On Tue, Oct 25, 2016 at 02:40:00PM +0530, meghanelogal wrote:
> Existing DDB algorithm divide the DDB wrt data rate,
> hence the planes with the less height but same width
> will be allocated less blocks and watermark are based
> on width which requires more DDB. With this data the flip
> may fail.
On Mon, Oct 24, 2016 at 09:18:36PM -0700, Dhinakaran Pandiyan wrote:
> Enabling DP audio stall fix is necessary to play audio over DP HBR2. So,
> let's set this bit right before enabling the audio codec. Playing audio
> without setting this bit results in pipe FIFO underruns.
Please insert a full
On Tue, Oct 25, 2016 at 06:19:59PM +, Pandiyan, Dhinakaran wrote:
> On Tue, 2016-10-25 at 12:14 +0300, Jani Nikula wrote:
> > On Tue, 25 Oct 2016, Jani Nikula wrote:
> > > On Tue, 25 Oct 2016, Dhinakaran Pandiyan
> > > wrote:
> > >> According to BSpec, cdclk has to be not less than 432 MHz w
On Tue, Oct 25, 2016 at 03:02:33PM +, Paneri, Praveen wrote:
> > So when you say that all Y tiling tests fail without this kernel hack,
> > which tests you are referring to?
> If I revert this IGT patch and do not make below kernel change, kms_draw_crc
> (ytiled cases, last patch in this seri
On Tue, Oct 25, 2016 at 09:08:37AM +0100, Tvrtko Ursulin wrote:
>
> On 24/10/2016 17:55, Praveen Paneri wrote:
> > This patch adds Y-tiling support for igt_draw_rect function.
> >
> > Change-Id: I139e9773b7df286febe9ffa3dce358df079dac14
>
> You can remove (and should) remove Gerrit tags when sen
On Tue, Oct 25, 2016 at 05:02:15PM -0200, Paulo Zanoni wrote:
> Em Seg, 2016-10-24 às 19:13 +0300, ville.syrj...@linux.intel.com
> escreveu:
> > From: Ville Syrjälä
> >
> > Pass the framebuffer size in .16 fixed point coordinates to
> > drm_rect_rotate() since that's what the source coordinates a
On Mon, Oct 24, 2016 at 11:08:31AM +0100, Chris Wilson wrote:
> As we allow userspace to set the dotclock, we should try to respect it!
> Userpsace will try to set its frametimings based upon the dotclock, so
> ideally it should match the measured vblank interval.
>
> Signed-off-by: Chris Wilson
On Mon, Oct 24, 2016 at 10:38:34AM +0100, Chris Wilson wrote:
> On Mon, Oct 24, 2016 at 11:14:31AM +0200, Daniel Vetter wrote:
> > On Mon, Oct 24, 2016 at 09:54:52AM +0100, Chris Wilson wrote:
> > Also with this patch we should be able to throw out the hacks for tv-out.
> > I only added those becau
On Mon, Oct 24, 2016 at 03:43:02PM +0530, Kamble, Sagar A wrote:
> Hi Daniel, Chris,
>
>
> Sorry for not clarifying the purpose and if any misunderstanding.
> This was to allow running at fixed frequency.
>
> Kindly clarify if it is intentional to have _store routine without S_IWUSR.
>
> On 10/
On Tue, Oct 25, 2016 at 06:16:34PM -0700, Manasi Navare wrote:
> A new optional connector property is added for keeping
> track of whether the link is good (link training passed) or
> link is bad (link training failed). If the link status property
> is Bad, then userspace should fire off a new mod
== Series Details ==
Series: drm/i915/dp: Debug log MST active links explicitly (rev3)
URL : https://patchwork.freedesktop.org/series/13627/
State : failure
== Summary ==
Series 13627v3 drm/i915/dp: Debug log MST active links explicitly
https://patchwork.freedesktop.org/api/1.0/series/13627/re
Jim,
Please let me know if the R-B is still good.
