Quoting Jani Nikula (2020-05-18 17:47:47)
> This is the first 3 patches of [1], because apparently patch 4 breaks
> the world. I've yet to pinpoint the issue, but these could move forward
> in the meanwhile.
It's not you, it's igt_params.c
-Chris
___
Int
Rather than going back and forth between the rb_node entry and the
virtual_engine type, store the ve local and reuse it. As the
container_of conversion from rb_node to virtual_engine requires a
variable offset, performing that conversion just once shaves off a bit
of code.
v2: Keep a single virtua
When we introduced the saturated workload detection to tell us to back
off from semaphore usage [semaphores have a noticeable impact on
contended bus cycles with the CPU for some heavy workloads], we first
introduced it as a per-context tracker. This allows individual contexts
to try and optimise t
Since we temporarily disable the heartbeat and restore back to the
default value, we can use the stored defaults on the engine and avoid
using a local.
Signed-off-by: Chris Wilson
---
drivers/gpu/drm/i915/gt/selftest_hangcheck.c | 25 +++
drivers/gpu/drm/i915/gt/selftest_lrc.c | 67 +++
If we decide to timeslice out the current virtual request, we will
unsubmit it while it is still busy (ve->context.inflight == sibling[0]).
If the virtual tasklet and then the other sibling tasklets run before we
completely schedule out the active virtual request for the preemption,
those other tas
Make sure that we can execute a virtual request on an already busy
engine, and conversely that we can execute a normal request if the
engines are already fully occupied by virtual requests.
Signed-off-by: Chris Wilson
---
drivers/gpu/drm/i915/gt/selftest_lrc.c | 200 -
1
When we look at i915_request_is_started() we must be careful in case we
are using a request that does not have the initial-breadcrumb and
instead the is-started is being compared against the end of the previous
request. This will make wait_for_submit() declare that a request has
been already submit
Having recognised that we do not change the sibling until we schedule
out, we can then defer the decision to resubmit the virtual engine from
the unwind of the active queue to scheduling out of the virtual context.
By keeping the unwind order intact on the local engine, we can preserve
data depend
We recorded the execlists->queue_priority_hint update for the inflight
request without kicking the tasklet. The next submitted request then
failed to be scheduled as it had a lower priority than the hint, leaving
the HW runnning with only the inflight request.
Fixes: 6cebcf746f3f ("drm/i915: Tweak
Once a virtual engine has been bound to a sibling, it will remain bound
until we finally schedule out the last active request. We can not rebind
the context to a new sibling while it is inflight as the context save
will conflict, hence we wait. As we cannot then use any other sibliing
while the con
If there are no internal levels and the user priority-shift is zero, we
can help the compiler eliminate some dead code:
Function old new delta
start_timeslice 169 154 -15
__execlists_submission_tasklet 4696
It was quite the oversight to only factor in the normal queue to decide
the timeslicing switch priority. By leaving out the next virtual request
from the priority decision, we would not timeslice the current engine if
there was an available virtual request.
Testcase: igt/gem_exec_balancer/sliced
F
Check for integer overflow in the priority chain, rather than against a
type-constricted max-priority check.
Signed-off-by: Chris Wilson
---
drivers/gpu/drm/i915/gt/selftest_lrc.c | 6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)
diff --git a/drivers/gpu/drm/i915/gt/selftest_lrc.c
b/
Hi all,
Today's linux-next merge of the tip tree got a conflict in:
drivers/gpu/drm/drm_dp_mst_topology.c
between commit:
a4292e52106b ("drm: Match drm_dp_send_clear_payload_id_table definition to
declaration")
from the drm-misc tree and commit:
53965dbe5396 ("drm: Make drm_dp_mst_dsc_
== Series Details ==
Series: drm/i915/mst: filter out the display mode exceed sink's capability
(rev3)
URL : https://patchwork.freedesktop.org/series/76095/
State : success
== Summary ==
CI Bug Log - changes from CI_DRM_8498_full -> Patchwork_17706_full
===
== Series Details ==
Series: drm/i915/ehl: Extend w/a 14010685332 to JSP/MCC
URL : https://patchwork.freedesktop.org/series/77382/
State : success
== Summary ==
CI Bug Log - changes from CI_DRM_8498_full -> Patchwork_17705_full
Summary
== Series Details ==
Series: drm/i915/mst: filter out the display mode exceed sink's capability
(rev3)
URL : https://patchwork.freedesktop.org/series/76095/
State : success
== Summary ==
CI Bug Log - changes from CI_DRM_8498 -> Patchwork_17706
=
So far, max dot clock rate for MST mode rely on physcial
bandwidth limitation. It would caused compatibility issue
if source display resolution exceed MST hub output ability.
