VERSION | 2 src/compiler/glsl/linker.cpp | 7 src/egl/drivers/dri2/platform_x11.c | 3 src/egl/main/eglconfig.c | 1 src/egl/main/egldefines.h | 7 src/egl/main/eglsurface.c | 6 src/gallium/auxiliary/draw/draw_pipe_stipple.c | 30 +-- src/gallium/auxiliary/tgsi/tgsi_text.c | 4 src/gallium/drivers/llvmpipe/lp_rast.h | 6 src/gallium/drivers/llvmpipe/lp_setup_tri.c | 2 src/gallium/drivers/nouveau/codegen/nv50_ir_emit_nv50.cpp | 4 src/gallium/drivers/nouveau/codegen/nv50_ir_from_tgsi.cpp | 12 - src/gallium/drivers/nouveau/codegen/nv50_ir_lowering_gm107.cpp | 18 +- src/gallium/drivers/nouveau/codegen/nv50_ir_lowering_nv50.cpp | 35 +++ src/gallium/drivers/nouveau/codegen/nv50_ir_lowering_nvc0.cpp | 57 +++++- src/gallium/drivers/nouveau/nv50/nv50_shader_state.c | 2 src/gallium/drivers/nouveau/nv50/nv50_state.c | 4 src/gallium/drivers/nouveau/nvc0/nvc0_shader_state.c | 1 src/gallium/drivers/nouveau/nvc0/nvc0_state.c | 6 src/gallium/drivers/nouveau/nvc0/nvc0_surface.c | 19 +- src/gallium/drivers/r600/r600_state_common.c | 24 +- src/gallium/drivers/r600/sb/sb_expr.cpp | 8 src/gallium/drivers/radeon/r600_buffer_common.c | 6 src/gallium/drivers/radeon/r600_texture.c | 10 - src/gallium/drivers/radeon/radeon_video.c | 3 src/gallium/drivers/radeonsi/si_descriptors.c | 1 src/gallium/drivers/softpipe/sp_tex_sample.c | 9 - src/gallium/state_trackers/clover/core/kernel.cpp | 2 src/gallium/state_trackers/omx/vid_dec.c | 21 +- src/gallium/state_trackers/omx/vid_dec.h | 7 src/gallium/state_trackers/omx/vid_dec_h264.c | 26 ++ src/gallium/state_trackers/omx/vid_dec_mpeg12.c | 6 src/gallium/state_trackers/omx/vid_enc.c | 44 ++-- src/gallium/winsys/svga/drm/vmw_screen_ioctl.c | 3 src/gallium/winsys/svga/drm/vmw_surface.c | 2 src/mesa/drivers/dri/i965/brw_fs.cpp | 20 ++ src/mesa/drivers/dri/i965/brw_fs.h | 1 src/mesa/drivers/dri/i965/brw_fs_visitor.cpp | 12 + src/mesa/drivers/dri/i965/brw_pipe_control.c | 2 src/mesa/drivers/dri/i965/brw_reg.h | 16 + src/mesa/drivers/dri/i965/brw_vec4.cpp | 1 src/mesa/drivers/dri/i965/brw_vec4_nir.cpp | 4 src/mesa/drivers/dri/i965/brw_vec4_tcs.cpp | 16 + src/mesa/drivers/dri/i965/brw_vec4_tes.cpp | 90 ++++++---- src/mesa/drivers/dri/i965/intel_copy_image.c | 6 src/mesa/drivers/dri/i965/intel_fbo.c | 23 ++ src/mesa/drivers/dri/i965/intel_tex_image.c | 2 src/mesa/state_tracker/st_cb_fbo.c | 22 +- src/mesa/state_tracker/st_cb_texture.c | 5 src/mesa/swrast/s_context.c | 5 50 files changed, 459 insertions(+), 164 deletions(-)
New commits: commit bcb7e8b084f334e101d3a95c8f1d78bc77d653c2 Author: Emil Velikov <[email protected]> Date: Wed Mar 30 00:10:07 2016 +0100 Update version to 11.2.0-rc4 Signed-off-by: Emil Velikov <[email protected]> diff --git a/VERSION b/VERSION index 1568ccd..0cdc786 100644 --- a/VERSION +++ b/VERSION @@ -1 +1 @@ -11.2.0-rc3 +11.2.0-rc4 commit 0efb7fb803f49b358a7976eba2d8b9f4956ab052 Author: Kenneth Graunke <[email protected]> Date: Fri Mar 25 15:33:35 2016 -0700 i965: Fix brw_render_cache_set_check_flush's PIPE_CONTROLs. Our driver uses the brw_render_cache mechanism to track buffers we've rendered to and are about to sample from. Previously, we did a single PIPE_CONTROL with the following bits set: - Render Target Flush - Depth Cache Flush - Texture Cache Invalidate - VF Cache Invalidate - Instruction Cache Invalidate - CS Stall This combined both "top of pipe" invalidations and "bottom of pipe" flushes, which isn't how the hardware is intended to be programmed. The "top of pipe" invalidations may happen right away, without any guarantees that rendering using those caches has completed. That rendering may continue altering the caches. The "bottom of pipe" flushes do wait for the rendering to complete. The CS stall also prevents further work from happening until data is flushed out. What we wanted to do was wait for rendering complete, flush the new data out of the render and depth caches, wait, then invalidate any stale data in read-only caches. We can accomplish this by doing the "bottom of pipe" flushes with a CS stall, then the "top of pipe" flushes as a second PIPE_CONTROL. The flushes will wait until the rendering is complete, and the CS stall will prevent the second PIPE_CONTROL with the invalidations from executing until the first is done. Fixes dEQP-GLES3.functional.texture.specification.teximage2d_pbo subtests on Braswell and Skylake. These tests hit the meta PBO texture upload path, which binds the PBO as a texture and samples from it, while rendering to the destination texture. The tests then sample from the texture. For now, we leave Gen4-5 alone. It probably needs work too, but apparently it hasn't even been setting the (G45+) TC invalidation bit at all... v2: Add Sandybridge post-sync non-zero workaround, for