https://bugs.freedesktop.org/show_bug.cgi?id=99987
--- Comment #13 from Hans de Goede ---
At least x2go and presumably also thinlinc can be fixed by doing:
sudo ln -s /usr/lib64/libGLX_mesa.so.0 /usr/lib64/libGLX_indirect.so.0
At least on a 64 bit Fedora, on 32 bit Fedora / other distros you w
On Fri, Mar 10, 2017 at 01:38:16PM +0100, Iago Toral Quiroga wrote:
> ---
> src/intel/vulkan/anv_blorp.c | 3 +++
> 1 file changed, 3 insertions(+)
>
> diff --git a/src/intel/vulkan/anv_blorp.c b/src/intel/vulkan/anv_blorp.c
> index d79c5e0..c871f03 100644
> --- a/src/intel/vulkan/anv_blorp.c
> +
On Fri, Mar 10, 2017 at 01:38:15PM +0100, Iago Toral Quiroga wrote:
> This situation can happen if we failed to allocate memory for the shader.
>
> v2:
> - We shouldn't see NULL shaders in anv_shader_bin_ref so we should not check
>for that (Jason). Make sure that callers don't attempt to cal
On Fri, Mar 10, 2017 at 01:38:21PM +0100, Iago Toral Quiroga wrote:
> ---
> src/intel/vulkan/anv_batch_chain.c | 9 +++--
> 1 file changed, 7 insertions(+), 2 deletions(-)
Reviewed-by: Topi Pohjolainen
>
> diff --git a/src/intel/vulkan/anv_batch_chain.c
> b/src/intel/vulkan/anv_batch_chai
On Fri, Mar 10, 2017 at 01:38:23PM +0100, Iago Toral Quiroga wrote:
> Growing the reloc list happens through calling anv_reloc_list_add() or
> anv_reloc_list_append(). Make sure that we call these through helpers
> that check the result and set the batch error status if needed.
>
> v2:
> - Handl
On Tue, 2017-03-14 at 10:12 +0200, Pohjolainen, Topi wrote:
> On Fri, Mar 10, 2017 at 01:38:16PM +0100, Iago Toral Quiroga wrote:
> >
> > ---
> > src/intel/vulkan/anv_blorp.c | 3 +++
> > 1 file changed, 3 insertions(+)
> >
> > diff --git a/src/intel/vulkan/anv_blorp.c
> > b/src/intel/vulkan/anv
Looks good to me :)
Reviewed-by: Plamena Manolova
On Tue, Mar 14, 2017 at 4:34 AM, Jason Ekstrand
wrote:
> The original version was very convoluted and tried way too hard to not
> just have the nested switch statement that it needs. Let's just write
> the obvious code and then we know it's cor
On Fri, Mar 10, 2017 at 01:38:34PM +0100, Iago Toral Quiroga wrote:
> Instead of asserting inside the function, and then use use that information
> to return early from its callers upon failure.
> ---
> src/intel/vulkan/anv_blorp.c | 35 ---
> src/intel/vulkan
On Tue, 2017-03-14 at 10:22 +0200, Pohjolainen, Topi wrote:
> On Fri, Mar 10, 2017 at 01:38:15PM +0100, Iago Toral Quiroga wrote:
> >
> > This situation can happen if we failed to allocate memory for the
> > shader.
> >
> > v2:
> > - We shouldn't see NULL shaders in anv_shader_bin_ref so we shou
Looks good to me.
Reviewed-by: Plamena Manolova
On Tue, Mar 14, 2017 at 2:00 AM, Mauro Rossi wrote:
> Fixes the following building error:
>
> out/target/product/x86_64/obj_x86/STATIC_LIBRARIES/libmesa_
> intel_compiler_intermediates/compiler/brw_nir_trig_workarounds.c:1:10:
> fatal error: 'brw_
Sorry Mauro seems like an identical patch has been approved and pushed a
couple of hours ago.
Thank you,
Plamena
On Tue, Mar 14, 2017 at 11:42 AM, Manolova, Plamena <
plamena.manol...@intel.com> wrote:
> Looks good to me.
