[Intel-gfx] ✓ Fi.CI.BAT: success for drm/i915: Record the current requests queue for execlists upon hang

2016-10-12 Thread Patchwork
== Series Details == Series: drm/i915: Record the current requests queue for execlists upon hang URL : https://patchwork.freedesktop.org/series/13660/ State : success == Summary == Series 13660v1 drm/i915: Record the current requests queue for execlists upon hang

Re: [Intel-gfx] ✗ Fi.CI.BAT: failure for drm/i915: Treat a framebuffer reference as an active reference whilst shrinking

2016-10-12 Thread Saarinen, Jani
> == Series Details == > > Series: drm/i915: Treat a framebuffer reference as an active reference whilst > shrinking > URL : https://patchwork.freedesktop.org/series/13648/ > State : failure > > == Summary == > > Series 13648v1 drm/i915: Treat a framebuffer reference as an active > reference

Re: [Intel-gfx] ✗ Fi.CI.BAT: warning for drm/i915/hsw: Fix GPU hang during resume from S3-devices state

2016-10-12 Thread Saarinen, Jani
> == Series Details == > > Series: drm/i915/hsw: Fix GPU hang during resume from S3-devices state > URL : https://patchwork.freedesktop.org/series/13654/ > State : warning > > == Summary == > > Series 13654v1 drm/i915/hsw: Fix GPU hang during resume from S3-devices > state >

Re: [Intel-gfx] ✗ Fi.CI.BAT: warning for Support for sustained capturing of GuC firmware logs (rev11)

2016-10-12 Thread Saarinen, Jani
> == Series Details == > > Series: Support for sustained capturing of GuC firmware logs (rev11) > URL : https://patchwork.freedesktop.org/series/7910/ > State : warning > > == Summary == > > Series 7910v11 Support for sustained capturing of GuC firmware logs >

Re: [Intel-gfx] drm/i915: WARN_ON_ONCE(!crtc_clock || cdclk < crtc_clock)

2016-10-12 Thread Paul Bolle
On Wed, 2016-10-12 at 14:06 +0200, Paul Bolle wrote: > That might take some time. Because bisecting always takes a long time > and especially since hitting this WARNING sometimes takes over an hour. > Anyhow, please prod me if I stay silent for too long. For the record: I just had to power cycle

[Intel-gfx] ✗ Fi.CI.BAT: warning for Add support for Legacy Hdmi audio

2016-10-12 Thread Patchwork
== Series Details == Series: Add support for Legacy Hdmi audio URL : https://patchwork.freedesktop.org/series/13661/ State : warning == Summary == Series 13661v1 Add support for Legacy Hdmi audio https://patchwork.freedesktop.org/api/1.0/series/13661/revisions/1/mbox/ Test

[Intel-gfx] ✗ Fi.CI.BAT: failure for series starting with [CI,1/3] drm/i915: Remove unused "valid" parameter from pte_encode

2016-10-12 Thread Patchwork
== Series Details == Series: series starting with [CI,1/3] drm/i915: Remove unused "valid" parameter from pte_encode URL : https://patchwork.freedesktop.org/series/13663/ State : failure == Summary == Series 13663v1 Series without cover letter

Re: [Intel-gfx] [PATCH v2] drm/i915: Allow compaction upto SWIOTLB max segment size

2016-10-12 Thread Konrad Rzeszutek Wilk
On Mon, Oct 10, 2016 at 11:27:00PM +0100, Chris Wilson wrote: > commit 1625e7e549c5 ("drm/i915: make compact dma scatter lists creation > work with SWIOTLB backend") took a heavy handed approach to undo the > scatterlist compaction in the face of SWIOTLB. (The compaction hit a bug > whereby we

Re: [Intel-gfx] [CI 3/3] drm/i915/gtt: Free unused lower-level page tables

2016-10-12 Thread Chris Wilson
On Wed, Oct 12, 2016 at 06:47:38PM +0200, Michał Winiarski wrote: > +static bool gen8_ppgtt_clear_pd(struct i915_address_space *vm, > struct i915_page_directory *pd, > uint64_t start, > uint64_t length) > {

Re: [Intel-gfx] [PATCH v2] drm/i915: Allow compaction upto SWIOTLB max segment size

2016-10-12 Thread Chris Wilson
On Wed, Oct 12, 2016 at 05:19:14PM -0400, Konrad Rzeszutek Wilk wrote: > On Mon, Oct 10, 2016 at 11:27:00PM +0100, Chris Wilson wrote: > > commit 1625e7e549c5 ("drm/i915: make compact dma scatter lists creation > > work with SWIOTLB backend") took a heavy handed approach to undo the > >

Re: [Intel-gfx] [PATCH RESEND 9/9] drm/i915: set proper N/M in modeset

2016-10-12 Thread Yang, Libin
> -Original Message- > From: Intel-gfx [mailto:intel-gfx-boun...@lists.freedesktop.org] On Behalf Of > Lin, Mengdong > Sent: Wednesday, October 12, 2016 10:46 AM > To: Nikula, Jani ; intel-gfx@lists.freedesktop.org > Cc: Nikula, Jani ;

Re: [Intel-gfx] [PATCH 04/19] drm/i915: Make HAS_RUNTIME_PM only take dev_priv

2016-10-12 Thread David Weinehall
On Tue, Oct 11, 2016 at 02:21:37PM +0100, Tvrtko Ursulin wrote: > From: Tvrtko Ursulin > > Saves 960 bytes of .rodata strings. > > v2: Add parantheses around dev_priv. (Ville Syrjala) > > Signed-off-by: Tvrtko Ursulin Reviewed-by: David

Re: [Intel-gfx] [PATCH 02/19] drm/i915: Make INTEL_PCH_TYPE & co only take dev_priv

