] drm/i915: Add ppgtt init/release trace
points
On Wed, Jun 18, 2014 at 05:16:39PM +0100, oscar.ma...@intel.com wrote:
From: Daniele Ceraolo Spurio daniele.ceraolospu...@intel.com
These tracepoints are useful for observing the creation and
destruction of Full PPGTTs.
Signed
On 04/08/15 09:56, Morton, Derek J wrote:
Bump.
Can this be merged? It only affects android and addresses an issue causing igt
to fail to build at all on android.
//Derek
-Original Message-
From: Morton, Derek J
Sent: Monday, July 27, 2015 11:31 AM
To:
tcase: igt/gem_concurrent_blit
Reported-by: Daniele Ceraolo Spurio <daniele.ceraolospu...@intel.com>
Signed-off-by: Michel Thierry <michel.thie...@intel.com>
---
The lock-up I was seeing goes away with these 2 patches applied, so:
Tested-by: Daniele Ceraolo Spurio <daniele.ceraolospu...@intel.com&
On 29/01/16 10:58, Chris Wilson wrote:
On Fri, Jan 29, 2016 at 09:21:33AM +, daniele.ceraolospu...@intel.com wrote:
From: Daniele Ceraolo Spurio <daniele.ceraolospu...@intel.com>
gem_require_ring will submit an execbuf using the provided flags and
skip the test if the ioctl
On 29/01/16 11:35, Chris Wilson wrote:
On Fri, Jan 29, 2016 at 11:16:37AM +, Daniele Ceraolo Spurio wrote:
On 29/01/16 10:58, Chris Wilson wrote:
On Fri, Jan 29, 2016 at 09:21:33AM +, daniele.ceraolospu...@intel.com wrote:
From: Daniele Ceraolo Spurio <daniele.ceraolo
On 27/01/16 16:39, Ville Syrjälä wrote:
On Wed, Jan 27, 2016 at 03:43:49PM +,daniele.ceraolospu...@intel.com wrote:
From: Daniele Ceraolo Spurio<daniele.ceraolospu...@intel.com>
While running some tests on the scheduler patches with rpm enabled I
came across a corr
On 25/02/16 11:32, Chris Wilson wrote:
On Thu, Feb 25, 2016 at 11:12:06AM +, Daniele Ceraolo Spurio wrote:
On 25/02/16 10:41, Chris Wilson wrote:
On Thu, Feb 25, 2016 at 10:32:11AM +, daniele.ceraolospu...@intel.com wrote:
+/* This test covers the case where we end up
On 25/02/16 10:41, Chris Wilson wrote:
On Thu, Feb 25, 2016 at 10:32:11AM +, daniele.ceraolospu...@intel.com wrote:
+/* This test covers the case where we end up in an uninitialised area of the
+ * ppgtt at an offset greater than the one where the last buffer is mapped.
This
+ * is
On 18/02/16 21:10, Chris Wilson wrote:
On Thu, Feb 18, 2016 at 05:34:50PM +, daniele.ceraolospu...@intel.com wrote:
+static void ppgtt_walking(void)
+{
+ memset(, 0, sizeof(execbuf));
+ execbuf.buffers_ptr = (uintptr_t)_exec;
+ execbuf.buffer_count = 1;
+
On 27/01/16 09:38, Chris Wilson wrote:
On Wed, Jan 27, 2016 at 08:55:40AM +, daniele.ceraolospu...@intel.com wrote:
From: Daniele Ceraolo Spurio <daniele.ceraolospu...@intel.com>
While running some tests on the scheduler patches with rpm enabled I
came across a corr
On 12/02/16 09:38, Derek Morton wrote:
Adds functions to create a number of different batch buffers to perform
several functions including:
Batch buffer which will run for a long duration to provide a delay on a
specified ring.
Function to calibrate the delay batch buffer to run for a
Hi,
first round of comments inline. A few things will probably have to
change based on the comments on the previous patch in the series so I'll
have a better look at the related logic once the new series is up.
Regards,
Daniele
On 12/02/16 09:38, Derek Morton wrote:
This is intended to
as
valid and invalid priority values we could use the values returned by
the debugfs forTEST_SUCCESS and the same values +/-1 for TEST_FAIL.