> Reviewed-by: Plamena Manolova
>
> On Tue, Mar 14, 2017 at 2:00 AM, Maur
On Tue, Mar 14, 2017 at 10:35:43AM +0100, Iago Toral wrote:
> On Tue, 2017-03-14 at 10:12 +0200, Pohjolainen, Topi wrote:
> > On Fri, Mar 10, 2017 at 01:38:16PM +0100, Iago Toral Quiroga wrote:
> > >
> > > ---
> > > src/intel/vulkan/anv_blorp.c | 3 +++
> > > 1 file changed, 3 insertions(+)
> > >
https://bugs.freedesktop.org/show_bug.cgi?id=99467
--- Comment #21 from Drago ---
Since workaround is not that big, can't it be committed and activated with
ENV var switch, until Id fixes the game. This will allow stressing the RADV
driver with the most demanding Vulkan game we have, and polishin
On 04/03/17 00:19, Francisco Jerez wrote:
> Samuel Iglesias Gonsálvez writes:
>
>> Add support to SEL instruction and add an assert to detect unsupported
>> instructions than do d2x conversions.
>>
>> Signed-off-by: Samuel Iglesias Gonsálvez
>> ---
>>
>> Curro, this patch legalizes SEL instruc
On 14/03/17 11:17, Samuel Iglesias Gonsálvez wrote:
>
>
> On 04/03/17 00:19, Francisco Jerez wrote:
>> Samuel Iglesias Gonsálvez writes:
>>
>>> Add support to SEL instruction and add an assert to detect unsupported
>>> instructions than do d2x conversions.
>>>
>>> Signed-off-by: Samuel Iglesia
On Tue, Mar 14, 2017 at 4:06 AM, Timothy Arceri wrote:
> This should help reduce any overhead added by the shader cache
> when programs are not found in the cache.
>
> To avoid creating any special function just for the sake of the
> tests we add a one second delay whenever we call dick_cache_put(
It's not a good practice to srand() in libraries [1]. Not sure what to
suggest though, maybe rand_r() but "POSIX.1-2008 marks rand_r() as
obsolete". Maybe somebody else (Emil?) has some ideas?
[1]
http://stackoverflow.com/questions/14883035/problems-of-using-srand-in-libraries
Gražvydas
On Tue
Hi,
On 11.03.2017 03:14, Kenneth Graunke wrote:
On systems without LLC, drm_intel_gem_bo_map_unsynchronized() has
had the surprising behavior of doing a synchronized GTT mapping.
This is obviously not what the user of the API wanted.
Eric left a comment indicating a valid concern: if the CPU an
On Tue, 2017-03-14 at 11:25 +0200, Pohjolainen, Topi wrote:
> On Fri, Mar 10, 2017 at 01:38:23PM +0100, Iago Toral Quiroga wrote:
> >
> > Growing the reloc list happens through calling anv_reloc_list_add()
> > or
> > anv_reloc_list_append(). Make sure that we call these through
> > helpers
> > tha
It seems new extensions will start to use aliases in enums, meaning
multiple enums with the same value. This forces us to track the values
of the enums to generate correct C code.
Signed-off-by: Lionel Landwerlin
Cc: Dylan Baker
---
src/vulkan/util/gen_enum_to_str.py | 49 ++
On Tue, 2017-03-14 at 12:01 +0100, Iago Toral wrote:
> On Tue, 2017-03-14 at 11:25 +0200, Pohjolainen, Topi wrote:
> >
> > On Fri, Mar 10, 2017 at 01:38:23PM +0100, Iago Toral Quiroga wrote:
> > >
> > >
> > > Growing the reloc list happens through calling
> > > anv_reloc_list_add()
> > > or
> >
On 14/03/17 21:27, Grazvydas Ignotas wrote:
On Tue, Mar 14, 2017 at 4:06 AM, Timothy Arceri wrote:
This should help reduce any overhead added by the shader cache
when programs are not found in the cache.
To avoid creating any special function just for the sake of the
tests we add a one secon
On Tue, 2017-03-14 at 11:39 +0200, Pohjolainen, Topi wrote:
> On Fri, Mar 10, 2017 at 01:38:34PM +0100, Iago Toral Quiroga wrote:
> >
> > Instead of asserting inside the function, and then use use that
> > information
> > to return early from its callers upon failure.
> > ---
> > src/intel/vulkan
On Tue, Mar 14, 2017 at 12:18:59PM +0100, Iago Toral wrote:
> On Tue, 2017-03-14 at 11:39 +0200, Pohjolainen, Topi wrote:
> > On Fri, Mar 10, 2017 at 01:38:34PM +0100, Iago Toral Quiroga wrote:
> > >
> > > Instead of asserting inside the function, and then use use that
> > > information
> > > to r
On Tue, Mar 14, 2017 at 4:08 AM, Timothy Arceri wrote:
> From: Alan Swanson
>
> Currently only a one in one out eviction so if at max_size and
> cache files were to constantly increase in size then so would the
> cache. Restrict to limit of 8 evictions per new cache entry.
