On Mon, 23 Jan 2017, Rodrigo Vivi wrote:
> Ander, can I carry your rv-b on this v2?
>
> Jani, with Ander confirmation can I merge these 2 patches?
Acked-by: Jani Nikula
>
>
> On Mon, Jan 23, 2017 at 11:24 AM, Patchwork
>
On Mon, 23 Jan 2017, Pierre-Louis Bossart
wrote:
>>> #include
>>> @@ -630,6 +631,10 @@ void intel_audio_codec_enable(struct intel_encoder
>>> *intel_encoder,
>>> if (acomp && acomp->audio_ops && acomp->audio_ops->pin_eld_notify)
>>>
Signed-off-by: Abdiel Janulgue
---
tests/gem_exec_schedule.c | 120 ++
1 file changed, 15 insertions(+), 105 deletions(-)
diff --git a/tests/gem_exec_schedule.c b/tests/gem_exec_schedule.c
index d8454d5..87d5a85 100644
From: "Navare, Manasi D"
This is the userspace component of the Displayport Compliance
testing software required for compliance testing of the I915
Display Port driver. This must be running in order to successfully
complete Display Port compliance testing. This app and
== Series Details ==
Series: drm/i915/perf: Improve handling of OA tail race
URL : https://patchwork.freedesktop.org/series/18448/
State : success
== Summary ==
Series 18448v1 drm/i915/perf: Improve handling of OA tail race
On Fri, Dec 30, 2016 at 3:02 PM, Rafael J. Wysocki wrote:
> On Fri, Dec 30, 2016 at 12:40 PM, Sedat Dilek wrote:
>> Hi,
>>
>> I have already reported this issue in [1].
>> One of the issue was solved.
>> Unfortunately, it looks like there is still a
There's no need for the driver to keep reading back the head pointer
from hardware since the hardware doesn't update it automatically. This
way we can treat any invalid head pointer value as a software/driver
bug instead of spurious hardware behaviour.
This change is also a small stepping stone
This avoids redundantly passing an (inout) head and tail pointer to
gen7_append_oa_reports() from gen7_oa_read which doesn't need to
reference either itself.
Moving the head/tail reads and writes into gen7_append_oa_reports should
have no functional effect except to avoid some redundant head
If the function for checking whether there is OA buffer data available
(during a poll or blocking read) has false positives then we want to
avoid a situation where the subsequent read() returns EAGAIN (after
a more accurate check) followed by a poll() immediately reporting
the same false positive
If I'm going to complain about a back-to-front convention then the least
I can do is not muddle the comment up too.
Signed-off-by: Robert Bragg
---
drivers/gpu/drm/i915/i915_perf.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git
There's a HW race condition between OA unit tail pointer register
updates and writes to memory whereby the tail pointer can sometimes get
ahead of what's been written out to the OA buffer so far (in terms of
what's visible to the CPU).
Although this can be observed explicitly while copying
While I was updating the i915 perf patches for gen8+ I found I also needed to
add a workaround for a tail pointer race, like we see for gen7 too. I was
feeling guilty knowing there are a few limitations with the current workaround
for gen7 and so want to address those before reusing the WA for
There were a couple of problems with both of these tests that could lead
to false negatives addressed by this patch.
1) The upper limit for the number of iterations missed a +1 to consider
that there might be a sample immediately available at the start of the
loop.
v2) The tests didn't
On Mon, Jan 23, 2017 at 12:50:07PM -0800, Manasi Navare wrote:
> Thanks Jani for reviewing this patch and for your feedback.
> Its very useful feedback and below are some of my comments:
>
> On Fri, Jan 20, 2017 at 05:05:03PM +0200, Jani Nikula wrote:
> > On Fri, 20 Jan 2017, Manasi Navare
Thanks for this patch. I have tested this on DPR 120 compliance test device
and verified that it works as expected.
Manasi
On Fri, Jan 20, 2017 at 07:04:06PM +0200, Jani Nikula wrote:
> There is no point in setting intel_dp->compliance.test_type, and
> proceeding with the autotests, if we're
== Series Details ==
Series: drm/i915: Move atomic state free from out of fence release
URL : https://patchwork.freedesktop.org/series/18431/
State : success
== Summary ==
Series 18431v1 drm/i915: Move atomic state free from out of fence release
Fences are required to support being released from under an atomic context.