safety. Cc: [email protected] Suggested-by: Francisco Jerez <[email protected]> Signed-off-by: Kenneth Graunke <[email protected]> Reviewed-by: Jason Ekstrand <[email protected]> Reviewed-by: Francisco Jerez <[email protected]> (cherry picked from commit 72473658c51d5e074ce219c1e6385a4cce29f467) diff --git a/src/mesa/drivers/dri/i965/brw_pipe_control.c b/src/mesa/drivers/dri/i965/brw_pipe_control.c index b41e28e..4672efd 100644 --- a/src/mesa/drivers/dri/i965/brw_pipe_control.c +++ b/src/mesa/drivers/dri/i965/brw_pipe_control.c @@ -338,8 +338,6 @@ brw_emit_mi_flush(struct brw_context *brw) } brw_emit_pipe_control_flush(brw, flags); } - - brw_render_cache_set_clear(brw); } int diff --git a/src/mesa/drivers/dri/i965/intel_fbo.c b/src/mesa/drivers/dri/i965/intel_fbo.c index 3a4a53a..91e8879 100644 --- a/src/mesa/drivers/dri/i965/intel_fbo.c +++ b/src/mesa/drivers/dri/i965/intel_fbo.c @@ -1065,7 +1065,28 @@ brw_render_cache_set_check_flush(struct brw_context *brw, drm_intel_bo *bo) if (!_mesa_set_search(brw->render_cache, bo)) return; - brw_emit_mi_flush(brw); + if (brw->gen >= 6) { + if (brw->gen == 6) { + /* [Dev-SNB{W/A}]: Before a PIPE_CONTROL with Write Cache + * Flush Enable = 1, a PIPE_CONTROL with any non-zero + * post-sync-op is required. + */ + brw_emit_post_sync_nonzero_flush(brw); + } + + brw_emit_pipe_control_flush(brw, + PIPE_CONTROL_DEPTH_CACHE_FLUSH | + PIPE_CONTROL_RENDER_TARGET_FLUSH | + PIPE_CONTROL_CS_STALL); + + brw_emit_pipe_control_flush(brw, + PIPE_CONTROL_TEXTURE_CACHE_INVALIDATE | + PIPE_CONTROL_CONST_CACHE_INVALIDATE); + } else { + brw_emit_mi_flush(brw); + } + + brw_render_cache_set_clear(brw); } /** commit 11654dda562188ba3cc743bcc286e7e48dde43e0 Author: Ilia Mirkin <[email protected]> Date: Sat Mar 26 22:32:43 2016 -0400 nvc0: disable primitive restart and index bias during blits Back in the dawn of time, we used to do immediate uploads for the vertex data, and all was well. However Maxwell dropped support for immediate vertex data, so we started feeding in a VBO (in all cases). But we forgot to disable some things that apply in such cases, specifically primitive restart and index bias. The latter was causing WoW and other Blizzard games trouble as they use a pattern where they draw with a base vertex (aka index bias), followed by texture uploads (aka blits, internally). Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=91526 Cc: "11.1 11.2" <[email protected]> Signed-off-by: Ilia Mirkin <[email protected]> Tested-by: Karol Herbst <[email protected]> (cherry picked from commit 41100b6b44e747b9003937f123fce571fd3dec46) diff --git a/src/gallium/drivers/nouveau/nvc0/nvc0_surface.c b/src/gallium/drivers/nouveau/nvc0/nvc0_surface.c index f4234b3..b33b6c3 100644 --- a/src/gallium/drivers/nouveau/nvc0/nvc0_surface.c +++ b/src/gallium/drivers/nouveau/nvc0/nvc0_surface.c @@ -1302,6 +1302,17 @@ nvc0_blit_3d(struct nvc0_context *nvc0, const struct pipe_blit_info *info) } nvc0->state.num_vtxelts = 2; + if (nvc0->state.prim_restart) { + IMMED_NVC0(push, NVC0_3D(PRIM_RESTART_ENABLE), 0); + nvc0->state.prim_restart = 0; + } + + if (nvc0->state.index_bias) { + IMMED_NVC0(push, NVC0_3D(VB_ELEMENT_BASE), 0); + IMMED_NVC0(push, NVC0_3D(VERTEX_ID_BASE), 0); + nvc0->state.index_bias = 0; + } + for (i = 0; i < info->dst.box.depth; ++i, z += dz) { if (info->dst.box.z + i) { BEGIN_NVC0(push, NVC0_3D(LAYER), 1); commit a715e4f40f4b2359cdd9712601a069e6e0f95607 Author: Ilia Mirkin <[email protected]> Date: Sun Mar 20 17:26:13 2016 -0400 nvc0/ir: fix picking of coordinates from tex instruction for textureGrad On Fermi, there's an argument in front of the coords that combines array and indirect handle, while on Kepler the array and the indirect handle are separate (and in front of the coords). We were previously only accounting for the array bit of it, if there were an indirect access it wouldn't be counted in the formula. Signed-off-by: Ilia Mirkin <[email protected]> Cc: "11.1 11.2" <[email protected]> (cherry picked from commit f667d15561820ee9dd8e836d43cce3ee52a4780e) diff --git a/src/gallium/drivers/nouveau/codegen/nv50_ir_lowering_nvc0.cpp b/src/gallium/drivers/nouveau/codegen/nv50_ir_lowering_nvc0.cpp index a9efe72..50454cb 100644 --- a/src/gallium/drivers/nouveau/codegen/nv50_ir_lowering_nvc0.cpp +++ b/src/gallium/drivers/nouveau/codegen/nv50_ir_lowering_nvc0.cpp @@ -874,7 +874,17 @@ NVC0LoweringPass::handleManualTXD(TexInstruction *i) Value *zero = bld.loadImm(bld.getSSA(), 0); int l, c; const int dim = i->tex.target.getDim() + i->tex.target.isCube(); - const int array = i->tex.target.isArray(); + + // This function is invoked after handleTEX lowering, so we have to expect + // the arguments in the order that the hw wants them. For Fermi, array