,9 +564,14 @@ draw_textured_quad(struct gl_context *ctx, GLint x, GLint y, GLfloat z, struct st_context *st = st_context(ctx); struct pipe_context *pipe = st->pipe; struct cso_context *cso = st->cso_context; + const unsigned fb_width = _mesa_geometric_width(ctx->DrawBuffer); + const unsigned fb_height = _mesa_geometric_height(ctx->DrawBuffer); GLfloat x0, y0, x1, y1; GLsizei maxSize; - boolean normalized = sv[0]->texture->target != PIPE_TEXTURE_RECT; + boolean normalized = sv[0]->texture->target == PIPE_TEXTURE_2D; + unsigned cso_state_mask; + + assert(sv[0]->texture->target == st->internal_target); /* limit checks */ /* XXX if DrawPixels image is larger than max texture size, break @@ -665,20 +582,19 @@ draw_textured_quad(struct gl_context *ctx, GLint x, GLint y, GLfloat z, assert(width <= maxSize); assert(height <= maxSize); - cso_save_rasterizer(cso); - cso_save_viewport(cso); - cso_save_samplers(cso); - cso_save_fragment_sampler_views(cso); - cso_save_fragment_shader(cso); - cso_save_stream_outputs(cso); - cso_save_vertex_shader(cso); - cso_save_geometry_shader(cso); - cso_save_vertex_elements(cso); - cso_save_vertex_buffers(cso); + cso_state_mask = (CSO_BIT_RASTERIZER | + CSO_BIT_VIEWPORT | + CSO_BIT_FRAGMENT_SAMPLERS | + CSO_BIT_FRAGMENT_SAMPLER_VIEWS | + CSO_BIT_STREAM_OUTPUTS | + CSO_BIT_VERTEX_ELEMENTS | + CSO_BIT_AUX_VERTEX_BUFFER_SLOT | + CSO_BITS_ALL_SHADERS); if (write_stencil) { - cso_save_depth_stencil_alpha(cso); - cso_save_blend(cso); + cso_state_mask |= (CSO_BIT_DEPTH_STENCIL_ALPHA | + CSO_BIT_BLEND); } + cso_save_state(cso, cso_state_mask); /* rasterizer state: just scissor */ { @@ -686,9 +602,10 @@ draw_textured_quad(struct gl_context *ctx, GLint x, GLint y, GLfloat z, memset(&rasterizer, 0, sizeof(rasterizer)); rasterizer.clamp_fragment_color = !st->clamp_frag_color_in_shader && ctx->Color._ClampFragmentColor; - rasterizer.gl_rasterization_rules = 1; + rasterizer.half_pixel_center = 1; + rasterizer.bottom_edge_rule = 1; rasterizer.depth_clip = !ctx->Transform.DepthClamp; - rasterizer.scissor = ctx->Scissor.Enabled; + rasterizer.scissor = ctx->Scissor.EnableFlags; cso_set_rasterizer(cso, &rasterizer); } @@ -724,12 +641,15 @@ draw_textured_quad(struct gl_context *ctx, GLint x, GLint y, GLfloat z, /* vertex shader state: position + texcoord pass-through */ cso_set_vertex_shader_handle(cso, driver_vp); - /* geometry shader state: disabled */ + /* disable other shaders */ + cso_set_tessctrl_shader_handle(cso, NULL); + cso_set_tesseval_shader_handle(cso, NULL); cso_set_geometry_shader_handle(cso, NULL); - /* texture sampling state: */ + /* user samplers, plus the drawpix samplers */ { struct pipe_sampler_state sampler; + memset(&sampler, 0, sizeof(sampler)); sampler.wrap_s = PIPE_TEX_WRAP_CLAMP; sampler.wrap_t = PIPE_TEX_WRAP_CLAMP; @@ -739,41 +659,62 @@ draw_textured_quad(struct gl_context *ctx, GLint x, GLint y, GLfloat z, sampler.mag_img_filter = PIPE_TEX_FILTER_NEAREST; sampler.normalized_coords = normalized; - cso_single_sampler(cso, 0, &sampler); - if (num_sampler_view > 1) { - cso_single_sampler(cso, 1, &sampler); + if (fpv) { + /* drawing a color image */ + const struct pipe_sampler_state *samplers[PIPE_MAX_SAMPLERS]; + uint num = MAX3(fpv->drawpix_sampler + 1, + fpv->pixelmap_sampler + 1, + st->state.num_samplers[PIPE_SHADER_FRAGMENT]); + uint i; + + for (i = 0; i < st->state.num_samplers[PIPE_SHADER_FRAGMENT]; i++) + samplers[i] = &st->state.samplers[PIPE_SHADER_FRAGMENT][i]; + + samplers[fpv->drawpix_sampler] = &sampler; + if (sv[1]) + samplers[fpv->pixelmap_sampler] = &sampler; + + cso_set_samplers(cso, PIPE_SHADER_FRAGMENT, num, samplers); + } else { + /* drawing a depth/stencil image */ + const struct pipe_sampler_state *samplers[2] = {&sampler, &sampler}; + + cso_set_samplers(cso, PIPE_SHADER_FRAGMENT, num_sampler_view, samplers); } - cso_single_sampler_done(cso); } - /* viewport state: viewport matching window dims */ - { - const float w = (float) ctx->DrawBuffer->Width; - const float h = (float) ctx->DrawBuffer->Height; - struct pipe_viewport_state vp; - vp.scale[0] = 0.5f * w; - vp.scale[1] = -0.5f * h; - vp.scale[2] = 0.5f; - vp.scale[3] = 