Re: [Mesa-dev] Fwd: [PATCH 5/6] st/mesa: implement blit-based ReadPixels
On Mon, 2013-03-18 at 01:06 +0100, Marek Olšák wrote: > > ([rant]We should really have an unmappable portion of VRAM as a > feature, because unmappable memory is cool: the memory can be > reclaimed immediately after texture deletion even if the texture is > busy, because the CPU can't access it, and thanks to that, ttm_bo_wait > could be a no-op for all unmappable resources[/rant]). FWIW, that doesn't require a separate unmappable domain but 'just' that TTM allows more asynchronous GPU operations. Thomas had plans for that a long time ago, don't know what happened to them. -- Earthling Michel Dänzer | http://www.amd.com Libre software enthusiast | Debian, X and DRI developer ___ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev
Re: [Mesa-dev] [PATCH 9/9] tgsi: add ArrayID documentation
Am 17.03.2013 20:59, schrieb Christoph Bumiller: On 17.03.2013 18:04, Christoph Bumiller wrote: On 17.03.2013 16:30, Christian König wrote: Am 15.03.2013 18:58, schrieb Christoph Bumiller: On 15.03.2013 13:08, Christian König wrote: Am 14.03.2013 15:53, schrieb Christoph Bumiller: On 14.03.2013 15:20, Christian König wrote: From: Christian König Signed-off-by: Christian König --- src/gallium/docs/source/tgsi.rst | 16 1 file changed, 16 insertions(+) diff --git a/src/gallium/docs/source/tgsi.rst b/src/gallium/docs/source/tgsi.rst index d9a7fe9..27fe039 100644 --- a/src/gallium/docs/source/tgsi.rst +++ b/src/gallium/docs/source/tgsi.rst @@ -1833,6 +1833,22 @@ If Interpolate flag is set to 1, a Declaration Interpolate token follows. If file is TGSI_FILE_RESOURCE, a Declaration Resource token follows. +If Array flag is set to 1, a Declaration Array token follows. + +Array Declaration + + +Declarations can optional have an ArrayID attribute which can be referred by +indirect addressing operands. An ArrayID of zero is reserved and treaded as +if no ArrayID is specified. + +If an indirect addressing operand refers to an specific declaration by using s/an/a Thx, fixed. +an ArrayID only the registers in this declaration are guaranteed to be +accessed, accessing any register outside this declaration results in undefined +behavior. + Note that the effective index is zero-based and not relative to the specified declaration. XXX: Is it ? Should it be ? Yes for compatibility reasons, otherwise we would need to change all drivers at once. + +If no ArrayID is specified with an indirect addressing operand the whole +register file might be accessed by this operand. + A practice which is strongly discouraged. Don't do this if you have more than 1 declaration for the file in question ! It will prevent packing of scalar/vec2 arrays and effective memory alias analysis. A bit shortened, but in general added the remark. Packing ? Yes ! We can pack arrays if they're declared as e.g. TEMP[0-3].xyzw TEMP[4-31].x And the caches will be very very thankful that we don't just access every 4th element of our 4 times larger than it needs to be buffer !!! And if your card can't do that, pleeease be nice and still make it possible for other drivers. :o3 It is probably possible with the new information to do so, but not priority for me cause I primary need it for our LLVM backend. At some point you'll be able to make use of the info in your backend, too, and then you'll regret having to refamiliarize with this code just because you didn't add the extra (estimated) 2 lines to set the UsageMask. I think you misunderstood me here, you don't need the UsageMask to generate those informations. It is possible by just scanning the shader to figure out which channels are used and which aren't. For temporaries that may be true ... and inputs/outputs are always vec4 sized to guarantee linkage, packing for GENERIC ones is handled at the mesa level. Additional to that I'm not convinced that using the UsageMask for this is 100% correct, to me it looks more like UsageMask is something we need for outputs to distinct between not writing to an output channel (and so still having the default) and not having an output channel at all. Actually, for gl_ClipDistance[] we use the UsageMask to specify if the clip distance was declared in the source (and thus should be enabled) instead of whether it's been written or not. I wanted to be able to distinguish between float gl_ClipDistance[8] or vec4 mesa_ClipDistance[2] with the UsageMask but I guess OUT[0..1].x, CLIPDIST might just as well mean that gl_ClipDistance[0 and 4] are being used ... Hm, we'll need a cap for that anway to tell st if it should lower ClipDistance to vec4s or not, and just assume that TGSI corresponds to what the cap says. And since this is the only case for IN/OUT where the driver's backend has to decide whether to pack or not ... ok, I'll just infer array width myself, too, and you can ignore the UsageMask. Also, NAK from me until array access/declarations for the other files follows suit. Sorry for being so ... pesky, but I'd really like this change to be 100% complete. Come on, doesn't it nag on your conscience if this is left to remain only a few smalls steps from perfection ? Declaring and accessing arrays for inputs/outputs are not so much of a problem, figuring out how to get this information to glsl_to_tgsi is the real problem. For temporaries changing the glsl_to_tgsi pass is pretty much sufficient, but for inputs and outputs you need to dig into the mesa state tracker, and I definitely don't intend to do so. Fine, then I'll have a look at that myself. * flaming eyes * Ok, had a look, had enough. It hurts. At least if you have never touched glsl-to-tgsi and go about it expecting it to be easy and straightforward. HURTS! Yes, that was also my initial reaction, and is also the
Re: [Mesa-dev] [PATCH 4/4] st/dri: don't create 24-bit depth, non-stencil visuals
On Thu, Mar 14, 2013 at 3:23 PM, Brian Paul wrote: > On 03/14/2013 08:18 AM, Marek Olšák wrote: >> >> Z24X8 actually does save memory, because Radeon HD 5000 and later >> cards don't support combined depth stencil formats, therefore Z24S8 is >> allocated internally as two separate buffers Z24X8 and S8. > > > Ah, OK. I'll drop this patch then. Thanks. > > However, I think we have some redundant visual/FBconfigs. I'm still > interested in trying to prune them down... Some configs might seem redundant, but glxinfo doesn't print all config properties. One such property is back_buffer_modes defined in dri_screen.c:99. Marek ___ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev
[Mesa-dev] [PATCH 1/9] tgsi/ureg: cleanup local temporary emission v2
From: Christian König
Instead of emitting each temporary separately, emit them in a chunk.
v2: keep separate function for emitting temps
Signed-off-by: Christian König
---
src/gallium/auxiliary/tgsi/tgsi_ureg.c | 30 +-
1 file changed, 13 insertions(+), 17 deletions(-)
diff --git a/src/gallium/auxiliary/tgsi/tgsi_ureg.c
b/src/gallium/auxiliary/tgsi/tgsi_ureg.c
index 3c2a923..72657ae 100644
--- a/src/gallium/auxiliary/tgsi/tgsi_ureg.c
+++ b/src/gallium/auxiliary/tgsi/tgsi_ureg.c
@@ -1260,24 +1260,23 @@ emit_decl_fs(struct ureg_program *ureg,
out[3].decl_semantic.Index = semantic_index;
}
-
-static void emit_decl( struct ureg_program *ureg,
- unsigned file,
- unsigned index,
- boolean local )
+static void
+emit_decl_temps( struct ureg_program *ureg,
+ unsigned first, unsigned last,
+ boolean local )
{
union tgsi_any_token *out = get_tokens( ureg, DOMAIN_DECL, 2 );
out[0].value = 0;
out[0].decl.Type = TGSI_TOKEN_TYPE_DECLARATION;
out[0].decl.NrTokens = 2;
- out[0].decl.File = file;
+ out[0].decl.File = TGSI_FILE_TEMPORARY;
out[0].decl.UsageMask = TGSI_WRITEMASK_XYZW;
out[0].decl.Local = local;
out[1].value = 0;
- out[1].decl_range.First = index;
- out[1].decl_range.Last = index;
+ out[1].decl_range.First = first;
+ out[1].decl_range.Last = last;
}
static void emit_decl_range( struct ureg_program *ureg,
@@ -1535,16 +1534,13 @@ static void emit_decls( struct ureg_program *ureg )
}
if (ureg->nr_temps) {
- if (util_bitmask_get_first_index(ureg->local_temps) ==
UTIL_BITMASK_INVALID_INDEX) {
- emit_decl_range( ureg,
- TGSI_FILE_TEMPORARY,
- 0, ureg->nr_temps );
+ for (i = 0; i < ureg->nr_temps;) {
+ boolean local = util_bitmask_get(ureg->local_temps, i);
+ unsigned first = i++;
+ while (i < ureg->nr_temps && local ==
util_bitmask_get(ureg->local_temps, i))
+++i;
- } else {
- for (i = 0; i < ureg->nr_temps; i++) {
-emit_decl( ureg, TGSI_FILE_TEMPORARY, i,
- util_bitmask_get(ureg->local_temps, i) );
- }
+ emit_decl_temps( ureg, first, i - 1, local );
}
}
--
1.7.9.5
___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/mesa-dev
[Mesa-dev] [PATCH 2/9] tgsi/ureg: implement support for array temporaries
From: Christian König
Don't bother with free temporaries, just allocate them at
the end and also emit them in their own declaration.
Signed-off-by: Christian König
---
src/gallium/auxiliary/tgsi/tgsi_ureg.c | 55
src/gallium/auxiliary/tgsi/tgsi_ureg.h | 38 +++---
2 files changed, 69 insertions(+), 24 deletions(-)
diff --git a/src/gallium/auxiliary/tgsi/tgsi_ureg.c
b/src/gallium/auxiliary/tgsi/tgsi_ureg.c
index 72657ae..88acdcb 100644
--- a/src/gallium/auxiliary/tgsi/tgsi_ureg.c
+++ b/src/gallium/auxiliary/tgsi/tgsi_ureg.c
@@ -153,6 +153,7 @@ struct ureg_program
struct util_bitmask *free_temps;
struct util_bitmask *local_temps;
+ struct util_bitmask *decl_temps;
unsigned nr_temps;
struct const_decl const_decls;
@@ -547,13 +548,18 @@ static struct ureg_dst alloc_temporary( struct
ureg_program *ureg,
/* Or allocate a new one.
*/
- if (i == UTIL_BITMASK_INVALID_INDEX)
+ if (i == UTIL_BITMASK_INVALID_INDEX) {
i = ureg->nr_temps++;
- util_bitmask_clear(ureg->free_temps, i);
+ if (local)
+ util_bitmask_set(ureg->local_temps, i);
- if (local)
- util_bitmask_set(ureg->local_temps, i);
+ /* Start a new declaration when the local flag changes */
+ if (!i || util_bitmask_get(ureg->local_temps, i - 1) != local)
+ util_bitmask_set(ureg->decl_temps, i);
+ }
+
+ util_bitmask_clear(ureg->free_temps, i);
return ureg_dst_register( TGSI_FILE_TEMPORARY, i );
}
@@ -568,6 +574,24 @@ struct ureg_dst ureg_DECL_local_temporary( struct
ureg_program *ureg )
return alloc_temporary(ureg, TRUE);
}
+struct ureg_dst ureg_DECL_array_temporary( struct ureg_program *ureg,
+ unsigned size,
+ boolean local )
+{
+ unsigned i = ureg->nr_temps;
+ struct ureg_dst dst = ureg_dst_register( TGSI_FILE_TEMPORARY, i );
+
+ if (local)
+ util_bitmask_set(ureg->local_temps, i);
+
+ util_bitmask_set(ureg->decl_temps, i);
+
+ ureg->nr_temps += size;
+ util_bitmask_set(ureg->decl_temps, ureg->nr_temps);
+
+ return dst;
+}
+
void ureg_release_temporary( struct ureg_program *ureg,
struct ureg_dst tmp )
{
@@ -856,11 +880,11 @@ ureg_emit_src( struct ureg_program *ureg,
}
if (src.Dimension) {
+ out[0].src.Dimension = 1;
+ out[n].dim.Dimension = 0;
+ out[n].dim.Padding = 0;
if (src.DimIndirect) {
- out[0].src.Dimension = 1;
out[n].dim.Indirect = 1;
- out[n].dim.Dimension = 0;
- out[n].dim.Padding = 0;
out[n].dim.Index = src.DimensionIndex;
n++;
out[n].value = 0;
@@ -871,10 +895,7 @@ ureg_emit_src( struct ureg_program *ureg,
out[n].src.SwizzleW = src.DimIndSwizzle;
out[n].src.Index = src.DimIndIndex;
} else {
- out[0].src.Dimension = 1;
out[n].dim.Indirect = 0;
- out[n].dim.Dimension = 0;
- out[n].dim.Padding = 0;
out[n].dim.Index = src.DimensionIndex;
}
n++;
@@ -1536,9 +1557,10 @@ static void emit_decls( struct ureg_program *ureg )
if (ureg->nr_temps) {
for (i = 0; i < ureg->nr_temps;) {
boolean local = util_bitmask_get(ureg->local_temps, i);
- unsigned first = i++;
- while (i < ureg->nr_temps && local ==
util_bitmask_get(ureg->local_temps, i))
-++i;
+ unsigned first = i;
+ i = util_bitmask_get_next_index(ureg->decl_temps, i + 1);
+ if (i == UTIL_BITMASK_INVALID_INDEX)
+i = ureg->nr_temps;
emit_decl_temps( ureg, first, i - 1, local );
}
@@ -1707,8 +1729,14 @@ struct ureg_program *ureg_create( unsigned processor )
if (ureg->local_temps == NULL)
goto no_local_temps;
+ ureg->decl_temps = util_bitmask_create();
+ if (ureg->decl_temps == NULL)
+ goto no_decl_temps;
+
return ureg;
+no_decl_temps:
+ util_bitmask_destroy(ureg->local_temps);
no_local_temps:
util_bitmask_destroy(ureg->free_temps);
no_free_temps:
@@ -1730,6 +1758,7 @@ void ureg_destroy( struct ureg_program *ureg )
util_bitmask_destroy(ureg->free_temps);
util_bitmask_destroy(ureg->local_temps);
+ util_bitmask_destroy(ureg->decl_temps);
FREE(ureg);
}
diff --git a/src/gallium/auxiliary/tgsi/tgsi_ureg.h
b/src/gallium/auxiliary/tgsi/tgsi_ureg.h
index fb663e9..cd140de 100644
--- a/src/gallium/auxiliary/tgsi/tgsi_ureg.h
+++ b/src/gallium/auxiliary/tgsi/tgsi_ureg.h
@@ -71,17 +71,17 @@ struct ureg_src
*/
struct ureg_dst
{
- unsigned File: 4; /* TGSI_FILE_ */
- unsigned WriteMask : 4; /* TGSI_WRITEMASK_ */
- unsigned Indirect: 1; /* BOOL */
- unsigned Saturate: 1; /* BOOL */
- unsigned Predicate : 1;
- unsigned PredNegate : 1; /* BOOL */
- unsigned PredSwizzleX: 2; /* TGSI_SWIZZLE_ */
- unsigned PredSwizzleY: 2; /* TGSI_SWIZZLE_ */
- un
[Mesa-dev] [PATCH 3/9] glsl_to_tgsi: use get_temp for all allocations
From: Christian König
Signed-off-by: Christian König
---
src/mesa/state_tracker/st_glsl_to_tgsi.cpp | 23 ++-
1 file changed, 10 insertions(+), 13 deletions(-)
diff --git a/src/mesa/state_tracker/st_glsl_to_tgsi.cpp
b/src/mesa/state_tracker/st_glsl_to_tgsi.cpp
index 0cef092..7562d11 100644
--- a/src/mesa/state_tracker/st_glsl_to_tgsi.cpp
+++ b/src/mesa/state_tracker/st_glsl_to_tgsi.cpp
@@ -1078,13 +1078,11 @@ glsl_to_tgsi_visitor::visit(ir_variable *ir)
*/
assert((int) ir->num_state_slots == type_size(ir->type));
- storage = new(mem_ctx) variable_storage(ir, PROGRAM_TEMPORARY,
-this->next_temp);
- this->variables.push_tail(storage);
- this->next_temp += type_size(ir->type);
+ dst = st_dst_reg(get_temp(ir->type));
+
+ storage = new(mem_ctx) variable_storage(ir, dst.file, dst.index);
- dst = st_dst_reg(st_src_reg(PROGRAM_TEMPORARY, storage->index,
- native_integers ? ir->type->base_type : GLSL_TYPE_FLOAT));
+ this->variables.push_tail(storage);
}
@@ -2052,11 +2050,11 @@ glsl_to_tgsi_visitor::visit(ir_dereference_variable *ir)
break;
case ir_var_auto:
case ir_var_temporary:
- entry = new(mem_ctx) variable_storage(var, PROGRAM_TEMPORARY,
- this->next_temp);
+ st_src_reg src = get_temp(var->type);
+
+ entry = new(mem_ctx) variable_storage(var, src.file, src.index);
this->variables.push_tail(entry);
- next_temp += type_size(var->type);
break;
}
@@ -2574,11 +2572,10 @@
glsl_to_tgsi_visitor::get_function_signature(ir_function_signature *sig)
storage = find_variable_storage(param);
assert(!storage);
- storage = new(mem_ctx) variable_storage(param, PROGRAM_TEMPORARY,
- this->next_temp);
- this->variables.push_tail(storage);
+ st_src_reg src = get_temp(param->type);
- this->next_temp += type_size(param->type);
+ storage = new(mem_ctx) variable_storage(param, src.file, src.index);
+ this->variables.push_tail(storage);
}
if (!sig->return_type->is_void()) {
--
1.7.9.5
___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/mesa-dev
[Mesa-dev] [PATCH 4/9] glsl_to_tgsi: allocate arrays separately v2
From: Christian König
Instead of allocating everything as temporaries, use the
new array allocation functions.
v2: fix bug in simplify_cmp, declare arrays on demand
Signed-off-by: Christian König
---
src/mesa/main/mtypes.h |1 +
src/mesa/state_tracker/st_glsl_to_tgsi.cpp | 89 ++--
2 files changed, 59 insertions(+), 31 deletions(-)
diff --git a/src/mesa/main/mtypes.h b/src/mesa/main/mtypes.h
index 83b6c89..8c38aa7 100644
--- a/src/mesa/main/mtypes.h
+++ b/src/mesa/main/mtypes.h
@@ -1796,6 +1796,7 @@ struct gl_transform_feedback_state
typedef enum
{
PROGRAM_TEMPORARY, /**< machine->Temporary[] */
+ PROGRAM_ARRAY, /**< Arrays & Matrixes */
PROGRAM_INPUT, /**< machine->Inputs[] */
PROGRAM_OUTPUT, /**< machine->Outputs[] */
PROGRAM_LOCAL_PARAM, /**< gl_program->LocalParams[] */
diff --git a/src/mesa/state_tracker/st_glsl_to_tgsi.cpp
b/src/mesa/state_tracker/st_glsl_to_tgsi.cpp
index 7562d11..2f4a5a3 100644
--- a/src/mesa/state_tracker/st_glsl_to_tgsi.cpp
+++ b/src/mesa/state_tracker/st_glsl_to_tgsi.cpp
@@ -85,6 +85,11 @@ extern "C" {
*/
#define MAX_TEMPS 4096
+/**
+ * Maximum number of arrays
+ */
+#define MAX_ARRAYS256
+
/* will be 4 for GLSL 4.00 */
#define MAX_GLSL_TEXTURE_OFFSET 1
@@ -315,6 +320,9 @@ public:
int next_temp;
+ unsigned array_sizes[MAX_ARRAYS];
+ unsigned next_array;
+
int num_address_regs;
int samplers_used;
bool indirect_addr_temps;
@@ -550,6 +558,7 @@ glsl_to_tgsi_visitor::emit(ir_instruction *ir, unsigned op,
if (dst.reladdr) {
switch(dst.file) {
case PROGRAM_TEMPORARY:
+ case PROGRAM_ARRAY:
this->indirect_addr_temps = true;
break;
case PROGRAM_LOCAL_PARAM:
@@ -571,6 +580,7 @@ glsl_to_tgsi_visitor::emit(ir_instruction *ir, unsigned op,
if(inst->src[i].reladdr) {
switch(inst->src[i].file) {
case PROGRAM_TEMPORARY:
+case PROGRAM_ARRAY:
this->indirect_addr_temps = true;
break;
case PROGRAM_LOCAL_PARAM:
@@ -1005,17 +1015,26 @@ glsl_to_tgsi_visitor::get_temp(const glsl_type *type)
st_src_reg src;
src.type = native_integers ? type->base_type : GLSL_TYPE_FLOAT;
- src.file = PROGRAM_TEMPORARY;
- src.index = next_temp;
src.reladdr = NULL;
- next_temp += type_size(type);
+ src.negate = 0;
+
+ if (type->is_array() || type->is_matrix()) {
+ src.file = PROGRAM_ARRAY;
+ src.index = next_array << 16 | 0x8000;
+ array_sizes[next_array] = type_size(type);
+ ++next_array;
+
+ } else {
+ src.file = PROGRAM_TEMPORARY;
+ src.index = next_temp;
+ next_temp += type_size(type);
+ }
if (type->is_array() || type->is_record()) {
src.swizzle = SWIZZLE_NOOP;
} else {
src.swizzle = swizzle_for_size(type->vector_elements);
}
- src.negate = 0;
return src;
}
@@ -2975,6 +2994,7 @@ glsl_to_tgsi_visitor::glsl_to_tgsi_visitor()
{
result.file = PROGRAM_UNDEFINED;
next_temp = 1;
+ next_array = 0;
next_signature_id = 1;
num_immediates = 0;
current_function = NULL;
@@ -3180,7 +3200,8 @@ glsl_to_tgsi_visitor::simplify_cmp(void)
assert(inst->dst.index < MAX_TEMPS);
prevWriteMask = tempWrites[inst->dst.index];
tempWrites[inst->dst.index] |= inst->dst.writemask;
- }
+ } else
+ break;
/* For a CMP to be considered a conditional write, the destination
* register and source register two must be the same. */
@@ -4011,6 +4032,7 @@ struct st_translate {
struct ureg_program *ureg;
struct ureg_dst temps[MAX_TEMPS];
+ struct ureg_dst arrays[MAX_ARRAYS];
struct ureg_src *constants;
struct ureg_src *immediates;
struct ureg_dst outputs[PIPE_MAX_SHADER_OUTPUTS];
@@ -4019,6 +4041,8 @@ struct st_translate {
struct ureg_src samplers[PIPE_MAX_SAMPLERS];
struct ureg_src systemValues[SYSTEM_VALUE_MAX];
+ unsigned array_sizes[MAX_ARRAYS];
+
const GLuint *inputMapping;
const GLuint *outputMapping;
@@ -4129,16 +4153,34 @@ dst_register(struct st_translate *t,
gl_register_file file,
GLuint index)
{
+ unsigned array;
+
switch(file) {
case PROGRAM_UNDEFINED:
return ureg_dst_undef();
case PROGRAM_TEMPORARY:
+ assert(index >= 0);
+ assert(index < (int) Elements(t->temps));
+
if (ureg_dst_is_undef(t->temps[index]))
t->temps[index] = ureg_DECL_local_temporary(t->ureg);
return t->temps[index];
+ case PROGRAM_ARRAY:
+ array = index >> 16;
+
+ assert(array >= 0);
+ assert(array < (int) Elements(t->arrays));
+
+ if (ureg_dst_is_undef(t->arrays[array]))
+ t->arrays[array] = ureg_DECL_array_temporary(
+t->ureg, t->array_sizes[array], TRUE);
+
+ return ureg_dst_array_offset(t->arrays[array],
+
[Mesa-dev] [PATCH 5/9] glsl_to_tgsi: remove indirect addressing limitations
From: Christian König
They shouldn't be necessary any more.