For example, source DUT had DP 1.2 output capability.
And MST docking just support HDMI 1.4 spec. When a HDMI 2.0
monitor c
== Series Details ==
Series: mm/gup, drm/i915: refactor gup_fast, convert to pin_user_pages()
URL : https://patchwork.freedesktop.org/series/77381/
State : failure
== Summary ==
CI Bug Log - changes from CI_DRM_8498_full -> Patchwork_17704_full
=
== Series Details ==
Series: series starting with [1/2] drm/i915: move trace_i915_reg_rw() to a
separate file
URL : https://patchwork.freedesktop.org/series/77378/
State : success
== Summary ==
CI Bug Log - changes from CI_DRM_8498_full -> Patchwork_17703_full
== Series Details ==
Series: drm/i915/ehl: Extend w/a 14010685332 to JSP/MCC
URL : https://patchwork.freedesktop.org/series/77382/
State : success
== Summary ==
CI Bug Log - changes from CI_DRM_8498 -> Patchwork_17705
Summary
---
**S
== Series Details ==
Series: series starting with [1/4] drm/i915: Don't set queue-priority hint when
supressing the reschedule
URL : https://patchwork.freedesktop.org/series/77377/
State : success
== Summary ==
CI Bug Log - changes from CI_DRM_8498_full -> Patchwork_17702_full
===
== Series Details ==
Series: mm/gup, drm/i915: refactor gup_fast, convert to pin_user_pages()
URL : https://patchwork.freedesktop.org/series/77381/
State : success
== Summary ==
CI Bug Log - changes from CI_DRM_8498 -> Patchwork_17704
Summa
This is a permanent w/a for JSL/EHL.This is to be applied to the
PCH types on JSL/EHL ie JSP/MCC
Bspec: 52888
Signed-off-by: Swathi Dhanavanthri
---
drivers/gpu/drm/i915/i915_irq.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/drivers/gpu/drm/i915/i915_irq.c b/drivers/g
== Series Details ==
Series: drm/i915: Resolve device hotunplug issues
URL : https://patchwork.freedesktop.org/series/77372/
State : success
== Summary ==
CI Bug Log - changes from CI_DRM_8498_full -> Patchwork_17701_full
Summary
---
== Series Details ==
Series: mm/gup, drm/i915: refactor gup_fast, convert to pin_user_pages()
URL : https://patchwork.freedesktop.org/series/77381/
State : warning
== Summary ==
$ dim sparse --fast origin/drm-tip
Sparse version: v0.6.0
Fast mode used, each commit won't be checked separately.
-
== Series Details ==
Series: mm/gup, drm/i915: refactor gup_fast, convert to pin_user_pages()
URL : https://patchwork.freedesktop.org/series/77381/
State : warning
== Summary ==
$ dim checkpatch origin/drm-tip
5dd288837ad8 mm/gup: move __get_user_pages_fast() down a few lines in gup.c
-:111: C
== Series Details ==
Series: series starting with [1/2] drm/i915: move trace_i915_reg_rw() to a
separate file
URL : https://patchwork.freedesktop.org/series/77378/
State : success
== Summary ==
CI Bug Log - changes from CI_DRM_8498 -> Patchwork_17703
==
== Series Details ==
Series: series starting with [1/2] drm/i915: move trace_i915_reg_rw() to a
separate file
URL : https://patchwork.freedesktop.org/series/77378/
State : warning
== Summary ==
$ dim checkpatch origin/drm-tip
caee6db994c4 drm/i915: move trace_i915_reg_rw() to a separate file
There were two nearly identical sets of code for gup_fast()
style of walking the page tables with interrupts disabled.
This has lead to the usual maintenance problems that arise from
having duplicated code.
