On 4/30/24 16:53, Jani Nikula wrote:
On Tue, 30 Apr 2024, Danilo Krummrich wrote:
After dropping linux/debugfs.h include from drm/drm_print.h the following
files in i915 miss the linux/debugfs.h include: i915_debugfs.c,
i915_debugfs_params.c and i915_gpu_error.c.
Add the include to fix
On 4/30/24 15:18, Chaitanya Kumar Borah wrote:
Add the missing break statement that causes the following build error
CC [M] drivers/gpu/drm/i915/display/intel_display_device.o
../drivers/gpu/drm/nouveau/nvkm/subdev/gsp/r535.c: In function
‘build_registry’:
t;drm/print: drop include debugfs.h and include where
needed")
Closes:
https://lore.kernel.org/oe-kbuild-all/202404260726.nyguvl59-...@intel.com/
Signed-off-by: Danilo Krummrich
---
drivers/gpu/drm/i915/i915_debugfs.c| 1 +
drivers/gpu/drm/i915/i915_debugfs_params.c | 1 +
drivers/gp
On 4/30/24 15:06, Lucas De Marchi wrote:
On Fri, Apr 26, 2024 at 06:08:19PM GMT, Danilo Krummrich wrote:
On 4/25/24 18:38, Timur Tabi wrote:
On Thu, 2024-04-25 at 15:22 +0200, Danilo Krummrich wrote:
+ size_t length;
+
+ /* Remove any whitespace from the parameter string
Now that we have vmemdup_array_user(), make use of it.
Signed-off-by: Danilo Krummrich
---
drivers/gpu/drm/nouveau/nouveau_drv.h | 5 +
1 file changed, 1 insertion(+), 4 deletions(-)
diff --git a/drivers/gpu/drm/nouveau/nouveau_drv.h
b/drivers/gpu/drm/nouveau/nouveau_drv.h
index
nouveau_uvmm_ioctl_vm_init() already has a pointer to struct drm_device,
no need to derive another one from struct drm_file.
Fixes: 266f7618e761 ("drm/nouveau: separately allocate struct nouveau_uvmm")
Signed-off-by: Danilo Krummrich
---
drivers/gpu/drm/nouveau/nouveau_uvmm.c | 3 +
On 4/11/24 03:15, Dave Airlie wrote:
From: Dave Airlie
Running a lot of VK CTS in parallel against nouveau, once every
few hours you might see something like this crash.
BUG: kernel NULL pointer dereference, address: 0008
PGD 800114e6e067 P4D 800114e6e067 PUD 109046067 PMD
On 4/11/24 13:08, Mikhail Kobuk wrote:
When Output Resource (dcb->or) value is assigned in
fabricate_dcb_output(), there may be out of bounds access to
dac_users array in case dcb->or is zero because ffs(dcb->or) is
used as index there.
The 'or' argument of fabricate_dcb_output() must be
On 4/10/24 17:39, Mikhail Kobuk wrote:
On 08/04/2024 16:23, Danilo Krummrich wrote:
On 4/5/24 22:05, Lyude Paul wrote:
On Fri, 2024-04-05 at 17:53 +0200, Danilo Krummrich wrote:
On 3/31/24 08:45, Mikhail Kobuk wrote:
Output Resource (dcb->or) value is not guaranteed to be non-zero
(
On 4/9/24 10:27, Lucas Stach wrote:
Am Dienstag, dem 09.04.2024 um 10:34 +1000 schrieb Dave Airlie:
From: Dave Airlie
Running a lot of VK CTS in parallel against nouveau, once every
few hours you might see something like this crash.
BUG: kernel NULL pointer dereference, address:
On 4/5/24 22:05, Lyude Paul wrote:
On Fri, 2024-04-05 at 17:53 +0200, Danilo Krummrich wrote:
On 3/31/24 08:45, Mikhail Kobuk wrote:
Output Resource (dcb->or) value is not guaranteed to be non-zero
(i.e.
in drivers/gpu/drm/nouveau/nouveau_bios.c, in
'fabricate_dcb_encoder_table()'
'
uveau/gsp: don't check devinit disable on GSP.")
