[Intel-gfx] [PATCH igt 4/4] lib: Free all internal buffers before measuring available memory

Release the internal caches (by flushing the idle_worker) to maximise the available memory for use by tests (and to reduce sporadic skipping when on the cusp). Signed-off-by: Chris Wilson --- lib/intel_os.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff

[Intel-gfx] ✓ Fi.CI.BAT: success for tests/kms_plane_lowres: Rework tests to work without fbcon. (rev2)

== Series Details == Series: tests/kms_plane_lowres: Rework tests to work without fbcon. (rev2) URL : https://patchwork.freedesktop.org/series/32016/ State : success == Summary == IGT patchset tested on top of latest successful build 9a3f6f59d4c208219f7d3178106bdef628ff68d0 tests/intel-ci:

[Intel-gfx] [PATCH v2] drm/i915: Flush the idle-worker for debugfs/i915_drop_caches

After being requested to idle the GPU, flush the idle worker to drop the residual active state, and any internal object caches. v2: By popular demand, introduce DROP_IDLE for fine-grained control from userspace, though it should be used as part of a DROP_ACTIVE | DROP_RETIRE | DROP_IDLE |

Re: [Intel-gfx] [PATCH i-g-t] kms_atomic_transition: Output more finegrained progress info to avoid CI watchdog timeout

On 10/18/2017 02:50 PM, Imre Deak wrote: On Wed, Oct 18, 2017 at 02:47:17PM +0300, Imre Deak wrote: On Wed, Oct 18, 2017 at 02:43:38PM +0300, Petri Latvala wrote: On Wed, Oct 18, 2017 at 02:29:33PM +0300, Imre Deak wrote: The CI software watchdog (owatch) will timeout if the test doesn't

[Intel-gfx] ✓ Fi.CI.BAT: success for kms_atomic_transition: Output more finegrained progress info to avoid CI watchdog timeout

== Series Details == Series: kms_atomic_transition: Output more finegrained progress info to avoid CI watchdog timeout URL : https://patchwork.freedesktop.org/series/32206/ State : success == Summary == IGT patchset tested on top of latest successful build

Re: [Intel-gfx] [PATCH v7 2/4] drm/i915/guc : Removing i915_modparams.enable_guc_loading module parameter

On Wed, 18 Oct 2017 00:50:47 +0200, Sujaritha Sundaresan wrote: We currently have two module parameters that control GuC: "enable_guc_loading" and "enable_guc_submission". Whenever we need i915_modparams.enable_guc_submission=1, we also need

Re: [Intel-gfx] [PATCH i-g-t] kms_atomic_transition: Output more finegrained progress info to avoid CI watchdog timeout

On Wed, Oct 18, 2017 at 02:47:17PM +0300, Imre Deak wrote: > On Wed, Oct 18, 2017 at 02:43:38PM +0300, Petri Latvala wrote: > > On Wed, Oct 18, 2017 at 02:29:33PM +0300, Imre Deak wrote: > > > The CI software watchdog (owatch) will timeout if the test doesn't > > > output anything for a long time

Re: [Intel-gfx] [PATCH i-g-t] kms_atomic_transition: Output more finegrained progress info to avoid CI watchdog timeout

On Wed, Oct 18, 2017 at 02:43:38PM +0300, Petri Latvala wrote: > On Wed, Oct 18, 2017 at 02:29:33PM +0300, Imre Deak wrote: > > The CI software watchdog (owatch) will timeout if the test doesn't > > output anything for a long time on standard out or error. At least the > >

Re: [Intel-gfx] [PATCH i-g-t] kms_atomic_transition: Output more finegrained progress info to avoid CI watchdog timeout

On Wed, Oct 18, 2017 at 02:29:33PM +0300, Imre Deak wrote: > The CI software watchdog (owatch) will timeout if the test doesn't > output anything for a long time on standard out or error. At least the > plane-all-modeset-transition and plane-all-modeset-transition-fences > subtests run without any

Re: [Intel-gfx] [PATCH i-g-t] kms_atomic_transition: Output more finegrained progress info to avoid CI watchdog timeout

Quoting Imre Deak (2017-10-18 12:29:33) > The CI software watchdog (owatch) will timeout if the test doesn't > output anything for a long time on standard out or error. At least the > plane-all-modeset-transition and plane-all-modeset-transition-fences > subtests run without any output longer than

[Intel-gfx] ✓ Fi.CI.BAT: success for tests/kms_cursor_legacy: Remove strict vblank checking on 2x-*-vs-cursor tests.