Signed-off-by: Christian König
---
src/mesa/state_tracker/st_glsl_to_tgsi.cpp | 36 +++-
1 file changed, 3 insertions(+), 33 deletions(-)
diff --git a/src/mesa/state_tracker/st_glsl_to_tgsi.cpp
b/src/mesa/state_tracker/st_glsl_to_tgsi.cpp
index 2f4a5a3..e3718ee 100644
--- a/src/mesa/state_tracker/st_glsl_to_tgsi.cpp
+++ b/src/mesa/state_tracker/st_glsl_to_tgsi.cpp
@@ -325,7 +325,6 @@ public:
int num_address_regs;
int samplers_used;
- bool indirect_addr_temps;
bool indirect_addr_consts;
int glsl_version;
@@ -557,10 +556,6 @@ glsl_to_tgsi_visitor::emit(ir_instruction *ir, unsigned op,
/* Update indirect addressing status used by TGSI */
if (dst.reladdr) {
switch(dst.file) {
- case PROGRAM_TEMPORARY:
- case PROGRAM_ARRAY:
- this->indirect_addr_temps = true;
- break;
case PROGRAM_LOCAL_PARAM:
case PROGRAM_ENV_PARAM:
case PROGRAM_STATE_VAR:
@@ -579,10 +574,6 @@ glsl_to_tgsi_visitor::emit(ir_instruction *ir, unsigned op,
for (i=0; i<3; i++) {
if(inst->src[i].reladdr) {
switch(inst->src[i].file) {
-case PROGRAM_TEMPORARY:
-case PROGRAM_ARRAY:
- this->indirect_addr_temps = true;
- break;
case PROGRAM_LOCAL_PARAM:
case PROGRAM_ENV_PARAM:
case PROGRAM_STATE_VAR:
@@ -3000,7 +2991,6 @@ glsl_to_tgsi_visitor::glsl_to_tgsi_visitor()
current_function = NULL;
num_address_regs = 0;
samplers_used = 0;
- indirect_addr_temps = false;
indirect_addr_consts = false;
glsl_version = 0;
native_integers = false;
@@ -3839,7 +3829,6 @@ get_pixel_transfer_visitor(struct st_fragment_program *fp,
v->next_temp = original->next_temp;
v->num_address_regs = original->num_address_regs;
v->samplers_used = prog->SamplersUsed = original->samplers_used;
- v->indirect_addr_temps = original->indirect_addr_temps;
v->indirect_addr_consts = original->indirect_addr_consts;
memcpy(&v->immediates, &original->immediates, sizeof(v->immediates));
v->num_immediates = original->num_immediates;
@@ -3970,7 +3959,6 @@ get_bitmap_visitor(struct st_fragment_program *fp,
v->next_temp = original->next_temp;
v->num_address_regs = original->num_address_regs;
v->samplers_used = prog->SamplersUsed = original->samplers_used;
- v->indirect_addr_temps = original->indirect_addr_temps;
v->indirect_addr_consts = original->indirect_addr_consts;
memcpy(&v->immediates, &original->immediates, sizeof(v->immediates));
v->num_immediates = original->num_immediates;
@@ -4840,17 +4828,6 @@ st_translate_program(
}
}
- if (program->indirect_addr_temps) {
- /* If temps are accessed with indirect addressing, declare temporaries
- * in sequential order. Else, we declare them on demand elsewhere.
- * (Note: the number of temporaries is equal to program->next_temp)
- */
- for (i = 0; i < (unsigned)program->next_temp; i++) {
- /* XXX use TGSI_FILE_TEMPORARY_ARRAY when it's supported by ureg */
- t->temps[i] = ureg_DECL_local_temporary(t->ureg);
- }
- }
-
/* Copy over array sizes
*/
memcpy(t->array_sizes, program->array_sizes, sizeof(unsigned) *
program->next_array);
@@ -5088,16 +5065,9 @@ get_mesa_program(struct gl_context *ctx,
v->copy_propagate();
while (v->eliminate_dead_code_advanced());
- /* FIXME: These passes to optimize temporary registers don't work when there
-* is indirect addressing of the temporary register space. We need proper
-* array support so that we don't have to give up these passes in every
-* shader that uses arrays.
-*/
- if (!v->indirect_addr_temps) {
- v->eliminate_dead_code();
- v->merge_registers();
- v->renumber_registers();
- }
+ v->eliminate_dead_code();
+ v->merge_registers();
+ v->renumber_registers();
/* Write the END instruction. */
v->emit(NULL, TGSI_OPCODE_END);
--
1.7.9.5
___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/mesa-dev
[Mesa-dev] [PATCH 6/9] tgsi: remove TGSI_FILE_(IMMEDIATE|TEMP)_ARRAY
From: Christian König
Nobody seems to be using it, and only nv50 had a partial implementation.
Signed-off-by: Christian König
---
src/gallium/auxiliary/tgsi/tgsi_build.c| 19 -
src/gallium/auxiliary/tgsi/tgsi_dump.c | 38 --
src/gallium/auxiliary/tgsi/tgsi_exec.c | 41 --
src/gallium/auxiliary/tgsi/tgsi_exec.h |2 -
src/gallium/auxiliary/tgsi/tgsi_parse.c| 11 ---
src/gallium/auxiliary/tgsi/tgsi_parse.h|6 --
src/gallium/auxiliary/tgsi/tgsi_sanity.c |2 -
src/gallium/auxiliary/tgsi/tgsi_strings.c |2 -
src/gallium/auxiliary/tgsi/tgsi_text.c | 43 ---
.../drivers/nv50/codegen/nv50_ir_from_tgsi.cpp | 80
src/gallium/include/pipe/p_shader_tokens.h |6 +-
11 files changed, 2 insertions(+), 248 deletions(-)
diff --git a/src/gallium/auxiliary/tgsi/tgsi_build.c
b/src/gallium/auxiliary/tgsi/tgsi_build.c
index cb7b9b2..33cbbd8 100644
--- a/src/gallium/auxiliary/tgsi/tgsi_build.c
+++ b/src/gallium/auxiliary/tgsi/tgsi_build.c
@@ -336,7 +336,6 @@ tgsi_default_full_declaration( void )
full_declaration.Range = tgsi_default_declaration_range();
full_declaration.Semantic = tgsi_default_declaration_semantic();
full_declaration.Interp = tgsi_default_declaration_interp();
- full_declaration.ImmediateData.u = NULL;
full_declaration.Resource = tgsi_default_declaration_resource();
full_declaration.SamplerView = tgsi_default_declaration_sampler_view();
@@ -425,24 +424,6 @@ tgsi_build_full_declaration(
header );
}
- if (full_decl->Declaration.File == TGSI_FILE_IMMEDIATE_ARRAY) {
- unsigned i, j;
- union tgsi_immediate_data *data;
-
- for (i = 0; i <= dr->Last; ++i) {
- for (j = 0; j < 4; ++j) {
-unsigned idx = i*4 + j;
-if (maxsize <= size)
- return 0;
-data = (union tgsi_immediate_data *) &tokens[size];
-++size;
-
-*data = full_decl->ImmediateData.u[idx];
-declaration_grow( declaration, header );
- }
- }
- }
-
if (full_decl->Declaration.File == TGSI_FILE_RESOURCE) {
struct tgsi_declaration_resource *dr;
diff --git a/src/gallium/auxiliary/tgsi/tgsi_dump.c
b/src/gallium/auxiliary/tgsi/tgsi_dump.c
index 3e6f76a..177be0f 100644
--- a/src/gallium/auxiliary/tgsi/tgsi_dump.c
+++ b/src/gallium/auxiliary/tgsi/tgsi_dump.c
@@ -347,44 +347,6 @@ iter_declaration(
TXT( ", INVARIANT" );
}
-
- if (decl->Declaration.File == TGSI_FILE_IMMEDIATE_ARRAY) {
- unsigned i;
- char range_indent[4];
-
- TXT(" {");
-
- if (decl->Range.Last < 10)
- range_indent[0] = '\0';
- else if (decl->Range.Last < 100) {
- range_indent[0] = ' ';
- range_indent[1] = '\0';
- } else if (decl->Range.Last < 1000) {
- range_indent[0] = ' ';
- range_indent[1] = ' ';
- range_indent[2] = '\0';
- } else {
- range_indent[0] = ' ';
- range_indent[1] = ' ';
- range_indent[2] = ' ';
- range_indent[3] = '\0';
- }
-
- dump_imm_data(iter, decl->ImmediateData.u,
-4, TGSI_IMM_FLOAT32);
- for(i = 1; i <= decl->Range.Last; ++i) {
- /* indent by strlen of:
- * "DCL IMMX[0..1] {" */
- CHR('\n');
- TXT( "" );
- TXT( range_indent );
- dump_imm_data(iter, decl->ImmediateData.u + i,
- 4, TGSI_IMM_FLOAT32);
- }
-
- TXT(" }");
- }
-
EOL();
return TRUE;
diff --git a/src/gallium/auxiliary/tgsi/tgsi_exec.c
b/src/gallium/auxiliary/tgsi/tgsi_exec.c
index 3df3ac3..1af2395 100644
--- a/src/gallium/auxiliary/tgsi/tgsi_exec.c
+++ b/src/gallium/auxiliary/tgsi/tgsi_exec.c
@@ -749,19 +749,6 @@ tgsi_exec_machine_bind_shader(
++mach->NumOutputs;
}
}
- if (parse.FullToken.FullDeclaration.Declaration.File ==
- TGSI_FILE_IMMEDIATE_ARRAY) {
-unsigned reg;
-struct tgsi_full_declaration *decl =
- &parse.FullToken.FullDeclaration;
-debug_assert(decl->Range.Last < TGSI_EXEC_NUM_IMMEDIATES);
-for (reg = decl->Range.First; reg <= decl->Range.Last; ++reg) {
- for( i = 0; i < 4; i++ ) {
- int idx = reg * 4 + i;
- mach->ImmArray[reg][i] = decl->ImmediateData.u[idx].Float;
- }
-}
- }
memcpy(declarations + numDeclarations,
&parse.FullToken.FullDeclaration,
sizeof(declarations[0]));
@@ -1116,16 +1103,6 @@ fetch_src_file_channel(const struct tgsi_exec_machine
*mach,
}
break;
- case TGSI_FILE_TEMPORARY_ARRAY:
- for (i = 0; i < TGSI_QUAD_SIZE; i++) {
- assert(index->i[i] < TGSI_EXEC_NUM_TEMPS);
- asser
[Mesa-dev] [PATCH 7/9] tgsi: add ArrayID to declarations
From: Christian König
Remember which declarations are declared as "arrays" and so
can be indirectly addressed. ArrayIDs start at 1, cause for
compatibility reasons zero is treaded as no array present.
Signed-off-by: Christian König
---
src/gallium/auxiliary/tgsi/tgsi_dump.c |6 ++
src/gallium/auxiliary/tgsi/tgsi_parse.c|4
src/gallium/auxiliary/tgsi/tgsi_parse.h|1 +
src/gallium/auxiliary/tgsi/tgsi_text.c | 22 +
src/gallium/auxiliary/tgsi/tgsi_ureg.c | 29 +---
src/gallium/include/pipe/p_shader_tokens.h |8 +++-
6 files changed, 66 insertions(+), 4 deletions(-)
diff --git a/src/gallium/auxiliary/tgsi/tgsi_dump.c
b/src/gallium/auxiliary/tgsi/tgsi_dump.c
index 177be0f..adca6af 100644
--- a/src/gallium/auxiliary/tgsi/tgsi_dump.c
+++ b/src/gallium/auxiliary/tgsi/tgsi_dump.c
@@ -272,6 +272,12 @@ iter_declaration(
ctx,
decl->Declaration.UsageMask );
+ if (decl->Declaration.Array) {
+ TXT( ", ARRAY(" );
+ SID(decl->Array.ArrayID);
+ CHR(')');
+ }
+
if (decl->Declaration.Local)
TXT( ", LOCAL" );
diff --git a/src/gallium/auxiliary/tgsi/tgsi_parse.c
b/src/gallium/auxiliary/tgsi/tgsi_parse.c
index 720d68d..29079ef 100644
--- a/src/gallium/auxiliary/tgsi/tgsi_parse.c
+++ b/src/gallium/auxiliary/tgsi/tgsi_parse.c
@@ -129,6 +129,10 @@ tgsi_parse_token(
next_token(ctx, &decl->SamplerView);
}
+ if( decl->Declaration.Array ) {
+ next_token(ctx, &decl->Array);
+ }
+
break;
}
diff --git a/src/gallium/auxiliary/tgsi/tgsi_parse.h
b/src/gallium/auxiliary/tgsi/tgsi_parse.h
index 78210ed..ae40f13 100644
--- a/src/gallium/auxiliary/tgsi/tgsi_parse.h
+++ b/src/gallium/auxiliary/tgsi/tgsi_parse.h
@@ -66,6 +66,7 @@ struct tgsi_full_declaration
struct tgsi_declaration_semantic Semantic;
struct tgsi_declaration_resource Resource;
struct tgsi_declaration_sampler_view SamplerView;
+ struct tgsi_declaration_array Array;
};
struct tgsi_full_immediate
diff --git a/src/gallium/auxiliary/tgsi/tgsi_text.c
b/src/gallium/auxiliary/tgsi/tgsi_text.c
index 7d580e6..8a2a760 100644
--- a/src/gallium/auxiliary/tgsi/tgsi_text.c
+++ b/src/gallium/auxiliary/tgsi/tgsi_text.c
@@ -1114,6 +1114,28 @@ static boolean parse_declaration( struct translate_ctx
*ctx )
cur = ctx->cur;
eat_opt_white( &cur );
+ if (*cur == ',') {
+ cur2 = cur;
+ cur2++;
+ eat_opt_white( &cur2 );
+ if (str_match_nocase_whole( &cur2, "ARRAY(" )) {
+ int arrayid;
+ eat_opt_white( &cur2 );
+ if (!parse_int( &cur2, &arrayid )) {
+report_error( ctx, "Expected `,'" );
+return FALSE;
+ }
+ eat_opt_white( &cur2 );
+ if (*cur2 != ')') {
+report_error( ctx, "Expected `,'" );
+return FALSE;
+ }
+ decl.Declaration.Array = 1;
+ decl.Array.ArrayID = arrayid;
+ cur = cur2;
+ }
+ }
+
if (*cur == ',' && !is_vs_input) {
uint i, j;
diff --git a/src/gallium/auxiliary/tgsi/tgsi_ureg.c
b/src/gallium/auxiliary/tgsi/tgsi_ureg.c
index 88acdcb..1e862cb 100644
--- a/src/gallium/auxiliary/tgsi/tgsi_ureg.c
+++ b/src/gallium/auxiliary/tgsi/tgsi_ureg.c
@@ -50,6 +50,7 @@ union tgsi_any_token {
struct tgsi_declaration_interp decl_interp;
struct tgsi_declaration_semantic decl_semantic;
struct tgsi_declaration_sampler_view decl_sampler_view;
+ struct tgsi_declaration_array array;
struct tgsi_immediate imm;
union tgsi_immediate_data imm_data;
struct tgsi_instruction insn;
@@ -78,6 +79,7 @@ struct ureg_tokens {
#define UREG_MAX_IMMEDIATE 256
#define UREG_MAX_ADDR 2
#define UREG_MAX_PRED 1
+#define UREG_MAX_ARRAY_TEMPS 256
struct const_decl {
struct {
@@ -156,6 +158,9 @@ struct ureg_program
struct util_bitmask *decl_temps;
unsigned nr_temps;
+ unsigned array_temps[UREG_MAX_ARRAY_TEMPS];
+ unsigned nr_array_temps;
+
struct const_decl const_decls;
struct const_decl const_decls2D[PIPE_MAX_CONSTANT_BUFFERS];
@@ -584,11 +589,17 @@ struct ureg_dst ureg_DECL_array_temporary( struct
ureg_program *ureg,
if (local)
util_bitmask_set(ureg->local_temps, i);
+ /* Always start a new declaration at the start */
util_bitmask_set(ureg->decl_temps, i);
ureg->nr_temps += size;
+
+ /* and also at the end of the array */
util_bitmask_set(ureg->decl_temps, ureg->nr_temps);
+ if (ureg->nr_array_temps < UREG_MAX_ARRAY_TEMPS)
+ ureg->array_temps[ureg->nr_array_temps++] = i;
+
return dst;
}
@@ -1284,9 +1295,11 @@ emit_decl_fs(struct ureg_program *ureg,
static void
emit_decl_temps( struct ureg_program *ureg,
unsigned first, unsigned last,
- boolean local )
+ boolean local,
+ unsigned arrayid )
{
- union tgsi_any_token *out = get_tokens( ureg, DOMAIN_DECL, 2 );
+ union tgsi_
[Mesa-dev] [PATCH 8/9] tgsi: use separate structure for indirect address v2
From: Christian König
To further improve the optimization of source and destination
indirect addressing we need the ability to store a reference
to the declaration of the addressed operands.
