swr: fix transform feedback logic

The shader that is used to copy vertex data out of the vs/gs shaders to
the user-specified buffer (streamout or SO shader) was not using the
correct offsets.

Adjust the offsets that are used just for the SO shader:
- Make sure that position is handled in the same special way
  as in the vs/gs shaders
- Use the correct offset to be passed in the core
- consolidate register slot mapping logic into one function, since it's
  been calculated in 2 different places (one for calcuating the slot mask,
  and one for the register offsets themselves

Also make room for all attibutes in the backend vertex area.

Fixes:
- all vtk GL2PS tests
- 18 piglit tests (16 ext_transform_feedback tests,
  arb-quads-follow-provoking-vertex and primitive-type gl_points

v2:

- take care of more SGV slots in slot mapping logic
- trim feState.vsVertexSize
- fix GS interface and incorporate GS while calculating vsVertexSize

Note that vsVertexSize is used in the core as the one parameter that
controls vertex size between all stages, so it has to be adjusted appropriately
for the whole vs/gs/fs pipeline.

Also note that GS and SO is not fully implemented.  This will be addressed
later.

fixes:
- fixes total of 20 piglit tests

CC: 17.2 <mesa-stable@lists.freedesktop.org>

Reviewed-by: Bruce Cherniak <bruce.cherniak@intel.com>
(cherry picked from commit 194ff5eed1)

src/gallium/drivers/swr/swr_draw.cpp

@@ -81,8 +81,11 @@ swr_draw_vbo(struct pipe_context *pipe, const struct pipe_draw_info *info)
                offsets[output_buffer] = so->output[i].dst_offset;
             }
 
+            unsigned attrib_slot = so->output[i].register_index;
+            attrib_slot = swr_so_adjust_attrib(attrib_slot, ctx->vs);
+
             state.stream.decl[num].bufferIndex = output_buffer;
-            state.stream.decl[num].attribSlot = so->output[i].register_index - 1;
+            state.stream.decl[num].attribSlot = attrib_slot;
             state.stream.decl[num].componentMask =
                ((1 << so->output[i].num_components) - 1)
                << so->output[i].start_component;
@@ -129,10 +132,36 @@ swr_draw_vbo(struct pipe_context *pipe, const struct pipe_draw_info *info)
     * XXX setup provokingVertex & topologyProvokingVertex */
    SWR_FRONTEND_STATE feState = {0};
 
-   feState.vsVertexSize =
-      VERTEX_ATTRIB_START_SLOT +
-      + ctx->vs->info.base.num_outputs
-      - (ctx->vs->info.base.writes_position ? 1 : 0);
+   // feState.vsVertexSize seeds the PA size that is used as an interface
+   // between all the shader stages, so it has to be large enough to
+   // incorporate all interfaces between stages
+
+   // max of gs and vs num_outputs
+   feState.vsVertexSize = ctx->vs->info.base.num_outputs;
+   if (ctx->gs &&
+       ctx->gs->info.base.num_outputs > feState.vsVertexSize) {
+      feState.vsVertexSize = ctx->gs->info.base.num_outputs;
+   }
+
+   if (ctx->vs->info.base.num_outputs) {
+      // gs does not adjust for position in SGV slot at input from vs
+      if (!ctx->gs)
+         feState.vsVertexSize--;
+   }
+
+   // other (non-SGV) slots start at VERTEX_ATTRIB_START_SLOT
+   feState.vsVertexSize += VERTEX_ATTRIB_START_SLOT;
+
+   // The PA in the clipper does not handle BE vertex sizes
+   // different from FE. Increase vertexsize only for the cases that needed it
+
+   // primid needs a slot
+   if (ctx->fs->info.base.uses_primid)
+      feState.vsVertexSize++;
+   // sprite coord enable
+   if (ctx->rasterizer->sprite_coord_enable)
+      feState.vsVertexSize++;
+
 
    if (ctx->rasterizer->flatshade_first) {
       feState.provokingVertex = {1, 0, 0};

src/gallium/drivers/swr/swr_shader.cpp

@@ -414,7 +414,10 @@ BuilderSWR::swr_gs_llvm_emit_vertex(const struct lp_build_tgsi_gs_iface *gs_base
        } else if (iface->info->output_semantic_name[attrib] == TGSI_SEMANTIC_POSITION) {
           attribSlot = VERTEX_POSITION_SLOT;
        } else {
-          attribSlot = VERTEX_ATTRIB_START_SLOT + attrib - 1;
+          attribSlot = VERTEX_ATTRIB_START_SLOT + attrib;
+          if (iface->info->writes_position) {
+             attribSlot--;
+          }
        }
 
 #if USE_SIMD16_FRONTEND
@@ -923,6 +926,33 @@ swr_compile_vs(struct swr_context *ctx, swr_jit_vs_key &key)
    return func;
 }
 
+unsigned
+swr_so_adjust_attrib(unsigned in_attrib,
+                     swr_vertex_shader *swr_vs)
+{
+   ubyte semantic_name;
+   unsigned attrib;
+
+   attrib = in_attrib + VERTEX_ATTRIB_START_SLOT;
+
+   if (swr_vs) {
+      semantic_name = swr_vs->info.base.output_semantic_name[in_attrib];
+      if (semantic_name == TGSI_SEMANTIC_POSITION) {
+         attrib = VERTEX_POSITION_SLOT;
+      } else if (semantic_name == TGSI_SEMANTIC_PSIZE) {
+         attrib = VERTEX_SGV_SLOT;
+      } else if (semantic_name == TGSI_SEMANTIC_LAYER) {
+         attrib = VERTEX_SGV_SLOT;
+      } else {
+         if (swr_vs->info.base.writes_position) {
+               attrib--;
+         }
+      }
+   }
+
+   return attrib;
+}
+
 static unsigned
 locate_linkage(ubyte name, ubyte index, struct tgsi_shader_info *info)
 {

src/gallium/drivers/swr/swr_shader.h

@@ -30,6 +30,9 @@ struct swr_jit_fs_key;
 struct swr_jit_vs_key;
 struct swr_jit_gs_key;
 
+unsigned swr_so_adjust_attrib(unsigned in_attrib,
+                              swr_vertex_shader *swr_vs);
+
 PFN_VERTEX_FUNC
 swr_compile_vs(struct swr_context *ctx, swr_jit_vs_key &key);
 

src/gallium/drivers/swr/swr_state.cpp

@@ -345,13 +345,14 @@ swr_create_vs_state(struct pipe_context *pipe,
       // soState.streamToRasterizer not used
 
       for (uint32_t i = 0; i < stream_output->num_outputs; i++) {
+         unsigned attrib_slot = stream_output->output[i].register_index;
+         attrib_slot = swr_so_adjust_attrib(attrib_slot, swr_vs);
          swr_vs->soState.streamMasks[stream_output->output[i].stream] |=
-            1 << (stream_output->output[i].register_index - 1);
+            (1 << attrib_slot);
       }
       for (uint32_t i = 0; i < MAX_SO_STREAMS; i++) {
         swr_vs->soState.streamNumEntries[i] =
              _mm_popcnt_u32(swr_vs->soState.streamMasks[i]);
-        swr_vs->soState.vertexAttribOffset[i] = VERTEX_ATTRIB_START_SLOT; // TODO: optimize
        }
    }