== Series Details == Series: tests/kms_cursor_legacy: Remove strict vblank checking on 2x-*-vs-cursor tests. URL : https://patchwork.freedesktop.org/series/32202/ State : success == Summary == IGT patchset tested on top of latest successful build 9a3f6f59d4c208219f7d3178106bdef628ff68d0

Re: [Intel-gfx] [PATCH i-g-t] tests/kms_cursor_legacy: Remove strict vblank checking on 2x-*-vs-cursor tests.

Op 18-10-17 om 12:47 schreef Chris Wilson: > Quoting Maarten Lankhorst (2017-10-18 11:43:25) >> There is no guarantee that on 2 screens the vblanks are in any way >> synchronized, so it's best to stop checking them, since doing an update >> involving multiple pipes may incur an arbitrary amount of

[Intel-gfx] [PATCH i-g-t] kms_atomic_transition: Output more finegrained progress info to avoid CI watchdog timeout

The CI software watchdog (owatch) will timeout if the test doesn't output anything for a long time on standard out or error. At least the plane-all-modeset-transition and plane-all-modeset-transition-fences subtests run without any output longer than the watchdog timeout, so output some more

[Intel-gfx] ✓ Fi.CI.BAT: success for drm/i915: Disable lazy PPGTT page table optimization for vGPU

== Series Details == Series: drm/i915: Disable lazy PPGTT page table optimization for vGPU URL : https://patchwork.freedesktop.org/series/32201/ State : success == Summary == Series 32201v1 drm/i915: Disable lazy PPGTT page table optimization for vGPU

Re: [Intel-gfx] [PATCH i-g-t] tests/pm_rps: Move some test logic out of boost function

On Wed, 2017-10-18 at 11:10 +0530, Sagar Arun Kamble wrote: > > On 10/17/2017 6:13 PM, Radoslaw Szwichtenberg wrote: > > Moving code out of the boost function will allow its usage > > in other/new test scenarios. > > > > Signed-off-by: Radoslaw Szwichtenberg >

Re: [Intel-gfx] [PATCH v7 2/4] drm/i915/guc : Removing i915_modparams.enable_guc_loading module parameter

On Tue, 2017-10-17 at 15:50 -0700, Sujaritha Sundaresan wrote: > We currently have two module parameters that control GuC: > "enable_guc_loading" and "enable_guc_submission". Whenever > we need i915_modparams.enable_guc_submission=1, we also need > enable_guc_loading=1. We also need

Re: [Intel-gfx] [PATCH i-g-t] tests/kms_cursor_legacy: Remove strict vblank checking on 2x-*-vs-cursor tests.

Quoting Maarten Lankhorst (2017-10-18 11:43:25) > There is no guarantee that on 2 screens the vblanks are in any way > synchronized, so it's best to stop checking them, since doing an update > involving multiple pipes may incur an arbitrary amount of delay. Ah, that's a shame. Not much point in

[Intel-gfx] [PATCH i-g-t] tests/kms_cursor_legacy: Remove strict vblank checking on 2x-*-vs-cursor tests.

There is no guarantee that on 2 screens the vblanks are in any way synchronized, so it's best to stop checking them, since doing an update involving multiple pipes may incur an arbitrary amount of delay. Signed-off-by: Maarten Lankhorst Bugzilla:

Re: [Intel-gfx] [PATCH] drm/i915: Disable lazy PPGTT page table optimization for vGPU

Quoting Joonas Lahtinen (2017-10-18 11:25:43) > When running under virtualization (vGPU active), we must disable > the azy PPGTT page table initialization optimization introduced by: > > 14826673247e ("drm/i915: Only initialize partially filled pagetables") > > We must do this because GVT-g

Re: [Intel-gfx] [PATCH v7 3/4] drm/i915/guc : Updating GuC and HuC FW select function

