On Fri, 10 Apr 2015 15:49:32 -0700
Jordan Justen jordan.l.jus...@intel.com wrote:
Sounds reasonable. Assuming you build tested, series
Reviewed-by: Jordan Justen jordan.l.jus...@intel.com
Same
r-b: bwidawsk...
Though I am not sure I see the utility of 4 patches instead of 1...
On
On Fri, Apr 10, 2015 at 07:39:51PM +0100, Neil Roberts wrote:
Commit e93566a15c61c33faa changed the message header code needed to
make Skylake use SIMD4x2 so that it uses a register with width 4
instead of 8 as the source register in the send message. However it
also changed the width for the
Starting with Skylake, the display engine is capable of scanning out from
Y-tiled buffers. As such, we can and should use Y-tiling for better efficiency.
Note that the buffer allocation done for mipmaps will already never allocate an
X-tiled buffer for GEN9.
Signed-off-by: Ben Widawsky
Reviewed-by: Jason Ekstrand jason.ekstr...@intel.com
On Sat, Apr 11, 2015 at 1:38 PM, Ben Widawsky
benjamin.widaw...@intel.com wrote:
Certain platforms support the ability to sample from a texture, and write it
out
to the file RT - thus saving a costly send instructions (note that this is a
Readding mesa-dev@ to Cc.
On Sat, Apr 11, 2015 at 11:06 AM, Darius Goad alegen...@gmail.com wrote:
Well, I did make a v2, but all that did was fix a bug. How does this not
come close?
Well, for instance the i965 driver supports ~5 generations of
hardware, three of them that don't support
From: Marek Olšák marek.ol...@amd.com
This hack is no longer needed. (see the previous commit)
---
src/mesa/state_tracker/st_glsl_to_tgsi.cpp | 24
1 file changed, 24 deletions(-)
diff --git a/src/mesa/state_tracker/st_glsl_to_tgsi.cpp
From: Marek Olšák marek.ol...@amd.com
---
src/mesa/program/prog_parameter.c | 72 ---
src/mesa/program/prog_parameter.h | 9 -
2 files changed, 81 deletions(-)
diff --git a/src/mesa/program/prog_parameter.c
b/src/mesa/program/prog_parameter.c
index
From: Marek Olšák marek.ol...@amd.com
This will allow removing the uniform storage re-association during
TGSI generation at draw time.
---
src/mesa/state_tracker/st_glsl_to_tgsi.cpp | 44 --
1 file changed, 24 insertions(+), 20 deletions(-)
diff --git
Thanks for the lengthy response :)
8. apr. 2015 01.52 skrev Connor Abbott cwabbo...@gmail.com:
Hi Thomas,
Thanks for submitting a proposal! Some comments/answers below.
On Tue, Apr 7, 2015 at 3:34 PM, Thomas Helland
thomashellan...@gmail.com wrote:
Hi,
For those that don't know I've
From: Marek Olšák marek.ol...@amd.com
Cc: 10.4 10.5 mesa-sta...@lists.freedesktop.org
---
src/mesa/state_tracker/st_glsl_to_tgsi.cpp | 8 +++-
1 file changed, 3 insertions(+), 5 deletions(-)
diff --git a/src/mesa/state_tracker/st_glsl_to_tgsi.cpp
From: Marek Olšák marek.ol...@amd.com
---
src/mesa/state_tracker/st_atom_shader.c| 3 +--
src/mesa/state_tracker/st_glsl_to_tgsi.cpp | 24
src/mesa/state_tracker/st_glsl_to_tgsi.h | 7 ---
3 files changed, 9 insertions(+), 25 deletions(-)
diff --git
From: Marek Olšák marek.ol...@amd.com
---
configure.ac | 4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)
diff --git a/configure.ac b/configure.ac
index 4ed4b74..113fb49 100644
--- a/configure.ac
+++ b/configure.ac
@@ -421,7 +421,9 @@ if test x$enable_debug = xyes; then
fi
From: Marek Olšák marek.ol...@amd.com
---
src/gallium/drivers/radeon/radeon_setup_tgsi_llvm.c | 6 ++
1 file changed, 6 insertions(+)
diff --git a/src/gallium/drivers/radeon/radeon_setup_tgsi_llvm.c
b/src/gallium/drivers/radeon/radeon_setup_tgsi_llvm.c
index 1690194..333f7ae 100644
---
On Wednesday, April 08, 2015 12:25:27 PM Martin Peres wrote:
On 08/04/15 10:06, Kenneth Graunke wrote:
Previously, we translated into NIR and did all the optimizations and
lowering as part of running fs_visitor. This meant that we did all of
that work twice for fragment shaders - once for
On Sat, Apr 11, 2015 at 11:25 AM, Ben Widawsky b...@bwidawsk.net wrote:
On Fri, 10 Apr 2015 15:49:32 -0700
Jordan Justen jordan.l.jus...@intel.com wrote:
Sounds reasonable. Assuming you build tested, series
Reviewed-by: Jordan Justen jordan.l.jus...@intel.com
Same
r-b: bwidawsk...