-DK
On Tue, 2016-10-25 at 21:37 -0700, Dhinakaran Pandiyan wrote:
> From: "Pandiyan, Dhinakaran"
>
> No functional change. Just printing the number of active links without
> stating what the number means is not very useful. So, add relevant text
From: "Pandiyan, Dhinakaran"
No functional change. Just printing the number of active links without
stating what the number means is not very useful. So, add relevant text.
v2: Included connector info (Chris)
Signed-off-by: Dhinakaran Pandiyan
Reviewed-by: Jim Bride
---
drivers/gpu/drm/i915/i
Existing DDB algorithm divide the DDB wrt data rate,
hence the planes with the less height but same width
will be allocated less blocks and watermark are based
on width which requires more DDB. With this data the flip
may fail.
In new DDB algorithm, the DDB is divided based on
watermark requiremen
== Series Details ==
Series: series starting with [1/2] drm: Add a new connector property for link
status
URL : https://patchwork.freedesktop.org/series/14374/
State : failure
== Summary ==
Series 14374v1 Series without cover letter
https://patchwork.freedesktop.org/api/1.0/series/14374/revis
Chris,
Would you be able to make the necessary changes in the suerspace
driver so I can do some testing tomorrow?
Manasi
On Tue, Oct 25, 2016 at 06:16:34PM -0700, Manasi Navare wrote:
> A new optional connector property is added for keeping
> track of whether the link is good (link training pass
The link status connector property is attached to the drm
object in DP initialization.
This also defines a helper function to set the property value.
This will be used to set the link sttaus to Bad in case
of link training failures.
Cc: dri-de...@lists.freedesktop.org
Cc: Jani Nikula
Cc: Daniel V
A new optional connector property is added for keeping
track of whether the link is good (link training passed) or
link is bad (link training failed). If the link status property
is Bad, then userspace should fire off a new modeset at the current
mode even if there have not been any changes in the
http://kisskb.ellerman.id.au/kisskb/buildresult/12840554/
Since the GVT stuff it looks like some divide should be a do_div.
Dave.
___
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
On Mon, 2016-10-03 at 11:42 -0700, Anusha Srivatsa wrote:
> From: Peter Antoine
>
> The HuC authentication is done by host2guc call. The HuC RSA keys
> are sent to GuC for authentication.
>
> v2: rebased on top of drm-intel-nightly.
> changed name format and upped version 1.7.
> v3: rebased
Mixing up git rebase and reset was not a good idea, will send the
corrected patch. Please ignore this.
On Tue, 2016-10-25 at 16:42 -0700, Dhinakaran Pandiyan wrote:
> Enabling DP audio stall fix is necessary to play audio over DP HBR2. So,
> let's set this bit right before enabling the audio codec
== Series Details ==
Series: DP audio fixes (rev4)
URL : https://patchwork.freedesktop.org/series/14314/
State : success
== Summary ==
Series 14314v4 DP audio fixes
https://patchwork.freedesktop.org/api/1.0/series/14314/revisions/4/mbox/
fi-bdw-5557u total:246 pass:231 dwarn:0 dfail:
On Tue, Oct 25, 2016 at 10:35 PM, Matthew Auld <
matthew.william.a...@gmail.com> wrote:
> On 25 October 2016 at 00:19, Robert Bragg wrote:
>
> > diff --git a/drivers/gpu/drm/i915/i915_drv.h
> b/drivers/gpu/drm/i915/i915_drv.h
> > index 3448d05..ea24814 100644
> > --- a/drivers/gpu/drm/i915/i91
Enabling DP audio stall fix is necessary to play audio over DP HBR2. So,
let's set this bit right before enabling the audio codec. Playing audio
without setting this bit results in pipe FIFO underruns.