Signed-off-by: Dave Airlie
Reviewed-by: Danilo Krummrich
---
drivers/gpu/drm/nouveau/nvkm/subdev/devinit/gm107.c | 12
drivers/gpu/drm/nouveau/nvkm/subdev/devinit/r535.c | 1 +
2 files changed, 9 insertions(+), 4 deletion
On 4/4/24 18:02, Arnd Bergmann wrote:
From: Arnd Bergmann
Calling a function through an incompatible pointer type causes breaks
kcfi, so clang warns about the assignment:
drivers/gpu/drm/nouveau/nvkm/subdev/bios/shadowof.c:73:10: error: cast from
'void (*)(const void *)' to 'void (*)(void
ode:
strings = (char *)>entries[NV_GSP_REG_NUM_ENTRIES];
...
memcpy(strings, r535_registry_entries[i].name, name_len);
Signed-off-by: Kees Cook
Applied to drm-misc-fixes, thanks!
---
Cc: Karol Herbst
Cc: Lyude Paul
Cc: Danilo Krummrich
Cc: David Airl
Hi Easwar,
On 3/29/24 18:00, Easwar Hariharan wrote:
I2C v7, SMBus 3.2, and I3C specifications have replaced "master/slave"
with more appropriate terms. Inspired by and following on to Wolfram's
series to fix drivers/i2c/[1], fix the terminology for users of
I2C_ALGOBIT bitbanging interface,
On 3/31/24 08:45, Mikhail Kobuk wrote:
Output Resource (dcb->or) value is not guaranteed to be non-zero (i.e.
in drivers/gpu/drm/nouveau/nouveau_bios.c, in 'fabricate_dcb_encoder_table()'
'dcb->or' is assigned value '0' in call to 'fabricate_dcb_output()').
I don't really know much about the
the prev + unmap range calcs to use start/end and map back to addr/range.
I like how using start/end instead fixes the issue and keeps it simple.
Signed-off-by: Dave Airlie
Fixes: b88baab82871 ("drm/nouveau: implement new VM_BIND uAPI")
Cc: Danilo Krummrich
Applied the patch to drm-
On 3/15/24 10:09, Colin Ian King wrote:
There is a statement with two semicolons. Remove the second one, it
is redundant.
Signed-off-by: Colin Ian King
Applied, thanks!
---
drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git
Hi all,
In this mail I briefly want to announce the Nova project and subsequently talk
about the first
efforts taken in order to upstream required Rust abstractions:
We just started to work on Nova, a Rust-based GSP-only driver for Nvidia GPUs.
Nova, in the long
term, is intended to serve as
On 3/19/24 18:38, Timothy Maden wrote:
Hello
Can this be merged please ?
Applied this to drm-misc-next-fixes a couple of hours ago.
- Danilo
On 3/14/24 03:45, Dave Airlie wrote:
From: Dave Airlie
GSP should be handling this and I can see no evidence in opengpu
driver that this register
c and nouveau maintainers, ack for merging this via drm-intel,
please?
Sure, please go ahead.
Thanks,
Danilo
BR,
Jani.
v2:
- Rename enumerators (Ville)
Cc: Arun R Murthy
Cc: Ville Syrjälä
Cc: Karol Herbst
Cc: Lyude Paul
Cc: Danilo Krummrich
Reviewed-by: Ville Syrjälä
Signed-off-by: Jani
Cc: Ville Syrjälä
Cc: Karol Herbst
Cc: Lyude Paul
Cc: Danilo Krummrich
Reviewed-by: Ville Syrjälä
Signed-off-by: Jani Nikula
Acked-by: Danilo Krummrich
---
drivers/gpu/drm/display/drm_dp_mst_topology.c | 20 ++--
drivers/gpu/drm/i915/display/intel_dp.c | 4 ++--
driver
evict device private memory during release")
Suggested-by: Danilo Krummrich
Signed-off-by: Duoming Zhou
Applied to drm-misc-fixes, thanks!
---
Changes in v3:
- Switch kcalloc() to kvcalloc().
drivers/gpu/drm/nouveau/nouveau_dmem.c | 12 ++--
1 file changed, 6 insert
Hi Duoming,
thanks for sending a V2.
On 3/5/24 15:39, Duoming Zhou wrote:
The kcalloc() in nouveau_dmem_evict_chunk() will return null if
the physical memory has run out. As a result, if we dereference
src_pfns, dst_pfns or dma_addrs, the null pointer dereference bugs
will happen.
Moreover,
Hi Duoming,
On 3/3/24 08:53, Duoming Zhou wrote:
The kcalloc() in nouveau_dmem_evict_chunk() will return null if
the physical memory has run out. As a result, if we dereference
src_pfns, dst_pfns or dma_addrs, the null pointer dereference bugs
will happen.