The drm_atomic_state struct may take a mutex when being released and so
we cannot drop a reference to the drm_atomic_state from the fence release
path directly, and so we need to defer that unreference to a worker.
[
Hi All,
Here is v2 of my cht i2c-pmic and i915-punit access coordination
patchset. New in this version are some minor spelling / style fixes
and more importantly the inclusion of 6 extra i2c-designware patches.
These 6 patches are ready for merging, they have Wolfram's (the i2c
maintainer's)
On my cherrytrail tablet with axp288 pmic, just doing a bunch of repeated
reads from the pmic, e.g. "i2cdump -y 14 0x34" would lookup the tablet in
1 - 3 runs guaranteed.
This seems to be causes by the cpu trying to enter C6 or C7 while we hold
the punit bus semaphore, at which point everything
One some systems the P-Unit accesses the PMIC to change various voltages
through the same bus as other kernel drivers use for e.g. battery
monitoring.
If a driver sends requests to the P-Unit which require the P-Unit to access
the PMIC bus while another driver is also accessing the PMIC bus
If (!shared_host) simply return 0, this avoids delaying the probe if
iosf_mbi_available() returns false when an i2c bus is not using the
punit semaphore.
Also move the if (!iosf_mbi_available()) check to above the
dev_info, so that we do not repeat the dev_info on every probe
until
Rename intel_uncore_early_sanitize to intel_uncore_resume, dropping the
(always true) restore_forcewake argument and add a new intel_uncore_resume
function to replace the intel_uncore_forcewake_reset(dev_priv, false)
calls done from the suspend / runtime_suspend functions and make
Make sure the P-Unit or the PMIC i2c bus is not in use when we send a
request to the P-Unit by calling iosf_mbi_punit_acquire() / _release()
around P-Unit write accesses.
BugLink: https://bugzilla.kernel.org/show_bug.cgi?id=155241
Signed-off-by: Hans de Goede
Tested-by:
Listen for PMIC bus access notifications and get FORCEWAKE_ALL while
the bus is accessed to avoid needing to do any forcewakes, which need
PMIC bus access, while the PMIC bus is busy:
This fixes errors like these showing up in dmesg, usually followed
by a gfx or system freeze:
Use iosf_mbi_modify instead of iosf_mbi_read + iosf_mbi_write so that
we keep the iosf_mbi_lock locked during the read-modify-write done to
reset the semaphore.
Signed-off-by: Hans de Goede
Reviewed-by: Andy Shevchenko
Acked-by: Jarkko
The cherrytrail punit has the pmic i2c bus access semaphore at a
different register address.
Signed-off-by: Hans de Goede
Reviewed-by: Takashi Iwai
Tested-by: Takashi Iwai
Reviewed-by: Andy Shevchenko
Call the iosf_mbi pmic_bus_access_notifier_chain on bus acquire / release.
BugLink: https://bugzilla.kernel.org/show_bug.cgi?id=155241
Signed-off-by: Hans de Goede
Tested-by: tagorereddy
Reviewed-by: Andy Shevchenko
Acquire P-Unit access to stop others from accessing the P-Unit while the
PMIC i2c bus is in use. This is necessary because accessing the P-Unit
from the kernel may result in the P-Unit trying to access the PMIC i2c
bus, which results in a hang when it happens while we own the PMIC i2c
bus
Pass dw_i2c_dev into the helper functions, this is a preparation patch
for the punit semaphore fixes done in the other patches in this set.
Signed-off-by: Hans de Goede
Reviewed-by: Takashi Iwai
Tested-by: Takashi Iwai
Reviewed-by: Andy
Some drivers may need to acquire P-Unit managed resources from interrupt
context, where they cannot call iosf_mbi_punit_acquire().
This commit adds a notifier chain which allows a driver to get notified
(in a process context) before other drivers start accessing the PMIC bus,
so that the driver
Rename accessor_flags to flags, so that we can use the field for
other flags too. This is a preparation patch for adding cherrytrail
support to the punit semaphore code.
Signed-off-by: Hans de Goede
Reviewed-by: Andy Shevchenko
Acked-by:
Ander, can I carry your rv-b on this v2?
Jani, with Ander confirmation can I merge these 2 patches?
