== Series Details ==
Series: drm/i915/glk: Add Quirk for GLK NUC HDMI port issues.
URL : https://patchwork.freedesktop.org/series/6/
State : success
== Summary ==
= CI Bug Log - changes from CI_DRM_4291_full -> Patchwork_9237_full =
== Summary - WARNING ==
Minor unknown changes coming
== Series Details ==
Series: drm/i915: inline skl_copy_ddb_for_pipe() to its only caller
URL : https://patchwork.freedesktop.org/series/5/
State : success
== Summary ==
= CI Bug Log - changes from CI_DRM_4291_full -> Patchwork_9236_full =
== Summary - WARNING ==
Minor unknown changes
== Series Details ==
Series: drm/i915/glk: Add Quirk for GLK NUC HDMI port issues.
URL : https://patchwork.freedesktop.org/series/6/
State : success
== Summary ==
= CI Bug Log - changes from CI_DRM_4291 -> Patchwork_9237 =
== Summary - WARNING ==
Minor unknown changes coming with
On 5/11/2018 2:33 PM, Vidya Srinivas wrote:
From: Chandra Konduru
This patch adds NV12 as supported format
to intel_framebuffer_init and performs various checks.
v2:
-Fix an issue in checks added (Chandra Konduru)
v3: rebased (me)
v4: Review comments by Ville addressed
Added platform check
Quoting Paulo Zanoni (2018-06-08 00:07:00)
> static void
> skl_print_wm_changes(const struct drm_atomic_state *state)
> {
> @@ -5381,7 +5370,10 @@ static void skl_initial_wm(struct intel_atomic_state
> *state,
> if (cstate->base.active_changed)
>
== Series Details ==
Series: drm/i915/glk: Add Quirk for GLK NUC HDMI port issues.
URL : https://patchwork.freedesktop.org/series/6/
State : warning
== Summary ==
$ dim sparse origin/drm-tip
Commit: drm/i915/glk: Add Quirk for GLK NUC HDMI port issues.
== Series Details ==
Series: drm/i915/glk: Add Quirk for GLK NUC HDMI port issues.
URL : https://patchwork.freedesktop.org/series/6/
State : warning
== Summary ==
$ dim checkpatch origin/drm-tip
60f585b542bc drm/i915/glk: Add Quirk for GLK NUC HDMI port issues.
-:28: CHECK:SPACING: spaces
== Series Details ==
Series: drm/i915: inline skl_copy_ddb_for_pipe() to its only caller
URL : https://patchwork.freedesktop.org/series/5/
State : success
== Summary ==
= CI Bug Log - changes from CI_DRM_4291 -> Patchwork_9236 =
== Summary - WARNING ==
Minor unknown changes coming
From: Clint Taylor
On GLK NUC platforms the HDMI retiming buffer needs additional disabled
time to correctly sync to a faster incoming signal.
When measured on a scope the highspeed lines of the HDMI clock turn off
for ~400uS during a normal resolution change. The HDMI retimer on the
GLK NUC
While things may have been different before, right now the function is
very simple and has a single caller. IMHO any possible benefits from
an abstraction here are gone and not worth the price of the current
indirection while reading the code.
Cc: Mahesh Kumar
Signed-off-by: Paulo Zanoni
---
== Series Details ==
Series: drm/i915/audio: Add 801Mhz clock entries to dp_aud_n_m table
URL : https://patchwork.freedesktop.org/series/44435/
State : success
== Summary ==
= CI Bug Log - changes from CI_DRM_4290_full -> Patchwork_9235_full =
== Summary - WARNING ==
Minor unknown changes
== Series Details ==
Series: series starting with [1/2] drm/i915: Cancel reset preparations on
failed resets (rev2)
URL : https://patchwork.freedesktop.org/series/44288/
State : success
== Summary ==
= CI Bug Log - changes from CI_DRM_4290_full -> Patchwork_9233_full =
== Summary - SUCCESS
== Series Details ==
Series: series starting with [CI,1/3] drm/i915: Prepare for non-object vma
(rev2)
URL : https://patchwork.freedesktop.org/series/44417/
State : success
== Summary ==
= CI Bug Log - changes from CI_DRM_4289_full -> Patchwork_9232_full =
== Summary - WARNING ==
Minor
Recently we discovered that we have a race between swapping and
suspend in our resume path (we might be trying to page in an object
after disabling the block devices). Let's try to exercise that by
exhausting all of system memory before suspend.
