static void brw_emit_vertices(struct brw_context *brw)
{
struct gl_context *ctx = &brw->ctx;
GLuint i, nr_elements;
brw_prepare_vertices(brw);
brw_emit_query_begin(brw);
nr_elements = brw->vb.nr_enabled + brw->vs.prog_data->uses_vertexid;
/* If the VS doesn't read any inputs (calculating vertex position from
* a state variable for some reason, for example), emit a single pad
* VERTEX_ELEMENT struct and bail.
*
* The stale VB state stays in place, but they don't do anything unless
* a VE loads from them.
*/
if (nr_elements == 0) {
BEGIN_BATCH(3);
OUT_BATCH((_3DSTATE_VERTEX_ELEMENTS << 16) | 1);
if (brw->gen >= 6) {
OUT_BATCH((0 << GEN6_VE0_INDEX_SHIFT) |
GEN6_VE0_VALID |
(BRW_SURFACEFORMAT_R32G32B32A32_FLOAT << BRW_VE0_FORMAT_SHIFT) |
(0 << BRW_VE0_SRC_OFFSET_SHIFT));
} else {
OUT_BATCH((0 << BRW_VE0_INDEX_SHIFT) |
BRW_VE0_VALID |
(BRW_SURFACEFORMAT_R32G32B32A32_FLOAT << BRW_VE0_FORMAT_SHIFT) |
(0 << BRW_VE0_SRC_OFFSET_SHIFT));
}
OUT_BATCH((BRW_VE1_COMPONENT_STORE_0 << BRW_VE1_COMPONENT_0_SHIFT) |
(BRW_VE1_COMPONENT_STORE_0 << BRW_VE1_COMPONENT_1_SHIFT) |
(BRW_VE1_COMPONENT_STORE_0 << BRW_VE1_COMPONENT_2_SHIFT) |
(BRW_VE1_COMPONENT_STORE_1_FLT << BRW_VE1_COMPONENT_3_SHIFT));
ADVANCE_BATCH();
return;
}
/* Now emit VB and VEP state packets.
*/
if (brw->vb.nr_buffers) {
if (brw->gen >= 6) {
assert(brw->vb.nr_buffers <= 33);
} else {
assert(brw->vb.nr_buffers <= 17);
}
BEGIN_BATCH(1 + 4*brw->vb.nr_buffers);
OUT_BATCH((_3DSTATE_VERTEX_BUFFERS << 16) | (4*brw->vb.nr_buffers - 1));
for (i = 0; i < brw->vb.nr_buffers; i++) {
struct brw_vertex_buffer *buffer = &brw->vb.buffers[i];
uint32_t dw0;
if (brw->gen >= 6) {
dw0 = buffer->step_rate
? GEN6_VB0_ACCESS_INSTANCEDATA
: GEN6_VB0_ACCESS_VERTEXDATA;
dw0 |= i << GEN6_VB0_INDEX_SHIFT;
} else {
dw0 = buffer->step_rate
? BRW_VB0_ACCESS_INSTANCEDATA
: BRW_VB0_ACCESS_VERTEXDATA;
dw0 |= i << BRW_VB0_INDEX_SHIFT;
}
if (brw->gen >= 7)
dw0 |= GEN7_VB0_ADDRESS_MODIFYENABLE;
if (brw->gen == 7)
dw0 |= GEN7_MOCS_L3 << 16;
WARN_ONCE(buffer->stride >= (brw->gen >= 5 ? 2048 : 2047),
"VBO stride %d too large, bad rendering may occur\n",
buffer->stride);
OUT_BATCH(dw0 | (buffer->stride << BRW_VB0_PITCH_SHIFT));
OUT_RELOC(buffer->bo, I915_GEM_DOMAIN_VERTEX, 0, buffer->offset);
if (brw->gen >= 5) {
OUT_RELOC(buffer->bo, I915_GEM_DOMAIN_VERTEX, 0, buffer->bo->size - 1);
} else
OUT_BATCH(0);
OUT_BATCH(buffer->step_rate);
}
ADVANCE_BATCH();
}
/* The hardware allows one more VERTEX_ELEMENTS than VERTEX_BUFFERS, presumably
* for VertexID/InstanceID.
*/
if (brw->gen >= 6) {
assert(nr_elements <= 34);
} else {
assert(nr_elements <= 18);
}
struct brw_vertex_element *gen6_edgeflag_input = NULL;
BEGIN_BATCH(1 + nr_elements * 2);
OUT_BATCH((_3DSTATE_VERTEX_ELEMENTS << 16) | (2 * nr_elements - 1));
for (i = 0; i < brw->vb.nr_enabled; i++) {
struct brw_vertex_element *input = brw->vb.enabled[i];
uint32_t format = brw_get_vertex_surface_type(brw, input->glarray);
uint32_t comp0 = BRW_VE1_COMPONENT_STORE_SRC;
uint32_t comp1 = BRW_VE1_COMPONENT_STORE_SRC;
uint32_t comp2 = BRW_VE1_COMPONENT_STORE_SRC;
uint32_t comp3 = BRW_VE1_COMPONENT_STORE_SRC;
if (input == &brw->vb.inputs[VERT_ATTRIB_EDGEFLAG]) {
/* Gen6+ passes edgeflag as sideband along with the vertex, instead
* of in the VUE. We have to upload it sideband as the last vertex
* element according to the B-Spec.
