nv50: add support for light-twoside
[mesa.git] / src / gallium / drivers / nv50 / nv50_vbo.c
index 584336682e443a76550d42df0687e6303a064fb3..eeed148c7b904bfd4b79c06ee846dd4359d47d3b 100644 (file)
@@ -22,6 +22,7 @@
 
 #include "pipe/p_context.h"
 #include "pipe/p_state.h"
+#include "pipe/p_inlines.h"
 
 #include "nv50_context.h"
 
@@ -48,26 +49,108 @@ nv50_prim(unsigned mode)
        return NV50TCL_VERTEX_BEGIN_POINTS;
 }
 
+static INLINE uint32_t
+nv50_vbo_type_to_hw(unsigned type)
+{
+       switch (type) {
+       case PIPE_FORMAT_TYPE_FLOAT:
+               return NV50TCL_VERTEX_ARRAY_ATTRIB_TYPE_FLOAT;
+       case PIPE_FORMAT_TYPE_UNORM:
+               return NV50TCL_VERTEX_ARRAY_ATTRIB_TYPE_UNORM;
+       case PIPE_FORMAT_TYPE_SNORM:
+               return NV50TCL_VERTEX_ARRAY_ATTRIB_TYPE_SNORM;
+       case PIPE_FORMAT_TYPE_USCALED:
+               return NV50TCL_VERTEX_ARRAY_ATTRIB_TYPE_USCALED;
+       case PIPE_FORMAT_TYPE_SSCALED:
+               return NV50TCL_VERTEX_ARRAY_ATTRIB_TYPE_SSCALED;
+       /*
+       case PIPE_FORMAT_TYPE_UINT:
+               return NV50TCL_VERTEX_ARRAY_ATTRIB_TYPE_UINT;
+       case PIPE_FORMAT_TYPE_SINT:
+               return NV50TCL_VERTEX_ARRAY_ATTRIB_TYPE_SINT; */
+       default:
+               return 0;
+       }
+}
+
+static INLINE uint32_t
+nv50_vbo_size_to_hw(unsigned size, unsigned nr_c)
+{
+       static const uint32_t hw_values[] = {
+               0, 0, 0, 0,
+               NV50TCL_VERTEX_ARRAY_ATTRIB_SIZE_8,
+               NV50TCL_VERTEX_ARRAY_ATTRIB_SIZE_8_8,
+               NV50TCL_VERTEX_ARRAY_ATTRIB_SIZE_8_8_8,
+               NV50TCL_VERTEX_ARRAY_ATTRIB_SIZE_8_8_8_8,
+               NV50TCL_VERTEX_ARRAY_ATTRIB_SIZE_16,
+               NV50TCL_VERTEX_ARRAY_ATTRIB_SIZE_16_16,
+               NV50TCL_VERTEX_ARRAY_ATTRIB_SIZE_16_16_16,
+               NV50TCL_VERTEX_ARRAY_ATTRIB_SIZE_16_16_16_16,
+               0, 0, 0, 0,
+               NV50TCL_VERTEX_ARRAY_ATTRIB_SIZE_32,
+               NV50TCL_VERTEX_ARRAY_ATTRIB_SIZE_32_32,
+               NV50TCL_VERTEX_ARRAY_ATTRIB_SIZE_32_32_32,
+               NV50TCL_VERTEX_ARRAY_ATTRIB_SIZE_32_32_32_32 };
+
+       /* we'd also have R11G11B10 and R10G10B10A2 */
+
+       assert(nr_c > 0 && nr_c <= 4);
+
+       if (size > 32)
+               return 0;
+       size >>= (3 - 2);
+
+       return hw_values[size + (nr_c - 1)];
+}
+
+static INLINE uint32_t
+nv50_vbo_vtxelt_to_hw(struct pipe_vertex_element *ve)
+{
+       uint32_t hw_type, hw_size;
+       enum pipe_format pf = ve->src_format;
+       unsigned size = pf_size_x(pf) << pf_exp2(pf);
+
+       hw_type = nv50_vbo_type_to_hw(pf_type(pf));
+       hw_size = nv50_vbo_size_to_hw(size, ve->nr_components);
+
+       if (!hw_type || !hw_size) {
+               NOUVEAU_ERR("unsupported vbo format: %s\n", pf_name(pf));
+               abort();
+               return 0x24e80000;
+       }
+
+       if (pf_swizzle_x(pf) == 2) /* BGRA */
+               hw_size |= (1 << 31); /* no real swizzle bits :-( */
+
+       return (hw_type | hw_size);
+}
+
 boolean
 nv50_draw_arrays(struct pipe_context *pipe, unsigned mode, unsigned start,
                 unsigned count)
 {
        struct nv50_context *nv50 = nv50_context(pipe);
+       struct nouveau_channel *chan = nv50->screen->tesla->channel;
+       struct nouveau_grobj *tesla = nv50->screen->tesla;
 
