nvidia-tegra-modules.git/include/uapi/linux, branch gpu-paging

gpu-paging: Functions to support initial working version

2022-05-25T01:16:30+00:00

- Buffer deallocation (+IOCTL)
- Buffer reallocation
- Private dmabuf user list and accessor

nvdla: kmd: Expose full set of limits for task parameters

2021-08-17T03:32:13+00:00

- Certain task parameter limits were not exposed to UMD / Users
- Expose these limits and rename to provide consistent KMD interface
across linux / qnx

Jira DLA-4467
JIRA DLA-4908
Bug 3352716

Change-Id: Ibcf1de5f4d442d9b1f3fefaf71195bc6e58fff5c
Signed-off-by: Anup Mahindre 
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvidia/+/2558254
(cherry picked from commit ccb23fdbb6da2908893cfebcdcd0d219f9b255e0)
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvidia/+/2572193
Tested-by: Bharat Nihalani 
Tested-by: mobile promotions 
Reviewed-by: Bharat Nihalani 
Reviewed-by: mobile promotions 
GVS: Gerrit_Virtual_Submit

nvdla: kmd: add support for stride signal action

2021-08-17T03:32:07+00:00

NVDEV_FENCE_SIGNAL_STRIDE fence action translates to
ACTION_INCREMENT_SEM firmware action.

Jira DLA-4445
JIRA DLA-4908
Bug 3352716

Change-Id: Idb537ea784614d031f66c570359dbc7dd74374af
Signed-off-by: Arvind M 
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvidia/+/2548309
(cherry picked from commit 55093de292c83d787ee3cb5d74bcb1c89a65e8fc)
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvidia/+/2572192
Tested-by: Bharat Nihalani 
Tested-by: mobile promotions 
Reviewed-by: Bharat Nihalani 
Reviewed-by: mobile promotions 
GVS: Gerrit_Virtual_Submit

nvdla: kmd: add bypass exec support with submit

2021-08-17T03:32:02+00:00

[1] NVDLA_SUBMIT_FLAGS_BYPASS_EXEC flag allows its clients
    (UMD or nvdla_kmd_sanity) to bypass execution for a
    submission. This flag is a property of submission and
    shall directly apply on all tasks within that submit.
[2] With above flag set, it propagates information to firmware
    through task descriptor and firmware shall be responsible
    for bypassing the execution.

Jira DLA-4443
JIRA DLA-4908
Bug 3352716

Change-Id: I70000ef486905c812fec65f265a378a99844c10a
Signed-off-by: Arvind M 
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvidia/+/2544802
(cherry picked from commit 72da3fde9aa1624722359573d37ec6e73017ca98)
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvidia/+/2572191
Tested-by: Bharat Nihalani 
Tested-by: mobile promotions 
Reviewed-by: Bharat Nihalani 
Reviewed-by: mobile promotions 
GVS: Gerrit_Virtual_Submit

video: tegra: host: nvdla: Add support for internal buffers

2021-08-17T03:31:56+00:00

- Add buffer type field to nvdla_mem_handle
- For internal buffers, skip pin and unpin operations and pass the
offset as final adress
- Also update error handling in nvdla_map_task_memory to match with
  convention

Jira DLA-4376
JIRA DLA-4908
Bug 3352716

Change-Id: I662da30cb9c606d2f67b792f09e026af391c89d1
Signed-off-by: Anup Mahindre 
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvidia/+/2503708
(cherry picked from commit a2cc66a3f318394e65f6f126bbb7eaa5ccdae3e2)
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvidia/+/2572190
Tested-by: Bharat Nihalani 
Tested-by: mobile promotions 
Reviewed-by: Bharat Nihalani 
Reviewed-by: mobile promotions 
GVS: Gerrit_Virtual_Submit

misc: tegra-profiler: add unique identifier

2021-01-16T02:40:09+00:00

Add unique identifier seqid (sequence number) for samples.

