To achieve permanent fault coverage, the CTAs launched by
each kernel in the mission and redundant contexts must execute
on different hardware resources.
This feature requires a change in software to make it possible
to modify the virtual SM id to TPC mapping across mission and
redundant contexts.
This CL adds only SM diversity flags which are exposed to
its clients through ioctl/devctl interfaces.
Actual virtual SM id to TPC mapping implementation
will be part of upcoming patch sets.
Added NvGpu CFLAGS to identify the safety build
"CONFIG_NVGPU_BUILD_CONFIGURATION_IS_SAFETY"
JIRA NVGPU-4133
Change-Id: I5a18256780e6726e399e39c1c8d155d2ef07d7bd
Signed-off-by: Lakshmanan M <lm@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/2250461
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Previously, unit interrupt enabling/disabling and corresponding MC level
interrupt enabling/disabling was not done at the same time.
With this change, stall and nonstall interrupt for units are programmed
at MC level along with individual unit interrupts. Kept access to MC
interrupt registers through mc.intr_lock spinlock.
For doing this separated CE and GR interrupt mask functions.
mc.intr_enable is only used when there is global interrupt
control to be set. Removed mc_gp10b.c as mc_gp10b_intr_enable
is now removed. Removed following functions - mc_gv100_intr_enable,
mc_gv11b_intr_enable & intr_tu104_enable. Removed intr_pmu_unit_config
as we can use the generic unit interrupt control function.
JIRA NVGPU-4336
Change-Id: Ibd296d4a60fda6ba930f18f518ee56ab3f9dacad
Signed-off-by: Sagar Kamble <skamble@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/2196178
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
The pmu init thread typically returns immediately
without calling nvgpu_thread_should_stop().
pmu_pg_kill_task() checks if the thread is running, and
if it is, calls nvgpu_thread_stop().
However, there's a race condition where the init thread could
have exited between the time that kill_task() checked the
running flag and the time we actually stop the thread, leading
to a kernel crash.
Fix this by making the running flag in the nvgpu_thread struct
atomic. Both the thread proxy function and the thread_stop()
function will set the flag to false.
In the case of nvgpu_thread_proxy(), if the flag is already false,
then nvgpu_thread_stop() has already reset it, at which point we
just wait for nvgpu_thread_should_stop() to return true.
In the case of nvgpu_thread_stop(), if the flag is already false,
then the thread proxy function has already exited, and there is
nothing more to do.
Bug 2591298
Change-Id: I9ba6b63c30a5c3e1df11e790094836b44373122b
Signed-off-by: Peter Daifuku <pdaifuku@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/2230358
GVS: Gerrit_Virtual_Submit
Reviewed-by: Thomas Fleury <tfleury@nvidia.com>
Reviewed-by: Alex Waterman <alexw@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
tegra_get_chip_id is going to be deprecated soon and this patch removes
calls to it. Chip-ID is already read via DT and call to
tegra_get_chip_id() can be avoided by adding metadata to store the
Chip-ID information in struct gk20a_platform.
Bug 200524194
Bug 200551105
Change-Id: I5f9f5abf679cf9afe98840e20144d76eb0238426
Signed-off-by: Debarshi Dutta <ddutta@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/2236311
Reviewed-by: Automatic_Commit_Validation_User
GVS: Gerrit_Virtual_Submit
Reviewed-by: Sagar Kamble <skamble@nvidia.com>
Reviewed-by: Vijayakumar Subbu <vsubbu@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Add support in nvgpu to parse and get the freq cap from DT.
The patch does below
Parse the DT and gets the freq cap value during probe.
During clk_arb init compare this with P0.Max and takes the lowest.
Send change_seq with the new value and set dgpu freq.
Use the lowest for "get points","get default","set VF".
Bug 200556366
Change-Id: Ie10243f9bf83cb5ae07ebcc4cdc8efaffa56c309
Signed-off-by: Abdul Salam <absalam@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/2204644
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
gk20a_pm_runtime_suspend can fail and invoke gk20a_pm_finalize_poweron
that can cause double mapping of the usermode mmap region via
io_remap_pfn_range(). Avoid this by using a boolean variable to track
whether the region is already mapped.
Bug 2707416
Change-Id: I4d8cbe427400a5b986348a19af145367cc08ffc6
Signed-off-by: Debarshi Dutta <ddutta@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/2229312
GVS: Gerrit_Virtual_Submit
Reviewed-by: Dinesh T <dt@nvidia.com>
Reviewed-by: Vijayakumar Subbu <vsubbu@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
The devinit executes in parallel with PCIE link training
to reduce exit latency. Therefore, all PCIE settings that
normally occur during devinit after the PCIE link is up are
deferred until nvgpu has resumed control.