On Mon, Jan 23, 2017 at 11:24 AM, Patchwork
wrote:
> == Series Details ==
>
> Series: series starting with [1/2] drm/i915/kbl: Apply
>
Thanks Jani for reviewing this patch and for your feedback.
Its very useful feedback and below are some of my comments:
On Fri, Jan 20, 2017 at 05:05:03PM +0200, Jani Nikula wrote:
> On Fri, 20 Jan 2017, Manasi Navare wrote:
> > This patch adds support to handle
== Series Details ==
Series: series starting with [1/2] drm/i915/kbl: Apply
WaIncreaseDefaultTLBEntries on KBL.
URL : https://patchwork.freedesktop.org/series/18421/
State : success
== Summary ==
Series 18421v1 Series without cover letter
On Mon, Jan 23, 2017 at 09:05:02AM -0800, Manasi Navare wrote:
> On Sat, Jan 21, 2017 at 05:16:42PM +0200, Jani Nikula wrote:
> > On Sat, 21 Jan 2017, Manasi Navare wrote:
> > > This patch series addresses all the review comments from the previous
> > > series:
> > >
Along with GLK it was introduced the .is_lp and IS_GEN9_LP.
So, following the same simplification standard we can
put Skylake and Kabylake under the same bucket for most
of the things.
So let's add the IS_GEN9_BC for "Big Core" (non Atom based
platforms).
The i915_drv.c was let out of this patch
According to wa_database this Wa persist on KBL as it was on SKL.
Cc: Tim Gore
Cc: Mika Kuoppala
Cc: Tvrtko Ursulin
Signed-off-by: Rodrigo Vivi
Reviewed-by: Ander Conselvan de Oliveira
On Sat, Jan 21, 2017 at 05:16:42PM +0200, Jani Nikula wrote:
> On Sat, 21 Jan 2017, Manasi Navare wrote:
> > This patch series addresses all the review comments from the previous
> > series:
> > https://patchwork.freedesktop.org/series/18256/
>
> It does not, and it
On Fri, Jan 13, 2017 at 09:04:12AM +0200, Jani Nikula wrote:
> On Fri, 13 Jan 2017, Rodrigo Vivi wrote:
> > Most of commiters already knows that by heart and also dim already
> > have this warning. So maybe this is just Bart writing to blackboard. Duh!
> >
> > Anyway for
#include
@@ -630,6 +631,10 @@ void intel_audio_codec_enable(struct intel_encoder
*intel_encoder,
if (acomp && acomp->audio_ops && acomp->audio_ops->pin_eld_notify)
acomp->audio_ops->pin_eld_notify(acomp->audio_ops->audio_ptr,
Hi,
On 23-01-17 11:36, Jani Nikula wrote:
On Fri, 20 Jan 2017, Hans de Goede wrote:
Hi,
On 31-12-16 17:00, Hans de Goede wrote:
Hi,
On 27-12-16 11:58, Jani Nikula wrote:
On Sun, 25 Dec 2016, Hans de Goede wrote:
If there is no OPREGION_ASLE_EXT
On Mon, Jan 23, 2017 at 03:01:20PM +, Matthew Auld wrote:
> On 23 January 2017 at 14:52, Chris Wilson wrote:
> > Since tweaking i915_vma_compare() we allowed constructors to skip
> > clearing the ggtt_view believing that we didn't access the unused
> > members. That,
On 2017-01-23 06:00 AM, Mika Kahola wrote:
With this patch applied, I received couple of issues during the
compilation
igt_kms.c:332:5: warning: no previous prototype for
‘kmstest_pipe_to_index’ [-Wmissing-prototypes]
int kmstest_pipe_to_index(char pipe)
and
kms_fence_pin_leak.c:118:41:
Op 23-01-17 om 14:26 schreef Mika Kahola:
> On Thu, 2017-01-12 at 12:37 +0100, Maarten Lankhorst wrote:
>> Hey,
>>
>> Op 12-01-17 om 11:28 schreef Mika Kahola:
>>> On Thu, 2017-01-12 at 11:11 +0100, Maarten Lankhorst wrote:
Op 30-12-16 om 13:00 schreef Mika Kahola:
> This test case adds
Chris Wilson writes:
> To complement the check in execlists_elsp_ready(), also assert that we
> don't submit the same context while it has a lite restore still pending.