There is already a core internal routine in gup.c for gup_fast(),
so just enhance it very s
This is the FOLL_PIN equivalent of __get_user_pages_fast(),
except with a more descriptive name, and gup_flags instead of
a boolean "write" in the argument list.
Signed-off-by: John Hubbard
---
include/linux/mm.h | 2 ++
mm/gup.c | 36
2 files chan
This code was using get_user_pages*(), in a "Case 2" scenario
(DMA/RDMA), using the categorization from [1]. That means that it's
time to convert the get_user_pages*() + put_page() calls to
pin_user_pages*() + unpin_user_pages() calls.
There is some helpful background in [2]: basically, this is a
This is in order to avoid a forward declaration of
internal_get_user_pages_fast(), in the next patch.
This is code movement only--all generated code should
be identical.
Signed-off-by: John Hubbard
---
mm/gup.c | 112 +++
1 file changed, 56 in
This needs to go through Andrew's -mm tree, due to adding a new gup.c
routine. However, I would really love to have some testing from the
drm/i915 folks, because I haven't been able to run-time test that part
of it.
Otherwise, though, the series has passed my basic run time testing:
some LTP tests
On Tue, May 19, 2020 at 12:35:02AM +0100, Chris Wilson wrote:
Quoting Lucas De Marchi (2020-05-19 00:30:49)
Now that we have the declaration of trace_i915_reg_rw() in a separate
header, start tracing intel_uncore_*_fw() mmio-accessors.
These were untraced on purpose.
humn... can you share wh
== Series Details ==
Series: series starting with [1/4] drm/i915: Don't set queue-priority hint when
supressing the reschedule
URL : https://patchwork.freedesktop.org/series/77377/
State : success
== Summary ==
CI Bug Log - changes from CI_DRM_8498 -> Patchwork_17702
=
Quoting Lucas De Marchi (2020-05-19 00:30:49)
> Now that we have the declaration of trace_i915_reg_rw() in a separate
> header, start tracing intel_uncore_*_fw() mmio-accessors.
These were untraced on purpose.
-Chris
___
Intel-gfx mailing list
Intel-gfx@
== Series Details ==
Series: series starting with [1/4] drm/i915: Don't set queue-priority hint when
supressing the reschedule
URL : https://patchwork.freedesktop.org/series/77377/
State : warning
== Summary ==
$ dim sparse --fast origin/drm-tip
Sparse version: v0.6.0
Fast mode used, each com
== Series Details ==
Series: series starting with [1/4] drm/i915: Don't set queue-priority hint when
supressing the reschedule
URL : https://patchwork.freedesktop.org/series/77377/
State : warning
== Summary ==
$ dim checkpatch origin/drm-tip
c3388b20997e drm/i915: Don't set queue-priority hi
The device always exist, so use it to derive the module name required to
lookup either the debugfs params directory or the sysfs module parameters.
Fixes: 2f5cee33ce55 ("igt/params: use igt_params_set_save for
igt_set_module_param*")
Signed-off-by: Chris Wilson
Cc: Jani Nikula
Cc: Juha-Pekka He
Don't second guess, require the user to provide the device that wish to
set the module parameter for.
Signed-off-by: Chris Wilson
---
lib/igt_params.c | 8
lib/igt_params.h | 4 ++--
lib/igt_psr.c | 18 -
lib/igt_ps
Now that we have the declaration of trace_i915_reg_rw() in a separate
header, start tracing intel_uncore_*_fw() mmio-accessors.
Signed-off-by: Lucas De Marchi
---
drivers/gpu/drm/i915/intel_uncore.h | 27 ++-
1 file changed, 22 insertions(+), 5 deletions(-)
diff --git a/
Currently we can't call trace_i915_reg_rw() from some headers due to
include order and i915_trace.h needing some struct definitions.
Move the declaration of trace_i915_reg_rw() to another file so it can be
included separately. Note that the trace points are still defined by
i915_trace_point.c whic
== Series Details ==
Series: drm/i915: Resolve device hotunplug issues
URL : https://patchwork.freedesktop.org/series/77372/
State : success
== Summary ==
CI Bug Log - changes from CI_DRM_8498 -> Patchwork_17701
Summary
---
**SUCCESS
== Series Details ==
Series: drm/i915/gvt: Use ARRAY_SIZE for vgpu_types
URL : https://patchwork.freedesktop.org/series/77369/
State : success
== Summary ==
CI Bug Log - changes from CI_DRM_8498_full -> Patchwork_17699_full
Summary
---
We recorded the execlists->queue_priority_hint update for the inflight
request without kicking the tasklet. The next submitted request then
failed to be scheduled as it had a lower priority than the hint, leaving
the HW runnning with only the inflight request.