This workaround is applicable only for audio sample rates up to 96kHz. For
frequencies above 96k
According to BSpec, cdclk has to be not less than 432 MHz with DP audio
enabled, port width x4, and link rate HBR2 (5.4 GHz)
Having a lower cdclk triggers pipe underruns, which then lead to displays
continuously cycling off and on. This is essential for DP MST audio as the
link is trained at HBR2
On Tue, 2016-10-25 at 18:19 +, Pandiyan, Dhinakaran wrote:
> On Tue, 2016-10-25 at 12:14 +0300, Jani Nikula wrote:
> > On Tue, 25 Oct 2016, Jani Nikula wrote:
> > > On Tue, 25 Oct 2016, Dhinakaran Pandiyan
> > > wrote:
> > >> According to BSpec, cdclk has to be not less than 432 MHz with DP
On Tue, Oct 25, 2016 at 12:19:29AM +0100, Robert Bragg wrote:
> +static int claim_specific_ctx(struct i915_perf_stream *stream)
> +{
> + struct drm_i915_private *dev_priv = stream->dev_priv;
> + struct i915_vma *vma;
> + int ret;
> +
> + ret = i915_mutex_lock_interruptible(&dev_priv
On Tue, Oct 25, 2016 at 5:09 AM, Jani Nikula
wrote:
> On Sat, 22 Oct 2016, Manasi Navare wrote:
>> This function provides a way for the driver to redo a
>> modeset on the current mode and retry the link training
>> at a lower link rate/lane count/bpp. This will get called
>> incase the link train
On Tue, Oct 25, 2016 at 03:09:39PM +0300, Jani Nikula wrote:
> On Sat, 22 Oct 2016, Manasi Navare wrote:
> > This function provides a way for the driver to redo a
> > modeset on the current mode and retry the link training
> > at a lower link rate/lane count/bpp. This will get called
> > incase th
Tested with HuC version 1.07.1398 on SKL.
Reviewed-by: Carlos Santa
Tested-by: Carlos Santa
On Mon, 2016-10-03 at 11:42 -0700, Anusha Srivatsa wrote:
> From: Peter Antoine
>
> Add debugfs entry for HuC loading status check.
>
> v2: rebase on-top of drm-intel-nightly.
> v3: rebased again.
> v
On 25 October 2016 at 00:19, Robert Bragg wrote:
> Being able to program OACONTROL from a non-privileged batch buffer is
> not sufficient to be able to configure the OA unit. This was originally
> allowed to help enable Mesa to expose OA counters via the
> INTEL_performance_query extension, but th
On 25 October 2016 at 00:19, Robert Bragg wrote:
> check_cmd() is checking whether a command adheres to certain
> restrictions that ensure it's safe to execute within a privileged batch
> buffer. Returning false implies a privilege problem, not that the
> command is invalid.
>
> The distinction ma
On Tue, Oct 25, 2016 at 05:27:21PM -0400, Sean Paul wrote:
> On Tue, Oct 25, 2016 at 3:46 PM, Chris Wilson
> wrote:
> > drm_property_lookup_blob() returns a reference to the returned blob, and
> > drm_atomic_replace_property_blob() takes a references to the blob it
> > stores, so afterwards we ar
On 25 October 2016 at 00:19, Robert Bragg wrote:
> Gen graphics hardware can be set up to periodically write snapshots of
> performance counters into a circular buffer via its Observation
> Architecture and this patch exposes that capability to userspace via the
> i915 perf interface.
>
> v2:
>
From: Felix Monninger
drm_property_lookup_blob() returns a reference to the returned blob, and
drm_atomic_replace_property_blob() takes a references to the blob it
stores, so afterwards we are left owning a reference to the new_blob that
we never release, and thus leak memory every time we update
On Tue, Oct 25, 2016 at 3:46 PM, Chris Wilson wrote:
> drm_property_lookup_blob() returns a reference to the returned blob, and
> drm_atomic_replace_property_blob() takes a references to the blob it
> stores, so afterwards we are left owning a reference to the new_blob that
> we never release, and
On Tue, Oct 25, 2016 at 11:44:36PM +0300, Marius Vlad wrote:
> +int
> +igt_pkill(int sig, const char *comm)
> +{
> + int err = 0;
> + PROCTAB *proc;
> + proc_t *proc_info;
> +
> + proc = openproc(PROC_FILLCOM | PROC_FILLSTAT | PROC_FILLARG);
> + igt_assert(proc != NULL);
> +
> +
On Tue, Oct 25, 2016 at 02:48:02PM +0200, Arkadiusz Hiler wrote:
> Comment mentioned use of intel_uncore_forcewake_irq{unlock, lock}
> functions which are nonexistent (and never were).