I think
On Saturday, April 11, 2015 01:38:44 PM Ben Widawsky wrote:
Certain platforms support the ability to sample from a texture, and write it
out
to the file RT - thus saving a costly send instructions (note that this is a
potnential win if one wanted to backport to a tag that didn't have the
On Sat, Apr 11, 2015 at 12:11 PM, Marek Olšák mar...@gmail.com wrote:
From: Marek Olšák marek.ol...@amd.com
---
configure.ac | 4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)
diff --git a/configure.ac b/configure.ac
index 4ed4b74..113fb49 100644
--- a/configure.ac
+++
On Sat, Apr 11, 2015 at 12:11 PM, Marek Olšák mar...@gmail.com wrote:
From: Marek Olšák marek.ol...@amd.com
---
Ugh, these two functions have been dead since 2012 and 2010. :(
Reviewed-by: Matt Turner matts...@gmail.com
___
mesa-dev mailing list
Certain platforms support the ability to sample from a texture, and write it out
to the file RT - thus saving a costly send instructions (note that this is a
potnential win if one wanted to backport to a tag that didn't have the patch
from Topi which removed excess MOVs from LOAD_PAYLOAD -
On Sat, Apr 11, 2015 at 10:23 AM, Matt Turner matts...@gmail.com wrote:
Commit e16c5c90 was replacing 'region' with 'mt', leaving this
nonsensical code.
---
src/mesa/drivers/dri/i965/intel_mipmap_tree.c | 6 ++
1 file changed, 2 insertions(+), 4 deletions(-)
diff --git
On Sat, Apr 11, 2015 at 09:40:36AM -0700, Matt Turner wrote:
On Sat, Apr 11, 2015 at 2:46 AM, Kenneth Graunke kenn...@whitecape.org
wrote:
new_prim was declared as a stack variable within a nested scope; we
tried to retain a pointer to that data beyond the scope, which is bogus.
GCC
On Sat, Apr 11, 2015 at 3:12 PM, Thomas Helland
thomashellan...@gmail.com wrote:
Yes, copy propagation probably won't be so useful once we have value
range propagation; the former is a special case of the latter. Note
that we have a nifty way of actually doing the constant folding
2015-04-08 18:03 GMT+02:00 Jason Ekstrand ja...@jlekstrand.net:
On Tue, Apr 7, 2015 at 4:52 PM, Connor Abbott cwabbo...@gmail.com wrote:
Hi Thomas,
Thanks for submitting a proposal! Some comments/answers below.
On Tue, Apr 7, 2015 at 3:34 PM, Thomas Helland
thomashellan...@gmail.com wrote:
On Sat, Apr 11, 2015 at 2:13 PM, Matt Turner matts...@gmail.com wrote:
Readding mesa-dev@ to Cc.
On Sat, Apr 11, 2015 at 11:06 AM, Darius Goad alegen...@gmail.com wrote:
Well, I did make a v2, but all that did was fix a bug. How does this not
come close?
Well, for instance the i965 driver
The problem is NDEBUG is defined if --enable-debug isn't used. If I
add -DDEBUG manually to CFLAGS, both DEBUG and NDEBUG will be defined.
That means Gallium assertions will be enabled (because DEBUG is
defined) and Mesa core assertion will be disabled (because NDEBUG is
defined).
I was about to
On Apr 10, 2015 1:42 PM, Rob Clark robdcl...@gmail.com wrote:
From: Rob Clark robcl...@freedesktop.org
I guess I was looking too much at how lower_system_values worked when
writing lower_idiv.