*/
if (brw->gen >= 6) {
gen6_edgeflag_input = input;
continue;
}
}
switch (input->glarray->Size) {
case 0: comp0 = BRW_VE1_COMPONENT_STORE_0;
case 1: comp1 = BRW_VE1_COMPONENT_STORE_0;
case 2: comp2 = BRW_VE1_COMPONENT_STORE_0;
case 3: comp3 = input->glarray->Integer ? BRW_VE1_COMPONENT_STORE_1_INT
: BRW_VE1_COMPONENT_STORE_1_FLT;
break;
}
if (brw->gen >= 6) {
OUT_BATCH((input->buffer << GEN6_VE0_INDEX_SHIFT) |
GEN6_VE0_VALID |
(format << BRW_VE0_FORMAT_SHIFT) |
(input->offset << BRW_VE0_SRC_OFFSET_SHIFT));
} else {
OUT_BATCH((input->buffer << BRW_VE0_INDEX_SHIFT) |
BRW_VE0_VALID |
(format << BRW_VE0_FORMAT_SHIFT) |
(input->offset << BRW_VE0_SRC_OFFSET_SHIFT));
}
if (brw->gen >= 5)
OUT_BATCH((comp0 << BRW_VE1_COMPONENT_0_SHIFT) |
(comp1 << BRW_VE1_COMPONENT_1_SHIFT) |
(comp2 << BRW_VE1_COMPONENT_2_SHIFT) |
(comp3 << BRW_VE1_COMPONENT_3_SHIFT));
else
OUT_BATCH((comp0 << BRW_VE1_COMPONENT_0_SHIFT) |
(comp1 << BRW_VE1_COMPONENT_1_SHIFT) |
(comp2 << BRW_VE1_COMPONENT_2_SHIFT) |
(comp3 << BRW_VE1_COMPONENT_3_SHIFT) |
((i * 4) << BRW_VE1_DST_OFFSET_SHIFT));
}
if (brw->gen >= 6 && gen6_edgeflag_input) {
uint32_t format =
brw_get_vertex_surface_type(brw, gen6_edgeflag_input->glarray);
OUT_BATCH((gen6_edgeflag_input->buffer << GEN6_VE0_INDEX_SHIFT) |
GEN6_VE0_VALID |
GEN6_VE0_EDGE_FLAG_ENABLE |
(format << BRW_VE0_FORMAT_SHIFT) |
(gen6_edgeflag_input->offset << BRW_VE0_SRC_OFFSET_SHIFT));
OUT_BATCH((BRW_VE1_COMPONENT_STORE_SRC << BRW_VE1_COMPONENT_0_SHIFT) |
(BRW_VE1_COMPONENT_STORE_0 << BRW_VE1_COMPONENT_1_SHIFT) |
(BRW_VE1_COMPONENT_STORE_0 << BRW_VE1_COMPONENT_2_SHIFT) |
(BRW_VE1_COMPONENT_STORE_0 << BRW_VE1_COMPONENT_3_SHIFT));
}
if (brw->vs.prog_data->uses_vertexid) {
uint32_t dw0 = 0, dw1 = 0;
dw1 = ((BRW_VE1_COMPONENT_STORE_VID << BRW_VE1_COMPONENT_0_SHIFT) |
(BRW_VE1_COMPONENT_STORE_IID << BRW_VE1_COMPONENT_1_SHIFT) |
(BRW_VE1_COMPONENT_STORE_0 << BRW_VE1_COMPONENT_2_SHIFT) |
(BRW_VE1_COMPONENT_STORE_0 << BRW_VE1_COMPONENT_3_SHIFT));
if (brw->gen >= 6) {
dw0 |= GEN6_VE0_VALID;
} else {
dw0 |= BRW_VE0_VALID;
dw1 |= (i * 4) << BRW_VE1_DST_OFFSET_SHIFT;
}
/* Note that for gl_VertexID, gl_InstanceID, and gl_PrimitiveID values,
* the format is ignored and the value is always int.
*/
OUT_BATCH(dw0);
OUT_BATCH(dw1);
}
ADVANCE_BATCH();
}
static void brw_emit_vertices(struct brw_context *brw)
{
struct gl_context *ctx = &brw->intel.ctx;
struct intel_context *intel = intel_context(ctx);
GLuint i, nr_elements;
brw_prepare_vertices(brw);
brw_emit_query_begin(brw);
/* If the VS doesn't read any inputs (calculating vertex position from
* a state variable for some reason, for example), emit a single pad
* VERTEX_ELEMENT struct and bail.
*
* The stale VB state stays in place, but they don't do anything unless
* a VE loads from them.
*/
if (brw->vb.nr_enabled == 0) {
BEGIN_BATCH(3);
OUT_BATCH((_3DSTATE_VERTEX_ELEMENTS << 16) | 1);
if (intel->gen >= 6) {
OUT_BATCH((0 << GEN6_VE0_INDEX_SHIFT) |
GEN6_VE0_VALID |
(BRW_SURFACEFORMAT_R32G32B32A32_FLOAT << BRW_VE0_FORMAT_SHIFT) |
(0 << BRW_VE0_SRC_OFFSET_SHIFT));
} else {
OUT_BATCH((0 << BRW_VE0_INDEX_SHIFT) |
BRW_VE0_VALID |
(BRW_SURFACEFORMAT_R32G32B32A32_FLOAT << BRW_VE0_FORMAT_SHIFT) |
(0 << BRW_VE0_SRC_OFFSET_SHIFT));
}
OUT_BATCH((BRW_VE1_COMPONENT_STORE_0 << BRW_VE1_COMPONENT_0_SHIFT) |
(BRW_VE1_COMPONENT_STORE_0 << BRW_VE1_COMPONENT_1_SHIFT) |
(BRW_VE1_COMPONENT_STORE_0 << BRW_VE1_COMPONENT_2_SHIFT) |
(BRW_VE1_COMPONENT_STORE_1_FLT << BRW_VE1_COMPONENT_3_SHIFT));
CACHED_BATCH();
return;
}
/* Now emit VB and VEP state packets.
