On 01/09/2015 11:07 PM, Kenneth Graunke wrote:
Sandybridge requires the post-sync non-zero workaround in a ton of
places, and if you ever miss one, the GPU usually hangs.
Currently, we try to track exactly when a workaround flush is
necessary (via the brw-batch.need_workaround_flush flag).
---
src/glsl/nir/nir.h | 37 +
1 file changed, 37 insertions(+)
diff --git a/src/glsl/nir/nir.h b/src/glsl/nir/nir.h
index 5ebfc5a..7b5794d 100644
--- a/src/glsl/nir/nir.h
+++ b/src/glsl/nir/nir.h
@@ -482,6 +482,43 @@ typedef struct {
bool is_ssa;
}
Designated initializers with anonymous unions don't work in MSVC or
GCC 4.6. With a couple of constructor methods, we don't need them any
more and the code is actually cleaner.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=88467
---
src/glsl/nir/nir_from_ssa.c | 30
https://bugs.freedesktop.org/show_bug.cgi?id=88467
--- Comment #11 from Jason Ekstrand ja...@jlekstrand.net ---
(In reply to Jason Ekstrand from comment #10)
(In reply to Dave Airlie from comment #6)
This is actually more annoying
with gcc 4.4
gcc -std=gnu99 supports anonymous
From: Ville Syrjälä ville.syrj...@linux.intel.com
max_wm_threads depends on the GT SKU on gen8. Update the values to
match the spec.
The max number of threads in 3DSTATE_PS is always programmed to 64
and the hardware internally scales that depending on the GT SKU. So
this doesn't change the max
From: Ville Syrjälä ville.syrj...@linux.intel.com
According to BSpec the correct number for min_vs_entries is 34 for CHV.
Signed-off-by: Ville Syrjälä ville.syrj...@linux.intel.com
---
src/mesa/drivers/dri/i965/brw_device_info.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git
NFC.
Signed-off-by: Jan Vesely jan.ves...@rutgers.edu
---
src/mesa/main/texcompress_bptc.c | 40 ++--
1 file changed, 22 insertions(+), 18 deletions(-)
diff --git a/src/mesa/main/texcompress_bptc.c b/src/mesa/main/texcompress_bptc.c
index c944ac2..56ca320
NFC.
Signed-off-by: Jan Vesely jan.ves...@rutgers.edu
---
src/mesa/state_tracker/st_extensions.c | 47 ++
1 file changed, 31 insertions(+), 16 deletions(-)
diff --git a/src/mesa/state_tracker/st_extensions.c
b/src/mesa/state_tracker/st_extensions.c
index
NFC.
Signed-off-by: Jan Vesely jan.ves...@rutgers.edu
---
This series fixes 38 -Wmissing-field-initializers warnings reported by gcc
src/mesa/main/get_hash_params.py | 8
1 file changed, 4 insertions(+), 4 deletions(-)
diff --git a/src/mesa/main/get_hash_params.py
https://bugs.freedesktop.org/show_bug.cgi?id=87886
Stéphane Travostino stephane.travost...@gmail.com changed:
What|Removed |Added
Status|NEEDINFO
From: Ville Syrjälä ville.syrj...@linux.intel.com
Increase the device info .urb.size for BDW GT3 and CHV to match the
default URB size for each.
Also add all missing platforms (BYT,BDW,CHV) to the comment describing
the default URB size in gen7_urb.c.
Signed-off-by: Ville Syrjälä
Assuming this actually compiles with GCC 4.4...
Reviewed-by: Connor Abbott cwabbo...@gmail.com
On Wed, Jan 21, 2015 at 2:14 PM, Jason Ekstrand ja...@jlekstrand.net wrote:
Designated initializers with anonymous unions don't work in MSVC or
GCC 4.6. With a couple of constructor methods, we
On 01/21/2015 10:33 AM, Jan Vesely wrote:
NFC.
NFC? I'm assuming this doesn't mean Near Field Communications or no
'fine' clue.