v2: Explicitly share the large memory area on
== Series Details ==
Series: drm/i915/icl: Add warn about unsupported CDCLK rates
URL : https://patchwork.freedesktop.org/series/44421/
State : success
== Summary ==
= CI Bug Log - changes from CI_DRM_4289_full -> Patchwork_9231_full =
== Summary - WARNING ==
Minor unknown changes coming
== Series Details ==
Series: drm/i915/audio: Add 801Mhz clock entries to dp_aud_n_m table
URL : https://patchwork.freedesktop.org/series/44435/
State : success
== Summary ==
= CI Bug Log - changes from CI_DRM_4290 -> Patchwork_9235 =
== Summary - SUCCESS ==
No regressions found.
From: "Sripada, Radhakrishna"
Expand the Maud/Naud table according to DP 1.4 spec to include entries for
810 MHz clock. This is required for audio to work with HBR3.
Cc: Dhinakaran Pandiyan
Cc: Jani Nikula
Signed-off-by: Radhakrishna Sripada
---
drivers/gpu/drm/i915/intel_audio.c | 10
== Series Details ==
Series: drm/i915: Mark the GPU as wedged without error on fault injection
URL : https://patchwork.freedesktop.org/series/44411/
State : success
== Summary ==
= CI Bug Log - changes from CI_DRM_4289_full -> Patchwork_9229_full =
== Summary - WARNING ==
Minor unknown
== Series Details ==
Series: series starting with [1/6] drm/i915: Store first production revid into
device info
URL : https://patchwork.freedesktop.org/series/44429/
State : failure
== Summary ==
= CI Bug Log - changes from CI_DRM_4290 -> Patchwork_9234 =
== Summary - FAILURE ==
Serious
Hi Chris,
Thank you for the patch! Perhaps something to improve:
[auto build test WARNING on drm-intel/for-linux-next]
[also build test WARNING on v4.17 next-20180607]
[if your patch is applied to the wrong git tree, please drop us a note to help
improve the system]
url:
https://github.com
== Series Details ==
Series: series starting with [1/6] drm/i915: Store first production revid into
device info
URL : https://patchwork.freedesktop.org/series/44429/
State : warning
== Summary ==
$ dim sparse origin/drm-tip
Commit: drm/i915: Store first production revid into device info
== Series Details ==
Series: series starting with [1/6] drm/i915: Store first production revid into
device info
URL : https://patchwork.freedesktop.org/series/44429/
State : warning
== Summary ==
$ dim checkpatch origin/drm-tip
ec78e5bf259c drm/i915: Store first production revid into device
== Series Details ==
Series: Queued/runnable/running engine stats (rev14)
URL : https://patchwork.freedesktop.org/series/36926/
State : success
== Summary ==
= CI Bug Log - changes from CI_DRM_4289_full -> Patchwork_9228_full =
== Summary - WARNING ==
Minor unknown changes coming with
== Series Details ==
Series: series starting with [1/2] drm/i915: Cancel reset preparations on
failed resets (rev2)
URL : https://patchwork.freedesktop.org/series/44288/
State : success
== Summary ==
= CI Bug Log - changes from CI_DRM_4290 -> Patchwork_9233 =
== Summary - WARNING ==
Make a new named define for each platform which signifies
the first production revid that platform has. Placing
it along the revision ids adds clarity.
Signed-off-by: Mika Kuoppala
---
drivers/gpu/drm/i915/i915_pci.c | 8
drivers/gpu/drm/i915/intel_chipset.h | 3 +++
2 files
Carve out chipset definitions into new intel_chipset.h
Suggested-by: Chris Wilson
Cc: Chris Wilson
Signed-off-by: Mika Kuoppala
---
drivers/gpu/drm/i915/i915_drv.h | 200 +-
drivers/gpu/drm/i915/intel_chipset.h | 208 +++
2 files changed,
If we are doing revision checks against a preproduction
range, when there is already a product, it is a sign
that there is code to be removed.
Signed-off-by: Mika Kuoppala
---
drivers/gpu/drm/i915/intel_chipset.h | 30 +---
1 file changed, 23 insertions(+), 7
We don't need to have distinct flag for alpha quality if
we agree that setting the first production revid to be the
epoch for stepping out from alpha quality on that platform.