*/
if (brw->vb.nr_buffers) {
if (intel->gen >= 6) {
assert(brw->vb.nr_buffers <= 33);
} else {
assert(brw->vb.nr_buffers <= 17);
}
BEGIN_BATCH(1 + 4*brw->vb.nr_buffers);
OUT_BATCH((_3DSTATE_VERTEX_BUFFERS << 16) | (4*brw->vb.nr_buffers - 1));
for (i = 0; i < brw->vb.nr_buffers; i++) {
struct brw_vertex_buffer *buffer = &brw->vb.buffers[i];
uint32_t dw0;
if (intel->gen >= 6) {
dw0 = buffer->step_rate
? GEN6_VB0_ACCESS_INSTANCEDATA
: GEN6_VB0_ACCESS_VERTEXDATA;
dw0 |= i << GEN6_VB0_INDEX_SHIFT;
} else {
dw0 = buffer->step_rate
? BRW_VB0_ACCESS_INSTANCEDATA
: BRW_VB0_ACCESS_VERTEXDATA;
dw0 |= i << BRW_VB0_INDEX_SHIFT;
}
if (intel->gen >= 7)
dw0 |= GEN7_VB0_ADDRESS_MODIFYENABLE;
OUT_BATCH(dw0 | (buffer->stride << BRW_VB0_PITCH_SHIFT));
OUT_RELOC(buffer->bo, I915_GEM_DOMAIN_VERTEX, 0, buffer->offset);
if (intel->gen >= 5) {
OUT_RELOC(buffer->bo, I915_GEM_DOMAIN_VERTEX, 0, buffer->bo->size - 1);
} else
OUT_BATCH(0);
OUT_BATCH(buffer->step_rate);
brw->vb.current_buffers[i].handle = buffer->bo->handle;
brw->vb.current_buffers[i].offset = buffer->offset;
brw->vb.current_buffers[i].stride = buffer->stride;
brw->vb.current_buffers[i].step_rate = buffer->step_rate;
}
brw->vb.nr_current_buffers = i;
ADVANCE_BATCH();
}
nr_elements = brw->vb.nr_enabled + brw->vs.prog_data->uses_vertexid;
/* The hardware allows one more VERTEX_ELEMENTS than VERTEX_BUFFERS, presumably
* for VertexID/InstanceID.
*/
if (intel->gen >= 6) {
assert(nr_elements <= 34);
} else {
assert(nr_elements <= 18);
}
BEGIN_BATCH(1 + nr_elements * 2);
OUT_BATCH((_3DSTATE_VERTEX_ELEMENTS << 16) | (2 * nr_elements - 1));
for (i = 0; i < brw->vb.nr_enabled; i++) {
struct brw_vertex_element *input = brw->vb.enabled[i];
uint32_t format = get_surface_type(input->glarray->Type,
input->glarray->Size,
input->glarray->Format,
input->glarray->Normalized,
input->glarray->Integer);
uint32_t comp0 = BRW_VE1_COMPONENT_STORE_SRC;
uint32_t comp1 = BRW_VE1_COMPONENT_STORE_SRC;
uint32_t comp2 = BRW_VE1_COMPONENT_STORE_SRC;
uint32_t comp3 = BRW_VE1_COMPONENT_STORE_SRC;
switch (input->glarray->Size) {
case 0: comp0 = BRW_VE1_COMPONENT_STORE_0;
case 1: comp1 = BRW_VE1_COMPONENT_STORE_0;
case 2: comp2 = BRW_VE1_COMPONENT_STORE_0;
case 3: comp3 = input->glarray->Integer ? BRW_VE1_COMPONENT_STORE_1_INT
: BRW_VE1_COMPONENT_STORE_1_FLT;
break;
}
if (intel->gen >= 6) {
OUT_BATCH((input->buffer << GEN6_VE0_INDEX_SHIFT) |
GEN6_VE0_VALID |
(format << BRW_VE0_FORMAT_SHIFT) |
(input->offset << BRW_VE0_SRC_OFFSET_SHIFT));
} else {
OUT_BATCH((input->buffer << BRW_VE0_INDEX_SHIFT) |
BRW_VE0_VALID |
(format << BRW_VE0_FORMAT_SHIFT) |
(input->offset << BRW_VE0_SRC_OFFSET_SHIFT));
}
if (intel->gen >= 5)
OUT_BATCH((comp0 << BRW_VE1_COMPONENT_0_SHIFT) |
(comp1 << BRW_VE1_COMPONENT_1_SHIFT) |
(comp2 << BRW_VE1_COMPONENT_2_SHIFT) |
(comp3 << BRW_VE1_COMPONENT_3_SHIFT));
else
OUT_BATCH((comp0 << BRW_VE1_COMPONENT_0_SHIFT) |
(comp1 << BRW_VE1_COMPONENT_1_SHIFT) |
(comp2 << BRW_VE1_COMPONENT_2_SHIFT) |
(comp3 << BRW_VE1_COMPONENT_3_SHIFT) |
((i * 4) << BRW_VE1_DST_OFFSET_SHIFT));
}
if (brw->vs.prog_data->uses_vertexid) {
uint32_t dw0 = 0, dw1 = 0;
dw1 = ((BRW_VE1_COMPONENT_STORE_VID << BRW_VE1_COMPONENT_0_SHIFT) |
(BRW_VE1_COMPONENT_STORE_IID << BRW_VE1_COMPONENT_1_SHIFT) |
(BRW_VE1_COMPONENT_STORE_0 << BRW_VE1_COMPONENT_2_SHIFT) |
(BRW_VE1_COMPONENT_STORE_0 << BRW_VE1_COMPONENT_3_SHIFT));
if (intel->gen >= 6) {
dw0 |= GEN6_VE0_VALID;
} else {
dw0 |= BRW_VE0_VALID;
dw1 |= (i * 4) << BRW_VE1_DST_OFFSET_SHIFT;
}
/* Note that for gl_VertexID, gl_InstanceID, and gl_PrimitiveID values,
* the format is ignored and the value is always int.