This patch uses stack variables to
On 3/4/24 19:31, Karol Herbst wrote:
Those are already de-facto UAPI, so let's just move it into the uapi
header.
Signed-off-by: Karol Herbst
Reviewed-by: Danilo Krummrich
---
drivers/gpu/drm/nouveau/nouveau_abi16.c | 20 +++-
drivers/gpu/drm/nouveau/nouveau_abi16.h
other nouveau ioctl call.
Fixes: b88baab82871 ("drm/nouveau: implement new VM_BIND uAPI")
Cc: Danilo Krummrich
Signed-off-by: Karol Herbst
Should add a stable tag for that one, otherwise:
Reviewed-by: Danilo Krummrich
---
drivers/gpu/drm/nouveau/nouveau_gem.c | 2 +-
1 file
On 3/4/24 18:55, Greg KH wrote:
On Mon, Mar 04, 2024 at 06:01:46PM +0100, Danilo Krummrich wrote:
Cc: # v6.7 only
You say 6.7 only, but this commit is in 6.6, so why not 6.6 also?
Good catch, I was sure I originally merged this for 6.7. This fix should indeed
be applied to 6.6 as well
rolherbst/a20eb0f937a06ed6aabe2ac2ca3d11b5/raw/9cd8b1dc5894872d0eeebbee3dd0fdd28bb576bc/gistfile1.txt
Fixes: b88baab82871 ("drm/nouveau: implement new VM_BIND uAPI")
Signed-off-by: Danilo Krummrich
---
drivers/gpu/drm/nouveau/nouveau_drm.c | 5 +++--
1 file changed, 3 insertions(+), 2 delet
On 3/1/24 04:42, Dave Airlie wrote:
From: Dave Airlie
It appears the client object tree has no locking unless I've missed
something else. Fix races around adding/removing client objects,
mostly vram bar mappings.
4562.099306] general protection fault, probably for non-canonical address
On Fri, Feb 23, 2024 at 10:14:53AM +1000, Dave Airlie wrote:
> On Fri, 23 Feb 2024 at 00:45, Danilo Krummrich wrote:
> >
> > Using the kernel global workqueue to signal fences can lead to
> > unexpected deadlocks. Some other work (e.g. from a different driver)
> > c
the fence from running.
While this seems fairly unlikely, it's potentially exploitable.
Fixes: 39126abc5e20 ("nouveau: offload fence uevents work to workqueue")
Signed-off-by: Danilo Krummrich
---
drivers/gpu/drm/nouveau/nouveau_drm.c | 13 +++--
drivers/gpu/drm/nouveau/nouveau_dr
On 2/20/24 20:45, Timur Tabi wrote:
On Mon, 2024-02-19 at 10:32 +0100, Danilo Krummrich wrote:
Looks like I spoke too soon, I just hit the problem with the drm-next tree.
Did you apply the patch to drm-next?
Ugh, you're right. I don't how I got confused, but I could have sworn that I
saw
On 2/16/24 17:41, Daniel Vetter wrote:
On Tue, Feb 13, 2024 at 06:39:20PM +0100, Danilo Krummrich wrote:
On 2/9/24 19:52, Daniel Vetter wrote:
On Fri, Feb 09, 2024 at 06:41:32PM +0100, Danilo Krummrich wrote:
On 2/6/24 15:03, Daniel Vetter wrote:
On Mon, Feb 05, 2024 at 11:00:04PM +0100
On 2/15/24 00:48, Timur Tabi wrote:
On Fri, 2024-02-02 at 17:14 +, Timur Tabi wrote:
On Fri, 2024-02-02 at 01:05 +0100, Danilo Krummrich wrote:
nouveau_abi16_ioctl_channel_alloc() and nouveau_cli_init() simply call
their corresponding *_fini() counterpart. This can lead
On 2/13/24 19:09, Dan Carpenter wrote:
If gf100_bar_new_() fails then "bar" is not initialized.
Fixes: 5bf0257136a2 ("drm/nouveau/mmu/r535: initial support")
Signed-off-by: Dan Carpenter
Applied to drm-misc-fixes, thanks!
