Move the message filtering logic to debug_is_message_filtered.
Signed-off-by: Chia-I Wu o...@lunarg.com
---
src/mesa/main/errors.c | 111 +++--
1 file changed, 61 insertions(+), 50 deletions(-)
diff --git a/src/mesa/main/errors.c b/src/mesa/main
Merge get_msg into the function.
Signed-off-by: Chia-I Wu o...@lunarg.com
---
src/mesa/main/errors.c | 99 +-
1 file changed, 26 insertions(+), 73 deletions(-)
diff --git a/src/mesa/main/errors.c b/src/mesa/main/errors.c
index 139e31d..5136f75
Move group clearing to debug_clear_group. Add debug_pop_group and
debug_destroy for use in _mesa_PopDebugGroup and _mesa_free_errors_data
respectively.
Signed-off-by: Chia-I Wu o...@lunarg.com
---
src/mesa/main/errors.c | 127 +
1 file changed, 64
Hi list,
This series adds a thread pool to the GLSL compiler, and a drirc option to
defer glCompileShader calls to the pool. The goal is to reduce the start-up
time of applications that are aware of this feature. That is, applications
that compile shaders first and check the compile status
Move gl_debug_state allocation to a new function, debug_create.
Signed-off-by: Chia-I Wu o...@lunarg.com
---
src/mesa/main/errors.c | 62 ++
1 file changed, 37 insertions(+), 25 deletions(-)
diff --git a/src/mesa/main/errors.c b/src/mesa/main
Move most of the code to debug_set_default_filter.
Signed-off-by: Chia-I Wu o...@lunarg.com
---
src/mesa/main/errors.c | 113 +++--
1 file changed, 62 insertions(+), 51 deletions(-)
diff --git a/src/mesa/main/errors.c b/src/mesa/main/errors.c
index
This allows them to be moved to .rodata, and allow us to be sure that they
will not be modified.
Signed-off-by: Chia-I Wu o...@lunarg.com
---
src/glsl/builtin_types.cpp | 2 +-
src/glsl/builtin_variables.cpp | 63 +
src/glsl/glsl_parser.yy
Move message logging to debug_log_message.
Signed-off-by: Chia-I Wu o...@lunarg.com
---
src/mesa/main/errors.c | 136 -
1 file changed, 77 insertions(+), 59 deletions(-)
diff --git a/src/mesa/main/errors.c b/src/mesa/main/errors.c
index c478aac
There may be two contexts compiling shaders at the same time. locale_t needs
to be protected.
Signed-off-by: Chia-I Wu o...@lunarg.com
---
src/glsl/glsl_lexer.ll | 1 +
src/glsl/ir_reader.cpp | 2 ++
src/glsl/strtod.c | 36
src/glsl/strtod.h | 3
Add validate_length and use it and log_msg directly.
Signed-off-by: Chia-I Wu o...@lunarg.com
---
src/mesa/main/errors.c | 46 --
1 file changed, 20 insertions(+), 26 deletions(-)
diff --git a/src/mesa/main/errors.c b/src/mesa/main/errors.c
index
Move group copying to debug_push_group. Save group message in the group
before pushing instead of after.
Signed-off-by: Chia-I Wu o...@lunarg.com
---
src/mesa/main/errors.c | 100 +
1 file changed, 59 insertions(+), 41 deletions(-)
diff --git
Merge set_message_state, control_messages, and control_app_messages into the
function.
Signed-off-by: Chia-I Wu o...@lunarg.com
---
src/mesa/main/errors.c | 83 +++---
1 file changed, 17 insertions(+), 66 deletions(-)
diff --git a/src/mesa/main
It will be used to implement threaded glCompileShader.
Signed-off-by: Chia-I Wu o...@lunarg.com
---
src/glsl/Makefile.am | 12 +-
src/glsl/Makefile.sources | 3 +-
src/glsl/tests/threadpool_test.cpp | 137 +
src/glsl/threadpool.c | 394
When GL_DEBUG_OUTPUT_SYNCHRONOUS is GL_TRUE, drivers are allowed to log debug
messages from other threads. The debug state needs to be protected. We are
about to change mesa to spawn threads for deferred glCompileShader calls and
we will need this groundwork.
Signed-off-by: Chia-I Wu o
In both call sites, it could be easily replaced by direct
debug_is_message_filtered calls.