*/
OUT_BATCH(dw0);
OUT_BATCH(dw1);
}
CACHED_BATCH();
}
static void brw_emit_vertices(struct brw_context *brw)
{
struct gl_context *ctx = &brw->intel.ctx;
struct intel_context *intel = intel_context(ctx);
GLuint i, nr_elements;
brw_prepare_vertices(brw);
brw_emit_query_begin(brw);
/* If the VS doesn't read any inputs (calculating vertex position from
* a state variable for some reason, for example), emit a single pad
* VERTEX_ELEMENT struct and bail.
*
* The stale VB state stays in place, but they don't do anything unless
* a VE loads from them.
*/
if (brw->vb.nr_enabled == 0) {
BEGIN_BATCH(3);
OUT_BATCH((_3DSTATE_VERTEX_ELEMENTS << 16) | 1);
if (intel->gen >= 6) {
OUT_BATCH((0 << GEN6_VE0_INDEX_SHIFT) |
GEN6_VE0_VALID |
(BRW_SURFACEFORMAT_R32G32B32A32_FLOAT << BRW_VE0_FORMAT_SHIFT) |
(0 << BRW_VE0_SRC_OFFSET_SHIFT));
} else {
OUT_BATCH((0 << BRW_VE0_INDEX_SHIFT) |
BRW_VE0_VALID |
(BRW_SURFACEFORMAT_R32G32B32A32_FLOAT << BRW_VE0_FORMAT_SHIFT) |
(0 << BRW_VE0_SRC_OFFSET_SHIFT));
}
OUT_BATCH((BRW_VE1_COMPONENT_STORE_0 << BRW_VE1_COMPONENT_0_SHIFT) |
(BRW_VE1_COMPONENT_STORE_0 << BRW_VE1_COMPONENT_1_SHIFT) |
(BRW_VE1_COMPONENT_STORE_0 << BRW_VE1_COMPONENT_2_SHIFT) |
(BRW_VE1_COMPONENT_STORE_1_FLT << BRW_VE1_COMPONENT_3_SHIFT));
CACHED_BATCH();
return;
}
/* Now emit VB and VEP state packets.
*/
if (brw->vb.nr_buffers) {
if (intel->gen >= 6) {
assert(brw->vb.nr_buffers <= 33);
} else {
assert(brw->vb.nr_buffers <= 17);
}
BEGIN_BATCH(1 + 4*brw->vb.nr_buffers);
OUT_BATCH((_3DSTATE_VERTEX_BUFFERS << 16) | (4*brw->vb.nr_buffers - 1));
for (i = 0; i < brw->vb.nr_buffers; i++) {
struct brw_vertex_buffer *buffer = &brw->vb.buffers[i];
uint32_t dw0;
if (intel->gen >= 6) {
dw0 = buffer->step_rate
? GEN6_VB0_ACCESS_INSTANCEDATA
: GEN6_VB0_ACCESS_VERTEXDATA;
dw0 |= i << GEN6_VB0_INDEX_SHIFT;
} else {
dw0 = buffer->step_rate
? BRW_VB0_ACCESS_INSTANCEDATA
: BRW_VB0_ACCESS_VERTEXDATA;
dw0 |= i << BRW_VB0_INDEX_SHIFT;
}
if (intel->gen >= 7)
dw0 |= GEN7_VB0_ADDRESS_MODIFYENABLE;
OUT_BATCH(dw0 | (buffer->stride << BRW_VB0_PITCH_SHIFT));
OUT_RELOC(buffer->bo, I915_GEM_DOMAIN_VERTEX, 0, buffer->offset);
if (intel->gen >= 5) {
OUT_RELOC(buffer->bo, I915_GEM_DOMAIN_VERTEX, 0, buffer->bo->size - 1);
} else
OUT_BATCH(0);
OUT_BATCH(buffer->step_rate);
brw->vb.current_buffers[i].handle = buffer->bo->handle;
brw->vb.current_buffers[i].offset = buffer->offset;
brw->vb.current_buffers[i].stride = buffer->stride;
brw->vb.current_buffers[i].step_rate = buffer->step_rate;
}
brw->vb.nr_current_buffers = i;
ADVANCE_BATCH();
}
nr_elements = brw->vb.nr_enabled + brw->vs.prog_data->uses_vertexid;
/* The hardware allows one more VERTEX_ELEMENTS than VERTEX_BUFFERS, presumably
* for VertexID/InstanceID.