---
It looks like this was intended to handle a failure from the
On 2/13/24 10:57, Arnd Bergmann wrote:
From: Arnd Bergmann
clang-16 warns about casting between incompatible function types:
drivers/gpu/drm/nouveau/nvkm/subdev/bios/shadow.c:161:10: error: cast from
'void (*)(const struct firmware *)' to 'void (*)(void *)' converts to
incompatible function
On 2/13/24 18:10, Timur Tabi wrote:
On Tue, 2024-02-13 at 17:57 +0100, Danilo Krummrich wrote:
+ struct debugfs_blob_wrapper blob_init;
+ struct debugfs_blob_wrapper blob_intr;
+ struct debugfs_blob_wrapper blob_rm;
+ struct debugfs_blob_wrapper blob_pmu;
+ struct
On 2/9/24 19:52, Daniel Vetter wrote:
On Fri, Feb 09, 2024 at 06:41:32PM +0100, Danilo Krummrich wrote:
On 2/6/24 15:03, Daniel Vetter wrote:
On Mon, Feb 05, 2024 at 11:00:04PM +0100, Danilo Krummrich wrote:
On 2/5/24 22:08, Dave Airlie wrote:
On Tue, 6 Feb 2024 at 02:22, Danilo Krummrich
Hi Christian,
What's the status of this effort? Was there ever a follow-up?
- Danilo
On 11/16/23 23:23, Danilo Krummrich wrote:
Hi Christian,
Thanks for sending an update of this patch!
On Thu, Nov 16, 2023 at 03:15:46PM +0100, Christian König wrote:
Start to improve the scheduler document
On 2/12/24 22:15, Timur Tabi wrote:
The LOGINIT, LOGINTR, LOGRM, and LOGPMU buffers are circular buffers
that have printf-like logs from GSP-RM and PMU encoded in them.
LOGINIT, LOGINTR, and LOGRM are allocated by Nouveau and their DMA
addresses are passed to GSP-RM during initialization. The
On 2/12/24 22:15, Timur Tabi wrote:
Add the NVreg_RegistryDwords command line parameter, which allows
specifying additional registry keys to be sent to GSP-RM. This
allows additional configuration, debugging, and experimentation
with GSP-RM, which uses these keys to alter its behavior.
Note
On 2/12/24 12:22, Arnd Bergmann wrote:
From: Arnd Bergmann
The conversion to kvcalloc() mixed up the object size and count
arguments, causing a warning:
drivers/gpu/drm/nouveau/nouveau_svm.c: In function
'nouveau_svm_fault_buffer_ctor':
drivers/gpu/drm/nouveau/nouveau_svm.c:1010:40: error:
On 2/6/24 15:03, Daniel Vetter wrote:
On Mon, Feb 05, 2024 at 11:00:04PM +0100, Danilo Krummrich wrote:
On 2/5/24 22:08, Dave Airlie wrote:
On Tue, 6 Feb 2024 at 02:22, Danilo Krummrich wrote:
On 1/29/24 02:50, Dave Airlie wrote:
From: Dave Airlie
This should break the deadlock between
On 2/5/24 22:08, Dave Airlie wrote:
On Tue, 6 Feb 2024 at 02:22, Danilo Krummrich wrote:
On 1/29/24 02:50, Dave Airlie wrote:
From: Dave Airlie
This should break the deadlock between the fctx lock and the irq lock.
This offloads the processing off the work from the irq into a workqueue
On 1/30/24 04:26, Dave Airlie wrote:
From: Dave Airlie
Timur pointed this out before, and it just slipped my mind,
but this might help some things work better, around pcie power
management.
Fixes: 8d55b0a940bb ("nouveau/gsp: add some basic registry entries.")
Signed-off-by: Dave Airlie
On 1/29/24 02:50, Dave Airlie wrote:
From: Dave Airlie
This should break the deadlock between the fctx lock and the irq lock.
This offloads the processing off the work from the irq into a workqueue.
Signed-off-by: Dave Airlie
Nouveau's scheduler uses a dedicated wq, hence from this
On 2/5/24 10:03, Boris Brezillon wrote:
+Danilo for the panthor gpuvm-needs update.
On Sun, 4 Feb 2024 09:14:44 +0800 (CST)
"Andy Yan" wrote:
Hi Boris：
I saw this warning sometimes（Run on a armbain based bookworm），not sure is a
know issue or something else。
[15368.293031]
On 2/2/24 18:14, Timur Tabi wrote:
On Fri, 2024-02-02 at 01:05 +0100, Danilo Krummrich wrote:
nouveau_abi16_ioctl_channel_alloc() and nouveau_cli_init() simply call
their corresponding *_fini() counterpart. This can lead to
nouveau_sched_fini() being called without struct nouveau_sched ever
Omit to create scheduler instances when using the legacy uAPI. When
using the legacy NOUVEAU_GEM_PUSHBUF ioctl no scheduler instance is
required, hence omit creating scheduler instances in
nouveau_abi16_ioctl_channel_alloc().