>
> The description was also incomplete and could cause confusion. Updated
> comment is more elaborate on usage a
v2:
- embedded gem_alive and gem_exec_store into test (Chris Wilson)
- int main() to igt_main (Chris Wilson)
- moved tests/gem_alive -> tools/gem_info (Chris Wilson)
- added to intel-ci/fast-feedback.testlist (Petri Latvala)
- added hda_dynamic_debug() (Petri Latvala)
- renamed from tests/drv_modul
v2:
- don't read cached values (Chris Wilson)
- warn on per connector, and fail per mean (Chris Wilson)
These are synthetic: 5us per connector, and 600us for all (as
threshold).
Signed-off-by: Marius Vlad
---
tests/Makefile.sources| 2 +-
tests/kms_sysfs_edid_timing | 25 --
This series adds some library support to help converting sh scripts to C
version. Based on that I've converted drv_module_reload_basic and
kms_sysfs_edid_timing. Other tests should follow. drv_module_reload requires
the most boilerplate code.
The reason for so many changes is the fact that some c
Previously under unbind_fbcon(), to disable/enable framebuffer console.
lib/igt_aux: Added helpers to help convert sh scripts to C version. libkmod
and procps interface.
v2:
- Renamed libkmod helpers (Chris Wilson)
- Removed SIGTERM/SIGKILL case where we repeatedly tried to terminate the
process:
== Series Details ==
Series: drm: Release reference from blob lookup after replacing property
URL : https://patchwork.freedesktop.org/series/14357/
State : success
== Summary ==
Series 14357v1 drm: Release reference from blob lookup after replacing property
https://patchwork.freedesktop.org/ap
On Tue, Oct 25, 2016 at 10:05:23PM +0530, akash.g...@intel.com wrote:
> From: Akash Goel
>
> Driver accesses the ringbuffer pages, via GMADR BAR, if the pages are
> pinned in mappable aperture portion of GGTT and for ringbuffer pages
> allocated from Stolen memory, access can only be done through
On Tue, Oct 25, 2016 at 05:40:35PM +0100, Tvrtko Ursulin wrote:
> From: Tvrtko Ursulin
>
> We do not need to set up a fence for the rotated view.
>
> Display does not need it and no one can access it.
>
> v2: Move code to __i915_vma_set_map_and_fenceable. (Chris Wilson)
>
> Signed-off-by: Tvrt
drm_property_lookup_blob() returns a reference to the returned blob, and
drm_atomic_replace_property_blob() takes a references to the blob it
stores, so afterwards we are left owning a reference to the new_blob that
we never release, and thus leak memory every time we update a property
such as duri
Em Seg, 2016-10-24 às 19:13 +0300, ville.syrj...@linux.intel.com
escreveu:
> From: Ville Syrjälä
>
> Pass the framebuffer size in .16 fixed point coordinates to
> drm_rect_rotate() since that's what the source coordinates are as
> well
> at this stage. We used to do this part of the computation i
On Tue, Oct 25, 2016 at 03:17:47PM +0300, Jani Nikula wrote:
> On Sat, 22 Oct 2016, Manasi Navare wrote:
> > If link training at a link rate optimal for a particular
> > mode fails during modeset's atomic commit phase, then we
> > let the modeset complete and then retry. We save the link rate
> >
On Mon, Oct 24, 2016 at 11:23:39PM -0700, Pandiyan, Dhinakaran wrote:
> On Fri, 2016-10-21 at 16:45 -0700, Manasi Navare wrote:
> > If link training at a link rate optimal for a particular
> > mode fails during modeset's atomic commit phase, then we
> > let the modeset complete and then retry. We s
On Tue, 2016-10-25 at 11:47 +0300, Jani Nikula wrote:
> On Tue, 25 Oct 2016, Dhinakaran Pandiyan
> wrote:
> > Enabling DP audio stall fix is necessary to play audio over DP HBR2. So,
> > let's set this bit right before enabling the audio codec. Playing audio
> > without setting this bit results i
On Tue, 2016-10-25 at 11:46 +0300, Ville Syrjälä wrote:
> On Mon, Oct 24, 2016 at 09:18:37PM -0700, Dhinakaran Pandiyan wrote:
> > According to BSpec, cdclk has to be not less than 432 MHz with DP audio
> > enabled, port width x4, and link rate HBR2 (5.4 GHz)
> >
> > Having a lower cdclk triggers
On Tue, 2016-10-25 at 12:14 +0300, Jani Nikula wrote:
> On Tue, 25 Oct 2016, Jani Nikula wrote:
> > On Tue, 25 Oct 2016, Dhinakaran Pandiyan
> > wrote:
> >> According to BSpec, cdclk has to be not less than 432 MHz with DP audio
> >> enabled, port width x4, and link rate HBR2 (5.4 GHz)
> >>
> >>
On Tue, Oct 25, 2016 at 03:17:47PM +0300, Jani Nikula wrote:
> On Sat, 22 Oct 2016, Manasi Navare wrote:
> > If link training at a link rate optimal for a particular
> > mode fails during modeset's atomic commit phase, then we
> > let the modeset complete and then retry. We save the link rate
> >
Reviewed-by: Rodrigo Vivi
On Mon, 2016-10-24 at 17:28 -0700, Anusha Srivatsa wrote:
> Currently, for display there is only one DMC image for KBL.
> Remove the stepping_info table for KBL and use the no_stepping_info
> array for loading the firmware.
>
> v2: Removed the block of code as pointed o
== Series Details ==
Series: series starting with [v2,1/2] drm/i915: Rotated view does not need a
fence (rev3)
URL : https://patchwork.freedesktop.org/series/14340/
State : failure
== Summary ==
Series 14340v3 Series without cover letter
https://patchwork.freedesktop.org/api/1.0/series/14340/
From: Tvrtko Ursulin
We do not need to set up a fence for the rotated view.
Display does not need it and no one can access it.
v2: Move code to __i915_vma_set_map_and_fenceable. (Chris Wilson)
Signed-off-by: Tvrtko Ursulin
Fixes: 05a20d098db1 ("drm/i915: Move map-and-fenceable tracking to the
From: Tvrtko Ursulin
Objects can have multiple VMAs used for display in which
case assertion that objects must not be pinned for display
more times than the current VMA is incorrect.
v2: Commit message update. (Chris Wilson)
Signed-off-by: Tvrtko Ursulin
Fixes: 058d88c4330f ("drm/i915: Track p
From: Akash Goel
Driver accesses the ringbuffer pages, via GMADR BAR, if the pages are
pinned in mappable aperture portion of GGTT and for ringbuffer pages
allocated from Stolen memory, access can only be done through GMADR BAR.
In case of GuC based submission, updates done in ringbuffer via GMAD
== Series Details ==
Series: series starting with [1/4] drm/i915: Don't try to initialize sprite
planes on pre-ilk
URL : https://patchwork.freedesktop.org/series/14347/
State : warning
== Summary ==
Series 14347v1 Series without cover letter
https://patchwork.freedesktop.org/api/1.0/series/14
> +
> +/* Note we copy the properties from userspace outside of the i915 perf
> + * mutex to avoid an awkward lockdep with mmap_sem.
> + *
> + * Note this function only validates properties in isolation it doesn't
> + * validate that the combination of properties makes sense or that all
> + * prope
From: Ville Syrjälä
Due to the plane->index not getting readjusted in drm_plane_cleanup(),
we can't continue initialization of some plane/crtc init fails.
Well, we sort of could I suppose if we left all initialized planes on
the list, but that would expose those planes to userspace as well.
But
From: Ville Syrjälä
We don't currently implement support for sprite planes on pre-ilk
platforms, so let's leave num_sprites at 0 so that we don't get
spurious errors during driver init.
Signed-off-by: Ville Syrjälä
---
drivers/gpu/drm/i915/intel_device_info.c | 5 +++--
drivers/gpu/drm/i915/in
From: Ville Syrjälä
Kill the switch statement from the sprite init code and replace with a
more straightforward if ladder. Now each significant evolution of the
sprite hardware is in its own neat box.