>
> V2: (Timothy Arceri
On Tue, Mar 14, 2017 at 12:06:16PM +0100, Iago Toral wrote:
> On Tue, 2017-03-14 at 12:01 +0100, Iago Toral wrote:
> > On Tue, 2017-03-14 at 11:25 +0200, Pohjolainen, Topi wrote:
> > >
> > > On Fri, Mar 10, 2017 at 01:38:23PM +0100, Iago Toral Quiroga wrote:
> > > >
> > > >
> > > > Growing the r
On 13 March 2017 at 19:24, Emil Velikov wrote:
> On 13 March 2017 at 17:14, Eric Engestrom wrote:
>> drm-common.h:63:49: warning: ‘struct egl’ declared inside parameter list
>> will not be
>> visible outside of this definition or declaration
>> int (*run)(const struct gbm *gbm, const struct eg
Namely:
AC_CONFIG_SRCDIR - it's used to "help" people who deliberatelly override
--srcdir. Nobody should be doing this to begin with, so just let them
get what they're asking for.
AC_CONFIG_HEADERS - we don't have any conditionals, function or macro
checks. Hence we don't need the header.
Signed-
Both flags are widely available and the rest of MAYBE_WARN are of little
interest. Since atm no flags were passed, we might as well not bother
with anything but the former two.
Signed-off-by: Emil Velikov
---
This will prompt some warnings, but we can sort this at a later stage.
---
Makefile.am
We're creating a single binary with no static/shared libraries.
We implicitly relied on the macro to pull AC_PROG_CC, so we should add
it now, otherwise configure will fail.
As a nice bonus, with the last three commits combined the execution time
of autogen.sh drops by more than half.
be
With the removal of libtool as of last commit we no longer need the m4
folder or any of the related infra.
Signed-off-by: Emil Velikov
---
Makefile.am | 2 --
configure.ac | 1 -
m4/.gitignore | 5 -
3 files changed, 8 deletions(-)
delete mode 100644 m4/.gitignore
diff --git a/Makefile.
Signed-off-by: Emil Velikov
---
.gitignore | 17 -
1 file changed, 17 deletions(-)
diff --git a/.gitignore b/.gitignore
index 3f95cde..1fc8950 100644
--- a/.gitignore
+++ b/.gitignore
@@ -5,26 +5,9 @@ Makefile
.deps
.libs
*.o
-*.lo
-*.la
-libtool
-*.pc
config.log
config.stat
On 4 March 2017 at 01:12, Jason Ekstrand wrote:
> Cc: "17.0 13.0"
> ---
> src/compiler/nir/nir_intrinsics.h | 4 +++-
> 1 file changed, 3 insertions(+), 1 deletion(-)
>
> diff --git a/src/compiler/nir/nir_intrinsics.h
> b/src/compiler/nir/nir_intrinsics.h
> index f45bfe2..5c8f283 100644
> --- a
https://bugs.freedesktop.org/show_bug.cgi?id=77449
Bug 77449 depends on bug 99194, which changed state.
Bug 99194 Summary: Saints Row IV and Unturned cause GPU resets
https://bugs.freedesktop.org/show_bug.cgi?id=99194
What|Removed |Added
Ilia, are you waiting for another person to review the patch or it's
no longer applicable ?
Thanks
Emil
On 10 March 2017 at 11:07, Marek Olšák wrote:
> Reviewed-by: Marek Olšák
>
> Marek
>
> On Sat, Mar 4, 2017 at 7:52 PM, Ilia Mirkin wrote:
>> This prevents textureQueryLevels, which maps as L
On 14 March 2017 at 01:45, Jason Ekstrand wrote:
> On Mon, Mar 13, 2017 at 5:34 PM, Emil Velikov
> wrote:
>>
>> On 13 March 2017 at 19:08, Jason Ekstrand wrote:
>> > On Mon, Mar 13, 2017 at 11:19 AM, Emil Velikov
>> >
>> > wrote:
>> >>
>> >> On 5 March 2017 at 16:36, Eduardo Lima Mitev wrote:
On Tue, 2017-03-14 at 11:53 +0200, Pohjolainen, Topi wrote:
> On Tue, Mar 14, 2017 at 10:35:43AM +0100, Iago Toral wrote:
> >
> > On Tue, 2017-03-14 at 10:12 +0200, Pohjolainen, Topi wrote:
> > >
> > > On Fri, Mar 10, 2017 at 01:38:16PM +0100, Iago Toral Quiroga
> > > wrote:
> > > >
> > > >
> >
On Tuesday, 2017-03-14 12:10:48 +, Emil Velikov wrote:
> Both flags are widely available and the rest of MAYBE_WARN are of little
> interest. Since atm no flags were passed, we might as well not bother
> with anything but the former two.