Cc: Joonas Lahtinen
Cc: Chris Wilson
Cc: Tomi Sarvela
Cc: Jani Nikula
Signed-off-by: Mika Kuoppala
---
Store first known production revid into the device info.
By doing this we then can do preliminary hardware
checks in code and also prepare to complain automatically
on outdated workarounds etc.
Uninitialized (zero) product revision id means that
there are no known preliminary hardware for this
We don't need kbl preprod workarounds anymore.
Signed-off-by: Mika Kuoppala
---
drivers/gpu/drm/i915/intel_lrc.c | 12
drivers/gpu/drm/i915/intel_workarounds.c | 5 -
2 files changed, 17 deletions(-)
diff --git a/drivers/gpu/drm/i915/intel_lrc.c
There is a problem with kbl up to rev E0 where a heavy
memory/fabric traffic from adjacent engine(s) can cause an engine
reset to fail. This traffic can be from normal memory accesses
or it can be from heavy polling on a semaphore wait.
For engine hogging causing a fail, we already fallback to
On Thu, Jun 07, 2018 at 08:01:14PM +0300, Ville Syrjälä wrote:
> On Thu, Jun 07, 2018 at 07:13:53PM +0300, Imre Deak wrote:
> > While checking workarounds related to the CDCLK PLL, I noticed that the
> > DMC firmware bits for WA#1183 are missing for SKL. After that I
> > clarified with HW people
== Series Details ==
Series: series starting with [CI,1/3] drm/i915: Prepare for non-object vma
(rev2)
URL : https://patchwork.freedesktop.org/series/44417/
State : success
== Summary ==
= CI Bug Log - changes from CI_DRM_4289 -> Patchwork_9232 =
== Summary - WARNING ==
Minor unknown
On Thu, Jun 07, 2018 at 07:13:53PM +0300, Imre Deak wrote:
> While checking workarounds related to the CDCLK PLL, I noticed that the
> DMC firmware bits for WA#1183 are missing for SKL. After that I
> clarified with HW people that it's not needed on SKL, since it doesn't
> support eDP1.4 which
Quoting Chris Wilson (2018-05-29 10:55:28)
> If we find a task that has already been selected for reaping, consider
> that it may still free some memory. Currently, we skip such tasks
> believing that we've already extracted as memory free pages as possible
> from before hitting a livelock. In
== Series Details ==
Series: series starting with [CI,1/3] drm/i915: Prepare for non-object vma
(rev2)
URL : https://patchwork.freedesktop.org/series/44417/
State : warning
== Summary ==
$ dim sparse origin/drm-tip
Commit: drm/i915: Prepare for non-object vma
Okay!
Commit: drm/i915:
== Series Details ==
Series: drm/i915/icl: Add warn about unsupported CDCLK rates
URL : https://patchwork.freedesktop.org/series/44421/
State : success
== Summary ==
= CI Bug Log - changes from CI_DRM_4289 -> Patchwork_9231 =
== Summary - WARNING ==
Minor unknown changes coming with
In the next patch, we will subclass the gen6 hw_ppgtt. In order, for the
two different generations of hw ppgtt stucts to be of different size,
push the allocation down to the constructor.
Signed-off-by: Chris Wilson
Cc: Joonas Lahtinen
Cc: Mika Kuoppala
Cc: Matthew Auld
Reviewed-by: Joonas
== Series Details ==
Series: series starting with [CI,1/3] drm/i915: Prepare for non-object vma
URL : https://patchwork.freedesktop.org/series/44417/
State : failure
== Summary ==
= CI Bug Log - changes from CI_DRM_4289 -> Patchwork_9230 =
== Summary - FAILURE ==
Serious unknown changes
While checking workarounds related to the CDCLK PLL, I noticed that the
DMC firmware bits for WA#1183 are missing for SKL. After that I
clarified with HW people that it's not needed on SKL, since it doesn't
support eDP1.4 which would be the only thing requiring the problematic
CDCLK clock rates.
== Series Details ==
Series: series starting with [CI,1/3] drm/i915: Prepare for non-object vma
URL : https://patchwork.freedesktop.org/series/44417/
State : warning
== Summary ==
$ dim sparse origin/drm-tip
Commit: drm/i915: Prepare for non-object vma
Okay!
Commit: drm/i915: Decouple vma
In order to allow ourselves to use VMA to wrap other entities other than
GEM objects, we need to allow for the vma->obj backpointer to be NULL.
