gallium: add void *user_buffer in pipe_vertex_buffer

This reduces CPU overhead in st_draw_vbo and removes a lot of unnecessary code
in that function which was required only to comply with the gallium interface,
but wasn't any useful really.

Adapted drivers: i915, llvmpipe, r300, softpipe.
No changes required in: r600, radeonsi.

User vertex buffers have been disabled in nv30, nv50, nvc0 and svga to keep
things working.

13 files changed, 42 insertions, 22 deletions

diff --git a/src/gallium/auxiliary/draw/draw_llvm.c b/src/gallium/auxiliary/draw/draw_llvm.c
index 620d6dc8de3..7bf9d5e5d7a 100644
--- a/src/gallium/auxiliary/draw/draw_llvm.c
+++ b/src/gallium/auxiliary/draw/draw_llvm.c
@@ -236,12 +236,13 @@ static LLVMTypeRef
 create_jit_vertex_buffer_type(struct gallivm_state *gallivm, const char *struct_name)
 {
    LLVMTargetDataRef target = gallivm->target;
-   LLVMTypeRef elem_types[3];
+   LLVMTypeRef elem_types[4];
    LLVMTypeRef vb_type;
 
    elem_types[0] =
    elem_types[1] = LLVMInt32TypeInContext(gallivm->context);
-   elem_types[2] = LLVMPointerType(LLVMInt8TypeInContext(gallivm->context), 0); /* vs_constants */
+   elem_types[2] =
+   elem_types[3] = LLVMPointerType(LLVMInt8TypeInContext(gallivm->context), 0); /* vs_constants */
 
 #if HAVE_LLVM >= 0x0300
    vb_type = LLVMStructCreateNamed(gallivm->context, struct_name);
diff --git a/src/gallium/auxiliary/util/u_draw.c b/src/gallium/auxiliary/util/u_draw.c
index d16575b7340..1f3eb887959 100644
--- a/src/gallium/auxiliary/util/u_draw.c
+++ b/src/gallium/auxiliary/util/u_draw.c
@@ -62,6 +62,10 @@ util_draw_max_index(
       const struct util_format_description *format_desc;
       unsigned format_size;
 
+      if (!buffer->buffer) {
+         continue;
+      }
+
       assert(buffer->buffer->height0 == 1);
       assert(buffer->buffer->depth0 == 1);
       buffer_size = buffer->buffer->width0;
diff --git a/src/gallium/auxiliary/util/u_vbuf.c b/src/gallium/auxiliary/util/u_vbuf.c
index 401c8a4a850..e60378c31c7 100644
--- a/src/gallium/auxiliary/util/u_vbuf.c
+++ b/src/gallium/auxiliary/util/u_vbuf.c
@@ -277,8 +277,8 @@ u_vbuf_translate_buffers(struct u_vbuf *mgr, struct translate_key *key,
          unsigned offset = vb->buffer_offset + vb->stride * start_vertex;
          uint8_t *map;
 
-         if (vb->buffer->user_ptr) {
-            map = vb->buffer->user_ptr + offset;
+         if (vb->user_buffer) {
+            map = (uint8_t*)vb->user_buffer + offset;
          } else {
             unsigned size = vb->stride ? num_vertices * vb->stride
                                        : sizeof(double)*4;
@@ -713,6 +713,7 @@ void u_vbuf_set_vertex_buffers(struct u_vbuf *mgr, unsigned count,
       struct pipe_vertex_buffer *real_vb = &mgr->real_vertex_buffer[i];
 
       pipe_resource_reference(&orig_vb->buffer, vb->buffer);
+      orig_vb->user_buffer = vb->user_buffer;
 
       real_vb->buffer_offset = orig_vb->buffer_offset = vb->buffer_offset;
       real_vb->stride = orig_vb->stride = vb->stride;
@@ -721,7 +722,7 @@ void u_vbuf_set_vertex_buffers(struct u_vbuf *mgr, unsigned count,
          mgr->nonzero_stride_vb_mask |= 1 << i;
       }
 