Signed-off-by: Chia-I Wu o...@lunarg.com
---
src/mesa/main/errors.c | 40 +++-
1 file changed, 11 insertions(+), 29 deletions(-)
diff --git a/src/mesa/main/errors.c b
Signed-off-by: Chia-I Wu o...@lunarg.com
---
src/mesa/main/errors.c | 8
1 file changed, 4 insertions(+), 4 deletions(-)
diff --git a/src/mesa/main/errors.c b/src/mesa/main/errors.c
index 0aa9c8a..b9f8fc6 100644
--- a/src/mesa/main/errors.c
+++ b/src/mesa/main/errors.c
@@ -924,9 +924,9
Setting it to a non-zero value N will cause glCompileShader to be deferred to
a thread pool. When N is greater than 1, it indicates the maximum number of
threads in the pool. When N is 1, the number of threads is up to the driver.
Signed-off-by: Chia-I Wu o...@lunarg.com
---
src/mesa/drivers
From: Chia-I Wu o...@lunarg.com
Threaded glCompileShader can be enabled for a context by calling
_mesa_enable_glsl_threadpool. It will initialize the singleton GLSL thread
pool and defer glCompileShader calls to the thread pool.
For applications to benefit from threaded glCompileShader
On Tue, Apr 22, 2014 at 8:10 PM, Timothy Arceri t_arc...@yahoo.com.au wrote:
On Tue, 2014-04-22 at 16:58 +0800, Chia-I Wu wrote:
Hi list,
This series adds a thread pool to the GLSL compiler, and a drirc option to
defer glCompileShader calls to the pool. The goal is to reduce the start-up
On Fri, Mar 28, 2014 at 5:40 AM, Ian Romanick i...@freedesktop.org wrote:
From: Ian Romanick ian.d.roman...@intel.com
This code was broken in some odd ways before. To much state was being
saved, it was being restored in the wrong order, and in the wrong way.
The biggest problem was that the
On Wed, Apr 16, 2014 at 3:18 AM, Eric Anholt e...@anholt.net wrote:
Kenneth Graunke kenn...@whitecape.org writes:
On 04/14/2014 05:33 PM, Eric Anholt wrote:
This manifested as rendering failures or sometimes GPU hangs in
compositors when they accidentally got MSAA visuals due to a bug in the
, then saw the table immediately
following with DEPTH_COMPONENT* formats, and assumed it was 3.17.
Based on a batch by Chia-I Wu, but without the driconf option to force
s/batch/patch/
Both patches look good to me. Unless I overlooked your patch for
piglit, this is needed
http
This unit test demonstrates a subtle bug fixed by
4ddf51db6af36736d5d42c1043eeea86e47459ce.
Signed-off-by: Chia-I Wu o...@lunarg.com
Cc: Eric Anholt e...@anholt.net
---
.../dri/i965/test_vec4_copy_propagation.cpp| 30 ++
1 file changed, 30 insertions(+)
diff --git
On Tue, Apr 8, 2014 at 2:52 AM, Eric Anholt e...@anholt.net wrote:
Chia-I Wu olva...@gmail.com writes:
From: Chia-I Wu o...@lunarg.com
Given
mov vgrf7, vgrf9.xyxz
add vgrf9.xyz, vgrf4.xyzw, vgrf5.xyzw
add vgrf10.x, vgrf6.xyzw, vgrf7.
the last instruction would be wrongly
From: Chia-I Wu o...@lunarg.com
Given
mov vgrf7, vgrf9.xyxz
add vgrf9.xyz, vgrf4.xyzw, vgrf5.xyzw
add vgrf10.x, vgrf6.xyzw, vgrf7.
the last instruction would be wrongly changed to
add vgrf10.x, vgrf6.xyzw, vgrf9.
during copy propagation.
The issue is that when deciding
On Fri, Mar 28, 2014 at 5:40 AM, Ian Romanick i...@freedesktop.org wrote:
From: Ian Romanick ian.d.roman...@intel.com
This will be used for GL_ARB_separate_shader_objects. That extension
not only allows separable shaders to rendezvous by location, but it also
allows traditionally linked
to detect this case, but we have to
paper over failures every time a new extension is added.
Reviewed-by: Chia-I Wu o...@lunarg.com
This change will cause a big pile of functions to disappear from
libGLESv2 and libGLESv1_CM.
libGLESv2 loses (20 functions):
glBindVertexArrayOES
They are not needed since 0da1a2cc369052643ccaea75a1722cc37652d82a.