In most cases, we know we are operating on a GEM object and its vma, but
we need the core code (such as i915_vma_pin/insert/bind/unbind) to work
In the next patch, we will subclass the gen6 hw_ppgtt. In order, for the
two different generations of hw ppgtt stucts to be of different size,
push the allocation down to the constructor.
Signed-off-by: Chris Wilson
Cc: Joonas Lahtinen
Cc: Mika Kuoppala
Cc: Matthew Auld
Reviewed-by: Joonas
To allow for future non-object backed vma, we need to be able to
specialise the callbacks for binding, et al, the vma. For example,
instead of calling vma->vm->bind_vma(), we now call
vma->ops->bind_vma(). This gives us the opportunity to later override the
operation for a custom vma.
v2: flip
Quoting Chris Wilson (2018-06-07 16:27:22)
> Batches are contained in their position within the GTT by the kernel,
constrained
> and if they are in an invalid poistion will be unbound and rebound
position
> before execution. In our test setup, we therefore need to place the
> batch into a valid
Batches are contained in their position within the GTT by the kernel,
and if they are in an invalid poistion will be unbound and rebound
before execution. In our test setup, we therefore need to place the
batch into a valid poistion within the GTT before we fill the ring with
busyspinners.
On Wed, 06 Jun 2018, Chris Wilson wrote:
> Quoting Mika Kuoppala (2018-06-06 15:18:24)
>> Update preproduction steppings detection so that
>> we get the warning and taint correctly for more
>> recent platforms.
>
> Before crying foul, I suggest we check INTEL_INFO()->is_alpha_support.
> We don't
Quoting Tvrtko Ursulin (2018-06-07 14:25:28)
> From: Tvrtko Ursulin
>
> We add a PMU counter to expose the number of requests currently executing
> on the GPU.
>
> This is useful to analyze the overall load of the system.
>
> v2:
> * Rebase.
> * Drop floating point constant. (Chris Wilson)
>
== Series Details ==
Series: drm/i915: Mark the GPU as wedged without error on fault injection
URL : https://patchwork.freedesktop.org/series/44411/
State : success
== Summary ==
= CI Bug Log - changes from CI_DRM_4289 -> Patchwork_9229 =
== Summary - WARNING ==
Minor unknown changes
Quoting Patchwork (2018-06-07 15:12:34)
> == Series Details ==
>
> Series: series starting with [01/20] drm/i915: Apply batch location
> restrictions before pinning
> URL : https://patchwork.freedesktop.org/series/44393/
> State : failure
>
> == Summary ==
>
> = CI Bug Log - changes from
== Series Details ==
Series: series starting with [01/20] drm/i915: Apply batch location
restrictions before pinning
URL : https://patchwork.freedesktop.org/series/44393/
State : failure
== Summary ==
= CI Bug Log - changes from CI_DRM_4289_full -> Patchwork_9227_full =
== Summary - FAILURE
== Series Details ==
Series: Queued/runnable/running engine stats (rev14)
URL : https://patchwork.freedesktop.org/series/36926/
State : success
== Summary ==
= CI Bug Log - changes from CI_DRM_4289 -> Patchwork_9228 =
== Summary - WARNING ==
Minor unknown changes coming with
Quoting Chris Wilson (2018-06-07 14:45:58)
> If we have been instructed (by CI) to inject a fault to load the module
> with a wedged GPU, do so quietly less we upset CI.
s/less/lest/
-Chris
___
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
If we have been instructed (by CI) to inject a fault to load the module
with a wedged GPU, do so quietly less we upset CI.
Signed-off-by: Chris Wilson
Cc: Michał Winiarski
Cc: Michal Wajdeczko
---
drivers/gpu/drm/i915/i915_gem.c | 2 ++
1 file changed, 2 insertions(+)
diff --git
From: Tvrtko Ursulin
We add a PMU counter to expose the number of requests currently executing
on the GPU.
This is useful to analyze the overall load of the system.
v2:
* Rebase.
* Drop floating point constant. (Chris Wilson)
v3:
* Change scale to 1024 for faster arithmetics. (Chris
From: Tvrtko Ursulin
We add a PMU counter to expose the number of requests with resolved
dependencies waiting for a slot on the GPU to run.
This is useful to analyze the overall load of the system.
v2: Don't limit to gen8+.
v3:
* Rebase for dynamic sysfs.