2016-10-12 Thread David Weinehall
On Tue, Oct 11, 2016 at 02:21:35PM +0100, Tvrtko Ursulin wrote: > From: Tvrtko Ursulin > > This saves 1872 bytes of .rodata strings. > > v2: > * Rebase. > * Add parantheses around dev_priv. (Ville Syrjala) > > Signed-off-by: Tvrtko Ursulin

Re: [Intel-gfx] [PATCH 12/19] drm/i915: Make IS_SKYLAKE only take dev_priv

2016-10-12 Thread David Weinehall
On Tue, Oct 11, 2016 at 02:21:45PM +0100, Tvrtko Ursulin wrote: > From: Tvrtko Ursulin > > Saves 1016 bytes of .rodata strings and couple hundred of .text. > > v2: Add parantheses around dev_priv. (Ville Syrjala) > > Signed-off-by: Tvrtko Ursulin

Re: [Intel-gfx] [PATCH 2/4] drm/i915: Respect alternate_ddc_pin for all DDI ports

2016-10-12 Thread Ville Syrjälä
On Tue, Oct 11, 2016 at 10:04:00PM +0200, Maarten Maathuis wrote: > My name does not include the word "show" (Tested-by tag). Sorry about that. Some copy-paste fail I suspect. I'll fix it up. And you actually tested the v1 patches, so I totally forgot to note that in the tested-by tags :( Care

Re: [Intel-gfx] [PATCH 19/19] drm/i915: Make IS_GEN macros only take dev_priv

2016-10-12 Thread David Weinehall
On Tue, Oct 11, 2016 at 02:21:52PM +0100, Tvrtko Ursulin wrote: > From: Tvrtko Ursulin > > Saves 1416 bytes of .rodata strings. > > v2: Add parantheses around dev_priv. (Ville Syrjala) > > Signed-off-by: Tvrtko Ursulin Reviewed-by: David

Re: [Intel-gfx] [PATCH i-g-t 1/5] tests: Use bash for debugfs_wedged and drm_lib.sh

2016-10-12 Thread David Weinehall
On Fri, Oct 07, 2016 at 12:54:03PM +0300, Joonas Lahtinen wrote: > On pe, 2016-10-07 at 10:38 +0300, Jani Nikula wrote: > > The "change" to use bash just reflects current reality. All the changes > > here look simple and sane, and immediately improve the results. The work > > is already done, no

Re: [Intel-gfx] [PATCH i-g-t 1/5] tests: Use bash for debugfs_wedged and drm_lib.sh

2016-10-12 Thread Joonas Lahtinen
On ke, 2016-10-12 at 13:12 +0300, David Weinehall wrote: > On Fri, Oct 07, 2016 at 12:54:03PM +0300, Joonas Lahtinen wrote: > > I'm against converting to bash/python for no > > benefit. > > +1, Insightful. > > Most of the bashisms seem to be simple cases of the superfluous > "function" in front

Re: [Intel-gfx] [PATCH] drm/i915: Add i915 perf infrastructure

2016-10-12 Thread Joonas Lahtinen
On ti, 2016-10-11 at 12:03 -0700, Robert Bragg wrote: > > > +               case DRM_I915_PERF_PROP_MAX: > > > +                       BUG(); > > > > We already handle this case above, but I guess we still need this in > > order to silence gcc... > > right, and preferable to having a default:

Re: [Intel-gfx] [PATCH 11/19] drm/i915: Make IS_KABYLAKE only take dev_priv

2016-10-12 Thread David Weinehall
On Tue, Oct 11, 2016 at 02:21:44PM +0100, Tvrtko Ursulin wrote: > From: Tvrtko Ursulin > > Saves 1320 bytes of .rodata strings. > > v2: Add parantheses around dev_priv. (Ville Syrjala) > > Signed-off-by: Tvrtko Ursulin Reviewed-by: David

[Intel-gfx] [PATCH 1/5] drm/i915: Use fence_write() from rpm resume

2016-10-12 Thread Chris Wilson
During rpm resume we restore the fences, but we do not have the protection of struct_mutex. This rules out updating the activity tracking on the fences, and requires us to rely on the rpm as the serialisation barrier instead. [ 350.298052] [drm:intel_runtime_resume [i915]] Resuming device [

Re: [Intel-gfx] [PATCH 2/5] drm/i915: Update debugfs describe_obj() to show fault-mappable

2016-10-12 Thread Joonas Lahtinen
On ke, 2016-10-12 at 12:16 +0100, Chris Wilson wrote: > The current meaning of whether an object has a GGTT vma is very > ill-defined (and note we don't check for any partials either), it just > means that at some point it was in the GGTT but it may not be now. The > information we really care

Re: [Intel-gfx] [PATCH 15/19] drm/i915: Make IS_G4X only take dev_priv

2016-10-12 Thread David Weinehall
On Tue, Oct 11, 2016 at 02:21:48PM +0100, Tvrtko Ursulin wrote: > From: Tvrtko Ursulin > > Saves 472 bytes of .rodata strings. > > v2: Add parantheses around dev_priv. (Ville Syrjala) > > Signed-off-by: Tvrtko Ursulin Reviewed-by: David

Re: [Intel-gfx] [PATCH 18/19] drm/i915: Make INTEL_GEN only take dev_priv

2016-10-12 Thread David Weinehall
On Tue, Oct 11, 2016 at 02:21:51PM +0100, Tvrtko Ursulin wrote: > From: Tvrtko Ursulin > > Saves 968 bytes of .rodata strings. > > v2: Add parantheses around dev_priv. (Ville Syrjala) > > Signed-off-by: Tvrtko Ursulin Reviewed-by: David

