p. See CG_FUNCS in drivers/gpu/drm/i915/intel_clock_gating.c
The callers are the ones calling intel_clock_gating_init(), which is
both on display and gem sides. On the GEM side there's already and
eternal TODO comment:
* FIXME: break up the workarounds and apply them at the right time!
Luca
hed:
1) Survivability mode: to allow communication with mei and
possibly unbrick a device. Right now in xe we handle this
separately though: the driver needs to be initialized in this
mode
2) devcoredump: if there's a crash to be debugged with the help
On Tue, Jan 21, 2025 at 12:19:15PM -0500, Liang, Kan wrote:
On 2025-01-21 11:59 a.m., Lucas De Marchi wrote:
On Tue, Jan 21, 2025 at 10:53:31AM -0500, Liang, Kan wrote:
On 2025-01-21 9:29 a.m., Lucas De Marchi wrote:
On Mon, Jan 20, 2025 at 08:42:41PM -0500, Liang, Kan wrote:
-static int
On Mon, Feb 10, 2025 at 06:16:58PM +0200, Ville Syrjälä wrote:
On Fri, Feb 07, 2025 at 04:41:11PM -0600, Lucas De Marchi wrote:
On Fri, Feb 07, 2025 at 11:54:03PM +0200, Ville Syrjälä wrote:
>From: Ville Syrjälä
>
>Something has changed in the hardware on LNL/BMG because
>HDM
{ 14, 0, &xe_lpdp_display },
{ 14, 1, &xe2_hpd_display },
...
}
So maybe we need to check for the full version >= 1401 instead?
+Matt Roper, +Gustavo who may know the right bspec to confirm this
change in behavior
Lucas De Marchi
Ville Syrjälä (3):
drm/i915: Fix scan
On Fri, 31 Jan 2025 15:11:32 -0800, Lucas De Marchi wrote:
> Since commit 4ba4f1afb6a9 ("perf: Generic hotplug support for a PMU with
> a scope"), there's generic support for system-wide counters and
> integration with cpu hotplug.
>
> The i915 counters ar
igned-off-by: Rik van Riel
Reviewed-by: Christoph Hellwig
Signed-off-by: Lucas De Marchi
---
drivers/scsi/scsi_scan.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/scsi/scsi_scan.c b/drivers/scsi/scsi_scan.c
index 087fcbfc9aaa3..96d7e1a9a7c7a 100644
--- a/drivers/scsi
On Sun, Feb 02, 2025 at 07:40:35PM +0900, Vincent Mailhol wrote:
Hi Lucas and Yury,
On 08/02/2024 at 16:45, Lucas De Marchi wrote:
ove the implementation of REG_GENMASK/REG_BIT to a more appropriate
place to be shared by i915, xe and possibly other parts of the kernel.
For now this re-defines
n we are close to remove the force_probe.
Lucas De Marchi
--
Gustavo Sousa
after that LGTM:
Reviewed-by: Krzysztof Karas
Krzysztof
use breakages.
Cc: Kan Liang
Cc: Peter Zijlstra (Intel)
Cc: Vinay Belgaumkar
Reviewed-by: Umesh Nerlige Ramappa
Signed-off-by: Lucas De Marchi
---
v2: Expand commit message explanation to clarify what was discussed with
Kan, Tvrtko and Umesh in
https://patchwork.freedesktop.org/patch/
: Lucas De Marchi
Cc: Riana Tauro
Cc: Rodrigo Vivi
Signed-off-by: Vinay Belgaumkar
---
tests/intel/xe_pmu.c | 176 +++
tests/meson.build| 1 +
2 files changed, 177 insertions(+)
create mode 100644 tests/intel/xe_pmu.c
diff --git a/tests/intel/xe_pmu.c
On Mon, Jan 27, 2025 at 02:33:00PM -0800, Vinay Belgaumkar wrote:
Functions to parse event ID and GT bit shift for PMU events.
v2: Review comments (Riana)
Cc: Riana Tauro
Cc: Lucas De Marchi
Cc: Kamil Konieczny
Cc: Rodrigo Vivi
Signed-off-by: Vinay Belgaumkar
---
lib/igt_perf.c | 70
TRACE_SYSTEM xe
looking forward to the day this will be intel_display or i915_display,
but until then
Reviewed-by: Lucas De Marchi
is tracpoints above intentional? I'd say it's a typo, but it's repeated
4 times.
