There is no need to hardcode it, we can just use blob_key[0].
This is needed because the next patches are going to change how cache
keys are computed.
Signed-off-by: Grazvydas Ignotas
---
src/compiler/glsl/tests/cache_test.c | 8 ++--
1 file changed, 2 insertions(+), 6 deletions(-)
diff
We no longer need a string, so can just pass the timestamps as a blob to
reduce amount of data that needs to be hashed into the keys. The
argument is also renamed to version_blob because the cache doesn't
really care if it's timestamps, build id or something else.
Signed-off-by: Grazvyd
This will allow to hash additional data into the cache keys or even
change the hashing algorithm easily, should we decide to do so.
v2: don't try to compute key (and crash) if cache is disabled
Signed-off-by: Grazvydas Ignotas
---
src/compiler/glsl/glsl_parser_extras.cpp
://bugs.freedesktop.org/show_bug.cgi?id=100091
Signed-off-by: Grazvydas Ignotas
---
src/compiler/glsl/tests/cache_test.c | 10 ++---
src/util/disk_cache.c| 71 ++--
2 files changed, 25 insertions(+), 56 deletions(-)
diff --git a/src/compiler/glsl/tests/cache_test.c
This can be used to deal with key hash collisions from different
version (should we find that to actually happen) and to find
which mesa version produced it the entry.
Signed-off-by: Grazvydas Ignotas
---
src/util/disk_cache.c | 24 ++--
1 file changed, 22 insertions(+), 2
On Thu, Mar 16, 2017 at 3:28 AM, Timothy Arceri wrote:
> On 16/03/17 10:09, Grazvydas Ignotas wrote:
>>
>> This can be used to deal with key hash collisions from different
>> version (should we find that to actually happen) and to find
>> which mesa version produced it
On Thu, Mar 16, 2017 at 3:00 AM, Timothy Arceri wrote:
>
>
> On 16/03/17 10:09, Grazvydas Ignotas wrote:
>>
>> Simplifies the write code a bit and handles EINTR.
>>
>> Signed-off-by: Grazvydas Ignotas
>> ---
>> src/util/disk_cache.c | 40 ++
On Thu, Mar 16, 2017 at 2:49 AM, Timothy Arceri wrote:
>
>
> On 16/03/17 10:09, Grazvydas Ignotas wrote:
>>
>> This allows to get rid of the arch and gpu name directories.
>> Also, gpu_name is now optional.
>>
>> Signed-off-by: Grazvydas Ignotas
>>
On Thu, Mar 16, 2017 at 3:14 PM, Timothy Arceri wrote:
>
>
> On 17/03/17 00:00, Grazvydas Ignotas wrote:
>>
>> On Thu, Mar 16, 2017 at 2:49 AM, Timothy Arceri
>> wrote:
>>>
>>>
>>>
>>> On 16/03/17 10:09, Grazvydas Ignotas wrote:
On Thu, Mar 16, 2017 at 7:36 PM, Alan Swanson wrote:
> Since switching to LRU eviction the only user of these predicate
> functions now resolves directory entry stats itself so pass them
> directly saving calling fstat and strlen twice (and the
> expensive strlen is skipped entirely if access time
On Thu, Mar 9, 2017 at 9:07 PM, Emil Velikov wrote:
> From: Jason Ekstrand
>
> v2 [Emil Velikov]
> - Various fixes and initial stab at the Android build.
> - Keep the generation rules/EXTRA_DIST outside the conditional
This has broken anv build for me, because I don't have vulkan.h
anywhere in
or directory
Fixes: ee8044fd "intel/vulkan: Get rid of recursive make"
Cc: Emil Velikov
Signed-off-by: Grazvydas Ignotas
---
src/intel/Makefile.vulkan.am | 1 +
1 file changed, 1 insertion(+)
diff --git a/src/intel/Makefile.vulkan.am b/src/intel/Makefile.vulkan.am
index 1857802..e7c1
ly if access time is newer).
>
> v2: Update for empty cache dir detection changes
> v3: Fix passing string length to predicate with the +1 for NULL
> termination and also pass sb as pointer
> v4: Missed ampersand for passing sb as pointer
Rev
: Grazvydas Ignotas
---
src/util/disk_cache.c | 4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)
diff --git a/src/util/disk_cache.c b/src/util/disk_cache.c
index 6c91f23..f6768d5 100644
--- a/src/util/disk_cache.c
+++ b/src/util/disk_cache.c
@@ -841,12 +841,14 @@ cache_put(void *job, int
I haven't seen this causing problems in practice, but for correctness
we should also check if rename succeeded to avoid breaking accounting
and leaving a .tmp file behind.