The patch as it is still LGTM, so with or without my suggested change:
Reviewed-by: Daniele Ceraolo Spurio <daniele.ceraolospu...@int
On 12/02/16 09:38, Derek Morton wrote:
Add subtests to test each ring to check batch buffers of a higher
priority will be executed before batch buffers of a lower priority.
Signed-off-by: Derek Morton
---
tests/gem_scheduler.c | 34
buffers that have write dependencies are
executed in submission order but the batch buffers without interdependencies
do not get held up.
v2: Addressed review comments from Daniele Ceraolo Spurio
v3: Added logic to use generic BSD ring if there is 1 and BSD1 / BSD2 if there
are 2.
Signed-off
On 10/03/16 11:03, Derek Morton wrote:
For tests that use multiple rings to test interactions it is
useful to know if a ring exists without triggering the test to skip.
Signed-off-by: Derek Morton
---
lib/ioctl_wrappers.c | 2 +-
lib/ioctl_wrappers.h | 1 +
2
for a specified period
of time.
Function to create a batch buffer which writes timestamps to a buffer object.
Function to compare timestamps allowing for wrapping of the values.
v2: Moved code to intel_batchbuffer (Daniel Vetter)
Addressed review comments from Daniele Ceraolo Spurio
Signed-off-by: Derek
On 10/03/16 11:03, Derek Morton wrote:
Add subtests to test each ring to check batch buffers of a higher
priority will be executed before batch buffers of a lower priority.
v2: Addressed review comments from Daniele Ceraolo Spurio
Signed-off-by: Derek Morton <derek.j.mor...@intel.
On 17/03/16 08:49, Daniele Ceraolo Spurio wrote:
On 10/03/16 11:03, Derek Morton wrote:
This is intended to test the scheduler behaviour is correct.
The subtests are
-basic
Tests that batch buffers of the same priority submitted to a ring
execute in the order they are submitted.
-read
On 17/05/16 14:34, Tvrtko Ursulin wrote:
From: Tvrtko Ursulin
Rather than asking itself "am I a Broadwell, am I a Cherryview,
or am I neither of the two" on on low level page table operations,
like inserting and clearing PTEs; add a new vfunc kunmap_page_dma
and set
Added a subtest for invalid FENCE_IN usage, updated invalid-flag subtest
and made the rsvd2 test skip when exec fences are available.
Signed-off-by: Daniele Ceraolo Spurio <daniele.ceraolospu...@intel.com>
---
lib/ioctl_wrappers.c| 29 +
lib/ioctl_wrap
Added a subtest for invalid FENCE_IN usage, updated invalid-flag subtest
and made the rsvd2 test skip when exec fences are available.
v2: add a test that tries to use the device fd as fence (Chris)
Cc: Chris Wilson <ch...@chris-wilson.co.uk>
Signed-off-by: Daniele Ceraolo
;ch...@chris-wilson.co.uk>
Signed-off-by: Daniele Ceraolo Spurio <daniele.ceraolospu...@intel.com>
---
drivers/gpu/drm/i915/i915_gem_execbuffer.c | 11 +--
1 file changed, 5 insertions(+), 6 deletions(-)
diff --git a/drivers/gpu/drm/i915/i915_gem_execbuffer.c
b/drivers/
aniele Ceraolo Spurio <daniele.ceraolospu...@intel.com>
---
drivers/gpu/drm/i915/i915_gem_execbuffer.c | 8 +++-
1 file changed, 3 insertions(+), 5 deletions(-)
diff --git a/drivers/gpu/drm/i915/i915_gem_execbuffer.c
b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
index a40ade6..7f
<tvrtko.ursu...@intel.com>
Signed-off-by: Daniele Ceraolo Spurio <daniele.ceraolospu...@intel.com>
---
drivers/gpu/drm/i915/intel_uncore.c | 40 ++---
1 file changed, 10 insertions(+), 30 deletions(-)
diff --git a/drivers/gpu/drm/i915/intel_uncore.c
b/dri
nd
one to track vma activity.