>
> Signed-off-by: Chris Wilson
Reviewed-by: Mika Kuoppala
== Series Details ==
Series: drm/i915: Pevent copying uninitialised garbage into vma->ggtt_view
URL : https://patchwork.freedesktop.org/series/18414/
State : success
== Summary ==
Series 18414v1 drm/i915: Pevent copying uninitialised garbage into
vma->ggtt_view
Chris Wilson writes:
> Mark when we run the execlist tasklet following the interrupt, so we
> don't probe a potentially initialised register when submitting the
> contexts multiple times.
>
> Signed-off-by: Chris Wilson
> ---
>
On 23 January 2017 at 14:52, Chris Wilson wrote:
> Since tweaking i915_vma_compare() we allowed constructors to skip
> clearing the ggtt_view believing that we didn't access the unused
> members. That, as it turns out, was not entirely true. In particular,
>
Op 27-12-16 om 15:48 schreef Daniel Vetter:
> On Thu, Dec 22, 2016 at 11:33:23AM +0100, Maarten Lankhorst wrote:
>> This reverts commit 527b6abe5fd2d2
>> (Revert "drm/i915: Use atomic commits for legacy page_flips")
>>
>> and reapplies commit ee042aa40b66d1.
>> ("drm/i915: Use atomic commits for
Since tweaking i915_vma_compare() we allowed constructors to skip
clearing the ggtt_view believing that we didn't access the unused
members. That, as it turns out, was not entirely true. In particular,
i915_gem_fault() uses
ret = remap_io_mapping(area,
area->vm_start +
On Mon, Jan 23, 2017 at 02:39:03PM +, Chris Wilson wrote:
> On Mon, Jan 23, 2017 at 02:22:38PM +, Matthew Auld wrote:
> > On 23 January 2017 at 14:02, Chris Wilson wrote:
> > > On Mon, Jan 23, 2017 at 01:51:54PM +, Matthew Auld wrote:
> > >> On 14 January
On Mon, Jan 23, 2017 at 02:22:38PM +, Matthew Auld wrote:
> On 23 January 2017 at 14:02, Chris Wilson wrote:
> > On Mon, Jan 23, 2017 at 01:51:54PM +, Matthew Auld wrote:
> >> On 14 January 2017 at 00:28, Chris Wilson wrote:
> >> > It
On Mon, Jan 23, 2017 at 01:52:02PM +, Chris Wilson wrote:
> On Mon, Jan 23, 2017 at 03:43:27PM +0200, Ville Syrjälä wrote:
> > On Mon, Jan 23, 2017 at 01:05:57PM +, Chris Wilson wrote:
> > > This w/a was only used for preproduction hw, which is no longer in use.
> > > Remove the workaround
== Series Details ==
Series: drm/i915: Assert that we don't submit to execlists whilst a preempt is
pending
URL : https://patchwork.freedesktop.org/series/18407/
State : warning
== Summary ==
Series 18407v1 drm/i915: Assert that we don't submit to execlists whilst a
preempt is pending
On 23 January 2017 at 14:02, Chris Wilson wrote:
> On Mon, Jan 23, 2017 at 01:51:54PM +, Matthew Auld wrote:
>> On 14 January 2017 at 00:28, Chris Wilson wrote:
>> > It is only being used to clear a struct and set the type, after which it
On Mon, Jan 23, 2017 at 01:51:54PM +, Matthew Auld wrote:
> On 14 January 2017 at 00:28, Chris Wilson wrote:
> > It is only being used to clear a struct and set the type, after which it
> > is overwritten. Since we no longer check the unset bits of the union,
> >
On Mon, Jan 23, 2017 at 01:51:54PM +, Matthew Auld wrote:
> On 14 January 2017 at 00:28, Chris Wilson wrote:
> > It is only being used to clear a struct and set the type, after which it
> > is overwritten. Since we no longer check the unset bits of the union,
> >
Chris Wilson writes:
> We need to prevent resubmission of the context immediately following an
> initial resubmit (which does a lite-restore preemption). Currently we do
> this by disabling all submission whilst the context is still active, but
> we can improve this by
== Series Details ==
Series: series starting with [1/6] drm/i915: Remove disable_lite_restore_wa
URL : https://patchwork.freedesktop.org/series/18406/
State : success
== Summary ==
Series 18406v1 Series without cover letter
On 14 January 2017 at 00:28, Chris Wilson wrote:
> It is only being used to clear a struct and set the type, after which it
> is overwritten. Since we no longer check the unset bits of the union,
> skipping the clear is permissible.