Lucas De Marchi
+#endif
#if !defined(__INTEL_DISPLAY_TRACE_
On Fri, Jan 24, 2025 at 04:46:21PM -0800, Umesh Nerlige Ramappa wrote:
Hi Lucas,
Mostly a bunch of questions since I think I am missing something.
On Tue, Jan 21, 2025 at 10:59:08AM -0600, Lucas De Marchi wrote:
On Tue, Jan 21, 2025 at 10:53:31AM -0500, Liang, Kan wrote:
On 2025-01-21 9:29
On Thu, Jan 23, 2025 at 06:06:29PM +, Tvrtko Ursulin wrote:
On 23/01/2025 16:27, Lucas De Marchi wrote:
On Thu, Jan 23, 2025 at 09:43:35AM +, Tvrtko Ursulin wrote:
On 20/01/2025 22:57, Lucas De Marchi wrote:
On Mon, Jan 20, 2025 at 10:08:39AM -0500, Liang, Kan wrote:
On 2025-01
On Thu, Jan 23, 2025 at 09:43:35AM +, Tvrtko Ursulin wrote:
On 20/01/2025 22:57, Lucas De Marchi wrote:
On Mon, Jan 20, 2025 at 10:08:39AM -0500, Liang, Kan wrote:
On 2025-01-16 5:24 p.m., Lucas De Marchi wrote:
Since commit 4ba4f1afb6a9 ("perf: Generic hotplug support for a PMU w
On Tue, Jan 21, 2025 at 10:53:31AM -0500, Liang, Kan wrote:
On 2025-01-21 9:29 a.m., Lucas De Marchi wrote:
On Mon, Jan 20, 2025 at 08:42:41PM -0500, Liang, Kan wrote:
-static int i915_pmu_cpu_offline(unsigned int cpu, struct hlist_node
*node)
-{
- struct i915_pmu *pmu = hlist_entry_safe
But it's still a behavior change. Please make it clear in the
description that the patch also changes/fixes the scope from core scope
to system-wide.
sure... do you have a suggestion how to test the hotplug? For testing
purposes, can I force the perf cpu assigned to be something other than
the
On Mon, Jan 20, 2025 at 10:08:39AM -0500, Liang, Kan wrote:
On 2025-01-16 5:24 p.m., Lucas De Marchi wrote:
Since commit 4ba4f1afb6a9 ("perf: Generic hotplug support for a PMU with
a scope"), there's generic support for system-wide counters and
integration with cpu hotplug. S
On Fri, Jan 17, 2025 at 05:55:57PM +0200, Imre Deak wrote:
On Thu, Jan 16, 2025 at 02:38:34PM -0600, Lucas De Marchi wrote:
On Mon, Jan 13, 2025 at 07:40:59PM +0200, Imre Deak wrote:
> On Mon, Jan 13, 2025 at 06:38:34PM +0200, Jani Nikula wrote:
> > On Mon, 13 Jan 2025, Imre De
iang
Cc: Peter Zijlstra (Intel)
Cc: Vinay Belgaumkar
Signed-off-by: Lucas De Marchi
---
drivers/gpu/drm/i915/i915_module.c | 2 -
drivers/gpu/drm/i915/i915_pmu.c| 114 +
drivers/gpu/drm/i915/i915_pmu.h| 11 ---
3 files changed, 1 insertion(+), 126 deletions(-
ons at all. Is it ever reasonable for the
user to disable this if USB4 is enabled?
On platforms that don't support DP tunneling, while supporting other
USB4 functionality (or for systems w/o any TypeC/DP connectors) it would
make sense to disable this option.
isn't this too fine grained? if we expose every single functionality of
the driver like this we will bury distros on configs and exponentially
explode the testing combination. And yes, this broke the build for me.
Lucas De Marchi
No such file or directory
12 | #include "intel_wakeref.h"
because we don't setup the right include directories.
We used to test in CI a display-disabled build, not sure what happened
with that.
Lucas De Marchi
+
obj-$(CONFIG_DRM_XE) += xe.o
obj-$(CONFIG_DRM_XE_KUNIT_TEST) += tests/
--
2.44.2
On Thu, Jan 16, 2025 at 03:30:51PM +0100, Francois Dugast wrote:
On Thu, Jan 16, 2025 at 08:26:36AM -0600, Lucas De Marchi wrote:
On Thu, Jan 16, 2025 at 01:45:32PM +0100, Francois Dugast wrote:
> Ensure all Xe driver files have a proper SPDX license identifier, add it
> in files where
otherwise, Reviewed-by: Lucas De Marchi
Lucas De Marchi
+ */
+
#ifndef _I915_GEM_STOLEN_H_
#define _I915_GEM_STOLEN_H_
--
2.43.0
perf event already has a default function that returns 0, no need to
override with the same thing.