Signed-off-by: Danilo Krummrich
---
drivers/gpu/drm/nouveau
rg/nouveau/20240131213917.1545604-1-tt...@nvidia.com/
Signed-off-by: Danilo Krummrich
---
drivers/gpu/drm/nouveau/nouveau_abi16.c | 10 ---
drivers/gpu/drm/nouveau/nouveau_abi16.h | 2 +-
drivers/gpu/drm/nouveau/nouveau_drm.c | 7 +++--
drivers/gpu/drm/nouveau/nouveau_drv.h | 2 +-
d
-Wstringop-overflow (already enabled by default)
Cc: David Airlie
Cc: Daniel Vetter
Cc: Maarten Lankhorst
Cc: Maxime Ripard
Cc: Thomas Zimmermann
Cc: Alex Deucher
Cc: Christian König
Cc: Pan, Xinhui
Cc: Karol Herbst
Cc: Lyude Paul
Cc: Danilo Krummrich
Cc: Rob Clark
Cc: Abhinav Kumar
Cc
On 1/10/24 18:39, Jani Nikula wrote:
Fix the W=1 warning -Wunused-but-set-variable.
Cc: Karol Herbst
Cc: Lyude Paul
Cc: Danilo Krummrich
Cc: nouv...@lists.freedesktop.org
Signed-off-by: Jani Nikula
Reviewed-by: Danilo Krummrich
---
drivers/gpu/drm/nouveau/nouveau_svm.c | 10
On 1/10/24 18:39, Jani Nikula wrote:
Fix the W=1 warning -Wunused-but-set-variable.
Cc: Karol Herbst
Cc: Lyude Paul
Cc: Danilo Krummrich
Cc: nouv...@lists.freedesktop.org
Signed-off-by: Jani Nikula
Reviewed-by: Danilo Krummrich
---
drivers/gpu/drm/nouveau/nvkm/subdev/acr/lsfw.c | 3
will give up this approach, unless Matt or Brian can see a way.
A few replies inline @Welty, Brian I had more thoughts inline to one of
your original question
-Original Message-
From: Danilo Krummrich
Sent: Tuesday, January 23, 2024 6:57 PM
To: Zeng, Oak ; Christian König
; Dave
the same thoughts after reading your
initial mail.
Please find some more comments below.
-Original Message-
From: Christian König
Sent: Tuesday, January 23, 2024 6:13 AM
To: Zeng, Oak ; Danilo Krummrich ;
Dave Airlie ; Daniel Vetter
Cc: Welty, Brian ; dri-devel@lists.freedesktop.org;
intel
On 1/16/24 13:31, Dan Carpenter wrote:
On Tue, Jan 16, 2024 at 11:16:09AM +, Colin Ian King wrote:
The variable ret is being assigned a value but it isn't being
read afterwards. The assignment is redundant and so ret can be
removed.
Cleans up clang scan build warning:
warning: Although the
On 1/19/24 10:05, Thomas Hellström wrote:
The relatively recently introduced drm/exec utility was using uint32_t
in its interface, which was then also carried over to drm/gpuvm.
Prefer u32 in new code and update drm/exec and drm/gpuvm accordingly.
Cc: Christian König
Cc: Danilo Krummrich
On 1/9/24 07:24, yang.gua...@zte.com.cn wrote:
From: Chen Haonan
Use kmemdup() helper instead of open-coding to
simplify the code.
Signed-off-by: Chen Haonan
Reviewed-by: Yang Guang
Applied to drm-misc-next, thanks!
---
drivers/gpu/drm/nouveau/nvif/outp.c | 3 +--
1 file changed, 1
On 1/9/24 10:59, Jani Nikula wrote:
On Mon, 08 Jan 2024, Danilo Krummrich wrote:
On 1/4/24 21:16, Jani Nikula wrote:
Including drm_edid.h from nouveau_connector.h causes the rebuild of 15
files when drm_edid.h is modified, while there are only a few files that
actually need to include
On 12/25/23 07:51, Vegard Nossum wrote:
As of commit b77fdd6a48e6 ("scripts/kernel-doc: restore warning for
Excess struct/union"), we see the following warnings when running 'make
htmldocs':
./include/uapi/drm/nouveau_drm.h:292: warning: Excess struct member
'DRM_NOUVEAU_VM_BIND_OP_MAP'
On 1/4/24 21:16, Jani Nikula wrote:
Including drm_edid.h from nouveau_connector.h causes the rebuild of 15
files when drm_edid.h is modified, while there are only a few files that
actually need to include drm_edid.h.
