The PMU pstate deinit was invoked part of gpu power off. This frees and clears
the pmgr_pmu struct which causes the pmu remove support to crash when it
tries to access the pmgr_pmu object for freeing up the pmu board objects.
Deferred pstate deinit to nvgpu driver removal as there is no reason for it be
invoked part of prepare poweroff sequence.
JIRA NVGPU-1618
Change-Id: I2eb52000f0732d0abed54946e0843367b119d443
Signed-off-by: Antony Clince Alex <aalex@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1971225
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Follow-up change to rename g->ops.mm.bar1_map (and implementations)
to more specific g->ops.mm.bar1_map_userd.
Also use nvgpu_big_zalloc() to allocate userd slabs memory descriptors.
Bug 2422486
Bug 200474793
Change-Id: Iceff3bd1d34d56d3bb9496c179fff1b876b224ce
Signed-off-by: Thomas Fleury <tfleury@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1970891
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
gk20a_fifo_recover_tsg does high-level software calls and
invokes gk20a_fifo_recover. This function belongs to the tsg unit and
is moved to tsg.c file. Also, the function is renamed to
nvgpu_tsg_recover.
Jira NVGPU-1237
Change-Id: Id1911fb182817b0cfc47b3219065cba6c4ca507a
Signed-off-by: Debarshi Dutta <ddutta@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1970034
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
gk20a_fifo_recover_ch does high-level calls and invokes
gk20a_fifo_recover. This function belongs to the channel unit and is
moved to the file channel.c. Also, the function is renamed to
nvgpu_channel_recover.
Jira NVGPU-1237
Change-Id: I31890f85fdb2c42648cc063dd9c4e7e35930dcef
Signed-off-by: Debarshi Dutta <ddutta@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1970033
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Any tsg specific functions that does high-level software-centric
operations below to the TSG unit and not the FIFO unit.
Move the below public functions as well as their dependent
static functions to common/fifo/tsg.c and also rename them to use the
prefix nvgpu_tsg_*
gk20a_fifo_set_ctx_mmu_error_tsg
gk20a_fifo_abort_tsg
gk20a_fifo_error_tsg
gk20a_fifo_check_tsg_ctxsw_timeout
Jira NVGPU-1237
Change-Id: I4e3da821a878d4b4a0a0b53fbb7f4c10f135f58d
Signed-off-by: Debarshi Dutta <ddutta@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1934299
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Any channel specific functions having high-level software-centric
operations belong to the channel unit and not the FIFO unit.
Move the below public functions as well as their dependent
static functions to common/fifo/channel.c. Also, rename the functions
to use the prefix nvgpu_channel_*.
gk20a_fifo_set_ctx_mmu_error_ch
gk20a_fifo_error_ch
gk20a_fifo_check_ch_ctxsw_timeout
Jira NVGPU-1237
Change-Id: Id6b6d69bbed193befbfc4c30ecda1b600d846199
Signed-off-by: Debarshi Dutta <ddutta@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1932358
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
The function gk20a_fifo_sched_disable_rw accesses HW directly. Rename it
and add a HAL indirection so that it can be called from chip-independent
code.
Also fix some trivial MISRA violations in the function.