Added inside the i915_vma_bind function (was outside before) to keep a
similar placement as trace_i915_vma_unbind.
v2: print bind_flags instead of flags (Chris)
Cc: Chris Wilson <ch...@chris-wilson.co.uk>
Signed-off-by: Daniele Ceraolo Spurio <daniele.ceraolos
Ping. can anyone review/comment on this?
Thanks,
Daniele
On 12/01/17 14:21, daniele.ceraolospu...@intel.com wrote:
From: Daniele Ceraolo Spurio <daniele.ceraolospu...@intel.com>
The call went away in:
commit 3b16525cc4c1a43e9053cfdc414356eea24bdfad
Author: Chris Wilson <ch
On 20/01/17 13:28, Chris Wilson wrote:
On Fri, Jan 20, 2017 at 01:16:13PM -0800, Daniele Ceraolo Spurio wrote:
Ping. can anyone review/comment on this?
Thanks,
Daniele
On 12/01/17 14:21, daniele.ceraolospu...@intel.com wrote:
From: Daniele Ceraolo Spurio <daniele.ceraolospu...@intel.
<joonas.lahti...@linux.intel.com>
Cc: Michal Wajdeczko <michal.wajdec...@intel.com>
Cc: Arkadiusz Hiler <arkadiusz.hi...@intel.com>
Cc: Oscar Mateo <oscar.ma...@intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursu...@intel.com>
Cc: Chris Wilson <ch...@chris-wilson.
On 16/02/17 06:18, Oscar Mateo wrote:
Starting with intel_guc_loader, down to intel_guc_submission
and finally to intel_guc_log.
Signed-off-by: Oscar Mateo
---
drivers/gpu/drm/i915/i915_guc_submission.c | 94 +
drivers/gpu/drm/i915/intel_guc_loader.c| 19
On 23/02/17 08:38, Oscar Mateo wrote:
On 02/23/2017 11:14 AM, Michał Winiarski wrote:
We're always using all engines and kernel context for guc clients, let's
remove those arguments from guc_client_alloc.
I am quite new to the GuC but, by the look of it, passing the ctx was
groundwork for
We could then have a "registration_id", "registration_pool", etc.
Whatever name we decide to go with, if it differs from whatever is in
the documentation I suggest to make sure that the relation is properly
documented in the code to avoid confusion in the future.
Cheer
Reviewed-by: Daniele Ceraolo Spurio <daniele.ceraolospu...@intel.com>
On 24/02/17 11:15, Kelvin Gardiner wrote:
The replay bit of the ring mode register is not a valid bit for Gen8+. Do not
write to this bit.
Signed-off-by: Kelvin Gardiner <kelvin.gardi...@intel.com>
Cc: Joo
), move comment about save/restore because it
is not related to the mmio_white_list field.
Signed-off-by: Daniele Ceraolo Spurio <daniele.ceraolospu...@intel.com>
Cc: Michał Winiarski <michal.winiar...@intel.com>
Cc: Michal Wajdeczko <michal.wajdec...@intel.com>
C
Reviewed-by: Daniele Ceraolo Spurio <daniele.ceraolospu...@intel.com>
On 16/02/17 06:31, Oscar Mateo wrote:
I guess no one has needed to change the verbosity level of the GuC logs.
Signed-off-by: Oscar Mateo <oscar.ma...@intel.com>
---
tools/intel_guc_logger.c | 2 +-
1 fil
+
/*
* Tell the GuC to allocate or deallocate a specific doorbell
*/
-static int guc_allocate_doorbell(struct intel_guc *guc,
-struct i915_guc_client *client)
+static int __create_doorbell_hw(struct i915_guc_client *client)
I would rather prefer to only
On 10/02/17 05:30, Joonas Lahtinen wrote:
Started adding proper teardown to guc_client_alloc, ended up removing
quite a few dead ends where errors communicating with the GuC were
silently ignored. There also seemed to be quite a few erronous
teardown actions performed in case of an error
On 15/02/17 23:44, Chris Wilson wrote:
On Wed, Feb 15, 2017 at 06:28:59PM -0800, Daniele Ceraolo Spurio wrote:
On 14/02/17 05:53, Joonas Lahtinen wrote:
-static void guc_disable_doorbell(struct intel_guc *guc,
-struct i915_guc_client *client)
+static int
On 15/12/16 07:47, Arkadiusz Hiler wrote:
Current version of intel_guc_load() does a lot:
- cares about submission
- loads huc
- implement WA
This change offloads some of the logic to intel_uc_load(), which now
cares about the above.