Bug 2661545
Change-Id: Ifdd4f645b2e1791d93567cc34d6ab0691a25d101
Signed-off-by: Abdul Salam <absalam@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/2210625
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Move graphics related defs and functions under CONFIG_NVGPU_GRAPHICS
switch.
Move classes not supported in GV11B under CONFIG_NVGPU_NON_FUSA
switch.
Add missing valid class numbers to gpu_class.is_valid HAL.
Also remove un-used class defs from class.h header.
Lot of qnx safety tests are still using graphics 3d class.
Until those tests got fixed, allowing 3d graphics class
as valid class for safety build.
JIRA NVGPU-4301
Change-Id: Ifd2a13bee3210821799c2bca10e7245eb3c79121
Signed-off-by: Seshendra Gadagottu <sgadagottu@nvidia.com>
Signed-off-by: Tejal Kudav <tkudav@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/2224658
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
gk20a.h will include gops_mc.h to contain the mc ops definitions. Add
doxygen comments for the HAL functions that are called directly.
Also move mc_gp10b_intr_pmu_unit_config to non-fusa HAL file.
JIRA NVGPU-2524
Change-Id: I4f326332d7842211b004b372d79fac9fe6ed40e7
Signed-off-by: Sagar Kamble <skamble@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/2226017
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Writing same value to railgate_enable_store should be treated as nop
and made successfully. Doing so is not only an optimization for the
operation but also convention that users expect for "settings". This
change is primary for fixing a peculiar situation in the driver:
root@localhost:/sys/devices/17000000.gp10b# cat railgate_enable
0
root@localhost:/sys/devices/17000000.gp10b# echo 0 > railgate_enable
bash: echo: write error: Invalid argument
Attempt to disable railgating on a platform where railgating isn't
supported shouldn't be treated as 'invalid'. It's disabled after all.
Bug 200562094
Change-Id: I3c04934bdbaf337c33d7de9cac6d53c96b4dacae
Signed-off-by: Leon Yu <leoyu@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/2225476
(cherry picked from commit 10b3b5b1d5)
Reviewed-on: https://git-master.nvidia.com/r/2226185
Reviewed-by: Automatic_Commit_Validation_User
GVS: Gerrit_Virtual_Submit
Reviewed-by: Bibek Basu <bbasu@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Currently nvgpu reads the temperature by reading the
NV_THERM_I2CS_SENSOR_00 register. Below are the issues
with current approach
1) NV_THERM_I2CS_SENSOR_00 doesn't support
fractional precision which is POR.
2) It doesn't support negative temperatures which
is required for Auto.
3) It doesn't take into account the right POR
sensor in VFE VBIOS tables.
From therm channel get status interface we can read the
current temperature from PMU.
NVBUG - 200549047
Change-Id: I2fb21926208876f3d3bebe3f2dee08edafedbc7d
Signed-off-by: rmylavarapu <rmylavarapu@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/2196224
Reviewed-by: svc-mobile-coverity <svc-mobile-coverity@nvidia.com>
GVS: Gerrit_Virtual_Submit
Reviewed-by: Vijayakumar Subbu <vsubbu@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
nvgpu.common.unit was just an enum used for passing to nvgpu.common.mc
APIs. So, move the enum into mc.h, and replace the include of unit.h
with mc.h where appropriate. And update the yaml arch.
JIRA NVGPU-4144
Change-Id: I210ea4d3b49cd494e43add1b52f3fbcdb020a1e3
Signed-off-by: Philip Elcan <pelcan@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/2216106
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
When virtualized, the guest OS has no direct access to
PMU functionality:
- Don't create debugfs entries that rely on PMU access
- Clean up PMU vgpu HAL entries that imply that PMU access
is supported
Bug 200543218
Change-Id: I12730b600802448a240f3de042760041d3ae7d29
Signed-off-by: Peter Daifuku <pdaifuku@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/2213650
Reviewed-by: Automatic_Commit_Validation_User
GVS: Gerrit_Virtual_Submit
Reviewed-by: Alex Waterman <alexw@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
MISRA Rule 17.1 forbids use of stdarg.h features which are defined for
variable arguments.
This patch modifies logging macros to use slogf function for QNX builds.
This avoids use of variable argument functions used for formatting log
message.
Jira NVGPU-4075
Change-Id: I5b6bb1107a7e431afaa960003858193a477b2ee6
Signed-off-by: Vedashree Vidwans <vvidwans@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/2192016
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
IRQs were not enabled before nvgpu_finalize_poweron, so debugging early
init issues such as MMU fault, invalid PRIV ring or bus access etc.
triggered during nvgpu power-on was cumbersome. Hence, Enable the
IRQs before nvgpu_finalize_poweron is called.
In HUB (MMU fault) ISR, MMU fault handling is only limited to snapped
in priv reg in case of fault during nvgpu power-on.
In HUB (MMU fault) ISR, access to fault buffers is synchronized as
nvgpu driver reads the fault buffer registers before proceeding
with fault handling. However, additional MMU fault handling
needs to be synchronized with GR/FIFO/quiesce/recovery setup
through nvgpu power-on state.
JIRA NVGPU-1592
Change-Id: I8a5f2fcd79cb7ad8e215359e7a9fad50bfd46d67
Signed-off-by: Sagar Kamble <skamble@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/2203861
Reviewed-by: svc-mobile-coverity <svc-mobile-coverity@nvidia.com>
Reviewed-by: svc-mobile-misra <svc-mobile-misra@nvidia.com>
Reviewed-by: Philip Elcan <pelcan@nvidia.com>
GVS: Gerrit_Virtual_Submit
Reviewed-by: Vijayakumar Subbu <vsubbu@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
IRQs can get triggered during nvgpu power-on due to MMU fault, invalid
PRIV ring or bus access etc. Handlers for those IRQs can't access the
full state related to the IRQ unless nvgpu is fully powered on.
In order to let the IRQ handlers know about the nvgpu power-on state
gk20a.power_on_state variable has to be protected through spinlock
to avoid the deadlock due to usage of earlier power_lock mutex.
Further the IRQs need to be disabled on local CPU while updating the
power state variable hence use spin_lock_irqsave and spin_unlock_-
irqrestore APIs for protecting the access.
JIRA NVGPU-1592
Change-Id: If5d1b5e2617ad90a68faa56ff47f62bb3f0b232b
Signed-off-by: Sagar Kamble <skamble@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/2203860
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Currently after sending change seq RPC, nvgpu waits for a fixed time
of 20ms.
This CL replaces this with pmu_wait_message_cond, which will return
immediately after getting change seq completion event.
Also added debug fs node to get the change seq execution time.
Bug 200545366
Change-Id: Iba283f65d4949858be9cbff88de4d21a8c92ff81
Signed-off-by: Abdul Salam <absalam@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/2202423
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
golden image size will be set when memory allocated.
See function:
- nvgpu_gr_obj_ctx_init
If golden image size is 0, gr_golden_image should be a NULL
pointer in most cases. So add NULL pointer checking in
tpc_pg_mask_store to avoid NULL pointer exception.
Bug 2403210
Change-Id: I14df5cd94d7a4418c3089c5f84b6eab93c485ba6
Signed-off-by: Sunny Li <sunnyl@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/2161280
Reviewed-by: Automatic_Commit_Validation_User
Reviewed-by: Debarshi Dutta <ddutta@nvidia.com>
GVS: Gerrit_Virtual_Submit
Reviewed-by: Deepak Nibade <dnibade@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
This patch addresses misra violations due to SDL error reporting
callbacks. In particular, it addresses the following misra violation:
- misra_c_2012_directive_4_7_violation: Calling function
"nvgpu_report_*_err()" which returns error information without testing
the error information.
JIRA NVGPU-4025
Change-Id: Ia10b6b3fd9c127a8c5189c3b6ba316f243cedf04
Signed-off-by: Rajesh Devaraj <rdevaraj@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/2196895
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
ifdef function prototypes with CONFIG_* defines. This fixes MISRA rule
8.6 violations which complain about undefined functions.
Also moved nvgpu_channel_get_from_file prototype to ioctl_channel.h &
nvgpu_probe to driver_common.h as those are linux specific. Define
nvgpu_init_soc_vars in posix/soc.c as it is implemented in QNX.