Jira NVGPU-1309
Change-Id: Icf320738d3d1d4baa40257a9da3ca2c6b7fefc0b
Signed-off-by: Konsta Holtta <kholtta@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1971274
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
g->ops.gr.alloc_gr_ctx HAL right now allocates graphics context and
also initializes preemption mode for various platforms
Separate out a new HAL g->ops.gr.init_ctxsw_preemption_mode that
initializes preemption mode and call it from gk20a_alloc_obj_ctx()
after context is created
g->ops.gr.alloc_gr_ctx now only allocates the context as the name
suggests
Jira NVGPU-1527
Change-Id: I8a44672d5ab2ebfe315e6334115265e4ee4f24f0
Signed-off-by: Deepak Nibade <dnibade@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1972254
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
g->ops.gr.enable_cde_in_fecs and g->ops.gr.update_boosted_ctx
are no longer required since we can directly call
g->ops.gr.ctxsw_prog.set_cde_enabled and
g->ops.gr.ctxsw_prog.set_pmu_options_boost_clock_frequencies
respectively
remove those functions and the ops
Jira NVGPU-1526
Change-Id: Idb0ad5f634e78aac44ec325ba2b7f59c612b29e8
Signed-off-by: Deepak Nibade <dnibade@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1972184
GVS: Gerrit_Virtual_Submit
Reviewed-by: Konsta Holtta <kholtta@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>
Commit ca611e4d0e (gpu: nvgpu: verify usermode mapping is at least
PAGE_SIZE) was not quite the right thing to do; do_mmap() rounds the
length up to a page boundary anyway, but the length must not be longer
than the size of the usermode region which is 64 KB to avoid leaking
access to other registers.
Bug 2441531
Change-Id: Ib1c88a6725db62c8276b6e8b880631227a4fc8cd
Signed-off-by: Konsta Holtta <kholtta@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1971339
Reviewed-by: Alex Waterman <alexw@nvidia.com>
GVS: Gerrit_Virtual_Submit
Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com>
Reviewed-by: Allen Martin <amartin@nvidia.com>
Reviewed-by: Debarshi Dutta <ddutta@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
MISRA rule 16.3 states all switch clause to have break statement.
Fixing the missing break statement for default case.
Two consecutive labels without statement need not have break.
So no break is added for case without any statement.
This also makes the switch statement well-formed covering 16.1.
JIRA NVGPU-1510
JIRA NVGPU-1536
Change-Id: I2c16888cbcb429f49f1f260e18b7c6eba55aa83c
Signed-off-by: Abdul Salam <absalam@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1964328
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
MISRA Rule 10.4 only allows the usage of arithmetic operations on
operands of the same essential type category.
Adding "U" at the end of the integer literals to have same type of
operands when an arithmetic operation is performed.
This fixes violation where an arithmetic operation is performed on
signed and unsigned int types.
JIRA NVGPU-992
Change-Id: Icb724f3424c8161c12b69d373ff08c7648f79e56
Signed-off-by: Sai Nikhil <snikhil@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1834225
GVS: Gerrit_Virtual_Submit
Reviewed-by: Adeel Raza <araza@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Remove gp106 and add gv100 unit test
This is as a part of removing gp106 support
This adds testing for gv100 to the fuse unit test
Removes check_sec/non_sec as they are not used
Removes delta SRAM fuse as they are not present in GV100
Bug 200457373
Change-Id: I9bb4b714500eae01d0df00bb9f6842d4d4fbfd12
Signed-off-by: Abdul Salam <absalam@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1960034
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>
GSPLITE falcon base address was being set without invoking hal api.
This patch defines gpu_ops.gsp.falcon_base_addr hal api to get this
base address.
JIRA NVGPU-1587
Change-Id: Id187b34d022f90c09b8762cdab7769323b607cc0
Signed-off-by: Sagar Kamble <skamble@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1969432
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>
GPCCS falcon base address was being set without invoking hal api. Remove
FALCON_GPCCS_BASE. This patch defines gpu_ops.gr.gpccs_falcon_base_addr
hal api to get this base address.
JIRA NVGPU-1587
Change-Id: Icfa7a26d1bb2d67c81f05a43f6ce906f59706b3d
Signed-off-by: Sagar Kamble <skamble@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1969431
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>
FECS falcon base address was being set without invoking hal api. Remove
FALCON_FECS_BASE. This patch defines gpu_ops.gr.fecs_falcon_base_addr hal
api to get this base address.
JIRA NVGPU-1587
Change-Id: I9c8e60be4ee81a154020c982893725a12ebb72ef
Signed-off-by: Sagar Kamble <skamble@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1969430
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>
SEC2 falcon base address was being set without invoking hal api. Remove
FALCON_SEC_BASE. This patch defines gpu_ops.sec2.falcon_base_addr hal api
to get this base address.