Since most of the fields in tgsi_src_register doesn't apply for
an indirect addressing operand replace it with a separate
tgsi_ind_register structure and so make room for extra information.
v2: rename Declaration to ArrayID, put the ArrayID into () instead of []
Signed-off-by: Christian König
---
src/gallium/auxiliary/gallivm/lp_bld_tgsi_soa.c|4 +-
src/gallium/auxiliary/tgsi/tgsi_build.c| 109 +++-
src/gallium/auxiliary/tgsi/tgsi_dump.c | 28 -
src/gallium/auxiliary/tgsi/tgsi_exec.c |8 +-
src/gallium/auxiliary/tgsi/tgsi_parse.c| 35 +--
src/gallium/auxiliary/tgsi/tgsi_parse.h|8 +-
src/gallium/auxiliary/tgsi/tgsi_text.c | 27 +++--
src/gallium/auxiliary/tgsi/tgsi_ureg.c | 36 +++
src/gallium/auxiliary/tgsi/tgsi_ureg.h |7 ++
src/gallium/auxiliary/tgsi/tgsi_util.c | 18
src/gallium/auxiliary/tgsi/tgsi_util.h |3 +
src/gallium/drivers/i915/i915_fpc.h|8 +-
src/gallium/drivers/nv30/nvfx_vertprog.c |2 +-
.../drivers/nv50/codegen/nv50_ir_from_tgsi.cpp |6 ++
src/gallium/drivers/r600/r600_llvm.c |2 +-
src/gallium/include/pipe/p_shader_tokens.h | 24 +++--
16 files changed, 193 insertions(+), 132 deletions(-)
diff --git a/src/gallium/auxiliary/gallivm/lp_bld_tgsi_soa.c
b/src/gallium/auxiliary/gallivm/lp_bld_tgsi_soa.c
index 9fe87c4..f157dc1 100644
--- a/src/gallium/auxiliary/gallivm/lp_bld_tgsi_soa.c
+++ b/src/gallium/auxiliary/gallivm/lp_bld_tgsi_soa.c
@@ -517,12 +517,12 @@ emit_mask_scatter(struct lp_build_tgsi_soa_context *bld,
static LLVMValueRef
get_indirect_index(struct lp_build_tgsi_soa_context *bld,
unsigned reg_file, unsigned reg_index,
- const struct tgsi_src_register *indirect_reg)
+ const struct tgsi_ind_register *indirect_reg)
{
LLVMBuilderRef builder = bld->bld_base.base.gallivm->builder;
struct lp_build_context *uint_bld = &bld->bld_base.uint_bld;
/* always use X component of address register */
- unsigned swizzle = indirect_reg->SwizzleX;
+ unsigned swizzle = indirect_reg->Swizzle;
LLVMValueRef base;
LLVMValueRef rel;
LLVMValueRef max_index;
diff --git a/src/gallium/auxiliary/tgsi/tgsi_build.c
b/src/gallium/auxiliary/tgsi/tgsi_build.c
index 33cbbd8..435d94a 100644
--- a/src/gallium/auxiliary/tgsi/tgsi_build.c
+++ b/src/gallium/auxiliary/tgsi/tgsi_build.c
@@ -816,6 +816,43 @@ tgsi_build_src_register(
return src_register;
}
+static struct tgsi_ind_register
+tgsi_default_ind_register( void )
+{
+ struct tgsi_ind_register ind_register;
+
+ ind_register.File = TGSI_FILE_NULL;
+ ind_register.Swizzle = TGSI_SWIZZLE_X;
+ ind_register.ArrayID = 0;
+
+ return ind_register;
+}
+
+static struct tgsi_ind_register
+tgsi_build_ind_register(
+ unsigned file,
+ unsigned swizzle,
+ unsigned arrayid,
+ int index,
+ struct tgsi_instruction *instruction,
+ struct tgsi_header *header )
+{
+ struct tgsi_ind_register ind_register;
+
+ assert( file < TGSI_FILE_COUNT );
+ assert( swizzle <= TGSI_SWIZZLE_W );
+ assert( index >= -0x8000 && index <= 0x7FFF );
+
+ ind_register.File = file;
+ ind_register.Swizzle = swizzle;
+ ind_register.Index = index;
+ ind_register.ArrayID = arrayid;
+
+ instruction_grow( instruction, header );
+
+ return ind_register;
+}
+
static struct tgsi_dimension
tgsi_default_dimension( void )
{
@@ -835,9 +872,9 @@ tgsi_default_full_src_register( void )
struct tgsi_full_src_register full_src_register;
full_src_register.Register = tgsi_default_src_register();
- full_src_register.Indirect = tgsi_default_src_register();
+ full_src_register.Indirect = tgsi_default_ind_register();
full_src_register.Dimension = tgsi_default_dimension();
- full_src_register.DimIndirect = tgsi_default_src_register();
+ full_src_register.DimIndirect = tgsi_default_ind_register();
return full_src_register;
}
@@ -910,9 +947,9 @@ tgsi_default_full_dst_register( void )
struct tgsi_full_dst_register full_dst_register;
full_dst_register.Register = tgsi_default_dst_register();
- full_dst_register.Indirect = tgsi_default_src_register();
+ full_dst_register.Indirect = tgsi_default_ind_register();
full_dst_register.Dimension = tgsi_default_dimension();
- full_dst_register.DimIndirect = tgsi_default_src_register();
+ full_dst_register.DimIndirect = tgsi_default_ind_register();
return full_dst_register;
}
@@ -1057,24 +1094,18 @@ tgsi_build_full_instruction(
header );
if( reg->Register.Indirect ) {
- struct tgsi_src_register *ind;
+ struct tgsi_ind_register
[Mesa-dev] [PATCH 9/9] tgsi: add ArrayID documentation v2
From: Christian König v2: further improve the text with comments from Christoph Bumiller. Signed-off-by: Christian König --- src/gallium/docs/source/tgsi.rst | 18 ++ 1 file changed, 18 insertions(+) diff --git a/src/gallium/docs/source/tgsi.rst b/src/gallium/docs/source/tgsi.rst index d9a7fe9..4c55e60 100644 --- a/src/gallium/docs/source/tgsi.rst +++ b/src/gallium/docs/source/tgsi.rst @@ -1833,6 +1833,24 @@ If Interpolate flag is set to 1, a Declaration Interpolate token follows. If file is TGSI_FILE_RESOURCE, a Declaration Resource token follows. +If Array flag is set to 1, a Declaration Array token follows. + +Array Declaration + + +Declarations can optional have an ArrayID attribute which can be referred by +indirect addressing operands. An ArrayID of zero is reserved and treaded as +if no ArrayID is specified. + +If an indirect addressing operand refers to a specific declaration by using +an ArrayID only the registers in this declaration are guaranteed to be +accessed, accessing any register outside this declaration results in undefined +behavior. Note that for compatibility the effective index is zero-based and +not relative to the specified declaration + +If no ArrayID is specified with an indirect addressing operand the whole +register file might be accessed by this operand. This is strongly discouraged +and will prevent packing of scalar/vec2 arrays and effective alias analysis. Declaration Semantic -- 1.7.9.5 ___ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev
[Mesa-dev] Error while compiling the MAPI directory
Hi, I am receiving the following error while compiling the code in the mapi directory. I am using mesa 7.5. (On compiling glapi_dispatch.c) OpenGL\mapi/glapi/glapitemp.h(86) : error C2375: 'glNewList' : redefinition; different linkage c:\poc\edk2\AppPkg\Include\GL/gl.h(944) : see declaration of 'glNewList' OpenGL\mapi/glapi/glapitemp.h(92) : error C2375: 'glEndList' : redefinition; different linkage c:\poc\edk2\AppPkg\Inclu de\GL/gl.h(946) : see declaration of 'glEndList' and so on (On compiling files inside mapi/mapi directory) mapi_tmp.h(48) : error C2006: '#include' : expected a filename, found 'constant' ___ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev
Re: [Mesa-dev] [PATCH 8/9] tgsi: use separate structure for indirect address v2
Am 18.03.2013 14:36, schrieb Christian König: > From: Christian König > > To further improve the optimization of source and destination > indirect addressing we need the ability to store a reference > to the declaration of the addressed operands. > > Since most of the fields in tgsi_src_register doesn't apply for > an indirect addressing operand replace it with a separate > tgsi_ind_register structure and so make room for extra information. > > v2: rename Declaration to ArrayID, put the ArrayID into () instead of [] > > Signed-off-by: Christian König > --- > diff --git a/src/gallium/auxiliary/tgsi/tgsi_ureg.h > b/src/gallium/auxiliary/tgsi/tgsi_ureg.h > index cd140de..be99ae0 100644 > --- a/src/gallium/auxiliary/tgsi/tgsi_ureg.h > +++ b/src/gallium/auxiliary/tgsi/tgsi_ureg.h > @@ -62,6 +62,7 @@ struct ureg_src > int IndirectIndex: 16; /* SINT */ > int DimensionIndex : 16; /* SINT */ > int DimIndIndex : 16; /* SINT */ > + unsigned ArrayID : 10; /* UINT */ > }; > > /* Very similar to a tgsi_dst_register, removing unsupported fields > @@ -84,6 +85,7 @@ struct ureg_dst > int Index : 16; /* SINT */ > int IndirectIndex : 16; /* SINT */ > int IndirectSwizzle : 2; /* TGSI_SWIZZLE_ */ > + unsigned ArrayID : 10; /* UINT */ > }; > Not sure I like the ordering of these fields... In any case it would be difficult (but possible) to make ureg_dst fit into a 64bit struct. FWIW (but that's not your fault) IndirectSwizzle really ought to be a unsigned too, looks to me like this misdeclaration really could cause bugs somewhere. Roland ___ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev
[Mesa-dev] [PATCH 1/2] build libgallium shared by default.
This is one of the 2 patches used in ubuntu for decreasing size of mesa build. The other one is more hacky, and links libmesagallium into libgallium, and then links libgallium against libdricore too for minimal duplication. This might mess up with static llvm, iirc static llvm is built wrong and linking libgallium against it will cause it to export all llvm symbols too. I believe that this is a bug in llvm though, not in mesa. The static libraries should not export all llvm symbols. I should probably change the default to not building shared gallium for the same reason. :-) Meant mostly for discussion purposes, although if there is no objection I'll change the default to disabling shared gallium, and just let every distro enable shared llvm and libgallium for themselves. >8--- Signed-off-by: Maarten Lankhorst --- a/configure.ac +++ b/configure.ac @@ -733,6 +733,19 @@ fi AM_CONDITIONAL(HAVE_SHARED_GLAPI, test "x$enable_shared_glapi" = xyes) +AC_ARG_ENABLE([shared-gallium], +[AS_HELP_STRING([--enable-shared-gallium], +[Enable shared gallium core @<:@default=yes@:>@])], +[enable_shared_gallium="$enableval"], +[enable_shared_gallium=yes]) + +SHARED_GALLIUM="0" +if test "x$enable_shared_gallium" = xyes; then +SHARED_GALLIUM="1" +fi +AC_SUBST([SHARED_GALLIUM]) +AM_CONDITIONAL(HAVE_SHARED_GALLIUM, test $SHARED_GALLIUM = 1) + dnl dnl Driver specific build directories dnl --- a/src/gallium/auxiliary/Makefile.am +++ b/src/gallium/auxiliary/Makefile.am @@ -3,13 +3,19 @@ include Makefile.sources include $(top_srcdir)/src/gallium/Automake.inc -noinst_LTLIBRARIES = libgallium.la - AM_CFLAGS = \ -I$(top_srcdir)/src/gallium/auxiliary/util \ - $(GALLIUM_CFLAGS) $(VISIBILITY_CFLAGS) + $(GALLIUM_CFLAGS) + +AM_CXXFLAGS = -AM_CXXFLAGS = $(VISIBILITY_CXXFLAGS) +if HAVE_SHARED_GALLIUM +lib_LTLIBRARIES = libgallium.la +else +noinst_LTLIBRARIES = libgallium.la +AM_CFLAGS += $(VISIBILITY_CFLAGS) +AM_CXXFLAGS += $(VISIBILITY_CXXFLAGS) +endif libgallium_la_SOURCES = \ $(C_SOURCES) \ @@ -28,6 +34,8 @@ AM_CXXFLAGS += -fno-rtti +libgallium_la_LIBADD = $(LLVM_LIBS) + endif libgallium_la_SOURCES += \ ___ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev
Re: [Mesa-dev] [PATCH 8/9] tgsi: use separate structure for indirect address v2
Am 18.03.2013 15:12, schrieb Roland Scheidegger: Am 18.03.2013 14:36, schrieb Christian König: From: Christian König To further improve the optimization of source and destination indirect addressing we need the ability to store a reference to the declaration of the addressed operands. Since most of the fields in tgsi_src_register doesn't apply for an indirect addressing operand replace it with a separate tgsi_ind_register structure and so make room for extra information. v2: rename Declaration to ArrayID, put the ArrayID into () instead of [] Signed-off-by: Christian König --- diff --git a/src/gallium/auxiliary/tgsi/tgsi_ureg.h b/src/gallium/auxiliary/tgsi/tgsi_ureg.h index cd140de..be99ae0 100644 --- a/src/gallium/auxiliary/tgsi/tgsi_ureg.h +++ b/src/gallium/auxiliary/tgsi/tgsi_ureg.h @@ -62,6 +62,7 @@ struct ureg_src int IndirectIndex: 16; /* SINT */ int DimensionIndex : 16; /* SINT */ int DimIndIndex : 16; /* SINT */ + unsigned ArrayID : 10; /* UINT */ }; /* Very similar to a tgsi_dst_register, removing unsupported fields @@ -84,6 +85,7 @@ struct ureg_dst int Index : 16; /* SINT */ int IndirectIndex : 16; /* SINT */ int IndirectSwizzle : 2; /* TGSI_SWIZZLE_ */ + unsigned ArrayID : 10; /* UINT */ }; Not sure I like the ordering of these fields... In any case it would be difficult (but possible) to make ureg_dst fit into a 64bit struct. FWIW (but that's not your fault) IndirectSwizzle really ought to be a unsigned too, looks to me like this misdeclaration really could cause bugs somewhere. Noticed the strange declaration of IndirectSwizzle also, and I have it on my todo list (as very low priority) to figure out if this causes any problems or not, but as far as I have seen we currently only use indirect addressing with swizzle X, and then it probably won't hurt. Regarding putting everything into a 64bit struct, well I'm pretty sure that I don't like that idea. Bitfields might make sense for TGSI tokens, but for ureg_src/ureg_dst saving some bytes of memory probably doesn't judge the extra overhead related to them. Christian. Roland ___ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev ___ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev
[Mesa-dev] [PATCH 2/2] (hack) build libgallium against libdricore
This is a hack ubuntu uses to reduce the size of mesa, it links libgallium
against libmesagallium and libdricore.
I think it's too ugly to commit in its current form, but it works with the
flags ubuntu passes to configure.
On 9.0 branch I had a similar patch, without shared libgallium and without this
hack, the stripped build was 36 mb. With just shared gallium it was 25 mb,
and with both applied it was 12 mb.
libmesagallium had to be moved to a subdir, to force the correct build order.
>8---
Not-signed-off-by: Maarten Lankhorst
--- a/src/mesa/libdricore/Makefile.am
+++ b/src/mesa/libdricore/Makefile.am
@@ -42,6 +42,7 @@
libdricore@VERSION@_la_LDFLAGS = -version-number 1:0
libdricore@VERSION@_la_LIBADD = \
../program/libdricore_program.la \
+$(top_builddir)/src/mapi/shared-glapi/libglapi.la
$()
if HAVE_X86_ASM
@@ -65,8 +66,10 @@
-I$(top_srcdir)/src/mesa/sparc
endif
+lib_LTLIBRARIES =
+
if HAVE_DRI
-lib_LTLIBRARIES = libdricore@VERSION@.la
+lib_LTLIBRARIES += libdricore@VERSION@.la
# Provide compatibility with scripts for the old Mesa build system for
# a while by putting a link to the driver into /lib of the build tree.
@@ -76,6 +79,17 @@
ln -sf libdricore@VERSION@.so.1
$(top_builddir)/$(LIB_DIR)/libdricore@VERSION@.so
endif
+libmesagallium_la_SOURCES = \
+ $(STATETRACKER_FILES)
+libmesagallium_la_CFLAGS = @LLVM_CFLAGS@
+libmesagallium_la_CXXFLAGS = @LLVM_CXXFLAGS@
+libmesagallium_la_LIBADD = libdricore@VERSION@.la $(LLVM_LIBS)
+
+if HAVE_GALLIUM
+noinst_LTLIBRARIES = libmesagallium.la
+endif
+
+
CLEANFILES = \
$(top_builddir)/$(LIB_DIR)/libdricore@VERSION@.so.1 \
$(top_builddir)/$(LIB_DIR)/libdricore@VERSION@.so
--- a/configure.ac
+++ b/configure.ac
@@ -928,6 +928,8 @@
AC_SUBST([GLESv2_PC_LIB_PRIV])
DRI_LIB_DEPS="\$(top_builddir)/src/mesa/libdricore/libdricore${VERSION}.la"
+MESAGALLIUM_LIBS="${DRI_LIB_DEPS}"
+AC_SUBST([MESAGALLIUM_LIBS], ${MESAGALLIUM_LIBS})
AC_SUBST([HAVE_XF86VIDMODE])
--- a/src/gallium/targets/dri-i915/Makefile.am
+++ b/src/gallium/targets/dri-i915/Makefile.am
@@ -49,7 +49,7 @@
i915_dri_la_LDFLAGS = -module -avoid-version -shared -no-undefined
i915_dri_la_LIBADD = \
- $(top_builddir)/src/mesa/libmesagallium.la \
+ @MESAGALLIUM_LIBS@ \
$(top_builddir)/src/gallium/auxiliary/libgallium.la \
$(top_builddir)/src/gallium/state_trackers/dri/drm/libdridrm.la \
$(top_builddir)/src/gallium/winsys/i915/drm/libi915drm.la \
--- a/src/gallium/targets/dri-nouveau/Makefile.am
+++ b/src/gallium/targets/dri-nouveau/Makefile.am
@@ -48,7 +48,7 @@
nouveau_dri_la_LDFLAGS = -module -avoid-version -shared -no-undefined
nouveau_dri_la_LIBADD = \
- $(top_builddir)/src/mesa/libmesagallium.la \
+ @MESAGALLIUM_LIBS@ \
$(top_builddir)/src/gallium/auxiliary/libgallium.la \
$(top_builddir)/src/gallium/state_trackers/dri/drm/libdridrm.la \
$(top_builddir)/src/gallium/winsys/nouveau/drm/libnouveaudrm.la \
--- a/src/gallium/targets/dri-r300/Makefile.am
+++ b/src/gallium/targets/dri-r300/Makefile.am
@@ -49,7 +49,7 @@
r300_dri_la_LDFLAGS = -module -avoid-version -shared -no-undefined
r300_dri_la_LIBADD = \
- $(top_builddir)/src/mesa/libmesagallium.la \
+ @MESAGALLIUM_LIBS@ \
$(top_builddir)/src/gallium/auxiliary/libgallium.la \
$(top_builddir)/src/gallium/state_trackers/dri/drm/libdridrm.la \
$(top_builddir)/src/gallium/winsys/radeon/drm/libradeonwinsys.la \
--- a/src/gallium/targets/dri-r600/Makefile.am
+++ b/src/gallium/targets/dri-r600/Makefile.am
@@ -48,7 +48,7 @@
r600_dri_la_LDFLAGS = -module -avoid-version -shared -no-undefined
r600_dri_la_LIBADD = \
- $(top_builddir)/src/mesa/libmesagallium.la \
+ @MESAGALLIUM_LIBS@ \
$(top_builddir)/src/gallium/auxiliary/libgallium.la \
$(top_builddir)/src/gallium/drivers/r600/libr600.la \
$(top_builddir)/src/gallium/state_trackers/dri/drm/libdridrm.la \
--- a/src/gallium/targets/dri-radeonsi/Makefile.am
+++ b/src/gallium/targets/dri-radeonsi/Makefile.am
@@ -49,7 +49,7 @@
radeonsi_dri_la_LDFLAGS = -module -avoid-version -shared -no-undefined
radeonsi_dri_la_LIBADD = \
- $(top_builddir)/src/mesa/libmesagallium.la \
+ @MESAGALLIUM_LIBS@ \
$(top_builddir)/src/gallium/auxiliary/libgallium.la \
$(top_builddir)/src/gallium/drivers/radeonsi/libradeonsi.la \
$(top_builddir)/src/gallium/state_trackers/dri/drm/libdridrm.la \
--- a/src/gallium/targets/dri-swrast/Makefile.am
+++ b/src/gallium/targets/dri-swrast/Makefile.am
@@ -49,7 +49,7 @@
swrast_dri_la_LDFLAGS = -module -avoid-version -shared -no-undefined
swrast_dri_la_LIBADD = \
- $(top_builddir)/src/mesa/libmesagallium.la \
+ @MESAGALLIUM_LIBS@ \
$(top_builddir)/src/gallium/auxiliary/libgallium.la \
$(top_builddir)/src/gallium/state_trackers/dri/sw/libdrisw.la \
$(top_builddir)/src/gallium/winsys/sw/dri
Re: [Mesa-dev] [PATCH 1/2] build libgallium shared by default.
2013/3/18 Maarten Lankhorst : > This is one of the 2 patches used in ubuntu for decreasing size of mesa build. > > The other one is more hacky, and links libmesagallium into libgallium, > and then links libgallium against libdricore too for minimal duplication. > > This might mess up with static llvm, iirc static llvm is built wrong and > linking libgallium against it will cause it to export all llvm symbols too. > I believe that this is a bug in llvm though, not in mesa. The static libraries > should not export all llvm symbols. > > I should probably change the default to not building shared gallium for the > same reason. :-) > > Meant mostly for discussion purposes, although if there is no objection I'll > change the default to disabling shared gallium, and just let every distro > enable > shared llvm and libgallium for themselves. > >>8--- > Signed-off-by: Maarten Lankhorst > > --- a/configure.ac > +++ b/configure.ac > @@ -733,6 +733,19 @@ > fi > AM_CONDITIONAL(HAVE_SHARED_GLAPI, test "x$enable_shared_glapi" = xyes) > > +AC_ARG_ENABLE([shared-gallium], > +[AS_HELP_STRING([--enable-shared-gallium], > +[Enable shared gallium core @<:@default=yes@:>@])], > +[enable_shared_gallium="$enableval"], > +[enable_shared_gallium=yes]) > + The following code could be simplified... > +SHARED_GALLIUM="0" > +if test "x$enable_shared_gallium" = xyes; then > +SHARED_GALLIUM="1" > +fi > +AC_SUBST([SHARED_GALLIUM]) > +AM_CONDITIONAL(HAVE_SHARED_GALLIUM, test $SHARED_GALLIUM = 1) > + ... to: AM_CONDITIONAL(HAVE_SHARED_GALLIUM, test "x$enable_shared_gallium" = xyes) > dnl > dnl Driver specific build directories > dnl > --- a/src/gallium/auxiliary/Makefile.am > +++ b/src/gallium/auxiliary/Makefile.am > @@ -3,13 +3,19 @@ > include Makefile.sources > include $(top_srcdir)/src/gallium/Automake.inc > > -noinst_LTLIBRARIES = libgallium.la > - > AM_CFLAGS = \ > -I$(top_srcdir)/src/gallium/auxiliary/util \ > - $(GALLIUM_CFLAGS) $(VISIBILITY_CFLAGS) > + $(GALLIUM_CFLAGS) > + > +AM_CXXFLAGS = > > -AM_CXXFLAGS = $(VISIBILITY_CXXFLAGS) > +if HAVE_SHARED_GALLIUM > +lib_LTLIBRARIES = libgallium.la > +else > +noinst_LTLIBRARIES = libgallium.la > +AM_CFLAGS += $(VISIBILITY_CFLAGS) > +AM_CXXFLAGS += $(VISIBILITY_CXXFLAGS) > +endif > > libgallium_la_SOURCES = \ > $(C_SOURCES) \ > @@ -28,6 +34,8 @@ > > AM_CXXFLAGS += -fno-rtti > > +libgallium_la_LIBADD = $(LLVM_LIBS) > + > endif > > libgallium_la_SOURCES += \ > > ___ > mesa-dev mailing list > mesa-dev@lists.freedesktop.org > http://lists.freedesktop.org/mailman/listinfo/mesa-dev ___ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev
Re: [Mesa-dev] [PATCH] gallivm: fix return opcode handling in main function of a shader
Looks good to me. Please add all necessary test cases to cover all these code
paths.