*/
if (intel->gen >= 6) {
assert(nr_elements <= 34);
} else {
assert(nr_elements <= 18);
}
struct brw_vertex_element *gen6_edgeflag_input = NULL;
BEGIN_BATCH(1 + nr_elements * 2);
OUT_BATCH((_3DSTATE_VERTEX_ELEMENTS << 16) | (2 * nr_elements - 1));
for (i = 0; i < brw->vb.nr_enabled; i++) {
struct brw_vertex_element *input = brw->vb.enabled[i];
uint32_t format = get_surface_type(input->glarray->Type,
input->glarray->Size,
input->glarray->Format,
input->glarray->Normalized,
input->glarray->Integer);
uint32_t comp0 = BRW_VE1_COMPONENT_STORE_SRC;
uint32_t comp1 = BRW_VE1_COMPONENT_STORE_SRC;
uint32_t comp2 = BRW_VE1_COMPONENT_STORE_SRC;
uint32_t comp3 = BRW_VE1_COMPONENT_STORE_SRC;
/* The gen4 driver expects edgeflag to come in as a float, and passes
* that float on to the tests in the clipper. Mesa's current vertex
* attribute value for EdgeFlag is stored as a float, which works out.
* glEdgeFlagPointer, on the other hand, gives us an unnormalized
* integer ubyte. Just rewrite that to convert to a float.
*/
if (input->attrib == VERT_ATTRIB_EDGEFLAG) {
/* Gen6+ passes edgeflag as sideband along with the vertex, instead
* of in the VUE. We have to upload it sideband as the last vertex
* element according to the B-Spec.
*/
if (intel->gen >= 6) {
gen6_edgeflag_input = input;
continue;
}
if (format == BRW_SURFACEFORMAT_R8_UINT)
format = BRW_SURFACEFORMAT_R8_SSCALED;
}
switch (input->glarray->Size) {
case 0: comp0 = BRW_VE1_COMPONENT_STORE_0;
case 1: comp1 = BRW_VE1_COMPONENT_STORE_0;
case 2: comp2 = BRW_VE1_COMPONENT_STORE_0;
case 3: comp3 = input->glarray->Integer ? BRW_VE1_COMPONENT_STORE_1_INT
: BRW_VE1_COMPONENT_STORE_1_FLT;
break;
}
if (intel->gen >= 6) {
OUT_BATCH((input->buffer << GEN6_VE0_INDEX_SHIFT) |
GEN6_VE0_VALID |
(format << BRW_VE0_FORMAT_SHIFT) |
(input->offset << BRW_VE0_SRC_OFFSET_SHIFT));
} else {
OUT_BATCH((input->buffer << BRW_VE0_INDEX_SHIFT) |
BRW_VE0_VALID |
(format << BRW_VE0_FORMAT_SHIFT) |
(input->offset << BRW_VE0_SRC_OFFSET_SHIFT));
}
if (intel->gen >= 5)
OUT_BATCH((comp0 << BRW_VE1_COMPONENT_0_SHIFT) |
(comp1 << BRW_VE1_COMPONENT_1_SHIFT) |
(comp2 << BRW_VE1_COMPONENT_2_SHIFT) |
(comp3 << BRW_VE1_COMPONENT_3_SHIFT));
else
OUT_BATCH((comp0 << BRW_VE1_COMPONENT_0_SHIFT) |
(comp1 << BRW_VE1_COMPONENT_1_SHIFT) |
(comp2 << BRW_VE1_COMPONENT_2_SHIFT) |
(comp3 << BRW_VE1_COMPONENT_3_SHIFT) |
((i * 4) << BRW_VE1_DST_OFFSET_SHIFT));
}
if (intel->gen >= 6 && gen6_edgeflag_input) {
uint32_t format = get_surface_type(gen6_edgeflag_input->glarray->Type,
gen6_edgeflag_input->glarray->Size,
gen6_edgeflag_input->glarray->Format,
gen6_edgeflag_input->glarray->Normalized,
gen6_edgeflag_input->glarray->Integer);
OUT_BATCH((gen6_edgeflag_input->buffer << GEN6_VE0_INDEX_SHIFT) |
GEN6_VE0_VALID |
GEN6_VE0_EDGE_FLAG_ENABLE |
(format << BRW_VE0_FORMAT_SHIFT) |
(gen6_edgeflag_input->offset << BRW_VE0_SRC_OFFSET_SHIFT));
OUT_BATCH((BRW_VE1_COMPONENT_STORE_SRC << BRW_VE1_COMPONENT_0_SHIFT) |
(BRW_VE1_COMPONENT_STORE_0 << BRW_VE1_COMPONENT_1_SHIFT) |
(BRW_VE1_COMPONENT_STORE_0 << BRW_VE1_COMPONENT_2_SHIFT) |
(BRW_VE1_COMPONENT_STORE_0 << BRW_VE1_COMPONENT_3_SHIFT));
}
if (brw->vs.prog_data->uses_vertexid) {
uint32_t dw0 = 0, dw1 = 0;
dw1 = ((BRW_VE1_COMPONENT_STORE_VID << BRW_VE1_COMPONENT_0_SHIFT) |
(BRW_VE1_COMPONENT_STORE_IID << BRW_VE1_COMPONENT_1_SHIFT) |
(BRW_VE1_COMPONENT_STORE_0 << BRW_VE1_COMPONENT_2_SHIFT) |
(BRW_VE1_COMPONENT_STORE_0 << BRW_VE1_COMPONENT_3_SHIFT));
if (intel->gen >= 6) {
dw0 |= GEN6_VE0_VALID;
} else {
dw0 |= BRW_VE0_VALID;
dw1 |= (i * 4) << BRW_VE1_DST_OFFSET_SHIFT;
}
/* Note that for gl_VertexID, gl_InstanceID, and gl_PrimitiveID values,
* the format is ignored and the value is always int.