1.0f; - vp.translate[0] = 0.5f * w; - vp.translate[1] = 0.5f * h; - vp.translate[2] = 0.5f; - vp.translate[3] = 0.0f; - cso_set_viewport(cso, &vp); + /* user textures, plus the drawpix textures */ + if (fpv) { + /* drawing a color image */ + struct pipe_sampler_view *sampler_views[PIPE_MAX_SAMPLERS]; + uint num = MAX3(fpv->drawpix_sampler + 1, + fpv->pixelmap_sampler + 1, + st->state.num_sampler_views[PIPE_SHADER_FRAGMENT]); + + memcpy(sampler_views, st->state.sampler_views[PIPE_SHADER_FRAGMENT], + sizeof(sampler_views)); + + sampler_views[fpv->drawpix_sampler] = sv[0]; + if (sv[1]) + sampler_views[fpv->pixelmap_sampler] = sv[1]; + cso_set_sampler_views(cso, PIPE_SHADER_FRAGMENT, num, sampler_views); + } else { + /* drawing a depth/stencil image */ + cso_set_sampler_views(cso, PIPE_SHADER_FRAGMENT, num_sampler_view, sv); } - cso_set_vertex_elements(cso, 3, st->velems_util_draw); - cso_set_stream_outputs(st->cso_context, 0, NULL, 0); + /* viewport state: viewport matching window dims */ + cso_set_viewport_dims(cso, fb_width, fb_height, TRUE); - /* texture state: */ - cso_set_fragment_sampler_views(cso, num_sampler_view, sv); + cso_set_vertex_elements(cso, 3, st->util_velems); + cso_set_stream_outputs(cso, 0, NULL, NULL); /* Compute Gallium window coords (y=0=top) with pixel zoom. * Recall that these coords are transformed by the current * vertex shader and viewport transformation. */ if (st_fb_orientation(ctx->DrawBuffer) == Y_0_BOTTOM) { - y = ctx->DrawBuffer->Height - (int) (y + height * ctx->Pixel.ZoomY); + y = fb_height - (int) (y + height * ctx->Pixel.ZoomY); invertTex = !invertTex; } @@ -783,27 +724,29 @@ draw_textured_quad(struct gl_context *ctx, GLint x, GLint y, GLfloat z, y1 = y + height * ctx->Pixel.ZoomY; /* convert Z from [0,1] to [-1,-1] to match viewport Z scale/bias */ - z = z * 2.0 - 1.0; + z = z * 2.0f - 1.0f; - draw_quad(ctx, x0, y0, z, x1, y1, color, invertTex, - normalized ? ((GLfloat) width / sv[0]->texture->width0) : (GLfloat)width, - normalized ? ((GLfloat) height / sv[0]->texture->height0) : (GLfloat)height); + { + const float clip_x0 = x0 / (float) fb_width * 2.0f - 1.0f; + const float clip_y0 = y0 / (float) fb_height * 2.0f - 1.0f; + const float clip_x1 = x1 / (float) fb_width * 2.0f - 1.0f; + const float clip_y1 = y1 / (float) fb_height * 2.0f - 1.0f; + const float maxXcoord = normalized ? + ((float) width / sv[0]->texture->width0) : (float) width; + const float maxYcoord = normalized + ? ((float) height / sv[0]->texture->height0) : (float) height; + const float sLeft = 0.0f, sRight = maxXcoord; + const float tTop = invertTex ? maxYcoord : 0.0f; + const float tBot = invertTex ? 0.0f : maxYcoord; + + if (!st_draw_quad(st, clip_x0, clip_y0, clip_x1, clip_y1, z, + sLeft, tBot, sRight, tTop, color, 0)) { + _mesa_error(ctx, GL_OUT_OF_MEMORY, "glDrawPixels"); + } + } /* restore state */ - cso_restore_rasterizer(cso); - cso_restore_viewport(cso); - cso_restore_samplers(cso); - cso_restore_fragment_sampler_views(cso); - cso_restore_fragment_shader(cso); - cso_restore_vertex_shader(cso); - cso_restore_geometry_shader(cso); - cso_restore_vertex_elements(cso); - cso_restore_vertex_buffers(cso); - cso_restore_stream_outputs(cso); - if (write_stencil) { - cso_restore_depth_stencil_alpha(cso); - cso_restore_blend(cso); - } + cso_restore_state(cso); } @@ -815,7 +758,7 @@ static void draw_stencil_pixels(struct gl_context *ctx, GLint x, GLint y, GLsizei width, GLsizei height, GLenum format, GLenum type, const struct gl_pixelstore_attrib *unpack, - const GLvoid *pixels) + const void *pixels) { struct st_context *st = st_context(ctx); struct pipe_context *pipe = st->pipe; @@ -852,25 +795,24 @@ draw_stencil_pixels(struct gl_context *ctx, GLint x, GLint y, usage = PIPE_TRANSFER_WRITE; } - pt = pipe_get_transfer(pipe, strb->texture, - strb->rtt_level, strb->rtt_face + strb->rtt_slice, - usage, x, y, - width, height); - - stmap = pipe_transfer_map(pipe, pt); + stmap = pipe_transfer_map(pipe, strb->texture, + strb->surface->u.tex.level, + strb->surface->u.tex.first_layer, + usage, x, y, + width, height, &pt); pixels = _mesa_map_pbo_source(ctx, &clippedUnpack, pixels); assert(pixels); - sValues = (GLubyte *) malloc(width * sizeof(GLubyte)); - zValues = (GLuint *) malloc(width * sizeof(GLuint)); + sValues = malloc(width * sizeof(GLubyte)); + zValues = malloc(width * sizeof(GLuint)); if (sValues && zValues) { GLint row; for (row = 0; row < height; row++) { GLfloat *zValuesFloat = (GLfloat*)zValues; GLenum destType = GL_UNSIGNED_BYTE; - const GLvoid *source = _mesa_image_address2d(&clippedUnpack, pixels, + const void *source = _mesa_image_address2d(&clippedUnpack, pixels, width, height, format, type, row, 0); @@ -985,7 +927,6 @@ draw_stencil_pixels(struct gl_context *ctx, GLint x, GLint y, /* unmap the stencil buffer */ pipe_transfer_unmap(pipe, pt); - pipe->transfer_destroy(pipe, pt); } @@ -1002,7 +943,7 @@ get_color_fp_variant(struct st_context *st) memset(&key, 0, sizeof(key)); - key.st = st; + key.st = st->has_shareable_shaders ? NULL : st; key.drawpixels = 1; key.scaleAndBias = (ctx->Pixel.RedBias != 0.0 || ctx->Pixel.RedScale != 1.0 || @@ -1014,7 +955,7 @@ get_color_fp_variant(struct st_context *st) ctx->Pixel.AlphaScale != 1.0); key.pixelMaps = ctx->Pixel.MapColorFlag; key.clamp_color = st->clamp_frag_color_in_shader && - st->ctx->Color._ClampFragmentColor; + ctx->Color._ClampFragmentColor; fpv = st_get_fp_variant(st, st->fp, &key); @@ -1023,26 +964,87 @@ get_color_fp_variant(struct st_context *st) /** - * Get fragment program variant for a glDrawPixels or glCopyPixels - * command for depth/stencil data. + * Clamp glDrawPixels width and height to the maximum texture size. */ -static struct st_fp_variant * -get_depth_stencil_fp_variant(struct st_context *st, GLboolean write_depth, - GLboolean write_stencil) +static void +clamp_size(struct pipe_context *pipe, GLsizei *width, GLsizei *height, + struct gl_pixelstore_attrib *unpack) { - struct st_fp_variant_key key; - struct st_fp_variant *fpv; + const int maxSize = + 1 << (pipe->screen->get_param(pipe->screen, + PIPE_CAP_MAX_TEXTURE_2D_LEVELS) - 1); + + if (*width > maxSize) { + if (unpack->RowLength == 0) + unpack->RowLength = *width; + *width = maxSize; + } + if (*height > maxSize) { + *height = maxSize; + } +} - memset(&key, 0, sizeof(key)); - key.st = st; - key.drawpixels = 1; - key.drawpixels_z = write_depth; - key.drawpixels_stencil = write_stencil; +/** + * Search the array of 4 swizzle components for the named component and return + * its position. + */ +static unsigned +search_swizzle(const unsigned char swizzle[4], unsigned component) +{ + unsigned i; + for (i = 0; i < 4; i++) { + if (swizzle[i] == component) + return i; + } + assert(!"search_swizzle() failed"); + return 0; +} - fpv = st_get_fp_variant(st, st->fp, &key); - return fpv; +/** + * Set the sampler view's swizzle terms. This is used to handle RGBA + * swizzling when the incoming image format isn't an exact match for + * the actual texture format. For example, if we have glDrawPixels( + * GL_RGBA, GL_UNSIGNED_BYTE) and we chose the texture format + * PIPE_FORMAT_B8G8R8A8 then we can do use the sampler view swizzle to + * avoid swizzling all the pixels in software in the texstore code. + */ +static void +setup_sampler_swizzle(struct pipe_sampler_view *sv, GLenum format, GLenum type) +{ + if ((format == GL_RGBA || format == GL_BGRA) && type == GL_UNSIGNED_BYTE) { + const struct util_format_description *desc = + util_format_description(sv->texture->format); + unsigned c0, c1, c2, c3; + + /* Every gallium driver supports at least one 32-bit packed RGBA format. + * We must have chosen one for (GL_RGBA, GL_UNSIGNED_BYTE). + */ + assert(desc->block.bits == 32); + + /* invert the format's swizzle to setup the sampler's swizzle */ + if (format == GL_RGBA) { + c0 = PIPE_SWIZZLE_X; + c1 = PIPE_SWIZZLE_Y; + c2 = PIPE_SWIZZLE_Z; + c3 = PIPE_SWIZZLE_W; + } + else { + assert(format == GL_BGRA); + c0 = PIPE_SWIZZLE_Z; + c1 = PIPE_SWIZZLE_Y; + c2 = PIPE_SWIZZLE_X; + c3 = PIPE_SWIZZLE_W; + } + sv->swizzle_r = search_swizzle(desc->swizzle, c0); + sv->swizzle_g = search_swizzle(desc->swizzle, c1); + sv->swizzle_b = search_swizzle(desc->swizzle, c2); + sv->swizzle_a = search_swizzle(desc->swizzle, c3); + } + else { + /* use the default sampler swizzle */ + } } @@ -1053,16 +1055,32 @@ static void st_DrawPixels(struct gl_context *ctx, GLint x, GLint y, GLsizei width, GLsizei height, GLenum format, GLenum type, - const struct gl_pixelstore_attrib *unpack, const GLvoid *pixels) + const struct gl_pixelstore_attrib *unpack, const void *pixels) { void *driver_vp, *driver_fp; struct st_context *st = st_context(ctx); - const GLfloat *color; struct pipe_context *pipe = st->pipe; GLboolean write_stencil = GL_FALSE, write_depth = GL_FALSE; - struct pipe_sampler_view *sv[2]; + struct pipe_sampler_view *sv[2] = { NULL }; int num_sampler_view = 1; - struct st_fp_variant *fpv; + struct gl_pixelstore_attrib clippedUnpack; + struct st_fp_variant *fpv = NULL; + struct pipe_resource *pt; + + /* Mesa state should be up to date by now */ + assert(ctx->NewState == 0x0); + + st_flush_bitmap_cache(st); + + st_validate_state(st, ST_PIPELINE_RENDER); + + /* Limit the size of the glDrawPixels to the max texture size. + * Strictly speaking, that's not correct but since we don't handle + * larger images yet, this is better than crashing. + */ + clippedUnpack = *unpack; + unpack = &clippedUnpack; + clamp_size(st->pipe, &width, &height, &clippedUnpack); if (format == GL_DEPTH_STENCIL) write_stencil = write_depth = GL_TRUE; @@ -1079,94 +1097,85 @@ st_DrawPixels(struct gl_context *ctx, GLint x, GLint y, return; } - /* Mesa state should be up to date by now */ - assert(ctx->NewState == 0x0); - - st_validate_state(st); - /* * Get vertex/fragment shaders */ if (write_depth || write_stencil) { - fpv = get_depth_stencil_fp_variant(st, write_depth, write_stencil); - - driver_fp = fpv->driver_shader; - + driver_fp = get_drawpix_z_stencil_program(st, write_depth, + write_stencil); driver_vp = make_passthrough_vertex_shader(st, GL_TRUE); - - color = ctx->Current.RasterColor; } else { fpv = get_color_fp_variant(st); driver_fp = fpv->driver_shader; - driver_vp = make_passthrough_vertex_shader(st, GL_FALSE); - color = NULL; - if (st->pixel_xfer.pixelmap_enabled) { - sv[1] = st->pixel_xfer.pixelmap_sampler_view; - num_sampler_view++; + if (ctx->Pixel.MapColorFlag) { + pipe_sampler_view_reference(&sv[1], + st->pixel_xfer.pixelmap_sampler_view); + num_sampler_view++; } + + /* compiling a new fragment shader variant added new state constants + * into the constant buffer, we need to update them + */ + st_upload_constants(st, st->fp->Base.Base.Parameters, + PIPE_SHADER_FRAGMENT); } - /* update fragment program constants */ - st_upload_constants(st, fpv->parameters, PIPE_SHADER_FRAGMENT); + /* Put glDrawPixels image into a texture */ + pt = make_texture(st, width, height, format, type, unpack, pixels); + if (!pt) { + _mesa_error(ctx, GL_OUT_OF_MEMORY, "glDrawPixels"); + return; + } - /* draw with textured quad */ - { - struct pipe_resource *pt - = make_texture(st, width, height, format, type, unpack, pixels); - if (pt) { - sv[0] = st_create_texture_sampler_view(st->pipe, pt); - - if (sv[0]) { - /* Create a second sampler view to read stencil. - * The stencil is written using the shader stencil export - * functionality. */ - if (write_stencil) { - enum pipe_format stencil_format = PIPE_FORMAT_NONE; - - switch (pt->format) { - case PIPE_FORMAT_Z24_UNORM_S8_UINT: - case PIPE_FORMAT_X24S8_UINT: - stencil_format = PIPE_FORMAT_X24S8_UINT; - break; - case PIPE_FORMAT_S8_UINT_Z24_UNORM: - case PIPE_FORMAT_S8X24_UINT: - stencil_format = PIPE_FORMAT_S8X24_UINT; - break; - case PIPE_FORMAT_Z32_FLOAT_S8X24_UINT: - case PIPE_FORMAT_X32_S8X24_UINT: - stencil_format = PIPE_FORMAT_X32_S8X24_UINT; - break; - case PIPE_FORMAT_S8_UINT: - stencil_format = PIPE_FORMAT_S8_UINT; - break; - default: - assert(0); - } + /* create sampler view for the image */ + sv[0] = st_create_texture_sampler_view(st->pipe, pt); + if (!sv[0]) { + _mesa_error(ctx, GL_OUT_OF_MEMORY, "glDrawPixels"); + pipe_resource_reference(&pt, NULL); + return; + } - sv[1] = st_create_texture_sampler_view_format(st->pipe, pt, - stencil_format); - num_sampler_view++; - } + /* Set up the sampler view's swizzle */ + setup_sampler_swizzle(sv[0], format, type); - draw_textured_quad(ctx, x, y, ctx->Current.RasterPos[2], - width, height, - ctx->Pixel.ZoomX, ctx->Pixel.ZoomY, - sv, - num_sampler_view, - driver_vp, - driver_fp, - color, GL_FALSE, write_depth, write_stencil); - pipe_sampler_view_reference(&sv[0], NULL); - if (num_sampler_view > 1) - pipe_sampler_view_reference(&sv[1], NULL); - } + /* Create a second sampler view to read stencil. The stencil is + * written using the shader stencil export functionality. + */ + if (write_stencil) { + enum pipe_format stencil_format = + util_format_stencil_only(pt->format); + /* we should not be doing pixel map/transfer (see above) */ + assert(num_sampler_view == 1); + sv[1] = st_create_texture_sampler_view_format(st->pipe, pt, + stencil_format); + if (!sv[1]) { + _mesa_error(ctx, GL_OUT_OF_MEMORY, "glDrawPixels"); pipe_resource_reference(&pt, NULL); + pipe_sampler_view_reference(&sv[0], NULL); + return; } + num_sampler_view++; } + + draw_textured_quad(ctx, x, y, ctx->Current.RasterPos[2], + width, height, + ctx->Pixel.ZoomX, ctx->Pixel.ZoomY, + sv, + num_sampler_view, + driver_vp, + driver_fp, fpv, + ctx->Current.RasterColor, + GL_FALSE, write_depth, write_stencil); + pipe_sampler_view_reference(&sv[0], NULL); + if (num_sampler_view > 1) + pipe_sampler_view_reference(&sv[1], NULL); + + /* free the texture (but may persist in the cache) */ + pipe_resource_reference(&pt, NULL); } @@ -1223,18 +1232,16 @@ copy_stencil_pixels(struct gl_context *ctx, GLint srcx, GLint srcy, dsty = rbDraw->Base.Height - dsty - height; } - ptDraw = pipe_get_transfer(pipe, - rbDraw->texture, - rbDraw->rtt_level, - rbDraw->rtt_face + rbDraw->rtt_slice, - usage, dstx, dsty, - width, height); - - assert(util_format_get_blockwidth(ptDraw->resource->format) == 1); - assert(util_format_get_blockheight(ptDraw->resource->format) == 1); + assert(util_format_get_blockwidth(rbDraw->texture->format) == 1); + assert(util_format_get_blockheight(rbDraw->texture->format) == 1); /* map the stencil buffer */ - drawMap = pipe_transfer_map(pipe, ptDraw); + drawMap = pipe_transfer_map(pipe, + rbDraw->texture, + rbDraw->surface->u.tex.level, + rbDraw->surface->u.tex.first_layer, + usage, dstx, dsty, + width, height, &ptDraw); /* draw */ /* XXX PixelZoom not handled yet */ @@ -1259,7 +1266,6 @@ copy_stencil_pixels(struct gl_context *ctx, GLint srcx, GLint srcy, /* unmap the stencil buffer */ pipe_transfer_unmap(pipe, ptDraw); - pipe->transfer_destroy(pipe, ptDraw); } @@ -1277,29 +1283,13 @@ st_get_color_read_renderbuffer(struct gl_context *ctx) } -/** Do the src/dest regions overlap? */ -static GLboolean -regions_overlap(GLint srcX, GLint srcY, GLint dstX, GLint dstY, - GLsizei width, GLsizei height) -{ - if (srcX + width <= dstX || - dstX + width <= srcX || - srcY + height <= dstY || - dstY + height <= srcY) - return GL_FALSE; - else - return GL_TRUE; -} - - /** * Try to do a glCopyPixels for simple cases with a blit by calling - * pipe->resource_copy_region(). + * pipe->blit(). * * We can do this when we're copying color pixels (depth/stencil * eventually) with no pixel zoom, no pixel transfer ops, no - * per-fragment ops, the src/dest regions don't overlap and the - * src/dest pixel formats are the same. + * per-fragment ops, and the src/dest regions don't overlap. */ static GLboolean blit_copy_pixels(struct gl_context *ctx, GLint srcx, GLint srcy, @@ -1308,8 +1298,9 @@ blit_copy_pixels(struct gl_context *ctx, GLint srcx, GLint srcy, { struct st_context *st = st_context(ctx); struct pipe_context *pipe = st->pipe; + struct pipe_screen *screen = pipe->screen; struct gl_pixelstore_attrib pack, unpack; - GLint readX, readY, readW, readH; + GLint readX, readY, readW, readH, drawX, drawY, drawW, drawH; if (type == GL_COLOR && ctx->Pixel.ZoomX == 1.0 && @@ -1317,17 +1308,18 @@ blit_copy_pixels(struct gl_context *ctx, GLint srcx, GLint srcy, ctx->_ImageTransferState == 0x0 && !ctx->Color.BlendEnabled && !ctx->Color.AlphaEnabled && + (!ctx->Color.ColorLogicOpEnabled || ctx->Color.LogicOp == GL_COPY) && !ctx->Depth.Test && !ctx->Fog.Enabled && !ctx->Stencil.Enabled && !ctx->FragmentProgram.Enabled && !ctx->VertexProgram.Enabled && - !ctx->Shader.CurrentFragmentProgram && - st_fb_orientation(ctx->ReadBuffer) == st_fb_orientation(ctx->DrawBuffer) && + !ctx->_Shader->CurrentProgram[MESA_SHADER_FRAGMENT] && + !ctx->ATIFragmentShader._Enabled && ctx->DrawBuffer->_NumColorDrawBuffers == 1 && - !ctx->Query.CondRenderQuery) { + !ctx->Query.CondRenderQuery && + !ctx->Query.CurrentOcclusionObject) { struct st_renderbuffer *rbRead, *rbDraw; - GLint drawX, drawY; /* * Clip the read region against the src buffer bounds. @@ -1354,29 +1346,65 @@ blit_copy_pixels(struct gl_context *ctx, GLint srcx, GLint srcy, readX = readX - pack.SkipPixels + unpack.SkipPixels; readY = readY - pack.SkipRows + unpack.SkipRows; + drawW = readW; + drawH = readH; + rbRead = st_get_color_read_renderbuffer(ctx); rbDraw = st_renderbuffer(ctx->DrawBuffer->_ColorDrawBuffers[0]); - if ((rbRead != rbDraw || - !regions_overlap(readX, readY, drawX, drawY, readW, readH)) && - rbRead->Base.Format == rbDraw->Base.Format) { - struct pipe_box srcBox; - - /* flip src/dst position if needed */ - if (st_fb_orientation(ctx->ReadBuffer) == Y_0_TOP) { - /* both buffers will have the same orientation */ - readY = ctx->ReadBuffer->Height - readY - readH; - drawY = ctx->DrawBuffer->Height - drawY - readH; - } + /* Flip src/dst position depending on the orientation of buffers. */ + if (st_fb_orientation(ctx->ReadBuffer) == Y_0_TOP) { + readY = rbRead->Base.Height - readY; + readH = -readH; + } - u_box_2d(readX, readY, readW, readH, &srcBox); + if (st_fb_orientation(ctx->DrawBuffer) == Y_0_TOP) { + /* We can't flip the destination for pipe->blit, so we only adjust + * its position and flip the source. + */ + drawY = rbDraw->Base.Height - drawY - drawH; + readY += readH; + readH = -readH; + } - pipe->resource_copy_region(pipe, - rbDraw->texture, - rbDraw->rtt_level, drawX, drawY, 0, - rbRead->texture, - rbRead->rtt_level, &srcBox); - return GL_TRUE; + if (rbRead != rbDraw || + !_mesa_regions_overlap(readX, readY, readX + readW, readY + readH, + drawX, drawY, drawX + drawW, drawY + drawH)) { + struct pipe_blit_info blit; + + memset(&blit, 0, sizeof(blit)); + blit.src.resource = rbRead->texture; + blit.src.level = rbRead->surface->u.tex.level; + blit.src.format = rbRead->texture->format; + blit.src.box.x = readX; + blit.src.box.y = readY; + blit.src.box.z = rbRead->surface->u.tex.first_layer; + blit.src.box.width = readW; + blit.src.box.height = readH; + blit.src.box.depth = 1; + blit.dst.resource = rbDraw->texture; + blit.dst.level = rbDraw->surface->u.tex.level; + blit.dst.format = rbDraw->texture->format; + blit.dst.box.x = drawX; + blit.dst.box.y = drawY; + blit.dst.box.z = rbDraw->surface->u.tex.first_layer; + blit.dst.box.width = drawW; + blit.dst.box.height = drawH; + blit.dst.box.depth = 1; + blit.mask = PIPE_MASK_RGBA; + blit.filter = PIPE_TEX_FILTER_NEAREST; + + if (screen->is_format_supported(screen, blit.src.format, + blit.src.resource->target, + blit.src.resource->nr_samples, + PIPE_BIND_SAMPLER_VIEW) && + screen->is_format_supported(screen, blit.dst.format, + blit.dst.resource->target, + blit.dst.resource->nr_samples, + PIPE_BIND_RENDER_TARGET)) { + pipe->blit(pipe, &blit); + return GL_TRUE; + } } } @@ -1395,17 +1423,18 @@ st_CopyPixels(struct gl_context *ctx, GLint srcx, GLint srcy, struct st_renderbuffer *rbRead; void *driver_vp, *driver_fp; struct pipe_resource *pt; - struct pipe_sampler_view *sv[2]; + struct pipe_sampler_view *sv[2] = { NULL }; + struct st_fp_variant *fpv = NULL; int num_sampler_view = 1; - GLfloat *color; - enum pipe_format srcFormat, texFormat; + enum pipe_format srcFormat; + unsigned srcBind; GLboolean invertTex = GL_FALSE; GLint readX, readY, readW, readH; - GLuint sample_count; struct gl_pixelstore_attrib pack = ctx->DefaultPacking; - struct st_fp_variant *fpv; - st_validate_state(st); + st_flush_bitmap_cache(st); + + st_validate_state(st, ST_PIPELINE_RENDER); if (type == GL_DEPTH_STENCIL) { /* XXX make this more efficient */ @@ -1435,61 +1464,80 @@ st_CopyPixels(struct gl_context *ctx, GLint srcx, GLint srcy, * Get vertex/fragment shaders */ if (type == GL_COLOR) { + fpv = get_color_fp_variant(st); + rbRead = st_get_color_read_renderbuffer(ctx); - color = NULL; - fpv = get_color_fp_variant(st); driver_fp = fpv->driver_shader; - driver_vp = make_passthrough_vertex_shader(st, GL_FALSE); - if (st->pixel_xfer.pixelmap_enabled) { - sv[1] = st->pixel_xfer.pixelmap_sampler_view; - num_sampler_view++; + if (ctx->Pixel.MapColorFlag) { + pipe_sampler_view_reference(&sv[1], + st->pixel_xfer.pixelmap_sampler_view); + num_sampler_view++; } + + /* compiling a new fragment shader variant added new state constants + * into the constant buffer, we need to update them + */ + st_upload_constants(st, st->fp->Base.Base.Parameters, + PIPE_SHADER_FRAGMENT); } else { assert(type == GL_DEPTH); rbRead = st_renderbuffer(ctx->ReadBuffer-> Attachment[BUFFER_DEPTH].Renderbuffer); - color = ctx->Current.Attrib[VERT_ATTRIB_COLOR0]; - - fpv = get_depth_stencil_fp_variant(st, GL_TRUE, GL_FALSE); - driver_fp = fpv->driver_shader; + driver_fp = get_drawpix_z_stencil_program(st, GL_TRUE, GL_FALSE); driver_vp = make_passthrough_vertex_shader(st, GL_TRUE); } - /* update fragment program constants */ - st_upload_constants(st, fpv->parameters, PIPE_SHADER_FRAGMENT); - - sample_count = rbRead->texture->nr_samples; - /* I believe this would be legal, presumably would need to do a resolve - for color, and for depth/stencil spec says to just use one of the - depth/stencil samples per pixel? Need some transfer clarifications. */ - assert(sample_count < 2); - + /* Choose the format for the temporary texture. */ srcFormat = rbRead->texture->format; + srcBind = PIPE_BIND_SAMPLER_VIEW | + (type == GL_COLOR ? PIPE_BIND_RENDER_TARGET : PIPE_BIND_DEPTH_STENCIL); - if (screen->is_format_supported(screen, srcFormat, st->internal_target, - sample_count, - PIPE_BIND_SAMPLER_VIEW)) { - texFormat = srcFormat; - } - else { - /* srcFormat can't be used as a texture format */ + if (!screen->is_format_supported(screen, srcFormat, st->internal_target, 0, + srcBind)) { + /* srcFormat is non-renderable. Find a compatible renderable format. */ if (type == GL_DEPTH) { - texFormat = st_choose_format(screen, GL_DEPTH_COMPONENT, - GL_NONE, GL_NONE, st->internal_target, - sample_count, PIPE_BIND_DEPTH_STENCIL); - assert(texFormat != PIPE_FORMAT_NONE); + srcFormat = st_choose_format(st, GL_DEPTH_COMPONENT, GL_NONE, + GL_NONE, st->internal_target, 0, + srcBind, FALSE); } else { - /* default color format */ - texFormat = st_choose_format(screen, GL_RGBA, - GL_NONE, GL_NONE, st->internal_target, - sample_count, PIPE_BIND_SAMPLER_VIEW); - assert(texFormat != PIPE_FORMAT_NONE); + assert(type == GL_COLOR); + + if (util_format_is_float(srcFormat)) { + srcFormat = st_choose_format(st, GL_RGBA32F, GL_NONE, + GL_NONE, st->internal_target, 0, + srcBind, FALSE); + } + else if (util_format_is_pure_sint(srcFormat)) { + srcFormat = st_choose_format(st, GL_RGBA32I, GL_NONE, + GL_NONE, st->internal_target, 0, + srcBind, FALSE); + } + else if (util_format_is_pure_uint(srcFormat)) { + srcFormat = st_choose_format(st, GL_RGBA32UI, GL_NONE, + GL_NONE, st->internal_target, 0, + srcBind, FALSE); + } + else if (util_format_is_snorm(srcFormat)) { + srcFormat = st_choose_format(st, GL_RGBA16_SNORM, GL_NONE, + GL_NONE, st->internal_target, 0, + srcBind, FALSE); + } + else { + srcFormat = st_choose_format(st, GL_RGBA, GL_NONE, + GL_NONE, st->internal_target, 0, + srcBind, FALSE); + } + } + + if (srcFormat == PIPE_FORMAT_NONE) { + assert(0 && "cannot choose a format for src of CopyPixels"); + return; } } @@ -1509,12 +1557,20 @@ st_CopyPixels(struct gl_context *ctx, GLint srcx, GLint srcy, readY = srcy; readW = width; readH = height; - _mesa_clip_readpixels(ctx, &readX, &readY, &readW, &readH, &pack); + if (!