-      if (!vb->buffer) {
+      if (!vb->buffer && !vb->user_buffer) {
          pipe_resource_reference(&real_vb->buffer, NULL);
          continue;
       }
@@ -733,7 +734,7 @@ void u_vbuf_set_vertex_buffers(struct u_vbuf *mgr, unsigned count,
          continue;
       }
 
-      if (!mgr->caps.user_vertex_buffers && vb->buffer->user_ptr) {
+      if (!mgr->caps.user_vertex_buffers && vb->user_buffer) {
          mgr->user_vb_mask |= 1 << i;
          pipe_resource_reference(&real_vb->buffer, NULL);
          continue;
@@ -798,9 +799,7 @@ u_vbuf_upload_buffers(struct u_vbuf *mgr,
          continue;
       }
 
-      assert(vb->buffer);
-
-      if (!vb->buffer->user_ptr) {
+      if (!vb->user_buffer) {
          continue;
       }
 
@@ -837,7 +836,7 @@ u_vbuf_upload_buffers(struct u_vbuf *mgr,
    for (i = 0; i < nr_vbufs; i++) {
       unsigned start, end = end_offset[i];
       struct pipe_vertex_buffer *real_vb;
-      uint8_t *ptr;
+      const uint8_t *ptr;
 
       if (!end) {
          continue;
@@ -847,7 +846,7 @@ u_vbuf_upload_buffers(struct u_vbuf *mgr,
       assert(start < end);
 
       real_vb = &mgr->real_vertex_buffer[i];
-      ptr = mgr->vertex_buffer[i].buffer->user_ptr;
+      ptr = mgr->vertex_buffer[i].user_buffer;
 
       u_upload_data(mgr->uploader, start, end - start, ptr + start,
                     &real_vb->buffer_offset, &real_vb->buffer);
diff --git a/src/gallium/drivers/i915/i915_state.c b/src/gallium/drivers/i915/i915_state.c
index 40cef5a9d86..d38d1ede5d6 100644
--- a/src/gallium/drivers/i915/i915_state.c
+++ b/src/gallium/drivers/i915/i915_state.c
@@ -983,7 +983,9 @@ static void i915_set_vertex_buffers(struct pipe_context *pipe,
 
    /* map new */
    for (i = 0; i < count; i++) {
-      void *buf = i915_buffer(buffers[i].buffer)->data;
+      const void *buf = buffers[i].user_buffer;
+      if (!buf)
+            buf = i915_buffer(buffers[i].buffer)->data;
       draw_set_mapped_vertex_buffer(draw, i, buf);
    }
 }
diff --git a/src/gallium/drivers/llvmpipe/lp_draw_arrays.c b/src/gallium/drivers/llvmpipe/lp_draw_arrays.c
index 239a596bce0..5a4a9bd6801 100644
--- a/src/gallium/drivers/llvmpipe/lp_draw_arrays.c
+++ b/src/gallium/drivers/llvmpipe/lp_draw_arrays.c
@@ -67,7 +67,9 @@ llvmpipe_draw_vbo(struct pipe_context *pipe, const struct pipe_draw_info *info)
     * Map vertex buffers
     */
    for (i = 0; i < lp->num_vertex_buffers; i++) {
-      void *buf = llvmpipe_resource_data(lp->vertex_buffer[i].buffer);
+      const void *buf = lp->vertex_buffer[i].user_buffer;
+      if (!buf)
+         buf = llvmpipe_resource_data(lp->vertex_buffer[i].buffer);
       draw_set_mapped_vertex_buffer(draw, i, buf);
    }
 
diff --git a/src/gallium/drivers/nv30/nv30_draw.c b/src/gallium/drivers/nv30/nv30_draw.c
index 61e324606f2..c6907fa806f 100644
--- a/src/gallium/drivers/nv30/nv30_draw.c
+++ b/src/gallium/drivers/nv30/nv30_draw.c
@@ -366,7 +366,7 @@ nv30_render_vbo(struct pipe_context *pipe, const struct pipe_draw_info *info)
 {
    struct nv30_context *nv30 = nv30_context(pipe);
    struct draw_context *draw = nv30->draw;
-   struct pipe_transfer *transfer[PIPE_MAX_ATTRIBS];
+   struct pipe_transfer *transfer[PIPE_MAX_ATTRIBS] = {NULL};
    struct pipe_transfer *transferi = NULL;
    int i;
 