On 10/18/2017 4:20 AM, Sujaritha Sundaresan wrote: Updating GuC and HuC firmware select function to support removing i915_modparams.enable_guc_loading module parameter. v2: Clarifying the commit message (Anusha) v3: Unify seq_puts messages, Re-factoring code as per review (Michal) v4:

[Intel-gfx] [PATCH] drm/i915: Disable lazy PPGTT page table optimization for vGPU

When running under virtualization (vGPU active), we must disable the azy PPGTT page table initialization optimization introduced by: 14826673247e ("drm/i915: Only initialize partially filled pagetables") We must do this because GVT-g makes unduly assumptions about guest behaviour, which this

[Intel-gfx] ✗ Fi.CI.BAT: failure for igt/pm_rc6_residency: Allow some leeway on the upper %% bound

== Series Details == Series: igt/pm_rc6_residency: Allow some leeway on the upper %% bound URL : https://patchwork.freedesktop.org/series/32112/ State : failure == Summary == Series 32112 revision 1 was fully merged or fully failed: no git log == Logs == For more details see:

[Intel-gfx] ✓ Fi.CI.BAT: success for drm/i915: Do not rely on wm preservation for ILK watermarks

== Series Details == Series: drm/i915: Do not rely on wm preservation for ILK watermarks URL : https://patchwork.freedesktop.org/series/32197/ State : success == Summary == Series 32197v1 drm/i915: Do not rely on wm preservation for ILK watermarks

[Intel-gfx] ✓ Fi.CI.IGT: success for HAX: kernel/swsusp: Show invalid swap signature (rev2)

== Series Details == Series: HAX: kernel/swsusp: Show invalid swap signature (rev2) URL : https://patchwork.freedesktop.org/series/32130/ State : success == Summary == shard-hswtotal:2551 pass:1441 dwarn:0 dfail:0 fail:9 skip:1101 time:9283s == Logs == For more details see:

Re: [Intel-gfx] linux-next: manual merge of the sound-asoc tree with the drm-misc tree

On Wed, Oct 18, 2017 at 10:57:33AM +0100, Mark Brown wrote: > I fixed it up (see below) and can carry the fix as necessary. This > is now fixed as far as linux-next is concerned, but any non trivial > conflicts should be mentioned to your upstream maintainer when your tree > is submitted for

Re: [Intel-gfx] [PATCH v5] drm/i915: Replace *_reference/unreference() or *_ref/unref with _get/put()

On Sun, Oct 15, 2017 at 12:06:44AM +0530, Harsha Sharma wrote: > Replace instances of drm_framebuffer_reference/unreference() with > *_get/put() suffixes and drm_dev_unref with *_put() suffix > because get/put is shorter and consistent with the > kernel use of *_get/put suffixes. > Done with

[Intel-gfx] linux-next: manual merge of the sound-asoc tree with the drm-misc tree

Hi all, Today's linux-next merge of the sound-asoc tree got a conflict in: drivers/gpu/drm/amd/include/amd_shared.h between commit: cfa289fd4986c ("drm/amdgpu: rename amdgpu_dpm_funcs to amd_pm_funcs") from the drm-misc tree and commit: 1e4448648333a ("drm/amdgpu Moving amdgpu asic

Re: [Intel-gfx] [PATCH igt] igt/pm_rc6_residency: Measure residency after checking for applicability

On Tue, Oct 17, 2017 at 09:49:54AM +0100, Chris Wilson wrote: > CI doesn't run in whole-test mode, but runs each subtest individually. > Tests that are designed to do a block of work to be shared between many > subtests end up running that work multiple times (once per subtest) and > worse, that

[Intel-gfx] ✓ Fi.CI.BAT: success for lib/i915: Move submission related helpers to lib/i915/gem_submission

== Series Details == Series: lib/i915: Move submission related helpers to lib/i915/gem_submission URL : https://patchwork.freedesktop.org/series/32184/ State : success == Summary == IGT patchset tested on top of latest successful build e66027ceef6266ef72fc7485567f518d2ce48817 tests: remove

Re: [Intel-gfx] [PATCH v2] drm/i915: Fixup userptr mmu notifier registration error handling

On 18/10/2017 10:07, Daniel Vetter wrote: On Tue, Oct 17, 2017 at 04:09:08PM +0100, Tvrtko Ursulin wrote: From: Tvrtko Ursulin Avoid dereferencing the error pointer and also avoid returning NULL from i915_mmu_notifier_find since the callers do not expect that. How

[Intel-gfx] ✓ Fi.CI.IGT: success for drm/i915: Don't recheck link status for eDP display.