and + // indirect are both in the leading arg, while for Kepler, array and + // indirect are separate (and both precede the coordinates). Maxwell is + // handled in a separate function. + unsigned array; + if (targ->getChipset() < NVISA_GK104_CHIPSET) + array = i->tex.target.isArray() || i->tex.rIndirectSrc >= 0; + else + array = i->tex.target.isArray() + (i->tex.rIndirectSrc >= 0); i->op = OP_TEX; // no need to clone dPdx/dPdy later commit afd477472120b2c0646a59525dd717dac9dee80c Author: Anuj Phogat <[email protected]> Date: Fri Mar 11 15:24:36 2016 -0800 i965: Fix assert conditions for src/dst x/y offsets Cc: [email protected] Signed-off-by: Anuj Phogat <[email protected]> Reviewed-by: Kenneth Graunke <[email protected]> (cherry picked from commit 4ba47f7b2adf71ed100cd390a1a9fbd4434e896a) diff --git a/src/mesa/drivers/dri/i965/intel_copy_image.c b/src/mesa/drivers/dri/i965/intel_copy_image.c index 08b7623..ccb82b6 100644 --- a/src/mesa/drivers/dri/i965/intel_copy_image.c +++ b/src/mesa/drivers/dri/i965/intel_copy_image.c @@ -140,9 +140,9 @@ copy_image_with_memcpy(struct brw_context *brw, _mesa_get_format_block_size(src_mt->format, &src_bw, &src_bh); assert(src_width % src_bw == 0); - assert(src_height % src_bw == 0); + assert(src_height % src_bh == 0); assert(src_x % src_bw == 0); - assert(src_y % src_bw == 0); + assert(src_y % src_bh == 0); /* If we are on the same miptree, same level, and same slice, then * intel_miptree_map won't let us map it twice. We have to do things a @@ -153,7 +153,7 @@ copy_image_with_memcpy(struct brw_context *brw, if (same_slice) { assert(dst_x % src_bw == 0); - assert(dst_y % src_bw == 0); + assert(dst_y % src_bh == 0); map_x1 = MIN2(src_x, dst_x); map_y1 = MIN2(src_y, dst_y); commit 864eba84dd8437e7f13473e68d43a9c1d3f9575f Author: xavier <[email protected]> Date: Wed Mar 9 09:58:48 2016 +0100 r600/sb: Do not distribute neg in expr_handler::fold_assoc() when folding multiplications. Previously it was doing this transformation for a Trine 3 shader: MUL R6.x.12, R13.x.23, 0.5|3f000000 - MULADD R4.x.12, -R6.x.12, 2|40000000, 1|3f800000 + MULADD R4.x.12, -R13.x.23, -1|bf800000, 1|3f800000 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=94412 Signed-off-by: Xavier Bouchoux <[email protected]> Cc: "11.0 11.1 11.2" <[email protected]> Reviewed-by: Glenn Kennard <[email protected]> Signed-off-by: Dave Airlie <[email protected]> (cherry picked from commit fce0b55ccbc33d320b9734a53c2a9f7886450c73) diff --git a/src/gallium/drivers/r600/sb/sb_expr.cpp b/src/gallium/drivers/r600/sb/sb_expr.cpp index 556a05d..3dd3a48 100644 --- a/src/gallium/drivers/r600/sb/sb_expr.cpp +++ b/src/gallium/drivers/r600/sb/sb_expr.cpp @@ -598,9 +598,13 @@ bool expr_handler::fold_assoc(alu_node *n) { unsigned op = n->bc.op; bool allow_neg = false, cur_neg = false; + bool distribute_neg = false; switch(op) { case ALU_OP2_ADD: + distribute_neg = true; + allow_neg = true; + break; case ALU_OP2_MUL: case ALU_OP2_MUL_IEEE: allow_neg = true; @@ -632,7 +636,7 @@ bool expr_handler::fold_assoc(alu_node *n) { if (v1->is_const()) { literal arg = v1->get_const_value(); apply_alu_src_mod(a->bc, 1, arg); - if (cur_neg) + if (cur_neg && distribute_neg) arg.f = -arg.f; if (a == n) @@ -660,7 +664,7 @@ bool expr_handler::fold_assoc(alu_node *n) { if (v0->is_const()) { literal arg = v0->get_const_value(); apply_alu_src_mod(a->bc, 0, arg); - if (cur_neg) + if (cur_neg && distribute_neg) arg.f = -arg.f; if (last_arg == 0) { commit 98380e02d526419d21dd7b48de3e3fcf7256c8db Author: Samuel Pitoiset <[email protected]> Date: Mon Mar 21 13:15:44 2016 +0100 nvc0: make sure to delete samplers used by compute shaders Signed-off-by: Samuel Pitoiset <[email protected]> Reviewed-by: Ilia Mirkin <[email protected]> Cc: "11.1 11.2" <[email protected]> (cherry picked from commit 9efd8b590f716bb7766ae6816bc080e7ef60a010) diff --git a/src/gallium/drivers/nouveau/nvc0/nvc0_state.c b/src/gallium/drivers/nouveau/nvc0/nvc0_state.c index cd33b4b..68116e4 100644 --- a/src/gallium/drivers/nouveau/nvc0/nvc0_state.c +++ b/src/gallium/drivers/nouveau/nvc0/nvc0_state.c @@ -413,7 +413,7 @@ nvc0_sampler_state_delete(struct pipe_context *pipe, void *hwcso) { unsigned s, i; - for (s = 0; s < 5; ++s) + for (s = 0; s < 6; ++s) for (i = 0; i < nvc0_context(pipe)->num_samplers[s]; ++i) if (nvc0_context(pipe)->samplers[s][i] == hwcso) nvc0_context(pipe)->samplers[s][i] = NULL; commit 7ce73f2608623c24cd36885df9ce92aa89706680 Author: Nicolai Hähnle <[email protected]> Date: Fri Mar 18 17:16:39 2016 -0500 st/mesa: use the texture view's format for render-to-texture Aside from the bug below, it fixes a simplistic test I've written locally, and I see no regression in Piglit for radeonsi. Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=94595 Cc: "11.0 11.1 11.2" <[email protected]> Reviewed-by: Ilia Mirkin <[email protected]> Reviewed-by: Marek Olšák <[email protected]> (cherry picked from commit a8b315b8271e867db30650dedb52e53d8dd9667c) diff --git a/src/mesa/state_tracker/st_cb_fbo.c b/src/mesa/state_tracker/st_cb_fbo.c index 82ab914..ff570e0 100644 --- a/src/mesa/state_tracker/st_cb_fbo.c +++ b/src/mesa/state_tracker/st_cb_fbo.c @@ -387,6 +387,7 @@ st_update_renderbuffer_surface(struct st_context *st, { struct pipe_context *pipe = st->pipe; struct pipe_resource *resource = strb->texture; + struct st_texture_object *stTexObj = NULL; unsigned rtt_width = strb->Base.Width; unsigned rtt_height = strb->Base.Height; unsigned rtt_depth = strb->Base.Depth; @@ -398,9 +399,18 @@ st_update_renderbuffer_surface(struct st_context *st, */ boolean enable_srgb = (st->ctx->Color.sRGBEnabled && _mesa_get_format_color_encoding(strb->Base.Format) == GL_SRGB); - enum pipe_format format = (enable_srgb) ? - util_format_srgb(resource->format) : - util_format_linear(resource->format); + enum pipe_format format = resource->format; + + if (strb->is_rtt) { + stTexObj = st_texture_object(strb->Base.TexImage->TexObject); + if (stTexObj->surface_based) + format = stTexObj->surface_format; + } + + format = (enable_srgb) ? + util_format_srgb(format) : + util_format_linear(format); + unsigned first_layer, last_layer, level; if (resource->target == PIPE_TEXTURE_1D_ARRAY) { @@ -431,8 +441,8 @@ st_update_renderbuffer_surface(struct st_context *st, /* Adjust for texture views */ if (strb->is_rtt && resource->array_size > 1 && - strb->Base.TexImage->TexObject->Immutable) { - struct gl_texture_object *tex = strb->Base.TexImage->TexObject; + stTexObj->base.Immutable) { + struct gl_texture_object *tex = &stTexObj->base; first_layer += tex->MinLayer; if (!strb->rtt_layered) last_layer += tex->MinLayer; @@ -492,8 +502,6 @@ st_render_texture(struct gl_context *ctx, st_update_renderbuffer_surface(st, strb); - strb->Base.Format = st_pipe_format_to_mesa_format(pt->format); - /* Invalidate buffer state so that the pipe's framebuffer state * gets updated. * That's where the new renderbuffer (which we just created) gets commit 3b6bda665a5a890f2c98e19d2939d7de92b8cb4c Author: Nishanth Peethambaran <[email protected]> Date: Fri Mar 11 01:23:00 2016 -0500 st/omx/dec: Correct the timestamping Attach the timestamp to the dpb buffer and use that timestamp while pushing buffer from dpb list to the omx client. Reviewed-by: Christian König <[email protected]> Signed-off-by: Nishanth Peethambaran <[email protected]> Cc: "11.1 11.2" <[email protected]> (cherry picked from commit eeb117a09d6c0eb2b4fa94d55e8015c8aa982727) diff --git a/src/gallium/state_trackers/omx/vid_dec.c b/src/gallium/state_trackers/omx/vid_dec.c index 9fcf20e..108a460 100644 --- a/src/gallium/state_trackers/omx/vid_dec.c +++ b/src/gallium/state_trackers/omx/vid_dec.c @@ -419,6 +419,7 @@ static OMX_ERRORTYPE vid_dec_DecodeBuffer(omx_base_PortType *port, OMX_BUFFERHEA priv->in_buffers[i] = buf; priv->sizes[i] = buf->nFilledLen; priv->inputs[i] = buf->pBuffer; + priv->timestamps[i] = buf->nTimeStamp; while (priv->num_in_buffers > (!!(buf->nFlags & OMX_BUFFERFLAG_EOS) ? 0 : 1)) { bool eos = !!(priv->in_buffers[0]->nFlags & OMX_BUFFERFLAG_EOS); @@ -469,6 +470,7 @@ static OMX_ERRORTYPE vid_dec_DecodeBuffer(omx_base_PortType *port, OMX_BUFFERHEA priv->in_buffers[0] = priv->in_buffers[1]; priv->sizes[0] = priv->sizes[1] - delta; priv->inputs[0] = priv->inputs[1] + delta; + priv->timestamps[0] = priv->timestamps[1]; } if (r) @@ -526,9 +528,13 @@ static void vid_dec_FrameDecoded(OMX_COMPONENTTYPE *comp, OMX_BUFFERHEADERTYPE* { vid_dec_PrivateType *priv = comp->pComponentPrivate; bool eos = !!(input->nFlags & OMX_BUFFERFLAG_EOS); + OMX_TICKS timestamp; - if (!input->pInputPortPrivate) - input->pInputPortPrivate = priv->Flush(priv); + if (!input->pInputPortPrivate) { + input->pInputPortPrivate = priv->Flush(priv, ×tamp); + if (timestamp != OMX_VID_DEC_TIMESTAMP_INVALID) + input->nTimeStamp = timestamp; + } if (input->pInputPortPrivate) { if (output->pInputPortPrivate) { @@ -539,6 +545,7 @@ static void vid_dec_FrameDecoded(OMX_COMPONENTTYPE *comp, OMX_BUFFERHEADERTYPE* vid_dec_FillOutput(priv, input->pInputPortPrivate, output); } output->nFilledLen = output->nAllocLen; + output->nTimeStamp = input->nTimeStamp; } if (eos && input->pInputPortPrivate) diff --git a/src/gallium/state_trackers/omx/vid_dec.h b/src/gallium/state_trackers/omx/vid_dec.h index 3b39826..649d745 100644 --- a/src/gallium/state_trackers/omx/vid_dec.h +++ b/src/gallium/state_trackers/omx/vid_dec.h @@ -59,6 +59,8 @@ #define OMX_VID_DEC_AVC_NAME "OMX.mesa.video_decoder.avc" #define