Reduce the size of memory allocations in the channel debug dump by
capturing only the necessary values from the instance block. This also
simplifies the allocation path slightly with the downside of having to
add a capture_channel_ram_dump HAL for reading the interesting parts
explicitly beforehand to the now smaller staging buffer.
Also rename struct ch_state to struct nvgpu_channel_dump_info.
Jira NVGPU-886
Change-Id: I5d7518d9d474b0b728b183383bc83d89ecf91b98
Signed-off-by: Konsta Holtta <kholtta@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1928207
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
MISRA rule 18.7 doesn't allow flexible array members. To work around
that, modify the instance block member in struct ch_state to be an
explicit pointer and allocate it separately for simplicity.
Jira NVGPU-886
Change-Id: I34299bec79bf7706f9cdfa42dee7fba765c9f312
Signed-off-by: Konsta Holtta <kholtta@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1928205
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
MISRA rule 10.4 doesn't allow arithmetic conversions on operands of
different essential type category.
Fix violations where an arithmetic conversion is performed on boolean
and non-boolean types.
JIRA NVGPU-994
Change-Id: I2af9937678462b632bb6ec6178e10d02104fc3bc
Signed-off-by: Amurthyreddy <amurthyreddy@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1832337
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
The context switch timeout works by triggering a hardware timeout at 10
Hz. When handling these, we check whether a channel has actually timed
out. Currently the timeout limit can be shorter than the 10 Hz interval
which always causes us to recover a channel but would also cause
detection of progress if there was any in the interval.
Handling both situations at the same time would reuse the channel
pointer local to the function after a loop has finished and would cause
memory corruption. Fix this by making the two branches mutually
exclusive, and move the recover case to happen first because that's how
our tests assume things to work.
Jira NVGPU-967
Change-Id: I26aa0fa7fd80ab42a9a1a93a6cca2cd29c9d3f3f
Signed-off-by: Konsta Holtta <kholtta@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1932449
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
There are two places where C standard identifiers are being
redefined. This is forbidden by MISRA rule 21.2. This patch will
fix two violations 'div' and 'exp' by renaming them 'divisor'
and 'exponent' respectively. Renamed 'man' as 'mantissa' for
uniformity.
JIRA NVGPU-1035
Change-Id: I0bbd38e5021c0047f9ce646dd6f90a30a3e4f3a5
Signed-off-by: smadhavan <smadhavan@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1852429
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
MISRA rule 17.2 forbids recursion as a hazard on the stack space. To
comply and additionally to make the code somewhat more straightforward
to read, rewrite the runlist construction with three explicit functions
that work as the three levels of the earlier recursion. These levels map
to the three priority levels of TSGs and having more than that is
unlikely.
When "runlist interleaving" is enabled, TSGs with higher priorities get
interleaved between the switch of each pair of lower-level priority
TSGs, so that the latency for a job at priority level X is no more than
all jobs' timeslices of priority X and higher, plus at most one job at a
lower level.
This can be illustrated as follows (low, medium, high TSGs 1 and 2):
L1 L2 (only low-priority TSGs)
H1 H2 (only high-priority TSGs)
H1 H2 M1 H1 H2 M2 (no low-priority TSGs)
M1 M2 L1 M1 M2 L2 (no high-priority TSGs)
H1 H2 L1 H1 H2 L2 (no medium-priority TSGs)
H1 H2 M1 H1 H2 M2 H1 H2 L1 H1 H2 M1 H1 H2 M2 H1 H2 L2 (no empty levels)
Without interleaving, the items are simply grouped by priority.
Jira NVGPU-1174
Change-Id: Ic3b5106945df7105633730ecd1d150af770a5e83
Signed-off-by: Konsta Holtta <kholtta@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1918226
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
MISRA rule-14.4 doesn't allow the usage of function pointers & integer
types as booleans in the controlling expression of an if statement or
an iteration statement.
Fix violations where a function pointer or a function whose return
value is an integer, is used as a boolean in the controlling expression
of if and loop statements.
JIRA NVGPU-1021
Change-Id: Ic5336268394ba4396ce80744c25930d2fb44dc42
Signed-off-by: Amurthyreddy <amurthyreddy@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1932147
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
The chid member of the channel_gk20a struct was being used as a unsigned
value. By being declared as an int, it was causing MISRA 10.3 violations
for implicit assignment of different types.
JIRA NVGPU-647
Change-Id: I7477fad6f0c837cf7ede1dba803158b1dda717af
Signed-off-by: Philip Elcan <pelcan@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1918470
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
MISRA 10.3 rule disallows implicit assignments between different
essential types. This adds casts to address some of these violations
in fifo_gk20a.
This also removes unnecessary bar1 test in
gk20a_fifo_handle_sched_error() (rather than add messy casting).