Also, don't initialize the base for non-supported falcons.
JIRA NVGPU-1587
Change-Id: Iad19a9987416076cf9090d30a48ff83369cf73c2
Signed-off-by: Sagar Kamble <skamble@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1969429
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>
PMU falcon base address was being set without invoking hal api. Remove
FALCON_PWR_BASE. This patch defines gpu_ops.pmu.falcon_base_addr hal api
to get this base address.
JIRA NVGPU-1587
Change-Id: I5c3f27e89bdcc775025bc8d4fa9cf0af11ceb002
Signed-off-by: Sagar Kamble <skamble@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1969428
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>
NVDEC falcon base address was being set without invoking hal api. Remove
FALCON_NVDEC_BASE. This patch defines gpu_ops.fb.falcon_base_addr hal api
to get this base address. Currently gp106 and tu104 have these
implemented. gv100 uses the gp106 hal interface.
Also, don't initialize the base for non-supported falcons.
JIRA NVGPU-1587
Change-Id: I0be759b8462ede9b85690a70431480afdee9602c
Signed-off-by: Sagar Kamble <skamble@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1969427
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>
For gk20a & gp106, nvgpu driver currently hardcodes base address for FECS,
GPCCS, NVDEC falcons. Instead, emit those register in the hw headers. Base
address is the address of IRQSSET register.
JIRA NVGPU-1587
Change-Id: I710c4006b5ef6ded7980491ac157f8c03399b831
Signed-off-by: Sagar Kamble <skamble@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1966643
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 unit test verifies that the PD cache is capable of packing
multiple PD allocations into a single page. The verification works
by using the DMA fault injection method to prevent more than one
DMA alloc from happening in the pd_cache code.
If an OOM error is detected by the pd_cache code before the max
number of PDs per page is allocated by the pd_cache then there's
a failure in the packing code.
The correct number of PDs per page is easy to calculate: it's
just PAGE_SIZE / pd_size.
JIRA NVGPU-1323
Change-Id: I2c5b58ecf32da48d503c576012d9e564a270144c
Signed-off-by: Alex Waterman <alexw@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1970876
Reviewed-by: Konsta Holtta <kholtta@nvidia.com>
GVS: Gerrit_Virtual_Submit
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Added gpc_ppc_count, pes_tpc_count and pes_tpc_mask to constants so they
are used to init the according structure in gr.
They are needed to support get_nonpes_aware_tpc.
Bug 200452721
Jira GVSCI-91
Change-Id: Idbb59daab58c9877eab373439771b19c28cbcef1
Signed-off-by: Richard Zhao <rizhao@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1967997
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
PMU counters #0 and #4 are used to count total cycles and busy cycles.
These counts are used by podgov to estimate GPU load.
PMU idle intr status register is used to monitor overflow. Overflow
rarely occurs because frequency governor reads and resets the counters
at a high cadence. When overflow occurs, 100% work load is reported to
frequency governor.
Bug 1963732
Change-Id: I046480ebde162e6eda24577932b96cfd91b77c69
Signed-off-by: Peng Liu <pengliu@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1939547
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
NVGPU_CTXSW_IOCTL_RING_SETUP can be used to setup a custom ring buffer
and it accepts size via arguments. nvgpu driver will return an error
if size requested is greater than 128 * 4096 but this value is hardcoded
and not exposed anywhere.
Add characteristics field in nvgpu.h to expose this size so that corresponding
nvrm_gpu API can use it.
Bug 2169674
Change-Id: Icf9465d4eec6ba3a307ea9490bd5da563944e4f6
Signed-off-by: Anup Mahindre <amahindre@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1967596
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
We had to force allocation of physically contiguous memory for
USERD in nvlink case, as a channel's USERD address is computed as
an offset from fifo->userd address, and nvlink bypasses SMMU.
With 4096 channels, it can become difficult to allocate 2MB of
physically contiguous sysmem for USERD on a busy system.