@@ -403,7 +403,9 @@ nv30_render_vbo(struct pipe_context *pipe, const struct pipe_draw_info *info)
    }
 
    for (i = 0; i < nv30->num_vtxbufs; i++) {
-      void *map = pipe_buffer_map(pipe, nv30->vtxbuf[i].buffer,
+      const void *map = nv30->vtxbuf[i].user_buffer;
+      if (!map)
+         map = pipe_buffer_map(pipe, nv30->vtxbuf[i].buffer,
                                   PIPE_TRANSFER_UNSYNCHRONIZED |
                                   PIPE_TRANSFER_READ, &transfer[i]);
       draw_set_mapped_vertex_buffer(draw, i, map);
@@ -425,7 +427,8 @@ nv30_render_vbo(struct pipe_context *pipe, const struct pipe_draw_info *info)
    if (info->indexed)
       pipe_buffer_unmap(pipe, transferi);
    for (i = 0; i < nv30->num_vtxbufs; i++)
-      pipe_buffer_unmap(pipe, transfer[i]);
+      if (transfer[i])
+         pipe_buffer_unmap(pipe, transfer[i]);
 
    nv30->draw_dirty = 0;
    nv30_state_release(nv30);
diff --git a/src/gallium/drivers/nv30/nv30_screen.c b/src/gallium/drivers/nv30/nv30_screen.c
index 26969ef938b..5d550d13b5e 100644
--- a/src/gallium/drivers/nv30/nv30_screen.c
+++ b/src/gallium/drivers/nv30/nv30_screen.c
@@ -78,10 +78,11 @@ nv30_screen_get_param(struct pipe_screen *pscreen, enum pipe_cap param)
    case PIPE_CAP_TGSI_FS_COORD_ORIGIN_LOWER_LEFT:
    case PIPE_CAP_TGSI_FS_COORD_PIXEL_CENTER_HALF_INTEGER:
    case PIPE_CAP_TGSI_FS_COORD_PIXEL_CENTER_INTEGER:
-   case PIPE_CAP_USER_VERTEX_BUFFERS:
    case PIPE_CAP_USER_INDEX_BUFFERS:
    case PIPE_CAP_USER_CONSTANT_BUFFERS:
       return 1;
+   case PIPE_CAP_USER_VERTEX_BUFFERS:
+      return 0;
    case PIPE_CAP_CONSTANT_BUFFER_OFFSET_ALIGNMENT:
       return 16;
    /* nv4x capabilities */
diff --git a/src/gallium/drivers/nv50/nv50_screen.c b/src/gallium/drivers/nv50/nv50_screen.c
index 20a135b6039..61925ac23c9 100644
--- a/src/gallium/drivers/nv50/nv50_screen.c
+++ b/src/gallium/drivers/nv50/nv50_screen.c
@@ -149,8 +149,8 @@ nv50_screen_get_param(struct pipe_screen *pscreen, enum pipe_cap param)
       return 1;
    case PIPE_CAP_TGSI_CAN_COMPACT_VARYINGS:
    case PIPE_CAP_TGSI_CAN_COMPACT_CONSTANTS:
-      return 0; /* state trackers will know better */
    case PIPE_CAP_USER_VERTEX_BUFFERS:
+      return 0; /* state trackers will know better */
    case PIPE_CAP_USER_INDEX_BUFFERS:
    case PIPE_CAP_USER_CONSTANT_BUFFERS:
       return 1;
diff --git a/src/gallium/drivers/nvc0/nvc0_screen.c b/src/gallium/drivers/nvc0/nvc0_screen.c
index 366e6fbb832..846b21ad6c9 100644
--- a/src/gallium/drivers/nvc0/nvc0_screen.c
+++ b/src/gallium/drivers/nvc0/nvc0_screen.c
@@ -139,8 +139,8 @@ nvc0_screen_get_param(struct pipe_screen *pscreen, enum pipe_cap param)
       return 1;
    case PIPE_CAP_TGSI_CAN_COMPACT_VARYINGS:
    case PIPE_CAP_TGSI_CAN_COMPACT_CONSTANTS:
-      return 0; /* state trackers will know better */
    case PIPE_CAP_USER_VERTEX_BUFFERS:
+      return 0; /* state trackers will know better */
    case PIPE_CAP_USER_INDEX_BUFFERS:
    case PIPE_CAP_USER_CONSTANT_BUFFERS:
       return 1;
diff --git a/src/gallium/drivers/r300/r300_render.c b/src/gallium/drivers/r300/r300_render.c
index 99302dc8b0a..76165f51a1b 100644
--- a/src/gallium/drivers/r300/r300_render.c
+++ b/src/gallium/drivers/r300/r300_render.c
@@ -873,7 +873,10 @@ static void r300_swtcl_draw_vbo(struct pipe_context* pipe,
             indexed ? 256 : 6);
 