Signed-off-by: Ville Syrjälä
---
drivers/gpu/drm/i915/intel_sprite.c | 70 ---
From: Ville Syrjälä
The zpos magic sorting uses the object ID to solve conflicting zpos
values. Let's initialize our planes in an order that makes the object
IDs agree with the normal primary->sprites->cursor z order.
Signed-off-by: Ville Syrjälä
---
drivers/gpu/drm/i915/intel_display.c | 16 +
On 25 October 2016 at 13:48, Arkadiusz Hiler wrote:
> Comment mentioned use of intel_uncore_forcewake_irq{unlock, lock}
> functions which are nonexistent (and never were).
>
> The description was also incomplete and could cause confusion. Updated
> comment is more elaborate on usage and caveats.
>
== Series Details ==
Series: series starting with [1/2] drm/i915/audio: drop extra crtc clock check
from HDMI audio N lookup
URL : https://patchwork.freedesktop.org/series/14342/
State : warning
== Summary ==
Series 14342v1 Series without cover letter
https://patchwork.freedesktop.org/api/1.0
On Tue, Oct 25, 2016 at 05:54:16PM +0300, Jani Nikula wrote:
> Remaining patch from https://patchwork.freedesktop.org/series/12754/
> plus a cleanup.
I'm not feeling like checking bspec for the nuts and bolts details on
the registers again, but at least the code looks sane to me and the
M/N values
== Series Details ==
Series: series starting with [1/2] drm/i915: Rotated view does not need a fence
URL : https://patchwork.freedesktop.org/series/14340/
State : warning
== Summary ==
Series 14340v1 Series without cover letter
https://patchwork.freedesktop.org/api/1.0/series/14340/revisions/1
On Mon, 24 Oct 2016, Imre Deak wrote:
> On my APL the LSPCON firmware resumes in PCON mode as opposed to the
> expected LS mode. It also appears to be in a state where AUX DPCD reads
> will succeed but return garbage recovering only after a few hundreds of
> milliseconds. After the recovery time D
On Tue, 25 Oct 2016, Imre Deak wrote:
> Extend the branch/sink descriptor info with the missing device ID
> field. While at it also read out all the descriptor registers in one
> transfer and make the debug print more compact.
>
> v2: (Jani)
> - Cache the descriptor in intel_dp.
> - Split out this
> So when you say that all Y tiling tests fail without this kernel hack, which
> tests you are referring to?
If I revert this IGT patch and do not make below kernel change, kms_draw_crc
(ytiled cases, last patch in this series) fail with following error.
Test assertion failure function igt_asser
Remaining patch from https://patchwork.freedesktop.org/series/12754/
plus a cleanup.
BR,
Jani.
___
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
The array contains the crtc clock, rely on that. While at it, debug log
the HDMI N value or automatic mode.
Cc: Ville Syrjälä
Cc: "Lin, Mengdong"
Cc: Libin Yang
Signed-off-by: Jani Nikula
---
drivers/gpu/drm/i915/intel_audio.c | 19 +--
1 file changed, 9 insertions(+), 10 dele
From: Libin Yang
When modeset occurs and the LS_CLK is set to some special values in DP
mode, the N/M need to be set manually if audio is playing. Otherwise the
first several seconds may be silent in audio playback.
The relationship of Maud and Naud is expressed in the following
equation:
Maud/
On Tue, Oct 25, 2016 at 04:27:26PM +0200, Saarinen, Jani wrote:
> > == Series Details ==
> >
> > Series: drm/i915: fix comment on I915_{READ,WRITE}_FW (rev2)
> > URL : https://patchwork.freedesktop.org/series/14334/
> > State : failure
> >
> > == Summary ==
> >
> > Series 14334v2 drm/i915: fix
== Series Details ==
Series: drm/i915/lspcon: Work around resume failure (rev2)
URL : https://patchwork.freedesktop.org/series/14280/
State : warning
== Summary ==
Series 14280v2 drm/i915/lspcon: Work around resume failure
https://patchwork.freedesktop.org/api/1.0/series/14280/revisions/2/mbox
On Tue, Oct 25, 2016 at 03:12:51PM +0100, Tvrtko Ursulin wrote:
>
> On 25/10/2016 15:09, Chris Wilson wrote:
> >On Tue, Oct 25, 2016 at 03:00:24PM +0100, Tvrtko Ursulin wrote:
> >>From: Tvrtko Ursulin
> >>
> >>Objects can have multiple VMAs used for display in which
> >>case assertion that object
On 25/10/2016 15:12, Ville Syrjälä wrote:
On Tue, Oct 25, 2016 at 03:00:23PM +0100, Tvrtko Ursulin wrote:
From: Tvrtko Ursulin
We do not need to set up a fence for the rotated view.