Signed-off-by: Jan Vesely jan.ves...@rutgers.edu
---
src/mesa/main/texcompress_bptc.c | 40
++--
1 file changed, 22
On Wednesday, January 21, 2015 08:17:34 PM ville.syrj...@linux.intel.com wrote:
From: Ville Syrjälä ville.syrj...@linux.intel.com
max_wm_threads depends on the GT SKU on gen8. Update the values to
match the spec.
The max number of threads in 3DSTATE_PS is always programmed to 64
and the
Reviewed-by: Connor Abbott cwabbo...@gmail.com
On Wed, Jan 21, 2015 at 2:14 PM, Jason Ekstrand ja...@jlekstrand.net wrote:
---
src/glsl/nir/nir.h | 37 +
1 file changed, 37 insertions(+)
diff --git a/src/glsl/nir/nir.h b/src/glsl/nir/nir.h
index
On Wed, Jan 21, 2015 at 1:12 PM, Tobias Klausmann
tobias.johannes.klausm...@mni.thm.de wrote:
Signed-off-by: Tobias Klausmann tobias.johannes.klausm...@mni.thm.de
---
src/glsl/glcpp/glcpp-parse.y | 19 ++-
1 file changed, 10 insertions(+), 9 deletions(-)
diff --git
On 21.01.2015 22:22, Matt Turner wrote:
On Wed, Jan 21, 2015 at 1:12 PM, Tobias Klausmann
tobias.johannes.klausm...@mni.thm.de wrote:
(snip)
We're messing up the indentation here. But really we want to get rid
of tabs, not add more. glcpp is unfortunately full of tabs.
Alright ignore this one
Signed-off-by: Tobias Klausmann tobias.johannes.klausm...@mni.thm.de
---
src/glsl/glcpp/glcpp-parse.y | 19 ++-
1 file changed, 10 insertions(+), 9 deletions(-)
diff --git a/src/glsl/glcpp/glcpp-parse.y b/src/glsl/glcpp/glcpp-parse.y
index e5bebe5..8b545da 100644
---
On Wednesday, January 21, 2015 08:17:35 PM ville.syrj...@linux.intel.com wrote:
From: Ville Syrjälä ville.syrj...@linux.intel.com
According to BSpec the correct number for min_vs_entries is 34 for CHV.
Signed-off-by: Ville Syrjälä ville.syrj...@linux.intel.com
---
Reviewed-by: Ian Romanick ian.d.roman...@intel.com
On 01/20/2015 09:45 PM, Tapani Pälli wrote:
Fixes currently failing Piglit case
interface-blocks-name-reused-globally.vert
v2: combine var declaration with assignment (Ian)
Signed-off-by: Tapani Pälli tapani.pa...@intel.com
---
On Tue, Jan 20, 2015 at 12:11 AM, Kenneth Graunke kenn...@whitecape.org wrote:
On Monday, January 19, 2015 03:31:05 PM Matt Turner wrote:
---
src/mesa/drivers/dri/i965/brw_eu.c | 22 ++
src/mesa/drivers/dri/i965/brw_eu.h | 1 +
2 files changed, 23 insertions(+)
diff
On Wed, Jan 21, 2015 at 2:52 PM, Kenneth Graunke kenn...@whitecape.org wrote:
On Wednesday, January 21, 2015 01:02:46 PM Matt Turner wrote:
On Tue, Jan 20, 2015 at 12:11 AM, Kenneth Graunke kenn...@whitecape.org
wrote:
On Monday, January 19, 2015 03:31:05 PM Matt Turner wrote:
---
On Wednesday, January 21, 2015 11:59:39 AM Chad Versace wrote:
On 01/09/2015 11:07 PM, Kenneth Graunke wrote:
Sandybridge requires the post-sync non-zero workaround in a ton of
places, and if you ever miss one, the GPU usually hangs.