Signed-off-by: Lucas De Marchi
---
drivers/gpu/drm/i915/i915_pmu.c | 6 --
1 file changed, 6 deletions(-)
diff --git a/drivers/gpu/drm/i915/i915_pmu.c b/drivers/gpu/drm/i915/i915_pmu.c
index
On Mon, Jan 06, 2025 at 09:27:34AM -0800, Matt Roper wrote:
On Mon, Jan 06, 2025 at 09:24:55AM -0800, Lucas De Marchi wrote:
This reverts commit adcfb264c3ed51fbbf5068ddf10d309a63683868.
It completely tanks the CI and nothing gets tested. This is being either
reverted upstream or the fix will
/g4sefofdrwu72ijhse7k57wuvrwhvn2eoqmc4jdoepkcgs7h5n@hmuhkwnye6pe/#t
References: https://gitlab.freedesktop.org/drm/i915/kernel/-/issues/13464
Signed-off-by: Lucas De Marchi
---
mm/vmstat.c | 3 +--
1 file changed, 1 insertion(+), 2 deletions(-)
diff --git a/mm/vmstat.c b/mm/vmstat.c
index 0889b75cef149..4d016314a56c9 100644
hat time we may then retire
intel_gpu_top and only use gputop as reference.
Lucas De Marchi
Lucas, Thomas, thoughts?
Or in case the conclusion ends up being "yes", then lets at least share some
more code between intel_gpu_top and this work. Ie. make it in a way gputop
completely subsu
only user. Now we already have 3, so I think it makes sense to go
through drm-misc... I looked at the possible conflicts in i915 and xe
and they are minimal.
thanks
Lucas De Marchi
[1]
https://lore.kernel.org/dri-devel/20241128153707.1294347-1-raag.ja...@intel.com/
André Almeida (1):
drm/a
t "lockdep: Enable PROVE_RAW_LOCK_NESTING with
PROVE_LOCKING."")
which should be available in drm-tip starting with
drm-tip: 2024y-12m-11d-16h-58m-54s UTC integration manifest
thanks
Lucas De Marchi
w I'm
looking for a real solution... ]
On Tue, Dec 10, 2024 at 05:55:35PM -0500, Rodrigo Vivi wrote:
On Tue, Dec 10, 2024 at 09:00:13AM -0800, Lucas De Marchi wrote:
On Mon, Dec 09, 2024 at 03:53:51PM +0200, Luca Coelho wrote:
> This reverts commit 560af5dc839eef08a273908f390cfefefb82aa0
;lock will actually be a good thing.
I will keep digging.
Lucas De Marchi
lib/Kconfig.debug | 12 ++--
1 file changed, 10 insertions(+), 2 deletions(-)
diff --git a/lib/Kconfig.debug b/lib/Kconfig.debug
index f3d723705879..de4ffe09323b 100644
--- a/lib/Kconfig.debug
+++ b/lib/
_show+0x46/0x170
dev_attr_show+0x1a/0x70
sysfs_kf_seq_show+0xaa/0x120
kernfs_seq_show+0x41/0x60
Just use the buffer size as argument to fix it.
Fixes: 226ff2e681d0 ("usb: typec: ucsi: Convert connector specific commands to
bitmaps")
Signed-off-by: Lucas De Marc
, Suresh Kumar ; Coelho, Luciano
; Nikula, Jani ; De Marchi, Lucas ;
intel-gfx@lists.freedesktop.org ; intel...@lists.freedesktop.org
; linux...@vger.kernel.org ; Sebastian Reichel
Subject: RE: Regression on linux-next (next-20241120) and drm-tip
+@Krogerus, Heikki
-Original Message-
From
005-power-supply-no-wakeup-source-v1-0-1d62bf9bc...@weissschuh.net/
it doesn's seem like these changes in power_supply are really the
culprit. I tried taking all the power_supply changes on top of v6.12
and I don't see any issue there. It's looking more like a memory
corruption
Could you point me to the full boot log in the drm-tip CI?