Good stuff guys. Very subtle code indeed.
Jose
- Original Message -
> From: Roland Scheidegger
>
> If we're in some conditional or loop we must not return, or the code
> after the condition is never executed.
> (v2): And, we also can't just continue as nothing happened, since the
> mask update code would later check if we actually have a mask, so we
> need to remember that there was a return in main where we didn't exit
> (to illustrate this, a ret in a if clause would cause a mask update
> which is still ok as we're in a conditional, but after the endif the
> mask update code would drop the mask hence bringing execution back to
> pixels which should have their execution mask set to zero by the ret).
> Thanks to Christoph Bumiller for figuring this out.
>
> This fixes https://bugs.freedesktop.org/show_bug.cgi?id=62357.
>
> Note: This is a candidate for the stable branches.
> ---
> src/gallium/auxiliary/gallivm/lp_bld_tgsi.h | 1 +
> src/gallium/auxiliary/gallivm/lp_bld_tgsi_soa.c | 20 +---
> 2 files changed, 18 insertions(+), 3 deletions(-)
>
> diff --git a/src/gallium/auxiliary/gallivm/lp_bld_tgsi.h
> b/src/gallium/auxiliary/gallivm/lp_bld_tgsi.h
> index dac97c3..6e65e12 100644
> --- a/src/gallium/auxiliary/gallivm/lp_bld_tgsi.h
> +++ b/src/gallium/auxiliary/gallivm/lp_bld_tgsi.h
> @@ -243,6 +243,7 @@ struct lp_exec_mask {
> struct lp_build_context *bld;
>
> boolean has_mask;
> + boolean ret_in_main;
>
> LLVMTypeRef int_vec_type;
>
> diff --git a/src/gallium/auxiliary/gallivm/lp_bld_tgsi_soa.c
> b/src/gallium/auxiliary/gallivm/lp_bld_tgsi_soa.c
> index 0dc26b5..965255a 100644
> --- a/src/gallium/auxiliary/gallivm/lp_bld_tgsi_soa.c
> +++ b/src/gallium/auxiliary/gallivm/lp_bld_tgsi_soa.c
> @@ -73,6 +73,7 @@ static void lp_exec_mask_init(struct lp_exec_mask *mask,
> struct lp_build_context
>
> mask->bld = bld;
> mask->has_mask = FALSE;
> + mask->ret_in_main = FALSE;
> mask->cond_stack_size = 0;
> mask->loop_stack_size = 0;
> mask->call_stack_size = 0;
> @@ -108,7 +109,7 @@ static void lp_exec_mask_update(struct lp_exec_mask
> *mask)
> } else
> mask->exec_mask = mask->cond_mask;
>
> - if (mask->call_stack_size) {
> + if (mask->call_stack_size || mask->ret_in_main) {
> mask->exec_mask = LLVMBuildAnd(builder,
> mask->exec_mask,
> mask->ret_mask,
> @@ -117,7 +118,8 @@ static void lp_exec_mask_update(struct lp_exec_mask
> *mask)
>
> mask->has_mask = (mask->cond_stack_size > 0 ||
> mask->loop_stack_size > 0 ||
> - mask->call_stack_size > 0);
> + mask->call_stack_size > 0 ||
> + mask->ret_in_main);
> }
>
> static void lp_exec_mask_cond_push(struct lp_exec_mask *mask,
> @@ -348,11 +350,23 @@ static void lp_exec_mask_ret(struct lp_exec_mask *mask,
> int *pc)
> LLVMBuilderRef builder = mask->bld->gallivm->builder;
> LLVMValueRef exec_mask;
>
> - if (mask->call_stack_size == 0) {
> + if (mask->cond_stack_size == 0 &&
> + mask->loop_stack_size == 0 &&
> + mask->call_stack_size == 0) {
> /* returning from main() */
> *pc = -1;
> return;
> }
> +
> + if (mask->call_stack_size == 0) {
> + /*
> + * This requires special handling since we need to ensure
> + * we don't drop the mask even if we have no call stack
> + * (e.g. after a ret in a if clause after the endif)
> + */
> + mask->ret_in_main = TRUE;
> + }
> +
> exec_mask = LLVMBuildNot(builder,
> mask->exec_mask,
> "ret");
> --
> 1.7.9.5
>
___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/mesa-dev
[Mesa-dev] [Bug 62425] Crash in vbo_save_playback_vertex_list when using glCallList in Cairo-Dock
https://bugs.freedesktop.org/show_bug.cgi?id=62425 --- Comment #1 from Brian Paul --- This sounds the same as fedora bug 918661: http://bugzilla.redhat.com/show_bug.cgi?id=918661 Can you see if the patch from that bug report helps? -- You are receiving this mail because: You are the assignee for the bug. ___ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev
Re: [Mesa-dev] glxgears is faster but 3D render is so slow
On 03/15/2013 05:33 PM, jupiter wrote:
Hi Brian,
On 3/15/13, Brian Paul wrote:
On 03/15/2013 05:39 AM, jupiter wrote:
Thanks Brian and Matt.
On 3/15/13, Matt Turner wrote:
On Thu, Mar 14, 2013 at 6:29 AM, Brian Paul wrote:
Hmm, I guess autoconf still has some unneeded dependencies on DRI when
it's
not needed. You might try adding --with-gallium-drivers=swrast so that
no
DRI drivers are selected.
Don't think so. He's just not setting --with-gallium-drivers= so he
gets a radeon driver by default.
I did try to setup --with-gallium-drivers=llvm, it was an error, the
gallium drivers does not have llvm. I now changed to
--with-gallium-drivers=swrast. Matt is right, I don't need libdrm
anymore.
What is the correct llvm for --with-gallium-drivers? I also seen that
dri-core were also built, that could cause the problems.
The package built on CentOS 6.2 32-bit machine now included
lib/gallium, but the libGL.so and libGL.so.1 did not link to
lib/gallium/libGL.so.1.5.0. After manually linking the lib/libGL.so
and libGL.so.1 to lib/gallium/libGL.so.1.5.0, although the glxinfo
OpenGL rendering string is now pointing to the llvmpipe, but it seems
broken the xlib driver. It stopped running my 3D application via VNC
connection. Does the LLVMPIPE use any DRI?
No.
Or is it still xlib driver?
llvmpipe uses Xlib only.
As the VNC can only use xlib, anything bypass the xlib will break the
VNC connection.
Do other OpenGL apps run OK with llvmpipe or is it just Chimera that's
not working? How exactly is Chimera failing/broken?
Please see following benchmark for using both xlib and llvm:
(1) xlib driver
$ glxinfo | grep "OpenGL renderer string"
OpenGL renderer string: Mesa X11
glxgears 1500 FPS
glxspheres
15 frames/sec - 15 Mpixels/sec
(2) llvm driver
$ glxinfo | grep "OpenGL renderer string"
OpenGL renderer string: Gallium 0.4 on llvmpipe (LLVM 3.2, 128 bits)
glxgears: 600 FPS
glxspheres: 1 frames/sec - 1 Mpixels/sec
I'm not familiar with glxspheres. But the xlib/swrast driver was
optimized for simple things like glxgears. llvm might be slower on
that kind of thing, but it should be much, much faster with modern
apps that uses shaders and texturing.
It is fair to say, if running llvm driver in my local machine (a
32-bit CentOS 6.2 without VNC connection), it was indeed faster than
the xlib driver.
Seems to me that the llvm driver broken the xlib VNC connection which
could be caused by either I haven't configure the llvm correctly, or
mesa llvm compile process may have bugs.
I don't understand what you mean by "llvm driver broken the xlib VNC
connection".
For your further examination, please see following configurations for
building each drivers:
(1) Compile xlib driver
${SOURCE}/${CONFIGURE} --prefix=${INSTALL} --enable-xlib-glx
--disable-dri --with-gallium-drivers=swrast
(2) Compile llvm driver
LLVM="/usr/local/libllvm/3.2"
${SOURCE}/${CONFIGURE} --prefix=${INSTALL} --enable-xlib-glx
--disable-dri --enable-gallium-llvm --with-gallium-drivers=swrast
--with-llvm-shared-libs=${LLVM}/lib --with-llvm-prefix=${LLVM}
Manually change libGL.so and libGL.so.1 to link lib/gallium/libGL.so.1.5.0.
Looks OK to me.
-Brian
___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/mesa-dev
[Mesa-dev] [Bug 61827] [softpipe] piglit fbo-clear-formats GL_ARB_depth_buffer_float stencil regression
https://bugs.freedesktop.org/show_bug.cgi?id=61827 Anuj Phogat changed: What|Removed |Added Status|NEW |RESOLVED Resolution|--- |DUPLICATE --- Comment #1 from Anuj Phogat --- This bug is duplicate of bug 59498. *** This bug has been marked as a duplicate of bug 59498 *** -- You are receiving this mail because: You are the assignee for the bug. ___ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev
[Mesa-dev] [Bug 62482] New: [softpipe] piglit amd_seamless_cubemap_per_texture regression
https://bugs.freedesktop.org/show_bug.cgi?id=62482
Priority: medium
Bug ID: 62482
Keywords: regression
CC: jfons...@vmware.com
Assignee: mesa-dev@lists.freedesktop.org
Summary: [softpipe] piglit amd_seamless_cubemap_per_texture
regression
Severity: normal
Classification: Unclassified
OS: Linux (All)
Reporter: v...@freedesktop.org
Hardware: x86-64 (AMD64)
Status: NEW
Version: git
Component: Other
Product: Mesa
mesa: 5a13e051d9043c2b95333cf778b242ce94be70d5 (master)
$ ./bin/amd_seamless_cubemap_per_texture -auto
Mesa warning: failed to remap glClampColorARB
Mesa warning: failed to remap glTexBufferARB
Mesa warning: failed to remap glFramebufferTextureARB
Mesa warning: failed to remap glVertexAttribDivisorARB
Mesa warning: failed to remap glProgramParameteriARB
Probe at (80,20)
Expected: 0.50 0.00 0.50
Observed: 1.00 0.00 0.00
Probe at (170,20)
Expected: 0.50 0.00 0.50
Observed: 0.00 0.00 1.00
Probe at (200,20)
Expected: 0.50 0.00 0.50
Observed: 1.00 0.00 0.00
Probe at (230,20)
Expected: 0.50 0.00 0.50
Observed: 0.00 0.00 1.00
Probe at (20,50)
Expected: 0.50 0.00 0.50
Observed: 1.00 0.00 0.00
Probe at (50,50)
Expected: 0.50 0.00 0.50
Observed: 0.00 0.00 1.00
Probe at (80,50)
Expected: 0.50 0.00 0.50
Observed: 1.00 0.00 0.00
Probe at (110,50)
Expected: 0.50 0.00 0.50
Observed: 0.00 0.00 1.00
Probe at (140,50)
Expected: 0.50 0.00 0.50
Observed: 1.00 0.00 0.00
Probe at (170,50)
Expected: 0.50 0.00 0.50
Observed: 0.00 0.00 1.00
Probe at (200,50)
Expected: 0.50 0.00 0.50
Observed: 1.00 0.00 0.00
Probe at (230,50)
Expected: 0.50 0.00 0.50
Observed: 0.00 0.00 1.00
PIGLIT: {'result': 'fail' }
6a3d77e13dbae17000e35ce16023532200e68d09 is the first bad commit
commit 6a3d77e13dbae17000e35ce16023532200e68d09
Author: José Fonseca
Date: Thu Mar 14 11:44:21 2013 +
softpipe: Shrink context size.
- each softpipe_tex_tile_cache 50*64*64*4*4 = 3,276,800 bytes
- each softpipe_context has 3*32 softpipe_tex_tile_cache, i.e, each
softpipe
context is 314,572,800 bytes, i.e, 300MB
That is, in a 32bits process (around 3GB virtual memory max), we can
only fit 10 contexts.
This change is a short-term hack to shrink the context size. Longer
term we'll need to change how the texture cache works.
Reviewed-by: Brian Paul
:04 04 64ff15d9847673c6545754f29cba7bcbdd4be457
727f8e3f8b282c58978ca3c4a9d637ebc4f7f21e Msrc
bisect run success
--
You are receiving this mail because:
You are the assignee for the bug.
___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/mesa-dev
Re: [Mesa-dev] Error while compiling the MAPI directory
- Original Message - > > > Hi, > > I am receiving the following error while compiling the code in the mapi > directory. I am using mesa 7.5. If you're compiling with MSVC I'd recommend using a recent Mesa release and save your self a world of trouble. It's known to build well there. If you must use this old release, then you'll likely need to search the MSVC build fixes and crossport them. Jose ___ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev
Re: [Mesa-dev] [PATCH 5/7] mesa: allow multisample texture targets in [Get]TexParameter*
Chris Forbes writes: > ARB_texture_storage_multisample allows texture parameters to be > queried for TEXTURE_2D_MULTISAMPLE and TEXTURE_2D_MULTISAMPLE_ARRAY > targets. > > Some parameters may also be set, with the following exceptions: > > - TEXTURE_BASE_LEVEL may not be set to a nonzero value; generates >INVALID_OPERATION > > - any state which appears in the `per-sampler` state table may not > be set; generates INVALID_OPERATION > > @@ -348,6 +392,11 @@ set_tex_parameteri(struct gl_context *ctx, > case GL_TEXTURE_MAX_LEVEL: >if (texObj->MaxLevel == params[0]) > return GL_FALSE; > + > + if ((texObj->Target == GL_TEXTURE_2D_MULTISAMPLE || > + texObj->Target == GL_TEXTURE_2D_MULTISAMPLE_ARRAY) && params[0] > != 0) > + goto invalid_operation; I don't see anything banning MAX_LEVEL != 0 -- it's not sampler state, and it's not otherwise called out. On the other hand, there's no useful thing you could do with MAX_LEVEL != 0, since it's a single-level texture type. I don't have an opinion either way, really. pgpdojtcqpxBG.pgp Description: PGP signature ___ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev
[Mesa-dev] [PATCH] mesa: Disable validate_ir_tree() on release builds.
Since half of ir_validate uses asserts() (the other using printf() then
abort()), there's not much use to calling it in a release build. Cuts
6.3% of the startup time of TF2.
NOTE: This is a candidate for the stable branches.
---
src/glsl/ir_validate.cpp |6 ++
src/glsl/linker.cpp |2 --
2 files changed, 6 insertions(+), 2 deletions(-)
diff --git a/src/glsl/ir_validate.cpp b/src/glsl/ir_validate.cpp
index 24ea506..699c192 100644
--- a/src/glsl/ir_validate.cpp
+++ b/src/glsl/ir_validate.cpp
@@ -701,6 +701,11 @@ check_node_type(ir_instruction *ir, void *data)
void
validate_ir_tree(exec_list *instructions)
{
+ /* We shouldn't have any reason to validate IR in a release build,
+* and it's half composed of assert()s anyway which wouldn't do
+* anything.
+*/
+#ifdef DEBUG
ir_validate v;
v.run(instructions);
@@ -710,4 +715,5 @@ validate_ir_tree(exec_list *instructions)
visit_tree(ir, check_node_type, NULL);
}
+#endif
}
diff --git a/src/glsl/linker.cpp b/src/glsl/linker.cpp
index 29856b0..2b30d2b 100644
--- a/src/glsl/linker.cpp
+++ b/src/glsl/linker.cpp
@@ -1067,13 +1067,11 @@ link_intrastage_shaders(void *mem_ctx,
free(linking_shaders);
-#ifdef DEBUG
/* At this point linked should contain all of the linked IR, so
* validate it to make sure nothing went wrong.
*/
if (linked)
validate_ir_tree(linked->ir);
-#endif
/* Make a pass over all variable declarations to ensure that arrays with
* unspecified sizes have a size specified. The size is inferred from the
--
1.7.10.4
___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/mesa-dev
Re: [Mesa-dev] [PATCH v3] i965: Avoid unnecessary copy when depthstencil workaround invoked by clear.
Paul Berry writes: > Since apps typically begin rendering with a call to glClear(), it is > likely that when brw_workaround_depthstencil_alignment() moves a > miplevel to a temporary buffer, it can avoid doing a blit, since the > contents of the miplevel are about to be erased. > > This patch adds the necessary plumbing to determine when > brw_workaround_depthstencil_alignment() is being called as a > consequence of glClear(), and avoids the unnecessary blit when it is > safe to do so. > > Reviewed-by: Chad Versace > Reviewed-by: Kenneth Graunke > > v2: Eliminate unnecessary call to _mesa_is_depthstencil_format(). Fix > handling of depth buffer in depth/stencil format. > > v3: Use correct bitfields for clear_mask. Fix handling of depth > buffer in depth/stencil format when hardware uses separate stencil. > When invalidating, make sure we still reassociate the image to the new > miptree. Reviewed-by: Eric Anholt pgpHmdVU47Gge.pgp Description: PGP signature ___ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev
Re: [Mesa-dev] [PATCH 1/7] mesa: add support for immutable textures to teximagemultisample()
Chris Forbes writes: > The new entrypoints will come later, but this adds the actual logic for > supporting immutable multisample textures: > > - The immutability flag is set as desired. > - Attempting to modify an immutable multisample texture produces > INVALID_OPERATION. > > Note: The extension spec does not mention adding this behavior to > TexImage*Multisample, but it seems like the reasonable thing to do. > > Signed-off-by: Chris Forbes You should also be calling tex_storage_error_check() in the immutable case, which covers a bunch of cases from p258 of the 4.2 core spec. Other than that, this series is: Reviewed-by: Eric Anholt pgprP217WqGZ6.pgp Description: PGP signature ___ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev
Re: [Mesa-dev] [PATCH 1/7] mesa: add support for immutable textures to teximagemultisample()
Chris Forbes writes: > The new entrypoints will come later, but this adds the actual logic for > supporting immutable multisample textures: > > - The immutability flag is set as desired. > - Attempting to modify an immutable multisample texture produces > INVALID_OPERATION. > > Note: The extension spec does not mention adding this behavior to > TexImage*Multisample, but it seems like the reasonable thing to do. I think that's because it's just inheriting the text on p261 about what you can't do to immutable textures. pgpMYWGjYJ2Nk.pgp Description: PGP signature ___ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev
Re: [Mesa-dev] [PATCH] R600: Use CONSTANT_BUFFER_0 address space for Implicit Parameters
Test case?
Chad
On Mar 16, 2013, at 9:45 AM, Vincent Lejeune wrote:
> It allows the backend to generate reads to constant cache which
> are faster that VTX_READ.
> ---
> lib/Target/R600/R600ISelLowering.cpp | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/lib/Target/R600/R600ISelLowering.cpp
> b/lib/Target/R600/R600ISelLowering.cpp
> index a0e27ea..caa1899 100644
> --- a/lib/Target/R600/R600ISelLowering.cpp
> +++ b/lib/Target/R600/R600ISelLowering.cpp
> @@ -522,7 +522,7 @@ SDValue
> R600TargetLowering::LowerImplicitParameter(SelectionDAG &DAG, EVT VT,
>unsigned DwordOffset)
> const {
> unsigned ByteOffset = DwordOffset * 4;
> PointerType * PtrType =
> PointerType::get(VT.getTypeForEVT(*DAG.getContext()),
> - AMDGPUAS::PARAM_I_ADDRESS);
> + AMDGPUAS::CONSTANT_BUFFER_0);
>
> // We shouldn't be using an offset wider than 16-bits for implicit
> parameters.
> assert(isInt<16>(ByteOffset));
> --
> 1.8.1.4
>
> ___
> llvm-commits mailing list
> llvm-comm...@cs.uiuc.edu
> http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits
___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/mesa-dev
[Mesa-dev] [Bug 61821] src/mesa/drivers/dri/common/xmlpool.h:96:29: fatal error: xmlpool/options.h
https://bugs.freedesktop.org/show_bug.cgi?id=61821 --- Comment #7 from Mike Stroyan --- Created attachment 76708 --> https://bugs.freedesktop.org/attachment.cgi?id=76708&action=edit configure.ac: Build dricommon for gallium swrast The 07f2dee7 commit added setting HAVE_COMMON_DRI in gallium_check_st(). But the dri-swrast case didn't use gallium_check_st(). I can get dri/common built when needed by using gallium_check_st instead of directly setting GALLIUM_TARGET_DIRS to add "dri-swrast". -- You are receiving this mail because: You are on the CC list for the bug. ___ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev
Re: [Mesa-dev] Google Summer of Code ideas needed
On 03/13/2013 10:11 AM, Tom Stellard wrote: Hi, It's time again for Google Summer of Code, so we need to start updating the X.Org ideas page (http://www.x.org/wiki/SummerOfCodeIdeas) with new ideas. Since there have been a few issues with the wikis lately, if you have any ideas please respond to this thread, and I will make sure they get onto the official ideas page (but still feel free to update the wiki page yourself if you can). A good project description should contain: I've been trying to update the wiki... I don't seem to have an account (seems unpossible), and the account creation page is broken. https://bugs.freedesktop.org/show_bug.cgi?id=62383 In the mean time, I have a couple ideas that someone could add. #1. Find common patterns in real GLSL shaders. Difficulty: Medium Skills Required: C, C++ Helpful, but optional skills: GLSL, compilers Possible Mentor: IanRomanick (idr on IRC) Where to ask questions: mesa-dev@lists.freedesktop.org, #dri-devel on irc.freedesktop.org Descrition: Using Mesa's stand-along GLSL compiler as a basis, generate a database of IR from a large number of existing shaders (e.g., from shaderdb). Write a piece of software that will mine this database for "large" patterns that commonly occur in shaders. This information will be used by people working on the GLSL compiler to improve code generation for these sequences. #2. Improved application of GLSL complier optimizations Difficulty: Easy Skills Required: C, C++ Helpful, but optional skills: GLSL, compilers Possible Mentor: IanRomanick (idr on IRC) Where to ask questions: mesa-dev@lists.freedesktop.org, #dri-devel on irc.freedesktop.org Descrition: Mesa's GLSL compiler contains a large number of optimization passes. Each pass may change the code of a shader, and this may result in opportunities for other passes to make more changes. As a result, we run all of our optimization passes in a loop until the shader code stabilizes. This is expensive, and, though we have never observed this in the wild, it is possible that a shader may never stabilize. Find a static ordering, with possible repeats, of optimization passes that does not compromise the quality of the generated code. Measure the before and after speed of compiling a large set of real-world shaders. - A brief description of the project - A difficulty rating (e.g. easy, medium, hard) - The skills / programming languages required Also, I am going to purge all the old ideas from the ideas page in the next week, so if there are any of the old ideas that you think are still relevant, let me know and I will keep it. The ideas page is used as one of the criteria by Google for selecting mentoring organizations and part of the reason X.Org was not selected last year was that the ideas page was not up to par, so if we want to participate in Google Summer of Code this year, it is important we have a good ideas page with lots of ideas. Thanks, Tom Stellard ___ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev ___ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev
Re: [Mesa-dev] [PATCH] i965: Don't use texture swizzling to force alpha to 1.0 if unnecessary.