Fixes: 6cebcf746f3f ("drm/i915: Tweak
Check for integer overflow in the priority chain, rather than against a
type-constricted max-priority check.
Signed-off-by: Chris Wilson
---
drivers/gpu/drm/i915/gt/selftest_lrc.c | 6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)
diff --git a/drivers/gpu/drm/i915/gt/selftest_lrc.c
b/
When we look at i915_request_is_started() we must be careful in case we
are using a request that does not have the initial-breadcrumb and
instead the is-started is being compared against the end of the previous
request. This will make wait_for_submit() declare that a request has
been already submit
Since we temporarily disable the heartbeat and restore back to the
default value, we can use the stored defaults on the engine and avoid
using a local.
Signed-off-by: Chris Wilson
---
drivers/gpu/drm/i915/gt/selftest_hangcheck.c | 25 +++
drivers/gpu/drm/i915/gt/selftest_lrc.c | 67 +++
== Series Details ==
Series: drm/i915: Resolve device hotunplug issues
URL : https://patchwork.freedesktop.org/series/77372/
State : warning
== Summary ==
$ dim sparse --fast origin/drm-tip
Sparse version: v0.6.0
Fast mode used, each commit won't be checked separately.
-
+drivers/gpu/drm/i915/
== Series Details ==
Series: drm/i915: Resolve device hotunplug issues
URL : https://patchwork.freedesktop.org/series/77372/
State : warning
== Summary ==
$ dim checkpatch origin/drm-tip
4d3df739b858 drm/i915: Drop user contexts on driver remove
-:19: WARNING:COMMIT_LOG_LONG_LINE: Possible unw
== Series Details ==
Series: drm/i915/selftests: Check for an initial-breadcrumb in wait_for_submit()
URL : https://patchwork.freedesktop.org/series/77371/
State : failure
== Summary ==
CI Bug Log - changes from CI_DRM_8498 -> Patchwork_17700
===
The device always exist, so use it to derive the module name required to
lookup either the debugfs params directory or the sysfs module parameters.
Fixes: 2f5cee33ce55 ("igt/params: use igt_params_set_save for
igt_set_module_param*")
Signed-off-by: Chris Wilson
Cc: Jani Nikula
Cc: Juha-Pekka He
gen8-unprivileged-single-counters appears to leak contexts, and does not
appear to be doing so intentionally.. Let's assume it's a bug in the
test and see if the contexts are released along with the device fd.
(If they are not released, that is more clearly an issue.)
Signed-off-by: Chris Wilson
We need each test in an isolated context, so that bad results from one
test do not interfere with the next. In particular, we want to clean up
the device and reset it to the defaults so that they are known for the
next test, and the test can focus on behaviour it wants to control.
Signed-off-by: C
Quoting Aishwarya Ramakrishnan (2020-05-18 16:03:36)
> Prefer ARRAY_SIZE instead of using sizeof
>
> Fixes coccicheck warning: Use ARRAY_SIZE
>
> Signed-off-by: Aishwarya Ramakrishnan
Reviewed-by: Chris Wilson
-Chris
___
Intel-gfx mailing list
Intel-g
== Series Details ==
Series: drm/i915/gvt: Use ARRAY_SIZE for vgpu_types
URL : https://patchwork.freedesktop.org/series/77369/
State : success
== Summary ==
CI Bug Log - changes from CI_DRM_8498 -> Patchwork_17699
Summary
---
**SUCCE
The device always exist, so use it to derive the module name required to
lookup either the debugfs params directory or the sysfs module parameters.
Fixes: 2f5cee33ce55 ("igt/params: use igt_params_set_save for
igt_set_module_param*")
Signed-off-by: Chris Wilson
Cc: Jani Nikula
Cc: Juha-Pekka He
== Series Details ==
Series: drm/i915: fix incorrect return of an error status
URL : https://patchwork.freedesktop.org/series/77368/
State : failure
== Summary ==
Applying: drm/i915: fix incorrect return of an error status
Using index info to reconstruct a base tree...