---
src/glsl/glsl_parser.yy| 1 -
src/glsl/glsl_symbol_table.cpp | 25 -
src/glsl/glsl_symbol_table.h | 2 --
3 files changed, 28 deletions(-)
diff --git a/src/glsl/glsl_parser.yy
They are not needed since 514f8c7ec7cc1ab18be93cebb5b9bf970b1955a9.
---
src/glsl/glsl_parser_extras.cpp | 10 --
src/glsl/glsl_parser_extras.h | 4
src/mesa/main/mtypes.h | 3 +--
3 files changed, 1 insertion(+), 16 deletions(-)
diff --git
On Sun, Mar 16, 2014 at 5:20 AM, Beren Minor beren.mi...@gmail.com wrote:
EGL 1.4 Specification says that
eglMakeCurrent(display, EGL_NO_SURFACE, EGL_NO_SURFACE, EGL_NO_CONTEXT) can
be used to release the
current thread's ownership on the surfaces and context.
MESA's egl implementation was
On Thu, Mar 13, 2014 at 3:35 PM, Emil Velikov emil.l.veli...@gmail.com wrote:
On 13/03/14 07:09, Steven Newbury wrote:
On Tue, 2014-03-04 at 21:12 +, Emil Velikov wrote:
Drop the duplication accross all dri targets.
...
diff --git a/src/gallium/targets/dri-ilo/Makefile.am
,
ilo_dri.so will never be loaded even enabled and installed. The change should
not create any more confusion.
Signed-off-by: Chia-I Wu olva...@gmail.com
Cc: Emil Velikov emil.l.veli...@gmail.com
---
src/gallium/targets/dri-ilo/Makefile.am | 7 +++
1 file changed, 3 insertions(+), 4 deletions
Hi Brian,
_glthread_GetID() was removed by this commit
commit 02cb04c68ffbdaffaf7513ddc951584cac29f209
Author: Brian Paul bri...@vmware.com
Date: Tue Mar 4 15:24:16 2014 -0700
mesa: remove remaining uses of _glthread_GetID()
It turns out the function, declared in glapi.h, is a part of
On Thu, Mar 6, 2014 at 11:03 PM, Brian Paul bri...@vmware.com wrote:
On 03/05/2014 08:37 PM, Chia-I Wu wrote:
On Thu, Mar 6, 2014 at 7:06 AM, Brian Paul bri...@vmware.com wrote:
---
src/mapi/mapi.c | 10 +-
src/mapi/stub.c |6 +++---
src/mapi/u_current.c |6
___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/mesa-dev
--
o...@lunarg.com
From a600bf561a9c3d5776fd60d3ae3d14bd4ab417e6 Mon Sep 17 00:00:00 2001
From: Chia-I Wu olva...@gmail.com
Date: Fri, 7 Mar 2014 11:39:24 +0800
Subject: [PATCH] st
On Fri, Mar 7, 2014 at 2:39 AM, Brian Paul bri...@vmware.com wrote:
To fix the problem that Jose worked around in 1d8e3067fd9.
Looks good to me.
---
src/mesa/state_tracker/st_format.c |4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)
diff --git
On Fri, Mar 7, 2014 at 11:56 AM, Chia-I Wu olva...@gmail.com wrote:
On Fri, Mar 7, 2014 at 2:04 AM, Jose Fonseca jfons...@vmware.com wrote:
- Original Message -
Am 06.03.2014 18:32, schrieb Jose Fonseca:
- Original Message -
- Original Message -
On 03/06
On Wed, Mar 5, 2014 at 10:48 PM, Brian Paul bri...@vmware.com wrote:
To make the functions more understandable.
Both patches are
Reviewed-by: Chia-I Wu o...@lunarg.com
---
src/mapi/glapi/glapi.c |2 +-
src/mapi/mapi_glapi.c |2 +-
src/mapi/u_current.c | 32
On Thu, Mar 6, 2014 at 7:06 AM, Brian Paul bri...@vmware.com wrote:
A user would have no idea what _glthread_ is. This removes the
last remaining instance of the _glthread_ string in Mesa.