>
> Signed-off-by: Chris Wilson
On Mon, Jan 23, 2017 at 03:43:27PM +0200, Ville Syrjälä wrote:
> On Mon, Jan 23, 2017 at 01:05:57PM +, Chris Wilson wrote:
> > This w/a was only used for preproduction hw, which is no longer in use.
> > Remove the workaround to simplify the code.
> >
> > Signed-off-by: Chris Wilson
To complement the check in execlists_elsp_ready(), also assert that we
don't submit the same context while it has a lite restore still pending.
Signed-off-by: Chris Wilson
---
drivers/gpu/drm/i915/intel_lrc.c | 1 +
1 file changed, 1 insertion(+)
diff --git
On Mon, Jan 23, 2017 at 01:05:57PM +, Chris Wilson wrote:
> This w/a was only used for preproduction hw, which is no longer in use.
> Remove the workaround to simplify the code.
>
> Signed-off-by: Chris Wilson
> Cc: Mika Kuoppala
>
On ma, 2017-01-23 at 13:05 +, Chris Wilson wrote:
> This w/a (WaSetDisablePixMaskCammingAndRhwoInCommonSliceChicken) was only
> used for preproduction hw, which is no longer in use. Remove the
> workaround to simplify the code.
>
> Signed-off-by: Chris Wilson
> Cc:
On ma, 2017-01-23 at 13:05 +, Chris Wilson wrote:
> This w/a was only used for preproduction hw, which is no longer in use.
> Remove the workaround to simplify the code.
>
> Signed-off-by: Chris Wilson
> Cc: Mika Kuoppala
> Cc: Joonas
On Mon, Jan 23, 2017 at 12:14:26PM +, Chris Wilson wrote:
> Mark when we run the execlist tasklet following the interrupt, so we
> don't probe a potentially initialised register when submitting the
s/initialised/uninitialised/!
-Chris
--
Chris Wilson, Intel Open Source Technology Centre
On Thu, 2017-01-12 at 12:37 +0100, Maarten Lankhorst wrote:
> Hey,
>
> Op 12-01-17 om 11:28 schreef Mika Kahola:
> >
> > On Thu, 2017-01-12 at 11:11 +0100, Maarten Lankhorst wrote:
> > >
> > > Op 30-12-16 om 13:00 schreef Mika Kahola:
> > > >
> > > > This test case adds TEST_ONLY flag to the
== Series Details ==
Series: series starting with [v2,1/5] drm/i915: Only disable execlist
preemption for the duration of the request
URL : https://patchwork.freedesktop.org/series/18403/
State : success
== Summary ==
Series 18403v1 Series without cover letter
On ma, 2017-01-23 at 13:06 +, Chris Wilson wrote:
> This w/a (WaClearTdlStateAckDirtyBits) was only used for preproduction hw,
> which is no longer in use. Remove the workaround to simplify the code.
>
> Signed-off-by: Chris Wilson
> Cc: Mika Kuoppala
On Mon, Jan 23, 2017 at 01:05:56PM +, Chris Wilson wrote:
> This w/a was only used for preproduction hw, which is no longer in use.
> Remove the workaround to simplify the code.
Note to self, remember to include WaEnableForceRestoreInCtxtDescForVCS
-Chris
--
Chris Wilson, Intel Open Source
On ma, 2017-01-23 at 13:05 +, Chris Wilson wrote:
> This w/a (WaDisableCtxRestoreArbitration) was only used for preproduction
> hw, which is no longer in use. Remove the workaround to simplify the code.
>
> Signed-off-by: Chris Wilson
> Cc: Mika Kuoppala
On ma, 2017-01-23 at 13:05 +, Chris Wilson wrote:
> This w/a was only used for preproduction hw, which is no longer in use.
> Remove the workaround to simplify the code.
>
> Signed-off-by: Chris Wilson
> Cc: Mika Kuoppala
> Cc: Joonas
This w/a (WaClearTdlStateAckDirtyBits) was only used for preproduction hw,
which is no longer in use. Remove the workaround to simplify the code.