* Drop currently executing
From: Tvrtko Ursulin
We add a PMU counter to expose the number of requests which have been
submitted from userspace but are not yet runnable due dependencies and
unsignaled fences.
This is useful to analyze the overall load of the system.
v2:
* Rebase for name change and re-order.
* Drop
The current method of checking for a failed module load is flawed, as we
only report the error on probing it is not being reported back by
modprobe. So we have to dig inside the module_parameters while the
module is still loaded to discover the error.
v2: Expect i915.inject_load_failure to be
The current method of checking for a failed module load is flawed, as we
only report the error on probing it is not being reported back by
modprobe. So we have to dig inside the module_parameters while the
module is still loaded to discover the error.
v2: Expect i915.inject_load_failure to be
/linux/commits/Chris-Wilson/drm-i915-Mark-i915-inject_load_failure-as-being-hit/20180607-174849
base: git://anongit.freedesktop.org/drm-intel for-linux-next
config: x86_64-randconfig-x018-201822 (attached as .config)
compiler: gcc-7 (Debian 7.3.0-16) 7.3.0
reproduce:
# save the attached
/linux/commits/Chris-Wilson/drm-i915-Mark-i915-inject_load_failure-as-being-hit/20180607-174849
base: git://anongit.freedesktop.org/drm-intel for-linux-next
config: i386-randconfig-n0-201822 (attached as .config)
compiler: gcc-7 (Debian 7.3.0-16) 7.3.0
reproduce:
# save the attached
The current method of checking for a failed module load is flawed, as we
only report the error on probing it is not being reported back by
modprobe. So we have to dig inside the module_parameters while the
module is still loaded to discover the error.
v2: Expect i915.inject_load_failure to be
== Series Details ==
Series: series starting with [1/2] drm/i915: Mark order of mmio to CCID/PP_DIR
with switch_context()
URL : https://patchwork.freedesktop.org/series/44392/
State : success
== Summary ==
= CI Bug Log - changes from CI_DRM_4287_full -> Patchwork_9226_full =
== Summary -
== Series Details ==
Series: series starting with [01/20] drm/i915: Apply batch location
restrictions before pinning
URL : https://patchwork.freedesktop.org/series/44393/
State : success
== Summary ==
= CI Bug Log - changes from CI_DRM_4289 -> Patchwork_9227 =
== Summary - WARNING ==
Quoting Matthew Auld (2018-06-07 11:17:14)
> On 7 June 2018 at 10:58, Chris Wilson wrote:
> > Pull the empty stubs together into the top level gen6_ppgtt_create, and
> > tear each one down on error in proper onion order (rather than use
> > Joonas' pet hate of calling the cleanup function in
== Series Details ==
Series: series starting with [01/20] drm/i915: Apply batch location
restrictions before pinning
URL : https://patchwork.freedesktop.org/series/44393/
State : warning
== Summary ==
$ dim sparse origin/drm-tip
Commit: drm/i915: Apply batch location restrictions before
On 7 June 2018 at 10:58, Chris Wilson wrote:
> Pull the empty stubs together into the top level gen6_ppgtt_create, and
> tear each one down on error in proper onion order (rather than use
> Joonas' pet hate of calling the cleanup function in indeterminable
> state).
>
> Signed-off-by: Chris
== Series Details ==
Series: series starting with [01/20] drm/i915: Apply batch location
restrictions before pinning
URL : https://patchwork.freedesktop.org/series/44393/
State : warning
== Summary ==
$ dim checkpatch origin/drm-tip
97656113bc83 drm/i915: Apply batch location restrictions
Hi All,
We are running v4.14.47 kernel and recently in one of our test cycle
we saw the below trace. I know this is not the usual way to raise a
BUG report, but since this was seen only once in one of the automated
test cycle so I donot have anything else apart from this trace.
Is this a known
As the most frequent PTE encoding is for the scratch page, cache it upon
creation.
Signed-off-by: Chris Wilson
Cc: Joonas Lahtinen
Cc: Mika Kuoppala
Cc: Matthew Auld
Reviewed-by: Matthew Auld
---
drivers/gpu/drm/i915/i915_gem_gtt.c | 20 ++--
If we will completely overwrite the PT with PTEs for the object, we can
forgo filling it with scratch entries.
References: 14826673247e ("drm/i915: Only initialize partially filled
pagetables")
Signed-off-by: Chris Wilson
Cc: Joonas Lahtinen
Cc: Mika Kuoppala
Cc: Matthew Auld
Reviewed-by:
We can stop asserting using WARN_ON as given sufficient CI coverage, we
can rely on using GEM_BUG_ON() to catch problems before merging.