Since ttn wasn't emitting load_var for sysvals and the only drivers
using lower_idiv were using
On Friday, April 10, 2015 03:52:53 PM Ben Widawsky wrote:
This optimization is disabled by default, and it is recommended to be enabled.
The docs do not provide much info.
Cc: Ian Romanick ian.d.roman...@intel.com
Cc: Neil S. Roberts n...@linux.intel.com
Signed-off-by: Ben Widawsky
On Friday, April 10, 2015 10:49:51 PM Ben Widawsky wrote:
On Fri, Apr 10, 2015 at 07:50:19PM -0700, Kenneth Graunke wrote:
On Friday, April 10, 2015 12:52:03 PM Ben Widawsky wrote:
Based originally on a patch from Ken in May 2014 of the same title. Things
changed enough that I didn't feel
new_prim was declared as a stack variable within a nested scope; we
tried to retain a pointer to that data beyond the scope, which is bogus.
GCC with -O1 eliminated most of the code that set new_prim's fields.
Move the declaration to fix the bug.
Bugzilla:
ARB_uniform_buffer_object spec says:
The total amount of buffer object storage available for any given uniform
block is subject to an implementation-dependent limit; the maximum amount
of available space, in basic machine units, can be queried by calling
GetIntegerv with the constant
Commit e16c5c90 was replacing 'region' with 'mt', leaving this
nonsensical code.
---
src/mesa/drivers/dri/i965/intel_mipmap_tree.c | 6 ++
1 file changed, 2 insertions(+), 4 deletions(-)
diff --git a/src/mesa/drivers/dri/i965/intel_mipmap_tree.c
Coverity is confused by the float int / 2 expression and suggests
casting MAX_GLUINT to unsigned, which I believe it was supposed to have
been already.
---
src/mesa/swrast/s_tritemp.h | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/src/mesa/swrast/s_tritemp.h
This patch set enables ARB_gpu_shader_fp64 for i965.
Darius Goad (3):
i965/fs: Add basic support for double-precision floating-point types
for Gen7+
i965: Add support for double-precision floating-point types on Gen8+.
ilo: Enable ARB_gpu_shader_fp64
---
src/mesa/drivers/dri/i965/brw_fs.cpp | 2 +-
src/mesa/drivers/dri/i965/brw_fs_visitor.cpp | 4 ++--
src/mesa/drivers/dri/i965/brw_shader.cpp | 3 ++-
src/mesa/drivers/dri/i965/brw_vec4_visitor.cpp | 1 +
4 files changed, 6 insertions(+), 4 deletions(-)
diff --git
---
src/gallium/drivers/ilo/ilo_screen.c | 2 ++
1 file changed, 2 insertions(+)
diff --git a/src/gallium/drivers/ilo/ilo_screen.c
b/src/gallium/drivers/ilo/ilo_screen.c
index 80ea4c7..4e6988f 100644
--- a/src/gallium/drivers/ilo/ilo_screen.c
+++ b/src/gallium/drivers/ilo/ilo_screen.c
@@ -147,6
On Sat, Apr 11, 2015 at 10:38 AM, Darius Goad alegen...@gmail.com wrote:
This patch set enables ARB_gpu_shader_fp64 for i965.
Darius Goad (3):
i965/fs: Add basic support for double-precision floating-point types
for Gen7+
i965: Add support for double-precision floating-point types on
On Sat, Apr 11, 2015 at 2:46 AM, Kenneth Graunke kenn...@whitecape.org wrote:
new_prim was declared as a stack variable within a nested scope; we
tried to retain a pointer to that data beyond the scope, which is bogus.
GCC with -O1 eliminated most of the code that set new_prim's fields.