*/
OUT_BATCH(dw0);
OUT_BATCH(dw1);
}
CACHED_BATCH();
}
static void brw_emit_vertices(struct brw_context *brw)
{
GLuint i;
brw_prepare_vertices(brw);
brw_prepare_shader_draw_parameters(brw);
brw_emit_query_begin(brw);
unsigned nr_elements = brw->vb.nr_enabled;
if (brw->vs.prog_data->uses_vertexid || brw->vs.prog_data->uses_instanceid)
++nr_elements;
/* If the VS doesn't read any inputs (calculating vertex position from
* a state variable for some reason, for example), emit a single pad
* VERTEX_ELEMENT struct and bail.
*
* The stale VB state stays in place, but they don't do anything unless
* a VE loads from them.
*/
if (nr_elements == 0) {
BEGIN_BATCH(3);
OUT_BATCH((_3DSTATE_VERTEX_ELEMENTS << 16) | 1);
if (brw->gen >= 6) {
OUT_BATCH((0 << GEN6_VE0_INDEX_SHIFT) |
GEN6_VE0_VALID |
(BRW_SURFACEFORMAT_R32G32B32A32_FLOAT << BRW_VE0_FORMAT_SHIFT) |
(0 << BRW_VE0_SRC_OFFSET_SHIFT));
} else {
OUT_BATCH((0 << BRW_VE0_INDEX_SHIFT) |
BRW_VE0_VALID |
(BRW_SURFACEFORMAT_R32G32B32A32_FLOAT << BRW_VE0_FORMAT_SHIFT) |
(0 << BRW_VE0_SRC_OFFSET_SHIFT));
}
OUT_BATCH((BRW_VE1_COMPONENT_STORE_0 << BRW_VE1_COMPONENT_0_SHIFT) |
(BRW_VE1_COMPONENT_STORE_0 << BRW_VE1_COMPONENT_1_SHIFT) |
(BRW_VE1_COMPONENT_STORE_0 << BRW_VE1_COMPONENT_2_SHIFT) |
(BRW_VE1_COMPONENT_STORE_1_FLT << BRW_VE1_COMPONENT_3_SHIFT));
ADVANCE_BATCH();
return;
}
/* Now emit VB and VEP state packets.
*/
unsigned nr_buffers =
brw->vb.nr_buffers + brw->vs.prog_data->uses_vertexid;
if (nr_buffers) {
if (brw->gen >= 6) {
assert(nr_buffers <= 33);
} else {
assert(nr_buffers <= 17);
}
BEGIN_BATCH(1 + 4 * nr_buffers);
OUT_BATCH((_3DSTATE_VERTEX_BUFFERS << 16) | (4 * nr_buffers - 1));
for (i = 0; i < brw->vb.nr_buffers; i++) {
struct brw_vertex_buffer *buffer = &brw->vb.buffers[i];
emit_vertex_buffer_state(brw, i, buffer->bo, buffer->bo->size - 1,
buffer->offset, buffer->stride,
buffer->step_rate);
}
if (brw->vs.prog_data->uses_vertexid) {
emit_vertex_buffer_state(brw, brw->vb.nr_buffers,
brw->draw.draw_params_bo,
brw->draw.draw_params_bo->size - 1,
brw->draw.draw_params_offset,
0, /* stride */
0); /* step rate */
}
ADVANCE_BATCH();
}
/* The hardware allows one more VERTEX_ELEMENTS than VERTEX_BUFFERS, presumably
* for VertexID/InstanceID.
*/
if (brw->gen >= 6) {
assert(nr_elements <= 34);
} else {
assert(nr_elements <= 18);
}
struct brw_vertex_element *gen6_edgeflag_input = NULL;
BEGIN_BATCH(1 + nr_elements * 2);
OUT_BATCH((_3DSTATE_VERTEX_ELEMENTS << 16) | (2 * nr_elements - 1));
for (i = 0; i < brw->vb.nr_enabled; i++) {
struct brw_vertex_element *input = brw->vb.enabled[i];
uint32_t format = brw_get_vertex_surface_type(brw, input->glarray);
uint32_t comp0 = BRW_VE1_COMPONENT_STORE_SRC;
uint32_t comp1 = BRW_VE1_COMPONENT_STORE_SRC;
uint32_t comp2 = BRW_VE1_COMPONENT_STORE_SRC;
uint32_t comp3 = BRW_VE1_COMPONENT_STORE_SRC;
if (input == &brw->vb.inputs[VERT_ATTRIB_EDGEFLAG]) {
/* Gen6+ passes edgeflag as sideband along with the vertex, instead
* of in the VUE. We have to upload it sideband as the last vertex
* element according to the B-Spec.