JIRA NVGPU-647
Change-Id: Ic8700459e47a59dc03e0149f6efb060efa4d4e42
Signed-off-by: Philip Elcan <pelcan@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1917635
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
MISRA 10.3 forbids assigning an object with a narrower essential type
or of a different essential type. This addresses the file
fifo_gk20a.c where constants were in violation.
JIRA NVGPU-647
Change-Id: I0ecf9b0ce40de76464efbde9e9c9b6aa69d80ec0
Signed-off-by: Philip Elcan <pelcan@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1917630
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
We right now acquire rulist_submit_mutex to submit runlist and also
to wait for submit completion
But locking is only needed to atomically configure the runlist submit
registers, hence move the locking to inside of
gk20a_fifo_runlist_hw_submit() where we program the registers
Also convert the mutex to spinlock at the same time
Note that similar locking is not required for
tu104_fifo_runlist_hw_submit() since the runlist submit registers
are per-runlist beginning Turing
Bug 200452543
Change-Id: I53d6179b80cb066466b64c6efa9393e55e381bfc
Signed-off-by: Deepak Nibade <dnibade@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1919058
Reviewed-by: svc-mobile-coverity <svc-mobile-coverity@nvidia.com>
Reviewed-by: svc-misra-checker <svc-misra-checker@nvidia.com>
GVS: Gerrit_Virtual_Submit
Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
We right now submit new runlist and wait for submit to complete in
gk20a_fifo_update_runlist_locked()
It is possible that multiple runlists are being updated in parallel
by multiple threads since the lock taken by parent of
gk20a_fifo_update_runlist_locked() is per-runlist
Note that the concurrent threads would still construct their runlists
into per-runlist buffer
But we still have a race condition while submitting these runlists
to hardware.
With an application that creates and destroys multiple contexts in
parallel this race condition gets realized and we see h/w reporting
an error interrupt NV_PFIFO_INTR_SCHED_ERROR_CODE_BAD_TSG which means
a bad TSG was submitted
Fix this by adding a global lock for runlist submit and wait sequence
This ensures that concurrent threads do not try to submit runlists
to the hardware at the same time
Bug 200452543
Bug 2405416
Change-Id: I2660a2e5d9af1da400e7f865361722dc0914f96f
Signed-off-by: Deepak Nibade <dnibade@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1851114
Reviewed-by: svc-mobile-coverity <svc-mobile-coverity@nvidia.com>
Reviewed-by: svc-misra-checker <svc-misra-checker@nvidia.com>
GVS: Gerrit_Virtual_Submit
Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
MISRA Rule-17.7 requires the return value of all functions to be used.
Fix is either to use the return value or change the function to return
void. This patch contains fix for calls to nvgpu_mutex_init and
improves related error handling.
JIRA NVGPU-677
Change-Id: I609fa138520cc7ccfdd5aa0e7fd28c8ca0b3a21c
Signed-off-by: Nicolas Benech <nbenech@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1805598
Reviewed-by: svc-misra-checker <svc-misra-checker@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>
This patch corrects a handful of MISRA 10.1 violations involving
illegal arithmetic operations (e.g. bitwise OR) on boolean values:
* fix to status handling in regops validation code
* fix to debugger event handling in gr code
* fix to entries_left tracking in runlist construct code
* fix to verbose channel dumping and reset tracking in fifo code
JIRA NVGPU-650
Change-Id: I3c3d9123b5a0e08fc936d0e63d51de99fc310ade
Signed-off-by: Scott Long <scottl@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1810957
Reviewed-by: svc-misra-checker <svc-misra-checker@nvidia.com>
GVS: Gerrit_Virtual_Submit
Reviewed-by: Alex Waterman <alexw@nvidia.com>
Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Fix MISRA rule 10.1 violations involving gk20a_nonstall_ops
enums by replacing them with with corresponding #defines.
Because these values can be used in expressions that require
unsigned values (e.g. bitwise OR) we cannot use enums.
The g->ce2.isr_nonstall() function was previously returning an
int that was a combination of gk20a_nonstall_ops enum bits which
led to the violations.
JIRA NVGPU-650
Change-Id: I6210aacec8829b3c8d339c5fe3db2f3069c67406
Signed-off-by: Scott Long <scottl@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1796242
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
In the current code, gk20a.h includes io.h which gets directly included
in a lot of other files. io.h contains methods which uses a struct
gk20a as a parameter leading to a circular dependency between io.h
and gk20a.h. This can be mitigated by removing io.h from gk20a.h as
part of larger effort to moving gk20a.h to nvgpu/gk20a.h
JIRA NVGPU-597
Change-Id: I93e504fa9371b88152737b342a75580c65e8f712
Signed-off-by: Debarshi Dutta <ddutta@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1787316
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
-For Si platforms, gk20a_get_gr_idle_timeout returns
3000 ms i.e. 3 sec. Currently this time is used for
preempt polling and this conflicts with channel
timeout if polling times out. Use fifo_eng_timeout_us converted
to ms for preempt polling.