Signed-off-by: Chris Wilson
Cc: Joonas Lahtinen
Cc: Mika Kuoppala
Cc: Matthew Auld
Reviewed-by: Matthew Auld
---
drivers/gpu/drm/i915/i915_gem_gtt.c | 2
As we cannot reliably change used page tables while the context is
active, the earliest opportunity we have to recover excess pages is when
the context becomes idle. So whenever we unbind the context (it must be
idle, and indeed being evicted) free the unused ptes.
Signed-off-by: Chris Wilson
If we know that the user cannot access the GGTT, by virtue of having a
segregated memory area, we can skip clearing the unused entries as they
cannot be accessed.
Signed-off-by: Chris Wilson
Cc: Joonas Lahtinen
Cc: Mika Kuoppala
Cc: Matthew Auld
---
drivers/gpu/drm/i915/i915_gem_gtt.c | 4
To allow ourselves to use a first class vma for the aliasing_ppgtt page
directory, we have to reorder the shutdown on module unload to remove
and unpin the aliasing_ppgtt before complaining about any objects left
in the GGTT.
Signed-off-by: Chris Wilson
Cc: Joonas Lahtinen
Cc: Mika Kuoppala
The legacy gen6 ppgtt needs a little more hand holding than gen8+, and
so requires a larger structure. As I intend to make this slightly more
complicated in the future, separate the gen6 from the core gen8 hw
struct by subclassing. This patch moves the gen6 only features out to
gen6_hw_ppgtt and
In the next patch, we will subclass the gen6 hw_ppgtt. In order, for the
two different generations of hw ppgtt stucts to be of different size,
push the allocation down to the constructor.
Signed-off-by: Chris Wilson
Cc: Joonas Lahtinen
Cc: Mika Kuoppala
Cc: Matthew Auld
Reviewed-by: Joonas
To allow for future non-object backed vma, we need to be able to
specialise the callbacks for binding, et al, the vma. For example,
instead of calling vma->vm->bind_vma(), we now call
vma->ops->bind_vma(). This gives us the opportunity to later override the
operation for a custom vma.
v2: flip
When we update the gen6 ppgtt page directories, we do so by writing the
new address into a reserved slot in the GGTT. It appears that when the
GPU reads that entry from the gsm, it uses its small cache and that we
need to invalidate that cache after writing. We don't see an issue
currently as we
In order to allow ourselves to use VMA to wrap other entities other than
GEM objects, we need to allow for the vma->obj backpointer to be NULL.
In most cases, we know we are operating on a GEM object and its vma, but
we need the core code (such as i915_vma_pin/insert/bind/unbind) to work
Pull the empty stubs together into the top level gen6_ppgtt_create, and
tear each one down on error in proper onion order (rather than use
Joonas' pet hate of calling the cleanup function in indeterminable
state).
Signed-off-by: Chris Wilson
Cc: Joonas Lahtinen
Cc: Mika Kuoppala
Cc: Matthew
As we were only supporting aliasing_ppgtt on gen7 for some time, we
saved a few checks by preallocating the page directories on creation.
However, since we need 2MiB of page directories for each ppgtt, to
support arbitrary numbers of user contexts, we need to be more prudent
in our allocations,
In order to be able to evict the gen6 ppgtt, we have to unpin it at some
point. We can simply use our context activity tracking to know when the
ppgtt is no longer in use by hardware, and so only keep it pinned while
being used a request.
For the kernel_context (and thus aliasing_ppgtt), it
Currently all page directories are bound at creation using an
unevictable node in the GGTT. This severely limits us as we cannot
remove any inactive ppgtt for new contexts, or under aperture pressure.
To fix this we need to make the page directory into a first class and
unbindable vma. Hence, the
We special case the position of the batch within the GTT to prevent
negative self-relocation deltas from underflowing. However, that
restriction is being applied after a trial pin of the batch in its
current position. Thus we are not rejecting an invalid location if the
batch has been before,
An issue encountered with switching mm on gen7 is that the GPU likes to
hang (with the VS unit busy) when told to force restore the current
context. We can simply workaround this by substituting the
MI_FORCE_RESTORE flag with a round-trip through the kernel_context,
forcing the context to be saved
When using CS commands, PP_DIR is not sampled until the context is
loaded, but when doing manual mmio after reset we load the context
before the mm. Let's switch this over for consistency.