Reviewed-by: Chia-I Wu o...@lunarg.com
---
src/mapi/u_thread.h |6 +++---
1 file changed, 3
generate)
{
- u_mutex_declare_static(dynamic_mutex);
+ static mtx_t dynamic_mutex = _MTX_INITIALIZER_NP
PTHREAD_MUTEX_INITIALIZER;
PTHREAD_MUTEX_INITIALIZER should be dropped. With that fixed,
Reviewed-by: Chia-I Wu o...@lunarg.com
struct mapi_stub *stub = NULL;
int count, i
The format is mapped to PIPE_FORMAT_B8G8R8X8_SRGB.
---
src/mesa/state_tracker/st_format.c | 11 +--
1 file changed, 9 insertions(+), 2 deletions(-)
diff --git a/src/mesa/state_tracker/st_format.c
b/src/mesa/state_tracker/st_format.c
index 5f951eb..25577ac 100644
---
The format is needed to represent an RGB-only winsys framebuffer that is
sRGB-capable.
---
src/mesa/main/format_pack.c | 18 ++
src/mesa/main/format_unpack.c | 15 +++
src/mesa/main/formats.c | 21 +
src/mesa/main/formats.h | 3 +++
Hi,
Until now, we do not support sRGB write for winsys framebuffers. This series
changes that so that winsys framebuffers are marked sRGB-capable when the pipe
driver supports it.
This is not to be confused with GLX_FRAMEBUFFER_SRGB_CAPABLE_ARB. The
attribute guarantees the created winsys
The texture formats of winsys fbo are always linear becase the st manager
(st/dri for example) could not know the colorspace used. But it does not mean
that we cannot make the fbo sRGB-capable. By
- setting rb-Visual.sRGBCapable to GL_TRUE when the pipe driver supports the
format in sRGB
On Wed, Mar 5, 2014 at 12:10 AM, Brian Paul bri...@vmware.com wrote:
Spotted by Chia-I Wu.
unpack_ubyte_ARGB_REV() needs the same care too. With that fixed,
this patch is
Reviewed-by: Chia-I Wu o...@lunarg.com
---
src/mesa/main/format_pack.c |2 +-
src/mesa/main
colorspace,
- giving rb an sRGB internal format, and
- updating code to check rb-Format instead of strb-texture-format,
we should be good.
Fixed bug 75226 for at least llvmpipe and ilo, with no piglit regression.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=75226
Signed-off-by: Chia-I
On Mon, Mar 3, 2014 at 6:06 PM, Michel Dänzer mic...@daenzer.net wrote:
On Mon, 2014-03-03 at 16:22 +0800, Chia-I Wu wrote:
The texture formats of winsys fbo are always linear becase the st manager
(st/dri for example) could not know the colorspace used. But it does not
mean
that we cannot
Hi list,
From formats.h, we have
/* Type P formats */ /* msb -- TEXEL BITS --- lsb */
/* */
MESA_FORMAT_A8B8G8R8_UNORM, /* */
But in format_unpack.c,
-significant position of the 32-bit word.
Two comments below, and the patch is
Reviewed-by: Chia-I Wu o...@lunarg.com
---
src/mesa/main/formats.h | 98
+++
1 file changed, 49 insertions(+), 49 deletions(-)
diff --git a/src/mesa/main/formats.h b/src
On Mon, Feb 24, 2014 at 5:03 PM, Ilia Mirkin imir...@alum.mit.edu wrote:
On Mon, Feb 24, 2014 at 3:38 AM, Dave Airlie airl...@gmail.com wrote:
I've taken a quick look at ARB_texture_query_levels,
and the HW generally seems to implement this by returning it in the W
channel of RESINFO, which
from intel_winsys to intel_winsys_info
and ultimatelly to ilo_dev_info, due to the design of the former two.
Cc: Chia-I Wu olva...@gmail.com
Signed-off-by: Emil Velikov emil.l.veli...@gmail.com
---
Hi Chia-I
I've looked at a cleaner way of doing things but I fell a bit short.
Perhaps you can
On Sat, Feb 22, 2014 at 11:03 AM, Emil Velikov emil.l.veli...@gmail.com wrote:
Signed-off-by: Emil Velikov emil.l.veli...@gmail.com
---
src/glx/drisw_glx.c | 41 +
src/glx/drisw_priv.h | 64
2 files
On Thu, Feb 20, 2014 at 7:03 AM, Kenneth Graunke kenn...@whitecape.org wrote:
On 02/19/2014 02:27 PM, Ian Romanick wrote:
On 02/19/2014 12:08 PM, Kenneth Graunke wrote:
On 02/18/2014 09:48 PM, Chia-I Wu wrote:
Since 73bc6061f5c3b6a3bb7a8114bb2e1ab77d23cfdb, Z16 support is
not advertised
of the
paddings. But it should be necessary, at least according to the PRM.