Kenneth Graunke writes: > This partially fixes a performance regression since commit 33599433c7. > More work is required to fully fix it in all cases. This at least helps > Warsow. Thanks, Ken. Reviewed-by: Carl Worth -Carl pgpEBf6hWs_2H.pgp Description: PGP signature ___ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev
Re: [Mesa-dev] Google Summer of Code ideas needed
On Wed, Mar 13, 2013 at 10:11 AM, Tom Stellard wrote: > - A brief description of the project > - A difficulty rating (e.g. easy, medium, hard) > - The skills / programming languages required GL/GLSL tests for GL 3.2, 3.3 in the piglit test suite Difficulty: Easy-Medium Skills Required: C Useful skills: OpenGL, GLSL programming Hardware/Software required: driver supporting >= OpenGL 3.2 Possible Mentor: JordanJusten (jljusten on IRC) Where to ask questions: mesa-dev@lists.freedesktop.org, #dri-devel on irc.freedesktop.org Description: Write tests for OpenGL 3.2 / GLSL 1.50 and newer for the piglit test suite. Proposal should identify: * GL/GLSL version/features you plan to focus on * Number of tests you estimate completing * Hardware/driver you have access to -Jordan ___ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev
[Mesa-dev] [PATCH] i965: Align all depth miplevels to 8 in the X direction.
On an INTEL_DEBUG=perf piglit run on IVB, reduces the instances of "HW workaround: blit" (the printouts from the misaligned-depth workaround blits) from 725 to 684 It doesn't totally eliminate the workaround blit, because we still have problems with Y offsets that we can't fix (since texturing can only align miplevels up to 2 or 4, not 8). No regressions on piglit/es3conform on IVB. --- src/mesa/drivers/dri/intel/intel_tex_layout.c | 10 +- 1 file changed, 9 insertions(+), 1 deletion(-) diff --git a/src/mesa/drivers/dri/intel/intel_tex_layout.c b/src/mesa/drivers/dri/intel/intel_tex_layout.c index 35030df..95242f8 100644 --- a/src/mesa/drivers/dri/intel/intel_tex_layout.c +++ b/src/mesa/drivers/dri/intel/intel_tex_layout.c @@ -77,7 +77,15 @@ intel_horizontal_texture_alignment_unit(struct intel_context *intel, if (format == MESA_FORMAT_S8) return 8; - if (intel->gen >= 7 && format == MESA_FORMAT_Z16) + /* The depth alignment requirements in the table above are for rendering to +* depth miplevels using the LOD control fields. We don't use LOD control +* fields, and instead use page offsets plus intra-tile x/y offsets, which +* require that the low 3 bits are zero. To reduce the number of x/y +* offset workaround blits we do, align the X to 8, which depth texturing +* can handle (sadly, it can't handle 8 in the Y direction). +*/ + if (intel->gen >= 4 && + _mesa_get_format_base_format(format) == GL_DEPTH_COMPONENT) return 8; return 4; -- 1.7.10.4 ___ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev
Re: [Mesa-dev] Redefinition error in glapitmp.h
On Sun, Mar 17, 2013 at 12:57 AM, Ritvik Sharma wrote: > Also in the file mapi_tmp.h I am getting #include error for MAPI_ABI_HEADER. Look at how the Makefiles define MAPI_ABI_HEADER (e.g., in src/mapi/es2api/Makefile.am). ___ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev
[Mesa-dev] [PATCH v2 00/15] GLSL 1.50 interface blocks support
git://people.freedesktop.org/~jljusten/mesa interface-blocks-v2 v2: * 3 new patches added to series * Add support for interface block instance arrays * Add support for rejecting unmatched interface blocks during the linking phase. * Known issue: fails new piglit glsl-1.50 test: execution/interface-blocks-same-uniform-varying-name.shader_test v1: * Initial support for GLSL 1.50 interface block support * Known issue: interface block arrays are not working * Known issue: rejection of unmatched interface blocks is not working * Piglit tests for known issues have been sent to piglit list Jordan Justen (15): glsl: rename ast_uniform_block to ast_interface_block glsl parser: rename uniform block to interface block glsl: parse in/out types for interface blocks glsl parser: reject VS+in & FS+out interface blocks glsl parser: on desktop GL require GLSL 150 for instance names glsl parser: allow in & out for interface block members glsl parser: handle interface block member qualifier glsl ast_to_hir: support in/out for interface blocks glsl ir: add as_dereference_record glsl linker: cross validate interface block types between stages glsl linker: remove interface block instance names glsl link_varyings: link interface blocks using the block name glsl linker: support arrays of interface block instances glsl linker: compare interface blocks during intrastage linking glsl linker: compare interface blocks during interstage linking src/glsl/Makefile.sources |2 + src/glsl/ast.h|4 +- src/glsl/ast_to_hir.cpp | 23 ++- src/glsl/glsl_parser.yy | 189 --- src/glsl/interface_blocks.cpp | 104 + src/glsl/interface_blocks.h | 35 + src/glsl/ir.h |6 + src/glsl/ir_optimization.h|1 + src/glsl/link_varyings.cpp| 33 +++- src/glsl/linker.cpp | 28 src/glsl/lower_named_interface_blocks.cpp | 237 + 11 files changed, 599 insertions(+), 63 deletions(-) create mode 100644 src/glsl/interface_blocks.cpp create mode 100644 src/glsl/interface_blocks.h create mode 100644 src/glsl/lower_named_interface_blocks.cpp -- 1.7.10.4 ___ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev
[Mesa-dev] [PATCH v2 01/15] glsl: rename ast_uniform_block to ast_interface_block
Signed-off-by: Jordan Justen
---
src/glsl/ast.h |4 ++--
src/glsl/ast_to_hir.cpp |6 +++---
src/glsl/glsl_parser.yy | 14 +++---
3 files changed, 12 insertions(+), 12 deletions(-)
diff --git a/src/glsl/ast.h b/src/glsl/ast.h
index fcc6b45..49c3939 100644
--- a/src/glsl/ast.h
+++ b/src/glsl/ast.h
@@ -805,9 +805,9 @@ public:
ast_compound_statement *body;
};
-class ast_uniform_block : public ast_node {
+class ast_interface_block : public ast_node {
public:
- ast_uniform_block(ast_type_qualifier layout,
+ ast_interface_block(ast_type_qualifier layout,
const char *instance_name,
ast_expression *array_size)
: layout(layout), block_name(NULL), instance_name(instance_name),
diff --git a/src/glsl/ast_to_hir.cpp b/src/glsl/ast_to_hir.cpp
index 92065f5..40f3188 100644
--- a/src/glsl/ast_to_hir.cpp
+++ b/src/glsl/ast_to_hir.cpp
@@ -4244,12 +4244,12 @@ ast_struct_specifier::hir(exec_list *instructions,
}
ir_rvalue *
-ast_uniform_block::hir(exec_list *instructions,
- struct _mesa_glsl_parse_state *state)
+ast_interface_block::hir(exec_list *instructions,
+ struct _mesa_glsl_parse_state *state)
{
YYLTYPE loc = this->get_location();
- /* The ast_uniform_block has a list of ast_declarator_lists. We
+ /* The ast_interface_block has a list of ast_declarator_lists. We
* need to turn those into ir_variables with an association
* with this uniform block.
*/
diff --git a/src/glsl/glsl_parser.yy b/src/glsl/glsl_parser.yy
index f52ed9b..cd33078 100644
--- a/src/glsl/glsl_parser.yy
+++ b/src/glsl/glsl_parser.yy
@@ -79,7 +79,7 @@ static void yyerror(YYLTYPE *loc, _mesa_glsl_parse_state *st,
const char *msg)
ast_case_label_list *case_label_list;
ast_case_statement *case_statement;
ast_case_statement_list *case_statement_list;
- ast_uniform_block *uniform_block;
+ ast_interface_block *interface_block;
struct {
ast_node *cond;
@@ -1900,7 +1900,7 @@ uniform_block:
}
| layout_qualifier basic_uniform_block
{
- ast_uniform_block *block = $2;
+ ast_interface_block *block = $2;
if (!block->layout.merge_qualifier(& @1, state, $1)) {
YYERROR;
}
@@ -1911,7 +1911,7 @@ uniform_block:
basic_uniform_block:
UNIFORM NEW_IDENTIFIER '{' member_list '}' instance_name_opt ';'
{
- ast_uniform_block *const block = $6;
+ ast_interface_block *const block = $6;
block->block_name = $2;
block->declarations.push_degenerate_list_at_head(& $4->link);
@@ -1944,19 +1944,19 @@ basic_uniform_block:
instance_name_opt:
/* empty */
{
- $$ = new(state) ast_uniform_block(*state->default_uniform_qualifier,
+ $$ = new(state)
ast_interface_block(*state->default_uniform_qualifier,
NULL,
NULL);
}
| NEW_IDENTIFIER
{
- $$ = new(state) ast_uniform_block(*state->default_uniform_qualifier,
+ $$ = new(state)
ast_interface_block(*state->default_uniform_qualifier,
$1,
NULL);
}
| NEW_IDENTIFIER '[' constant_expression ']'
{
- $$ = new(state) ast_uniform_block(*state->default_uniform_qualifier,
+ $$ = new(state)
ast_interface_block(*state->default_uniform_qualifier,
$1,
$3);
}
@@ -1965,7 +1965,7 @@ instance_name_opt:
_mesa_glsl_error(& @1, state,
"instance block arrays must be explicitly sized\n");
- $$ = new(state) ast_uniform_block(*state->default_uniform_qualifier,
+ $$ = new(state)
ast_interface_block(*state->default_uniform_qualifier,
$1,
NULL);
}
--
1.7.10.4
___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/mesa-dev
[Mesa-dev] [PATCH v2 02/15] glsl parser: rename uniform block to interface block
Signed-off-by: Jordan Justen
---
src/glsl/glsl_parser.yy | 24
1 file changed, 12 insertions(+), 12 deletions(-)
diff --git a/src/glsl/glsl_parser.yy b/src/glsl/glsl_parser.yy
index cd33078..33b74ea 100644
--- a/src/glsl/glsl_parser.yy
+++ b/src/glsl/glsl_parser.yy
@@ -115,7 +115,7 @@ static void yyerror(YYLTYPE *loc, _mesa_glsl_parse_state
*st, const char *msg)
%token STRUCT VOID_TOK WHILE
%token IDENTIFIER TYPE_IDENTIFIER NEW_IDENTIFIER
%type any_identifier
-%type instance_name_opt
+%type instance_name_opt
%token FLOATCONSTANT
%token INTCONSTANT UINTCONSTANT BOOLCONSTANT
%token FIELD_SELECTION
@@ -164,7 +164,7 @@ static void yyerror(YYLTYPE *loc, _mesa_glsl_parse_state
*st, const char *msg)
%type interpolation_qualifier
%type layout_qualifier
%type layout_qualifier_id_list layout_qualifier_id
-%type uniform_block_layout_qualifier
+%type interface_block_layout_qualifier
%type type_specifier
%type type_specifier_no_prec
%type type_specifier_nonarray
@@ -223,8 +223,8 @@ static void yyerror(YYLTYPE *loc, _mesa_glsl_parse_state
*st, const char *msg)
%type declaration
%type declaration_statement
%type jump_statement
-%type uniform_block
-%type basic_uniform_block
+%type interface_block
+%type basic_interface_block
%type struct_specifier
%type struct_declaration_list
%type struct_declaration
@@ -784,7 +784,7 @@ declaration:
$3->is_precision_statement = true;
$$ = $3;
}
- | uniform_block
+ | interface_block
{
$$ = $1;
}
@@ -1140,7 +1140,7 @@ layout_qualifier_id:
}
}
- /* See also uniform_block_layout_qualifier. */
+ /* See also interface_block_layout_qualifier. */
if (!$$.flags.i && state->ARB_uniform_buffer_object_enable) {
if (strcmp($1, "std140") == 0) {
$$.flags.q.std140 = 1;
@@ -1211,7 +1211,7 @@ layout_qualifier_id:
"identifier `%s' used\n", $1);
}
}
- | uniform_block_layout_qualifier
+ | interface_block_layout_qualifier
{
$$ = $1;
/* Layout qualifiers for ARB_uniform_buffer_object. */
@@ -1232,7 +1232,7 @@ layout_qualifier_id:
* most qualifiers. See the any_identifier path of
* layout_qualifier_id for the others.
*/
-uniform_block_layout_qualifier:
+interface_block_layout_qualifier:
ROW_MAJOR
{
memset(& $$, 0, sizeof($$));
@@ -1893,12 +1893,12 @@ function_definition:
;
/* layout_qualifieropt is packed into this rule */
-uniform_block:
- basic_uniform_block
+interface_block:
+ basic_interface_block
{
$$ = $1;
}
- | layout_qualifier basic_uniform_block
+ | layout_qualifier basic_interface_block
{
ast_interface_block *block = $2;
if (!block->layout.merge_qualifier(& @1, state, $1)) {
@@ -1908,7 +1908,7 @@ uniform_block:
}
;
-basic_uniform_block:
+basic_interface_block:
UNIFORM NEW_IDENTIFIER '{' member_list '}' instance_name_opt ';'
{
ast_interface_block *const block = $6;
--
1.7.10.4
___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/mesa-dev
[Mesa-dev] [PATCH v2 03/15] glsl: parse in/out types for interface blocks
Previously only 'uniform' was allowed for uniform blocks.
Now, in/out can be parsed, but it will only be allowed for
GLSL >= 150.
Signed-off-by: Jordan Justen
---
src/glsl/glsl_parser.yy | 60 ++-
1 file changed, 49 insertions(+), 11 deletions(-)
diff --git a/src/glsl/glsl_parser.yy b/src/glsl/glsl_parser.yy
index 33b74ea..7adc06d 100644
--- a/src/glsl/glsl_parser.yy
+++ b/src/glsl/glsl_parser.yy
@@ -165,6 +165,7 @@ static void yyerror(YYLTYPE *loc, _mesa_glsl_parse_state
*st, const char *msg)
%type layout_qualifier
%type layout_qualifier_id_list layout_qualifier_id
%type interface_block_layout_qualifier
+%type interface_qualifier
%type type_specifier
%type type_specifier_no_prec
%type type_specifier_nonarray
@@ -1215,11 +1216,11 @@ layout_qualifier_id:
{
$$ = $1;
/* Layout qualifiers for ARB_uniform_buffer_object. */
- if (!state->ARB_uniform_buffer_object_enable) {
+ if ($$.flags.q.uniform && !state->ARB_uniform_buffer_object_enable) {
_mesa_glsl_error(& @1, state,
"#version 140 / GL_ARB_uniform_buffer_object "
"layout qualifier `%s' is used\n", $1);
- } else if (state->ARB_uniform_buffer_object_warn) {
+ } else if ($$.flags.q.uniform &&
state->ARB_uniform_buffer_object_warn) {
_mesa_glsl_warning(& @1, state,
"#version 140 / GL_ARB_uniform_buffer_object "
"layout qualifier `%s' is used\n", $1);
@@ -1909,21 +1910,29 @@ interface_block:
;
basic_interface_block:
- UNIFORM NEW_IDENTIFIER '{' member_list '}' instance_name_opt ';'
+ interface_qualifier NEW_IDENTIFIER '{' member_list '}'
instance_name_opt ';'
{
ast_interface_block *const block = $6;
block->block_name = $2;
block->declarations.push_degenerate_list_at_head(& $4->link);
- if (!state->ARB_uniform_buffer_object_enable) {
- _mesa_glsl_error(& @1, state,
- "#version 140 / GL_ARB_uniform_buffer_object "
- "required for defining uniform blocks\n");
- } else if (state->ARB_uniform_buffer_object_warn) {
- _mesa_glsl_warning(& @1, state,
-"#version 140 / GL_ARB_uniform_buffer_object "
-"required for defining uniform blocks\n");
+ if ($1.flags.q.uniform) {
+ if (!state->ARB_uniform_buffer_object_enable) {
+_mesa_glsl_error(& @1, state,
+ "#version 140 / GL_ARB_uniform_buffer_object "
+ "required for defining uniform blocks\n");
+ } else if (state->ARB_uniform_buffer_object_warn) {
+_mesa_glsl_warning(& @1, state,
+ "#version 140 /
GL_ARB_uniform_buffer_object "
+ "required for defining uniform blocks\n");
+ }
+ } else {
+ if (state->es_shader || state->language_version < 150) {
+_mesa_glsl_error(& @1, state,
+"#version 150 required for using "
+"interface blocks.\n");
+ }
}
/* Since block arrays require names, and both features are added in
@@ -1937,10 +1946,39 @@ basic_interface_block:
"blocks with an instance name\n");
}
+ unsigned interface_type_mask, interface_type_flags;
+ struct ast_type_qualifier temp_type_qualifier;
+
+ temp_type_qualifier.flags.i = 0;
+ temp_type_qualifier.flags.q.uniform = true;
+ temp_type_qualifier.flags.q.in = true;
+ temp_type_qualifier.flags.q.out = true;
+ interface_type_mask = temp_type_qualifier.flags.i;
+ interface_type_flags = $1.flags.i & interface_type_mask;
+ block->layout.flags.i |= interface_type_flags;
+
$$ = block;
}
;
+interface_qualifier:
+ IN_TOK
+ {
+ memset(& $$, 0, sizeof($$));
+ $$.flags.q.in = 1;
+ }
+ | OUT_TOK
+ {
+ memset(& $$, 0, sizeof($$));
+ $$.flags.q.out = 1;
+ }
+ | UNIFORM
+ {
+ memset(& $$, 0, sizeof($$));
+ $$.flags.q.uniform = 1;
+ }
+ ;
+
instance_name_opt:
/* empty */
{
--
1.7.10.4
___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/mesa-dev
[Mesa-dev] [PATCH v2 04/15] glsl parser: reject VS+in & FS+out interface blocks
Signed-off-by: Jordan Justen
---
src/glsl/glsl_parser.yy | 14 ++
1 file changed, 14 insertions(+)
diff --git a/src/glsl/glsl_parser.yy b/src/glsl/glsl_parser.yy
index 7adc06d..8e6b04d 100644
--- a/src/glsl/glsl_parser.yy
+++ b/src/glsl/glsl_parser.yy
@@ -1935,6 +1935,20 @@ basic_interface_block:
}
}
+ /* GLSLangSpec.1.50.11, 4.3.7 Interface Blocks:
+ * "It is illegal to have an input block in a vertex shader
+ * or an output block in a fragment shader"
+ */
+ if ((state->target == vertex_shader) && $1.flags.q.in) {
+ _mesa_glsl_error(& @1, state,
+ "in interface block is not allowed for "
+ "a vertex shader\n");
+ } else if ((state->target == fragment_shader) && $1.flags.q.out) {
+ _mesa_glsl_error(& @1, state,
+ "out interface block is not allowed for "
+ "a fragment shader\n");
+ }
+
/* Since block arrays require names, and both features are added in
* the same language versions, we don't have to explicitly
* version-check both things.
--
1.7.10.4
___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/mesa-dev
[Mesa-dev] [PATCH v2 05/15] glsl parser: on desktop GL require GLSL 150 for instance names
Interface blocks in GLSL 150 allow an instance name to be used.
Signed-off-by: Jordan Justen
---
src/glsl/glsl_parser.yy | 15 ++-
1 file changed, 10 insertions(+), 5 deletions(-)
diff --git a/src/glsl/glsl_parser.yy b/src/glsl/glsl_parser.yy
index 8e6b04d..1fd8cc2 100644
--- a/src/glsl/glsl_parser.yy
+++ b/src/glsl/glsl_parser.yy
@@ -1953,11 +1953,16 @@ basic_interface_block:
* the same language versions, we don't have to explicitly
* version-check both things.