Display does not need it and no one can access it.
Signed-off-by: Tvrtko Ursulin
Cc: Chris Wilson
Cc: Joona
On 25/10/2016 15:12, Tvrtko Ursulin wrote:
On 25/10/2016 15:09, Chris Wilson wrote:
On Tue, Oct 25, 2016 at 03:00:24PM +0100, Tvrtko Ursulin wrote:
From: Tvrtko Ursulin
Objects can have multiple VMAs used for display in which
case assertion that objects must not be pinned for display
more t
On 25/10/2016 15:09, Chris Wilson wrote:
On Tue, Oct 25, 2016 at 03:00:24PM +0100, Tvrtko Ursulin wrote:
From: Tvrtko Ursulin
Objects can have multiple VMAs used for display in which
case assertion that objects must not be pinned for display
more times than their single display related VMA is
> == Series Details ==
>
> Series: drm/i915: fix comment on I915_{READ,WRITE}_FW (rev2)
> URL : https://patchwork.freedesktop.org/series/14334/
> State : failure
>
> == Summary ==
>
> Series 14334v2 drm/i915: fix comment on I915_{READ,WRITE}_FW
> https://patchwork.freedesktop.org/api/1.0/serie
On Tue, Oct 25, 2016 at 03:17:09PM +0100, Chris Wilson wrote:
> On Tue, Oct 25, 2016 at 05:12:22PM +0300, Ville Syrjälä wrote:
> > On Tue, Oct 25, 2016 at 03:00:23PM +0100, Tvrtko Ursulin wrote:
> > > From: Tvrtko Ursulin
> > >
> > > We do not need to set up a fence for the rotated view.
> > >
>
On Tue, Oct 25, 2016 at 05:12:22PM +0300, Ville Syrjälä wrote:
> On Tue, Oct 25, 2016 at 03:00:23PM +0100, Tvrtko Ursulin wrote:
> > From: Tvrtko Ursulin
> >
> > We do not need to set up a fence for the rotated view.
> >
> > Display does not need it and no one can access it.
> >
> > Signed-off-
On Tue, Oct 25, 2016 at 03:00:23PM +0100, Tvrtko Ursulin wrote:
> From: Tvrtko Ursulin
>
> We do not need to set up a fence for the rotated view.
>
> Display does not need it and no one can access it.
>
> Signed-off-by: Tvrtko Ursulin
> Cc: Chris Wilson
> Cc: Joonas Lahtinen
Broken by 05a20
On Tue, Oct 25, 2016 at 03:00:24PM +0100, Tvrtko Ursulin wrote:
> From: Tvrtko Ursulin
>
> Objects can have multiple VMAs used for display in which
> case assertion that objects must not be pinned for display
> more times than their single display related VMA is incorrect.
But each of those vma
On Tue, Oct 25, 2016 at 03:00:23PM +0100, Tvrtko Ursulin wrote:
> From: Tvrtko Ursulin
>
> We do not need to set up a fence for the rotated view.
>
> Display does not need it and no one can access it.
See __i915_vma_set_map_and_fenceable. Stop the bit at source. And
include the above line as a
On 24/10/2016 17:13, ville.syrj...@linux.intel.com wrote:
From: Ville Syrjälä
Pass the framebuffer size in .16 fixed point coordinates to
drm_rect_rotate() since that's what the source coordinates are as well
at this stage. We used to do this part of the computation in integer
coordinates, but
From: Tvrtko Ursulin
We do not need to set up a fence for the rotated view.