No piglit regression on Sandy Bridge and Haswell.
Signed-off-by: Chia-I Wu o...@lunarg.com
---
src/mesa/drivers/dri/i965/intel_mipmap_tree.c | 59 ---
1 file changed, 44 insertions(+), 15 deletions(-)
diff
HiZ buffers are opaque to the driver, and keeping the miptree around does not
gain us anything.
Signed-off-by: Chia-I Wu o...@lunarg.com
---
src/mesa/drivers/dri/i965/brw_misc_state.c| 9 -
src/mesa/drivers/dri/i965/gen6_blorp.cpp | 2 +-
src/mesa/drivers/dri/i965
It worked only because depth region and hiz region have the same cpp, tiling,
and pitch.
Signed-off-by: Chia-I Wu o...@lunarg.com
---
src/mesa/drivers/dri/i965/brw_misc_state.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/src/mesa/drivers/dri/i965/brw_misc_state.c
b/src
On Tue, Feb 18, 2014 at 5:02 PM, Juha-Pekka Heikkilä
juhapekka.heikk...@gmail.com wrote:
On Tue, Feb 18, 2014 at 9:40 AM, Chia-I Wu olva...@gmail.com wrote:
On Tue, Feb 18, 2014 at 12:21 AM, Juha-Pekka Heikkila
juhapekka.heikk...@gmail.com wrote:
_eglInitResource() was used to memset entire
. In light of that, and per Ian's suggestion, stop advertising Z16
support by default, and add a drirc option, gl30_sized_format_rules, so that
users can override.
This regresses required-sized-texture-formats on GL 3.0.
Signed-off-by: Chia-I Wu o...@lunarg.com
Cc: Ian Romanick ian.d.roman
From: Chia-I Wu o...@lunarg.com
The information was lost during linking, causing the layout to be treated as
FRAG_DEPTH_LAYOUT_NONE.
Signed-off-by: Chia-I Wu o...@lunarg.com
---
src/mesa/main/shaderapi.c | 5 +
1 file changed, 5 insertions(+)
diff --git a/src/mesa/main/shaderapi.c b/src
On Tue, Feb 18, 2014 at 12:21 AM, Juha-Pekka Heikkila
juhapekka.heikk...@gmail.com wrote:
_eglInitResource() was used to memset entire _EGLSync by
writing more than size of pointed target. This does work
as long as Resource is the first element in _EGLSync,
this patch fixes such dependency.
Hi,
I am looking at performance issues for some benchmark for a while, and am able
to identify three issues so far, all related to depth test.
The first issue is slow 16-bit depth buffer. This is already known and is
fixed for GLES contexts by commit 73bc6061f5c3b6a3bb7a8114bb2e1a. It is not
is
Reviewed-by: Chia-I Wu olva...@gmail.com
v2 (idr): Use 64 instead of 4096
Tested-by: Ian Romanick ian.d.roman...@intel.com
Cc: Chia-I Wu olva...@gmail.com
---
src/gallium/drivers/ilo/ilo_screen.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/src/gallium/drivers/ilo
On Wed, Jan 8, 2014 at 8:23 PM, Steven Newbury st...@snewbury.org.uk wrote:
On Fri, 2014-01-03 at 13:14 +0800, Chia-I Wu wrote:
On Thu, Jan 2, 2014 at 10:39 PM, Steven Newbury st...@snewbury.org.uk
wrote:
Forgot to add signed-off-by...
In trying to get gallium-nine working with the ilo
On Thu, Jan 2, 2014 at 10:39 PM, Steven Newbury st...@snewbury.org.uk wrote:
Forgot to add signed-off-by...
In trying to get gallium-nine working with the ilo Gallium driver I
noticed there's no ilo pipe-loader driver being built.
This patch simply puts in place the missing pieces.