M drivers/gpu/drm/
== Series Details ==
Series: series starting with [1/3] drm/i915/params: don't expose
inject_probe_failure in debugfs
URL : https://patchwork.freedesktop.org/series/77366/
State : failure
== Summary ==
CI Bug Log - changes from CI_DRM_8497 -> Patchwork_17697
==
== Series Details ==
Series: drm/i915/selftests: Measure dispatch latency (rev8)
URL : https://patchwork.freedesktop.org/series/77308/
State : success
== Summary ==
CI Bug Log - changes from CI_DRM_8495_full -> Patchwork_17695_full
Summary
== Series Details ==
Series: series starting with [1/3] drm/i915/params: don't expose
inject_probe_failure in debugfs
URL : https://patchwork.freedesktop.org/series/77366/
State : warning
== Summary ==
$ dim sparse --fast origin/drm-tip
Sparse version: v0.6.0
Fast mode used, each commit won't
== Series Details ==
Series: series starting with [1/7] drm/i915: Move saturated workload detection
back to the context
URL : https://patchwork.freedesktop.org/series/77365/
State : failure
== Summary ==
CI Bug Log - changes from CI_DRM_8497 -> Patchwork_17696
== Series Details ==
Series: series starting with [1/3] drm/i915/params: don't expose
inject_probe_failure in debugfs
URL : https://patchwork.freedesktop.org/series/77366/
State : warning
== Summary ==
$ dim checkpatch origin/drm-tip
960f6e1f205d drm/i915/params: don't expose inject_probe_fai
== Series Details ==
Series: series starting with [1/7] drm/i915: Move saturated workload detection
back to the context
URL : https://patchwork.freedesktop.org/series/77365/
State : warning
== Summary ==
$ dim sparse --fast origin/drm-tip
Sparse version: v0.6.0
Fast mode used, each commit won
== Series Details ==
Series: series starting with [1/7] drm/i915: Move saturated workload detection
back to the context
URL : https://patchwork.freedesktop.org/series/77365/
State : warning
== Summary ==
$ dim checkpatch origin/drm-tip
a7577cd7f63b drm/i915: Move saturated workload detection
On Fri, May 15, 2020 at 10:59:57PM +0300, Ville Syrjälä wrote:
> On Fri, May 15, 2020 at 12:18:22PM -0700, Manasi Navare wrote:
> > On Tue, May 12, 2020 at 08:41:43PM +0300, Ville Syrjala wrote:
> > > From: Ville Syrjälä
> > >
> > > The DP spec says:
> > > "When the combination of the requested p
The idea is to revoke DMA mappings on driver remove in order to work
around intel-iommu judging late unmapping on driver release after an
open device is removed as bugs. That also resolves runtime power
management warnings on late object removal.
Janusz Krzysztofik (4):
drm/i915: Drop user cont
GGTT including its scratch page is not cleaned up until driver release.
Since DMA mappings still exist before scratch page cleanup, unmapping
them on last device close after the driver has been already removed may
be judged by intel-iommu code as a bug and result in kernel panic.
Since we abort re
Contexts associated with open device file descriptors together with
their assigned address spaces are now closed on device file close. On
address space closure its associated DMA mappings are revoked. If the
device is removed while being open, subsequent attempts to revoke
those mappings while cl
UC firmware is stored in a GEM object. Clean it up on driver remove to
avoid intel-iommu triggered kernel panic on late DMA unmapping or even
an RPM related warning on object late removal in no IOMMU setups.