Here is one:
https://intel-gfx-ci.01.org/tree/intel-xe/xe-2310-5779fb3c12faf12589054127d60b1d36d56ba219/bat-lnl-1/boot0.txt
You can get more, with slightly different signatures by heading to
https://intel-gfx-ci.01.org/tree/intel-xe/bat-all.html?testfilter=igt@runner@aborted
- click in one of the red cells for bat-lnl and then boot0
Lucas De Marchi
Hi Dave and Simona,
A few more drm-xe fixes for this week.
thanks
Lucas De Marchi
drm-xe-fixes-2024-11-14:
Driver Changes:
- Fix unlock on exec ioctl error path (Matthew Brost)
- Fix hibernation on LNL due to ggtt getting lost
(Matthew Brost / Matthew Auld)
- Fix missing runtime PM in OA
Hi Dave and Simona,
drm-xe-fixes for 6.12-rc7. Still busier than I'd like for an rc7, but
needed particularly for LNL.
thanks,
Lucas De Marchi
drm-xe-fixes-2024-11-08:
Driver Changes:
- Fix ccs_mode setting for Xe2 and later (Balasubramani)
- Synchronize ccs_mode setting with client cre
This is already done inside perf_pmu_unregister() - no need to do it
before.
Reviewed-by: Matt Roper
Signed-off-by: Lucas De Marchi
---
drivers/gpu/drm/i915/i915_pmu.c | 7 +--
1 file changed, 1 insertion(+), 6 deletions(-)
diff --git a/drivers/gpu/drm/i915/i915_pmu.c b/drivers/gpu/drm
Both the documentation and most of other users call the return of
cpuhp_setup_state_multi() as "state". Follow that.
Reviewed-by: Matt Roper
Signed-off-by: Lucas De Marchi
---
drivers/gpu/drm/i915/i915_pmu.c | 14 +++---
1 file changed, 7 insertions(+), 7 deletions(-)
di
69111...@infradead.org/
With that, we can drop a few patches, keeping just the generic
refactors my series included. And in future, with a fixed
perf_pmu_unregister() we may simplify our checks by
pmu->registered too.
Lucas De Marchi (4):
drm/i915/pmu: Rename cpuhp_slot to cpuhp_state
drm/
_common.c.
Reviewed-by: Matt Roper
Signed-off-by: Lucas De Marchi
---
drivers/gpu/drm/i915/i915_pmu.c | 25 +
drivers/gpu/drm/i915/i915_pmu.h | 4 ++--
2 files changed, 15 insertions(+), 14 deletions(-)
diff --git a/drivers/gpu/drm/i915/i915_pmu.c b/drivers/gpu/drm/i
ady a "closed" boolean, use that instead and extend
it's meaning to unregistered/unregistering.
Reviewed-by: Matt Roper
Signed-off-by: Lucas De Marchi
---
drivers/gpu/drm/i915/i915_pmu.c | 18 ++
1 file changed, 6 insertions(+), 12 deletions(-)
diff --git a/
Hi Dave and Simona,
drm-xe-fixes for 6.12-rc6. Two important fixes for LNL and one missing
hpd enabling. This last one brought in a refactor as dependency,
otherwise it would cause some headache with conflicts.
thanks
Lucas De Marchi
drm-xe-fixes-2024-10-31:
Driver Changes:
- Fix missing HPD
By Peter:
git://git.kernel.org/pub/scm/linux/kernel/git/peterz/queue.git
perf/pmu-unregister
https://lore.kernel.org/all/20241022215210.ga31...@noisy.programming.kicks-ass.net/
Signed-off-by: Lucas De Marchi
---
include/linux/idr.h| 17 +
include/linux/perf_event.h | 35 ++-
kernel
Try again, this time with the correct igt series.
Test-with: 20241023050502.3049664-1-lucas.demar...@intel.com
Lucas De Marchi (1):
pmu changes
include/linux/idr.h| 17 +
include/linux/perf_event.h | 35 ++-
kernel/events/core.c | 620 +
3
, we have at least 2
projects to try to convince them that's a bad idea :)
Lucas De Marchi
On Wed, Oct 23, 2024 at 02:18:33PM -0700, Lucas De Marchi wrote:
This reverts commit efeb42fe3b7c58abbe674515252cfdb71c09eba5.
That machine is not in CI anymore, let's drop the hack.
Signed-off-by: Lucas De Marchi
Failures unrelated to these hacks being removed.
Thanks Jani for th
/** @unordered_wq: used to serialize unordered work, mostly display */
struct workqueue_struct *unordered_wq;
... which is, actually, just display.