Signed-off-by: Chris Wilson
Cc: Mika Kuoppala
Cc: Joonas Lahtinen
This w/a (WaSetDisablePixMaskCammingAndRhwoInCommonSliceChicken) was only
used for preproduction hw, which is no longer in use. Remove the
workaround to simplify the code.
Signed-off-by: Chris Wilson
Cc: Mika Kuoppala
Cc: Joonas Lahtinen
This w/a (WaDisableCtxRestoreArbitration) was only used for preproduction
hw, which is no longer in use. Remove the workaround to simplify the code.
Signed-off-by: Chris Wilson
Cc: Mika Kuoppala
Cc: Joonas Lahtinen
This w/a was only used for preproduction hw, which is no longer in use.
Remove the workaround to simplify the code.
Signed-off-by: Chris Wilson
Cc: Mika Kuoppala
Cc: Joonas Lahtinen
Cc: Tvrtko Ursulin
Stop creating and emitting the wa_ctx per-bb as it is no longer used.
Signed-off-by: Chris Wilson
---
drivers/gpu/drm/i915/intel_lrc.c | 37 +
1 file changed, 13 insertions(+), 24 deletions(-)
diff --git
This w/a was only used for preproduction hw, which is no longer in use.
Remove the workaround to simplify the code.
Signed-off-by: Chris Wilson
Cc: Mika Kuoppala
Cc: Joonas Lahtinen
Cc: Tvrtko Ursulin
On Mon, Jan 23, 2017 at 02:33:20PM +0200, Mika Kuoppala wrote:
> Chris Wilson writes:
>
> > On Mon, Jan 23, 2017 at 12:14:25PM +, Chris Wilson wrote:
> >> @@ -558,7 +558,8 @@ static bool execlists_elsp_ready(struct
> >> intel_engine_cs *engine)
> >>int port;
>
Chris Wilson writes:
> On Mon, Jan 23, 2017 at 12:14:25PM +, Chris Wilson wrote:
>> @@ -558,7 +558,8 @@ static bool execlists_elsp_ready(struct intel_engine_cs
>> *engine)
>> int port;
>>
>> port = 1; /* wait for a free slot */
>> -if
On Mon, Jan 23, 2017 at 12:22:52PM -, Patchwork wrote:
> == Series Details ==
>
> Series: series starting with [CI,1/2] drm/i915: Assert that the
> kernel_context is hw-id 0
> URL : https://patchwork.freedesktop.org/series/18402/
> State : warning
>
> == Summary ==
>
> Series 18402v1
On Mon, Jan 23, 2017 at 12:14:27PM +, Chris Wilson wrote:
> If the CSB head/tail pointers are unchanged, we can skip the update of
> the CSB register afterwards.
>
> Signed-off-by: Chris Wilson
Reviewed-by: Joonas Lahtinen
--
On Mon, Jan 23, 2017 at 12:14:25PM +, Chris Wilson wrote:
> @@ -558,7 +558,8 @@ static bool execlists_elsp_ready(struct intel_engine_cs
> *engine)
> int port;
>
> port = 1; /* wait for a free slot */
> - if (engine->disable_lite_restore_wa || engine->preempt_wa)
> + if
== Series Details ==
Series: series starting with [CI,1/2] drm/i915: Assert that the kernel_context
is hw-id 0
URL : https://patchwork.freedesktop.org/series/18402/
State : warning
== Summary ==
Series 18402v1 Series without cover letter
If the CSB head/tail pointers are unchanged, we can skip the update of
the CSB register afterwards.
Signed-off-by: Chris Wilson
---
drivers/gpu/drm/i915/intel_lrc.c | 7 +--
1 file changed, 5 insertions(+), 2 deletions(-)
diff --git
When introduced, I thought that reducing client latency from the
signaler was the priority. Since its inception the signaler has become
responsible for keeping the execlists full, via the dma-fence. As this
is very important to minimise overall execution time, signal the
dma-fence first and then
If the second ELSP port is available, schedule the execlists tasklet to
see if the new request can occupy it.
Signed-off-by: Chris Wilson
Cc: Tvrtko Ursulin
Reviewed-by: Tvrtko Ursulin
---
Mark when we run the execlist tasklet following the interrupt, so we
don't probe a potentially initialised register when submitting the
contexts multiple times.