>
> Signed-off-by: Emil Velikov
> ---
> This will promp
On Tue, 2017-03-14 at 14:04 +0200, Pohjolainen, Topi wrote:
> On Tue, Mar 14, 2017 at 12:06:16PM +0100, Iago Toral wrote:
> >
> > On Tue, 2017-03-14 at 12:01 +0100, Iago Toral wrote:
> > >
> > > On Tue, 2017-03-14 at 11:25 +0200, Pohjolainen, Topi wrote:
> > > >
> > > >
> > > > On Fri, Mar 10,
BTW, in case it is useful, I have the series available here, including
fixes to review feedback obtained so far:
https://github.com/Igalia/mesa/tree/vk-oom
Iago
On Fri, 2017-03-10 at 13:38 +0100, Iago Toral Quiroga wrote:
> I think this series addresses all the issues raised for v1, except
> mak
I am waiting to have time to do proper testing on the patch like I
promised. Thus far, the time to do so has failed to materialize.
On Mar 14, 2017 8:27 AM, "Emil Velikov" wrote:
> Ilia, are you waiting for another person to review the patch or it's
> no longer applicable ?
>
> Thanks
> Emil
>
>
Signed-off-by: Eric Engestrom
---
cube-tex.c | 6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)
diff --git a/cube-tex.c b/cube-tex.c
index 1e7741d..a543e83 100644
--- a/cube-tex.c
+++ b/cube-tex.c
@@ -227,7 +227,7 @@ static int get_fd_rgba(uint32_t *pstride)
map = gbm_bo_map(b
`ret` isn't used by anything, so remove it as well.
Signed-off-by: Eric Engestrom
---
kmscube.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/kmscube.c b/kmscube.c
index 91e0104..8057e66 100644
--- a/kmscube.c
+++ b/kmscube.c
@@ -67,7 +67,7 @@ int main(int argc, char *a
Signed-off-by: Eric Engestrom
---
drm-common.c | 1 -
1 file changed, 1 deletion(-)
diff --git a/drm-common.c b/drm-common.c
index 764ffa2..0262816 100644
--- a/drm-common.c
+++ b/drm-common.c
@@ -35,7 +35,6 @@ drm_fb_destroy_callback(struct gbm_bo *bo, void *data)
{
int drm_fd = gbm_de
Signed-off-by: Eric Engestrom
---
drm-atomic.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drm-atomic.c b/drm-atomic.c
index 3362eac..01fa35a 100644
--- a/drm-atomic.c
+++ b/drm-atomic.c
@@ -362,7 +362,7 @@ const struct drm * init_drm_atomic(const char *device)
get
Signed-off-by: Eric Engestrom
---
drm-atomic.c | 3 +--
1 file changed, 1 insertion(+), 2 deletions(-)
diff --git a/drm-atomic.c b/drm-atomic.c
index 01fa35a..5d64e50 100644
--- a/drm-atomic.c
+++ b/drm-atomic.c
@@ -108,7 +108,6 @@ static int drm_atomic_commit(uint32_t fb_id, uint32_t flags)
Signed-off-by: Eric Engestrom
---
I feel like I might be missing something here; is copying the fields
into these vars of any use when debugging?
---
drm-atomic.c | 4
1 file changed, 4 deletions(-)
diff --git a/drm-atomic.c b/drm-atomic.c
index 5d64e50..27c6b1e 100644
--- a/drm-atomic.c
++
Signed-off-by: Eric Engestrom
---
drm-legacy.c | 3 +++
1 file changed, 3 insertions(+)
diff --git a/drm-legacy.c b/drm-legacy.c
index 2392a3d..8e49075 100644
--- a/drm-legacy.c
+++ b/drm-legacy.c
@@ -33,6 +33,9 @@ static struct drm drm;
static void page_flip_handler(int fd, unsigned int frame,
On 14 March 2017 at 13:16, Ilia Mirkin wrote:
> I am waiting to have time to do proper testing on the patch like I promised.