*/
- if (block->instance_name != NULL
- && !(state->language_version == 300 && state->es_shader)) {
- _mesa_glsl_error(& @1, state,
- "#version 300 es required for using uniform "
- "blocks with an instance name\n");
+ if (block->instance_name != NULL) {
+ if(state->es_shader && state->language_version < 300) {
+_mesa_glsl_error(& @1, state,
+"#version 300 es required for using uniform "
+"blocks with an instance name\n");
+ } else if (!state->es_shader && state->language_version < 150) {
+_mesa_glsl_error(& @1, state,
+"#version 150 required for using uniform "
+"blocks with an instance name\n");
+ }
}
unsigned interface_type_mask, interface_type_flags;
--
1.7.10.4
___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/mesa-dev
[Mesa-dev] [PATCH v2 07/15] glsl parser: handle interface block member qualifier
An interface block member may specify the type:
in {
in vec4 in_var_with_qualifier;
};
In this case it must match the interface block type.
It can also omit the qualifier:
uniform {
vec4 uniform_var_without_qualifier;
};
In this case, it should use the same type as the interface block.
Signed-off-by: Jordan Justen
---
src/glsl/glsl_parser.yy | 17 +
1 file changed, 17 insertions(+)
diff --git a/src/glsl/glsl_parser.yy b/src/glsl/glsl_parser.yy
index d21ca2d..ac1d8ed 100644
--- a/src/glsl/glsl_parser.yy
+++ b/src/glsl/glsl_parser.yy
@@ -1976,6 +1976,23 @@ basic_interface_block:
interface_type_flags = $1.flags.i & interface_type_mask;
block->layout.flags.i |= interface_type_flags;
+ foreach_list_typed (ast_declarator_list, member, link,
&block->declarations) {
+ ast_type_qualifier& qualifier = member->type->qualifier;
+ if ((qualifier.flags.i & interface_type_mask) == 0) {
+qualifier.flags.i |= interface_type_flags;
+ } else if ((qualifier.flags.i & interface_type_mask) !=
+interface_type_flags) {
+/* GLSLangSpec.1.50.11, 4.3.7 Interface Blocks:
+ * "Input variables, output variables, and uniform variables
can only
+ * be in in blocks, out blocks, and uniform blocks,
respectively."
+ */
+_mesa_glsl_error(& @1, state,
+ "uniform/in/out qualifier on "
+ "interface block member does not match "
+ "the interface block\n");
+ }
+ }
+
$$ = block;
}
;
--
1.7.10.4
___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/mesa-dev
[Mesa-dev] [PATCH v2 06/15] glsl parser: allow in & out for interface block members
Previously uniform blocks allowed for the 'uniform' keyword
to be used with members of a uniform blocks. With interface
blocks 'in' can be used on 'in' interface block members and
'out' can be used on 'out' interface block members.
The basic_interface_block rule will verify that the same
qualifier type is used with the block and each member.
Signed-off-by: Jordan Justen
---
src/glsl/glsl_parser.yy | 47 ++-
1 file changed, 30 insertions(+), 17 deletions(-)
diff --git a/src/glsl/glsl_parser.yy b/src/glsl/glsl_parser.yy
index 1fd8cc2..d21ca2d 100644
--- a/src/glsl/glsl_parser.yy
+++ b/src/glsl/glsl_parser.yy
@@ -2041,41 +2041,54 @@ member_list:
}
;
-/* Specifying "uniform" inside of a uniform block is redundant. */
-uniformopt:
- /* nothing */
- | UNIFORM
- ;
-
member_declaration:
- layout_qualifier uniformopt type_specifier struct_declarator_list ';'
+ layout_qualifier fully_specified_type struct_declarator_list ';'
{
void *ctx = state;
- ast_fully_specified_type *type = new(ctx) ast_fully_specified_type();
+ ast_fully_specified_type *type = $2;
type->set_location(yylloc);
- type->qualifier = $1;
- type->qualifier.flags.q.uniform = true;
- type->specifier = $3;
+ if (!type->qualifier.merge_qualifier(& @1, state, $1)) {
+ YYERROR;
+ }
+
+ if (type->qualifier.flags.q.attribute) {
+ _mesa_glsl_error(& @1, state,
+ "keyword 'attribute' cannot be used with "
+ "interface block member\n");
+ } else if (type->qualifier.flags.q.varying) {
+ _mesa_glsl_error(& @1, state,
+ "keyword 'varying' cannot be used with "
+ "interface block member\n");
+ }
+
$$ = new(ctx) ast_declarator_list(type);
$$->set_location(yylloc);
$$->ubo_qualifiers_valid = true;
- $$->declarations.push_degenerate_list_at_head(& $4->link);
+ $$->declarations.push_degenerate_list_at_head(& $3->link);
}
- | uniformopt type_specifier struct_declarator_list ';'
+ | fully_specified_type struct_declarator_list ';'
{
void *ctx = state;
- ast_fully_specified_type *type = new(ctx) ast_fully_specified_type();
+ ast_fully_specified_type *type = $1;
type->set_location(yylloc);
- type->qualifier.flags.q.uniform = true;
- type->specifier = $2;
+ if (type->qualifier.flags.q.attribute) {
+ _mesa_glsl_error(& @1, state,
+ "keyword 'attribute' cannot be used in "
+ "interface block\n");
+ } else if (type->qualifier.flags.q.varying) {
+ _mesa_glsl_error(& @1, state,
+ "keyword 'varying' cannot be used in "
+ "interface block\n");
+ }
+
$$ = new(ctx) ast_declarator_list(type);
$$->set_location(yylloc);
$$->ubo_qualifiers_valid = true;
- $$->declarations.push_degenerate_list_at_head(& $3->link);
+ $$->declarations.push_degenerate_list_at_head(& $2->link);
}
;
--
1.7.10.4
___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/mesa-dev
[Mesa-dev] [PATCH v2 08/15] glsl ast_to_hir: support in/out for interface blocks
Signed-off-by: Jordan Justen
---
src/glsl/ast_to_hir.cpp | 17 ++---
1 file changed, 14 insertions(+), 3 deletions(-)
diff --git a/src/glsl/ast_to_hir.cpp b/src/glsl/ast_to_hir.cpp
index 40f3188..ee54c70 100644
--- a/src/glsl/ast_to_hir.cpp
+++ b/src/glsl/ast_to_hir.cpp
@@ -4293,6 +4293,17 @@ ast_interface_block::hir(exec_list *instructions,
*/
assert(declared_variables.is_empty());
+ ir_variable_mode var_mode;
+ if (this->layout.flags.q.in) {
+ var_mode = ir_var_shader_in;
+ } else if (this->layout.flags.q.out) {
+ var_mode = ir_var_shader_out;
+ } else if (this->layout.flags.q.uniform) {
+ var_mode = ir_var_uniform;
+ } else {
+ assert(!"interface block layout qualifier not found!");
+ }
+
/* Page 39 (page 45 of the PDF) of section 4.3.7 in the GLSL ES 3.00 spec
* says:
*
@@ -4309,11 +4320,11 @@ ast_interface_block::hir(exec_list *instructions,
var = new(state) ir_variable(block_array_type,
this->instance_name,
- ir_var_uniform);
+ var_mode);
} else {
var = new(state) ir_variable(block_type,
this->instance_name,
- ir_var_uniform);
+ var_mode);
}
var->interface_type = block_type;
@@ -4329,7 +4340,7 @@ ast_interface_block::hir(exec_list *instructions,
ir_variable *var =
new(state) ir_variable(fields[i].type,
ralloc_strdup(state, fields[i].name),
- ir_var_uniform);
+ var_mode);
var->interface_type = block_type;
state->symbols->add_variable(var);
--
1.7.10.4
___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/mesa-dev
[Mesa-dev] [PATCH v2 09/15] glsl ir: add as_dereference_record
Signed-off-by: Jordan Justen
---
src/glsl/ir.h |6 ++
1 file changed, 6 insertions(+)
diff --git a/src/glsl/ir.h b/src/glsl/ir.h
index bbfec69..ee27dea 100644
--- a/src/glsl/ir.h
+++ b/src/glsl/ir.h
@@ -120,6 +120,7 @@ public:
virtual class ir_dereference * as_dereference() { return NULL; }
virtual class ir_dereference_array *as_dereference_array() { return
NULL; }
virtual class ir_dereference_variable *as_dereference_variable() { return
NULL; }
+ virtual class ir_dereference_record *as_dereference_record() { return NULL;
}
virtual class ir_expression *as_expression() { return NULL; }
virtual class ir_rvalue *as_rvalue() { return NULL; }
virtual class ir_loop * as_loop() { return NULL; }
@@ -1738,6 +1739,11 @@ public:
virtual ir_constant *constant_expression_value(struct hash_table
*variable_context = NULL);
+ virtual ir_dereference_record *as_dereference_record()
+ {
+ return this;
+ }
+
/**
* Get the variable that is ultimately referenced by an r-value
*/
--
1.7.10.4
___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/mesa-dev
[Mesa-dev] [PATCH v2 11/15] glsl linker: remove interface block instance names
Convert interface blocks with instance names into flat
interface blocks without an instance name.
Signed-off-by: Jordan Justen
---
src/glsl/Makefile.sources |1 +
src/glsl/ir_optimization.h|1 +
src/glsl/linker.cpp |6 +
src/glsl/lower_named_interface_blocks.cpp | 202 +
4 files changed, 210 insertions(+)
create mode 100644 src/glsl/lower_named_interface_blocks.cpp
diff --git a/src/glsl/Makefile.sources b/src/glsl/Makefile.sources
index c294aa4..86ae43e 100644
--- a/src/glsl/Makefile.sources
+++ b/src/glsl/Makefile.sources
@@ -62,6 +62,7 @@ LIBGLSL_FILES = \
$(GLSL_SRCDIR)/lower_mat_op_to_vec.cpp \
$(GLSL_SRCDIR)/lower_noise.cpp \
$(GLSL_SRCDIR)/lower_packed_varyings.cpp \
+ $(GLSL_SRCDIR)/lower_named_interface_blocks.cpp \
$(GLSL_SRCDIR)/lower_packing_builtins.cpp \
$(GLSL_SRCDIR)/lower_texture_projection.cpp \
$(GLSL_SRCDIR)/lower_variable_index_to_cond_assign.cpp \
diff --git a/src/glsl/ir_optimization.h b/src/glsl/ir_optimization.h
index 2454bbe..c0850ce 100644
--- a/src/glsl/ir_optimization.h
+++ b/src/glsl/ir_optimization.h
@@ -105,6 +105,7 @@ void lower_ubo_reference(struct gl_shader *shader,
exec_list *instructions);
void lower_packed_varyings(void *mem_ctx, unsigned location_base,
unsigned locations_used, ir_variable_mode mode,
gl_shader *shader);
+void lower_named_interface_blocks(void *mem_ctx, gl_shader *shader);
bool optimize_redundant_jumps(exec_list *instructions);
bool optimize_split_arrays(exec_list *instructions, bool linked);
diff --git a/src/glsl/linker.cpp b/src/glsl/linker.cpp
index acbdd2f..003a331 100644
--- a/src/glsl/linker.cpp
+++ b/src/glsl/linker.cpp
@@ -1748,6 +1748,12 @@ link_shaders(struct gl_context *ctx, struct
gl_shader_program *prog)
prog->LinkStatus = true;
}
+
+ for (unsigned int i = 0; i < MESA_SHADER_TYPES; i++) {
+ if (prog->_LinkedShaders[i] != NULL)
+ lower_named_interface_blocks(mem_ctx, prog->_LinkedShaders[i]);
+ }
+
/* Implement the GLSL 1.30+ rule for discard vs infinite loops Do
* it before optimization because we want most of the checks to get
* dropped thanks to constant propagation.
diff --git a/src/glsl/lower_named_interface_blocks.cpp
b/src/glsl/lower_named_interface_blocks.cpp
new file mode 100644
index 000..2e0c322
--- /dev/null
+++ b/src/glsl/lower_named_interface_blocks.cpp
@@ -0,0 +1,202 @@
+/*
+ * Copyright (c) 2013 Intel Corporation
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice (including the next
+ * paragraph) shall be included in all copies or substantial portions of the
+ * Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
+ * DEALINGS IN THE SOFTWARE.
+ */
+
+/**
+ * \file lower_named_interface_blocks.cpp
+ *
+ * This lowering pass converts all interface blocks with instance names
+ * into interface blocks without an interface name.
+ *
+ * For example, the following shader:
+ *
+ * out block {
+ * float block_var;
+ * } inst_name;
+ *
+ * main()
+ * {
+ * inst_name.block_var = 0.0;
+ * }
+ *
+ * Is rewritten to:
+ *
+ * out block {
+ * float block_var;
+ * };
+ *
+ * main()
+ * {
+ * block_var = 0.0;
+ * }
+ *
+ * This takes place after the shader code has already been verified with
+ * the interface name in place.
+ *
+ * The linking phase will use the interface name rather than the
+ * interface's instance name.
+ *
+ * This modification to the ir allows our currently existing dead code
+ * elimination to work with interface blocks without changes.
+ */
+
+#include "glsl_symbol_table.h"
+#include "ir.h"
+#include "ir_optimization.h"
+#include "ir_rvalue_visitor.h"
+#include "program/hash_table.h"
+
+class flatten_named_interface_blocks_declarations : public ir_rvalue_visitor
+{
+public:
+ void * const mem_ctx;
+ hash_table *interface_namespace;
+
+ flatten_named_interface_blocks_declarations(void *mem_ctx)
+ : mem_ctx(mem_ctx)
+ {
+ }
+
+ voi
[Mesa-dev] [PATCH v2 10/15] glsl linker: cross validate interface block types between stages
Signed-off-by: Jordan Justen
---
src/glsl/linker.cpp | 13 +
1 file changed, 13 insertions(+)
diff --git a/src/glsl/linker.cpp b/src/glsl/linker.cpp
index 29856b0..acbdd2f 100644
--- a/src/glsl/linker.cpp
+++ b/src/glsl/linker.cpp
@@ -478,6 +478,7 @@ cross_validate_globals(struct gl_shader_program *prog,
* them.
*/
glsl_symbol_table variables;
+ glsl_symbol_table interfaces;
for (unsigned i = 0; i < num_shaders; i++) {
if (shader_list[i] == NULL)
continue;
@@ -488,6 +489,18 @@ cross_validate_globals(struct gl_shader_program *prog,
if (var == NULL)
continue;
+ if (var->is_in_uniform_block()) {
+const glsl_type *type =
interfaces.get_type(var->interface_type->name);
+if (type == NULL) {
+ interfaces.add_type(var->interface_type->name,
var->interface_type);
+} else if (type != var->interface_type) {
+ linker_error(prog, "interface block `%s' is inconsistent "
+"between shader stages\n",
+type->name);
+ return false;
+}
+ }
+
if (uniforms_only && (var->mode != ir_var_uniform))
continue;
--
1.7.10.4
___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/mesa-dev
[Mesa-dev] [PATCH v2 12/15] glsl link_varyings: link interface blocks using the block name
Signed-off-by: Jordan Justen
---
src/glsl/link_varyings.cpp | 33 +
1 file changed, 29 insertions(+), 4 deletions(-)
diff --git a/src/glsl/link_varyings.cpp b/src/glsl/link_varyings.cpp
index 04c9fdd..b374049 100644
--- a/src/glsl/link_varyings.cpp
+++ b/src/glsl/link_varyings.cpp
@@ -965,6 +965,8 @@ assign_varying_locations(struct gl_context *ctx,
= hash_table_ctor(0, hash_table_string_hash, hash_table_string_compare);
hash_table *consumer_inputs
= hash_table_ctor(0, hash_table_string_hash, hash_table_string_compare);
+ hash_table *consumer_interface_inputs
+ = hash_table_ctor(0, hash_table_string_hash, hash_table_string_compare);
/* Operate in a total of three passes.
*
@@ -983,8 +985,17 @@ assign_varying_locations(struct gl_context *ctx,
((ir_instruction *) node)->as_variable();
if ((input_var != NULL) && (input_var->mode == ir_var_shader_in)) {
-hash_table_insert(consumer_inputs, input_var,
- ralloc_strdup(mem_ctx, input_var->name));
+if (input_var->interface_type != NULL) {
+ char *const iface_field_name =
+ ralloc_asprintf(mem_ctx, "%s.%s",
+ input_var->interface_type->name,
+ input_var->name);
+ hash_table_insert(consumer_interface_inputs, input_var,
+ iface_field_name);
+} else {
+ hash_table_insert(consumer_inputs, input_var,
+ ralloc_strdup(mem_ctx, input_var->name));
+}
}
}
}
@@ -998,8 +1009,19 @@ assign_varying_locations(struct gl_context *ctx,
tfeedback_candidate_generator g(mem_ctx, tfeedback_candidates);
g.process(output_var);
- ir_variable *input_var =
- (ir_variable *) hash_table_find(consumer_inputs, output_var->name);
+ ir_variable *input_var;
+ if (output_var->interface_type != NULL) {
+ char *const iface_field_name =
+ralloc_asprintf(mem_ctx, "%s.%s",
+output_var->interface_type->name,
+output_var->name);
+ input_var =
+(ir_variable *) hash_table_find(consumer_interface_inputs,
+iface_field_name);
+ } else {
+ input_var =
+(ir_variable *) hash_table_find(consumer_inputs, output_var->name);
+ }
if (input_var && input_var->mode != ir_var_shader_in)
input_var = NULL;
@@ -1019,6 +1041,7 @@ assign_varying_locations(struct gl_context *ctx,
if (matched_candidate == NULL) {
hash_table_dtor(tfeedback_candidates);
hash_table_dtor(consumer_inputs);
+ hash_table_dtor(consumer_interface_inputs);
return false;
}
@@ -1036,12 +1059,14 @@ assign_varying_locations(struct gl_context *ctx,
if (!tfeedback_decls[i].assign_location(ctx, prog)) {
hash_table_dtor(tfeedback_candidates);
hash_table_dtor(consumer_inputs);
+ hash_table_dtor(consumer_interface_inputs);
return false;
}
}
hash_table_dtor(tfeedback_candidates);
hash_table_dtor(consumer_inputs);
+ hash_table_dtor(consumer_interface_inputs);
if (ctx->Const.DisableVaryingPacking) {
/* Transform feedback code assumes varyings are packed, so if the driver
--
1.7.10.4
___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/mesa-dev
[Mesa-dev] [PATCH v2 13/15] glsl linker: support arrays of interface block instances
With this change we now support interface block arrays.
For example, cases like this:
out block_name {
float f;
} block_instance[2];
This allows Mesa to pass the piglit glsl-1.50 test:
* execution/interface-blocks-complex-vs-fs.shader_test
Signed-off-by: Jordan Justen
---
src/glsl/lower_named_interface_blocks.cpp | 53 -
1 file changed, 44 insertions(+), 9 deletions(-)
diff --git a/src/glsl/lower_named_interface_blocks.cpp
b/src/glsl/lower_named_interface_blocks.cpp
index 2e0c322..405e7a9 100644
--- a/src/glsl/lower_named_interface_blocks.cpp
+++ b/src/glsl/lower_named_interface_blocks.cpp
@@ -107,22 +107,47 @@
flatten_named_interface_blocks_declarations::run(exec_list *instructions)
if (var->mode == ir_var_uniform)
continue;
- const glsl_type *const t = var->type;
+ const glsl_type * iface_t = var->type;
+ const glsl_type * array_t = NULL;
exec_node *insert_pos = var;
- char *iface_field_name;
- for (unsigned i = 0; i < t->length; i++) {
-iface_field_name = ralloc_asprintf(mem_ctx, "%s.%s", t->name,
- t->fields.structure[i].name);
+
+ if (iface_t->is_array()) {
+array_t = iface_t;
+iface_t = array_t->fields.array;
+ }
+
+ assert (iface_t->is_interface());
+
+ for (unsigned i = 0; i < iface_t->length; i++) {
+const char * field_name = iface_t->fields.structure[i].name;
+char *iface_field_name =
+ ralloc_asprintf(mem_ctx, "%s.%s",
+ iface_t->name, field_name);
ir_variable *found_var =
(ir_variable *) hash_table_find(interface_namespace,
iface_field_name);
if (!found_var) {
ir_variable *new_var =
- new(mem_ctx) ir_variable(t->fields.structure[i].type,
- ralloc_strdup(mem_ctx,
t->fields.structure[i].name),
+ new(mem_ctx) ir_variable(iface_t->fields.structure[i].type,
+ ralloc_strdup(mem_ctx,
iface_t->fields.structure[i].name),
(ir_variable_mode) var->mode);
- new_var->interface_type = t;
+ if (array_t != NULL) {
+ const glsl_type *new_array_type =
+ glsl_type::get_array_instance(
+iface_t->fields.structure[i].type,
+array_t->length);
+ char *array_var_name =
+ ralloc_asprintf(mem_ctx, "%s[%d]",
+ new_var->name, array_t->length);
+ ir_variable *new_array_var =
+ new(mem_ctx) ir_variable(new_array_type,
+ array_var_name,
+ (ir_variable_mode) var->mode);
+ new_var = new_array_var;
+ }
+
+ new_var->interface_type = iface_t;
hash_table_insert(interface_namespace, new_var,
iface_field_name);
insert_pos->insert_after(new_var);
@@ -187,9 +212,19 @@
flatten_named_interface_blocks_declarations::handle_rvalue(ir_rvalue **rvalue)
(ir_variable *) hash_table_find(interface_namespace,
iface_field_name);
assert(found_var);
+
ir_dereference_variable *deref_var =
new(mem_ctx) ir_dereference_variable(found_var);
- *rvalue = deref_var;
+
+ ir_dereference_array *deref_array =
+ ir->record->as_dereference_array();
+ if (deref_array != NULL) {
+ *rvalue =
+new(mem_ctx) ir_dereference_array(deref_var,
+ deref_array->array_index);
+ } else {
+ *rvalue = deref_var;
+ }
}
}
--
1.7.10.4
___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/mesa-dev
[Mesa-dev] [PATCH v2 14/15] glsl linker: compare interface blocks during intrastage linking
Verify that interface blocks match when combining compilation
units at the same stage. (For example, when merging all vertex
shaders.)