        nv50_state_validate(nv50);
 
-       BEGIN_RING(tesla, 0x142c, 1);
-       OUT_RING  (0);
-       BEGIN_RING(tesla, 0x142c, 1);
-       OUT_RING  (0);
-
-       BEGIN_RING(tesla, NV50TCL_VERTEX_BEGIN, 1);
-       OUT_RING  (nv50_prim(mode));
-       BEGIN_RING(tesla, NV50TCL_VERTEX_BUFFER_FIRST, 2);
-       OUT_RING  (start);
-       OUT_RING  (count);
-       BEGIN_RING(tesla, NV50TCL_VERTEX_END, 1);
-       OUT_RING  (0);
+       BEGIN_RING(chan, tesla, 0x142c, 1);
+       OUT_RING  (chan, 0);
+       BEGIN_RING(chan, tesla, 0x142c, 1);
+       OUT_RING  (chan, 0);
+       BEGIN_RING(chan, tesla, 0x1440, 1);
+       OUT_RING  (chan, 0);
+       BEGIN_RING(chan, tesla, 0x1334, 1);
+       OUT_RING  (chan, 0);
+
+       BEGIN_RING(chan, tesla, NV50TCL_VERTEX_BEGIN, 1);
+       OUT_RING  (chan, nv50_prim(mode));
+       BEGIN_RING(chan, tesla, NV50TCL_VERTEX_BUFFER_FIRST, 2);
+       OUT_RING  (chan, start);
+       OUT_RING  (chan, count);
+       BEGIN_RING(chan, tesla, NV50TCL_VERTEX_END, 1);
+       OUT_RING  (chan, 0);
 
        pipe->flush(pipe, 0, NULL);
        return TRUE;
@@ -77,11 +160,14 @@ static INLINE void
 nv50_draw_elements_inline_u08(struct nv50_context *nv50, uint8_t *map,
                              unsigned start, unsigned count)
 {
+       struct nouveau_channel *chan = nv50->screen->tesla->channel;
+       struct nouveau_grobj *tesla = nv50->screen->tesla;
+
        map += start;
 
        if (count & 1) {
-               BEGIN_RING(tesla, 0x15e8, 1);
-               OUT_RING  (map[0]);
+               BEGIN_RING(chan, tesla, 0x15e8, 1);
+               OUT_RING  (chan, map[0]);
                map++;
                count--;
        }
@@ -90,9 +176,9 @@ nv50_draw_elements_inline_u08(struct nv50_context *nv50, uint8_t *map,
                unsigned nr = count > 2046 ? 2046 : count;
                int i;
 