Cc: Anusha Srivatsa
Cc:
+
+fail:
+ /*
+* We've failed to load the firmware :(
+*
+* Decide whether to disable GuC submission and fall back to
+* execlist mode, and whether to hide the error by returning
+* zero or to return -EIO, which the caller will treat as a
+
to trigger this assert. Let's add a proper bias.
Cc: Chris Wilson <ch...@chris-wilson.co.uk>
Cc: Daniele Ceraolo Spurio <daniele.ceraolospu...@intel.com>
Signed-off-by: Michał Winiarski <michal.winiar...@intel.com>
---
drivers/gpu/drm/i915/intel_guc_loader.c | 3 ++-
1 file
On 21/12/16 06:11, Chris Wilson wrote:
The GuC would like to own the upper portion of the GTT for itself, so
exclude it from our drm_mm to prevent us using it.
I had a quick chat with a GuC dev and he is pretty sure that GuC can't
reserve any GTT areas for itself (and that is why we do the
On 15/12/16 07:47, Arkadiusz Hiler wrote:
Let intel_guc_init() focus on determining and fetching the correct
firmware.
This patch introduces intel_sanitize_uc_params() that is called from
intel_uc_init().
Then, if we have GuC, we can call intel_guc_init() conditionally and we
do not have to
at the
bottom of the GGTT, and couple it into the assertion that we don't feed
unmappable addresses to the GuC.
Signed-off-by: Chris Wilson <ch...@chris-wilson.co.uk>
Cc: Michal Wajdeczko <michal.wajdec...@intel.com>
Cc: Arkadiusz Hiler <arkadiusz.hi...@intel.com>
Cc: Dani
On 22/12/16 08:38, Arkadiusz Hiler wrote:
On Thu, Dec 22, 2016 at 03:18:08PM +, Chris Wilson wrote:
On Thu, Dec 22, 2016 at 03:53:15PM +0100, Arkadiusz Hiler wrote:
On Wed, Dec 21, 2016 at 07:35:04PM +0100, Srivatsa, Anusha wrote:
With enable_guc_loading=2 and enable_guc_submission=0 I
On 22/12/16 14:23, Patchwork wrote:
== Series Details ==
Series: drm/i915: request ring to be pinned above GUC_WOPCM_TOP
URL : https://patchwork.freedesktop.org/series/17147/
State : failure
== Summary ==
Series 17147v1 drm/i915: request ring to be pinned above GUC_WOPCM_TOP
The forcewake_get call in the guc_send_mmio function was added to
avoid getting and releasing forcewake on each register access.
While this makes sense, all GuC registers are in the blitter range
so no need to wake all the wells.
Signed-off-by: Daniele Ceraolo Spurio <daniele.ceraolo
On 13/03/17 01:56, Oscar Mateo wrote:
On 03/13/2017 04:46 AM, Chris Wilson wrote:
On Fri, Mar 10, 2017 at 08:28:55AM -0800, Oscar Mateo wrote:
Doorbell release flow requires that we wait for GEN8_DRB_VALID bit to go
to zero after updating db_status before we call the GuC to release the
On 06/04/17 12:15, Rodrigo Vivi wrote:
From: Ben Widawsky
The docs are not yet correct, so I cannot provide a reference to it. In the
current docs, the size is actually smaller than SKL. This seems unlikely given
that in another part of the docs there are clearly
to me:
Reviewed-by: Daniele Ceraolo Spurio <daniele.ceraolospu...@intel.com>
On a related note, we only allocate the pages for saving the GuC state
if i915.enable_guc_submission=1 (they're part of the ADS), but from what
I can see we call intel_guc_suspend/resume unconditionally.
Since t
On 24/03/17 18:30, Michel Thierry wrote:
From: Arun Siluvery
GuC expects a list of registers from the driver which are saved/restored
during engine reset. The type of value to be saved is controlled by
flags. We provide a minimal set of registers that we want
g them after internal GuC reset
(Daniele).