OMX_VID_DEC_AVC_ROLE "video_decoder.avc" +#define OMX_VID_DEC_TIMESTAMP_INVALID ((OMX_TICKS) -1) + struct vl_vlc; DERIVEDCLASS(vid_dec_PrivateType, omx_base_filter_PrivateType) @@ -69,7 +71,7 @@ DERIVEDCLASS(vid_dec_PrivateType, omx_base_filter_PrivateType) struct pipe_video_codec *codec; \ void (*Decode)(vid_dec_PrivateType *priv, struct vl_vlc *vlc, unsigned min_bits_left); \ void (*EndFrame)(vid_dec_PrivateType *priv); \ - struct pipe_video_buffer *(*Flush)(vid_dec_PrivateType *priv); \ + struct pipe_video_buffer *(*Flush)(vid_dec_PrivateType *priv, OMX_TICKS *timestamp); \ struct pipe_video_buffer *target, *shadow; \ union { \ struct { \ @@ -100,6 +102,9 @@ DERIVEDCLASS(vid_dec_PrivateType, omx_base_filter_PrivateType) OMX_BUFFERHEADERTYPE *in_buffers[2]; \ const void *inputs[2]; \ unsigned sizes[2]; \ + OMX_TICKS timestamps[2]; \ + OMX_TICKS timestamp; \ + bool first_buf_in_frame; \ bool frame_finished; \ bool frame_started; \ unsigned bytes_left; \ diff --git a/src/gallium/state_trackers/omx/vid_dec_h264.c b/src/gallium/state_trackers/omx/vid_dec_h264.c index 75f27d2..9aab6d1 100644 --- a/src/gallium/state_trackers/omx/vid_dec_h264.c +++ b/src/gallium/state_trackers/omx/vid_dec_h264.c @@ -45,6 +45,7 @@ struct dpb_list { struct list_head list; struct pipe_video_buffer *buffer; + OMX_TICKS timestamp; unsigned poc; }; @@ -82,7 +83,7 @@ static const uint8_t Default_8x8_Inter[64] = { static void vid_dec_h264_Decode(vid_dec_PrivateType *priv, struct vl_vlc *vlc, unsigned min_bits_left); static void vid_dec_h264_EndFrame(vid_dec_PrivateType *priv); -static struct pipe_video_buffer *vid_dec_h264_Flush(vid_dec_PrivateType *priv); +static struct pipe_video_buffer *vid_dec_h264_Flush(vid_dec_PrivateType *priv, OMX_TICKS *timestamp); void vid_dec_h264_Init(vid_dec_PrivateType *priv) { @@ -94,6 +95,7 @@ void vid_dec_h264_Init(vid_dec_PrivateType *priv) LIST_INITHEAD(&priv->codec_data.h264.dpb_list); priv->picture.h264.field_order_cnt[0] = priv->picture.h264.field_order_cnt[1] = INT_MAX; + priv->first_buf_in_frame = true; } static void vid_dec_h264_BeginFrame(vid_dec_PrivateType *priv) @@ -104,6 +106,9 @@ static void vid_dec_h264_BeginFrame(vid_dec_PrivateType *priv) return; vid_dec_NeedTarget(priv); + if (priv->first_buf_in_frame) + priv->timestamp = priv->timestamps[0]; + priv->first_buf_in_frame = false; priv->picture.h264.num_ref_frames = priv->picture.h264.pps->sps->max_num_ref_frames; @@ -127,7 +132,8 @@ static void vid_dec_h264_BeginFrame(vid_dec_PrivateType *priv) priv->frame_started = true; } -static struct pipe_video_buffer *vid_dec_h264_Flush(vid_dec_PrivateType *priv) +static struct pipe_video_buffer *vid_dec_h264_Flush(vid_dec_PrivateType *priv, + OMX_TICKS *timestamp) { struct dpb_list *entry, *result = NULL; struct pipe_video_buffer *buf; @@ -146,6 +152,8 @@ static struct pipe_video_buffer *vid_dec_h264_Flush(vid_dec_PrivateType *priv) return NULL; buf = result->buffer; + if (timestamp) + *timestamp = result->timestamp; --priv->codec_data.h264.dpb_num; LIST_DEL(&result->list); @@ -159,6 +167,7 @@ static void vid_dec_h264_EndFrame(vid_dec_PrivateType *priv) struct dpb_list *entry; struct pipe_video_buffer *tmp; bool top_field_first; + OMX_TICKS timestamp; if (!priv->frame_started) return; @@ -181,7 +190,9 @@ static void vid_dec_h264_EndFrame(vid_dec_PrivateType *priv) if (!entry) return; + priv->first_buf_in_frame = true; entry->buffer = priv->target; + entry->timestamp = priv->timestamp; entry->poc = MIN2(priv->picture.h264.field_order_cnt[0], priv->picture.h264.field_order_cnt[1]); LIST_ADDTAIL(&entry->list, &priv->codec_data.h264.dpb_list); ++priv->codec_data.h264.dpb_num; @@ -192,7 +203,8 @@ static void vid_dec_h264_EndFrame(vid_dec_PrivateType *priv) return; tmp = priv->in_buffers[0]->pInputPortPrivate; - priv->in_buffers[0]->pInputPortPrivate = vid_dec_h264_Flush(priv); + priv->in_buffers[0]->pInputPortPrivate = vid_dec_h264_Flush(priv, ×tamp); + priv->in_buffers[0]->nTimeStamp = timestamp; priv->target = tmp; priv->frame_finished = priv->in_buffers[0]->pInputPortPrivate != NULL; } diff --git a/src/gallium/state_trackers/omx/vid_dec_mpeg12.c b/src/gallium/state_trackers/omx/vid_dec_mpeg12.c index bef83ec..7b2df8f4 100644 --- a/src/gallium/state_trackers/omx/vid_dec_mpeg12.c +++ b/src/gallium/state_trackers/omx/vid_dec_mpeg12.c @@ -61,7 +61,7 @@ static uint8_t default_non_intra_matrix[64] = { static void vid_dec_mpeg12_Decode(vid_dec_PrivateType *priv, struct vl_vlc *vlc, unsigned min_bits_left); static void vid_dec_mpeg12_EndFrame(vid_dec_PrivateType *priv); -static struct pipe_video_buffer *vid_dec_mpeg12_Flush(vid_dec_PrivateType *priv); +static struct pipe_video_buffer *vid_dec_mpeg12_Flush(vid_dec_PrivateType *priv, OMX_TICKS *timestamp); void