Jani, should we move this wq to display where it belongs, with the right
flags, rather than queueing it on system_unbound_wq?
Lucas D
The current reading of engine utilization has same races. This should
fix most of them while also drastically reducing the update rate needed
on "normal apps".
I left tests/xe_drm_fdinfo running on 2 systems and saw no failures
after 100 iterations.
Lucas De Marchi (3):
drm/xe: Ad
Help debugging when LRC timestamp is updated for a exec queue.
Signed-off-by: Lucas De Marchi
---
drivers/gpu/drm/xe/Makefile | 1 +
drivers/gpu/drm/xe/xe_lrc.c | 3 ++
drivers/gpu/drm/xe/xe_trace_lrc.c | 9 ++
drivers/gpu/drm/xe/xe_trace_lrc.h | 52
equency in which the
update is needed.
Fixes: 2149ded63079 ("drm/xe: Fix use after free when client stats are
captured")
Signed-off-by: Lucas De Marchi
---
drivers/gpu/drm/xe/xe_exec_queue.c | 6 ++
drivers/gpu/drm/xe/xe_guc_submit.c | 2 --
2 files changed, 6 insertions(+), 2 deletions
the queue was removed from the array, and the value wasn't updated
yet by the free_job().
Explicitly accumulate the engine utilization so the right value is
visible after the ioctl return.
Link: https://gitlab.freedesktop.org/drm/xe/kernel/-/issues/2667
Cc: Jonathan Cavitt
Signed-off-by:
xe_gt_err(gt, "Timed out wait for G2H, fence %u, action %04x,
done %s",
|g2h_fence.seqno, action[0],
str_yes_no(g2h_fence.done));
| xa_erase(&ct->fence_lookup, g2h_fence.seqno);
| ->>>>>>> c9ff14d0339a (
extra \ from the Makefile (Jani)
Signed-off-by: Matthew Auld
Cc: Jonathan Cavitt
Cc: Matt Roper
Cc: Lucas De Marchi
Cc: Vinod Govindapillai
Cc: Jani Nikula
Cc: intel-gfx@lists.freedesktop.org
Reviewed-by: Jonathan Cavitt
Acked-by: Rodrigo Vivi
Link:
https://patchwork.freedesktop.org/patch
t Nautiyal
Signed-off-by: Ankit Nautiyal
(added prefix drm in subject)
Link:
https://patchwork.freedesktop.org/patch/msgid/20240730040941.396862-1-mitulkumar.ajitkumar.gol...@intel.com
Signed-off-by: Lucas De Marchi
---
drivers/gpu/drm/i915/display/intel_dp.c | 2 +-
1 file changed, 1 insertion(+),
This reverts commit efeb42fe3b7c58abbe674515252cfdb71c09eba5.
That machine is not in CI anymore, let's drop the hack.
Signed-off-by: Lucas De Marchi
---
drivers/acpi/sleep.c | 20
1 file changed, 20 deletions(-)
diff --git a/drivers/acpi/sleep.c b/drivers/acpi/sl
This reverts commit 852510ccd334063bec2c2c3e5a61a071599ac015.
This should be solved already since commit e6d0c13e9f46 ("ALSA: hda:
i915: Remove extra argument from snd_hdac_i915_init").
We can try to drop it and see if anything explodes on dg1 land.
Signed-off-by: Lucas De Marchi
Signed-off-by: Jani Nikula
Acked-by: Imre Deak
Reviewed-by: Lucas De Marchi
And applied. Or better, removed the patch from topic/core-for-CI.