Currently, we try to track exactly when a workaround
On Wed, Jan 21, 2015 at 3:19 PM, Kenneth Graunke kenn...@whitecape.org wrote:
Okay, I see what you're doing now. It's definitely not what my
brw_negate_cmod function did. This is why you absolutely need
doxygen comments, ideally with an example. The functions for
(a cmp1 b) - !(a cmp1 b)
On Wednesday, January 21, 2015 01:02:46 PM Matt Turner wrote:
On Tue, Jan 20, 2015 at 12:11 AM, Kenneth Graunke kenn...@whitecape.org
wrote:
On Monday, January 19, 2015 03:31:05 PM Matt Turner wrote:
---
src/mesa/drivers/dri/i965/brw_eu.c | 22 ++
On Wednesday, January 21, 2015 03:05:03 PM Matt Turner wrote:
On Wed, Jan 21, 2015 at 2:52 PM, Kenneth Graunke kenn...@whitecape.org
wrote:
On Wednesday, January 21, 2015 01:02:46 PM Matt Turner wrote:
On Tue, Jan 20, 2015 at 12:11 AM, Kenneth Graunke kenn...@whitecape.org
wrote:
On
On Wed, 2015-01-21 at 12:34 -0800, Ian Romanick wrote:
On 01/21/2015 10:33 AM, Jan Vesely wrote:
NFC.
NFC? I'm assuming this doesn't mean Near Field Communications or no
'fine' clue.
No functional change. seen that in other commits somwhere, thoght it was
universal
Signed-off-by: Jan
On Wed, Jan 21, 2015 at 3:49 PM, Jan Vesely jan.ves...@rutgers.edu wrote:
On Wed, 2015-01-21 at 12:34 -0800, Ian Romanick wrote:
On 01/21/2015 10:33 AM, Jan Vesely wrote:
NFC.
NFC? I'm assuming this doesn't mean Near Field Communications or no
'fine' clue.
No functional change. seen that
https://bugs.freedesktop.org/show_bug.cgi?id=84566
Jason Ekstrand ja...@jlekstrand.net changed:
What|Removed |Added
Status|NEW |RESOLVED
From: Dave Airlie airl...@redhat.com
anonymous structs/union don't work with c99 but do work with gnu99
on gcc 4.4.
This on top of Jason's designated initialisers changes, make
Mesa build on RHEL6 again.
Signed-off-by: Dave Airlie airl...@redhat.com
---
configure.ac | 12 +++-
1
On 01/21/2015 05:35 PM, Matt Turner wrote:
On Wed, Jan 21, 2015 at 5:28 PM, Dave Airlie airl...@gmail.com wrote:
From: Dave Airlie airl...@redhat.com
anonymous structs/union don't work with c99 but do work with gnu99
on gcc 4.4.
This on top of Jason's designated initialisers changes, make
On Wed, Jan 21, 2015 at 5:28 PM, Dave Airlie airl...@gmail.com wrote:
From: Dave Airlie airl...@redhat.com
anonymous structs/union don't work with c99 but do work with gnu99
on gcc 4.4.
This on top of Jason's designated initialisers changes, make
Mesa build on RHEL6 again.
Signed-off-by:
On 01/21/2015 03:05 PM, Matt Turner wrote:
On Wed, Jan 21, 2015 at 2:52 PM, Kenneth Graunke kenn...@whitecape.org
wrote:
On Wednesday, January 21, 2015 01:02:46 PM Matt Turner wrote:
On Tue, Jan 20, 2015 at 12:11 AM, Kenneth Graunke kenn...@whitecape.org
wrote:
On Monday, January 19, 2015
On Wed, Jan 21, 2015 at 6:57 PM, Ian Romanick i...@freedesktop.org wrote:
On 01/21/2015 03:05 PM, Matt Turner wrote:
On Wed, Jan 21, 2015 at 2:52 PM, Kenneth Graunke kenn...@whitecape.org
wrote:
On Wednesday, January 21, 2015 01:02:46 PM Matt Turner wrote:
Okay, I changed the name, and in
Reviewed-by: Connor Abbott cwabbo...@gmail.com
On Wed, Jan 21, 2015 at 11:23 PM, Matt Turner matts...@gmail.com wrote:
Fixes a couple of warnings in the process.
---
src/glsl/nir/glsl_to_nir.cpp | 57
--
src/glsl/nir/nir.h | 3
Fixes a couple of warnings in the process.