The
improvement is 1.32134% +/- 0.161878% (n=13).
v2:
- unset hiz_cleared automatically in intel_miptree_slice_set_needs_hiz_resolve
- set/unset hiz_cleared with intel_renderbuffer_att_set_needs_depth_resolve
Signed-off-by: Chia-I Wu o...@lunarg.com
---
src/mesa/drivers/dri/i965/brw_clear.c
On Fri, Dec 27, 2013 at 7:25 AM, Chad Versace
chad.vers...@linux.intel.com wrote:
On 12/10/2013 09:54 PM, Chia-I Wu wrote:
When the depth buffer is already cleared, skip GEN6_HIZ_OP_DEPTH_CLEAR.
This
is made possible by tracking which slices have been cleared in
struct intel_mipmap_level
When the depth buffer is already cleared, skip GEN6_HIZ_OP_DEPTH_CLEAR. This
is made possible by tracking which slices have been cleared in
struct intel_mipmap_level. The hiz_cleared flag is unset when the depth
buffer is rendered to or when a HiZ resolve is needed.
For Unigine Tropics, the FPS
On Tue, Oct 22, 2013 at 3:05 AM, Eric Anholt e...@anholt.net wrote:
Chia-I Wu olva...@gmail.com writes:
On Thu, Oct 17, 2013 at 3:24 AM, Matt Turner matts...@gmail.com wrote:
On Mon, Oct 14, 2013 at 4:14 PM, Eric Anholt e...@anholt.net wrote:
Previously, the best thing we had was to schedule
On Thu, Oct 17, 2013 at 3:24 AM, Matt Turner matts...@gmail.com wrote:
On Mon, Oct 14, 2013 at 4:14 PM, Eric Anholt e...@anholt.net wrote:
Previously, the best thing we had was to schedule the things unblocked by
the current instruction, on the hope that it would be consuming two values
at the
On Tue, Oct 15, 2013 at 7:14 AM, Eric Anholt e...@anholt.net wrote:
Previously, the best thing we had was to schedule the things unblocked by
the current instruction, on the hope that it would be consuming two values
at the end of their live intervals while only producing one new value.
On Thu, Oct 17, 2013 at 1:53 PM, Chia-I Wu olva...@gmail.com wrote:
Hi Eric,
On Sat, Oct 12, 2013 at 3:18 AM, Eric Anholt e...@anholt.net wrote:
Chia-I Wu olva...@gmail.com writes:
Hi Eric,
The frame rate of Unigine Tropics (with low shader quality) dropped
from 40.8 to 23.5 after
On Thu, Oct 17, 2013 at 3:29 PM, Chia-I Wu olva...@gmail.com wrote:
On Thu, Oct 17, 2013 at 1:53 PM, Chia-I Wu olva...@gmail.com wrote:
Hi Eric,
On Sat, Oct 12, 2013 at 3:18 AM, Eric Anholt e...@anholt.net wrote:
Chia-I Wu olva...@gmail.com writes:
Hi Eric,
The frame rate of Unigine
On Wed, Oct 16, 2013 at 2:08 AM, Ian Romanick i...@freedesktop.org wrote:
On 10/09/2013 12:06 AM, Chia-I Wu wrote:
From: Chia-I Wu o...@lunarg.com
FS_OPCODE_OVERWRITE_DST is used to indicate that the destination register is
(completely) overwritten. No code is emitted, but the liveness
From: Chia-I Wu o...@lunarg.com
When two partial writes write the first and second halves of a variable
respectively before the variable is used, the variable can be added to the def
bitset.
v2: no change
v3: no longer rely on hints from by the visitor
Signed-off-by: Chia-I Wu o...@lunarg.com
From: Chia-I Wu o...@lunarg.com
Add fs_visitor::emit_texture, which is used to emit the texture instruction
after the message payload has been set up.
v2: rebased because of texture gather changes
v3: rebased because of texture-from-GRF changes
Signed-off-by: Chia-I Wu o...@lunarg.com
---
src
From: Chia-I Wu o...@lunarg.com
Add fs_visitor::emit_dual_texture_gen7 that emulate SIMD16 sample_d with dual
SIMD8 sample_d on gen7+.
No piglit quick.tests regression on Ivy Bridge and Haswell.
Improved Xonotic with Ultra effects by 6.0209% +/- 0.396586% (N=11) on
Haswell.
v2: no change
v3
On Wed, Oct 16, 2013 at 6:26 AM, Eric Anholt e...@anholt.net wrote:
Chia-I Wu olva...@gmail.com writes:
From: Chia-I Wu o...@lunarg.com
FS_OPCODE_OVERWRITE_DST is used to indicate that the destination register is
(completely) overwritten. No code is emitted, but the liveness analysis can
Hi Eric,
On Sat, Oct 12, 2013 at 3:18 AM, Eric Anholt e...@anholt.net wrote:
Chia-I Wu olva...@gmail.com writes:
Hi Eric,
The frame rate of Unigine Tropics (with low shader quality) dropped
from 40.8 to 23.5 after this change.