vid_dec_mpeg12_Init(vid_dec_PrivateType *priv) { @@ -131,10 +131,12 @@ static void vid_dec_mpeg12_EndFrame(vid_dec_PrivateType *priv) priv->in_buffers[0]->pInputPortPrivate = done; } -static struct pipe_video_buffer *vid_dec_mpeg12_Flush(vid_dec_PrivateType *priv) +static struct pipe_video_buffer *vid_dec_mpeg12_Flush(vid_dec_PrivateType *priv, OMX_TICKS *timestamp) { struct pipe_video_buffer *result = priv->picture.mpeg12.ref[1]; priv->picture.mpeg12.ref[1] = NULL; + if (timestamp) + *timestamp = OMX_VID_DEC_TIMESTAMP_INVALID; return result; } commit c1763ec83cb502e4124365efa09d7f5272456891 Author: Nishanth Peethambaran <[email protected]> Date: Tue Mar 15 01:56:18 2016 -0400 st/omx: Remove trailing spaces Reviewed-by: Christian König <[email protected]> Signed-off-by: Nishanth Peethambaran <[email protected]> Cc: "11.1 11.2" <[email protected]> (cherry picked from commit 46de6bbb775602ab237d0054e5351b0fc90d942b) diff --git a/src/gallium/state_trackers/omx/vid_dec.c b/src/gallium/state_trackers/omx/vid_dec.c index 5584348..9fcf20e 100644 --- a/src/gallium/state_trackers/omx/vid_dec.c +++ b/src/gallium/state_trackers/omx/vid_dec.c @@ -140,7 +140,7 @@ static OMX_ERRORTYPE vid_dec_Constructor(OMX_COMPONENTTYPE *comp, OMX_STRING nam r = omx_base_filter_Constructor(comp, name); if (r) - return r; + return r; priv->profile = PIPE_VIDEO_PROFILE_UNKNOWN; @@ -268,7 +268,7 @@ static OMX_ERRORTYPE vid_dec_SetParameter(OMX_HANDLETYPE handle, OMX_INDEXTYPE i r = checkHeader(param, sizeof(OMX_PARAM_COMPONENTROLETYPE)); if (r) return r; - + if (!strcmp((char *)role->cRole, OMX_VID_DEC_MPEG2_ROLE)) { priv->profile = PIPE_VIDEO_PROFILE_MPEG2_MAIN; } else if (!strcmp((char *)role->cRole, OMX_VID_DEC_AVC_ROLE)) { @@ -321,7 +321,7 @@ static OMX_ERRORTYPE vid_dec_GetParameter(OMX_HANDLETYPE handle, OMX_INDEXTYPE i strcpy((char *)role->cRole, OMX_VID_DEC_MPEG2_ROLE); else if (priv->profile == PIPE_VIDEO_PROFILE_MPEG4_AVC_HIGH) strcpy((char *)role->cRole, OMX_VID_DEC_AVC_ROLE); - + break; } @@ -474,7 +474,7 @@ static OMX_ERRORTYPE vid_dec_DecodeBuffer(omx_base_PortType *port, OMX_BUFFERHEA if (r) return r; } - + return OMX_ErrorNone; } @@ -513,7 +513,7 @@ static void vid_dec_FillOutput(vid_dec_PrivateType *priv, struct pipe_video_buff box.width = def->nFrameWidth / 2; box.height = def->nFrameHeight / 2; - + src = priv->pipe->transfer_map(priv->pipe, views[1]->texture, 0, PIPE_TRANSFER_READ, &box, &transfer); util_copy_rect(dst, views[1]->texture->format, def->nStride, 0, 0, diff --git a/src/gallium/state_trackers/omx/vid_dec_h264.c b/src/gallium/state_trackers/omx/vid_dec_h264.c index b453682..75f27d2 100644 --- a/src/gallium/state_trackers/omx/vid_dec_h264.c +++ b/src/gallium/state_trackers/omx/vid_dec_h264.c @@ -91,7 +91,7 @@ void vid_dec_h264_Init(vid_dec_PrivateType *priv) priv->Decode = vid_dec_h264_Decode; priv->EndFrame = vid_dec_h264_EndFrame; priv->Flush = vid_dec_h264_Flush; - + LIST_INITHEAD(&priv->codec_data.h264.dpb_list); priv->picture.h264.field_order_cnt[0] = priv->picture.h264.field_order_cnt[1] = INT_MAX; } @@ -829,7 +829,7 @@ static void slice_header(vid_dec_PrivateType *priv, struct vl_rbsp *rbsp, priv->picture.h264.field_order_cnt[0] = expectedPicOrderCnt + priv->codec_data.h264.delta_pic_order_cnt[0]; priv->picture.h264.field_order_cnt[1] = priv->picture.h264.field_order_cnt[0] + sps->offset_for_top_to_bottom_field + priv->codec_data.h264.delta_pic_order_cnt[1]; - + } else if (!priv->picture.h264.bottom_field_flag) priv->picture.h264.field_order_cnt[0] = expectedPicOrderCnt + priv->codec_data.h264.delta_pic_order_cnt[0]; else @@ -859,7 +859,7 @@ static void slice_header(vid_dec_PrivateType *priv, struct vl_rbsp *rbsp, if (!priv->picture.h264.field_pic_flag) { priv->picture.h264.field_order_cnt[0] = tempPicOrderCnt; priv->picture.h264.field_order_cnt[1] = tempPicOrderCnt; - + } else if (!priv->picture.h264.bottom_field_flag) priv->picture.h264.field_order_cnt[0] = tempPicOrderCnt; else @@ -876,7 +876,7 @@ static void slice_header(vid_dec_PrivateType *priv, struct vl_rbsp *rbsp, priv->picture.h264.num_ref_idx_l0_active_minus1 = pps->num_ref_idx_l0_default_active_minus1; priv->picture.h264.num_ref_idx_l1_active_minus1 = pps->num_ref_idx_l1_default_active_minus1; - + if (slice_type == PIPE_H264_SLICE_TYPE_P || slice_type == PIPE_H264_SLICE_TYPE_SP || slice_type == PIPE_H264_SLICE_TYPE_B) { diff --git a/src/gallium/state_trackers/omx/vid_enc.c b/src/gallium/state_trackers/omx/vid_enc.c index df22a97..4505fe1 100644 --- a/src/gallium/state_trackers/omx/vid_enc.c +++ b/src/gallium/state_trackers/omx/vid_enc.c @@ -179,7 +179,7 @@ static OMX_ERRORTYPE vid_enc_Constructor(OMX_COMPONENTTYPE *comp, OMX_STRING nam if (!screen->get_video_param(screen, PIPE_VIDEO_PROFILE_MPEG4_AVC_HIGH, PIPE_VIDEO_ENTRYPOINT_ENCODE, PIPE_VIDEO_CAP_SUPPORTED)) return OMX_ErrorBadParameter; - + priv->stacked_frames_num = screen->get_video_param(screen, PIPE_VIDEO_PROFILE_MPEG4_AVC_HIGH, PIPE_VIDEO_ENTRYPOINT_ENCODE, @@ -242,7 +242,7 @@ static OMX_ERRORTYPE vid_enc_Constructor(OMX_COMPONENTTYPE *comp, OMX_STRING nam port->Port_AllocateBuffer = vid_enc_AllocateOutBuffer; port->Port_FreeBuffer = vid_enc_FreeOutBuffer; - + priv->bitrate.eControlRate = OMX_Video_ControlRateDisable; priv->bitrate.nTargetBitrate = 0; @@ -253,7 +253,7 @@ static OMX_ERRORTYPE vid_enc_Constructor(OMX_COMPONENTTYPE *comp, OMX_STRING nam priv->profile_level.eProfile = OMX_VIDEO_AVCProfileBaseline; priv->profile_level.eLevel = OMX_VIDEO_AVCLevel42; - priv->force_pic_type.IntraRefreshVOP = OMX_FALSE; + priv->force_pic_type.IntraRefreshVOP = OMX_FALSE; priv->frame_num = 0; priv->pic_order_cnt = 0; priv->restricted_b_frames = debug_get_bool_option("OMX_USE_RESTRICTED_B_FRAMES", FALSE); @@ -380,7 +380,7 @@ static OMX_ERRORTYPE vid_enc_SetParameter(OMX_HANDLETYPE handle, OMX_INDEXTYPE i port = (omx_base_video_PortType *)priv->ports[OMX_BASE_FILTER_OUTPUTPORT_INDEX]; port->sPortParam.nBufferSize = framesize * 512 / (16*16); - + priv->frame_rate = def->format.video.xFramerate; priv->callbacks->EventHandler(comp, priv->callbackData, OMX_EventPortSettingsChanged, @@ -532,10 +532,10 @@ static OMX_ERRORTYPE vid_enc_SetConfig(OMX_HANDLETYPE handle, OMX_INDEXTYPE idx, vid_enc_PrivateType *priv = comp->pComponentPrivate; OMX_ERRORTYPE r; int i; - + if (!config) return OMX_ErrorBadParameter; - + switch(idx) { case OMX_IndexConfigVideoIntraVOPRefresh: { OMX_CONFIG_INTRAREFRESHVOPTYPE *type = config; @@ -543,9 +543,9 @@ static OMX_ERRORTYPE vid_enc_SetConfig(OMX_HANDLETYPE handle, OMX_INDEXTYPE idx, r = checkHeader(config, sizeof(OMX_CONFIG_INTRAREFRESHVOPTYPE)); if (r) return r; - + priv->force_pic_type = *type; - + break; } case OMX_IndexConfigCommonScale: { @@ -568,11 +568,11 @@ static OMX_ERRORTYPE vid_enc_SetConfig(OMX_HANDLETYPE handle, OMX_INDEXTYPE idx, priv->scale = *scale; if (priv->scale.xWidth != 0xffffffff && priv->scale.xHeight != 0xffffffff) { struct pipe_video_buffer templat = {}; - + templat.buffer_format = PIPE_FORMAT_NV12; templat.chroma_format = PIPE_VIDEO_CHROMA_FORMAT_420; - templat.width = priv->scale.xWidth; - templat.height = priv->scale.xHeight; + templat.width = priv->scale.xWidth; + templat.height = priv->scale.xHeight; templat.interlaced = false; for (i = 0; i < OMX_VID_ENC_NUM_SCALING_BUFFERS; ++i) { priv->scale_buffer[i] = priv->s_pipe->create_video_buffer(priv->s_pipe, &templat); @@ -615,7 +615,7 @@ static OMX_ERRORTYPE vid_enc_GetConfig(OMX_HANDLETYPE handle, OMX_INDEXTYPE idx, default: return omx_base_component_GetConfig(handle, idx, config); } - + return OMX_ErrorNone; } @@ -1010,10 +1010,10 @@ static void enc_ControlPicture(omx_base_PortType *port, struct pipe_h264_enc_pic switch (priv->bitrate.eControlRate) { case OMX_Video_ControlRateVariable: rate_ctrl->rate_ctrl_method = PIPE_H264_ENC_RATE_CONTROL_METHOD_VARIABLE; - break; + break; case OMX_Video_ControlRateConstant: rate_ctrl->rate_ctrl_method = PIPE_H264_ENC_RATE_CONTROL_METHOD_CONSTANT; - break; + break; case OMX_Video_ControlRateVariableSkipFrames: rate_ctrl->rate_ctrl_method = PIPE_H264_ENC_RATE_CONTROL_METHOD_VARIABLE_SKIP; break; @@ -1023,8 +1023,8 @@ static void enc_ControlPicture(omx_base_PortType *port, struct pipe_h264_enc_pic default: rate_ctrl->rate_ctrl_method = PIPE_H264_ENC_RATE_CONTROL_METHOD_DISABLE; break; - } - + } + rate_ctrl->frame_rate_den = OMX_VID_ENC_CONTROL_FRAME_RATE_DEN_DEFAULT; rate_ctrl->frame_rate_num = ((priv->frame_rate) >> 16) * rate_ctrl->frame_rate_den; @@ -1035,7 +1035,7 @@ static void enc_ControlPicture(omx_base_PortType *port, struct pipe_h264_enc_pic rate_ctrl->target_bitrate = priv->bitrate.nTargetBitrate; else rate_ctrl->target_bitrate = OMX_VID_ENC_BITRATE_MAX; - rate_ctrl->peak_bitrate = rate_ctrl->target_bitrate; + rate_ctrl->peak_bitrate = rate_ctrl->target_bitrate; if (rate_ctrl->target_bitrate < OMX_VID_ENC_BITRATE_MEDIAN) rate_ctrl->vbv_buffer_size = MIN2((rate_ctrl->target_bitrate * 2.75), OMX_VID_ENC_BITRATE_MEDIAN); else @@ -1051,7 +1051,7 @@ static void enc_ControlPicture(omx_base_PortType *port, struct pipe_h264_enc_pic rate_ctrl->peak_bits_picture_integer = rate_ctrl->target_bits_picture; rate_ctrl->peak_bits_picture_fraction = 0; } - + picture->quant_i_frames = priv->quant.nQpI; picture->quant_p_frames = priv->quant.nQpP; picture->quant_b_frames = priv->quant.nQpB; @@ -1069,7 +1069,7 @@ static void enc_HandleTask(omx_base_PortType *port, struct encode_task *task, unsigned size = priv->ports[OMX_BASE_FILTER_OUTPUTPORT_INDEX]->sPortParam.nBufferSize; struct pipe_video_buffer *vbuf = task->buf; struct pipe_h264_enc_picture_desc picture = {}; - + /* -------------- scale input image --------- */ enc_ScaleInput(port, &vbuf, &size); priv->s_pipe->flush(priv->s_pipe, NULL, 0); @@ -1160,7 +1160,7 @@ static OMX_ERRORTYPE vid_enc_EncodeFrame(omx_base_PortType *port, OMX_BUFFERHEAD priv->force_pic_type.IntraRefreshVOP) { enc_ClearBframes(port, inp); picture_type = PIPE_H264_ENC_PICTURE_TYPE_IDR; - priv->force_pic_type.IntraRefreshVOP = OMX_FALSE; + priv->force_pic_type.IntraRefreshVOP = OMX_FALSE; priv->frame_num = 0; } else if (priv->codec->profile == PIPE_VIDEO_PROFILE_MPEG4_AVC_BASELINE || !