diff options
author | keithw <keithw@keithw-laptop.(none)> | 2007-09-27 16:20:31 +0100 |
---|---|---|
committer | keithw <keithw@keithw-laptop.(none)> | 2007-09-27 16:20:31 +0100 |
commit | e0271e16da6427436ed755f170e9b402ada5dd7c (patch) | |
tree | 5bdb1e37ad9e07fd45b241221a79f33b2c4f8511 | |
parent | 48385cd13ac9410e07856924a3cd113b78f198db (diff) |
remove dead code
-rw-r--r-- | src/mesa/state_tracker/st_cb_texture.c | 83 |
1 files changed, 0 insertions, 83 deletions
diff --git a/src/mesa/state_tracker/st_cb_texture.c b/src/mesa/state_tracker/st_cb_texture.c index 13c803c6a7..e7be3df722 100644 --- a/src/mesa/state_tracker/st_cb_texture.c +++ b/src/mesa/state_tracker/st_cb_texture.c @@ -487,8 +487,6 @@ try_pbo_upload(GLcontext *ctx, dst_stride = stImage->mt->pitch; - intelFlush(&intel->ctx); - LOCK_HARDWARE(intel); { struct _DriBufferObject *src_buffer = intel_bufferobj_buffer(intel, pbo, INTEL_READ); @@ -505,10 +503,7 @@ try_pbo_upload(GLcontext *ctx, dst_stride, dst_buffer, dst_offset, 0, 0, 0, 0, width, height, GL_COPY); - - intel_batchbuffer_flush(intel->batch); } - UNLOCK_HARDWARE(intel); return GL_TRUE; #endif @@ -556,10 +551,6 @@ st_TexImage(GLcontext * ctx, DBG("%s target %s level %d %dx%dx%d border %d\n", __FUNCTION__, _mesa_lookup_enum_by_nr(target), level, width, height, depth, border); -#if 0 - intelFlush(ctx); -#endif - stImage->face = target_to_face(target); stImage->level = level; @@ -937,10 +928,6 @@ st_TexSubimage(GLcontext * ctx, _mesa_lookup_enum_by_nr(target), level, xoffset, yoffset, width, height); -#if 0 - intelFlush(ctx); -#endif - pixels = _mesa_validate_pbo_teximage(ctx, dims, width, height, depth, format, type, pixels, packing, "glTexSubImage2D"); @@ -1209,12 +1196,6 @@ do_copy_texsubimage(GLcontext *ctx, * vs. user-created FBO */ -#if 00 /* XXX FIX flush/locking */ - intelFlush(ctx); - /* XXX still need the lock ? */ - LOCK_HARDWARE(intel); -#endif - #if 0 /* A bit of fiddling to get the blitter to work with -ve * pitches. But we get a nice inverted blit this way, so it's @@ -1230,7 +1211,6 @@ do_copy_texsubimage(GLcontext *ctx, dest_offset, x, y + height, dstx, dsty, width, height, GL_COPY); /* ? */ - intel_batchbuffer_flush(intel->batch); #else pipe->region_copy(pipe, @@ -1245,10 +1225,6 @@ do_copy_texsubimage(GLcontext *ctx, /* size */ width, height); #endif - -#if 0 - UNLOCK_HARDWARE(intel); -#endif } else { fallback_copy_texsubimage(ctx, target, level, @@ -1583,70 +1559,11 @@ st_finalize_mipmap_tree(GLcontext *ctx, } } - /** - if (need_flush) - intel_batchbuffer_flush(intel->batch); - **/ return GL_TRUE; } -#if 0 /* unused? */ -void -st_tex_map_images(struct pipe_context *pipe, - struct st_texture_object *stObj) -{ - GLuint nr_faces = (stObj->base.Target == GL_TEXTURE_CUBE_MAP) ? 6 : 1; - GLuint face, i; - - DBG("%s\n", __FUNCTION__); - - for (face = 0; face < nr_faces; face++) { - for (i = stObj->firstLevel; i <= stObj->lastLevel; i++) { - struct st_texture_image *stImage = - st_texture_image(stObj->base.Image[face][i]); - - if (stImage->mt) { - stImage->base.Data - = st_miptree_image_map(pipe, - stImage->mt, - stImage->face, - stImage->level, - &stImage->base.RowStride, - stImage->base.ImageOffsets); - /* convert stride to texels, not bytes */ - stImage->base.RowStride /= stImage->mt->cpp; -/* stImage->base.ImageStride /= stImage->mt->cpp; */ - } - } - } -} - - - -void -st_tex_unmap_images(struct pipe_context *pipe, - struct st_texture_object *stObj) -{ - GLuint nr_faces = (stObj->base.Target == GL_TEXTURE_CUBE_MAP) ? 6 : 1; - GLuint face, i; - - for (face = 0; face < nr_faces; face++) { - for (i = stObj->firstLevel; i <= stObj->lastLevel; i++) { - struct st_texture_image *stImage = - st_texture_image(stObj->base.Image[face][i]); - - if (stImage->mt) { - st_miptree_image_unmap(pipe, stImage->mt); - stImage->base.Data = NULL; - } - } - } -} -#endif - - void |