<4> [93.335282] [ cut here ]
<4> [93.335515] pm_runtime_get_sync(
GT scratch page is now released and its DMA mappings revoked on driver
release. If a device is removed while its file descriptor is still
open, the driver is not released until last device file descriptor
closure. In that case intel-iommu code may judge late DMA unmapping as
a bug and kernel pani
== Series Details ==
Series: drm/i915/selftests: Measure dispatch latency (rev8)
URL : https://patchwork.freedesktop.org/series/77308/
State : success
== Summary ==
CI Bug Log - changes from CI_DRM_8495 -> Patchwork_17695
Summary
---
When we look at i915_request_is_started() we must be careful in case we
are using a request that does not have the initial-breadcrumb and
instead the is-started is being compared against the end of the previous
request. This will make wait_for_submit() declare that a request has
been already submit
Prefer ARRAY_SIZE instead of using sizeof
Fixes coccicheck warning: Use ARRAY_SIZE
Signed-off-by: Aishwarya Ramakrishnan
---
drivers/gpu/drm/i915/gvt/vgpu.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/gpu/drm/i915/gvt/vgpu.c b/drivers/gpu/drm/i915/gvt/vgpu.c
inde
On Mon, May 18, 2020 at 02:14:15PM +0100, Chris Wilson wrote:
> Quoting Ville Syrjala (2020-05-18 13:13:54)
> > From: Ville Syrjälä
> >
> > The current dbuf slice computation only happens when there are
> > active pipes. If we are turning off all the pipes we just leave
> > the dbuf slice mask at
== Series Details ==
Series: drm/i915: Fix dbuf slice mask when turning off all the pipes (rev2)
URL : https://patchwork.freedesktop.org/series/77322/
State : failure
== Summary ==
CI Bug Log - changes from CI_DRM_8494_full -> Patchwork_17690_full
==
From: Colin Ian King
Currently when a call to intel_atomic_get_dbuf_state fails the
error value being returns is a potentially uninitialized value
in variable ret. Fix this by returning the error value of
new_dbuf_state.
Addresses-Coverity: ("Uninitialized scalar value")
Fixes: 3cf43cdc63fb ("d
fake_lmem_start does not need to be mutable via module param sysfs. It's
only used during driver probe.
Fixes: 1629224324b6 ("drm/i915/lmem: add the fake lmem region")
Cc: Matthew Auld
Cc: Joonas Lahtinen
Cc: Chris Wilson
Reviewed-by: Rodrigo Vivi
Signed-off-by: Jani Nikula
---
drivers/gpu/d
Only support runtime changes through the debugfs.
i915.verbose_state_checks remains an exception, and is not exposed via
debugfs.
This depends on IGT having been updated to use the debugfs for modifying
the parameters.
Cc: Juha-Pekka Heikkilä
Cc: Venkata Sandeep Dhanalakota
Reviewed-by: Juha-P
The parameter only makes sense as a module parameter only.
Fixes: c43c5a8818d4 ("drm/i915/params: add i915 parameters to debugfs")
Cc: Juha-Pekka Heikkilä
Cc: Venkata Sandeep Dhanalakota
Reviewed-by: Juha-Pekka Heikkila
Signed-off-by: Jani Nikula
---
drivers/gpu/drm/i915/i915_params.h | 2 +-
This is the first 3 patches of [1], because apparently patch 4 breaks
the world. I've yet to pinpoint the issue, but these could move forward
in the meanwhile.
BR,
Jani.
[1] https://patchwork.freedesktop.org/series/77272/
___
Intel-gfx mailing list
If we decide to timeslice out the current virtual request, we will
unsubmit it while it is still busy (ve->context.inflight == sibling[0]).
If the virtual tasklet and then the other sibling tasklets run before we
completely schedule out the active virtual request for the preemption,
those other tas
It was quite the oversight to only factor in the normal queue to decide
the timeslicing switch priority. By leaving out the next virtual request
from the priority decision, we would not timeslice the current engine if
there was an available virtual request.
Testcase: igt/gem_exec_balancer/sliced
F
Having recognised that we do not change the sibling until we schedule
out, we can then defer the decision to resubmit the virtual engine from
the unwind of the active queue to scheduling out of the virtual context.
By keeping the unwind order intact on the local engine, we can preserve
data depend
Rather than going back and forth between the rb_node entry and the
virtual_engine type, store the ve local and reuse it. As the
container_of conversion from rb_node to virtual_engine requires a
variable offset, performing that conversion just once shaves off a bit
of code.
v2: Keep a single virtua
Make sure that we can execute a virtual request on an already busy
engine, and conversely that we can execute a normal request if the
engines are already fully occupied by virtual requests.