Cc: Daniele Ceraolo Spurio <daniele.ceraolospu...@intel.com>
Signed-off-by: Arun Siluvery <arun.siluv...@linux.intel.com>
Signed-off-by: Jeff McGee <jeff.mc...@intel.com>
Signed-off-by: Michel Thierry <michel.thie...@intel.com>
@@
On 24/03/17 18:30, Michel Thierry wrote:
Final enablement patch for GPU hang detection using watchdog timeout.
Using the gem_context_setparam ioctl, users can specify the desired
timeout value in microseconds, and the driver will do the conversion to
'timestamps'.
The recommended default
On 18/04/17 13:23, Michel Thierry wrote:
Final enablement patch for GPU hang detection using watchdog timeout.
Using the gem_context_setparam ioctl, users can specify the desired
timeout value in microseconds, and the driver will do the conversion to
'timestamps'.
The recommended default
On 21/04/17 13:07, Michel Thierry wrote:
On 20/04/17 10:29, Michel Thierry wrote:
On 20/04/17 09:39, Daniele Ceraolo Spurio wrote:
On 20/04/17 04:33, Joonas Lahtinen wrote:
On ke, 2017-04-19 at 11:35 -0700, Michel Thierry wrote:
From: Arun Siluvery <arun.siluv...@linux.intel.
intel_uc_init_early back to the top (Michal).
Signed-off-by: Michel Thierry <michel.thie...@intel.com>
Signed-off-by: Michal Wajdeczko <michal.wajdec...@intel.com>
Signed-off-by: Oscar Mateo <oscar.ma...@intel.com>
Reviewed-by: Daniele Ceraolo Spurio <daniele.ceraolospu...@intel.
s Lahtinen <joonas.lahti...@linux.intel.com>
Cc: Michal Wajdeczko <michal.wajdec...@intel.com>
Cc: Arkadiusz Hiler <arkadiusz.hi...@intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursu...@intel.com>
Cc: Chris Wilson <ch...@chris-wilson.co.uk>
Cc: Daniele Ceraolo Spurio <da
On 10/03/17 08:28, Oscar Mateo wrote:
While at it, fix a typo (s/ring_lcra/ring_lrca) and improve the naming of one
firware interface field (s/ring_tail/submit_element_info, since it can contain
more than just the ring tail).
No change in functionality.
Signed-off-by: Oscar Mateo
every time we try to load GuC.
Cc: Jeff McGee <jeff.mc...@intel.com>
Cc: Anusha Srivatsa <anusha.sriva...@intel.com>
Signed-off-by: Daniele Ceraolo Spurio <daniele.ceraolospu...@intel.com>
---
drivers/gpu/drm/i915/intel_guc_loader.c | 4
drivers/gpu/drm/i915/inte
On 24/03/17 18:30, Michel Thierry wrote:
For watchdog / media reset, the firmware must know the address of the shared
data page (the first page of the default context).
This information should be in DWORD 9 of the GUC_CTL structure.
Signed-off-by: Michel Thierry
of the default ctx as we needed for
suspend/resume/reset (Daniele).
Cc: Chris Wilson <ch...@chris-wilson.co.uk>
Cc: Daniele Ceraolo Spurio <daniele.ceraolospu...@intel.com>
Signed-off-by: Michel Thierry <michel.thie...@intel.com>
Reviewed-by: Daniele Ceraolo Spurio <daniele.ce
urrent value from RING_MODE reg instead; no need to preserve
head/tail either, be extra paranoid and save whitelisted registers (Daniele).
Cc: Daniele Ceraolo Spurio <daniele.ceraolospu...@intel.com>
Signed-off-by: Arun Siluvery <arun.siluv...@linux.intel.com>
Signed-off-by: J
<michal.wajdec...@intel.com>
Cc: Daniele Ceraolo Spurio <daniele.ceraolospu...@intel.com>
Cc: Chris Wilson <ch...@chris-wilson.co.uk>
Acked-by: Daniele Ceraolo Spurio <daniele.ceraolospu...@intel.com>
-Daniele
___
Intel-gf
On 06/07/17 11:51, Michel Thierry wrote:
On 7/6/2017 10:59 AM, Chris Wilson wrote:
> If there are no conflicts, then just skip the patch, and really there
> shouldn't be since the writes we care about are ordered and the writes
> we don't, we don't. We will need to move to per-context hwsp
On 12/07/17 15:58, Chris Wilson wrote:
The engine provides also provides mirror of the CSB write pointer in the
HWSP, but not of our read pointer. To take advantage of this we need to
remember where we read up to on the last interrupt and continue off from
there. This poses a problem following
the new macros to prevent abusive overuse of the old ones (Chris).