Re: [Intel-gfx] [PATCH 1/5] drm/i915: Use fence_write() from rpm resume

2016-10-12 Thread Joonas Lahtinen
This was already; Reviewed-by: Joonas Lahtinen Regards, Joonas -- Joonas Lahtinen Open Source Technology Center Intel Corporation ___ Intel-gfx mailing list Intel-gfx@lists.freedesktop.org

[Intel-gfx] [PATCH] drm/i915/dp: Debug log MST active links explicitly

2016-10-12 Thread Dhinakaran Pandiyan
No functional change. Just printing the number of active links without stating what the number means is not very useful. So, add relevant text. Signed-off-by: Dhinakaran Pandiyan --- drivers/gpu/drm/i915/intel_dp_mst.c | 8 1 file changed, 4

Re: [Intel-gfx] [PATCH 13/19] drm/i915: Make IS_BROXTON only take dev_priv

2016-10-12 Thread David Weinehall
On Tue, Oct 11, 2016 at 02:21:46PM +0100, Tvrtko Ursulin wrote: > From: Tvrtko Ursulin > > Saves 1392 bytes of .rodata strings. > > v2: Add parantheses around dev_priv. (Ville Syrjala) > > Signed-off-by: Tvrtko Ursulin This patch does quite

Re: [Intel-gfx] [PATCH 1/2] drm/i915: Move common code out of i915_gpu_error.c

2016-10-12 Thread Joonas Lahtinen
On ti, 2016-10-11 at 14:32 +0100, Chris Wilson wrote: > In the next patch, I want to conditionally compile i915_gpu_error.c and > that requires moving the functions used by debug out of > i915_gpu_error.c! > > Signed-off-by: Chris Wilson Reviewed-by: Joonas Lahtinen

Re: [Intel-gfx] [PATCH 07/19] drm/i915: Make INTEL_DEVID only take dev_priv

2016-10-12 Thread David Weinehall
On Tue, Oct 11, 2016 at 02:21:40PM +0100, Tvrtko Ursulin wrote: > From: Tvrtko Ursulin > > Saves 4472 bytes of .rodata strings. > > v2: Add parantheses around dev_priv. (Ville Syrjala) Reviewed-by: David Weinehall > > Signed-off-by:

Re: [Intel-gfx] [PATCH i-g-t 1/5] tests: Use bash for debugfs_wedged and drm_lib.sh

2016-10-12 Thread Joonas Lahtinen
On ke, 2016-10-12 at 14:16 +0300, Jani Nikula wrote: > If you really care, go ahead and send the patches to make these Bourne > shell compatible, but then do also sign up for testing them on non-bash > shells. The CI won't. I don't think it's worth the trouble, but YMMV. If they're re-written

[Intel-gfx] [PATCH] drm/i915: GMBUS don't need no forcewake

2016-10-12 Thread ville . syrjala
From: Ville Syrjälä GMBUS is part of the display engine, and thus has no need for forcewake. Let's not bother trying to grab it then. I don't recall if the display engine suffers from system hangs due to multiple accesses to the same "cacheline" in mmio space. I

Re: [Intel-gfx] [PATCH 10/19] drm/i915: Make IS_HASWELL only take dev_priv

2016-10-12 Thread David Weinehall
On Tue, Oct 11, 2016 at 02:21:43PM +0100, Tvrtko Ursulin wrote: > From: Tvrtko Ursulin > > Saves 2432 bytes of .rodata strings. > > v2: Add parantheses around dev_priv. (Ville Syrjala) > > Signed-off-by: Tvrtko Ursulin Reviewed-by: David

Re: [Intel-gfx] ✗ Fi.CI.BAT: warning for drm/i915: Trust VBT aux/ddc information

2016-10-12 Thread Ville Syrjälä
On Tue, Oct 11, 2016 at 06:49:46PM -, Patchwork wrote: > == Series Details == > > Series: drm/i915: Trust VBT aux/ddc information > URL : https://patchwork.freedesktop.org/series/13600/ > State : warning > > == Summary == > > Series 13600v1 drm/i915: Trust VBT aux/ddc information >

Re: [Intel-gfx] [PATCH 14/19] drm/i915: Make HAS_L3_DPF only take dev_priv

2016-10-12 Thread David Weinehall
On Tue, Oct 11, 2016 at 02:21:47PM +0100, Tvrtko Ursulin wrote: > From: Tvrtko Ursulin > > Saves 472 bytes of .rodata strings. > > v2: Add parantheses around dev_priv. (Ville Syrjala) > > Signed-off-by: Tvrtko Ursulin Reviewed-by: David

[Intel-gfx] ✗ Fi.CI.BAT: failure for series starting with [CI,1/6] drm/i915: Move common code out of i915_gpu_error.c

2016-10-12 Thread Patchwork
== Series Details == Series: series starting with [CI,1/6] drm/i915: Move common code out of i915_gpu_error.c URL : https://patchwork.freedesktop.org/series/13630/ State : failure == Summary == Series 13630v1 Series without cover letter

[Intel-gfx] [PATCH] drm/i915: Make sure the base lives at offset 0 of all kms objects

2016-10-12 Thread ville . syrjala
From: Ville Syrjälä We occasionally depend on eg. to_intel_crtc(NULL) being NULL as well. Sprinkle in some BUILD_BUG_ON()s to make sure we don't accidentally change things in a way that would violate this assumption. Signed-off-by: Ville Syrjälä