Signed-off-by: Chris Wilson
---
drivers/gpu/drm/i915/gt/selftest_lrc.c | 188 -
1
When we introduced the saturated workload detection to tell us to back
off from semaphore usage [semaphores have a noticeable impact on
contended bus cycles with the CPU for some heavy workloads], we first
introduced it as a per-context tracker. This allows individual contexts
to try and optimise t
Once a virtual engine has been bound to a sibling, it will remain bound
until we finally schedule out the last active request. We can not rebind
the context to a new sibling while it is inflight as the context save
will conflict, hence we wait. As we cannot then use any other sibliing
while the con
On 2020-05-18 at 10:32:09 -0400, Sean Paul wrote:
> On Fri, May 15, 2020 at 10:48 AM Ramalingam C wrote:
> >
> > On 2020-04-29 at 15:54:46 -0400, Sean Paul wrote:
> > > From: Sean Paul
> > >
> > > Changes in v6:
> > > -Rebased on -tip
> > > -Disabled HDCP over MST on GEN12
> > > -Addressed Lyude'
== Series Details ==
Series: HDCP minor refactoring (rev2)
URL : https://patchwork.freedesktop.org/series/77224/
State : success
== Summary ==
CI Bug Log - changes from CI_DRM_8494_full -> Patchwork_17688_full
Summary
---
**SUCCESS**
== Series Details ==
Series: series starting with [1/8] drm/i915: Move saturated workload detection
back to the context (rev2)
URL : https://patchwork.freedesktop.org/series/77344/
State : failure
== Summary ==
CI Bug Log - changes from CI_DRM_8494 -> Patchwork_17693
=
A useful metric of the system's health is how fast we can tell the GPU
to do various actions, so measure our latency.
v2: Refactor all the instruction building into emitters.
Signed-off-by: Chris Wilson
Cc: Mika Kuoppala
Cc: Joonas Lahtinen
---
drivers/gpu/drm/i915/selftests/i915_request.c |
== Series Details ==
Series: drm/i915/selftests: Measure dispatch latency (rev7)
URL : https://patchwork.freedesktop.org/series/77308/
State : failure
== Summary ==
Applying: drm/i915/selftests: Measure dispatch latency
error: corrupt patch at line 23
error: could not build fake ancestor
hint:
== Series Details ==
Series: series starting with [1/8] drm/i915: Move saturated workload detection
back to the context (rev2)
URL : https://patchwork.freedesktop.org/series/77344/
State : warning
== Summary ==
$ dim sparse --fast origin/drm-tip
Sparse version: v0.6.0
Fast mode used, each com
== Series Details ==
Series: series starting with [1/8] drm/i915: Move saturated workload detection
back to the context (rev2)
URL : https://patchwork.freedesktop.org/series/77344/
State : warning
== Summary ==
$ dim checkpatch origin/drm-tip
ebeb414a1e8b drm/i915: Move saturated workload det
== Series Details ==
Series: drm/i915/dsb: Pre allocate and late cleanup of cmd buffer (rev9)
URL : https://patchwork.freedesktop.org/series/73036/
State : failure
== Summary ==
CI Bug Log - changes from CI_DRM_8494 -> Patchwork_17692
Summa
Quoting Chris Wilson (2020-05-18 16:40:15)
> Quoting Tvrtko Ursulin (2020-05-18 15:55:46)
> >
> > On 18/05/2020 14:00, Chris Wilson wrote:
> > > Quoting Tvrtko Ursulin (2020-05-18 13:53:29)
> > >>
> > >> On 18/05/2020 09:14, Chris Wilson wrote:
> > >>> Once a virtual engine has been bound to a sib
== Series Details ==
Series: drm/i915/selftests: Refactor sibling selection
URL : https://patchwork.freedesktop.org/series/77352/
State : success
== Summary ==
CI Bug Log - changes from CI_DRM_8494_full -> Patchwork_17687_full
Summary
-
Quoting Tvrtko Ursulin (2020-05-18 15:55:46)
>
> On 18/05/2020 14:00, Chris Wilson wrote:
> > Quoting Tvrtko Ursulin (2020-05-18 13:53:29)
> >>
> >> On 18/05/2020 09:14, Chris Wilson wrote:
> >>> Once a virtual engine has been bound to a sibling, it will remain bound
> >>> until we finally schedul
1 - 100 of 190 matches
Mail list logo