Thanks
Lucas De Marchi
By Peter:
git://git.kernel.org/pub/scm/linux/kernel/git/peterz/queue.git
perf/pmu-unregister
https://lore.kernel.org/all/20241022215210.ga31...@noisy.programming.kicks-ass.net/
Signed-off-by: Lucas De Marchi
---
include/linux/idr.h| 17 +
include/linux/perf_event.h | 35 ++-
kernel
Test-with: 20241023042909.3038309-2-lucas.demar...@intel.com
Lucas De Marchi (1):
pmu changes
include/linux/idr.h| 17 +
include/linux/perf_event.h | 35 ++-
kernel/events/core.c | 620 +
3 files changed, 455 insertions(+), 217 deletions
By Peter:
git://git.kernel.org/pub/scm/linux/kernel/git/peterz/queue.git
perf/pmu-unregister
https://lore.kernel.org/all/20241022215210.ga31...@noisy.programming.kicks-ass.net/
Signed-off-by: Lucas De Marchi
---
include/linux/idr.h| 17 +
include/linux/perf_event.h | 35 ++-
kernel
By Peter:
git://git.kernel.org/pub/scm/linux/kernel/git/peterz/queue.git
perf/pmu-unregister
https://lore.kernel.org/all/20241022215210.ga31...@noisy.programming.kicks-ass.net/
To be tested with additional igt tests
Signed-off-by: Lucas De Marchi
---
include/linux/idr.h| 17
Hi Dave and Simona,
drm-xe-fixes for 6.12-rc4. Mostly some error path fixes and locking
adjustements. Timestamp bit width fixes delta time calculations in
userspace and one display fix for tile4 modifier in LNL/BMG.
thanks
Lucas De Marchi
drm-xe-fixes-2024-10-17:
Driver Changes:
- New
ady a "closed" boolean, use that instead and extend
it's meaning to unregistered/unregistering.
Signed-off-by: Lucas De Marchi
---
drivers/gpu/drm/i915/i915_pmu.c | 19 ++-
1 file changed, 6 insertions(+), 13 deletions(-)
diff --git a/drivers/gpu/drm/i915/i915_pmu.c b/
pl_common.c.
Signed-off-by: Lucas De Marchi
---
drivers/gpu/drm/i915/i915_pmu.c | 25 +
drivers/gpu/drm/i915/i915_pmu.h | 4 ++--
2 files changed, 15 insertions(+), 14 deletions(-)
diff --git a/drivers/gpu/drm/i915/i915_pmu.c b/drivers/gpu/drm/i915/i915_pmu.c
.508569383 0 ns i915__03_00.0/rc6-residency/
3.009405012 ns i915__03_00.0/rc6-residency/
Signed-off-by: Lucas De Marchi
---
drivers/gpu/drm/i915/i915_pmu.c | 35 -
drivers/gpu/drm/i915/i915_pmu.h | 6 ++
2 files changed, 40 insert
up change.
Signed-off-by: Lucas De Marchi
---
drivers/gpu/drm/i915/i915_pmu.c | 21 -
1 file changed, 16 insertions(+), 5 deletions(-)
diff --git a/drivers/gpu/drm/i915/i915_pmu.c b/drivers/gpu/drm/i915/i915_pmu.c
index 67b6cbdeff1da..4d05d98f51b8e 100644
--- a/drivers/gpu/drm/i915/
Patches 2 and 3 of
https://lore.kernel.org/all/20241008183501.1354695-1-lucas.demar...@intel.com/
with the unused module var removed (pending for v2 after review).
Signed-off-by: Lucas De Marchi
---
include/linux/perf_event.h | 12
kernel/events/core.c | 37
This is already done inside perf_pmu_unregister() - no need to do it
before.
Signed-off-by: Lucas De Marchi
---
drivers/gpu/drm/i915/i915_pmu.c | 7 +--
1 file changed, 1 insertion(+), 6 deletions(-)
diff --git a/drivers/gpu/drm/i915/i915_pmu.c b/drivers/gpu/drm/i915/i915_pmu.c
index
Both the documentation and most of other users call the return of
cpuhp_setup_state_multi() as "state". Follow that.
Signed-off-by: Lucas De Marchi
---
drivers/gpu/drm/i915/i915_pmu.c | 14 +++---
1 file changed, 7 insertions(+), 7 deletions(-)
diff --git a/drivers/gp
pmu
object to be alive.
Instead of pigging back on the event->destroy() to take and put the
device reference, implement the new get()/put() on the pmu object for
that purpose.
(c) is only done to have a flag to avoid some function entrypoints when
pmu is unregistered.
Cc: sta...@vger.kernel.org # 5.1
00:03:00.0: [drm] *ERROR* Failed to setup i915 fbdev emulation
(ret=-512)
and
- RIP: 0010:drm_mode_config_cleanup+0x2f4/0x310 [drm]
Lucas De Marchi (8):
squashed series - 20241008183501.1354695-1-lucas.demar...@intel.com
drm/i915/pmu: Let resource survive unbind
drm/i91
This reverts commit 80e846398c695dc3d0d2eac6da91772ff2aade65.
Let's see what CI says about reverting this 4 ~ 5 years old commit. We
shouldn't be in the business of changing perf/core in our "for CI" drm
branch.