*/
if (brw->gen >= 6) {
gen6_edgeflag_input = input;
continue;
}
}
switch (input->glarray->Size) {
case 0: comp0 = BRW_VE1_COMPONENT_STORE_0;
case 1: comp1 = BRW_VE1_COMPONENT_STORE_0;
case 2: comp2 = BRW_VE1_COMPONENT_STORE_0;
case 3: comp3 = input->glarray->Integer ? BRW_VE1_COMPONENT_STORE_1_INT
: BRW_VE1_COMPONENT_STORE_1_FLT;
break;
}
if (brw->gen >= 6) {
OUT_BATCH((input->buffer << GEN6_VE0_INDEX_SHIFT) |
GEN6_VE0_VALID |
(format << BRW_VE0_FORMAT_SHIFT) |
(input->offset << BRW_VE0_SRC_OFFSET_SHIFT));
} else {
OUT_BATCH((input->buffer << BRW_VE0_INDEX_SHIFT) |
BRW_VE0_VALID |
(format << BRW_VE0_FORMAT_SHIFT) |
(input->offset << BRW_VE0_SRC_OFFSET_SHIFT));
}
if (brw->gen >= 5)
OUT_BATCH((comp0 << BRW_VE1_COMPONENT_0_SHIFT) |
(comp1 << BRW_VE1_COMPONENT_1_SHIFT) |
(comp2 << BRW_VE1_COMPONENT_2_SHIFT) |
(comp3 << BRW_VE1_COMPONENT_3_SHIFT));
else
OUT_BATCH((comp0 << BRW_VE1_COMPONENT_0_SHIFT) |
(comp1 << BRW_VE1_COMPONENT_1_SHIFT) |
(comp2 << BRW_VE1_COMPONENT_2_SHIFT) |
(comp3 << BRW_VE1_COMPONENT_3_SHIFT) |
((i * 4) << BRW_VE1_DST_OFFSET_SHIFT));
}
if (brw->gen >= 6 && gen6_edgeflag_input) {
uint32_t format =
brw_get_vertex_surface_type(brw, gen6_edgeflag_input->glarray);
OUT_BATCH((gen6_edgeflag_input->buffer << GEN6_VE0_INDEX_SHIFT) |
GEN6_VE0_VALID |
GEN6_VE0_EDGE_FLAG_ENABLE |
(format << BRW_VE0_FORMAT_SHIFT) |
(gen6_edgeflag_input->offset << BRW_VE0_SRC_OFFSET_SHIFT));
OUT_BATCH((BRW_VE1_COMPONENT_STORE_SRC << BRW_VE1_COMPONENT_0_SHIFT) |
(BRW_VE1_COMPONENT_STORE_0 << BRW_VE1_COMPONENT_1_SHIFT) |
(BRW_VE1_COMPONENT_STORE_0 << BRW_VE1_COMPONENT_2_SHIFT) |
(BRW_VE1_COMPONENT_STORE_0 << BRW_VE1_COMPONENT_3_SHIFT));
}
if (brw->vs.prog_data->uses_vertexid || brw->vs.prog_data->uses_instanceid) {
uint32_t dw0 = 0, dw1 = 0;
uint32_t comp0 = BRW_VE1_COMPONENT_STORE_0;
uint32_t comp1 = BRW_VE1_COMPONENT_STORE_0;
uint32_t comp2 = BRW_VE1_COMPONENT_STORE_0;
uint32_t comp3 = BRW_VE1_COMPONENT_STORE_0;
if (brw->vs.prog_data->uses_vertexid) {
comp0 = BRW_VE1_COMPONENT_STORE_SRC;
comp2 = BRW_VE1_COMPONENT_STORE_VID;
}
if (brw->vs.prog_data->uses_instanceid) {
comp3 = BRW_VE1_COMPONENT_STORE_IID;
}
dw1 = (comp0 << BRW_VE1_COMPONENT_0_SHIFT) |
(comp1 << BRW_VE1_COMPONENT_1_SHIFT) |
(comp2 << BRW_VE1_COMPONENT_2_SHIFT) |
(comp3 << BRW_VE1_COMPONENT_3_SHIFT);
if (brw->gen >= 6) {
dw0 |= GEN6_VE0_VALID |
brw->vb.nr_buffers << GEN6_VE0_INDEX_SHIFT |
BRW_SURFACEFORMAT_R32_UINT << BRW_VE0_FORMAT_SHIFT;
} else {
dw0 |= BRW_VE0_VALID |
brw->vb.nr_buffers << BRW_VE0_INDEX_SHIFT |
BRW_SURFACEFORMAT_R32_UINT << BRW_VE0_FORMAT_SHIFT;
dw1 |= (i * 4) << BRW_VE1_DST_OFFSET_SHIFT;
}
/* Note that for gl_VertexID, gl_InstanceID, and gl_PrimitiveID values,
* the format is ignored and the value is always int.
*/
OUT_BATCH(dw0);
OUT_BATCH(dw1);
}
ADVANCE_BATCH();
}
static void brw_emit_vertices(struct brw_context *brw)
{
struct gl_context *ctx = &brw->intel.ctx;
struct intel_context *intel = intel_context(ctx);
GLuint i;
brw_emit_query_begin(brw);
/* If the VS doesn't read any inputs (calculating vertex position from
* a state variable for some reason, for example), emit a single pad
* VERTEX_ELEMENT struct and bail.
*
* The stale VB state stays in place, but they don't do anything unless
* a VE loads from them.