Fixes piglit glsl-1.50 test:
* linker/interface-blocks-multiple-vs-member-count-mismatch.shader_test
Signed-off-by: Jordan Justen
---
src/glsl/Makefile.sources |1 +
src/glsl/interface_blocks.cpp | 90 +
src/glsl/interface_blocks.h | 32 +++
src/glsl/linker.cpp |6 +++
4 files changed, 129 insertions(+)
create mode 100644 src/glsl/interface_blocks.cpp
create mode 100644 src/glsl/interface_blocks.h
diff --git a/src/glsl/Makefile.sources b/src/glsl/Makefile.sources
index 86ae43e..ba37ad2 100644
--- a/src/glsl/Makefile.sources
+++ b/src/glsl/Makefile.sources
@@ -25,6 +25,7 @@ LIBGLSL_FILES = \
$(GLSL_SRCDIR)/glsl_types.cpp \
$(GLSL_SRCDIR)/glsl_symbol_table.cpp \
$(GLSL_SRCDIR)/hir_field_selection.cpp \
+ $(GLSL_SRCDIR)/interface_blocks.cpp \
$(GLSL_SRCDIR)/ir_basic_block.cpp \
$(GLSL_SRCDIR)/ir_builder.cpp \
$(GLSL_SRCDIR)/ir_clone.cpp \
diff --git a/src/glsl/interface_blocks.cpp b/src/glsl/interface_blocks.cpp
new file mode 100644
index 000..f3daccf
--- /dev/null
+++ b/src/glsl/interface_blocks.cpp
@@ -0,0 +1,90 @@
+/*
+ * Copyright (c) 2013 Intel Corporation
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice (including the next
+ * paragraph) shall be included in all copies or substantial portions of the
+ * Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
+ * DEALINGS IN THE SOFTWARE.
+ */
+
+/**
+ * \file interface_blocks.cpp
+ * GLSL interface blocks support
+ */
+
+#include "ir.h"
+#include "program/hash_table.h"
+
+bool
+cross_validate_interface_blocks(struct gl_shader **shader_list,
+unsigned num_shaders)
+{
+ bool ok = true;
+
+ assert(num_shaders >= 1);
+
+ hash_table *ifaces
+ = hash_table_ctor(0, hash_table_string_hash, hash_table_string_compare);
+ const char *iface_name = NULL;
+
+ for (unsigned int i = 0; ok && i < num_shaders; i++) {
+ if (shader_list[i] == NULL)
+ continue;
+
+ foreach_list(node, shader_list[i]->ir) {
+ ir_variable *var = ((ir_instruction *) node)->as_variable();
+ if (!var)
+continue;
+
+ const glsl_type *iface_type = var->interface_type;
+
+ if (iface_type == NULL)
+continue;
+
+ /* We've already looked at this interface within this shader
+ * compilation unit, so there is no need to check it again.
+ */
+ if (iface_name != NULL &&
+ strcmp(iface_name, iface_type->name) == 0)
+continue;
+
+ iface_name = iface_type->name;
+
+ const glsl_type *old_iface_type =
+(glsl_type *) hash_table_find(ifaces, iface_name);
+
+ /* This is the first time we've seen the interface, so save
+ * it into our hash table.
+ */
+ if (old_iface_type == NULL) {
+hash_table_insert(ifaces, (void*) iface_type, iface_name);
+continue;
+ }
+
+ ok &= old_iface_type == iface_type;
+ if (!ok)
+break;
+ }
+
+ iface_name = NULL;
+ }
+
+ hash_table_dtor(ifaces);
+
+ return ok;
+}
+
diff --git a/src/glsl/interface_blocks.h b/src/glsl/interface_blocks.h
new file mode 100644
index 000..76ab725
--- /dev/null
+++ b/src/glsl/interface_blocks.h
@@ -0,0 +1,32 @@
+/*
+ * Copyright (c) 2013 Intel Corporation
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice (including the next
+ * parag
[Mesa-dev] [PATCH v2 15/15] glsl linker: compare interface blocks during interstage linking
Verify that interface blocks match when linking separate shader
stages into a program.
Fixes piglit glsl-1.50 tests:
* linker/interface-blocks-vs-fs-member-count-mismatch.shader_test
* linker/interface-blocks-vs-fs-member-order-mismatch.shader_test
Signed-off-by: Jordan Justen
---
src/glsl/interface_blocks.cpp | 14 ++
src/glsl/interface_blocks.h |3 +++
src/glsl/linker.cpp |3 +++
3 files changed, 20 insertions(+)
diff --git a/src/glsl/interface_blocks.cpp b/src/glsl/interface_blocks.cpp
index f3daccf..4da40b0 100644
--- a/src/glsl/interface_blocks.cpp
+++ b/src/glsl/interface_blocks.cpp
@@ -88,3 +88,17 @@ cross_validate_interface_blocks(struct gl_shader
**shader_list,
return ok;
}
+bool
+validate_program_interface_blocks(const struct gl_shader_program *prog)
+{
+ bool ok = true;
+ struct gl_shader *shader_list[2];
+
+ shader_list[0] = prog->_LinkedShaders[MESA_SHADER_VERTEX];
+ shader_list[1] = prog->_LinkedShaders[MESA_SHADER_FRAGMENT];
+
+ ok = cross_validate_interface_blocks((struct gl_shader **) &shader_list, 2);
+
+ return ok;
+}
+
diff --git a/src/glsl/interface_blocks.h b/src/glsl/interface_blocks.h
index 76ab725..fa4f873 100644
--- a/src/glsl/interface_blocks.h
+++ b/src/glsl/interface_blocks.h
@@ -29,4 +29,7 @@ bool
cross_validate_interface_blocks(struct gl_shader **shader_list,
unsigned num_shaders);
+bool
+validate_program_interface_blocks(const struct gl_shader_program *prog);
+
#endif /* GLSL_INTERFACE_BLOCKS_H */
diff --git a/src/glsl/linker.cpp b/src/glsl/linker.cpp
index 5bc5132..45c60f6 100644
--- a/src/glsl/linker.cpp
+++ b/src/glsl/linker.cpp
@@ -1724,6 +1724,9 @@ link_shaders(struct gl_context *ctx, struct
gl_shader_program *prog)
sh);
}
+ if (!validate_program_interface_blocks(prog))
+ goto done;
+
/* Here begins the inter-stage linking phase. Some initial validation is
* performed, then locations are assigned for uniforms, attributes, and
* varyings.
--
1.7.10.4
___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/mesa-dev
Re: [Mesa-dev] [PATCH] i965: Align all depth miplevels to 8 in the X direction.
On 03/18/2013 03:56 PM, Eric Anholt wrote: On an INTEL_DEBUG=perf piglit run on IVB, reduces the instances of "HW workaround: blit" (the printouts from the misaligned-depth workaround blits) from 725 to 684 It doesn't totally eliminate the workaround blit, because we still have problems with Y offsets that we can't fix (since texturing can only align miplevels up to 2 or 4, not 8). No regressions on piglit/es3conform on IVB. --- src/mesa/drivers/dri/intel/intel_tex_layout.c | 10 +- 1 file changed, 9 insertions(+), 1 deletion(-) diff --git a/src/mesa/drivers/dri/intel/intel_tex_layout.c b/src/mesa/drivers/dri/intel/intel_tex_layout.c index 35030df..95242f8 100644 --- a/src/mesa/drivers/dri/intel/intel_tex_layout.c +++ b/src/mesa/drivers/dri/intel/intel_tex_layout.c @@ -77,7 +77,15 @@ intel_horizontal_texture_alignment_unit(struct intel_context *intel, if (format == MESA_FORMAT_S8) return 8; - if (intel->gen >= 7 && format == MESA_FORMAT_Z16) + /* The depth alignment requirements in the table above are for rendering to +* depth miplevels using the LOD control fields. We don't use LOD control +* fields, and instead use page offsets plus intra-tile x/y offsets, which +* require that the low 3 bits are zero. To reduce the number of x/y +* offset workaround blits we do, align the X to 8, which depth texturing +* can handle (sadly, it can't handle 8 in the Y direction). +*/ + if (intel->gen >= 4 && + _mesa_get_format_base_format(format) == GL_DEPTH_COMPONENT) return 8; return 4; I'm really excited to see this happen on Ivybridge. However, the HALIGN_8 surface state bit doesn't exist on Gen4-6, so this will break there. This patch but with intel->gen >= 7 instead of 4 would get a R-b. ___ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev
Re: [Mesa-dev] [PATCH 2/3] Add dri image entry point for creating image from fd
On Fri, Mar 01, 2013 at 02:03:20PM +0200, Ander Conselvan de Oliveira wrote:
> On 02/28/2013 03:52 AM, Kristian Høgsberg wrote:
> >---
> > include/GL/internal/dri_interface.h| 14 +++-
> > src/mesa/drivers/dri/intel/intel_regions.c | 33 +++
> > src/mesa/drivers/dri/intel/intel_regions.h | 6
> > src/mesa/drivers/dri/intel/intel_screen.c | 53
> > --
> > 4 files changed, 102 insertions(+), 4 deletions(-)
> >
>
> [...]
>
> >diff --git a/src/mesa/drivers/dri/intel/intel_screen.c
> >b/src/mesa/drivers/dri/intel/intel_screen.c
> >index 277b133..3ee0d69 100644
> >--- a/src/mesa/drivers/dri/intel/intel_screen.c
> >+++ b/src/mesa/drivers/dri/intel/intel_screen.c
> >@@ -546,6 +546,9 @@ intel_query_image(__DRIimage *image, int attrib, int
> >*value)
> > return false;
> >*value = image->planar_format->components;
> >return true;
> >+ case __DRI_IMAGE_ATTRIB_FD:
> >+ drm_intel_bo_gem_export_to_prime(image->region->bo, value);
> >+ return true;
>
> Need to check the return value of drm_intel_bo_gem_export_to_prime().
Right, thanks.
> [...]
>
> > .createImageFromRenderbuffer=
> > intel_create_image_from_renderbuffer,
> >@@ -705,6 +751,7 @@ static struct __DRIimageExtensionRec intelImageExtension
> >= {
> > .createImageFromNames = intel_create_image_from_names,
> > .fromPlanar = intel_from_planar,
> > .createImageFromTexture = intel_create_image_from_texture
> >+.createImageFromFd = intel_create_image_from_fd
>
> There's missing comma above, as you pointed out in IRC.
Yeah, that was a little odd, I think I had the compile fix in an uncommitted
patch.
> Other than this, the patch seems fine to me.
Thanks,
Kristian
___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/mesa-dev
Re: [Mesa-dev] [PATCH 3/3] wayland: Add prime fd passing as a buffer sharing mechanism
On Fri, Mar 01, 2013 at 02:06:03PM +0200, Ander Conselvan de Oliveira wrote:
> On 02/28/2013 03:52 AM, Kristian Høgsberg wrote:
> >diff --git a/src/egl/drivers/dri2/platform_wayland.c
> >b/src/egl/drivers/dri2/platform_wayland.c
> >index b5cd04a..1b42a98 100644
> >--- a/src/egl/drivers/dri2/platform_wayland.c
> >+++ b/src/egl/drivers/dri2/platform_wayland.c
> >@@ -451,6 +451,46 @@ static const struct wl_callback_listener frame_listener
> >= {
> > wayland_frame_callback
> > };
> >
> >+static void
> >+create_wl_buffer(struct dri2_egl_surface *dri2_surf)
> >+{
> >+ struct dri2_egl_display *dri2_dpy =
> >+ dri2_egl_display(dri2_surf->base.Resource.Display);
> >+ int fd;
> >+
> >+ if (dri2_surf->current->wl_buffer != NULL)
> >+ return;
> >+
> >+ if (dri2_dpy->capabilities & WL_DRM_CAPABILITY_PRIME) {
> [...]
> >+ } else {
>
> There's an indentation error here.
>
> There's also lines replacing tabs with spaces, but I don't know what
> is supposed to be the coding style of those files. Indent is done
> both ways in them.
Uhm, yeah... fixed. The mesa policy is to use spaces, but it's not
strictly enforced, but moving towards spaces can't be bad. It comes from
the .dir-locals.el in the mesa top-level directory which makes my emacs
do it.
> Anyway, the series looks good to me, feel free to add my R-b,
> although you might want someone else to review the intel_screen.c
> changes too. I'm not that familiar with that code.
Thanks,
Kristian
___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/mesa-dev
[Mesa-dev] Basic GLSL 1.50 support: new types & built-in functions
This series improves our basic GLSL 1.50 support. First, it makes 1.50 shaders using built-in functions work. Secondly, it adds the new types and built-ins from the ARB_texture_multisample extension (which are now part of core). It also makes it possible to run Unigine Valley and Heaven 4.0 with MESA_GL_VERSION_OVERRIDE=3.2 and MESA_GLSL_VERSION_OVERRIDE=150; i965 also needs shader_precompile=false to avoid an assertion failure which I haven't tracked down yet. ___ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev
[Mesa-dev] [PATCH 1/3] glsl: Bump standalone compiler versions to 1.50.
The version bumps are necessary in order to compile built-ins for 1.50.
Signed-off-by: Kenneth Graunke
---
src/glsl/builtins/tools/generate_builtins.py | 4 ++--
src/glsl/main.cpp| 2 +-
2 files changed, 3 insertions(+), 3 deletions(-)
diff --git a/src/glsl/builtins/tools/generate_builtins.py
b/src/glsl/builtins/tools/generate_builtins.py
index 6da29c0..748a689 100755
--- a/src/glsl/builtins/tools/generate_builtins.py
+++ b/src/glsl/builtins/tools/generate_builtins.py
@@ -174,7 +174,7 @@ read_builtins(GLenum target, const char *protos, const char
**functions, unsigne
{
struct gl_context fakeCtx;
fakeCtx.API = API_OPENGL_COMPAT;
- fakeCtx.Const.GLSLVersion = 140;
+ fakeCtx.Const.GLSLVersion = 150;
fakeCtx.Extensions.ARB_ES2_compatibility = true;
fakeCtx.Extensions.ARB_ES3_compatibility = true;
fakeCtx.Const.ForceGLSLExtensionsWarn = false;
@@ -182,7 +182,7 @@ read_builtins(GLenum target, const char *protos, const char
**functions, unsigne
struct _mesa_glsl_parse_state *st =
new(sh) _mesa_glsl_parse_state(&fakeCtx, target, sh);
- st->language_version = 140;
+ st->language_version = 150;
st->symbols->separate_function_namespace = false;
st->ARB_texture_rectangle_enable = true;
st->EXT_texture_array_enable = true;
diff --git a/src/glsl/main.cpp b/src/glsl/main.cpp
index ed6f122..ce084b4 100644
--- a/src/glsl/main.cpp
+++ b/src/glsl/main.cpp
@@ -47,7 +47,7 @@ initialize_context(struct gl_context *ctx, gl_api api)
/* The standalone compiler needs to claim support for almost
* everything in order to compile the built-in functions.
*/
- ctx->Const.GLSLVersion = 140;
+ ctx->Const.GLSLVersion = 150;
ctx->Extensions.ARB_ES3_compatibility = true;
ctx->Const.MaxClipPlanes = 8;
--
1.8.2
___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/mesa-dev
[Mesa-dev] [PATCH 2/3] glsl: Add sampler2DMS/sampler2DMSArray types to GLSL 1.50.
GLSL 1.50 includes support for the new sampler types introduced by
the ARB_texture_multisample extension.
Signed-off-by: Kenneth Graunke
---
src/glsl/glsl_types.cpp | 12 +++-
src/glsl/glsl_types.h | 1 +
2 files changed, 12 insertions(+), 1 deletion(-)
diff --git a/src/glsl/glsl_types.cpp b/src/glsl/glsl_types.cpp
index a783dcc..8b0a248 100644
--- a/src/glsl/glsl_types.cpp
+++ b/src/glsl/glsl_types.cpp
@@ -292,6 +292,14 @@ glsl_type::generate_140_types(glsl_symbol_table *symtab)
void
+glsl_type::generate_150_types(glsl_symbol_table *symtab)
+{
+ generate_140_types(symtab);
+ generate_ARB_texture_multisample_types(symtab, false);
+}
+
+
+void
glsl_type::generate_ARB_texture_rectangle_types(glsl_symbol_table *symtab,
bool warn)
{
@@ -385,9 +393,11 @@ _mesa_glsl_initialize_types(struct _mesa_glsl_parse_state
*state)
glsl_type::generate_130_types(state->symbols, true, skip_1d);
break;
case 140:
- case 150:
glsl_type::generate_140_types(state->symbols);
break;
+ case 150:
+ glsl_type::generate_150_types(state->symbols);
+ break;
default:
assert(!"Unexpected language version");
break;
diff --git a/src/glsl/glsl_types.h b/src/glsl/glsl_types.h
index 7930426..2f3b19f 100644
--- a/src/glsl/glsl_types.h
+++ b/src/glsl/glsl_types.h
@@ -583,6 +583,7 @@ private:
static void generate_130_types(glsl_symbol_table *, bool add_deprecated,
bool skip_1d);
static void generate_140_types(glsl_symbol_table *);
+ static void generate_150_types(glsl_symbol_table *);
static void generate_ARB_texture_rectangle_types(glsl_symbol_table *, bool);
static void generate_EXT_texture_array_types(glsl_symbol_table *, bool);
static void generate_OES_texture_3D_types(glsl_symbol_table *, bool);
--
1.8.2
___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/mesa-dev
[Mesa-dev] [PATCH 3/3] glsl: Add built-in functions for GLSL 1.50.