Reported-by: Daniele Ceraolo Spurio <daniele.ceraolospu...@intel.com>
Signed-off-by: Michel Thierry <michel.thie...@intel.com>
Cc: Chris Wilson <ch...@chris-wilson.co.uk>
Cc: Daniele Ceraolo Spurio <daniele.ce
Hi,
I'm a bit late to the party, but as I already mentioned at [1] the CNL
render ctx size is smaller than the gen9 one according to the specs. The
ctx format in the specs contains 16752 dwords, i.e. 17 pages, to which
we need to add 1 page for the PPHWSP, so 18 pages total. If we re-use
the
below 0x30 in an area we don't access,
but that is not guaranteed to be true for future platform.
To avoid the conflict, instead of re-using the PPHWSP of the kernel
ctx we can allocate a separate page for the HWSP like what happens for
pre-execlists platform.
Signed-off-by: Daniele Ceraolo Spurio
On 26/04/17 02:57, Zhi Wang wrote:
Uh...sorry for not mentioning that before:), and stolen memory is not my
business. :(
Actually we root-caused it.
This is how we found this case:
The story is W driver directly allocated the ring buffer after the
context image, and the context image size
gt;
Cc: Joonas Lahtinen <joonas.lahti...@linux.intel.com>
Cc: Daniele Ceraolo Spurio <daniele.ceraolospu...@intel.com>
Reviewed-by: Daniele Ceraolo Spurio <daniele.ceraolospu...@intel.com>
___
Intel-gfx mailing list
Intel-gfx@lis
Wajdeczko <michal.wajdec...@intel.com>
Cc: Oscar Mateo <oscar.ma...@intel.com>
Cc: Michel Thierry <michel.thie...@intel.com>
Cc: Daniele Ceraolo Spurio <daniele.ceraolospu...@intel.com>
---
drivers/gpu/drm/i915/intel_guc_ct.c | 14 +++---
drivers/gpu/drm/i9
On 07/08/17 09:14, Michal Wajdeczko wrote:
During debug we may want to investigate all communication
from the Guc. Add proper tracing macros in debug config.
v2: convert remaining DRM_DEBUG into new CT_DEBUG (Michal)
Signed-off-by: Michal Wajdeczko
---
On 07/08/17 09:14, Michal Wajdeczko wrote:
Flags bits are different in G2H message.
Signed-off-by: Michal Wajdeczko <michal.wajdec...@intel.com>
Cc: Daniele Ceraolo Spurio <daniele.ceraolospu...@intel.com>
Cc: Oscar Mateo <oscar.ma...@intel.com>
Cc: Kelvin Gardiner <kelv
On 05/05/17 16:23, Daniele Ceraolo Spurio wrote:
We're currently deleting the GuC logs if the FW fails to load, but those
are still useful to understand why the loading failed. Keeping the
object around allows us to access them after driver load is completed.