Signed-off-by: Grazvydas Ignotas
---
src/util/disk_cache.c | 8 ++--
1 file changed, 6 insertions(+), 2 dele
16s
- disk_cache_has_key() hit rate: ~50% -> ~96%
Signed-off-by: Grazvydas Ignotas
---
src/util/disk_cache.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/src/util/disk_cache.c b/src/util/disk_cache.c
index 7d10ba2..b88ef52 100644
--- a/src/util/disk_cache.c
+++ b/src/util/disk_c
On Sun, Mar 19, 2017 at 3:03 PM, Emil Velikov wrote:
> Hi Grazvydas,
>
> On 17 March 2017 at 22:05, Grazvydas Ignotas wrote:
>> Fixes build without vulkan.h installed in system header locations:
>> CC vulkan/vulkan_libvulkan_intel_la-anv_gem.lo
>> I
On Sun, Mar 19, 2017 at 9:30 PM, Jason Ekstrand wrote:
> On March 19, 2017 9:33:02 AM Grazvydas Ignotas wrote:
>
>> On Sun, Mar 19, 2017 at 3:03 PM, Emil Velikov
>> wrote:
>>>
>>> Hi Grazvydas,
>>>
>>> On 17 March 2017 at 22:05, Grazvydas
On Sun, Mar 19, 2017 at 11:21 PM, Emil Velikov wrote:
> On 19 March 2017 at 19:30, Jason Ekstrand wrote:
>> On March 19, 2017 9:33:02 AM Grazvydas Ignotas wrote:
>>
>>> On Sun, Mar 19, 2017 at 3:03 PM, Emil Velikov
>>> wrote:
>>>>
>>&
On Sun, Mar 19, 2017 at 11:27 PM, Emil Velikov wrote:
> On 19 March 2017 at 16:32, Grazvydas Ignotas wrote:
>
>> With that, it's picking the earlier installed
>> /opt/xorg/include/vulkan/vulkan_intel.h instead of
>> ../../include/vulkan/vulkan_intel.h , and that one
lly annoying, and using Emil's workarounds from
bug 100259 is cumbersome.
Tested-by: Grazvydas Ignotas
>
> It looks to me that in
>
> https://bugs.freedesktop.org/show_bug.cgi?id=100259
>
> Emil et al are discussing similar case with gbm.
>
> CC: Matt Turner
> CC
On Mon, Mar 20, 2017 at 9:43 PM, Pohjolainen, Topi
wrote:
> On Mon, Mar 20, 2017 at 04:45:22PM +, Emil Velikov wrote:
>> On 20 March 2017 at 10:27, Grazvydas Ignotas wrote:
>> > On Mon, Mar 20, 2017 at 11:28 AM, Topi Pohjolainen
>> > wrote:
>> >> Rec
d (or at least not in the same directory as
> vulkan_intel.h) the build will fail.
>
> In the future we might want to upstream the vulkan_intel.h within
> vulkan.h, but in either case the more robust thing is to rely on our own
> copy.
>
> Cc: Grazvydas Ignotas
> Fixes: ee80
ng libarary/dependency lists instead of
>> introducing duplicates (GL_LIB_DEPS contains DRI_LIB_DEPS).
>>
>> CC: Matt Turner
>> CC: Emil Velikov
>> Tested-by: Grazvydas Ignotas (v1)
>> Signed-off-by: Topi Pohjolainen
>> ---
>> src/glx/Makefile.am | 4
It might make sense to give more attention to cmake just because many
mesa-related projects are already using it: llvm, piglit, vulkan
loader and demos, mesa-demos, etc. Not sure how well it supports BSDs
and windows, but everyone building graphics stacks or contributing to
mesa should already be c
On Tue, Mar 21, 2017 at 10:57 AM, Timothy Arceri wrote:
> ---
> src/gallium/drivers/radeon/r600_test_dma.c | 3 +-
> src/util/rand_xor.c| 62
> --
> src/util/rand_xor.h| 6 ++-
> 3 files changed, 64 insertions(+), 7 de
On Tue, Mar 21, 2017 at 2:05 PM, Emil Velikov wrote:
> On 20 March 2017 at 21:51, Jason Ekstrand wrote:
>> On Mon, Mar 20, 2017 at 8:18 AM, Emil Velikov
>> wrote:
>>> diff --git a/src/intel/Makefile.vulkan.am b/src/intel/Makefile.vulkan.am
>>> index 185780246e2..7aacdf4e094 100644
>>> --- a/src/
This doesn't build for me:
make[4]: Entering directory
'/home/notaz/src/radeon/mesa/src/gallium/winsys/amdgpu/drm'
CC amdgpu_bo.lo
amdgpu_bo.c: In function ‘amdgpu_buffer_get_metadata’:
amdgpu_bo.c:615:53: error: ‘AMDGPU_TILING_SWIZZLE_MODE_SHIFT’
undeclared (first use in this function)
Rather than using 3 different ways to wrap _mesa_sha1_*() to SHA1*()
functions (a macro, prototype with implementation in .c and an inline
function), make all 3 inline functions.
Signed-off-by: Grazvydas Ignotas
---
src/util/mesa-sha1.c | 6 --
src/util/mesa-sha1.h | 13 ++---
2
bably be good enough for dir selection
purposes. For the series:
Reviewed-by: Grazvydas Ignotas
___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/mesa-dev
> v3: remove left over var from v1.
Reviewed-by: Grazvydas Ignotas
___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/mesa-dev
These are likely false positives, but are also annoying because they
show up on every "make install", which causes ac_nir_to_llvm to be
rebuilt here. Initializing those variables to NULL should be harmless
even when unnecessary.
Signed-off-by: Grazvydas Ignotas
---
src/
nformant radv vulkan driver"
Signed-off-by: Grazvydas Ignotas
---
src/amd/vulkan/radv_pipeline_cache.c | 3 +++
1 file changed, 3 insertions(+)
diff --git a/src/amd/vulkan/radv_pipeline_cache.c
b/src/amd/vulkan/radv_pipeline_cache.c
index 99a614d..beed35b 100644
--- a/src/amd/vulkan/radv_p
amd/common/ and amd/vulkan/ are using tabs for indent, which doesn't
match the settings in root .editorconfig, so let's override.
Signed-off-by: Grazvydas Ignotas
---
src/amd/common/.editorconfig | 3 +++
src/amd/vulkan/.editorconfig | 3 +++
2 files changed, 6 insertions(+)
create m
Valgrind reports it's being used uninitialized.
Fixes: f4e499ec791 "radv: add initial non-conformant radv vulkan driver"
Signed-off-by: Grazvydas Ignotas
---
src/amd/vulkan/radv_descriptor_set.c | 1 +
1 file changed, 1 insertion(+)
diff --git a/src/amd/vulkan/radv_descriptor_s
Wouldn't it be better to set AMDGPU_GEM_CREATE_VRAM_CLEARED when
creating the bo? That would need new RADEON_FLAG_* for winsys.
If not, at least make 4096 a define to use for both
ws->buffer_create() and memset.
Gražvydas
On Wed, Sep 13, 2017 at 12:26 PM, Samuel Pitoiset
wrote:
> To avoid rando
It was because on some platforms ./configure test looking for
__atomic_* passes because it's a compile (not link) test, even when
the compiler inserts library calls instead of actual implementation.
It looks like the correct solution is to link against libatomic (which
I was not aware of at that ti
On Fri, Sep 15, 2017 at 12:04 PM, Grazvydas Ignotas wrote:
> It was because on some platforms ./configure test looking for
> __atomic_* passes because it's a compile (not link) test, even when
> the compiler inserts library calls instead of actual implementation.
> It looks
On Fri, Oct 5, 2018 at 3:38 AM Dave Airlie wrote:
>
> From: Dave Airlie
>
> There may be some security or sandbox reason this might fail, so
> check and fail appropriately.
> ---
> src/amd/vulkan/radv_meta.c | 4 +++-
> 1 file changed, 3 insertions(+), 1 deletion(-)
>
> diff --git a/src/amd/vulk
Trivial.
---
src/amd/vulkan/radv_debug.c | 6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)
diff --git a/src/amd/vulkan/radv_debug.c b/src/amd/vulkan/radv_debug.c
index 08fc80c12ab..e81b9cccb57 100644
--- a/src/amd/vulkan/radv_debug.c
+++ b/src/amd/vulkan/radv_debug.c
@@ -320,11 +320,11
On Fri, Apr 6, 2018 at 3:28 PM, Samuel Pitoiset
wrote:
> Simple extension that only returns information for AMD hw.
>
> Signed-off-by: Samuel Pitoiset
> ---
> src/amd/vulkan/radv_device.c | 71
> +++
> src/amd/vulkan/radv_extensions.py | 1 +
> 2 files
301 - 341 of 341 matches
Mail list logo