This makes basic built-in functions work in GLSL 1.50. It supports everything except the new Geometry Shader functions. The new 150.glsl file is 140.glsl plus ARB_texture_multisample.glsl; 150.frag is identical to 140.frag except for the #version bump. Signed-off-by: Kenneth Graunke --- src/glsl/builtins/profiles/150.frag | 140 + src/glsl/builtins/profiles/150.glsl | 1005 +++ 2 files changed, 1145 insertions(+) create mode 100644 src/glsl/builtins/profiles/150.frag create mode 100644 src/glsl/builtins/profiles/150.glsl diff --git a/src/glsl/builtins/profiles/150.frag b/src/glsl/builtins/profiles/150.frag new file mode 100644 index 000..7798e25 --- /dev/null +++ b/src/glsl/builtins/profiles/150.frag @@ -0,0 +1,140 @@ +#version 150 +/* texture - bias variants */ + vec4 texture( sampler1D sampler, float P, float bias); +ivec4 texture(isampler1D sampler, float P, float bias); +uvec4 texture(usampler1D sampler, float P, float bias); + + vec4 texture( sampler2D sampler, vec2 P, float bias); +ivec4 texture(isampler2D sampler, vec2 P, float bias); +uvec4 texture(usampler2D sampler, vec2 P, float bias); + + vec4 texture( sampler3D sampler, vec3 P, float bias); +ivec4 texture(isampler3D sampler, vec3 P, float bias); +uvec4 texture(usampler3D sampler, vec3 P, float bias); + + vec4 texture( samplerCube sampler, vec3 P, float bias); +ivec4 texture(isamplerCube sampler, vec3 P, float bias); +uvec4 texture(usamplerCube sampler, vec3 P, float bias); + +float texture(sampler1DShadow sampler, vec3 P, float bias); +float texture(sampler2DShadow sampler, vec3 P, float bias); +float texture(samplerCubeShadow sampler, vec4 P, float bias); + + vec4 texture( sampler1DArray sampler, vec2 P, float bias); +ivec4 texture(isampler1DArray sampler, vec2 P, float bias); +uvec4 texture(usampler1DArray sampler, vec2 P, float bias); + + vec4 texture( sampler2DArray sampler, vec3 P, float bias); +ivec4 texture(isampler2DArray sampler, vec3 P, float bias); +uvec4 texture(usampler2DArray sampler, vec3 P, float bias); + +float texture(sampler1DArrayShadow sampler, vec3 P, float bias); + +/* textureProj - bias variants */ + vec4 textureProj( sampler1D sampler, vec2 P, float bias); +ivec4 textureProj(isampler1D sampler, vec2 P, float bias); +uvec4 textureProj(usampler1D sampler, vec2 P, float bias); + vec4 textureProj( sampler1D sampler, vec4 P, float bias); +ivec4 textureProj(isampler1D sampler, vec4 P, float bias); +uvec4 textureProj(usampler1D sampler, vec4 P, float bias); + + vec4 textureProj( sampler2D sampler, vec3 P, float bias); +ivec4 textureProj(isampler2D sampler, vec3 P, float bias); +uvec4 textureProj(usampler2D sampler, vec3 P, float bias); + vec4 textureProj( sampler2D sampler, vec4 P, float bias); +ivec4 textureProj(isampler2D sampler, vec4 P, float bias); +uvec4 textureProj(usampler2D sampler, vec4 P, float bias); + + vec4 textureProj( sampler3D sampler, vec4 P, float bias); +ivec4 textureProj(isampler3D sampler, vec4 P, float bias); +uvec4 textureProj(usampler3D sampler, vec4 P, float bias); + +float textureProj(sampler1DShadow sampler, vec4 P, float bias); +float textureProj(sampler2DShadow sampler, vec4 P, float bias); + +/* textureOffset - bias variants */ + vec4 textureOffset( sampler1D sampler, float P, int offset, float bias); +ivec4 textureOffset(isampler1D sampler, float P, int offset, float bias); +uvec4 textureOffset(usampler1D sampler, float P, int offset, float bias); + + vec4 textureOffset( sampler2D sampler, vec2 P, ivec2 offset, float bias); +ivec4 textureOffset(isampler2D sampler, vec2 P, ivec2 offset, float bias); +uvec4 textureOffset(usampler2D sampler, vec2 P, ivec2 offset, float bias); + + vec4 textureOffset( sampler3D sampler, vec3 P, ivec3 offset, float bias); +ivec4 textureOffset(isampler3D sampler, vec3 P, ivec3 offset, float bias); +uvec4 textureOffset(usampler3D sampler, vec3 P, ivec3 offset, float bias); + +float textureOffset(sampler1DShadow sampler, vec3 P, int offset, float bias); +float textureOffset(sampler2DShadow sampler, vec3 P, ivec2 offset, float bias); + + vec4 textureOffset( sampler1DArray sampler, vec2 P, int offset, float bias); +ivec4 textureOffset(isampler1DArray sampler, vec2 P, int offset, float bias); +uvec4 textureOffset(usampler1DArray sampler, vec2 P, int offset, float bias); + + vec4 textureOffset( sampler2DArray sampler, vec3 P, ivec2 offset, float bias); +ivec4 textureOffset(isampler2DArray sampler, vec3 P, ivec2 offset, float bias); +uvec4 textureOffset(usampler2DArray sampler, vec3 P, ivec2 offset, float bias); + +float textureOffset(sampler1DArrayShadow samp, vec3 P, int offset, float bias); + +/* textureProjOffsetOffset - bias variants */ + vec4 textureProjOffset( sampler1D sampler, vec2 P, int offset, float bias); +ivec4 textureProjOffset(isampler1D sampler, vec2 P, int offset, float bias); +uvec4 textureProjOffset(usampler1D sampler, vec2 P, int offset, float bias); + vec4 textureProjOffset( sampler1D sampler, vec4 P
Re: [Mesa-dev] Google Summer of Code ideas needed
On Mon, Mar 18, 2013 at 01:45:26PM -0700, Ian Romanick wrote: > On 03/13/2013 10:11 AM, Tom Stellard wrote: > > Hi, > > > > It's time again for Google Summer of Code, so we need to start updating > > the X.Org ideas page (http://www.x.org/wiki/SummerOfCodeIdeas) with new > > ideas. Since there have been a few issues with the wikis lately, if you > > have any ideas please respond to this thread, and I will make sure they > > get onto the official ideas page (but still feel free to update the wiki > > page yourself if you can). A good project description should contain: > > I've been trying to update the wiki... I don't seem to have an account > (seems unpossible), and the account creation page is broken. > > https://bugs.freedesktop.org/show_bug.cgi?id=62383 > This is due to the recent spam problems on the wiki, keithp can re-enable your account. > In the mean time, I have a couple ideas that someone could add. > > #1. Find common patterns in real GLSL shaders. > > Difficulty: Medium > Skills Required: C, C++ > Helpful, but optional skills: GLSL, compilers > Possible Mentor: IanRomanick (idr on IRC) > Where to ask questions: mesa-dev@lists.freedesktop.org, #dri-devel on > irc.freedesktop.org > Descrition: > > Using Mesa's stand-along GLSL compiler as a basis, generate a database > of IR from a large number of existing shaders (e.g., from shaderdb). > Write a piece of software that will mine this database for "large" > patterns that commonly occur in shaders. This information will be used > by people working on the GLSL compiler to improve code generation for > these sequences. > > #2. Improved application of GLSL complier optimizations > > Difficulty: Easy > Skills Required: C, C++ > Helpful, but optional skills: GLSL, compilers > Possible Mentor: IanRomanick (idr on IRC) > Where to ask questions: mesa-dev@lists.freedesktop.org, #dri-devel on > irc.freedesktop.org > Descrition: > > Mesa's GLSL compiler contains a large number of optimization passes. > Each pass may change the code of a shader, and this may result in > opportunities for other passes to make more changes. As a result, we > run all of our optimization passes in a loop until the shader code > stabilizes. This is expensive, and, though we have never observed this > in the wild, it is possible that a shader may never stabilize. > > Find a static ordering, with possible repeats, of optimization passes > that does not compromise the quality of the generated code. Measure the > before and after speed of compiling a large set of real-world shaders. > Thanks, I've added your ideas to the wiki page. -Tom > > - A brief description of the project > > - A difficulty rating (e.g. easy, medium, hard) > > - The skills / programming languages required > > > > Also, I am going to purge all the old ideas from the ideas page in the > > next week, so if there are any of the old ideas that you think are > > still relevant, let me know and I will keep it. > > > > The ideas page is used as one of the criteria by Google for selecting > > mentoring organizations and part of the reason X.Org was not selected > > last year was that the ideas page was not up to par, so if we want to > > participate in Google Summer of Code this year, it is important we > > have a good ideas page with lots of ideas. > > > > Thanks, > > Tom Stellard > > ___ > > mesa-dev mailing list > > mesa-dev@lists.freedesktop.org > > http://lists.freedesktop.org/mailman/listinfo/mesa-dev > > > ___ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev
Re: [Mesa-dev] [PATCH] i965: Align all depth miplevels to 8 in the X direction.
Kenneth Graunke writes: > On 03/18/2013 03:56 PM, Eric Anholt wrote: >> On an INTEL_DEBUG=perf piglit run on IVB, reduces the instances of "HW >> workaround: blit" (the printouts from the misaligned-depth workaround >> blits) from 725 to 684 >> >> It doesn't totally eliminate the workaround blit, because we still have >> problems with Y offsets that we can't fix (since texturing can only align >> miplevels up to 2 or 4, not 8). >> >> No regressions on piglit/es3conform on IVB. >> --- >> src/mesa/drivers/dri/intel/intel_tex_layout.c | 10 +- >> 1 file changed, 9 insertions(+), 1 deletion(-) >> >> diff --git a/src/mesa/drivers/dri/intel/intel_tex_layout.c >> b/src/mesa/drivers/dri/intel/intel_tex_layout.c >> index 35030df..95242f8 100644 >> --- a/src/mesa/drivers/dri/intel/intel_tex_layout.c >> +++ b/src/mesa/drivers/dri/intel/intel_tex_layout.c >> @@ -77,7 +77,15 @@ intel_horizontal_texture_alignment_unit(struct >> intel_context *intel, >> if (format == MESA_FORMAT_S8) >> return 8; >> >> - if (intel->gen >= 7 && format == MESA_FORMAT_Z16) >> + /* The depth alignment requirements in the table above are for rendering >> to >> +* depth miplevels using the LOD control fields. We don't use LOD >> control >> +* fields, and instead use page offsets plus intra-tile x/y offsets, >> which >> +* require that the low 3 bits are zero. To reduce the number of x/y >> +* offset workaround blits we do, align the X to 8, which depth texturing >> +* can handle (sadly, it can't handle 8 in the Y direction). >> +*/ >> + if (intel->gen >= 4 && >> + _mesa_get_format_base_format(format) == GL_DEPTH_COMPONENT) >> return 8; >> >> return 4; > > I'm really excited to see this happen on Ivybridge. > > However, the HALIGN_8 surface state bit doesn't exist on Gen4-6, so this > will break there. > > This patch but with intel->gen >= 7 instead of 4 would get a R-b. Good catch, thanks. pgp8UfMS0i1Db.pgp Description: PGP signature ___ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev
[Mesa-dev] [RFC] preparation for qxl sw wrapper driver
So we have this virtual GPU with nothing approaching a 3D engine, so we are currently running llvmpipe with drisw on it. However this incurs some overheads that now that we have a kernel driver, I believe we can remove. The main overheads are putimage for all rendering from a 3D compositor and getimage calls for all texture-from-pixmap operations. So implementing a dri2 qxl driver that wraps llvmpipe operations using the sw winsys wrapper, seemed like a good idea. So I ran into two problems implementing this, a) QXL rendering upside down, so when you texture from pixmap, what you get is upside down. Now we can handle this my supporting negative strides in enough places so texture sampling works. I've done this in patch 1, it just changes a bunch of unsigned->int, and works for me with the QXL driver once it adjust the transfer map to point at the other end of the texture. b) So DRI2 expects the hw/kernel to deal with the TFP ordering of the DDX rendering and the 3D driver using the output. However when your 3D driver is actually a sw renderer, you need some way to make sure the rendering on the GPU is finished before sampling. Now I expected we'd get a texture transfer for every access, but it appears we set the texture transfers up for dri2 buffers once in their lifetime, which kinda makes sense. So I saw there is an update_tex_buffer hook that looked like I could use it. So I added new hooks into the pipe screen and sw winsys to make it all happen. These are both RFCs, I'd like to know if people have any better ideas on how to handle these, I think the second case might be something someone who knows the dri state trackers a bit better might come up with something decent for. The first case I do feel like handling negative strides is useful in some cases, but I'm not sure how far we'd need to expand it (I don't think rendering to negative strides works) and test coverage could be fun. Dave. ___ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev
[Mesa-dev] [PATCH 2/2] gallium: add texture update hook to screen
From: Dave Airlie
We have cases on virtual GPU hw that renders using a wrapped llvmpipe,
that we can't produce the normal DRI2 semantics having the kernel/hw
enforce synchronisation semantics. It is also suboptimal to flush
all surfaces on the server side in the block handler. So there is
already a dri/st hook to update the texture, so create a screen interface
and call into it for the texture resource. The sw driver can then
call into the displaytarget wrapper which will call into the outer
wrapper driver to actually cause the flush to happen.
This contains the llvmpipe and wrapper sw changes.
Signed-off-by: Dave Airlie
---
src/gallium/drivers/llvmpipe/lp_texture.c | 13 +
src/gallium/drivers/softpipe/sp_texture.h | 2 +-
src/gallium/include/pipe/p_screen.h | 3 +++
src/gallium/include/state_tracker/sw_winsys.h | 6 ++
src/gallium/state_trackers/dri/drm/dri2.c | 5 -
src/gallium/winsys/sw/wrapper/wrapper_sw_winsys.c | 11 +++
6 files changed, 38 insertions(+), 2 deletions(-)
diff --git a/src/gallium/drivers/llvmpipe/lp_texture.c
b/src/gallium/drivers/llvmpipe/lp_texture.c
index f923292..48242d5 100644
--- a/src/gallium/drivers/llvmpipe/lp_texture.c
+++ b/src/gallium/drivers/llvmpipe/lp_texture.c
@@ -1375,6 +1375,18 @@ llvmpipe_print_resources(void)
}
#endif
+static void
+llvmpipe_texture_update(struct pipe_screen *pscreen,
+struct pipe_resource *pt)
+{
+ struct llvmpipe_screen *screen = llvmpipe_screen(pscreen);
+ struct llvmpipe_resource *lpr = llvmpipe_resource(pt);
+ struct sw_winsys *winsys = screen->winsys;
+ if (!lpr->dt)
+ return;
+
+ winsys->displaytarget_update(winsys, lpr->dt);
+}
void
llvmpipe_init_screen_resource_funcs(struct pipe_screen *screen)
@@ -1396,6 +1408,7 @@ llvmpipe_init_screen_resource_funcs(struct pipe_screen
*screen)
screen->resource_from_handle = llvmpipe_resource_from_handle;
screen->resource_get_handle = llvmpipe_resource_get_handle;
screen->can_create_resource = llvmpipe_can_create_resource;
+ screen->texture_update = llvmpipe_texture_update;
}
diff --git a/src/gallium/drivers/softpipe/sp_texture.h
b/src/gallium/drivers/softpipe/sp_texture.h
index 533d625..04c6eaf 100644
--- a/src/gallium/drivers/softpipe/sp_texture.h
+++ b/src/gallium/drivers/softpipe/sp_texture.h
@@ -46,7 +46,7 @@ struct softpipe_resource
struct pipe_resource base;
unsigned long level_offset[SP_MAX_TEXTURE_2D_LEVELS];
- unsigned stride[SP_MAX_TEXTURE_2D_LEVELS];
+ int stride[SP_MAX_TEXTURE_2D_LEVELS];
/**
* Display target, only valid for PIPE_TEXTURE_2D with the
diff --git a/src/gallium/include/pipe/p_screen.h
b/src/gallium/include/pipe/p_screen.h
index 754cfd7..680274c 100644
--- a/src/gallium/include/pipe/p_screen.h
+++ b/src/gallium/include/pipe/p_screen.h
@@ -206,6 +206,9 @@ struct pipe_screen {
struct pipe_fence_handle *fence,
uint64_t timeout );
+
+ void (*texture_update)(struct pipe_screen *,
+ struct pipe_resource *res);
};
diff --git a/src/gallium/include/state_tracker/sw_winsys.h
b/src/gallium/include/state_tracker/sw_winsys.h
index 71d2943..880a6d5 100644
--- a/src/gallium/include/state_tracker/sw_winsys.h
+++ b/src/gallium/include/state_tracker/sw_winsys.h
@@ -130,6 +130,12 @@ struct sw_winsys
(*displaytarget_display)( struct sw_winsys *ws,
struct sw_displaytarget *dt,
void *context_private );
+
+ /* some virtual GPUs have got hw to do sync with so add an
+ interface to allow it to happen */
+ void
+ (*displaytarget_update)( struct sw_winsys *ws,
+struct sw_displaytarget *dt );
void
(*displaytarget_destroy)( struct sw_winsys *ws,
diff --git a/src/gallium/state_trackers/dri/drm/dri2.c
b/src/gallium/state_trackers/dri/drm/dri2.c
index f8d311c..bfcfe3f 100644
--- a/src/gallium/state_trackers/dri/drm/dri2.c
+++ b/src/gallium/state_trackers/dri/drm/dri2.c
@@ -433,7 +433,10 @@ dri2_update_tex_buffer(struct dri_drawable *drawable,
struct dri_context *ctx,
struct pipe_resource *res)
{
- /* no-op */
+ struct pipe_context *pipe = ctx->st->pipe;
+
+ if (pipe->screen->texture_update)
+ pipe->screen->texture_update(pipe->screen, res);
}
static __DRIimage *
diff --git a/src/gallium/winsys/sw/wrapper/wrapper_sw_winsys.c
b/src/gallium/winsys/sw/wrapper/wrapper_sw_winsys.c
index 3761ae0..71d64c5 100644
--- a/src/gallium/winsys/sw/wrapper/wrapper_sw_winsys.c
+++ b/src/gallium/winsys/sw/wrapper/wrapper_sw_winsys.c
@@ -246,6 +246,16 @@ wsw_dt_unmap(struct sw_winsys *ws,
wdt->transfer = NULL;
}
+static void
+wsw_dt_update( struct sw_winsys *ws,
+ struct sw_displaytarget *dt )
+{
+ struct wrapper_sw_displaytarget *wdt = wrapper_sw_displaytarget(
[Mesa-dev] [PATCH 1/2] gallium/sw: allow for negative strides in some places
From: Dave Airlie
This is for a specific use-case for the QXL driver (which wraps
llvmpipe/softpipe). Texture it get via TFP are inverted, so
we need to use a negative stride to sample from them. So
this converts a bunch of cases where we use unsigned stride
to a signed value. I can't see the single bit less limiting
us anytime soon, and so far the underlying llvm code has
just worked properly and my texture samples end the right way up.
Signed-off-by: Dave Airlie
---
src/gallium/auxiliary/draw/draw_context.c | 2 +-
src/gallium/auxiliary/draw/draw_context.h | 2 +-
src/gallium/auxiliary/draw/draw_llvm.c| 2 +-
src/gallium/auxiliary/draw/draw_llvm.h| 4 ++--
src/gallium/drivers/llvmpipe/lp_jit.h | 2 +-
src/gallium/drivers/llvmpipe/lp_scene.h | 2 +-
src/gallium/drivers/llvmpipe/lp_state_sampler.c | 2 +-
src/gallium/drivers/llvmpipe/lp_texture.h | 2 +-
src/gallium/include/pipe/p_state.h| 2 +-
src/gallium/include/state_tracker/sw_winsys.h | 4 ++--
src/gallium/winsys/sw/wrapper/wrapper_sw_winsys.c | 8
src/gallium/winsys/sw/xlib/xlib_sw_winsys.c | 6 +++---
12 files changed, 19 insertions(+), 19 deletions(-)
diff --git a/src/gallium/auxiliary/draw/draw_context.c
b/src/gallium/auxiliary/draw/draw_context.c
index 045bb6b..0903cfb 100644
--- a/src/gallium/auxiliary/draw/draw_context.c
+++ b/src/gallium/auxiliary/draw/draw_context.c
@@ -807,7 +807,7 @@ draw_set_mapped_texture(struct draw_context *draw,
uint32_t width, uint32_t height, uint32_t depth,
uint32_t first_level, uint32_t last_level,
const void *base_ptr,
-uint32_t row_stride[PIPE_MAX_TEXTURE_LEVELS],
+int32_t row_stride[PIPE_MAX_TEXTURE_LEVELS],
uint32_t img_stride[PIPE_MAX_TEXTURE_LEVELS],
uint32_t mip_offsets[PIPE_MAX_TEXTURE_LEVELS])
{
diff --git a/src/gallium/auxiliary/draw/draw_context.h
b/src/gallium/auxiliary/draw/draw_context.h
index e8e2d94..4927ca4 100644
--- a/src/gallium/auxiliary/draw/draw_context.h
+++ b/src/gallium/auxiliary/draw/draw_context.h
@@ -155,7 +155,7 @@ draw_set_mapped_texture(struct draw_context *draw,
uint32_t width, uint32_t height, uint32_t depth,
uint32_t first_level, uint32_t last_level,
const void *base,
-uint32_t row_stride[PIPE_MAX_TEXTURE_LEVELS],
+int32_t row_stride[PIPE_MAX_TEXTURE_LEVELS],
uint32_t img_stride[PIPE_MAX_TEXTURE_LEVELS],
uint32_t mip_offsets[PIPE_MAX_TEXTURE_LEVELS]);
diff --git a/src/gallium/auxiliary/draw/draw_llvm.c
b/src/gallium/auxiliary/draw/draw_llvm.c
index 8e46687..3e7d497 100644
--- a/src/gallium/auxiliary/draw/draw_llvm.c
+++ b/src/gallium/auxiliary/draw/draw_llvm.c
@@ -1501,7 +1501,7 @@ draw_llvm_set_mapped_texture(struct draw_context *draw,
uint32_t width, uint32_t height, uint32_t depth,
uint32_t first_level, uint32_t last_level,
const void *base_ptr,
- uint32_t row_stride[PIPE_MAX_TEXTURE_LEVELS],
+ int32_t row_stride[PIPE_MAX_TEXTURE_LEVELS],
uint32_t img_stride[PIPE_MAX_TEXTURE_LEVELS],
uint32_t mip_offsets[PIPE_MAX_TEXTURE_LEVELS])
{
diff --git a/src/gallium/auxiliary/draw/draw_llvm.h
b/src/gallium/auxiliary/draw/draw_llvm.h
index c9f125b..748a01b 100644
--- a/src/gallium/auxiliary/draw/draw_llvm.h
+++ b/src/gallium/auxiliary/draw/draw_llvm.h
@@ -49,7 +49,7 @@ struct draw_jit_texture
uint32_t first_level;
uint32_t last_level;
const void *base;
- uint32_t row_stride[PIPE_MAX_TEXTURE_LEVELS];
+ int32_t row_stride[PIPE_MAX_TEXTURE_LEVELS];
uint32_t img_stride[PIPE_MAX_TEXTURE_LEVELS];
uint32_t mip_offsets[PIPE_MAX_TEXTURE_LEVELS];
};
@@ -337,7 +337,7 @@ draw_llvm_set_mapped_texture(struct draw_context *draw,
uint32_t width, uint32_t height, uint32_t depth,
uint32_t first_level, uint32_t last_level,
const void *base_ptr,
- uint32_t row_stride[PIPE_MAX_TEXTURE_LEVELS],
+ int32_t row_stride[PIPE_MAX_TEXTURE_LEVELS],
uint32_t img_stride[PIPE_MAX_TEXTURE_LEVELS],
uint32_t mip_offsets[PIPE_MAX_TEXTURE_LEVELS]);
diff --git a/src/gallium/drivers/llvmpipe/lp_jit.h
b/src/gallium/drivers/llvmpipe/lp_jit.h
index 4eddb2a..ea89124 100644
--- a/src/gallium/drivers/llvmpipe/lp_jit.h
+++ b/src/gallium/drivers/llvmpipe/lp_jit.h
@@ -55,7 +55,7 @@ struct lp_jit_texture
uint32_t first_level;
uint32_t last
Re: [Mesa-dev] Basic GLSL 1.50 support: new types & built-in functions
Reviewed-by: Jordan Justen On Mon, Mar 18, 2013 at 6:57 PM, Kenneth Graunke wrote: > This series improves our basic GLSL 1.50 support. First, it makes 1.50 > shaders using built-in functions work. Secondly, it adds the new types > and built-ins from the ARB_texture_multisample extension (which are now > part of core). > > It also makes it possible to run Unigine Valley and Heaven 4.0 with > MESA_GL_VERSION_OVERRIDE=3.2 and MESA_GLSL_VERSION_OVERRIDE=150; i965 > also needs shader_precompile=false to avoid an assertion failure which > I haven't tracked down yet. > > ___ > mesa-dev mailing list > mesa-dev@lists.freedesktop.org > http://lists.freedesktop.org/mailman/listinfo/mesa-dev ___ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev
