nv-tegra-mirror/linux-nvgpu
1
0
Fork 0
mirror of git://nv-tegra.nvidia.com/linux-nvgpu.git synced 2025-12-22 09:12:24 +03:00
Code Issues Packages Projects Releases Wiki Activity
9,884 Commits 57 Branches 84 Tags
6f4b7d5cc23128814ba9182f88c2c9b208bb0a08
Commit Graph

3421 Commits

Author SHA1 Message Date
prsethi
6f4b7d5cc2 gpu: nvgpu: fix the memory corruption issue 
Memory for tpc_index in gpc_tpc_physical_id_map array is allocated only
for number of tpcs while it should be number of tpcs*size of index.

Change fixes the memory allocation to avoid memory corruption.

Bug 3994374

Change-Id: Ibc593b1d0baba980787ae50f02ea20072525888c
Signed-off-by: prsethi <prsethi@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2906890
Reviewed-by: svcacv <svcacv@nvidia.com>
Reviewed-by: svc-mobile-coverity <svc-mobile-coverity@nvidia.com>
Reviewed-by: svc-mobile-cert <svc-mobile-cert@nvidia.com>
Reviewed-by: Dinesh T <dt@nvidia.com>
Reviewed-by: Kishan Palankar <kpalankar@nvidia.com>
Reviewed-by: Vaibhav Kachore <vkachore@nvidia.com>
GVS: Gerrit_Virtual_Submit <buildbot_gerritrpt@nvidia.com>
 2023-05-20 07:32:57 -07:00
Santosh BS
54b01e881b gpu: nvgpu: multimedia engine enumeration changes 
- Changes to fetch and expose supported multimedia engines to umd
- Unit and litter defines for multimedia engines
- Add functions to get runlist id

Jira NVGPU-9429
Bug 3962979

Signed-off-by: Santosh BS <santoshb@nvidia.com>
Change-Id: I072b4aac803c4a70d3659857cb0d804755c5dbd7
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2900765
Reviewed-by: svc-mobile-coverity <svc-mobile-coverity@nvidia.com>
Reviewed-by: svc-mobile-cert <svc-mobile-cert@nvidia.com>
Reviewed-by: svc-mobile-misra <svc-mobile-misra@nvidia.com>
Reviewed-by: Rajesh Devaraj <rdevaraj@nvidia.com>
Reviewed-by: V M S Seeta Rama Raju Mudundi <srajum@nvidia.com>
Reviewed-by: Vijayakumar Subbu <vsubbu@nvidia.com>
GVS: Gerrit_Virtual_Submit <buildbot_gerritrpt@nvidia.com>
 2023-05-18 23:40:19 -07:00
prsethi
24a533c9dc nvgpu: print the caller name with quiesce 
Currently quiesce method does not print the caller name which makes it
difficult to find the reason behind the issue.

Change prints caller name and invocation line number.

Bug 4098984

Change-Id: I34a0f557c411f997022668e187060c1c1247b15f
Signed-off-by: prsethi <prsethi@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2900585
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
 2023-05-15 15:54:43 -07:00
Rajesh Devaraj
a321679a5d gpu: nvgpu: add is_gsp_supported flag 
This patch adds is_gsp_supported flag and initializes it for GA10B,
TU104. Further, this flag is checked before initializaing GSP LITE
falcon.

JIRA NVGPU-9983

Change-Id: If0a4a3095c15cac113895f3d114e731f35211c5d
Signed-off-by: Rajesh Devaraj <rdevaraj@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2902651
Reviewed-by: svc-mobile-coverity <svc-mobile-coverity@nvidia.com>
Reviewed-by: svc-mobile-misra <svc-mobile-misra@nvidia.com>
Reviewed-by: svc-mobile-cert <svc-mobile-cert@nvidia.com>
Reviewed-by: Lakshmanan M <lm@nvidia.com>
Reviewed-by: Vijayakumar Subbu <vsubbu@nvidia.com>
GVS: Gerrit_Virtual_Submit <buildbot_gerritrpt@nvidia.com>
 2023-05-15 01:19:26 -07:00
mpoojary
a9b995bc3f gpu: nvgpu: Copy correct struct from RPC payload for pmu acr 
Size of nv_pmu_rpc_struct_acr_bootstrap_gr_falcons is copied from the
RPC payload for pmu acr instead of nv_pmu_rpc_header in pmu rpc handler.
This causes KSAN slab out-of-bounds error.

Bug 3727012

Change-Id: I633dac9167f9ed896dba956dc56e4081aaab6465
Signed-off-by: mpoojary <mpoojary@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2891392
Reviewed-by: svc-mobile-coverity <svc-mobile-coverity@nvidia.com>
Reviewed-by: svc-mobile-misra <svc-mobile-misra@nvidia.com>
Reviewed-by: svc-mobile-cert <svc-mobile-cert@nvidia.com>
Reviewed-by: Mahantesh Kumbar <mkumbar@nvidia.com>
GVS: Gerrit_Virtual_Submit <buildbot_gerritrpt@nvidia.com>
 2023-05-09 16:08:42 -07:00
vivekku
bd5ab81ccc gpu: nvgpu: nvs: queue direction update 
Changes:
-  update nvgpu_nvs_ctrl_queue to have queue direction as it is required
by gsp scheduler to erase queue individually
- queue direction is updated during ioctl call to create queue and is
used only by gsp scheduler. So no other moduler should be affected by
it.

- need to pass the size of struct which is u32 so downgrading it from
u64 to u32 is intentional, misra C violation 10.3 can be ignored here

Bug 4027512

Change-Id: I6ef6e4b06124e25da3d004a2d8822516c3ac2105
Signed-off-by: vivekku <vivekku@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2881804
Reviewed-by: svcacv <svcacv@nvidia.com>
Reviewed-by: svc-mobile-coverity <svc-mobile-coverity@nvidia.com>
Reviewed-by: svc-mobile-cert <svc-mobile-cert@nvidia.com>
Reviewed-by: Mahantesh Kumbar <mkumbar@nvidia.com>
GVS: Gerrit_Virtual_Submit <buildbot_gerritrpt@nvidia.com>
 2023-05-08 23:37:56 -07:00
Rajesh Devaraj
01d3ed09b0 gpu: nvgpu: update get_access_map 
This patch re-names the variables used in gr_init_get_access_map API:
whitelist - gr_access_map
num_entries - gr_access_map_num_entries
wl_addr_*[] - gr_access_map_*[]

JIRA NVGPU-9849

Change-Id: I3a0a59410af8983867af5bc2f9ff200e56e190c4
Signed-off-by: Rajesh Devaraj <rdevaraj@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2891567
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
 2023-05-05 02:31:10 -07:00
Shashank Singh
9512b9f1de gpu: nvgpu: remove user managed addr space capability flag 
Remove NVGPU_GPU_IOCTL_ALLOC_AS_FLAGS_USERSPACE_MANAGED and
NVGPU_AS_ALLOC_USERSPACE_MANAGED flags which are used for supporting
userspace managed address-space. This functionality is not implemented
fully in kernel neither going to be implemented in near future.

Jira NVGPU-9832
Bug 4034184

Change-Id: I3787d92c44682b02d440e52c7a0c8c0553742dcc
Signed-off-by: Shashank Singh <shashsingh@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2882168
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
 2023-05-04 11:39:30 -07:00
Martin Radev
84bb919909 gpu: nvgpu: Setup GFX-capable TPCs 
This patch dispatches to the appropriate HAL to
select the GFX-capable TPCs.

Bug 3944931

Change-Id: Ifb7338bea2cd59581133b7a2ba723f5d8bfa507c
Signed-off-by: Martin Radev <mradev@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2891725
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
 2023-05-04 11:38:54 -07:00
Divya
c728f09c18 gpu: nvgpu: add sysfs node for golden img status 
- Add a sysfs node "golden_img_status" to show
  if golden_image size and ptr are already initialized
  or not.
- This node helps to know golden image status before
  attempting to modify gpc/tpc/fbp masks.

Bug 3960290

Change-Id: I3c3de69b369bcaf2f0127e897d06e21cb8e2d68e
Signed-off-by: Divya <dsinghatwari@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2868729
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
 2023-05-04 11:36:37 -07:00
prsethi
c49ac865de gpu: nvgpu: init golden ctx image during nvgpu poweron 
Safety build temporal requirement is that on FECS power up it should go
through entire initialization methods.
init_golden_image callback is being called from devctl/ioctl path and
triggers FECS method 10 and 11. As these methods are part of APP init,
not being called during resume and causing quiesce on safety build.
To fix this issue, calling the callback from poweron API.

Bug 4082813
Bug 4037712

Change-Id: I2d27203d3cb4326ae7d8bd6025693fd61d5237df
Signed-off-by: prsethi <prsethi@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2893218
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
 2023-05-04 03:14:19 -07:00
Divya
3e5424bee3 gpu: nvgpu: gv11b: ap_compute fix 
- During nvgpu_poweron, PERFMON_INIT RPC and
  ACR_INIT_WPR_REGION command is sent to PMU in two different threads.
- For perfmon RPC method is used and for ACR, CMD-MSG queue is used.
- Since the pmu thread and poweron thread run in parallel, the
  pmu sequence acquired by both can have the same seq_id.
- For Perfmon RPC, nvgpu_pmu_seq_free_release() is called
  followed by nvgpu_pmu_seq_release().
- This causes clearing of sequence for the next command.
- To resolve this, instead of nvgpu_pmu_seq_free_release(),
  just free the rpc-payload after getting ack for perfmon and
  then do sequence release.
- This ensures that the ACR cmd sent just after perfmon RPC does
  not get the same seq_id and the sequence is not cleared.

Bug 4074021

Change-Id: Id9972cb719458062d8c7d9e226a25599026c052b
Signed-off-by: Divya <dsinghatwari@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2889840
Reviewed-by: svcacv <svcacv@nvidia.com>
Reviewed-by: svc-mobile-coverity <svc-mobile-coverity@nvidia.com>
Reviewed-by: svc-mobile-cert <svc-mobile-cert@nvidia.com>
Reviewed-by: Mahantesh Kumbar <mkumbar@nvidia.com>
GVS: Gerrit_Virtual_Submit <buildbot_gerritrpt@nvidia.com>
 2023-04-28 03:32:42 -07:00
srajum
53941baa93 gpu: nvgpu: fixing unit tests for ga10b 
- Add support for unit tests to run on orin platform.