> Thus far, the time to do so has failed to materialize.
>
Right - I seems to have misread your reply s/will do some testing
tonight/have done some testing/ :-\
-Emil
_
On Tue, Mar 14, 2017 at 02:08:10PM +0100, Iago Toral wrote:
> On Tue, 2017-03-14 at 14:04 +0200, Pohjolainen, Topi wrote:
> > On Tue, Mar 14, 2017 at 12:06:16PM +0100, Iago Toral wrote:
> > >
> > > On Tue, 2017-03-14 at 12:01 +0100, Iago Toral wrote:
> > > >
> > > > On Tue, 2017-03-14 at 11:25 +0
> Subject: kmscube.c: remove uninitialized var returned
Huh? Not sure what my brain was thinking; changed that locally to
"don't return uninitialized variable" :)
On Tuesday, 2017-03-14 13:33:51 +, Eric Engestrom wrote:
> `ret` isn't used by anything, so remove it as well.
>
> Signed-off-by
On Tue, Mar 14, 2017 at 04:02:17PM +0200, Pohjolainen, Topi wrote:
> On Tue, Mar 14, 2017 at 02:08:10PM +0100, Iago Toral wrote:
> > On Tue, 2017-03-14 at 14:04 +0200, Pohjolainen, Topi wrote:
> > > On Tue, Mar 14, 2017 at 12:06:16PM +0100, Iago Toral wrote:
> > > >
> > > > On Tue, 2017-03-14 at 1
Hi,
On 14.03.2017 12:48, Eero Tamminen wrote:
On 11.03.2017 03:14, Kenneth Graunke wrote:
On systems without LLC, drm_intel_gem_bo_map_unsynchronized() has
had the surprising behavior of doing a synchronized GTT mapping.
This is obviously not what the user of the API wanted.
Eric left a commen
---
src/intel/vulkan/anv_device.c | 7 +--
1 file changed, 5 insertions(+), 2 deletions(-)
diff --git a/src/intel/vulkan/anv_device.c b/src/intel/vulkan/anv_device.c
index d2d9117..c73b9af 100644
--- a/src/intel/vulkan/anv_device.c
+++ b/src/intel/vulkan/anv_device.c
@@ -412,8 +412,11 @@ VkRe
On Tue, Mar 14, 2017 at 5:23 AM, Emil Velikov
wrote:
> On 14 March 2017 at 01:45, Jason Ekstrand wrote:
> > On Mon, Mar 13, 2017 at 5:34 PM, Emil Velikov
> > wrote:
> >>
> >> On 13 March 2017 at 19:08, Jason Ekstrand wrote:
> >> > On Mon, Mar 13, 2017 at 11:19 AM, Emil Velikov
> >> >
> >> > w
https://bugs.freedesktop.org/show_bug.cgi?id=99467
--- Comment #22 from Vedran Miletić ---
(In reply to Drago from comment #21)
> Since workaround is not that big, can't it be committed and activated with
> ENV var switch, until Id fixes the game. This will allow stressing the RADV
> driver with
This makes it so that you don't get an "Implement gen7 HiZ" perf warning
when you manually disable HiZ on gen8.
---
src/intel/vulkan/anv_image.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/src/intel/vulkan/anv_image.c b/src/intel/vulkan/anv_image.c
index 5f17351..931ee2
The GL driver had a driconf option (which doesn't make much sense) and
the Vulkan driver had a hand-rolled environment variable. Instead,
let's tie both into the INTEL_DEBUG mechanism and unify things.