PBDMA does not require any sort of packing or contiguous USERD
allocation, as each channel has a direct pointer to that channel's
512B USERD region. When BAR1 is supported we only need the GPU VAs
to be contiguous, to setup the BAR1 inst block.
- Add slab allocator for USERD.
- Slabs are allocated in SYSMEM, using PAGE_SIZE for slab size.
- Contiguous channels share the same page (16 channels per slab).
- ch->userd_mem points to related nvgpu_mem descriptor
- ch->userd_offset is the offset from the beginning of the slab
- Pre-allocate GPU VAs for the whole BAR1
- Add g->ops.mm.bar1_map() method
- gk20a_mm_bar1_map() uses fixed mapping in BAR1 region
- vgpu_mm_bar1_map() passes the offset in TEGRA_VGPU_CMD_MAP_BAR1
- TEGRA_VGPU_CMD_MAP_BAR1 is called for each slab.
Bug 2422486
Bug 200474793
Change-Id: I202699fe55a454c1fc6d969e7b6196a46256d704
Signed-off-by: Thomas Fleury <tfleury@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1959032
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
This adds APIs to the POSIX sgt module to allow unit tests to create
sgt's and sgl's more easily. The new APIs allow the caller to pass in a
table of sgl's and creates the new sgt/sgl as needed.
Since this there is a new API to create an sgl, this also includes a new
nvgpu_sgl_free() API.
JIRA NVGPU-1563
Change-Id: I27ab7654289cbd2212a75625b6dd24d2b742e3bf
Signed-off-by: Philip Elcan <pelcan@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1966151
Reviewed-by: svc-mobile-coverity <svc-mobile-coverity@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>
There were several places were nvgpu_mem structs were declared on the
stack and not initialized. Some fields could be used erroneously by
underlying APIs.
To allow the initialization of an array of nvgpu_mem in
test_dma_alloc_fi_delayed_enable(), the array length had to be changed
from a const to a macro.
JIRA NVGPU-1563
Change-Id: I6c23445af7d932710898df2e23a8fd892bee77b3
Signed-off-by: Philip Elcan <pelcan@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1967966
Reviewed-by: svc-mobile-coverity <svc-mobile-coverity@nvidia.com>
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>
Add separate new unit gr/ctxsw_prog that provides interface to access
h/w header files hw_ctxsw_prog_*.h
Add below chip specific files that access above h/w unit and provide
interface through g->ops.gr.ctxsw_prog.*() HAL for rest of the units
common/gr/ctxsw_prog/ctxsw_prog_gm20b.c
common/gr/ctxsw_prog/ctxsw_prog_gp10b.c
common/gr/ctxsw_prog/ctxsw_prog_gv11b.c
Remove all the h/w header includes from rest of the units and code.
Remove direct calls to h/w headers ctxsw_prog_*() and use HALs
g->ops.gr.ctxsw_prog.*() instead
In gr_gk20a_find_priv_offset_in_ext_buffer(), h/w header
ctxsw_prog_extended_num_smpc_quadrants_v() is only defined on gk20a
And since we don't support gk20a remove corresponding code
Add missing h/w header ctxsw_prog_main_image_pm_mode_ctxsw_f() for
some chips
Add new h/w header ctxsw_prog_gpccs_header_stride_v()
Jira NVGPU-1526
Change-Id: I170f5c0da26ada833f94f5479ff299c0db56a732
Signed-off-by: Deepak Nibade <dnibade@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1966111
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
MISRA Rule 10.4 only allows the usage of arithmetic operations on
operands of the same essential type category.
Adding "U" at the end of the integer literals or casting operands
to have same type of operands when an arithmetic operation is
performed.
This fixes violations where an arithmetic operation is performed on
signed and unsigned int types.
JIRA NVGPU-992
Change-Id: I27e3e59c3559c377b4bd3cbcfced90fdf90350f2
Signed-off-by: Sai Nikhil <snikhil@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1921459
Reviewed-by: svc-misra-checker <svc-misra-checker@nvidia.com>
GVS: Gerrit_Virtual_Submit
Reviewed-by: Adeel Raza <araza@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