JIRA NVGPU-9909 

Change-Id: I60a059840fd0d2733b0a1f2b3c1f722f8616868e
Signed-off-by: srajum <srajum@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2892228
Reviewed-by: svcacv <svcacv@nvidia.com>
Reviewed-by: svc-mobile-coverity <svc-mobile-coverity@nvidia.com>
Reviewed-by: svc-mobile-misra <svc-mobile-misra@nvidia.com>
Reviewed-by: svc-mobile-cert <svc-mobile-cert@nvidia.com>
Reviewed-by: Dinesh T <dt@nvidia.com>
Reviewed-by: Ankur Kishore <ankkishore@nvidia.com>
GVS: Gerrit_Virtual_Submit <buildbot_gerritrpt@nvidia.com>
 2023-04-28 02:08:14 -07:00
Sagar Kamble
4ee71f9852 gpu: nvgpu: guard ecc sysfs remove with NVGPU_DISABLE_ECC_STATS 
Following error is seen on unloading nvgpu on platforms with
NVGPU_DISABLE_ECC_STATS set to true.

[ 3712.384639] Internal error: Oops: 96000004 [#1] PREEMPT SMP
[ 3712.388479] pc : sysfs_remove_file_ns+0x28/0x50
[ 3712.389119] lr : sysfs_remove_file_ns+0x28/0x50
...
[ 3712.400640]  sysfs_remove_file_ns+0x28/0x50
[ 3712.401280]  device_remove_file+0x34/0x50
[ 3712.414720]  nvgpu_ecc_sysfs_remove+0x74/0xc0 [nvgpu]
[ 3712.428800]  nvgpu_ecc_remove_support+0x38/0x80 [nvgpu]
[ 3712.442240]  nvgpu_put+0xb8/0x160 [nvgpu]
[ 3712.456319]  nvgpu_pci_remove+0x168/0x250 [nvgpu]
[ 3712.456959]  pci_device_remove+0x4c/0x100

This is happening as ecc sysfs files are not created, however
their removal is attempted. Add NVGPU_DISABLE_ECC_STATS check
for ecc sysfs removal.

Bug 3495440

Change-Id: I9726cf43a740c2b591ca39bdc572e8f4ff5684d3
Signed-off-by: Sagar Kamble <skamble@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2891876
Reviewed-by: svc-mobile-coverity <svc-mobile-coverity@nvidia.com>
Reviewed-by: svc-mobile-cert <svc-mobile-cert@nvidia.com>
Reviewed-by: svc-mobile-misra <svc-mobile-misra@nvidia.com>
Reviewed-by: Rajesh Devaraj <rdevaraj@nvidia.com>
Reviewed-by: Divya Singhatwaria <dsinghatwari@nvidia.com>
Reviewed-by: Martin Radev <mradev@nvidia.com>
Reviewed-by: Seema Khowala <seemaj@nvidia.com>
GVS: Gerrit_Virtual_Submit <buildbot_gerritrpt@nvidia.com>
 2023-04-26 20:08:45 -07:00
Martin Radev
924dd58da0 gpu: nvgpu: remove IO_COHERENT flag 
This patch removes the IO_COHERENT flag as IO
coherence is the default setting.

Bug 3959027

Change-Id: I9800c2b8b161f7bdc2d6856639dd03488881882d
Signed-off-by: Martin Radev <mradev@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2887630
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
 2023-04-21 11:32:05 -07:00
Martin Radev
81d95456b9 gpu: nvgpu: Rename PLATFORM_ATOMIC to SYSTEM_COHERENT 
To support current and future usecases, it would be
beneficial to select the SYSTEM_COHERENT aperture explicitly.

The benefits are:
- platform atomic code is cleaned-up.
- userspace can select the SYSTEM_COHERENT aperture for any
  specific usecases.

Bug 3959027

Change-Id: I6489ebe87fa75cc760930277bad5e0cacca80eb6
Signed-off-by: Martin Radev <mradev@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2864177
Reviewed-by: svc-mobile-coverity <svc-mobile-coverity@nvidia.com>
Reviewed-by: svc-mobile-cert <svc-mobile-cert@nvidia.com>
Reviewed-by: svc-mobile-misra <svc-mobile-misra@nvidia.com>
Reviewed-by: Seema Khowala <seemaj@nvidia.com>
GVS: Gerrit_Virtual_Submit <buildbot_gerritrpt@nvidia.com>
 2023-04-21 11:31:53 -07:00
Shashank Singh
21cb70f58d gpu: nvgpu: remove kind control capability 
Kind is controlled by nvgpu userspace library so related capability
flags can be removed from kernel and uapi interface.

Jira NVGPU-9832
Bug 4034184

Change-Id: Id2b0a4e1cd784638362116b8d99177467fba998b
Signed-off-by: Shashank Singh <shashsingh@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2880391
Reviewed-by: svc-mobile-coverity <svc-mobile-coverity@nvidia.com>
Reviewed-by: svc-mobile-misra <svc-mobile-misra@nvidia.com>
Reviewed-by: svc-mobile-cert <svc-mobile-cert@nvidia.com>
Reviewed-by: Dinesh T <dt@nvidia.com>
Reviewed-by: Ankur Kishore <ankkishore@nvidia.com>
GVS: Gerrit_Virtual_Submit <buildbot_gerritrpt@nvidia.com>
 2023-04-13 12:21:46 -07:00
Divya
7a4fff4b17 gpu: nvgpu: add hal for pmu sequence cleanup 
- On older chips, PMU uses CMD-MSG queue method to
  communicate with NvGPU.
- From Turing onwards, PMU uses RPC method for this.
- During poweroff, we release pmu_sequence and reset the
  members of the structure.
- For chips that use RPC, we need to free the payload as well
  and then reset the members.
- Add pmu_seq_cleanup hal for this.

Bug 4019694
Bug 4059157

Change-Id: Ieb474fe4ed81f54d78480214cde53b51d45652c6
Signed-off-by: Divya <dsinghatwari@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2882267
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
 2023-04-12 16:28:52 -07:00
Divya
db9a411a06 gpu: nvgpu: sync free of rpc_payload 
- During driver unload, shutdown or RG path as part of
  pmu destroy, pmu sequences have to be cleaned up to
  free payload memory and allocation info which is stored
  as part of pmu_sequence.
- While doing so there can be race condition with pmu_isr
  or nvgpu_pmu_rpc_execute path where it waits for fw ack.
- This race condition can lead to freeing of payload memory
  before nvgpu_pmu_sequences_cleanup() does.
- This can lead to memory corruption or double free issue
  when the cleanup code again tries to free the payload mem.
- To resolve this add a new function nvgpu_pmu_seq_free_release()
  which will check for seq->id in pmu seq tbl before freeing the
  memory and other info from pmu_sequence.
- Use this nvgpu_pmu_seq_free_release() in non-blocking RPC calls
  and also when fw ack fails or driver is dying scenario.
- For blocking call, synchronise freeing of rpc payload memory by
  using a new boolean seq_free_status.

Bug 4019694
Bug 4059157

Change-Id: Id45a6914a2d383a654539a87861c471a77fb6850
Signed-off-by: Divya <dsinghatwari@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2882210
Reviewed-by: svcacv <svcacv@nvidia.com>
Reviewed-by: svc-mobile-coverity <svc-mobile-coverity@nvidia.com>
Reviewed-by: svc-mobile-cert <svc-mobile-cert@nvidia.com>
Reviewed-by: Vijayakumar Subbu <vsubbu@nvidia.com>
GVS: Gerrit_Virtual_Submit <buildbot_gerritrpt@nvidia.com>
 2023-04-12 16:28:41 -07:00
Prathap Kumar Valsan
d0ed86ab1e gpu: nvgpu: update PT lvl array to six levels 
Increase the size of page table level array to index six levels.

Jira NVGPU-9760

Change-Id: I482639fd028ecd504ee8bc313c39f3bd710e81a9
Signed-off-by: Prathap Kumar Valsan <prathapk@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2868918
Reviewed-by: svcacv <svcacv@nvidia.com>
Reviewed-by: svc-mobile-coverity <svc-mobile-coverity@nvidia.com>
Reviewed-by: svc-mobile-misra <svc-mobile-misra@nvidia.com>
Reviewed-by: svc-mobile-cert <svc-mobile-cert@nvidia.com>
Reviewed-by: Seema Khowala <seemaj@nvidia.com>
GVS: Gerrit_Virtual_Submit <buildbot_gerritrpt@nvidia.com>
 2023-04-07 21:08:45 -07:00
Kishan
c6d5fb348c gpu: nvgpu: Capture thread name for every channel created 
This change ensures that in scenarios where GPU enters a bad
state because of the work submitted by a misbehaved thread,
we should be able to capture thread name as part of our
1st set of failure logs.
Changes for QNX env is pending.

JIRA NVGPU-7783

Change-Id: I65d55a6ade749ff91739458e0642ed2dafaae5cc
Signed-off-by: Kishan <kpalankar@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2879197
Reviewed-by: svc-mobile-coverity <svc-mobile-coverity@nvidia.com>
Reviewed-by: svc-mobile-misra <svc-mobile-misra@nvidia.com>
Reviewed-by: svc-mobile-cert <svc-mobile-cert@nvidia.com>
Reviewed-by: Tejal Kudav <tkudav@nvidia.com>
Reviewed-by: Ankur Kishore <ankkishore@nvidia.com>
GVS: Gerrit_Virtual_Submit <buildbot_gerritrpt@nvidia.com>
 2023-04-06 10:12:48 -07:00
Ramalingam C
af48120169 gpu: nvgpu: configure CWD sm id regs before ucode load 
GPCCS ucode is expecting the SM ID programmed in GPM and CWD registers
to be in sync. So create a hal called gr.init.load_sm_id_config()
for the sm_id programming for the CWD registers and invoke them before
the ucode load.

Initialize this hal only for the required GPUs.

JIRA NVGPU-9757

Change-Id: Ib0984fd6326c37e0c2a06123041032575a23ec04
Signed-off-by: Ramalingam C <ramalingamc@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2864999
Reviewed-by: svc-mobile-coverity <svc-mobile-coverity@nvidia.com>
Reviewed-by: svc-mobile-cert <svc-mobile-cert@nvidia.com>
Reviewed-by: Seema Khowala <seemaj@nvidia.com>
Reviewed-by: Rajesh Devaraj <rdevaraj@nvidia.com>
GVS: Gerrit_Virtual_Submit <buildbot_gerritrpt@nvidia.com>
 2023-04-06 10:09:14 -07:00
Sagar Kamble
821699d3a3 gpu: nvgpu: unset async subctx VM with correct index 
On deleting the subcontext, tsg->subctx_vms[] entries are set to NULL as
per the subcontext id. For async subcontexts the index logic was used
from that of tsg->async_veids bitmask. However subctx_vms is an array
shared by all subcontexts hence index should be subcontext id aka veid.

Also update the description of function nvgpu_tsg_validate_ch_subctx_vm
as some of the functionality is now moved to another function
nvgpu_tsg_create_sync_subcontext_internal.

Bug 3979886

Change-Id: Ic290fb175b34988c6ffabe9c9dc4ec124d2c70af
Signed-off-by: Sagar Kamble <skamble@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2879025
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
 2023-03-31 13:33:45 -07:00
Sagar Kamble
a5640d61bd gpu: nvgpu: free VEID if the channel is closed 
In case of process crash or forceful closure of the channels, userspace
may not release the VEID. In that case, creating further subcontexts
may not be possible.

Hence, when the channel is closed forcibly (linux), release the VEID on
closure of the last channel in the subcontext.

With this, normally on linux, channel close will not relase the VEID
However, on qnx it will release the VEID. So delete subcontext devctl
call on qnx will be nop in normal case hence changed the error print
and error return to success.

Also added check in the subcontext delete ioctl fn that all channels
are unbound before deleting the subcontext. This is to ensure that
channels don't refer to dangling subcontext pointer.

Bug 3979886

Change-Id: I434944b01740720011abce3664394ae8cb0d4e2e
Signed-off-by: Sagar Kamble <skamble@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2858060
Reviewed-by: svc-mobile-coverity <svc-mobile-coverity@nvidia.com>
Reviewed-by: svc-mobile-cert <svc-mobile-cert@nvidia.com>
Reviewed-by: Vijayakumar Subbu <vsubbu@nvidia.com>
GVS: Gerrit_Virtual_Submit <buildbot_gerritrpt@nvidia.com>
 2023-03-31 13:25:53 -07:00
Santosh BS
2a865e3aad gpu: nvgpu: NVENC support on TU104 
This patch adds nvenc support for TU104
- Fetch engine/dev info for nvenc
- Falcon NS boot (fw loading) support
- Engine context creation for nvenc
- Skip golden image for multimedia engines
- Avoid subctx for nvenc as it is a non-VEID engine
- Job submission/flow changes for nvenc
- Code refactoring to scale up the support for other multimedia
  engines in future.

Bug 3763551

Change-Id: I03d4e731ebcef456bcc5ce157f3aa39883270dc0
Signed-off-by: Santosh BS <santoshb@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2859416
Reviewed-by: svc-mobile-coverity <svc-mobile-coverity@nvidia.com>
Reviewed-by: svc-mobile-cert <svc-mobile-cert@nvidia.com>
Reviewed-by: Rajesh Devaraj <rdevaraj@nvidia.com>
Reviewed-by: Vijayakumar Subbu <vsubbu@nvidia.com>
GVS: Gerrit_Virtual_Submit <buildbot_gerritrpt@nvidia.com>
 2023-03-24 17:07:49 -07:00
Martin Radev
ac9a59075e gpu: nvgpu: Print flags after all flags are set 
Without this change, nvgpu would print out some
flags as disabled in dmesg but enable them shortly after.
This leads to confusion when examining UMD and nvgpu
reporting in UMDs.

This patch adds code to print out the flags after
all flags are set.

Bug 4031904

Change-Id: I67b9a4567886fd5e076f7ac3b8f284b52c03d7e4
Signed-off-by: Martin Radev <mradev@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2871606
Reviewed-by: svcacv <svcacv@nvidia.com>
Reviewed-by: svc-mobile-coverity <svc-mobile-coverity@nvidia.com>
Reviewed-by: svc-mobile-misra <svc-mobile-misra@nvidia.com>
Reviewed-by: svc-mobile-cert <svc-mobile-cert@nvidia.com>
Reviewed-by: Alex Waterman <alexw@nvidia.com>
Reviewed-by: Vijayakumar Subbu <vsubbu@nvidia.com>
Reviewed-by: Seema Khowala <seemaj@nvidia.com>
GVS: Gerrit_Virtual_Submit <buildbot_gerritrpt@nvidia.com>
 2023-03-21 09:45:13 -07:00
srajum
02834f8739 gpu: nvgpu: fix CERT-C issues 
- CID 10165014
  Dereference before null check

- CID 10166579
  Unused value

Bug 3952896

Change-Id: I6a7f2b97b4a6519272607e560d09c138048bd665
Signed-off-by: srajum <srajum@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2872276
Reviewed-by: svc-mobile-coverity <svc-mobile-coverity@nvidia.com>
Reviewed-by: svc-mobile-cert <svc-mobile-cert@nvidia.com>
Reviewed-by: Vaibhav Kachore <vkachore@nvidia.com>
GVS: Gerrit_Virtual_Submit <buildbot_gerritrpt@nvidia.com>
 2023-03-21 02:37:08 -07:00
Richard Zhao
f791adf880 gpu: nvgpu: move .runlist.hw_submit to use runlist_id 
Use detailed function parameters runlist_id, iova/aperture and count, so
the HAL could be reused on server side.

Jira GVSCI-15773

Change-Id: I28f68682b9eea4e798af5c850c87840bd9b79970
Signed-off-by: Richard Zhao <rizhao@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2863444
Reviewed-by: svc-mobile-coverity <svc-mobile-coverity@nvidia.com>
Reviewed-by: svc-mobile-cert <svc-mobile-cert@nvidia.com>
Reviewed-by: Ramesh Mylavarapu <rmylavarapu@nvidia.com>
Reviewed-by: Vijayakumar Subbu <vsubbu@nvidia.com>
GVS: Gerrit_Virtual_Submit <buildbot_gerritrpt@nvidia.com>
 2023-03-21 02:31:29 -07:00
Richard Zhao
a587d94f5a gpu: nvgpu: init nvs scheduler for vf 
nvs does not have a clean cut. runlist submit path uses nvs worker no
matter whether the feature is enabled.

Jira GVSCI-15773

Change-Id: I6f6db1e766b8079ad6ca4a6b530b3ec27094f840
Signed-off-by: Richard Zhao <rizhao@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2863443
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
 2023-03-21 02:31:21 -07:00
Richard Zhao
da1da8f563 gpu: nvgpu: move .preempt_trigger/.is_preempt_pending to IDs 
.preempt_tsg uses .preempt_trigger/.is_preempt_pending, so they both
have to use runlist_id and tsgid too.

Jira GVSCI-15770

Change-Id: Ida24d160c362ea1348d7c19e6d0352bb390d0a64
Signed-off-by: Richard Zhao <rizhao@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2863442
Reviewed-by: svc-mobile-coverity <svc-mobile-coverity@nvidia.com>
Reviewed-by: svc-mobile-cert <svc-mobile-cert@nvidia.com>
Reviewed-by: Ramesh Mylavarapu <rmylavarapu@nvidia.com>
Reviewed-by: Vijayakumar Subbu <vsubbu@nvidia.com>
GVS: Gerrit_Virtual_Submit <buildbot_gerritrpt@nvidia.com>
 2023-03-21 02:30:57 -07:00
Richard Zhao
8f5adab299 gpu: nvgpu: .preempt_tsg move to use runlist_id/tsgid 
It's for making .preempt_tsg reusable on server side.

Jira GVSCI-15770

Change-Id: Id9f477baa29cb63fb0e1d1650f4b1e6a2fa248c0
Signed-off-by: Richard Zhao <rizhao@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2863441
Reviewed-by: svc-mobile-coverity <svc-mobile-coverity@nvidia.com>
Reviewed-by: svc-mobile-cert <svc-mobile-cert@nvidia.com>
Reviewed-by: Ramesh Mylavarapu <rmylavarapu@nvidia.com>
Reviewed-by: Vijayakumar Subbu <vsubbu@nvidia.com>
GVS: Gerrit_Virtual_Submit <buildbot_gerritrpt@nvidia.com>
 2023-03-21 02:30:49 -07:00
Rajesh Devaraj
4f96f59c15 gpu: nvgpu: perform support_ls_pmu check 
Perform support_ls_pmu check before dereferencing pmu from gpu struct g
to avoid the possibility of kernel panic when LS PMU support is not
enabled.

JIRA NVGPU-9283

Change-Id: I65caac449f884164d797dedc2041d6ee4292e326
Signed-off-by: Rajesh Devaraj <rdevaraj@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2868250
Reviewed-by: svc-mobile-coverity <svc-mobile-coverity@nvidia.com>
Reviewed-by: svc-mobile-cert <svc-mobile-cert@nvidia.com>
Reviewed-by: svc-mobile-misra <svc-mobile-misra@nvidia.com>
Reviewed-by: Mahantesh Kumbar <mkumbar@nvidia.com>
Reviewed-by: Ramalingam C <ramalingamc@nvidia.com>
Reviewed-by: Seema Khowala <seemaj@nvidia.com>
GVS: Gerrit_Virtual_Submit <buildbot_gerritrpt@nvidia.com>
 2023-03-17 10:04:59 -07:00
Richard Zhao
4ec683975a gpu: nvgpu: obj_ctx: fix possible mem leak 
When generate golden image, subctx_mask memory was not freed on fail.
It was detected by code coverity checker.

Bug 3952896

Change-Id: Iae0c78b11003980c6b09ec0e72bebfda0a244b17
Signed-off-by: Richard Zhao <rizhao@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2868150
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
 2023-03-17 10:04:45 -07:00
prsethi
505690f505 gpu: nvgpu: add validation check for domain name 
Currently there is no validation checks for domain name used in domain
create command which can cause some security risk.
Patch enable the validation for domain name by only allowing char from
([a-z], [A-Z], [0-9], -, _) list.

Bug 3994374

Change-Id: Ia2cb6f533ed136e74e7a72934ad5267803d1236d
Signed-off-by: prsethi <prsethi@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2871515
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
 2023-03-17 04:17:14 -07:00
Richard Zhao
41823694a3 gpu: nvgpu: add .init_golden_image HAL 
golden image is created differently on native and VF.

Jira GVSCI-15772

Change-Id: I8d78d1214d8aac1d39d6529b68adef1dd6f8a516
Signed-off-by: Richard Zhao <rizhao@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2863440
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
 2023-03-17 04:04:33 -07:00
Richard Zhao
067e3590d5 gpu: nvgpu: runlist: init engine info of runlist for VF 
- init engine info for VF which is needed to setup ramfc
- avoid register access in nvgpu_runlist_get_device_id. It could use
rleng_id.
- alloc physical addressed memory for vf runlist mem.

Jira GVSCI-15773

Change-Id: I63494b306a2f56d090a61ea1fa581083224d1cb6
Signed-off-by: Richard Zhao <rizhao@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2863432
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
 2023-03-17 04:04:21 -07:00
Richard Zhao
64e22ee54b gpu: nvgpu: vgpu: add rleng_id to constants 
rl_eng_id is used to construct ram_fc_target_w. VF creates inst_block
and ramfc on client side.

Jira GVSCI-15769

Change-Id: Id641e644e829bfaf6a3bb0bb758c142f0a514db3
Signed-off-by: Richard Zhao <rizhao@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2863431
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
 2023-03-17 04:04:09 -07:00
vivekku
ce4293ab20 gpu: nvgpu: gsp: disabling multiple gsp firmware read 
Changes:
- disabled gsp firmware release during railgating
- firmware read happen only during power on

Bug 3935433

Change-Id: I9156c015ab7f90ab640c33ca99dc7f3e289b7659
Signed-off-by: vivekku <vivekku@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2870170
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
 2023-03-17 03:56:36 -07:00
vivekku
9773e7ca68 gpu: nvgpu: gsp: erase queue command for safety scheduler 
Changes:
- implemented erase queue command for safety scheduler to depopulate
control fifo parameters in safety scheduler FW.

NVGPU-9590
Bug 3935433

Change-Id: I2cd6cd967ac4dba61992dd285e45b18f34dda2ca
Signed-off-by: vivekku <vivekku@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2858533
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
 2023-03-17 03:55:31 -07:00
vivekku
a2a86eed27 gpu: nvgpu: gsp: migration from KMD to GSP 
Changes:
- submit shadow domain for legacy used cases in case user domain is not
present.
- disabling config flags for KMD to submit user domain.

Bug 3935433
NVGPU-9664

Change-Id: I498226df36d0b482d1af369526adb369d921b6ca
Signed-off-by: vivekku <vivekku@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2843968
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
 2023-03-17 03:55:20 -07:00
vivekku
35960f8f40 gpu: nvgpu: gsp: call runlist update and send ctrl fifo info 
Changes:
- function calls to add and delete domains
- updating runlist
- integrating control fifo changes with ioctls to send queue info to GSP
FW

Bug 3884011

Change-Id: I5ad29eb9501cc2df66843c074ee6a00aae91af23
Signed-off-by: vivekku <vivekku@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2826482
Reviewed-by: Ramesh Mylavarapu <rmylavarapu@nvidia.com>
Reviewed-by: Mahantesh Kumbar <mkumbar@nvidia.com>
GVS: Gerrit_Virtual_Submit <buildbot_gerritrpt@nvidia.com>
 2023-03-17 03:55:08 -07:00
prsethi
6b2c080f8f gpu:nvgpu: add enable flag for KMD_SCHEDULING_WORKER_THREAD support 
Currently KMD_SCHEDULING_WORKER_THREAD can be enabled/disabled using
compile time flag but this flag does give ability to control the
feature based on the chip.
GSP is enabled only on ga10b where KMD_SCHEDULING_WORKER_THREAD should
be disabled while should be enabled for other chips at the same time
to support GVS tests.
Change adds enabled flag to control KMD_SCHEDULING_WORKER_THREAD based
on the chip.

Bug 3935433

Change-Id: I9d2f34cf172d22472bdc4614073d1fb88ea204d7
Signed-off-by: prsethi <prsethi@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2867023
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
 2023-03-17 03:55:02 -07:00
Dinesh Kamalakannan
a8581f0283 gpu: nvgpu: mailbox dump before reporting to SDL 
In case of ACR Bootstrap failure, print the mailbox
values for the error code, before reporting the error
to Safety Service. In some cases, once the error is
reported to SDL, only the nvgpu_sw_quiesce debugs are
there in sloginfo.

Bug 3960710

Signed-off-by: Dinesh Kamalakannan <dineshka@nvidia.com>
Change-Id: Ib72bb4280cc80b39c5fb2e57f26c7b0e9ab6a4ad
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2867592
Reviewed-by: Mahantesh Kumbar <mkumbar@nvidia.com>
Reviewed-by: svc-mobile-coverity <svc-mobile-coverity@nvidia.com>
Reviewed-by: svc-mobile-misra <svc-mobile-misra@nvidia.com>
Reviewed-by: svc-mobile-cert <svc-mobile-cert@nvidia.com>
GVS: Gerrit_Virtual_Submit <buildbot_gerritrpt@nvidia.com>
 2023-03-13 04:59:08 -07:00
Richard Zhao
84ddb23633 gpu: nvgpu: move .force_ctx_reload to use runlist_id and chid 
Moving to use IDs rather than struct makes it reusable on server side.