Cc: Karol Herbst
Cc: Lyude Paul
Cc: Danilo Krummrich
Cc: nouv
* Sets up registers for the given mode/adjusted_mode pair.
crtc.c:629: warning: missing initial short description on line:
* Sets up registers for the given mode/adjusted_mode pair.
Signed-off-by: Randy Dunlap
Cc: Karol Herbst
Cc: Lyude Paul
Cc: Danilo Krummrich
Cc: nouv...@lists.freedes
totype for INIT_DONE(). Prototype was for
init_done() instead
Signed-off-by: Randy Dunlap
Cc: Karol Herbst
Cc: Lyude Paul
Cc: Danilo Krummrich
Cc: dri-devel@lists.freedesktop.org
Cc: nouv...@lists.freedesktop.org
Cc: Maarten Lankhorst
Cc: Maxime Ripard
Cc: Thomas Zimmermann
Cc: David Airli
Hi Wang,
there is another patch [1] to fix this, which already made it upstream.
- Danilo
[1]
https://patchwork.freedesktop.org/patch/msgid/20231122004926.84933-1-yang@linux.alibaba.com
On 12/15/23 11:02, Wang Jinchao wrote:
rm second including of chid.h
Signed-off-by: Wang Jinchao
Hi Yang,
On 12/14/23 13:03, yang.gua...@zte.com.cn wrote:
From: Yang Guang
Use kmemdup() helper instead of open-coding to
simplify the code.
Signed-off-by: Chen Haonan
Please add your "Signed-off-by" tag to this patch. If the above is intended to
indicate
that Chan was involved in
On Tue, Dec 05, 2023 at 04:35:00PM +0100, Maxime Ripard wrote:
> Hi,
>
> On Wed, Nov 29, 2023 at 11:07:59PM +0100, Danilo Krummrich wrote:
> > Some major GPUVM changes landed just before v8 of the PowerVR series. Since
> > v8
> > went in rather quickly (review proces
drm_gpuvm_bo_put()
to report when the object is destroyed.
Signed-off-by: Boris Brezillon
Reviewed-by: Danilo Krummrich
---
drivers/gpu/drm/drm_gpuvm.c | 8 ++--
include/drm/drm_gpuvm.h | 2 +-
2 files changed, 7 insertions(+), 3 deletions(-)
diff --git a/drivers/gpu/drm/drm_gpuvm.c b/drivers
to remind users of
this function to ensure that args != NULL when calling this function
for a map operation as it was suggested by Danilo [1]
[1] https://lore.kernel.org/lkml/6a1ebcef-bade-45a0-9bd9-c05f0226e...@redhat.com
Suggested-by: Danilo Krummrich
Signed-off-by: Yuran Pereira
Applied to drm
Hi Abhinav,
Thanks for sending this follow-up patch.
On 11/26/23 15:57, Abhinav Singh wrote:
Fix a sparse warning with this message
"warning:dereference of noderef expression". In this context it means we
are dereferencing a __rcu tagged pointer directly.
We should not be directly
d GEM and VM related code")
Signed-off-by: Danilo Krummrich
---
drivers/gpu/drm/imagination/pvr_vm.c | 91 +++-
1 file changed, 36 insertions(+), 55 deletions(-)
diff --git a/drivers/gpu/drm/imagination/pvr_vm.c
b/drivers/gpu/drm/imagination/pvr_vm.c
index e0
Fall back to drm_exec_lock_obj() if num_fences is zero for the
drm_gpuvm_prepare_* function family.
Otherwise dma_resv_reserve_fences() would actually allocate slots even
though num_fences is zero.
Cc: Christian König
Signed-off-by: Danilo Krummrich
---
drivers/gpu/drm/drm_gpuvm.c | 43
drm_gpuvm_exec wrappers
- drop 3 patches which were applied already
Danilo Krummrich (2):
drm/gpuvm: fall back to drm_exec_lock_obj()
drm/imagination: vm: make use of GPUVM's drm_exec helper
drivers/gpu/drm/drm_gpuvm.c | 43 +++--
drivers/gpu/drm/imagination/pvr_vm.c | 91
On 11/29/23 02:06, Gustavo A. R. Silva wrote:
On 11/28/23 19:01, Danilo Krummrich wrote:
On 11/16/23 20:55, Timur Tabi wrote:
On Thu, 2023-11-16 at 20:45 +0100, Danilo Krummrich wrote:
As I already mentioned for Timur's patch [2], I'd prefer to get a fix
upstream
(meaning [1] in this case
On 11/27/23 13:56, Dan Carpenter wrote:
This unlock doesn't belong here and it leads to a double unlock in
the caller, r535_gsp_rpc_push().