Bug 3216659

Signed-off-by: Igor Nabirushkin 
Change-Id: I84b595f976f1f1955ecb278ababaf45591cdf5e0
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvidia/+/2464500
(cherry picked from commit ec284fc0b7b94871565c612358ea9e78e9ae4731)
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvidia/+/2469351
Reviewed-by: Roman Rybalko 
Reviewed-by: Bibek Basu 
Reviewed-by: mobile promotions 
GVS: Gerrit_Virtual_Submit
Tested-by: mobile promotions

nvdla: kmd: handle {RELEASE,ALLOC}_QUEUE ioctls

2021-01-05T10:26:12+00:00

[1] This commit separates the allocation and deallocation
    of mission critical resources away from open() and close()
[2] It is achieved through introduction of NVDLA_IOCTL_{ALLOC,RELEASE}
    IOCTLs.
[3] nvdla_buffer_* APIs introduced to facilitate easier split.

Jira DLA-4175
Bug 3200422
Bug 200628173

Change-Id: I3fb07ecaff69c62ec5eb9e5bea39b07ae1624240
Signed-off-by: Arvind M 
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvidia/+/2403689
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvidia/+/2412585
GVS: Gerrit_Virtual_Submit
Tested-by: Bharat Nihalani 
Tested-by: mobile promotions 
Reviewed-by: Mitch Harwell 
Reviewed-by: Bharat Nihalani 
Reviewed-by: mobile promotions

video: tegra: host: pin with specific permission

2021-01-05T10:26:07+00:00

[1] Facilitate pinning of buffer with user requested access
    permission
[2] This feature allows to restrict access to buffers shared
    with the device.

Jira DLA-2502
Jira DLA-4175
Bug 3200422

Change-Id: I175ad10922480689de57eb832040a7d59b29b9bd
Signed-off-by: Arvind M 
Reviewed-on: https://git-master.nvidia.com/r/2209415
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvidia/+/2412573
GVS: Gerrit_Virtual_Submit
Tested-by: Bharat Nihalani 
Tested-by: mobile promotions 
Reviewed-by: Mitch Harwell 
Reviewed-by: Bharat Nihalani 
Reviewed-by: mobile promotions

drivers: tegra: nvmap: Add NVMAP_IOC_PARAMETERS ioctl

2020-07-06T16:10:33+00:00

Add NVMAP_IOC_PARAMETERS to return nvmap handle's

contig; Phys_Contig
import_id; NA
handle;
heap_number; Only valid for IVM carveout
access_flags; NA
heap; Heap Type
align; Alignement
coherency; Coherency flag
size; Handle Size

Bug 3038325

Change-Id: I6fcfce7c3808e21084b543fa87ff8e48431b19f5
Signed-off-by: puneet saxena 
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvidia/+/2368252
Reviewed-by: automaticguardword 
Reviewed-by: Kai Zhang (SW-TEGRA) 
Reviewed-by: Tuomas Kulve 
Reviewed-by: Bibek Basu 
Reviewed-by: mobile promotions 
Tested-by: mobile promotions 
GVS: Gerrit_Virtual_Submit

drivers: staging: Add Generic timestamping support

2020-02-24T15:24:48+00:00

Adds generic hardware timestamping (GTE) support for the T194 SoC.
There are certain applications like robotics which may require more
accurate recording of the occurance of the certain events. GTE driver
will help monitor AON GPIOs and LIC interrupts.

GTE driver also implements char driver for userspace to monitor GPIO
through IOCTL.

Other kernel APIs are exported and defined in the tegra-gte.h file for
kernel client drivers to use. APIs are not stable and subject to change
in the future. It is for this reason GTE driver is provided in the
staging directory.

Since its APIs are not stable, it should not break existing nvpps driver
code from the stage-main which also uses GTE but in limited way. It is
for this reason this driver will not be selected if nvpps is enabled.

Bug 2757864

Change-Id: I0947f2b90232eb6c2a31163e33ec5ad45b7bd415
Signed-off-by: Dipen Patel 
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvidia/+/2287452
Reviewed-by: Bibek Basu 
Reviewed-by: mobile promotions 
Tested-by: mobile promotions 
GVS: Gerrit_Virtual_Submit