-               BEGIN_RING(tesla, 0x400015f0, nr >> 1);
+               BEGIN_RING(chan, tesla, 0x400015f0, nr >> 1);
                for (i = 0; i < nr; i += 2)
-                       OUT_RING  ((map[1] << 16) | map[0]);
+                       OUT_RING  (chan, (map[i + 1] << 16) | map[i]);
 
                count -= nr;
                map += nr;
@@ -103,11 +189,14 @@ static INLINE void
 nv50_draw_elements_inline_u16(struct nv50_context *nv50, uint16_t *map,
                              unsigned start, unsigned count)
 {
+       struct nouveau_channel *chan = nv50->screen->tesla->channel;
+       struct nouveau_grobj *tesla = nv50->screen->tesla;
+
        map += start;
 
        if (count & 1) {
-               BEGIN_RING(tesla, 0x15e8, 1);
-               OUT_RING  (map[0]);
+               BEGIN_RING(chan, tesla, 0x15e8, 1);
+               OUT_RING  (chan, map[0]);
                map++;
                count--;
        }
@@ -116,9 +205,9 @@ nv50_draw_elements_inline_u16(struct nv50_context *nv50, uint16_t *map,
                unsigned nr = count > 2046 ? 2046 : count;
                int i;
 
-               BEGIN_RING(tesla, 0x400015f0, nr >> 1);
+               BEGIN_RING(chan, tesla, 0x400015f0, nr >> 1);
                for (i = 0; i < nr; i += 2)
-                       OUT_RING  ((map[1] << 16) | map[0]);
+                       OUT_RING  (chan, (map[i + 1] << 16) | map[i]);
 
                count -= nr;
                map += nr;
@@ -126,16 +215,19 @@ nv50_draw_elements_inline_u16(struct nv50_context *nv50, uint16_t *map,
 }
 
 static INLINE void
-nv50_draw_elements_inline_u32(struct nv50_context *nv50, uint8_t *map,
+nv50_draw_elements_inline_u32(struct nv50_context *nv50, uint32_t *map,
                              unsigned start, unsigned count)
 {
+       struct nouveau_channel *chan = nv50->screen->tesla->channel;
+       struct nouveau_grobj *tesla = nv50->screen->tesla;
+
        map += start;
 
        while (count) {
                unsigned nr = count > 2047 ? 2047 : count;
 
-               BEGIN_RING(tesla, 0x400015e8, nr);
-               OUT_RINGp (map, nr);
+               BEGIN_RING(chan, tesla, 0x400015e8, nr);
+               OUT_RINGp (chan, map, nr);
 
                count -= nr;
                map += nr;
@@ -148,18 +240,22 @@ nv50_draw_elements(struct pipe_context *pipe,
                   unsigned mode, unsigned start, unsigned count)
 {
        struct nv50_context *nv50 = nv50_context(pipe);
-       struct pipe_winsys *ws = pipe->winsys;
-       void *map = ws->buffer_map(ws, indexBuffer, PIPE_BUFFER_USAGE_CPU_READ);
+       struct nouveau_channel *chan = nv50->screen->tesla->channel;
+       struct nouveau_grobj *tesla = nv50->screen->tesla;
+       struct pipe_screen *pscreen = pipe->screen;
+       void *map;
+       
+       map = pipe_buffer_map(pscreen, indexBuffer, PIPE_BUFFER_USAGE_CPU_READ);
 
        nv50_state_validate(nv50);
 
-       BEGIN_RING(tesla, 0x142c, 1);
-       OUT_RING  (0);
-       BEGIN_RING(tesla, 0x142c, 1);
-       OUT_RING  (0);
+       BEGIN_RING(chan, tesla, 0x142c, 1);
+       OUT_RING  (chan, 0);
+       BEGIN_RING(chan, tesla, 0x142c, 1);
+       OUT_RING  (chan, 0);
 
-       BEGIN_RING(tesla, NV50TCL_VERTEX_BEGIN, 1);
-       OUT_RING  (nv50_prim(mode));
+       BEGIN_RING(chan, tesla, NV50TCL_VERTEX_BEGIN, 1);
+       OUT_RING  (chan, nv50_prim(mode));
        switch (indexSize) {
        case 1:
                nv50_draw_elements_inline_u08(nv50, map, start, count);
@@ -173,65 +269,132 @@ nv50_draw_elements(struct pipe_context *pipe,
        default:
                assert(0);
        }
-       BEGIN_RING(tesla, NV50TCL_VERTEX_END, 1);
-       OUT_RING  (0);
+       BEGIN_RING(chan, tesla, NV50TCL_VERTEX_END, 1);
+       OUT_RING  (chan, 0);
 