Signed-off-by: Chris Wilson
Cc: Joonas Lahtinen
Cc: Mika Kuoppala
Cc: Matthew Auld
---
If we interrupt the context switch and unwind, we leave the to_mm
believing that we have cleared the dirty bit for this engine (but the
LRI will never take place). Just in case we immediately reload the same
context, mark this engine as dirty so that we force the LRI to reload
the PD.
Let's see if we have all the kinks worked out and full-ppgtt now works
reliably on gen7 (Ivybridge, Valleyview/Baytrail and Haswell). If we can
let userspace have full control over their own ppgtt, it makes softpinning
far more effective, in turn making GPU dispatch far more efficient and
more
In this edition, I think I have solved the mystery mesa hang; having
identified a suspect in MI_SET_CONTEXT | FORCE_RESTORE. But the hang on
reset with a dirty PD is still there, and I'm still fighting byt which
fails to switch mm once every few hours.
-Chris
Hi,
On 06-06-18 20:12, Rodrigo Vivi wrote:
On Wed, Jun 06, 2018 at 02:12:36PM +0200, Hans de Goede wrote:
Hi,
On 06-06-18 12:43, Maarten Lankhorst wrote:
Op 06-06-18 om 11:54 schreef Hans de Goede:
Hi,
On 06-06-18 11:36, Maarten Lankhorst wrote:
Op 06-06-18 om 11:09 schreef Hans de Goede:
Quoting Mika Kuoppala (2018-06-05 19:03:57)
> There is a problem with kbl up to rev E0 where a heavy
> memory/fabric traffic from adjacent engine(s) can cause an engine
> reset to fail. This traffic can be from normal memory accesses
> or it can be from heavy polling on a semaphore wait.
>
> For
== Series Details ==
Series: series starting with [1/2] drm/i915: Mark order of mmio to CCID/PP_DIR
with switch_context()
URL : https://patchwork.freedesktop.org/series/44392/
State : success
== Summary ==
= CI Bug Log - changes from CI_DRM_4287 -> Patchwork_9226 =
== Summary - SUCCESS ==
== Series Details ==
Series: series starting with [1/2] drm/i915: Mark order of mmio to CCID/PP_DIR
with switch_context()
URL : https://patchwork.freedesktop.org/series/44392/
State : warning
== Summary ==
$ dim checkpatch origin/drm-tip
b79de1db57c5 drm/i915: Mark order of mmio to
Quoting Matthew Auld (2018-06-06 23:39:07)
> On 6 June 2018 at 22:45, Chris Wilson wrote:
> > In preparation for vm_fault_t becoming a distinct type, convert the
> > fault handler (i915_gem_fault()) over to the new interface.
> >
> > Based on a patch by Souptick Joarder
> >
> > References:
Quoting Michał Winiarski (2018-06-07 08:39:39)
> On Wed, Jun 06, 2018 at 03:54:41PM +0100, Chris Wilson wrote:
> > As part of our GEM initialisation now, we send a request to the hardware
> > in order to record the initial GPU state. This coupled with deferred
> > idle workers, makes aborting on
On Wed, Jun 06, 2018 at 03:54:41PM +0100, Chris Wilson wrote:
> As part of our GEM initialisation now, we send a request to the hardware
> in order to record the initial GPU state. This coupled with deferred
> idle workers, makes aborting on error tricky. We already have the
> mechanism in place
Mark order? I have no idea. Let me go get some coffee.
Quoting Chris Wilson (2018-06-07 08:30:24)
> When using CS commands, PP_DIR is not sampled until the context is
> loaded, but when doing manual mmio after reset we load the context
> before the mm. Let's switch this over for consistency.
>
>
An issue encountered with switching mm on gen7 is that the GPU likes to
hang (with the VS unit busy) when told to force restore the current
context. We can simply workaround this by substituting the
MI_FORCE_RESTORE flag with a round-trip through the kernel_context,
forcing the context to be saved
When using CS commands, PP_DIR is not sampled until the context is
loaded, but when doing manual mmio after reset we load the context
before the mm. Let's switch this over for consistency.
Signed-off-by: Chris Wilson
Cc: Joonas Lahtinen
Cc: Mika Kuoppala
Cc: Matthew Auld
---
100 matches
Mail list logo