---
src/glsl/nir/glsl_to_nir.cpp | 57 --
src/glsl/nir/nir.h | 3 +-
src/glsl/nir/nir_lower_system_values.c | 3 +-
src/glsl/nir/nir_to_ssa.c | 2 +-
4 files changed, 22
Overall this looks correct. I've got a few nits below and I'd like to take
a look at it with fresh eyes before giving an R-B as it's complicated
especially with all of the stuff to handle non-ssa. Not sure if it's
really worth doing non-ssa now that I see how much more complicated it
makes
https://bugs.freedesktop.org/show_bug.cgi?id=87137
Matt Turner matts...@gmail.com changed:
What|Removed |Added
Status|NEW |RESOLVED
On Thursday, January 22, 2015 12:12:18 AM marius.pre...@intel.com wrote:
From: Marius Predut marius.pre...@intel.com
On 32-bit, for floating point operations is used x86 FPU registers instead
SSE,
reason for when reinterprets an integer as a float result is unexpected
(modify floats when
On 22 January 2015 at 06:19, Jason Ekstrand ja...@jlekstrand.net wrote:
On Wed, Jan 21, 2015 at 12:19 PM, Connor Abbott cwabbo...@gmail.com wrote:
Assuming this actually compiles with GCC 4.4...
I don't know for sure, but it certainly gets us closer
--Jason
Much closer,
CC
From: Marius Predut marius.pre...@intel.com
On 32-bit, for floating point operations is used x86 FPU registers instead SSE,
reason for when reinterprets an integer as a float result is unexpected
(modify floats when they are written to memory).
Bugzilla:
Reviewed-by: Connor Abbbott cwabbo...@gmail.com
On Wed, Jan 21, 2015 at 8:25 PM, Jason Ekstrand ja...@jlekstrand.net wrote:
---
src/glsl/nir/nir_search.c | 24 ++--
1 file changed, 10 insertions(+), 14 deletions(-)
diff --git a/src/glsl/nir/nir_search.c
From: Michel Dänzer michel.daen...@amd.com
That's what device::ir_target() returns. Fixes reading beyond allocated
memory:
==1936== Invalid read of size 1
==1936==at 0x4C2C1B4: strlen (vg_replace_strmem.c:412)
==1936==by 0x9E00C30: std::basic_stringchar, std::char_traitschar,
From: Michel Dänzer michel.daen...@amd.com
Fixes writing beyond the allocated buffer:
==31855== Invalid write of size 1
==31855==at 0x50AB2A9: vsprintf (iovsprintf.c:43)
==31855==by 0x508F6F6: sprintf (sprintf.c:32)
==31855==by 0xB59C7EC: r600_get_compute_param
On Wed, Jan 21, 2015 at 5:25 PM, Eric Anholt e...@anholt.net wrote:
It's nice to have this present in your default cases so you can see what
instruction is triggering an abort.
---
src/glsl/nir/nir.h | 1 +
src/glsl/nir/nir_print.c | 14 --
2 files changed, 13
---
src/glsl/Makefile.sources | 1 +
src/glsl/nir/nir.h | 2 +
src/glsl/nir/nir_lower_phis_to_scalar.c | 238
3 files changed, 241 insertions(+)
create mode 100644 src/glsl/nir/nir_lower_phis_to_scalar.c
diff --git
Now that we can scalarize with NIR, there's no need for all this code
anymore. Let's get rid of it and just do scalar operations.