Signed-off-by: Chris Wilson
---
drivers/gpu/drm/i915/i915_irq.c | 7 +--
We need to prevent resubmission of the context immediately following an
initial resubmit (which does a lite-restore preemption). Currently we do
this by disabling all submission whilst the context is still active, but
we can improve this by limiting the restriction to only until we
receive
On ma, 2017-01-23 at 09:57 +, Chris Wilson wrote:
> On Mon, Jan 23, 2017 at 11:41:12AM +0200, Joonas Lahtinen wrote:
> >
> > On la, 2017-01-21 at 14:50 +, Chris Wilson wrote:
> > >
> > > In order to reset the GPU early on in the module load sequence, we need
> > > to allocate the basic
On Mon, Jan 23, 2017 at 11:50:47AM +, Tvrtko Ursulin wrote:
>
> On 23/01/2017 11:16, Chris Wilson wrote:
> >On Mon, Jan 23, 2017 at 10:51:28AM +, Tvrtko Ursulin wrote:
> >>
> >>On 21/01/2017 09:25, Chris Wilson wrote:
> >>>In order to maximise concurrency between engines, if we queue a
On 23/01/2017 11:31, Chris Wilson wrote:
When execlists signals the context completion, it also provides the
context id for the status event. Assert that id matches the one we expect.
v2: The upper dword of the context status is a duplicate of the upper
dword from elsp submission (i.e.
On 23/01/2017 11:50, Chris Wilson wrote:
On Mon, Jan 23, 2017 at 11:41:03AM +, Tvrtko Ursulin wrote:
On 23/01/2017 10:51, Chris Wilson wrote:
On Mon, Jan 23, 2017 at 10:43:10AM +, Tvrtko Ursulin wrote:
@@ -3285,6 +3291,7 @@ int i915_gem_object_set_cache_level(struct
== Series Details ==
Series: drm/color: un-inline drm_color_lut_extract()
URL : https://patchwork.freedesktop.org/series/18391/
State : failure
== Summary ==
Series 18391v1 drm/color: un-inline drm_color_lut_extract()
https://patchwork.freedesktop.org/api/1.0/series/18391/revisions/1/mbox/
On 23/01/2017 11:16, Chris Wilson wrote:
On Mon, Jan 23, 2017 at 10:51:28AM +, Tvrtko Ursulin wrote:
On 21/01/2017 09:25, Chris Wilson wrote:
In order to maximise concurrency between engines, if we queue a request
to a current idle ring, reorder its dependencies to execute that request
On Mon, Jan 23, 2017 at 11:41:03AM +, Tvrtko Ursulin wrote:
>
> On 23/01/2017 10:51, Chris Wilson wrote:
> >On Mon, Jan 23, 2017 at 10:43:10AM +, Tvrtko Ursulin wrote:
> >>>@@ -3285,6 +3291,7 @@ int i915_gem_object_set_cache_level(struct
> >>>drm_i915_gem_object *obj,
> >>> ret
On ma, 2017-01-23 at 11:31 +, Chris Wilson wrote:
> When execlists signals the context completion, it also provides the
> context id for the status event. Assert that id matches the one we expect.
>
> v2: The upper dword of the context status is a duplicate of the upper
> dword from elsp
On Mon, Jan 23, 2017 at 01:33:32PM +0200, Joonas Lahtinen wrote:
> On la, 2017-01-21 at 09:28 +, Chris Wilson wrote:
> > If the CSB head/tail pointers are unchanged, we can skip the update of
> > the CSB register afterwards.
> >
> > Signed-off-by: Chris Wilson
>
>
On 23/01/2017 10:51, Chris Wilson wrote:
On Mon, Jan 23, 2017 at 10:43:10AM +, Tvrtko Ursulin wrote:
@@ -3285,6 +3291,7 @@ int i915_gem_object_set_cache_level(struct
drm_i915_gem_object *obj,
ret = i915_gem_object_wait(obj,
On la, 2017-01-21 at 09:28 +, Chris Wilson wrote:
> When introduced, I thought that reducing client latency from the
> signaler was the priority. Since its inception the signaler has become
> responsible for keeping the execlists full, via the dma-fence. As this
> is very important to minimise
1 - 100 of 162 matches
Mail list logo