-In case of preempt timeout, do not issue recovery
for si platform. ctxsw timeout will trigger recovery
if needed. For non si platforms, issue preempt timeout rc
if preempt times out.
Bug 2113657
Bug 2064553
Bug 2038366
Bug 2028993
Bug 200426402
Change-Id: I8d9f58be9ac634e94defa92a20fb737bf256d841
Signed-off-by: Seema Khowala <seemaj@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1762076
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>
-Recovery can be called for various types of faults. Acquire
runlist_lock for all runlists so that current teardown is done
before proceeding to next one.
-For legacy chips teardown is done by triggering mmu fault so
make sure runlist_locks are acquired during teardown and also
during handling mmu fault.
-gk20a_fifo_handle_mmu_fault is renamed as
gk20a_fifo_handle_mmu_fault_locked
-gk20a_fifo_handle_mmu_fault called from gk20a_fifo_teardown_ch_tsg
is replaced with gk20a_fifo_handle_mmu_fault_locked
-gk20a_fifo_handle_mmu_fault acquires/release runlist_lock for all
runlists and calls gk20a_fifo_handle_mmu_fault_locked
Bug 2113657
Bug 2064553
Bug 2038366
Bug 2028993
Change-Id: I973d7ddb6924b50bae2d095152867e99c87e780a
Signed-off-by: Seema Khowala <seemaj@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1761197
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 order to avoid the circular dependencies,
rearrange the static inline functions from
gk20a.h file.
Moved gk20a_gr_flush_channel_tlb function to
gr_gk20a.c and removed the #include gr_gk20a.h
from gk20a.h
Added a helper function utils.h to
move all generic static inline functions which
have no reference to gpu related structures.
ptimer related functions are moved to
ptimer.h
Implementations for as and pmu are moved to
corresponding files.
JIRA NVGPU-624
Change-Id: I4e956326e773ba037bf3a1696cc4c462085dbbe5
Signed-off-by: Vinod G <vinodg@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1781941
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
This reverts commit 0b02c8589d.
Originally change was reverted as it was making ap_compute test on
embedded-qnx-hv e3550-t194 fail. With fixes related to replacing tsg
preempt with runlist preempt during teardown, preempt timeout set to
100 ms (earlier this was set to 1000ms for t194 and 3000ms for legacy
chips) and not issuing preempt timeout recovery if preempt fails, helped
resolve the issue.
Bug 200426402
Change-Id: If9a68d028a155075444cc1bdf411057e3388d48e
Signed-off-by: Seema Khowala <seemaj@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1762563
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
-Add support for aborting runlist/s. Aborting runlist/s,
will abort all active tsgs and associated active channels
within these active tsgs
-Bare channels are no longer supported. Remove recovery
support for bare channels. In case there are bare
channels, recovery will trigger runlist abort
Bug 2125776
Bug 2108544
Bug 2105322
Bug 2092051
Bug 2048824
Bug 2043838
Bug 2039587
Bug 2028993
Bug 2029245
Bug 2065990
Bug 1945121
Bug 200401707
Bug 200393631
Bug 200327596
Change-Id: I6bec8a0004508cf65ea128bf641a26bf4c2f236d
Signed-off-by: Seema Khowala <seemaj@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1640567
Reviewed-by: svc-mobile-coverity <svc-mobile-coverity@nvidia.com>
GVS: Gerrit_Virtual_Submit
Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
-is_preempt_pending hal does not need timeout_rc_type input param as
for volta, reset_eng_bitmask is saved if preempt times out. For
legacy chips, recovery triggers mmu fault and mmu fault handler
takes care of resetting engines.
-For volta, no special input param needed to differentiate between
preempt polling during normal scenario and preempt polling during
recovery. Recovery path uses preempt_ch_tsg hal to issue preempt.
This hal does not issue recovery if preempt times out.
Bug 2125776
Bug 2108544
Bug 2105322
Bug 2092051
Bug 2048824
Bug 2043838
Bug 2039587
Bug 2028993
Bug 2029245
Bug 2065990
Bug 1945121
Bug 200401707
Bug 200393631
Bug 200327596
Change-Id: Ie76a18ae0be880cfbeee615859a08179fb974fa8
Signed-off-by: Seema Khowala <seemaj@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1709799
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Add g->fifo_eng_timeout_us to define engine timeout in microseconds.
It is initialized with GRFIFO_TIMEOUT_CHECK_PERIOD_US. In RM server
case, it can be overriden with value defined in device tree.
Jira EVLR-2674
Change-Id: I69ac2ce779fe575566c8ba48e8cd2d0e6b2d93cf
Signed-off-by: Thomas Fleury <tfleury@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1728391
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