---
src/intel/common/gen_debug.c | 1 +
src/intel/common/gen_debug.h |
---
src/intel/vulkan/anv_private.h | 59 ++
src/intel/vulkan/genX_cmd_buffer.c | 48 ++-
2 files changed, 62 insertions(+), 45 deletions(-)
diff --git a/src/intel/vulkan/anv_private.h b/src/intel/vulkan/anv_private.h
index b11adf
---
src/intel/vulkan/anv_pass.c | 20 +++-
1 file changed, 11 insertions(+), 9 deletions(-)
diff --git a/src/intel/vulkan/anv_pass.c b/src/intel/vulkan/anv_pass.c
index 4a1a340..8d1768d 100644
--- a/src/intel/vulkan/anv_pass.c
+++ b/src/intel/vulkan/anv_pass.c
@@ -59,10 +59,8 @@ V
---
src/intel/vulkan/anv_pass.c| 87 ++
src/intel/vulkan/anv_private.h | 2 +
2 files changed, 89 insertions(+)
diff --git a/src/intel/vulkan/anv_pass.c b/src/intel/vulkan/anv_pass.c
index 8d1768d..d6d3794 100644
--- a/src/intel/vulkan/anv_pass.c
+++ b
Instead of figuring it all out ourselves, just use the information given
to us by the client.
---
src/intel/vulkan/anv_blorp.c | 88 --
src/intel/vulkan/genX_cmd_buffer.c | 10 +
2 files changed, 18 insertions(+), 80 deletions(-)
diff --git a/src/inte
Need to flush before updating the buffer to ensure that the copy is
ordered after previous accesses (assuming the app has performed the
appropriate barriers).
This fixes potential issues due to draws prior to an update reading
the new buffer content, despite having the necessary barriers between
t
Hi,
On a imx6q-sabresd board running kernel 4.9.14, mesa 17.0.1 and QT5.8:
# export QT_QPA_EGLFS_KMS_CONFIG=/media/sabresd.json
where sabresd.json contains:
{
"device": "/dev/dri/card1",
"hwcursor": false,
"pbuffers": true,
"outputs": [
{
"name": "HDMI-1",
"mode": "off"
https://bugs.freedesktop.org/show_bug.cgi?id=100199
Bug ID: 100199
Summary: clover: implement non-blocking clEnqueueWriteBuffer
and clEnqueueReadBuffer
Product: Mesa
Version: git
Hardware: Other
OS: All
https://bugs.freedesktop.org/show_bug.cgi?id=100199
Vedran Miletić changed:
What|Removed |Added
Assignee|mesa-dev@lists.freedesktop. |ved...@miletic.net
|
On Tue, Mar 14, 2017 at 4:04 AM, Lionel Landwerlin <
lionel.g.landwer...@intel.com> wrote:
> It seems new extensions will start to use aliases in enums, meaning
> multiple enums with the same value. This forces us to track the values
> of the enums to generate correct C code.
>
Seriously? That s
https://bugs.freedesktop.org/show_bug.cgi?id=100199
--- Comment #1 from Vedran Miletić ---
Created attachment 130217
--> https://bugs.freedesktop.org/attachment.cgi?id=130217&action=edit
Attempt at a fix
--
You are receiving this mail because:
You are the QA Contact for the bug.__
https://bugs.freedesktop.org/show_bug.cgi?id=99987
--- Comment #14 from Pierre Ossman ---
That does make things work. But I'm hesitant to consider the issue closed
though. There are some outstanding questions:
a) I'm still getting direct rendering, despite the file being called
libGLX_indirect.s
On Mon, Mar 13, 2017 at 10:24 PM, Ben Widawsky wrote:
> The idea behind modifiers like this is that the user of GBM will have
> some mechanism to query what properties the hardware supports for its BO
> or surface. This information is directly passed in (and stored) so that
> the DRI implementati
The comments below are my only remaining questions on this 6-patch series.
All of the *other* patches are
Reviewed-by: Jason Ekstrand
On Tue, Mar 14, 2017 at 8:53 AM, Jason Ekstrand
wrote:
> On Mon, Mar 13, 2017 at 10:24 PM, Ben Widawsky wrote:
>
>> The idea behind modifiers like this is that
https://bugs.freedesktop.org/show_bug.cgi?id=100201
Bug ID: 100201
Summary: llvmpipe Windows scons build with Visual Studio 2015
toolchain and LLVM 4.0 fails
Product: Mesa
Version: 17.0
Hardware: All
OS: Win
On Tue, Mar 14, 2017 at 07:55:01AM -0700, Jason Ekstrand wrote:
> This makes it so that you don't get an "Implement gen7 HiZ" perf warning
> when you manually disable HiZ on gen8.