v2: keep the object around
eo <oscar.ma...@intel.com>
Cc: Michal Wajdeczko <michal.wajdec...@intel.com>
Signed-off-by: Daniele Ceraolo Spurio <daniele.ceraolospu...@intel.com>
---
drivers/gpu/drm/i915/i915_debugfs.c | 35 ++-
drivers/gpu/drm/i915/i915_drv.c | 3 +++
driver
On 15/05/17 10:41, Michal Wajdeczko wrote:
On Mon, May 15, 2017 at 09:35:30AM -0700, Daniele Ceraolo Spurio wrote:
We're currently deleting the GuC logs if the FW fails to load, but those
are still useful to understand why the loading failed. Keeping the
object around allows us to access them
to intel_guc, move functions to intel_uc.c, don't
clear obj on new GuC load, free object only if enable_guc_loading
is set (Michal)
Cc: Chris Wilson <ch...@chris-wilson.co.uk>
Cc: Oscar Mateo <oscar.ma...@intel.com>
Cc: Michal Wajdeczko <michal.wajdec...@intel.com>
Signed-off-b
On 22/05/17 11:21, Patchwork wrote:
== Series Details ==
Series: drm/i915/guc: capture GuC logs if FW fails to load (rev6)
URL : https://patchwork.freedesktop.org/series/23982/
State : warning
== Summary ==
Series 23982v6 drm/i915/guc: capture GuC logs if FW fails to load
son.co.uk>
Cc: Daniele Ceraolo Spurio <daniele.ceraolospu...@intel.com>
Cc: Jeff McGee <jeff.mc...@intel.com>
Cc: Michal Wajdeczko <michal.wajdec...@intel.com>
Cc: Oscar Mateo <oscar.ma...@intel.com>
Signed-off-by: Michał Winiarski <michal.winiar...@intel.com>
---
d
bert.beck...@intel.com>
Daniele Ceraolo Spurio <daniele.ceraolospu...@intel.com>
v2: use gem_object_pin_map (Chris)
don't use DEBUG_RATELIMITED (Chris)
don't track action stats (Chris)
simplify next fence (Chris)
use READ_ONCE (Chris)
move blob allocation to new functi
+
+static int ctch_init(struct intel_guc *guc,
+struct intel_guc_ct_channel *ctch)
+{
+ struct i915_vma *vma;
+ void *blob;
+ int err;
+ int i;
+
+ GEM_BUG_ON(ctch->vma);
+
+ /* We allocate 1 page to hold both descriptors and both
On 22/05/17 12:56, Michal Wajdeczko wrote:
On Mon, May 22, 2017 at 10:50:28AM -0700, Daniele Ceraolo Spurio wrote:
We're currently deleting the GuC logs if the FW fails to load, but those
are still useful to understand why the loading failed. Keeping the
object around allows us to access them
On 18/05/17 10:08, Michal Wajdeczko wrote:
On Tue, May 16, 2017 at 02:52:53PM -0700, Daniele Ceraolo Spurio wrote:
We're currently deleting the GuC logs if the FW fails to load, but those
are still useful to understand why the loading failed. Keeping the
object around allows us to access them
ports. We also need to stop pretending
that we're doing "lite-restore" in GuC submission (we would create a
workitem each time we hit this condition).
v2: Also coalesce when replaying on reset (Daniele)
Cc: Chris Wilson <ch...@chris-wilson.co.uk>
Cc: Daniele Ceraolo Spurio <
static void nested_enable_signaling(struct drm_i915_gem_request *rq)
@@ -1223,6 +1155,9 @@ int i915_guc_submission_enable(struct drm_i915_private
*dev_priv)
enum intel_engine_id id;
int err;
+ BUILD_BUG_ON(ARRAY_SIZE(engine->execlist_port) *
+
;ch...@chris-wilson.co.uk>
Cc: Oscar Mateo <oscar.ma...@intel.com>
Cc: Michal Wajdeczko <michal.wajdec...@intel.com>
Signed-off-by: Daniele Ceraolo Spurio <daniele.ceraolospu...@intel.com>
---
drivers/gpu/drm/i915/i915_debugfs.c | 38 +++---
c: Daniele Ceraolo Spurio <daniele.ceraolospu...@intel.com>
Cc: Joonas Lahtinen <joonas.lahti...@linux.intel.com>
Cc: Oscar Mateo <oscar.ma...@intel.com>
Signed-off-by: Michel Thierry <michel.thie...@intel.com>
---
Reviewed-by: Daniele Ceraolo Spurio <daniele.cera
You'll also need to change the order of operations in intel_uc_fini_hw
to make sure guc_free_load_err_log is called before the
i915.enable_guc_loading check, because that log exists exactly when GuC
loading failed.