Display does not need it and no one can access it.
Signed-off-by: Tvrtko Ursulin
Cc: Chris Wilson
Cc: Joonas Lahtinen
---
drivers/gpu/drm/i915/i915_gem_gtt.h | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
From: Tvrtko Ursulin
Objects can have multiple VMAs used for display in which
case assertion that objects must not be pinned for display
more times than their single display related VMA is incorrect.
Signed-off-by: Tvrtko Ursulin
Cc: Chris Wilson
Cc: Joonas Lahtinen
Cc: Ville Syrjälä
---
dr
== Series Details ==
Series: drm/i915: fix comment on I915_{READ,WRITE}_FW (rev2)
URL : https://patchwork.freedesktop.org/series/14334/
State : failure
== Summary ==
Series 14334v2 drm/i915: fix comment on I915_{READ,WRITE}_FW
https://patchwork.freedesktop.org/api/1.0/series/14334/revisions/2/
== Series Details ==
Series: drm/i915: Include the kernel uptime in the error state (rev2)
URL : https://patchwork.freedesktop.org/series/14320/
State : success
== Summary ==
Series 14320v2 drm/i915: Include the kernel uptime in the error state
https://patchwork.freedesktop.org/api/1.0/series/
Extend the branch/sink descriptor info with the missing device ID
field. While at it also read out all the descriptor registers in one
transfer and make the debug print more compact.
v2: (Jani)
- Cache the descriptor in intel_dp.
- Split out this change into a separate patch.
v3: (Jani)
- Fix retu
Comment mentioned use of intel_uncore_forcewake_irq{unlock, lock}
functions which are nonexistent (and never were).
The description was also incomplete and could cause confusion. Updated
comment is more elaborate on usage and caveats.
v2: mention __locked variant of intel_uncore_forcewake_{get,pu
== Series Details ==
Series: dma-buf: Rename struct fence to dma_fence (rev3)
URL : https://patchwork.freedesktop.org/series/14324/
State : warning
== Summary ==
Series 14324v3 dma-buf: Rename struct fence to dma_fence
https://patchwork.freedesktop.org/api/1.0/series/14324/revisions/3/mbox/
T
On 25/10/2016 13:06, Paneri, Praveen wrote:
Hi Tvrtko,
Along with this change I made following change in the kernel side. I was not
sure if this is a hack of a legitimate change. Could you please give me a
pointer about how to move fwd from here? Without this all Y-tiling tests would
fail.
On Tue, Oct 25, 2016 at 02:15:23PM +0200, Arkadiusz Hiler wrote:
> Comment mentioned use of intel_uncore_forcewake_irq{unlock, lock}
> functions which are nonexistent (and never were).
>
> The description was also incomplete and could cause confusion. Updated
> comment is more elaborate on usage a
On Sat, 22 Oct 2016, Manasi Navare wrote:
> If link training at a link rate optimal for a particular
> mode fails during modeset's atomic commit phase, then we
> let the modeset complete and then retry. We save the link rate
> value at which link training failed and use a lower link rate
> to prun
== Series Details ==
Series: dma-buf: Rename struct fence to dma_fence (rev2)
URL : https://patchwork.freedesktop.org/series/14324/
State : success
== Summary ==
Series 14324v2 dma-buf: Rename struct fence to dma_fence
https://patchwork.freedesktop.org/api/1.0/series/14324/revisions/2/mbox/
T
Comment mentioned use of intel_uncore_forcewake_irq{unlock, lock}
functions which are nonexistent (and never were).
The description was also incomplete and could cause confusion. Updated
comment is more elaborate on usage and caveats.
Cc: Chris Wilson
Cc: Matthew Auld
Cc: Mika Kuoppala
Signed-
As well as knowing when the error occurred, it is more interesting to me
to know how long after booting the error occurred, and for good measure
record the time since last hw initialisation.
Signed-off-by: Chris Wilson
Reviewed-by: Matthew Auld
---
drivers/gpu/drm/i915/i915_drv.h | 4 +++
1 - 100 of 131 matches
Mail list logo