Both:
Reviewed-by: Topi Pohjolainen
> ---
> src/intel/vulkan/anv_image.c | 4 ++--
> 1 file changed, 2 insertions(
https://bugs.freedesktop.org/show_bug.cgi?id=100202
Bug ID: 100202
Summary: llvmpipe Windows scons build can't detect Visual
Studio 2017 toolchain
Product: Mesa
Version: 17.0
Hardware: All
OS: Windows (All)
Hi Otavio,
On Tue, Mar 14, 2017 at 1:58 PM, Otavio Salvador
wrote:
> On Tue, Mar 14, 2017 at 12:31 PM, Fabio Estevam wrote:
> ...
>> The cover of the movies are just black rectangles instead of showing
>> the actual movie pictures.
> ...
>
> I am attaching the two patches I am applying on mesa l
https://bugs.freedesktop.org/show_bug.cgi?id=99987
--- Comment #15 from Dennis Schridde ---
(In reply to Pierre Ossman from comment #14)
> b) If it is indeed a fallback, why is that needed? I.e. why isn't Mesa being
> chosen? It does not seem robust and future proof to rely on a fallback
> mechan
Hi Fabio
2017-03-14 18:16 GMT+01:00 Fabio Estevam :
> Hi Otavio,
>
> On Tue, Mar 14, 2017 at 1:58 PM, Otavio Salvador
> wrote:
>> On Tue, Mar 14, 2017 at 12:31 PM, Fabio Estevam wrote:
>> ...
>>> The cover of the movies are just black rectangles instead of showing
>>> the actual movie pictures.
On Tue, Mar 14, 2017 at 06:39:32PM +0100, Christian Gmeiner wrote:
> > By only reverting:
> >
> > commit 6c89a728d9e5d072cb504453e73077564c6523d3
> > Author: Wladimir J. van der Laan
> > Date: Wed Dec 7 12:59:54 2016 +
> >
> > etnaviv: Cannot render to rb-swapped formats
> >
> > Expo
Hi Wladimir,
On Tue, Mar 14, 2017 at 2:47 PM, Wladimir J. van der Laan
wrote:
> Yea, variants would be the way to render to RGBA succesfully
> on vivantes.
>
> Until then it's best to revert those two patches.
>
> Reverting only the one (latest) patch can give red/blue swapped issues when
> doin
https://bugs.freedesktop.org/show_bug.cgi?id=100199
--- Comment #2 from Vedran Miletić ---
Benchmarking ftp://ftp.gromacs.org/pub/benchmarks/water_GMX50_bare.tar.gz with
GROMACS 2016.3 I get the following results:
Without the patch:
Launch GPU ops.18 66001 16.90546
On Tue, 2017-03-14 at 05:22 +0100, Federico Dossena wrote:
> The game runs smoothly if you decompress the textures, I'm getting 30-40
> fps with llvmpipe and no "grass in the sky" artifacts.
> The problem is frame buffer effects and soft shadows, were they enabled
> when you tested it? If so, how
https://bugs.freedesktop.org/show_bug.cgi?id=94739
Luca Weiss changed:
What|Removed |Added
CC||bugzi...@z3ntu.xyz
--- Comment #6 from Luca
On 17-03-14 08:53:45, Jason Ekstrand wrote:
On Mon, Mar 13, 2017 at 10:24 PM, Ben Widawsky wrote:
The idea behind modifiers like this is that the user of GBM will have
some mechanism to query what properties the hardware supports for its BO
or surface. This information is directly passed in (a
The flushes could be due to TRANSFER barriers.
Signed-off-by: Bas Nieuwenhuizen
Cc: 17.0
---
src/amd/vulkan/si_cmd_buffer.c | 3 +++
1 file changed, 3 insertions(+)
diff --git a/src/amd/vulkan/si_cmd_buffer.c b/src/amd/vulkan/si_cmd_buffer.c
index 5d35287f8e3..b808052ddb2 100644
--- a/src/amd/
---
src/compiler/nir/nir_constant_expressions.py | 199 ++-
1 file changed, 101 insertions(+), 98 deletions(-)
diff --git a/src/compiler/nir/nir_constant_expressions.py
b/src/compiler/nir/nir_constant_expressions.py
index c6745f1..ad841e3 100644
--- a/src/compiler/nir/nir
On 17-03-14 13:31:01, Ben Widawsky wrote:
On 17-03-14 08:53:45, Jason Ekstrand wrote:
On Mon, Mar 13, 2017 at 10:24 PM, Ben Widawsky wrote:
The idea behind modifiers like this is that the user of GBM will have
some mechanism to query what properties the hardware supports for its BO
or surface
For opcodes such as the nir_op_pack_64_2x32 for which all sources and
destinations have explicit sizes, the bit_size parameter to the evaluate
function is pointless and *should* do nothing. Previously, we were
always switching on the bit_size and asserting if it isn't one of the
sizes in the list.