Re: [Intel-gfx] [PATCH 03/19] drm/i915: Make HAS_GMCH_DISPLAY only take dev_priv

2016-10-12 Thread Tvrtko Ursulin
On 12/10/2016 09:17, David Weinehall wrote: On Tue, Oct 11, 2016 at 02:21:36PM +0100, Tvrtko Ursulin wrote: From: Tvrtko Ursulin More .rodata string saving by avoid __I915__ magic inside WARNs. v2: Add parantheses around dev_priv. (Ville Syrjala) Signed-off-by:

[Intel-gfx] [PATCH] drm/i915: Fix misplaced '\n' in printing the GPU error's RING_HEAD

2016-10-12 Thread Chris Wilson
'\n' is supposed to be at the end of the line, not in the middle. Fixes: cdb324bde570 ("drm/i915: Show bounds of active request in the ring...") Signed-off-by: Chris Wilson Cc: Mika Kuoppala --- drivers/gpu/drm/i915/i915_gpu_error.c | 2 +- 1

Re: [Intel-gfx] [PATCH] drm/i915: Remove unused BSM_MASK causing warning

2016-10-12 Thread Joonas Lahtinen
On ke, 2016-10-12 at 08:42 +0100, Chris Wilson wrote: > greps ok. > Reviewed-by: Chris Wilson Merged, thanks for the review. Regards, Joonas -- Joonas Lahtinen Open Source Technology Center Intel Corporation ___ Intel-gfx

[Intel-gfx] [CI 1/6] drm/i915: Move common code out of i915_gpu_error.c

2016-10-12 Thread Chris Wilson
In the next patch, I want to conditionally compile i915_gpu_error.c and that requires moving the functions used by debug out of i915_gpu_error.c! Signed-off-by: Chris Wilson Reviewed-by: Joonas Lahtinen ---

[Intel-gfx] [CI 6/6] drm/i915: Compress GPU objects in error state

2016-10-12 Thread Chris Wilson
Our error states are quickly growing, pinning kernel memory with them. The majority of the space is taken up by the error objects. These compress well using zlib and without decode are mostly meaningless, so encoding them does not hinder quickly parsing the error state for familiarity. v2: Make

[Intel-gfx] [CI 4/6] drm/i915: Always use the GTT for error capture

2016-10-12 Thread Chris Wilson
Since the GTT provides universal access to any GPU page, we can use it to reduce our plethora of read methods to just one. It also has the important characteristic of being exactly what the GPU sees - if there are incoherency problems, seeing the batch as executed (rather than as trapped inside

[Intel-gfx] [CI 3/6] drm/i915: Stop the machine whilst capturing the GPU crash dump

2016-10-12 Thread Chris Wilson
The error state is purposefully racy as we expect it to be called at any time and so have avoided any locking whilst capturing the crash dump. However, with multi-engine GPUs and multiple CPUs, those races can manifest into OOPSes as we attempt to chase dangling pointers freed on other CPUs. Under

[Intel-gfx] [CI 5/6] drm/i915: Consolidate error object printing

2016-10-12 Thread Chris Wilson
Leave all the pretty printing to userspace and simplify the error capture to only have a single common object printer. It makes the kernel code more compact, and the refactoring allows us to apply more complex transformations like compressing the output. Signed-off-by: Chris Wilson

[Intel-gfx] [CI 2/6] drm/i915: Allow disabling error capture

2016-10-12 Thread Chris Wilson
We currently capture the GPU state after we detect a hang. This is vital for us to both triage and debug hangs in the wild (post-mortem debugging). However, it comes at the cost of running some potentially dangerous code (since it has to make very few assumption about the state of the driver) that

Re: [Intel-gfx] [PATCH 03/19] drm/i915: Make HAS_GMCH_DISPLAY only take dev_priv

2016-10-12 Thread David Weinehall
On Wed, Oct 12, 2016 at 09:43:02AM +0100, Tvrtko Ursulin wrote: > > On 12/10/2016 09:17, David Weinehall wrote: > > On Tue, Oct 11, 2016 at 02:21:36PM +0100, Tvrtko Ursulin wrote: > > > From: Tvrtko Ursulin > > > > > > More .rodata string saving by avoid __I915__ magic

[Intel-gfx] drm/i915: WARN_ON_ONCE(!crtc_clock || cdclk < crtc_clock)

2016-10-12 Thread Paul Bolle
On a laptop that tracks the latest stable release (Ie, it now runs v4.8.1) I see this WARNING     WARN_ON_ONCE(!crtc_clock || cdclk < crtc_clock) Full trace pasted below. I never saw this WARNING before v4.8. Since v4.8 I've had it in all (four, actually) boots. What am I expected to do about

Re: [Intel-gfx] [PATCH 01/19] drm/i915: Make HAS_DDI and HAS_PCH_LPT_LP only take dev_priv

2016-10-12 Thread David Weinehall
On Tue, Oct 11, 2016 at 02:21:34PM +0100, Tvrtko Ursulin wrote: > From: Tvrtko Ursulin > > This saves 3248 bytes of .rodata strings. > > v2: Add parantheses around dev_priv. (Ville Syrjala) Reviewed-by: David Weinehall How come you

Re: [Intel-gfx] [PATCH 08/19] drm/i915: Make IS_IVYBRIDGE only take dev_priv

2016-10-12 Thread David Weinehall
On Tue, Oct 11, 2016 at 02:21:41PM +0100, Tvrtko Ursulin wrote: > From: Tvrtko Ursulin > > Saves 848 bytes of .rodata strings. > > v2: Add parantheses around dev_priv. (Ville Syrjala) > > Signed-off-by: Tvrtko Ursulin Reviewed-by: David

Re: [Intel-gfx] drm/i915: WARN_ON_ONCE(!crtc_clock || cdclk < crtc_clock)

2016-10-12 Thread Joonas Lahtinen
On ke, 2016-10-12 at 11:56 +0200, Paul Bolle wrote: > On a laptop that tracks the latest stable release (Ie, it now runs > v4.8.1) I see this WARNING >     WARN_ON_ONCE(!crtc_clock || cdclk < crtc_clock) > > Full trace pasted below. I never saw this WARNING before v4.8. Since > v4.8 I've had it

Re: [Intel-gfx] [PATCH 17/19] drm/i915: Make IS_VALLEYVIEW only take dev_priv

2016-10-12 Thread David Weinehall
On Tue, Oct 11, 2016 at 02:21:50PM +0100, Tvrtko Ursulin wrote: > From: Tvrtko Ursulin > > Saves 944 bytes of .rodata strings and 128 bytes of .text. > > v2: Add parantheses around dev_priv. (Ville Syrjala) > > Signed-off-by: Tvrtko Ursulin

Re: [Intel-gfx] [PATCH] drm/i915: Remove unused BSM_MASK causing warning

2016-10-12 Thread Chris Wilson
On Wed, Oct 12, 2016 at 10:18:54AM +0300, Joonas Lahtinen wrote: > Remove never used BSM{,_MASK}. BSM_MASK #define also causes a warning. > > include/drm/i915_drm.h:96:34: warning: result of ‘65535 << 20’ > requires 37 bits to represent, but ‘int’ only has 32 bits > [-Wshiftoverflow=] >

Re: [Intel-gfx] [PATCH 03/19] drm/i915: Make HAS_GMCH_DISPLAY only take dev_priv

2016-10-12 Thread David Weinehall
On Tue, Oct 11, 2016 at 02:21:36PM +0100, Tvrtko Ursulin wrote: > From: Tvrtko Ursulin > > More .rodata string saving by avoid __I915__ magic inside WARNs. > > v2: Add parantheses around dev_priv. (Ville Syrjala) > > Signed-off-by: Tvrtko Ursulin

[Intel-gfx] ✗ Fi.CI.BAT: warning for drm/i915/dp: Debug log MST active links explicitly

2016-10-12 Thread Patchwork
== Series Details == Series: drm/i915/dp: Debug log MST active links explicitly URL : https://patchwork.freedesktop.org/series/13627/ State : warning == Summary == Series 13627v1 drm/i915/dp: Debug log MST active links explicitly

[Intel-gfx] [PATCH] kernel/cpu: Distinct name for cpu_hotplug.dep_map

2016-10-12 Thread Joonas Lahtinen
Hi Linus/Andrew, This patch has been waiting for feedback or merging for ~8 months now. It was posted to linux-ker...@vger.kernel.org initially and got an immediate Acked-by. After no other feedback, I revised the assumed maintainers and still no reactions. Next I changed over to

[Intel-gfx] [PATCH 5/5] drm/i915: Remove superfluous locking around userfault_list

2016-10-12 Thread Chris Wilson
Now that we have reduced the access to the list to either (a) under the struct_mutex whilst holding the RPM wakeref (so that concurrent writers to the list are serialised by struct_mutex) and (b) under the atomic runtime suspend (which cannot run concurrently with any other accessor due to the

[Intel-gfx] [PATCH 3/5] drm/i915: Move user fault tracking to a separate list

2016-10-12 Thread Chris Wilson
We want to decouple RPM and struct_mutex, but currently RPM has to walk the list of bound objects and remove userspace mmapping before we suspend (otherwise userspace may continue to access the GTT whilst it is powered down). This currently requires the struct_mutex to walk the bound_list, but if

[Intel-gfx] [PATCH 2/5] drm/i915: Update debugfs describe_obj() to show fault-mappable

2016-10-12 Thread Chris Wilson
The current meaning of whether an object has a GGTT vma is very ill-defined (and note we don't check for any partials either), it just means that at some point it was in the GGTT but it may not be now. The information we really care about here is whether it is taking up precious mappable aperture

Re: [Intel-gfx] [PATCH i-g-t 1/5] tests: Use bash for debugfs_wedged and drm_lib.sh

2016-10-12 Thread Jani Nikula
On Wed, 12 Oct 2016, David Weinehall wrote: > On Fri, Oct 07, 2016 at 12:54:03PM +0300, Joonas Lahtinen wrote: >> On pe, 2016-10-07 at 10:38 +0300, Jani Nikula wrote: >> > The "change" to use bash just reflects current reality. All the changes >> > here look

[Intel-gfx] [PATCH 4/5] drm/i915: Use RPM as the barrier for controlling user mmap access

2016-10-12 Thread Chris Wilson
We can remove the false coupling between RPM and struct mutex by the observation that we can use the RPM wakeref as the barrier around user mmap access. That is as we tear down the user's PTE atomically from within rpm suspend and then to fault in new PTE requires the rpm wakeref, means that no

[Intel-gfx] [CI 2/2] drm/i915: Update debugfs describe_obj() to show fault-mappable

2016-10-12 Thread Chris Wilson
The current meaning of whether an object has a GGTT vma is very ill-defined (and note we don't check for any partials either), it just means that at some point it was in the GGTT but it may not be now. The information we really care about here is whether it is taking up precious mappable aperture

[Intel-gfx] [CI 1/2] drm/i915: Use fence_write() from rpm resume

2016-10-12 Thread Chris Wilson
During rpm resume we restore the fences, but we do not have the protection of struct_mutex. This rules out updating the activity tracking on the fences, and requires us to rely on the rpm as the serialisation barrier instead. [ 350.298052] [drm:intel_runtime_resume [i915]] Resuming device [

[Intel-gfx] ✗ Fi.CI.BAT: failure for series starting with [CI,1/2] drm/i915: Use fence_write() from rpm resume

2016-10-12 Thread Patchwork
== Series Details == Series: series starting with [CI,1/2] drm/i915: Use fence_write() from rpm resume URL : https://patchwork.freedesktop.org/series/13642/ State : failure == Summary == Series 13642v1 Series without cover letter

[Intel-gfx] [PATCH igt] igt/gem_wait: Use explicit timers

2016-10-12 Thread Chris Wilson
Rather than guestimating a workload that should take a certain amount of time, use a sigitimer to terminate a batch (and so complete the wait) after an exact amount of time. And in the process expand testing to cover multiple rings and hangcheck. Signed-off-by: Chris Wilson

Re: [Intel-gfx] [PATCH 06/19] drm/i915: Make IS_GEN-range macro only take dev_priv

2016-10-12 Thread David Weinehall
On Tue, Oct 11, 2016 at 02:21:39PM +0100, Tvrtko Ursulin wrote: > From: Tvrtko Ursulin > > Saves 944 bytes of .rodata strings. > > v2: Add parantheses around dev_priv. (Ville Syrjala) > > Signed-off-by: Tvrtko Ursulin Reviewed-by: David

Re: [Intel-gfx] [PATCH i-g-t] tests/kms_plane_multiple: CRC based atomic correctness test

2016-10-12 Thread Maarten Lankhorst
Op 07-10-16 om 13:45 schreef Mika Kahola: > This is a testcase with multiple planes. The idea here is the following > > - draw a uniform frame with blue color > - grab crc for reference > - put planes randomly on top with the same blue color > - punch holes with black color into the primary

Re: [Intel-gfx] drm/i915: WARN_ON_ONCE(!crtc_clock || cdclk < crtc_clock)

2016-10-12 Thread Jani Nikula
On Wed, 12 Oct 2016, Paul Bolle wrote: > On Wed, 2016-10-12 at 14:08 +0300, Joonas Lahtinen wrote: >> Bisecting the offending commit between v4.8 and v4.8.1 would be a good >> start. > > That would be between v4.7 and v4.8. (I guess my report was ambiguous.) > > That might

[Intel-gfx] [PATCH igt] igt: Add exerciser for execbuf fence-out <-> fence-in

2016-10-12 Thread Chris Wilson
When execbuf2 supports explicit fencing with sync_file in/out fences (via a fence-fd), we can control execution via the fence. Signed-off-by: Chris Wilson --- tests/Makefile.sources | 1 + tests/gem_exec_fence.c | 377 +

Re: [Intel-gfx] [PATCH] drm/i915/hsw: Fix GPU hang during resume from S3-devices state

2016-10-12 Thread Chris Wilson
On Wed, Oct 12, 2016 at 05:46:37PM +0300, Imre Deak wrote: > Currently resuming on HSW from S3 pm_test/devices state leads to an > unrecoverable GPU hang. Resetting the GPU during suspend fixes this. For > a full S3 cycle this change only means the reset happens earlier (before > reaching S3). For

[Intel-gfx] ✗ Fi.CI.BAT: failure for series starting with [CI,1/3] drm/i915: Remove unused "valid" parameter from pte_encode

2016-10-12 Thread Patchwork
== Series Details == Series: series starting with [CI,1/3] drm/i915: Remove unused "valid" parameter from pte_encode URL : https://patchwork.freedesktop.org/series/13646/ State : failure == Summary == LD fs/btrfs/built-in.o LD arch/x86/kernel/cpu/built-in.o LD

[Intel-gfx] [PATCH] drm/i915/hsw: Fix GPU hang during resume from S3-devices state

2016-10-12 Thread Imre Deak
Currently resuming on HSW from S3 pm_test/devices state leads to an unrecoverable GPU hang. Resetting the GPU during suspend fixes this. For a full S3 cycle this change only means the reset happens earlier (before reaching S3). For S4 the reset will happen now both during the freeze and quiesce

Re: [Intel-gfx] drm/i915: WARN_ON_ONCE(!crtc_clock || cdclk < crtc_clock)

2016-10-12 Thread Paul Bolle
On Wed, 2016-10-12 at 17:34 +0300, Jani Nikula wrote: > In the mean time, please file a bug over at [1] so we don't lose > track. Done:  https://bugs.freedesktop.org/show_bug.cgi?id=98214 Paul Bolle ___ Intel-gfx mailing list

[Intel-gfx] [PATCH i-g-t] intel-ci: Remove kms_psr_sink_crc@psr_basic from BAT

2016-10-12 Thread Petri Latvala
The test is producing a lot of CI noise. Signed-off-by: Petri Latvala --- Will be pushed shortly. Visible in CI results tomorrow-ish. tests/intel-ci/fast-feedback.testlist | 1 - 1 file changed, 1 deletion(-) diff --git a/tests/intel-ci/fast-feedback.testlist

[Intel-gfx] [CI] drm/i915: Treat a framebuffer reference as an active reference whilst shrinking

2016-10-12 Thread Chris Wilson
Treat a framebuffer reference with the same priority as an active reference whilst shrinking. Framebuffers are likely to be reused and typically cost more to migrate to and from GPU memory (on LLC architectures we need to clflush), so defer the temptation to purge them during a kswapd run until we

Re: [Intel-gfx] [PATCH i-g-t 1/5] tests: Use bash for debugfs_wedged and drm_lib.sh

2016-10-12 Thread Jani Nikula
On Wed, 12 Oct 2016, Joonas Lahtinen wrote: > On ke, 2016-10-12 at 14:16 +0300, Jani Nikula wrote: >> If you really care, go ahead and send the patches to make these Bourne >> shell compatible, but then do also sign up for testing them on non-bash >> shells. The

Re: [Intel-gfx] ✗ Fi.CI.BAT: warning for drm/i915: Fix misplaced '\n' in printing the GPU error's RING_HEAD

2016-10-12 Thread Saarinen, Jani
> Test kms_flip: > Subgroup basic-flip-vs-modeset: > pass -> DMESG-WARN (fi-skl-6770hq) [ 468.807117] [drm:skl_set_cdclk [i915]] *ERROR* failed to inform PCU about cdclk change [ 468.816844] [drm:intel_cpu_fifo_underrun_irq_handler [i915]] *ERROR* CPU pipe C FIFO

[Intel-gfx] [PATCH 1/8] drm/i915/skl+: Prepare for removing data rate from skl watermark state

2016-10-12 Thread Maarten Lankhorst
Caching is not required, drm_atomic_crtc_state_for_each_plane_state can be used to inspect all plane_states that are assigned to the current crtc_state, so we can just recalculate every time. Signed-off-by: Maarten Lankhorst --- drivers/gpu/drm/i915/intel_pm.c

[Intel-gfx] [PATCH 2/8] drm/i915/skl+: Remove data_rate from watermark struct.

2016-10-12 Thread Maarten Lankhorst
It's only used in one function, and can be calculated without caching it in the global struct by using drm_atomic_crtc_state_for_each_plane_state. Signed-off-by: Maarten Lankhorst --- drivers/gpu/drm/i915/intel_drv.h | 4 drivers/gpu/drm/i915/intel_pm.c

[Intel-gfx] [PATCH 0/8] drm/i915/gen9+: Atomic wm fixes.

2016-10-12 Thread Maarten Lankhorst
This patch series applies on top of Lyude's patches. They clean up the remainder of SKL style wm's, and finally makes SKL watermarks ready for nonblocking modeset by using the crtc_state for watermarks as much as possible. Maarten Lankhorst (8): drm/i915/skl+: Prepare for removing data rate

[Intel-gfx] [PATCH 5/8] drm/i915: Add a atomic evasion step to watermark programming.

2016-10-12 Thread Maarten Lankhorst
Allow the driver to write watermarks during atomic evasion. This will make it possible to write the watermarks in a cleaner way on gen9+. Signed-off-by: Maarten Lankhorst --- drivers/gpu/drm/i915/i915_drv.h | 6 --

[Intel-gfx] [PATCH 6/8] drm/i915/gen9+: Use the watermarks from crtc_state for everything.

2016-10-12 Thread Maarten Lankhorst
There's no need to keep a duplicate skl_pipe_wm around any more, everything can be discovered from crtc_state, which we pass around correctly now even in case of plane disable. Signed-off-by: Maarten Lankhorst --- drivers/gpu/drm/i915/intel_display.c | 2 +-

[Intel-gfx] ✗ Fi.CI.BAT: warning for drm/i915: GMBUS don't need no forcewake

2016-10-12 Thread Patchwork
== Series Details == Series: drm/i915: GMBUS don't need no forcewake URL : https://patchwork.freedesktop.org/series/13641/ State : warning == Summary == Series 13641v1 drm/i915: GMBUS don't need no forcewake https://patchwork.freedesktop.org/api/1.0/series/13641/revisions/1/mbox/ Test

Re: [Intel-gfx] [PATCH 09/19] drm/i915: Make IS_BROADWELL only take dev_priv

2016-10-12 Thread David Weinehall
On Tue, Oct 11, 2016 at 02:21:42PM +0100, Tvrtko Ursulin wrote: > From: Tvrtko Ursulin > > Saves 1808 bytes of .rodata strings. > > v2: Add parantheses around dev_priv. (Ville Syrjala) > > Signed-off-by: Tvrtko Ursulin Reviewed-by: David

Re: [Intel-gfx] [RFC i-g-t PATCH 2/3] igt/gem_wait: Use new igt_dummyload api

2016-10-12 Thread Chris Wilson
On Wed, Oct 12, 2016 at 02:59:53PM +0300, Abdiel Janulgue wrote: > Signed-off-by: Abdiel Janulgue > --- > tests/gem_wait.c | 77 > +--- > 1 file changed, 12 insertions(+), 65 deletions(-) We can do so much

[Intel-gfx] [RFC i-g-t PATCH 2/3] igt/gem_wait: Use new igt_dummyload api

2016-10-12 Thread Abdiel Janulgue
Signed-off-by: Abdiel Janulgue --- tests/gem_wait.c | 77 +--- 1 file changed, 12 insertions(+), 65 deletions(-) diff --git a/tests/gem_wait.c b/tests/gem_wait.c index 461efdb..24a5f5e 100644 ---

[Intel-gfx] [RFC i-g-t PATCH 1/3] lib: add igt_dummyload

2016-10-12 Thread Abdiel Janulgue
Generalized from auto-tuned GPU dummy workload in gem_wait and kms_flip Signed-off-by: Abdiel Janulgue --- lib/Makefile.sources | 2 + lib/igt.h| 1 + lib/igt_dummyload.c | 419 +++

[Intel-gfx] [RFC i-g-t PATCH 3/3] igt/kms_flip: Use new igt_dummyload api

2016-10-12 Thread Abdiel Janulgue
Signed-off-by: Abdiel Janulgue --- tests/kms_flip.c | 191 +++ 1 file changed, 10 insertions(+), 181 deletions(-) diff --git a/tests/kms_flip.c b/tests/kms_flip.c index 065ad66..93cf391 100644 ---

[Intel-gfx] [RFC i-g-t] Extract autotuned dummy load into lib

2016-10-12 Thread Abdiel Janulgue
A lot of igt testcases need some dummy load to make sure a race window is big enough. Unfortunately having a fixed amount of workload leads to spurious test failures or overtly long runtimes on some fast/slow platforms. This library contains functionality to submit GPU workloads that is

Re: [Intel-gfx] [PATCH 13/19] drm/i915: Make IS_BROXTON only take dev_priv

2016-10-12 Thread Tvrtko Ursulin
On 12/10/2016 12:52, David Weinehall wrote: On Tue, Oct 11, 2016 at 02:21:46PM +0100, Tvrtko Ursulin wrote: From: Tvrtko Ursulin Saves 1392 bytes of .rodata strings. v2: Add parantheses around dev_priv. (Ville Syrjala) Signed-off-by: Tvrtko Ursulin

Re: [Intel-gfx] [PATCH] drm/i915: GMBUS don't need no forcewake

2016-10-12 Thread Chris Wilson
On Wed, Oct 12, 2016 at 02:44:47PM +0300, ville.syrj...@linux.intel.com wrote: > From: Ville Syrjälä > > GMBUS is part of the display engine, and thus has no need for > forcewake. Let's not bother trying to grab it then. > > I don't recall if the display engine

[Intel-gfx] ✗ Fi.CI.BAT: warning for series starting with [1/5] drm/i915: Use fence_write() from rpm resume

2016-10-12 Thread Patchwork
== Series Details == Series: series starting with [1/5] drm/i915: Use fence_write() from rpm resume URL : https://patchwork.freedesktop.org/series/13636/ State : warning == Summary == Series 13636v1 Series without cover letter

Re: [Intel-gfx] [PATCH 20/42] drm/i915: Implement pread without struct-mutex

2016-10-12 Thread Joonas Lahtinen
On pe, 2016-10-07 at 10:46 +0100, Chris Wilson wrote: > @@ -871,7 +845,7 @@ shmem_clflush_swizzled_range(char *addr, unsigned long > length, >  /* Only difference to the fast-path function is that this can handle bit17 >   * and uses non-atomic copy and kmap functions. */ >  static int >

[Intel-gfx] ✗ Fi.CI.BAT: warning for drm/i915: Fix misplaced '\n' in printing the GPU error's RING_HEAD

2016-10-12 Thread Patchwork
== Series Details == Series: drm/i915: Fix misplaced '\n' in printing the GPU error's RING_HEAD URL : https://patchwork.freedesktop.org/series/13639/ State : warning == Summary == Series 13639v1 drm/i915: Fix misplaced ' ' in printing the GPU error's RING_HEAD

Re: [Intel-gfx] [PATCH i-g-t 1/5] tests: Use bash for debugfs_wedged and drm_lib.sh

2016-10-12 Thread David Weinehall
On Wed, Oct 12, 2016 at 04:04:34PM +0300, Jani Nikula wrote: > On Wed, 12 Oct 2016, Joonas Lahtinen wrote: > > On ke, 2016-10-12 at 14:16 +0300, Jani Nikula wrote: > >> If you really care, go ahead and send the patches to make these Bourne > >> shell compatible,

[Intel-gfx] [CI 3/3] drm/i915/gtt: Free unused lower-level page tables

2016-10-12 Thread Michał Winiarski
Since "Dynamic page table allocations" were introduced, our page tables can grow (being dynamically allocated) with address space range usage. Unfortunately, their lifetime is bound to vm. This is not a huge problem when we're not using softpin - drm_mm is creating an upper bound on used range by

[Intel-gfx] [CI 1/3] drm/i915: Remove unused "valid" parameter from pte_encode

2016-10-12 Thread Michał Winiarski
We never used any invalid ptes, those were put in place for a possibility of doing gpu faults. However our batchbuffers are not restricted in length, so everything needs to be pointing to something and thus out-of-bounds is pointing to scratch. Remove the valid flag as it is always true. v2:

[Intel-gfx] [CI 2/3] drm/i915/gtt: Split gen8_ppgtt_clear_pte_range

2016-10-12 Thread Michał Winiarski
Let's use more top-down approach, where each gen8_ppgtt_clear_* function is responsible for clearing the struct passed as an argument and calling relevant clear_range functions on lower-level tables. Doing this rather than operating on PTE ranges makes the implementation of shrinking page tables

Re: [Intel-gfx] [PATCH] drm/i915: GMBUS don't need no forcewake

2016-10-12 Thread Ville Syrjälä
On Wed, Oct 12, 2016 at 12:58:34PM +0100, Chris Wilson wrote: > On Wed, Oct 12, 2016 at 02:44:47PM +0300, ville.syrj...@linux.intel.com wrote: > > From: Ville Syrjälä > > > > GMBUS is part of the display engine, and thus has no need for > > forcewake. Let's not

Re: [Intel-gfx] [PATCH 13/19] drm/i915: Make IS_BROXTON only take dev_priv

2016-10-12 Thread David Weinehall
On Wed, Oct 12, 2016 at 01:06:51PM +0100, Tvrtko Ursulin wrote: > > On 12/10/2016 12:52, David Weinehall wrote: > > On Tue, Oct 11, 2016 at 02:21:46PM +0100, Tvrtko Ursulin wrote: > > > From: Tvrtko Ursulin > > > > > > Saves 1392 bytes of .rodata strings. > > > > > >

Re: [Intel-gfx] [PATCH] drm/i915: GMBUS don't need no forcewake

2016-10-12 Thread Chris Wilson
On Wed, Oct 12, 2016 at 03:39:47PM +0300, Ville Syrjälä wrote: > On Wed, Oct 12, 2016 at 12:58:34PM +0100, Chris Wilson wrote: > > On Wed, Oct 12, 2016 at 02:44:47PM +0300, ville.syrj...@linux.intel.com > > wrote: > > > From: Ville Syrjälä > > > > > > GMBUS is

  1   2   >