_mesa_clip_readpixels(ctx, &readX, &readY, &readW, &readH, &pack)) { + /* The source region is completely out of bounds. Do nothing. + * The GL spec says "Results of copies from outside the window, + * or from regions of the window that are not exposed, are + * hardware dependent and undefined." + */ + return; + } + readW = MAX2(0, readW); readH = MAX2(0, readH); - /* alloc temporary texture */ - pt = alloc_texture(st, width, height, texFormat); + /* Allocate the temporary texture. */ + pt = alloc_texture(st, width, height, srcFormat, srcBind); if (!pt) return; @@ -1524,67 +1580,33 @@ st_CopyPixels(struct gl_context *ctx, GLint srcx, GLint srcy, return; } - /* Make temporary texture which is a copy of the src region. - */ - if (srcFormat == texFormat) { - struct pipe_box src_box; - u_box_2d(readX, readY, readW, readH, &src_box); - /* copy source framebuffer surface into mipmap/texture */ - pipe->resource_copy_region(pipe, - pt, /* dest tex */ - 0, /* dest lvl */ - pack.SkipPixels, pack.SkipRows, 0, /* dest pos */ - rbRead->texture, /* src tex */ - rbRead->rtt_level, /* src lvl */ - &src_box); - - } - else { - /* CPU-based fallback/conversion */ - struct pipe_transfer *ptRead = - pipe_get_transfer(st->pipe, rbRead->texture, - rbRead->rtt_level, - rbRead->rtt_face + rbRead->rtt_slice, - PIPE_TRANSFER_READ, - readX, readY, readW, readH); - struct pipe_transfer *ptTex; - enum pipe_transfer_usage transfer_usage; - - if (ST_DEBUG & DEBUG_FALLBACK) - debug_printf("%s: fallback processing\n", __FUNCTION__); - - if (type == GL_DEPTH && util_format_is_depth_and_stencil(pt->format)) - transfer_usage = PIPE_TRANSFER_READ_WRITE; - else - transfer_usage = PIPE_TRANSFER_WRITE; - - ptTex = pipe_get_transfer(st->pipe, pt, 0, 0, transfer_usage, - 0, 0, width, height); - - /* copy image from ptRead surface to ptTex surface */ - if (type == GL_COLOR) { - /* alternate path using get/put_tile() */ - GLfloat *buf = (GLfloat *) malloc(width * height * 4 * sizeof(GLfloat)); - enum pipe_format readFormat, drawFormat; - readFormat = util_format_linear(rbRead->texture->format); - drawFormat = util_format_linear(pt->format); - pipe_get_tile_rgba_format(pipe, ptRead, 0, 0, readW, readH, - readFormat, buf); - pipe_put_tile_rgba_format(pipe, ptTex, pack.SkipPixels, pack.SkipRows, - readW, readH, drawFormat, buf); - free(buf); - } - else { - /* GL_DEPTH */ - GLuint *buf = (GLuint *) malloc(width * height * sizeof(GLuint)); - pipe_get_tile_z(pipe, ptRead, 0, 0, readW, readH, buf); - pipe_put_tile_z(pipe, ptTex, pack.SkipPixels, pack.SkipRows, - readW, readH, buf); - free(buf); - } - - pipe->transfer_destroy(pipe, ptRead); - pipe->transfer_destroy(pipe, ptTex); + /* Copy the src region to the temporary texture. */ + { + struct pipe_blit_info blit; + + memset(&blit, 0, sizeof(blit)); + blit.src.resource = rbRead->texture; + blit.src.level = rbRead->surface->u.tex.level; + blit.src.format = rbRead->texture->format; + blit.src.box.x = readX; + blit.src.box.y = readY; + blit.src.box.z = rbRead->surface->u.tex.first_layer; + blit.src.box.width = readW; + blit.src.box.height = readH; + blit.src.box.depth = 1; + blit.dst.resource = pt; + blit.dst.level = 0; + blit.dst.format = pt->format; + blit.dst.box.x = pack.SkipPixels; + blit.dst.box.y = pack.SkipRows; + blit.dst.box.z = 0; + blit.dst.box.width = readW; + blit.dst.box.height = readH; + blit.dst.box.depth = 1; + blit.mask = util_format_get_mask(pt->format) & ~PIPE_MASK_S; + blit.filter = PIPE_TEX_FILTER_NEAREST; + + pipe->blit(pipe, &blit); } /* OK, the texture 'pt' contains the src image/pixels. Now draw a @@ -1595,8 +1617,9 @@ st_CopyPixels(struct gl_context *ctx, GLint srcx, GLint srcy, sv, num_sampler_view, driver_vp, - driver_fp, - color, invertTex, GL_FALSE, GL_FALSE); + driver_fp, fpv, + ctx->Current.Attrib[VERT_ATTRIB_COLOR0], + invertTex, GL_FALSE, GL_FALSE); pipe_resource_reference(&pt, NULL); pipe_sampler_view_reference(&sv[0], NULL); @@ -1616,16 +1639,14 @@ st_destroy_drawpix(struct st_context *st) { GLuint i; - for (i = 0; i < Elements(st->drawpix.shaders); i++) { - if (st->drawpix.shaders[i]) - _mesa_reference_fragprog(st->ctx, &st->drawpix.shaders[i], NULL); + for (i = 0; i < ARRAY_SIZE(st->drawpix.zs_shaders); i++) { + if (st->drawpix.zs_shaders[i]) + cso_delete_fragment_shader(st->cso_context, + st->drawpix.zs_shaders[i]); } - st_reference_fragprog(st, &st->pixel_xfer.combined_prog, NULL); if (st->drawpix.vert_shaders[0]) cso_delete_vertex_shader(st->cso_context, st->drawpix.vert_shaders[0]); if (st->drawpix.vert_shaders[1]) cso_delete_vertex_shader(st->cso_context, st->drawpix.vert_shaders[1]); } - -#endif /* FEATURE_drawpix */