== Series Details == Series: drm/i915: Don't recheck link status for eDP display. URL : https://patchwork.freedesktop.org/series/32164/ State : success == Summary == Test kms_flip: Subgroup flip-vs-rmfb: pass -> DMESG-WARN (shard-hsw) fdo#102614 +1 Test

[Intel-gfx] [PATCH] drm/i915: Do not rely on wm preservation for ILK watermarks

The original intent was to preserve watermarks as much as possible in intel_pipe_wm.raw_wm, and put the validated ones in intel_pipe_wm.wm. It seems this approach is insufficient and we don't always preserve the raw watermarks, so just use the atomic iterator we're already using to get a const

[Intel-gfx] linux-next: manual merge of the drm-intel tree with the drm-intel-fixes tree

Hi all, Today's linux-next merge of the drm-intel tree got a conflict in: drivers/gpu/drm/i915/i915_gem_evict.c between commit: 99b169d3c2052 ("drm/i915: Fix eviction when the GGTT is idle but full") from the drm-intel-fixes tree and commit: 9c1477e83e629 ("drm/i915/selftests: Exercise

Re: [Intel-gfx] [PATCH v2] drm/i915: Fixup userptr mmu notifier registration error handling

On Tue, Oct 17, 2017 at 04:09:08PM +0100, Tvrtko Ursulin wrote: > From: Tvrtko Ursulin > > Avoid dereferencing the error pointer and also avoid returning NULL > from i915_mmu_notifier_find since the callers do not expect that. How did my patch return a NULL value? At

Re: [Intel-gfx] [PATCH] drm/i915: Debugfs to disable context banning

On Tue, Oct 17, 2017 at 09:23:43AM -0700, Jeff McGee wrote: > On Tue, Oct 17, 2017 at 05:36:54PM +0200, Daniel Vetter wrote: > > On Tue, Oct 17, 2017 at 08:25:33AM -0700, jeff.mc...@intel.com wrote: > > > From: Jeff McGee > > > > > > Useful for stress testing various reset

Re: [Intel-gfx] [PATCH] drm/i915: No need for RING_MAX_NONPRIV_SLOTS space

Quoting Oscar Mateo (2017-10-17 21:27:51) > Now that we write RING_FORCE_TO_NONPRIV registers directly to hardware, > [commit 32ced39 ("drm/i915: Transform whitelisting WAs into a simple reg > write")] there is no need to save space for them in the list of context > workarounds. > > v2: Refer to

[Intel-gfx] ✓ Fi.CI.BAT: success for drm/i915/crt: split compute_config hook by platforms

== Series Details == Series: drm/i915/crt: split compute_config hook by platforms URL : https://patchwork.freedesktop.org/series/32135/ State : success == Summary == Series 32135v1 drm/i915/crt: split compute_config hook by platforms

Re: [Intel-gfx] [PATCH i-g-t] lib/i915: Move submission related helpers to lib/i915/gem_submission

Quoting Michał Winiarski (2017-10-18 09:02:12) > Since I accidentally broke the build for some, by putting the pretty > printer for submission inside ifdef HAVE_PROCPS, it's time to move the > whole thing into lib/i915 while fixing this mistake. > Let's also rename the pretty printer and add a doc

Re: [Intel-gfx] [PATCH i-g-t] tests: remove kms_fbc_crc

On Tue, Oct 17, 2017 at 05:10:21PM -0200, Paulo Zanoni wrote: > Em Ter, 2017-10-17 às 18:59 +0300, Ville Syrjälä escreveu: > > On Tue, Oct 17, 2017 at 05:16:40PM +0200, Daniel Vetter wrote: > > > This was the predecessor test to kms_frontbuffer_tracking, which > > > not > > > just tests fbc, and

[Intel-gfx] ✗ Fi.CI.IGT: failure for drm/i915: No need for RING_MAX_NONPRIV_SLOTS space

== Series Details == Series: drm/i915: No need for RING_MAX_NONPRIV_SLOTS space URL : https://patchwork.freedesktop.org/series/32159/ State : failure == Summary == Test drv_suspend: Subgroup sysfs-reader-hibernate: fail -> DMESG-FAIL (shard-hsw) k.org#196691 Test

Re: [Intel-gfx] ✓ Fi.CI.BAT: success for drm/i915: Fixup userptr mmu notifier registration error handling (rev2)

On 17/10/2017 17:07, Patchwork wrote: == Series Details == Series: drm/i915: Fixup userptr mmu notifier registration error handling (rev2) URL : https://patchwork.freedesktop.org/series/32136/ State : success == Summary == Series 32136v2 drm/i915: Fixup userptr mmu notifier registration

[Intel-gfx] ✓ Fi.CI.BAT: success for drm/i915: remove g4x lowfreq_avail and has_pipe_cxsr

== Series Details == Series: drm/i915: remove g4x lowfreq_avail and has_pipe_cxsr URL : https://patchwork.freedesktop.org/series/32134/ State : success == Summary == Series 32134v1 drm/i915: remove g4x lowfreq_avail and has_pipe_cxsr

Re: [Intel-gfx] [PATCH] drm/atomic-helper: check that drivers call drm_crtc_vblank_off

On Tue, Oct 17, 2017 at 06:32:52PM +0300, Ville Syrjälä wrote: > On Tue, Oct 17, 2017 at 05:27:14PM +0200, Daniel Vetter wrote: > > At least when they have vblank support they need to call this, or the > > vblank core will happily call into their crtc->enable_vblank callback > > even when the crtc

[Intel-gfx] [PATCH i-g-t] lib/i915: Move submission related helpers to lib/i915/gem_submission

Since I accidentally broke the build for some, by putting the pretty printer for submission inside ifdef HAVE_PROCPS, it's time to move the whole thing into lib/i915 while fixing this mistake. Let's also rename the pretty printer and add a doc to it as well as the section. Fixes: f6dfe556659f

[Intel-gfx] ✓ Fi.CI.IGT: success for drm/i915: Use a mask when applying WaProgramL3SqcReg1Default

== Series Details == Series: drm/i915: Use a mask when applying WaProgramL3SqcReg1Default URL : https://patchwork.freedesktop.org/series/32158/ State : success == Summary == Test kms_setmode: Subgroup basic: fail -> PASS (shard-hsw) fdo#99912 fdo#99912

[Intel-gfx] ✓ Fi.CI.BAT: success for GuC Interrupts/Log updates

== Series Details == Series: GuC Interrupts/Log updates URL : https://patchwork.freedesktop.org/series/32179/ State : success == Summary == Series 32179v1 GuC Interrupts/Log updates https://patchwork.freedesktop.org/api/1.0/series/32179/revisions/1/mbox/ Test chamelium: Subgroup

Re: [Intel-gfx] [PATCH v7 1/4] drm/i915 : Unifying seq_puts messages for feature support

On Wed, 18 Oct 2017 00:50:46 +0200, Sujaritha Sundaresan wrote: Unifying the various seq_puts messages in debugfs to the simplest one for feature support. v2: Clarifying the commit message (Anusha) v3: Re-factoring code as per review (Michal) v4: Rebase

[Intel-gfx] [PATCH 00/11] GuC Interrupts/Log updates

This series addresses following features/fixes: 1. Restructuring to control GuC interrupts from guc.c functions 2. Making GuC interrupts enable/disable reference based and tying up with Logging at all places. 3. Handle suspend/resume/reset for GuC interrupts. 4. Logging fixes about RPM wakeref

[Intel-gfx] [PATCH 07/11] drm/i915/guc: Grab RPM wakelock while disabling GuC interrupts

Disabling GuC interrupts involves access to GuC IRQ control registers hence ensure device is RPM awake. v2: Add comment about need to synchronize flush work and log runtime destroy Signed-off-by: Sagar Arun Kamble Cc: Michal Wajdeczko

[Intel-gfx] [PATCH 05/11] drm/i915/guc: Make GuC log related functions depend only on log level

With guc_log_level parameter sanitized we can remove the GuC submission checks from flush_guc_logs and i915_guc_log_register/unregister and intel_uc_fini_hw. It is important to note that GuC log runtime/relay channel data has to be freed during driver unregister. Freeing of that data can't be

[Intel-gfx] [PATCH 09/11] drm/i915/guc: Fix GuC interrupts disabling with Logging

With guc_log_unregister disabling runtime logging and interrupts there is no need to disable interrupts during uc_fini_hw hence it is removed. With GuC CT buffer mechanism, interrupt disabling can be added at a point where CT mechanism ceases. Signed-off-by: Sagar Arun Kamble

[Intel-gfx] [PATCH 03/11] drm/i915/guc: Pass intel_guc struct parameter to GuC interrupts functions

GuC interrupts handling functions are GuC specific functions hence update the parameter from dev_priv to intel_guc struct. Signed-off-by: Sagar Arun Kamble Cc: Michal Wajdeczko Cc: Daniele Ceraolo Spurio Cc:

[Intel-gfx] [PATCH 10/11] drm/i915/guc: Skip interrupt enabling if logging is already enabled

To balance the GuC interrupt references we don't allow enabling interrupts If logging was enabled earlier with different verbosity. We allow request to change log parameters to be sent to GuC though as user may want to just update the verbosity level at runtime. Signed-off-by: Sagar Arun Kamble

[Intel-gfx] [PATCH 06/11] drm/i915/guc: Only release GuC log object during submission_fini

GuC log runtime/relay channel data is released during i915 unregister, So only GuC log vma needs to be released during submission_fini. Signed-off-by: Sagar Arun Kamble Cc: Michal Wajdeczko Cc: Daniele Ceraolo Spurio

[Intel-gfx] [PATCH 02/11] drm/i915/guc: Move GuC interrupts related functions from i915_irq.c to intel_guc.c

GuC interrupts handling is core GuC functionality. Better to keep it with other core functions in intel_guc.c. Since they are used from uC functions, GuC logging, i915 irq handling keeping them grouped in intel_guc.c instead of intel_uc.c. Suggested-by: Michal Wajdeczko

[Intel-gfx] [PATCH 04/11] drm/i915/guc: Sanitize module parameter guc_log_level

Parameter guc_log_level needs to be sanitized based on GuC support and enable_guc_loading parameter since it depends on them like enable_guc_submission. This will make GuC logging paths independent of enable_guc_submission parameter in further patches. v2: Added documentation to intel_guc_log.c

[Intel-gfx] [PATCH 08/11] drm/i915/guc: Add client support to enable/disable GuC interrupts

This patch adds support to enable/disable GuC interrupts for different features without impacting others needs. Currently GuC log capture and CT buffer receive mechanisms use the GuC interrupts. GuC interrupts are currently enabled and disabled in different Logging scenarios all gated by log

[Intel-gfx] [PATCH 01/11] drm/i915: Export low level PM IRQ functions to control from GuC functions

In order to separate GuC IRQ handling functions from i915_irq.c we need to export the low level pm irq handlers. Export pm_iir, reset_pm_iir and enable/disable_pm_irq functions. Suggested-by: Michal Wajdeczko Signed-off-by: Sagar Arun Kamble

[Intel-gfx] [PATCH 11/11] drm/i915/guc: Restore GuC interrupts across suspend/reset if enabled

In order to override the disable/enable control of GuC interrupts during suspend/reset cycle we are creating two new functions suspend/restore guc_interrupts which check if interrupts were enabled and disable them on suspend and enable them on resume. They are used to restore interrupts across

Re: [Intel-gfx] ✓ Fi.CI.IGT: success for igt/gem_eio: Check hang/eio recovery during suspend

On Thu, Oct 12, 2017 at 09:20:46AM +0100, Chris Wilson wrote: > Quoting Arkadiusz Hiler (2017-10-12 09:15:33) > > On Sat, Oct 07, 2017 at 03:38:09AM +, Patchwork wrote: > > > == Series Details == > > > > > > Series: igt/gem_eio: Check hang/eio recovery during suspend > > > URL :

<    1   2