Signed-off-by: Lucas De Marchi
---
kerne
Hi Dave and Simona,
A few drm-xe fixes for 6.12-rc3.
thanks
Lucas De Marchi
drm-xe-fixes-2024-10-10:
Driver Changes:
- Fix error checking with xa_store() (Matthe Auld)
- Fix missing freq restore on GSC load error (Vinay)
- Fix wedged_mode file permission (Matt Roper)
- Fix use-after-free in ct
unsigned long wedge_recovery;
>
> hmm, so before the driver can ask for a reboot as a recovery method from
> wedge it has to somehow add 'reboot' as available method? why it that?
It's for consumers to use as fallbacks in case the preferred recovery method
(sent alo
hrough this path is very odd when it has checks
like phys_base >= mem.vram.usable_size
Lucas De Marchi
u64 __iomem *gte = tile0->mem.ggtt->gsm;
u64 pte;
@@ -83,7 +83,6 @@ initial_plane_bo(struct xe_device *xe,
}
all. You are just trading the
direct access ggtt->pt_ops->pte_encode_bo with a function that returns a
function pointer and you all it later.
Lucas De Marchi
u32 column, row;
/* TODO: Maybe rewrite so we can traverse the bo addresses sequentially,
@@ -30,8 +31,8 @@ write
Hi Dave and Simona,
drm-xe-fixes for 6.12-rc2. A few fixes, particularly on error paths and
corner cases. We have more than usual as I decided to skip last week
pull request. Some tuning for Xe2 were missing and also got updated to
match the spec.
Thanks
Lucas De Marchi
drm-xe-fixes-2024-10-03
them to logs.)
I believe that I like the lowercase more too.
ack
We could have a map for the debug prints, but I would delay this, because
that could end up in discussions with marketing brand strings... :/
And just the platform without "is_" prefix, i.e. *not*:
if (di
ready have in xe, would
be great IMO:
if (display->platform == PLATFORM_LUNARLAKE)
it may be verbose, but shouldn't be much used to matter in the end.
Lucas De Marchi
- Go through the trouble of making the existing IS_FOO() macros _Generic
and accept either i915 or display pointe
On Thu, Sep 19, 2024 at 09:56:47PM GMT, Lucas De Marchi wrote:
Hi Dave and Simona,
A few fixes for 6.11-rc1.
oops, I meant 6.12-rc1, of course :)
Lucas De Marchi
Hi Dave and Simona,
A few fixes for 6.11-rc1.
Thanks
Lucas De Marchi
drm-xe-next-fixes-2024-09-19:
Driver Changes:
- Fix macro for checking minimum GuC version (Michal Wajdeczko)
- Fix CCS offset calculation for some BMG SKUs (Matthew Auld)
- Fix locking on memory usage reporting via fdinfo
On Tue, Sep 17, 2024 at 07:13:49PM GMT, Jani Nikula wrote:
It's really struct xe_bo, and struct drm_i915_gem_object only works
because of -Ddrm_i915_gem_object=xe_bo in xe Makefile.
Signed-off-by: Jani Nikula
---
Acked-by: Lucas De Marchi
for merging through drm-intel.
Lucas De Marchi
On Tue, Sep 17, 2024 at 07:14:00PM GMT, Jani Nikula wrote:
Some compat macros have inadvertently been sprinkled in xe core
headers. Remove the final users and the macros.
Signed-off-by: Jani Nikula
Acked-by: Lucas De Marchi
for merging through drm-intel.
Lucas De Marchi
Acked-by: Lucas De Marchi
for merging through drm-intel.
Lucas De Marchi
On Tue, Sep 17, 2024 at 07:14:01PM GMT, Jani Nikula wrote:
We've now completely stopped using drm_i915_gem_object in display code
that gets built for xe. Kill off the -Ddrm_i915_gem_object=xe_bo
hack. Good riddance.
Signed-off-by: Jani Nikula
Acked-by: Lucas De Marchi
for merging th
On Tue, Sep 17, 2024 at 07:13:59PM GMT, Jani Nikula wrote:
Now that we've switched to struct drm_gem_object and the intel_bo_*
interfaces, we no longer need most of the compat gem headers. Remove.
Signed-off-by: Jani Nikula
Acked-by: Lucas De Marchi
for merging through drm-intel.