Thanks for the note. I see the regression as well, and I see
On Sat, Oct 12, 2013 at 3:18 AM, Eric Anholt e...@anholt.net wrote:
Chia-I Wu olva...@gmail.com writes:
Hi Eric,
The frame rate of Unigine Tropics (with low shader quality) dropped
from 40.8 to 23.5 after this change.
Thanks for the note. I see the regression as well, and I see a shader
On Wed, Oct 9, 2013 at 3:35 AM, Eric Anholt e...@anholt.net wrote:
Chia-I Wu olva...@gmail.com writes:
From: Chia-I Wu o...@lunarg.com
FS_OPCODE_OVERWRITE_DST is used to indicate that the destination register is
(completely) overwritten. No code is emitted, but the liveness analysis can
From: Chia-I Wu o...@lunarg.com
FS_OPCODE_OVERWRITE_DST is used to indicate that the destination register is
(completely) overwritten. No code is emitted, but the liveness analysis can
use it as a hint to add the destination register to DEF bitset. This is
needed because it is hard to figure
From: Chia-I Wu o...@lunarg.com
Add fs_visitor::emit_texture, which is used to emit the texture instruction
after the message payload has been set up.
Signed-off-by: Chia-I Wu o...@lunarg.com
---
src/mesa/drivers/dri/i965/brw_fs.h | 10 +-
src/mesa/drivers/dri/i965/brw_fs_fp.cpp
From: Chia-I Wu o...@lunarg.com
Add fs_visitor::emit_dual_texture_gen7 that emulate SIMD16 sample_d with dual
SIMD8 sample_d on gen7+.
No piglit quick.tests regression on Ivy Bridge and Haswell.
Improved Xonotic with Ultra effects by 6.76479% +/- 0.619064% (at 95.0%
confidence) on Haswell
the
problem. I have posted my other series that emulates SIMD16 sample_d
with dual SIMD8 sample_d for review.
On Wed, Oct 2, 2013 at 6:38 AM, Ian Romanick i...@freedesktop.org wrote:
On 09/30/2013 10:54 PM, Chia-I Wu wrote:
From: Chia-I Wu o...@lunarg.com
I agree with both of Ken's comments
On Tue, Oct 1, 2013 at 12:16 PM, Kenneth Graunke kenn...@whitecape.org wrote:
On 09/30/2013 07:16 PM, Ian Romanick wrote:
On 09/11/2013 10:00 PM, Chia-I Wu wrote:
From: Chia-I Wu o...@lunarg.com
Replicate the gradient of the top-left pixel to the other three pixels in
the
subspan, as how
From: Chia-I Wu o...@lunarg.com
Hi,
This series of patches implements the emulation SIMD16 sample_d with dual
SIMD8 sample_d. Before the changes, the compiler would fail to generate
SIMD16 code for fragment shaders that use textureGrad. And that hurts the
performance.
The first four patches
From: Chia-I Wu o...@lunarg.com
SIMD8 sampler messages are allowed in SIMD16 mode, and they could not work
without BRW_COMPRESSION_2NDHALF. Later PRMs (gen5 and later) do not
explicitly state whether BRW_COMPRESSION_2NDHALF is allowed, but they do have
examples using send with SecHalf
From: Chia-I Wu o...@lunarg.com
When the instruction to send the sampler message is forced uncompressed or
sechalf, send SIMD8 one even in SIMD16 mode.
Signed-off-by: Chia-I Wu o...@lunarg.com
---
src/mesa/drivers/dri/i965/brw_fs_generator.cpp | 3 ++-
1 file changed, 2 insertions(+), 1
From: Chia-I Wu o...@lunarg.com
FS_OPCODE_OVERWRITE_DST is used to indicate that the destination register is
(completely) overwritten. No code is emitted, but the liveness analysis can
use it as a hint to add the destination register to DEF bitset. This is
needed because it is hard to figure
From: Chia-I Wu o...@lunarg.com
Copy sechalf to the new register, otherwise we would read wrong HW registers.