JIRA NVGPU-3873
Change-Id: I5d2b238e1b5d1318867cd2416ac5f03cc6ab7c6a
Signed-off-by: Sagar Kamble <skamble@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/2196794
Reviewed-by: svc-mobile-coverity <svc-mobile-coverity@nvidia.com>
Reviewed-by: Deepak Nibade <dnibade@nvidia.com>
GVS: Gerrit_Virtual_Submit
Reviewed-by: Alex Waterman <alexw@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
In some cases, we would get deadlock issue due to there are two locks
acquisition on common clk driver's lock and nvgpu driver's locks. At
the bug, inconsistent lock ordering problem will come with one thread
gets "nvgpu lock -> clk lock" and the other thread gets "clk lock ->
nvgpu lock".
Slove the latter path with one-time initializing clk_parent entry
and use cached data afterward.
Bug 2555115
Change-Id: I31c5c2728f406307e7cfd4e555f4db0c163234d8
Signed-off-by: Jeremy Ho <jeremyh@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/2146727
(cherry picked from commit 42c2bdfb9f)
Reviewed-on: https://git-master.nvidia.com/r/2160290
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Replaced ch->mmu_debug_mode_enabled with ch->mmu_debug_mode_refcnt.
If channel is enabled multiple times by userspace, then ref count is
updated accordingly. There is an expectation that enable/disable
calls are balanced for setting channel's mmu debug mode.
When unbinding the channel, decrease refcnt for the channel until it
reaches 0.
Also, removed tsg parameter from nvgpu_tsg_set_mmu_debug_mode as it
can be retrieved from ch.
Bug 2515097
Change-Id: If334e374a55bd14ae219edbfd3b1fce5ff25c226
Signed-off-by: Thomas Fleury <tfleury@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/2184702
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
This patch adds nvgpu API in linux and posix to query vpr resize.
The new API nvgpu_is_vpr_resize_enabled() is used in
nvgpu_submit_channel_gpfifo().
Previously, if non-deterministic channel has timeout disabled and
GPU cannot railgate on some platform, then channel doesn't power ref
count and results in video freeze. To resolve non-determinstic channel
job tracking needs to be enabled if vpr resize is supported or if GPU
can railgate.
Bug 200532122
Change-Id: Icfbff6253762b195b2f5955749343974b1a7a269
Signed-off-by: Vedashree Vidwans <vvidwans@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/2171093
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
For safety build, nvgpu driver should enter SW quiesce state
in case an uncorrectable error has occurred. In this state, any
activity on the GPU should be prevented, without powering off the GPU.
Also, a minimal set of operations should be used to enter SW quiesce
state.
Entering SW quiesce state does the following:
- set sw_quiesce_pending: when this flag is set, interrupt
handlers exit after masking interrupts. This should help mitigate
an interrupt storm.
- wake up thread to complete quiescing.
The thread performs the following:
- set NVGPU_DRIVER_IS_DYING to prevent allocation of new resources
- disable interrupts
- disable fifo scheduling
- preempt all runlists
- set error notifier for all active channels
Note: for channels with usermode submit enabled, userspace can
still ring doorbell, but this will not trigger any work on
engines since fifo scheduling is disabled.
Jira NVGPU-3493
Change-Id: I639a32da754d8833f54dcec1fa23135721d8d89a
Signed-off-by: Thomas Fleury <tfleury@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/2172391
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Changes added to support "rmmod nvgpu" in dgpu simulation after gpu
poweron.
nvgpu_engine-wait_for_idle got stuck in busy mode for nvdec and nvec
engines in simulation as simulation doesnt support timeout.
These engines are not valid engines in nvgpu engine list.
Add nvgpu_engine_check_valid_id before checking engine status.
Simulation crash on accessing 0xb81604 top interrupt register.
Add func_priv_cpu_intr_top__size_1_v() function to get the supported
size than using default MAX_INTR_TOP_REGS.
nvlink is not supprted in dgpu simulation. Avoid warning for
-ENODEV return.
Avoid register read following gpu power off completion.
Bug 2498574
Change-Id: I9f9f1cf1ac4620242bda1d2cc0f29f51f81a6711
Signed-off-by: vinodg <vinodg@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/2179930
Reviewed-by: Seshendra Gadagottu <sgadagottu@nvidia.com>
GVS: Gerrit_Virtual_Submit
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
CE app functionality from nvgpu is non-safe for igpu. CE engines init
/reset/cg related functionality is required in safety. Hence move the
CE app logic under CONFIG_NVGPU_DGPU flag and update the sources
accordingly.
JIRA NVGPU-3814
Change-Id: I37aa00b1184baccd5fe569ec315be60ac42dac9b
Signed-off-by: Sagar Kamble <skamble@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/2168956
GVS: Gerrit_Virtual_Submit
Reviewed-by: Deepak Nibade <dnibade@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