Move
---
src/mesa/drivers/dri/i965/brw_reg.h | 1 +
src/mesa/drivers/dri/i965/brw_shader.cpp | 19 +--
2 files changed, 14 insertions(+), 6 deletions(-)
diff --git a/src/mesa/drivers/dri/i965/brw_reg.h
b/src/mesa/drivers/dri/i965/brw_reg.h
index 924b059..143a667 100644
---
---
src/gallium/drivers/ilo/ilo_screen.c | 2 ++
1 file changed, 2 insertions(+)
diff --git a/src/gallium/drivers/ilo/ilo_screen.c
b/src/gallium/drivers/ilo/ilo_screen.c
index 80ea4c7..e6a8354 100644
--- a/src/gallium/drivers/ilo/ilo_screen.c
+++ b/src/gallium/drivers/ilo/ilo_screen.c
@@ -147,6
---
src/mesa/drivers/dri/i965/brw_reg.h | 1 +
src/mesa/drivers/dri/i965/brw_shader.cpp | 19 +--
2 files changed, 14 insertions(+), 6 deletions(-)
diff --git a/src/mesa/drivers/dri/i965/brw_reg.h
b/src/mesa/drivers/dri/i965/brw_reg.h
index 924b059..143a667 100644
---
---
src/mesa/drivers/dri/i965/brw_fs.cpp | 2 +-
src/mesa/drivers/dri/i965/brw_fs_visitor.cpp | 4 ++--
src/mesa/drivers/dri/i965/brw_shader.cpp | 3 ++-
src/mesa/drivers/dri/i965/brw_vec4_visitor.cpp | 1 +
4 files changed, 6 insertions(+), 4 deletions(-)
diff --git
The description should be ... for people not using --enable-debug and
using -DDEBUG.
Marek
On Sat, Apr 11, 2015 at 9:57 PM, Matt Turner matts...@gmail.com wrote:
On Sat, Apr 11, 2015 at 12:11 PM, Marek Olšák mar...@gmail.com wrote:
From: Marek Olšák marek.ol...@amd.com
---
configure.ac | 4
https://bugs.freedesktop.org/show_bug.cgi?id=79706
Vinson Lee v...@freedesktop.org changed:
What|Removed |Added
Depends on||89330
--
You are
https://bugs.freedesktop.org/show_bug.cgi?id=79706
Vinson Lee v...@freedesktop.org changed:
What|Removed |Added
Depends on||89773
--
You are
https://bugs.freedesktop.org/show_bug.cgi?id=89773
Vinson Lee v...@freedesktop.org changed:
What|Removed |Added
Keywords||bisected
https://bugs.freedesktop.org/show_bug.cgi?id=79706
Vinson Lee v...@freedesktop.org changed:
What|Removed |Added
Depends on||89960
--
You are
https://bugs.freedesktop.org/show_bug.cgi?id=89960
Vinson Lee v...@freedesktop.org changed:
What|Removed |Added
Blocks||79706
--
You are
https://bugs.freedesktop.org/show_bug.cgi?id=89018
Sami Liedes sami.lie...@iki.fi changed:
What|Removed |Added
Status|NEEDINFO|NEW
CC|
https://bugs.freedesktop.org/show_bug.cgi?id=89018
--- Comment #7 from Jason Ekstrand ja...@jlekstrand.net ---
(In reply to Sami Liedes from comment #6)
I see this AFAICT exactly same issue with both radeonsi and swrast on recent
HEAD of mesa, so this is not Intel-specific. FWIW, I also tried
Add a test to exercise a worst case collision scenario
that may cause us to not be able to find an empty
slot in the table even though it is not full.
This hits the bug in my last revision of the series
converting the hash table to quadratic probing.
Signed-off-by: Thomas Helland
This should give better cache locality, less memory consumption,
less code, and should also be faster since we avoid modulo operations.
Also change table size to be power of two.
This gives better performance as we can do bitmasking instead of
modulo operations for fitting the hash in the address
The same rationale applies here as for the hash table.
Power of two size should give better performance,
and using the algorithm hash = sh + i/2 + i*i/2
should result in only distinct hash values when hitting collisions.
Difference at 95.0% confidence
-7.9505 +/- 2.44011
-5.04357% +/-
The performance numbers (shader-db runtime) are:
Difference at 95.0% confidence
-14.7608 +/- 3.36786
-9.05064% +/- 2.06501%
(Original runtime was 160 seconds)
While the profile data looked promising for increasing the
table size we start with, decreasing load factor, and integer
https://bugs.freedesktop.org/show_bug.cgi?id=89330
Vinson Lee v...@freedesktop.org changed:
What|Removed |Added
Blocks||79706
--
You are
Extends the syntax of GALLIUM_HUD environment variable to:
- Add options to set the size and exact location of each pane.
- Add an option to limit the maximum allowed value of the X axis on a
pane, clamping the graph down to not go above this value.
- Add an option to auto-adjust the value of
55 matches
Mail list logo