Signed-off-by: Chia-I Wu o...@lunarg.com
---
src/mesa/drivers/dri/i965/brw_fs.cpp | 1 +
1 file changed, 1 insertion(+)
diff --git a/src/mesa/drivers/dri/i965/brw_fs.cpp
b/src/mesa
From: Chia-I Wu o...@lunarg.com
Add fs_visitor::emit_texture, which is used to emit the texture instruction
after the message payload has been set up.
Signed-off-by: Chia-I Wu o...@lunarg.com
---
src/mesa/drivers/dri/i965/brw_fs.h | 10 ++-
src/mesa/drivers/dri/i965/brw_fs_fp.cpp
From: Chia-I Wu o...@lunarg.com
Add fs_visitor::emit_dual_texture_gen7 that emulate SIMD16 sample_d with dual
SIMD8 sample_d on gen7+. Fix fs_generator::generate_tex to send SIMD8
messages when force_uncompressed or force_sechalf is set.
No piglit quick.tests regression on Ivy Bridge
From: Chia-I Wu o...@lunarg.com
Consider only the top-left and top-right pixels to approximate DDX in a 2x2
subspan, unless the application requests a more accurate approximation via
GL_FRAGMENT_SHADER_DERIVATIVE_HINT or this optimization is disabled from the
new driconf option
On Mon, Sep 23, 2013 at 12:09 PM, Chia-I Wu olva...@gmail.com wrote:
On Fri, Sep 20, 2013 at 10:50 PM, Paul Berry stereotype...@gmail.com wrote:
On 17 September 2013 19:54, Chia-I Wu olva...@gmail.com wrote:
Hi Paul,
On Mon, Sep 16, 2013 at 3:46 PM, Chia-I Wu olva...@gmail.com wrote
On Fri, Sep 20, 2013 at 10:50 PM, Paul Berry stereotype...@gmail.com wrote:
On 17 September 2013 19:54, Chia-I Wu olva...@gmail.com wrote:
Hi Paul,
On Mon, Sep 16, 2013 at 3:46 PM, Chia-I Wu olva...@gmail.com wrote:
On Sat, Sep 14, 2013 at 5:15 AM, Paul Berry stereotype...@gmail.com
wrote
On Wed, Sep 18, 2013 at 6:27 AM, Mark Mueller markkmuel...@gmail.com wrote:
On Mon, Sep 16, 2013 at 1:31 AM, Chia-I Wu olva...@gmail.com wrote:
On Mon, Sep 16, 2013 at 4:12 PM, Chia-I Wu olva...@gmail.com wrote:
On Mon, Sep 16, 2013 at 3:50 AM, Mark Mueller markkmuel...@gmail.com
wrote
Hi Paul,
On Mon, Sep 16, 2013 at 3:46 PM, Chia-I Wu olva...@gmail.com wrote:
On Sat, Sep 14, 2013 at 5:15 AM, Paul Berry stereotype...@gmail.com wrote:
On 12 September 2013 22:06, Chia-I Wu olva...@gmail.com wrote:
From: Chia-I Wu o...@lunarg.com
Consider only the top-left and top-right
On Sat, Sep 14, 2013 at 5:15 AM, Paul Berry stereotype...@gmail.com wrote:
On 12 September 2013 22:06, Chia-I Wu olva...@gmail.com wrote:
From: Chia-I Wu o...@lunarg.com
Consider only the top-left and top-right pixels to approximate DDX in a
2x2
subspan, unless the application or the user
On Mon, Sep 16, 2013 at 3:50 AM, Mark Mueller markkmuel...@gmail.com wrote:
On Fri, Sep 13, 2013 at 2:15 PM, Paul Berry stereotype...@gmail.com wrote:
On 12 September 2013 22:06, Chia-I Wu olva...@gmail.com wrote:
From: Chia-I Wu o...@lunarg.com
Consider only the top-left and top-right
On Mon, Sep 16, 2013 at 4:12 PM, Chia-I Wu olva...@gmail.com wrote:
On Mon, Sep 16, 2013 at 3:50 AM, Mark Mueller markkmuel...@gmail.com wrote:
On Fri, Sep 13, 2013 at 2:15 PM, Paul Berry stereotype...@gmail.com wrote:
On 12 September 2013 22:06, Chia-I Wu olva...@gmail.com wrote:
From
101 - 200 of 604 matches
Mail list logo