Fixes: 176fdcbddfd2 ("drm/nouveau/gsp/r535: add support for booting GSP-RM")
Signed-off-by: Dan Carpenter
Good catch - applied to drm-misc-fixes.
-
On 11/16/23 20:55, Timur Tabi wrote:
On Thu, 2023-11-16 at 20:45 +0100, Danilo Krummrich wrote:
As I already mentioned for Timur's patch [2], I'd prefer to get a fix
upstream
(meaning [1] in this case). Of course, that's probably more up to Timur to
tell
if this will work out.
Don't count
On 11/28/23 11:47, Donald Robson wrote:
Hi Danilo,
Apologies - I guess I should have submitted a patch to handle zero fences in
your
locking functions with the final patch series.
On Sat, 2023-11-25 at 00:36 +0100, Danilo Krummrich wrote:
*** CAUTION: This email originates from a source
Use local references to other documentation where possible
(Bagas Sanjaya)
- General documentation fixes and typos (Danilo Krummrich and
Boris Brezillon)
- Improve the documentation around locks that need to be grabbed from the
dm-fence critical section (Boris Brezillon)
- Add more referen
Hi,
On 11/27/23 14:52, Christian König wrote:
Am 25.11.23 um 00:40 schrieb Danilo Krummrich:
Hi Christian,
do you think it makes sense to handle this in drm_exec_prepare_obj() or
even dma_resv_reserve_fences() even?
IIRC for drm_exec the discussion has gone a bit back and forth between
Hi Christian,
do you think it makes sense to handle this in drm_exec_prepare_obj() or
even dma_resv_reserve_fences() even?
- Danilo
On 11/25/23 00:36, Danilo Krummrich wrote:
Fall back to drm_exec_lock_obj() if num_fences is zero for the
drm_gpuvm_prepare_* function family.
Otherwise
d GEM and VM related code")
Signed-off-by: Danilo Krummrich
---
drivers/gpu/drm/imagination/pvr_vm.c | 16 +---
1 file changed, 5 insertions(+), 11 deletions(-)
diff --git a/drivers/gpu/drm/imagination/pvr_vm.c
b/drivers/gpu/drm/imagination/pvr_vm.c
index e0d74d9a6190..3f7888f5c
Fall back to drm_exec_lock_obj() if num_fences is zero for the
drm_gpuvm_prepare_* function family.
Otherwise dma_resv_reserve_fences() would actually allocate slots even
though num_fences is zero.
Cc: Christian König
Signed-off-by: Danilo Krummrich
---
drivers/gpu/drm/drm_gpuvm.c | 36
, this prevents
lifetime issues such as the VM being freed as long as drm_gpuvm_bo
structures still hold references to the VM.
Fixes: ff5f643de0bf ("drm/imagination: Add GEM and VM related code")
Signed-off-by: Danilo Krummrich
---
drivers/gpu/drm/imagination/pvr_vm.c | 11 ---
1 fi
the drm_gpuvm_prepare_* helpers useful for PowerVR.
- Danilo
Danilo Krummrich (5):
drm/imagination: vm: prevent duplicate drm_gpuvm_bo instances
drm/imagination: vm: check for drm_gpuvm_range_valid()
drm/imagination: vm: fix drm_gpuvm reference count
drm/gpuvm: fall back
Extend pvr_device_addr_and_size_are_valid() by the corresponding GPUVM
sanity checks. This includes a, previously missing, overflow check for
the base address and size of the requested mapping.
Fixes: ff5f643de0bf ("drm/imagination: Add GEM and VM related code")
Signed-off-by: Danilo
of struct drm_gpuvm_bo and hence leave us with duplicates.
Fixes: ff5f643de0bf ("drm/imagination: Add GEM and VM related code")
Signed-off-by: Danilo Krummrich
---
drivers/gpu/drm/imagination/pvr_vm.c | 10 ++
1 file changed, 6 insertions(+), 4 deletions(-)
diff --git a/drive
On 11/13/23 20:13, Abhinav Singh wrote:
This patch fixes a sparse warning with this message
"warning:dereference of noderef expression". In this context it means we
are dereferencing a __rcu tagged pointer directly.