(priv->pic_order_cnt % OMX_VID_ENC_P_PERIOD_DEFAULT) || @@ -1169,7 +1169,7 @@ static OMX_ERRORTYPE vid_enc_EncodeFrame(omx_base_PortType *port, OMX_BUFFERHEAD } else { picture_type = PIPE_H264_ENC_PICTURE_TYPE_B; } - + task->pic_order_cnt = priv->pic_order_cnt++; if (picture_type == PIPE_H264_ENC_PICTURE_TYPE_B) { @@ -1245,7 +1245,7 @@ static void vid_enc_BufferEncoded(OMX_COMPONENTTYPE *comp, OMX_BUFFERHEADERTYPE* output->pBuffer = priv->t_pipe->transfer_map(priv->t_pipe, outp->bitstream, 0, PIPE_TRANSFER_READ_WRITE, &box, &outp->transfer); - + /* ------------- get size of result ----------------- */ priv->codec->get_feedback(priv->codec, task->feedback, &size); commit b4045c43ae7781aa2004b72353b98d09008f2d62 Author: Ilia Mirkin <[email protected]> Date: Sun Mar 20 13:43:43 2016 -0400 nv50/ir: fix indirect texturing for non-array textures on nvc0 If a layer parameter is provided, we want to flip it to position 0 (and combine it with any indirect params). However if the target is not an array, there is no layer, so we have to shift all of the arguments down by one to make room for it. This fixes situations where there were non-coordinate parameters, such as bias, lod, depth compare, explicit derivatives. Instead of adding a new parameter at the front for the indirect reference, we would swap one of those in its place. Fixes dEQP-GLES31.functional.shaders.opaque_type_indexing.sampler.uniform.compute.*shadow Signed-off-by: Ilia Mirkin <[email protected]> Reported-by: Samuel Pitoiset <[email protected]> Tested-by: Samuel Pitoiset <[email protected]> Cc: "11.1 11.2" <[email protected]> (cherry picked from commit 7d98bfedd73d632041d27ff12ccf7c7be74a2ddd) diff --git a/src/gallium/drivers/nouveau/codegen/nv50_ir_lowering_nvc0.cpp b/src/gallium/drivers/nouveau/codegen/nv50_ir_lowering_nvc0.cpp index 289ba7c..a9efe72 100644 --- a/src/gallium/drivers/nouveau/codegen/nv50_ir_lowering_nvc0.cpp +++ b/src/gallium/drivers/nouveau/codegen/nv50_ir_lowering_nvc0.cpp @@ -746,9 +746,13 @@ NVC0LoweringPass::handleTEX(TexInstruction *i) } Value *arrayIndex = i->tex.target.isArray() ? i->getSrc(lyr) : NULL; - for (int s = dim; s >= 1; --s) - i->setSrc(s, i->getSrc(s - 1)); - i->setSrc(0, arrayIndex); + if (arrayIndex) { + for (int s = dim; s >= 1; --s) + i->setSrc(s, i->getSrc(s - 1)); + i->setSrc(0, arrayIndex); + } else { + i->moveSources(0, 1); + } if (arrayIndex) { int sat = (i->op == OP_TXF) ? 1 : 0; commit 9bcedb95535299f05e5a97bea8b16494c1f03825 Author: Ilia Mirkin <[email protected]> Date: Sat Mar 19 21:25:36 2016 -0400 st/mesa: only minify depth for 3d targets We make sure that that image depth matches the level's depth before copying it into place. However we should only be minifying the first level's depth for 3d textures - array textures have the same depth for all levels. This fixes tests such as dEQP-GLES3.functional.texture.specification.texsubimage3d_depth.* and I suspect account for a number of other odd situations I've run into where level > 0 of array textures was messed up. Signed-off-by: Ilia Mirkin <[email protected]> Reviewed-by: Dave Airlie <[email protected]> Cc: "11.1 11.2" <[email protected]> (cherry picked from commit adb40a739943d62508b9c79cbd85e3c67ee3b43b) diff --git a/src/mesa/state_tracker/st_cb_texture.c b/src/mesa/state_tracker/st_cb_texture.c index cfec627..32c9139 100644 --- a/src/mesa/state_tracker/st_cb_texture.c +++ b/src/mesa/state_tracker/st_cb_texture.c @@ -2886,10 +2886,13 @@ st_finalize_texture(struct gl_context *ctx, /* Need to import images in main memory or held in other textures. */ if (stImage && stObj->pt != stImage->pt) { + GLuint depth = stObj->depth0; + if (stObj->base.Target == GL_TEXTURE_3D) + depth = u_minify(depth, level); if (level == 0 || (stImage->base.Width == u_minify(stObj->width0, level) && stImage->base.Height == u_minify(stObj->height0, level) && - stImage->base.Depth == u_minify(stObj->depth0, level))) { + stImage->base.Depth == depth)) { /* src image fits expected dest mipmap level size */ copy_image_data_to_texture(st, stObj, level, stImage); }