*/
if (brw->vb.nr_enabled == 0) {
BEGIN_BATCH(3);
OUT_BATCH((_3DSTATE_VERTEX_ELEMENTS << 16) | 1);
if (intel->gen >= 6) {
OUT_BATCH((0 << GEN6_VE0_INDEX_SHIFT) |
GEN6_VE0_VALID |
(BRW_SURFACEFORMAT_R32G32B32A32_FLOAT << BRW_VE0_FORMAT_SHIFT) |
(0 << BRW_VE0_SRC_OFFSET_SHIFT));
} else {
OUT_BATCH((0 << BRW_VE0_INDEX_SHIFT) |
BRW_VE0_VALID |
(BRW_SURFACEFORMAT_R32G32B32A32_FLOAT << BRW_VE0_FORMAT_SHIFT) |
(0 << BRW_VE0_SRC_OFFSET_SHIFT));
}
OUT_BATCH((BRW_VE1_COMPONENT_STORE_0 << BRW_VE1_COMPONENT_0_SHIFT) |
(BRW_VE1_COMPONENT_STORE_0 << BRW_VE1_COMPONENT_1_SHIFT) |
(BRW_VE1_COMPONENT_STORE_0 << BRW_VE1_COMPONENT_2_SHIFT) |
(BRW_VE1_COMPONENT_STORE_1_FLT << BRW_VE1_COMPONENT_3_SHIFT));
CACHED_BATCH();
return;
}
/* Now emit VB and VEP state packets.
*/
if (brw->vb.nr_buffers) {
BEGIN_BATCH(1 + 4*brw->vb.nr_buffers);
OUT_BATCH((_3DSTATE_VERTEX_BUFFERS << 16) | (4*brw->vb.nr_buffers - 1));
for (i = 0; i < brw->vb.nr_buffers; i++) {
struct brw_vertex_buffer *buffer = &brw->vb.buffers[i];
uint32_t dw0;
if (intel->gen >= 6) {
dw0 = GEN6_VB0_ACCESS_VERTEXDATA | (i << GEN6_VB0_INDEX_SHIFT);
} else {
dw0 = BRW_VB0_ACCESS_VERTEXDATA | (i << BRW_VB0_INDEX_SHIFT);
}
if (intel->gen >= 7)
dw0 |= GEN7_VB0_ADDRESS_MODIFYENABLE;
OUT_BATCH(dw0 | (buffer->stride << BRW_VB0_PITCH_SHIFT));
OUT_RELOC(buffer->bo, I915_GEM_DOMAIN_VERTEX, 0, buffer->offset);
if (intel->gen >= 5) {
OUT_RELOC(buffer->bo, I915_GEM_DOMAIN_VERTEX, 0, buffer->bo->size - 1);
} else
OUT_BATCH(0);
OUT_BATCH(0); /* Instance data step rate */
brw->vb.current_buffers[i].handle = buffer->bo->handle;
brw->vb.current_buffers[i].offset = buffer->offset;
brw->vb.current_buffers[i].stride = buffer->stride;
}
brw->vb.nr_current_buffers = i;
ADVANCE_BATCH();
}
BEGIN_BATCH(1 + brw->vb.nr_enabled * 2);
OUT_BATCH((_3DSTATE_VERTEX_ELEMENTS << 16) | (2*brw->vb.nr_enabled - 1));
for (i = 0; i < brw->vb.nr_enabled; i++) {
struct brw_vertex_element *input = brw->vb.enabled[i];
uint32_t format = get_surface_type(input->glarray->Type,
input->glarray->Size,
input->glarray->Format,
input->glarray->Normalized);
uint32_t comp0 = BRW_VE1_COMPONENT_STORE_SRC;
uint32_t comp1 = BRW_VE1_COMPONENT_STORE_SRC;
uint32_t comp2 = BRW_VE1_COMPONENT_STORE_SRC;
uint32_t comp3 = BRW_VE1_COMPONENT_STORE_SRC;
switch (input->glarray->Size) {
case 0: comp0 = BRW_VE1_COMPONENT_STORE_0;
case 1: comp1 = BRW_VE1_COMPONENT_STORE_0;
case 2: comp2 = BRW_VE1_COMPONENT_STORE_0;
case 3: comp3 = BRW_VE1_COMPONENT_STORE_1_FLT;
break;
}
if (intel->gen >= 6) {
OUT_BATCH((input->buffer << GEN6_VE0_INDEX_SHIFT) |
GEN6_VE0_VALID |
(format << BRW_VE0_FORMAT_SHIFT) |
(input->offset << BRW_VE0_SRC_OFFSET_SHIFT));
} else {
OUT_BATCH((input->buffer << BRW_VE0_INDEX_SHIFT) |
BRW_VE0_VALID |
(format << BRW_VE0_FORMAT_SHIFT) |
(input->offset << BRW_VE0_SRC_OFFSET_SHIFT));
}
if (intel->gen >= 5)
OUT_BATCH((comp0 << BRW_VE1_COMPONENT_0_SHIFT) |
(comp1 << BRW_VE1_COMPONENT_1_SHIFT) |
(comp2 << BRW_VE1_COMPONENT_2_SHIFT) |
(comp3 << BRW_VE1_COMPONENT_3_SHIFT));
else
OUT_BATCH((comp0 << BRW_VE1_COMPONENT_0_SHIFT) |
(comp1 << BRW_VE1_COMPONENT_1_SHIFT) |
(comp2 << BRW_VE1_COMPONENT_2_SHIFT) |
(comp3 << BRW_VE1_COMPONENT_3_SHIFT) |
((i * 4) << BRW_VE1_DST_OFFSET_SHIFT));
}
CACHED_BATCH();
}