---
src/mesa/drivers/dri/i965/brw_fs.h | 15 -
src/mesa/drivers/dri/i965/brw_fs_nir.cpp | 489 ++-
2 files changed, 158
---
src/mapi/glapi/gen/ARB_direct_state_access.xml | 5 ++
src/mesa/main/bufferobj.c | 87 ++
src/mesa/main/bufferobj.h | 7 +++
src/mesa/main/tests/dispatch_sanity.cpp| 1 +
4 files changed, 47 insertions(+), 53
---
src/mapi/glapi/gen/ARB_direct_state_access.xml | 18 +++
src/mesa/main/bufferobj.c | 156 ++---
src/mesa/main/bufferobj.h | 34 --
src/mesa/main/tests/dispatch_sanity.cpp| 2 +
---
src/mapi/glapi/gen/ARB_direct_state_access.xml | 6 ++
src/mesa/main/bufferobj.c | 76 +-
src/mesa/main/bufferobj.h | 10
src/mesa/main/tests/dispatch_sanity.cpp| 1 +
4 files changed, 67 insertions(+), 26
---
include/GL/glext.h | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/include/GL/glext.h b/include/GL/glext.h
index 3f141a2..925df18 100644
--- a/include/GL/glext.h
+++ b/include/GL/glext.h
@@ -2738,7 +2738,7 @@ GLAPI void APIENTRY glCreateBuffers (GLsizei n, GLuint
---
include/GL/glext.h | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/include/GL/glext.h b/include/GL/glext.h
index bca1b6e..cf0ea3f 100644
--- a/include/GL/glext.h
+++ b/include/GL/glext.h
@@ -2744,7 +2744,7 @@ GLAPI void APIENTRY glClearNamedBufferSubData (GLuint
buffer,
---
include/GL/glext.h | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/include/GL/glext.h b/include/GL/glext.h
index cf0ea3f..ceb8b2f 100644
--- a/include/GL/glext.h
+++ b/include/GL/glext.h
@@ -2748,7 +2748,7 @@ GLAPI void APIENTRY glFlushMappedNamedBufferRange (GLuint
---
src/mapi/glapi/gen/ARB_direct_state_access.xml | 6 ++
src/mesa/main/bufferobj.c | 27 +++---
src/mesa/main/bufferobj.h | 4
src/mesa/main/tests/dispatch_sanity.cpp| 1 +
4 files changed, 35 insertions(+), 3
---
src/mapi/glapi/gen/ARB_direct_state_access.xml | 12 ++
src/mesa/main/bufferobj.c | 170 +
src/mesa/main/bufferobj.h | 7 +
src/mesa/main/tests/dispatch_sanity.cpp| 2 +
4 files changed, 111 insertions(+), 80
---
src/mapi/glapi/gen/ARB_direct_state_access.xml | 7 +++
src/mesa/main/bufferobj.c | 24 +++-
src/mesa/main/bufferobj.h | 8 ++--
src/mesa/main/tests/dispatch_sanity.cpp| 1 +
4 files changed, 37 insertions(+), 3
---
src/mapi/glapi/gen/ARB_direct_state_access.xml | 14 ++
src/mesa/main/bufferobj.c | 327 -
src/mesa/main/bufferobj.h | 20 +-
src/mesa/main/tests/dispatch_sanity.cpp| 2 +
4 files changed, 189 insertions(+), 174
---
include/GL/glext.h | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/include/GL/glext.h b/include/GL/glext.h
index 925df18..2543d1b 100644
--- a/include/GL/glext.h
+++ b/include/GL/glext.h
@@ -2740,7 +2740,7 @@ GLAPI void APIENTRY glNamedBufferData (GLuint buffer,
---
src/mapi/glapi/gen/ARB_direct_state_access.xml | 8 ++
src/mesa/main/bufferobj.c | 102 -
src/mesa/main/bufferobj.h | 12 +++
src/mesa/main/tests/dispatch_sanity.cpp| 1 +
4 files changed, 89 insertions(+), 34
Originally, I used a nir_alu_src because I was lazy. However, we only need
part of that datastructure and carying the whole thing around implies
things we don't want such as the possibility of using non-SSA sources.
---
src/glsl/nir/nir_search.c | 21 +
1 file changed, 13
---
src/glsl/nir/nir_algebraic.py | 14 ++
src/glsl/nir/nir_search.c | 6 +-
src/glsl/nir/nir_search.h | 7 +++
3 files changed, 26 insertions(+), 1 deletion(-)
diff --git a/src/glsl/nir/nir_algebraic.py b/src/glsl/nir/nir_algebraic.py
index f9b246d..37eb523 100644
On 21.01.2015 21:12, Marek Olšák wrote:
We also had a case when the CPU accidentally corrupted shaders,
because the shaders were mapped after textures and a CPU texture
upload overflowed and overwrote shaders. I suppose we should have
unmapped the shaders.