+       pipe_buffer_unmap(pscreen, indexBuffer);
        pipe->flush(pipe, 0, NULL);
        return TRUE;
 }
 
+static INLINE boolean
+nv50_vbo_static_attrib(struct nv50_context *nv50, unsigned attrib,
+                      struct nouveau_stateobj **pso,
+                      struct pipe_vertex_element *ve,
+                      struct pipe_vertex_buffer *vb)
+
+{
+       struct nouveau_stateobj *so;
+       struct nouveau_grobj *tesla = nv50->screen->tesla;
+       struct nouveau_bo *bo = nouveau_bo(vb->buffer);
+       float *v;
+       int ret;
+       enum pipe_format pf = ve->src_format;
+
+       if ((pf_type(pf) != PIPE_FORMAT_TYPE_FLOAT) ||
+           (pf_size_x(pf) << pf_exp2(pf)) != 32)
+               return FALSE;
+
+       ret = nouveau_bo_map(bo, NOUVEAU_BO_RD);
+       if (ret)
+               return FALSE;
+       v = (float *)(bo->map + (vb->buffer_offset + ve->src_offset));
+
+       so = *pso;
+       if (!so)
+               *pso = so = so_new(nv50->vtxelt_nr * 5, 0);
+
+       switch (ve->nr_components) {
+       case 4:
+               so_method(so, tesla, NV50TCL_VTX_ATTR_4F_X(attrib), 4);
+               so_data  (so, fui(v[0]));
+               so_data  (so, fui(v[1]));
+               so_data  (so, fui(v[2]));
+               so_data  (so, fui(v[3]));
+               break;
+       case 3:
+               so_method(so, tesla, NV50TCL_VTX_ATTR_3F_X(attrib), 3);
+               so_data  (so, fui(v[0]));
+               so_data  (so, fui(v[1]));
+               so_data  (so, fui(v[2]));
+               break;
+       case 2:
+               so_method(so, tesla, NV50TCL_VTX_ATTR_2F_X(attrib), 2);
+               so_data  (so, fui(v[0]));
+               so_data  (so, fui(v[1]));
+               break;
+       case 1:
+               so_method(so, tesla, NV50TCL_VTX_ATTR_1F(attrib), 1);
+               so_data  (so, fui(v[0]));
+               break;
+       default:
+               nouveau_bo_unmap(bo);
+               return FALSE;
+       }
+
+       nouveau_bo_unmap(bo);
+       return TRUE;
+}
+
 void
 nv50_vbo_validate(struct nv50_context *nv50)
 {
        struct nouveau_grobj *tesla = nv50->screen->tesla;
-       struct nouveau_stateobj *vtxbuf, *vtxfmt;
-       int i, vpi = 0;
+       struct nouveau_stateobj *vtxbuf, *vtxfmt, *vtxattr;
+       unsigned i;
+
+       /* don't validate if Gallium took away our buffers */
+       if (nv50->vtxbuf_nr == 0)
+               return;
 
-       vtxbuf = so_new(nv50->vtxelt_nr * 4, nv50->vtxelt_nr * 2);
+       vtxattr = NULL;
+       vtxbuf = so_new(nv50->vtxelt_nr * 7, nv50->vtxelt_nr * 4);
        vtxfmt = so_new(nv50->vtxelt_nr + 1, 0);
-       so_method(vtxfmt, tesla, 0x1ac0, nv50->vtxelt_nr);
+       so_method(vtxfmt, tesla, NV50TCL_VERTEX_ARRAY_ATTRIB(0),
+               nv50->vtxelt_nr);
 
        for (i = 0; i < nv50->vtxelt_nr; i++) {
                struct pipe_vertex_element *ve = &nv50->vtxelt[i];
                struct pipe_vertex_buffer *vb =
                        &nv50->vtxbuf[ve->vertex_buffer_index];
+               struct nouveau_bo *bo = nouveau_bo(vb->buffer);
+               uint32_t hw = nv50_vbo_vtxelt_to_hw(ve);
 