Jira GVSCI-15770

Change-Id: Id4e815e9cf78a43156449d0e77e8e331fc906725
Signed-off-by: Richard Zhao <rizhao@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2863439
Reviewed-by: svc-mobile-coverity <svc-mobile-coverity@nvidia.com>
Reviewed-by: svc-mobile-cert <svc-mobile-cert@nvidia.com>
GVS: Gerrit_Virtual_Submit <buildbot_gerritrpt@nvidia.com>
 2023-03-13 04:56:10 -07:00
Richard Zhao
c8d6a91de6 gpu: nvgpu: update .channel.enable/disable to use runlist_id and chid 
Moving to use IDs rather than struct makes it reusable on server side.

Jira GVSCI-15770

Change-Id: Ibd94ab8c9f0492bd6d20243525905d637eb8de66
Signed-off-by: Richard Zhao <rizhao@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2863438
Reviewed-by: svc-mobile-coverity <svc-mobile-coverity@nvidia.com>
Reviewed-by: svc-mobile-cert <svc-mobile-cert@nvidia.com>
GVS: Gerrit_Virtual_Submit <buildbot_gerritrpt@nvidia.com>
 2023-03-13 04:56:04 -07:00
Richard Zhao
d9c8d317f0 gpu: nvgpu: update .read_state to use runlist_id and chid 
Moving to use IDs rather than struct makes it reusable on server side.

Jira GVSCI-15770

Change-Id: Ia5e30ebb0e8092b9cdc4c3f3cd524f585fd4b410
Signed-off-by: Richard Zhao <rizhao@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2863437
Reviewed-by: svc-mobile-coverity <svc-mobile-coverity@nvidia.com>
Reviewed-by: svc-mobile-cert <svc-mobile-cert@nvidia.com>
Reviewed-by: Dinesh T <dt@nvidia.com>
GVS: Gerrit_Virtual_Submit <buildbot_gerritrpt@nvidia.com>
 2023-03-13 04:55:58 -07:00
Richard Zhao
2ff110f722 gpu: nvgpu: update .clear to use runlist_id and chid 
- Moving to use IDs rather than struct makes it reusable on server side.
- move channel bind/unbind to use .enable/.clear HALs

Jira GVSCI-15770

Change-Id: I86d4aae2953024e537e32a35fe9cabb1b91cd201
Signed-off-by: Richard Zhao <rizhao@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2863436
Reviewed-by: svc-mobile-coverity <svc-mobile-coverity@nvidia.com>
Reviewed-by: svc-mobile-misra <svc-mobile-misra@nvidia.com>
Reviewed-by: svc-mobile-cert <svc-mobile-cert@nvidia.com>
Reviewed-by: Ramesh Mylavarapu <rmylavarapu@nvidia.com>
GVS: Gerrit_Virtual_Submit <buildbot_gerritrpt@nvidia.com>
 2023-03-13 04:55:53 -07:00
Richard Zhao
7cd377568f gpu: nvgpu: vgpu: init ctx buffers for vf driver 
VF needs to allocate gr ctx buffers on gr init, since VF will manage the
gr ctx.

Jira GVSCI-15769

Change-Id: Ifd09e6b09306c0fd36bddc60caa3d0d56f2b29cb
Signed-off-by: Richard Zhao <rizhao@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2863434
Reviewed-by: svc-mobile-coverity <svc-mobile-coverity@nvidia.com>
Reviewed-by: svc-mobile-cert <svc-mobile-cert@nvidia.com>
Reviewed-by: Dinesh T <dt@nvidia.com>
GVS: Gerrit_Virtual_Submit <buildbot_gerritrpt@nvidia.com>
 2023-03-13 04:55:35 -07:00
Rajesh Devaraj
e5c5fee01e gpu: nvgpu: skip falcon sw init for pmu 
Perform falcon sw init for PMU only if it is supported in the platform.

JIRA NVGPU-9283

Change-Id: I697e389a7cd812c7cb941eac010549f541074bb4
Signed-off-by: Rajesh Devaraj <rdevaraj@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2848436
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
 2023-03-12 19:47:32 -07:00
Richard Zhao
a23e574de0 gpu: nvgpu: vf: init syncpt mem 
Since gmmu map is moved to VF, the syncpt mem map is also on VF clients.

Jira GVSCI-15733

Change-Id: Iaf68070da860616f5301a822ce98581b8a1a6629
Signed-off-by: Richard Zhao <rizhao@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2863445
Reviewed-by: svc-mobile-coverity <svc-mobile-coverity@nvidia.com>
Reviewed-by: svc-mobile-misra <svc-mobile-misra@nvidia.com>
Reviewed-by: svc-mobile-cert <svc-mobile-cert@nvidia.com>
Reviewed-by: svcacv <svcacv@nvidia.com>
Reviewed-by: Prathap Kumar Valsan <prathapk@nvidia.com>
GVS: Gerrit_Virtual_Submit <buildbot_gerritrpt@nvidia.com>
 2023-03-12 08:13:52 -07:00