Sounds like a good idea.
Tom, for
FYI: I just pushed the NIR patches with Connor's R-B so as soon as we're
done with configure.ac, we should be building on RHEL6 again.
--Jason
On Wed, Jan 21, 2015 at 5:37 PM, Ian Romanick i...@freedesktop.org wrote:
On 01/21/2015 05:35 PM, Matt Turner wrote:
On Wed, Jan 21, 2015 at 5:28 PM,
On Wed, Jan 21, 2015 at 5:26 PM, Eric Anholt e...@anholt.net wrote:
There aren't many users yet, but I wanted to do this from my scalarizing
pass.
---
src/glsl/nir/nir.c | 18 ++
src/glsl/nir/nir.h | 5 -
Yes, please. I've made that mistake enough myself.
Reviewed-by: Jason Ekstrand jason.ekstr...@intel.com
On Wed, Jan 21, 2015 at 5:25 PM, Eric Anholt e...@anholt.net wrote:
Almost all instructions we nir_ssa_def_init() for are nir_dests, and you
have to keep from forgetting to set is_ssa when
---
src/mapi/glapi/gen/ARB_direct_state_access.xml | 7 ++
src/mesa/main/bufferobj.c | 150 -
src/mesa/main/bufferobj.h | 13 ++-
src/mesa/main/tests/dispatch_sanity.cpp| 1 +
4 files changed, 113 insertions(+), 58
---
include/GL/glext.h | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/include/GL/glext.h b/include/GL/glext.h
index 1ffe576..0bc9e98 100644
--- a/include/GL/glext.h
+++ b/include/GL/glext.h
@@ -2736,7 +2736,7 @@ GLAPI void APIENTRY glGetTransformFeedbacki_v (GLuint
xfb,
---
src/mapi/glapi/gen/ARB_direct_state_access.xml | 7 +++
src/mesa/main/bufferobj.c | 63 +++---
src/mesa/main/bufferobj.h | 9
src/mesa/main/tests/dispatch_sanity.cpp| 1 +
4 files changed, 64 insertions(+), 16
---
include/GL/glext.h | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/include/GL/glext.h b/include/GL/glext.h
index 0bc9e98..3f141a2 100644
--- a/include/GL/glext.h
+++ b/include/GL/glext.h
@@ -2737,7 +2737,7 @@ GLAPI void APIENTRY glGetTransformFeedbacki64_v (GLuint
xfb,
---
include/GL/glext.h | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/include/GL/glext.h b/include/GL/glext.h
index 2543d1b..bca1b6e 100644
--- a/include/GL/glext.h
+++ b/include/GL/glext.h
@@ -2742,7 +2742,7 @@ GLAPI void APIENTRY glCopyNamedBufferSubData (GLuint
---
src/mapi/glapi/gen/ARB_direct_state_access.xml | 9 ++-
src/mesa/main/bufferobj.c | 77 --
src/mesa/main/bufferobj.h | 13 -
src/mesa/main/tests/dispatch_sanity.cpp| 1 +
4 files changed, 69 insertions(+), 31
---
src/mapi/glapi/gen/ARB_direct_state_access.xml | 7 +++
src/mesa/main/bufferobj.c | 65 --
src/mesa/main/bufferobj.h | 5 +-
src/mesa/main/tests/dispatch_sanity.cpp| 1 +
4 files changed, 64 insertions(+), 14
---
src/mesa/main/bufferobj.c | 19 +++
src/mesa/main/bufferobj.h | 4
2 files changed, 23 insertions(+)
diff --git a/src/mesa/main/bufferobj.c b/src/mesa/main/bufferobj.c
index 0c1ce98..aea7f1d 100644
--- a/src/mesa/main/bufferobj.c
+++ b/src/mesa/main/bufferobj.c
@@
---
include/GL/glext.h | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/include/GL/glext.h b/include/GL/glext.h
index 88505b0..1ffe576 100644
--- a/include/GL/glext.h
+++ b/include/GL/glext.h
@@ -2735,7 +2735,7 @@ GLAPI void APIENTRY glGetTransformFeedbackiv (GLuint xfb,
On 01/21/2015 05:40 PM, Laura Ekstrand wrote:
---
src/mapi/glapi/gen/ARB_direct_state_access.xml | 9 ++-
src/mesa/main/bufferobj.c | 77
--
src/mesa/main/bufferobj.h | 13 -
src/mesa/main/tests/dispatch_sanity.cpp
On 01/21/2015 05:40 PM, Laura Ekstrand wrote:
---
src/mapi/glapi/gen/ARB_direct_state_access.xml | 7 ++
src/mesa/main/bufferobj.c | 150
-
src/mesa/main/bufferobj.h | 13 ++-
src/mesa/main/tests/dispatch_sanity.cpp
Connor wrote *exactly* the same patch:
http://lists.freedesktop.org/archives/mesa-dev/2015-January/074522.html
This version can have my R-B too. And, fwiw, I like your commit message
better.