Thanks,
Daniele
On 02/06/17 16:46, Michel Thierry wrote:
And prevent calling
On 04/05/17 14:26, Srivatsa, Anusha wrote:
-Original Message-
From: Intel-gfx [mailto:intel-gfx-boun...@lists.freedesktop.org] On Behalf Of
Daniele Ceraolo Spurio
Sent: Thursday, May 4, 2017 11:52 AM
To: intel-gfx@lists.freedesktop.org
Subject: [Intel-gfx] [RFC] drm/i915/guc
On 04/05/17 14:31, Chris Wilson wrote:
On Thu, May 04, 2017 at 09:26:35PM +, Srivatsa, Anusha wrote:
+void i915_guc_load_error_log_capture(struct drm_i915_private *i915) {
+ void *log, *buf;
+ struct i915_vma *vma = i915->guc.log.vma;
+
+ if
the logs in the gpu_error struct (Chris)
add a check on guc_log_level to avoid snapshotting empty logs
Cc: Chris Wilson <ch...@chris-wilson.co.uk>
Cc: Oscar Mateo <oscar.ma...@intel.com>
Cc: Michal Wajdeczko <michal.wajdec...@intel.com>
Signed-off-by: Daniele Ceraolo Spurio &
com>
Signed-off-by: Daniele Ceraolo Spurio <daniele.ceraolospu...@intel.com>
---
drivers/gpu/drm/i915/i915_debugfs.c | 35 ++-
drivers/gpu/drm/i915/i915_drv.c | 3 +++
drivers/gpu/drm/i915/intel_guc_log.c | 17 +
drivers/gpu/drm/i
On 02/05/17 05:39, Michal Wajdeczko wrote:
Prepare for alternate GuC notification mechanism.
Signed-off-by: Michal Wajdeczko <michal.wajdec...@intel.com>
Cc: Joonas Lahtinen <joonas.lahti...@linux.intel.com>
Cc: Daniele Ceraolo Spurio <daniele.ceraolospu...@intel.com>
--
eraolo Spurio <daniele.ceraolospu...@intel.com>
Cc: Daniele Ceraolo Spurio <daniele.ceraolospu...@intel.com>
Cc: Joonas Lahtinen <joonas.lahti...@linux.intel.com>
---
drivers/gpu/drm/i915/intel_uc.c | 42 +
drivers/gpu/drm/i915/intel_uc
deczko <michal.wajdec...@intel.com>
Signed-off-by: Daniele Ceraolo Spurio <daniele.ceraolospu...@intel.com>
---
drivers/gpu/drm/i915/i915_debugfs.c | 36 ---
drivers/gpu/drm/i915/i915_drv.c | 3 +++
drivers/gpu/drm/i915/i915_drv.h | 6 ++
dr
bert.beck...@intel.com>
Daniele Ceraolo Spurio <daniele.ceraolospu...@intel.com>
v2: use gem_object_pin_map (Chris)
don't use DEBUG_RATELIMITED (Chris)
don't track action stats (Chris)
simplify next fence (Chris)
use READ_ONCE (Chris)
move blob allocation to new functio
;ch...@chris-wilson.co.uk>
Cc: Daniele Ceraolo Spurio <daniele.ceraolospu...@intel.com>
Cc: Michal Wajdeczko <michal.wajdec...@intel.com>
Cc: Oscar Mateo <oscar.ma...@intel.com>
Signed-off-by: Michał Winiarski <michal.winiar...@intel.com>
Reviewed-by: Chris Wilson <ch...@
uC objects
Cc: Michal Wajdeczko <michal.wajdec...@intel.com>
Cc: Oscar Mateo <oscar.ma...@intel.com>
Cc: Anusha Srivatsa <anusha.sriva...@intel.com>
Cc: Daniele Ceraolo Spurio <daniele.ceraolospu...@intel.com>
Signed-off-by: Sujaritha Sundaresan <sujaritha.sundare...@i
On 06/10/17 05:35, Michał Winiarski wrote:
On Thu, Oct 05, 2017 at 05:02:39PM +, Daniele Ceraolo Spurio wrote:
On 05/10/17 02:33, Chris Wilson wrote:
Quoting Michał Winiarski (2017-10-05 10:13:40)
We're using first page of kernel context state to share data with GuC,
let's precompute
The feature was never merged and there has been no progress in the
last year. The subtests are currently skipping on all platforms by
checking a field in the get_aperture ioctl structure that doesn't
exist in the kernel version of the struct.
Signed-off-by: Daniele Ceraolo Spurio
1 - 100 of 1795 matches
Mail list logo