On Thu, Sep 12, 2024 at 03:42:52PM GMT, Lucas De Marchi wrote:
unbind function is this:
function unbind {
vga="0300"
display="0380"
pci_vendor="8086"
while read -r pci_slot class devid xxx; do
sysdev=/sys/
s starting with this
release, we are being more careful and adding them when they are really
needed.
Thanks
Lucas De Marchi
drm-xe-next-fixes-2024-09-12:
Driver Changes:
- Fix usefafter-free when provisioning VF (Matthew Auld)
- Suppress rpm warning on false positive (Rodrigo)
- Fix memleak on i
On Thu, Sep 12, 2024 at 11:58:37AM GMT, Bommu, Krishnaiah wrote:
-Original Message-
From: De Marchi, Lucas
Sent: Wednesday, September 11, 2024 9:49 PM
To: Bommu, Krishnaiah
Cc: Vivi, Rodrigo ; intel...@lists.freedesktop.org;
intel-
g...@lists.freedesktop.org; Kamil Konieczny
IS_DISPLAY_VER(i915, md->display_ver.from, md->display_ver.until)
if so, with commit message rewording,
Reviewed-by: Lucas De Marchi
Lucas De Marchi
- if (md->modifier == I915_FORMAT_MOD_4_TILED_LNL_CCS &&
- (GRAPHICS_VER(i915) < 20 || IS_DGFX(i915
implified with:
if (rotation & DRM_MODE_REFLECT_X &&
intel_fb_is_tile4_modifier(fb->modifier) &&
DISPLAY_VER(dev_priv) >= 20)
Lucas De Marchi
+ drm_dbg_kms(&dev_priv->drm,
+ "horizontal flip is not s
+ linux-modules
+ Luis
On Wed, Sep 11, 2024 at 01:00:47AM GMT, Bommu, Krishnaiah wrote:
-Original Message-
From: De Marchi, Lucas
Sent: Tuesday, September 10, 2024 9:13 PM
To: Vivi, Rodrigo
Cc: Bommu, Krishnaiah ; intel-
x...@lists.freedesktop.org; intel-gfx@lists.freedesktop.org
On Tue, Sep 10, 2024 at 06:53:19PM GMT, Raag Jadav wrote:
On Mon, Sep 09, 2024 at 03:01:50PM -0500, Lucas De Marchi wrote:
On Sun, Sep 08, 2024 at 11:08:39PM GMT, Asahi Lina wrote:
> On 9/8/24 12:07 AM, Lucas De Marchi wrote:
> > On Sat, Sep 07, 2024 at 08:38:30PM GMT, Asahi L
MEI module handling across different
hardware
platforms, ensuring that the i915/Xe driver can be cleanly unloaded and reloaded
without issues.
v2: updated commit message
Signed-off-by: Bommu Krishnaiah
Cc: Kamil Konieczny
Cc: Daniele Ceraolo Spurio
Cc: Lucas De Marchi
Cc: Tejas Upadhyay
On Tue, Sep 03, 2024 at 10:32:03AM GMT, Jani Nikula wrote:
On Mon, 02 Sep 2024, Lucas De Marchi wrote:
On Mon, Sep 02, 2024 at 08:13:59PM GMT, Jani Nikula wrote:
There's considerable overlap in i915 and xe PCI ID macros, and (as can
be seen in this series) they get updated out of sync.
Hi Peter,
On Wed, Jul 24, 2024 at 10:39:57AM GMT, Lucas De Marchi wrote:
On Wed, Jul 24, 2024 at 02:41:05PM GMT, Peter Zijlstra wrote:
On Tue, Jul 23, 2024 at 10:30:08AM -0500, Lucas De Marchi wrote:
On Tue, Jul 23, 2024 at 09:03:25AM GMT, Tvrtko Ursulin wrote:
On 22/07/2024 22:06, Lucas De
There's no reason to hardcode checking for integrated graphics on a
specific pci slot. That information is already available per platform an
can be checked with IS_DGFX().
Reviewed-by: Ashutosh Dixit
Signed-off-by: Lucas De Marchi
---
drivers/gpu/drm/i915/i915_pmu.c | 17 +++
i915 pointer is not needed in this function and all the others simply
calculate the i915_pmu container based on the event->pmu. Follow the
same logic as in other functions.
Reviewed-by: Ashutosh Dixit
Signed-off-by: Lucas De Marchi
---
drivers/gpu/drm/i915/i915_pmu.c | 5 ++---
1 file chan
1 - 100 of 1589 matches
Mail list logo