On Wed, Jan 21, 2015 at 5:25 PM, Eric Anholt e...@anholt.net wrote:
brw_fs_nir has only seen scalar
brw_fs_nir has only seen scalar bools so far, thanks to vector splitting,
and the ralloc of in glsl_to_nir.cpp will *usually* get you a 0-filled
chunk of memory, so reading too large of a value will usually get you the
right bool value. But once we start doing vector bools in a few commits,
we
This is the equivalent of brw_fs_channel_expressions.cpp, which I wanted
for vc4.
---
This series, plus a commit to make i965 use it instead of channel_expressions,
is on the nir-scalarize branch of my mesa tree. With the whole series, there
are 6 regressions, 3 of which are due to
Almost all instructions we nir_ssa_def_init() for are nir_dests, and you
have to keep from forgetting to set is_ssa when you do. Just provide the
simpler helper, instead.
---
src/glsl/nir/glsl_to_nir.cpp| 14 --
src/glsl/nir/nir.c | 8
---
src/glsl/nir/nir_search.c | 24 ++--
1 file changed, 10 insertions(+), 14 deletions(-)
diff --git a/src/glsl/nir/nir_search.c b/src/glsl/nir/nir_search.c
index a7bd051..e69fdfd 100644
--- a/src/glsl/nir/nir_search.c
+++ b/src/glsl/nir/nir_search.c
@@ -221,13 +221,11 @@
There aren't many users yet, but I wanted to do this from my scalarizing
pass.
---
src/glsl/nir/nir.c | 18 ++
src/glsl/nir/nir.h | 5 -
src/glsl/nir/nir_lower_vec_to_movs.c | 7 ++-
src/glsl/nir/nir_opt_peephole_select.c | 5
It's nice to have this present in your default cases so you can see what
instruction is triggering an abort.
---
src/glsl/nir/nir.h | 1 +
src/glsl/nir/nir_print.c | 14 --
2 files changed, 13 insertions(+), 2 deletions(-)
diff --git a/src/glsl/nir/nir.h b/src/glsl/nir/nir.h
Originally, get_alu_src was supposed to handle resolving swizzles and
things like that. However, now that every instruction we have only takes
scalar sources, we don't really need it anymore. The only case where it's
still marginally useful is for the MOV's that are generated from the
out-of-ssa
---
src/glsl/Makefile.sources | 1 +
src/glsl/nir/nir.h | 2 +
src/glsl/nir/nir_lower_alu_to_scalar.c | 155 +
3 files changed, 158 insertions(+)
create mode 100644 src/glsl/nir/nir_lower_alu_to_scalar.c
diff --git
Yes, I'm fully aware that some of this is a repeat of what eric just sent.
However, most of the repeated work is in the first patch (and the 5th)
which I'm not too worried about. There's two solutions to scalarizing
things floating around and we can decide what we like best.