From: Dave Airlie
LLVM 4.0 released with a pretty messy regression, that hopefully
get fixed in the future.
This work around was proposed by Tom, and it fixes the CTS regressions
here at least, I'm not sure if this will cause any major side effects,
but correctness over speed and all that.
rade
On 15 March 2017 at 07:17, Dave Airlie wrote:
> From: Dave Airlie
>
> LLVM 4.0 released with a pretty messy regression, that hopefully
> get fixed in the future.
>
> This work around was proposed by Tom, and it fixes the CTS regressions
> here at least, I'm not sure if this will cause any major s
The idea behind modifiers like this is that the user of GBM will have
some mechanism to query what properties the hardware supports for its BO
or surface. This information is directly passed in (and stored) so that
the DRI implementation can create an image with the appropriate
attributes.
A gette
I've skimmed to changes from 1.0.5 to 1.0.42 and I think we have all
changes. We're still not conformant ofcourse, but this should not
regress stuff,
Signed-off-by: Bas Nieuwenhuizen
---
src/amd/vulkan/radv_device.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/src/amd/vulk
Thanks. Pushed.
On Tue, Mar 14, 2017 at 4:26 PM, Alex Smith wrote:
> Need to flush before updating the buffer to ensure that the copy is
> ordered after previous accesses (assuming the app has performed the
> appropriate barriers).
>
> This fixes potential issues due to draws prior to an update r
I couldn't really find an encoding in the spec. I'm not sure it
prescribes VK_MAKE_VERSION format, but vulkan.gpuinfo.org interprets
it that way by default. vulkaninfo gives the raw number, so we could
alternatively do something like 17001000, but that doesn't show
up right on vulkan.gpuinfo.org ag
I think we just don't want to store those shaders.
Signed-off-by: Samuel Pitoiset
---
run.c | 1 +
1 file changed, 1 insertion(+)
diff --git a/run.c b/run.c
index 74f87fd..12a78c6 100644
--- a/run.c
+++ b/run.c
@@ -392,6 +392,7 @@ main(int argc, char **argv)
setenv("allow_glsl_extension_
14.03.2017 u 23:08, Bas Nieuwenhuizen je napisao/la:
I couldn't really find an encoding in the spec. I'm not sure it
prescribes VK_MAKE_VERSION format, but vulkan.gpuinfo.org interprets
it that way by default. vulkaninfo gives the raw number, so we could
alternatively do something like 17001000,
On Mon, Mar 13, 2017 at 03:27:58PM -0700, Chad Versace wrote:
> isl has a giant comment that explains the hardware's padding
> requirements. (Hint: Cache lines and page faults). But the comment is in
> the wrong place, in isl_calc_linear_row_pitch(), which is unrelated to
> padding.
>
> The import
On 03/14/2017 10:21 PM, Dave Airlie wrote:
On 15 March 2017 at 07:17, Dave Airlie wrote:
From: Dave Airlie
LLVM 4.0 released with a pretty messy regression, that hopefully
get fixed in the future.
This work around was proposed by Tom, and it fixes the CTS regressions
here at least, I'm not
Reviewed-by: Marek Olšák
Marek
On Wed, Mar 15, 2017 at 12:17 AM, Samuel Pitoiset
wrote:
> I think we just don't want to store those shaders.
>
> Signed-off-by: Samuel Pitoiset
> ---
> run.c | 1 +
> 1 file changed, 1 insertion(+)
>
> diff --git a/run.c b/run.c
> index 74f87fd..12a78c6 100644
On Mon, Mar 13, 2017 at 03:27:59PM -0700, Chad Versace wrote:
> The calculations of row_pitch, the row pitch's alignment, surface size,
> and base_alignment were mixed together. This patch moves the calculation
> of row_pitch and its alignment to occur before the calculation of
> surface_size and b
Initially this was a workaround for a bug introduced in LLVM 4.0
in the SimplifyCFG pass that caused image instrinsics to disappear
(because they were badly sunk). Finally, this is a win because it
decreases SGPR spilling and increases the number of waves a bit.
Although, shader-db results are goo
1 - 100 of 120 matches
Mail list logo