We should not be directly dereferencing a rcu pointer. To get a normal
(non
Hi Christian,
Thanks for sending an update of this patch!
On Thu, Nov 16, 2023 at 03:15:46PM +0100, Christian König wrote:
> Start to improve the scheduler document. Especially document the
> lifetime of each of the objects as well as the restrictions around
> DMA-fence handling and userspace
Hi Gustavo,
On Thu, Nov 16, 2023 at 12:11:43PM -0600, Gustavo A. R. Silva wrote:
> Fake flexible arrays (zero-length and one-element arrays) are deprecated,
> and should be replaced by flexible-array members. So, replace
> zero-length array with a flexible-array member in `struct
>
On 11/16/23 12:48, Thomas Hellström wrote:
+Locks used and locking orders
+=
+
+One of the benefits of VM_BIND is that local GEM objects share the
gpu_vm's
+dma_resv object and hence the dma_resv lock. So even with a huge
+number of local GEM objects, only one lock
On 11/15/23 17:04, Thomas Hellström wrote:
Hi, Danilo,
On Wed, 2023-11-15 at 16:11 +0100, Danilo Krummrich wrote:
On Wed, Nov 15, 2023 at 01:49:37PM +0100, Thomas Hellström wrote:
Add the first version of the VM_BIND locking document which is
intended to be part of the xe driver upstreaming
On Wed, Nov 15, 2023 at 01:49:37PM +0100, Thomas Hellström wrote:
> Add the first version of the VM_BIND locking document which is
> intended to be part of the xe driver upstreaming agreement.
>
> The document describes and discuss the locking used during exec-
> functions, evicton and for
("drm/nouveau/gsp/r535: add support for booting GSP-RM")
Signed-off-by: Dan Carpenter
Reviewed-by: Danilo Krummrich
---
drivers/gpu/drm/nouveau/nvkm/subdev/gsp/r535.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/gsp
On 11/5/23 21:37, Dave Airlie wrote:
From: Dave Airlie
If we get back ENODEV don't fail load.
Maybe worth to note why this is OK in this case, might not be obvious
to future readers of the code.
Fixes: 15740541e8f0 ("drm/nouveau/devinit/tu102-: prepare for GSP-RM")
Maybe I'm missing
()
- Improved documentation
Changes since v6:
- Remove use of __always_inline
Signed-off-by: Donald Robson
Signed-off-by: Sarah Walker
Reviewed-by: Danilo Krummrich
Want me to apply the patch?
---
include/drm/drm_gpuvm.h | 28
1 file changed, 28 insertions
Hi Yuran,
On 10/26/23 19:03, Yuran Pereira wrote:
There are instances where the "args" argument passed to
nouveau_uvmm_sm_prepare() is NULL.
I.e. when nouveau_uvmm_sm_prepare() is called from
nouveau_uvmm_sm_unmap_prepare()
```
static int
nouveau_uvmm_sm_unmap_prepare(struct nouveau_uvmm
Make use of the scheduler's credit limit and scheduler job's credit
count to account for the actual size of a job, such that we fill up the
ring efficiently.
Signed-off-by: Danilo Krummrich
---
drivers/gpu/drm/nouveau/nouveau_abi16.c | 3 ++-
drivers/gpu/drm/nouveau/nouveau_drm.c | 2
/20231031032439.1558703-1-matthew.br...@intel.com/T/
Signed-off-by: Danilo Krummrich
---
drivers/gpu/drm/nouveau/nouveau_abi16.c | 18 +-
drivers/gpu/drm/nouveau/nouveau_abi16.h | 2 +-
drivers/gpu/drm/nouveau/nouveau_drm.c | 31 ++--
drivers/gpu/drm/nouveau/nouveau_drv.h | 9 +-
drivers/gpu
the GPUVM's GEM objects. Hence, make us of
it.
Signed-off-by: Danilo Krummrich
---
Originally, this patch was part of [1]. However, while applying the series,
I noticed that this patch needs another iteration, hence I held it back to
rework it and send it separately.
Changes since detached from [1
On 11/13/23 19:55, Abhinav Singh wrote:
On 11/14/23 00:19, Danilo Krummrich wrote:
Hi,
thanks for sending a v2.
On 11/13/23 19:42, Abhinav Singh wrote:
This patch fixes a sparse warning with this message
"warning:dereference of noderef expression". In this context i
1 - 100 of 799 matches
Mail list logo