-               switch (ve->src_format) {
-               case PIPE_FORMAT_R32G32B32A32_FLOAT:
-                       so_data(vtxfmt, 0x7e080000 | i);
-                       break;
-               case PIPE_FORMAT_R32G32B32_FLOAT:
-                       so_data(vtxfmt, 0x7e100000 | i);
-                       break;
-               case PIPE_FORMAT_R32G32_FLOAT:
-                       so_data(vtxfmt, 0x7e200000 | i);
-                       break;
-               case PIPE_FORMAT_R32_FLOAT:
-                       so_data(vtxfmt, 0x7e900000 | i);
-                       break;
-               case PIPE_FORMAT_R8G8B8A8_UNORM:
-                       so_data(vtxfmt, 0x24500000 | i);
-                       break;
-               default:
-               {
-                       NOUVEAU_ERR("invalid vbo format %s\n",
-                                   pf_name(ve->src_format));
-                       assert(0);
-                       return;
-               }
+               if (!vb->stride &&
+                   nv50_vbo_static_attrib(nv50, i, &vtxattr, ve, vb)) {
+                       so_data(vtxfmt, hw | (1 << 4));
+
+                       so_method(vtxbuf, tesla,
+                                 NV50TCL_VERTEX_ARRAY_FORMAT(i), 1);
+                       so_data  (vtxbuf, 0);
+                       continue;
                }
+               so_data(vtxfmt, hw | i);
 
-               so_method(vtxbuf, tesla, 0x900 + (i * 16), 3);
-               so_data  (vtxbuf, 0x20000000 | vb->pitch);
-               so_reloc (vtxbuf, vb->buffer, vb->buffer_offset +
+               so_method(vtxbuf, tesla, NV50TCL_VERTEX_ARRAY_FORMAT(i), 3);
+               so_data  (vtxbuf, 0x20000000 | vb->stride);
+               so_reloc (vtxbuf, bo, vb->buffer_offset +
                          ve->src_offset, NOUVEAU_BO_VRAM | NOUVEAU_BO_GART |
                          NOUVEAU_BO_RD | NOUVEAU_BO_HIGH, 0, 0);
-               so_reloc (vtxbuf, vb->buffer, vb->buffer_offset +
+               so_reloc (vtxbuf, bo, vb->buffer_offset +
                          ve->src_offset, NOUVEAU_BO_VRAM | NOUVEAU_BO_GART |
                          NOUVEAU_BO_RD | NOUVEAU_BO_LOW, 0, 0);
+
+               /* vertex array limits */
+               so_method(vtxbuf, tesla, 0x1080 + (i * 8), 2);
+               so_reloc (vtxbuf, bo, vb->buffer->size - 1,
+                         NOUVEAU_BO_VRAM | NOUVEAU_BO_GART | NOUVEAU_BO_RD |
+                         NOUVEAU_BO_HIGH, 0, 0);
+               so_reloc (vtxbuf, bo, vb->buffer->size - 1,
+                         NOUVEAU_BO_VRAM | NOUVEAU_BO_GART | NOUVEAU_BO_RD |
+                         NOUVEAU_BO_LOW, 0, 0);
        }
 
        so_ref (vtxfmt, &nv50->state.vtxfmt);
        so_ref (vtxbuf, &nv50->state.vtxbuf);
+       so_ref (vtxattr, &nv50->state.vtxattr);
+       so_ref (NULL, &vtxbuf);
+       so_ref (NULL, &vtxfmt);
+       so_ref (NULL, &vtxattr);
 }