     for (i = 0; i < r300->nr_vertex_buffers; i++) {
-        if (r300->vertex_buffer[i].buffer) {
+        if (r300->vertex_buffer[i].user_buffer) {
+            draw_set_mapped_vertex_buffer(r300->draw, i,
+                                          r300->vertex_buffer[i].user_buffer);
+        } else if (r300->vertex_buffer[i].buffer) {
             void *buf = pipe_buffer_map(pipe,
                                   r300->vertex_buffer[i].buffer,
                                   PIPE_TRANSFER_READ |
@@ -899,7 +902,8 @@ static void r300_swtcl_draw_vbo(struct pipe_context* pipe,
 
     for (i = 0; i < r300->nr_vertex_buffers; i++) {
         if (r300->vertex_buffer[i].buffer) {
-            pipe_buffer_unmap(pipe, vb_transfer[i]);
+            if (vb_transfer[i])
+                pipe_buffer_unmap(pipe, vb_transfer[i]);
             draw_set_mapped_vertex_buffer(r300->draw, i, NULL);
         }
     }
diff --git a/src/gallium/drivers/softpipe/sp_draw_arrays.c b/src/gallium/drivers/softpipe/sp_draw_arrays.c
index 27004071f02..f2ffe590f85 100644
--- a/src/gallium/drivers/softpipe/sp_draw_arrays.c
+++ b/src/gallium/drivers/softpipe/sp_draw_arrays.c
@@ -77,7 +77,9 @@ softpipe_draw_vbo(struct pipe_context *pipe,
 
    /* Map vertex buffers */
    for (i = 0; i < sp->num_vertex_buffers; i++) {
-      void *buf = softpipe_resource(sp->vertex_buffer[i].buffer)->data;
+      const void *buf = sp->vertex_buffer[i].user_buffer;
+      if (!buf)
+         buf = softpipe_resource(sp->vertex_buffer[i].buffer)->data;
       draw_set_mapped_vertex_buffer(draw, i, buf);
    }
 
diff --git a/src/gallium/drivers/svga/svga_screen.c b/src/gallium/drivers/svga/svga_screen.c
index 51c85838fe2..f37d1724657 100644
--- a/src/gallium/drivers/svga/svga_screen.c
+++ b/src/gallium/drivers/svga/svga_screen.c
@@ -161,6 +161,7 @@ svga_get_param(struct pipe_screen *screen, enum pipe_cap param)
    case PIPE_CAP_TEXTURE_SWIZZLE:
       return 1;
    case PIPE_CAP_USER_VERTEX_BUFFERS:
+      return 0;
    case PIPE_CAP_USER_INDEX_BUFFERS:
    case PIPE_CAP_USER_CONSTANT_BUFFERS:
       return 1;
diff --git a/src/gallium/include/pipe/p_state.h b/src/gallium/include/pipe/p_state.h
index 42510a685bf..f281791eead 100644
--- a/src/gallium/include/pipe/p_state.h
+++ b/src/gallium/include/pipe/p_state.h
@@ -447,6 +447,7 @@ struct pipe_vertex_buffer
    unsigned stride;    /**< stride to same attrib in next vertex, in bytes */
    unsigned buffer_offset;  /**< offset to start of data in buffer, in bytes */
    struct pipe_resource *buffer;  /**< the actual buffer */
+   const void *user_buffer;  /**< pointer to a user buffer if buffer == NULL */
 };