The important parts
---
src/glsl/Makefile.am | 6 +++
src/glsl/Makefile.sources| 1 +
src/glsl/nir/nir.h | 2 +
src/glsl/nir/nir_lower_alu_reductions.py | 78
4 files changed, 87 insertions(+)
create mode 100644
---
src/mapi/glapi/gen/ARB_direct_state_access.xml | 8 +++
src/mesa/main/tests/dispatch_sanity.cpp| 1 +
src/mesa/main/teximage.c | 92 +-
src/mesa/main/teximage.h | 4 ++
4 files changed, 104 insertions(+), 1
---
include/GL/glext.h | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/include/GL/glext.h b/include/GL/glext.h
index d3cfbb5..88505b0 100644
--- a/include/GL/glext.h
+++ b/include/GL/glext.h
@@ -2773,7 +2773,7 @@ GLAPI void APIENTRY glNamedRenderbufferStorageMultisample
On 01/21/2015 05:40 PM, Laura Ekstrand wrote:
---
include/GL/glext.h | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/include/GL/glext.h b/include/GL/glext.h
index 88505b0..1ffe576 100644
--- a/include/GL/glext.h
+++ b/include/GL/glext.h
@@ -2735,7 +2735,7 @@ GLAPI
On Wed, Jan 21, 2015 at 12:19 PM, Connor Abbott cwabbo...@gmail.com wrote:
Assuming this actually compiles with GCC 4.4...
I don't know for sure, but it certainly gets us closer
--Jason
Reviewed-by: Connor Abbott cwabbo...@gmail.com
On Wed, Jan 21, 2015 at 2:14 PM, Jason Ekstrand
On Wednesday, January 21, 2015 08:17:36 PM ville.syrj...@linux.intel.com wrote:
From: Ville Syrjälä ville.syrj...@linux.intel.com
Increase the device info .urb.size for BDW GT3 and CHV to match the
default URB size for each.
Also add all missing platforms (BYT,BDW,CHV) to the comment
On Wed, Jan 21, 2015 at 3:03 AM, Michel Dänzer mic...@daenzer.net wrote:
On 20.01.2015 22:39, Marek Olšák wrote:
The problem with CPDMA (DMA_DATA and WRITE_DATA) is that the ordering
of flushes must be correct. First, partial flushes must be done, so
that the shaders are idle.
That's only
https://bugs.freedesktop.org/show_bug.cgi?id=87886
--- Comment #26 from Eero Tamminen eero.t.tammi...@intel.com ---
as root, mount debugfs and monitor CAGF (actual GPU frequency) value from
/sys/kernel/debug/dri/0/i915_frequency_info, or if you have older kernel from
Matt and I noticed that one of the shaders hurt by INTEL_USE_NIR=1 had
load_input and load_uniform intrinsics repeated several times, with the
same parameters, but each one generating a distinct SSA value. This
made ALU operations on those values appear distinct as well.
Generating distinct SSA
In general I don't think that modifying shader code with the GPU is such
a good idea. We already had quite a number of occasions where the GPU
accidentally corrupted the shader data which usually leads to lockups.
Because of this I already wanted to suggest that everything that the GPU
https://bugs.freedesktop.org/show_bug.cgi?id=88467
--- Comment #9 from Dave Airlie airl...@freedesktop.org ---
that doesn't help with our internal builds for inclusion in RHEL, they have to
build with the system compiler.
--
You are receiving this mail because:
You are the QA Contact for the
https://bugs.freedesktop.org/show_bug.cgi?id=88662
Bug ID: 88662
Summary: unaligned access to gl_dlist_node
Product: Mesa
Version: 10.4
Hardware: SPARC
OS: OpenBSD
Status: NEW
Severity: normal
On Jan 21, 2015 3:29 AM, Kenneth Graunke kenn...@whitecape.org wrote:
Matt and I noticed that one of the shaders hurt by INTEL_USE_NIR=1 had
load_input and load_uniform intrinsics repeated several times, with the
same parameters, but each one generating a distinct SSA value. This
made ALU
https://bugs.freedesktop.org/show_bug.cgi?id=88467
--- Comment #10 from Jason Ekstrand ja...@jlekstrand.net ---
(In reply to Dave Airlie from comment #6)
This is actually more annoying
with gcc 4.4
gcc -std=gnu99 supports anonymous structs, but doesn't support designated
initialisers for
1 - 100 of 106 matches
Mail list logo