Delete the struct nvgpu_engine_info as it's essentially identical to
struct nvgpu_device. Duplicating data structures is not ideal as it's
terribly confusing what does what.
Update all uses of nvgpu_engine_info to use struct nvgpu_device. This
is often a fairly straight forward replacement. Couple of places though
where things got interesting:
- The enum_type that engine_info uses is defined in engines.h and
has a bit of SW abstraction - in particular the GRCE type. The only
place this seemed to be actually relevant (the IOCTL providing device
info to userspace) the GRCE engines can be worked out by comparing
runlist ID.
- Addition of masks based on intr_id and reset_id; those can be
computed easily enough using BIT32() but this is an area that
could be improved on.
This reaches into a lot of extraneous code that traverses the fifo
active engines list and dramtically simplifies this. Now, instead of
having to go through a table of engine IDs that point to the list of
all host engines, the active engine list is just a list of pointers to
valid engines. It's now trivial to do a for-all-active-engines type
loop. This could even be turned into a generic macro or otherwise
abstracted in the future.
JIRA NVGPU-5421
Change-Id: I3a810deb55a7dd8c09836fd2dae85d3e28eb23cf
Signed-off-by: Alex Waterman <alexw@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2319895
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
struct nvgpu_gr is right now initialized during probe and from OS
specific code. To support multiple instances of graphics engine,
nvgpu needs to initialize nvgpu_gr after number of engine instances
have been enumerated in poweron path.
Hence move nvgpu_gr_alloc() to poweron path and after gr manager has
been initialized.
Some of the members of nvgpu_gr are initialized in probe path and they
too are in OS specific code. Move them to common code in
nvgpu_gr_alloc()
Add field fecs_feature_override_ecc_val to struct gk20a to store the
override flag read from device tree. This flag is later copied to
nvgpu_gr in poweron path.
Update tpc_pg_mask_store() to check for g->gr being NULL before
accessing golden image pointer.
Update tpc_fs_mask_store() to return error if g->gr is not initialized.
This path needs nvgpu_gr struct initialized. Also fix the incorrect
NULL pointer check in tpc_fs_mask_store() which breaks the write path
to this sysfs.
Jira NVGPU-5648
Change-Id: Ifa2f66f3663dc2f7c8891cb03b25e997e148ab06
Signed-off-by: Deepak Nibade <dnibade@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2397259
Reviewed-by: automaticguardword <automaticguardword@nvidia.com>
Reviewed-by: svc-mobile-coverity <svc-mobile-coverity@nvidia.com>
Reviewed-by: svc-mobile-cert <svc-mobile-cert@nvidia.com>
Reviewed-by: Lakshmanan M <lm@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
GVS: Gerrit_Virtual_Submit
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Currently, NVGPU_SUPPORT_FECS_CTXSW_TRACE enabled flag is set to true
when fecs_trace s/w setup is executed successfully. Sometimes,
fecs_trace is required to be disabled for debugging. This change will
help disable/enable fecs_trace feature by modifying one of the enabled
flags.
Enable NVGPU_SUPPORT_FECS_CTXSW_TRACE during chip specific hal init.
Control fec_trace init and ctxsw dev open depending on
NVGPU_SUPPORT_FECS_CTXSW_TRACE flag status.
JIRA NVGPU-5616
Change-Id: Id0754a5af7cd95a67a1f0ae5de36115d44e1111b
Signed-off-by: Vedashree Vidwans <vvidwans@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2357501
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
The driver configures the sm hww global, warp ESR report masks during poweron
as part of gops_gr.gr_init_support. However, during golden context init, these
are overwritten with default entries from sw_ctx_load list; this leaves the
report masks in a state inconsistent with the driver expectation.
The driver should configure the sm hww warp, global ESR report masks during
golden context init and not before it; Hence, move set_hww_esr_report_mask from
power-on path to golden context init.
In addition, update set_hww_esr_report_mask to do RMW, so as to retain the
values loaded from sw_ctx_load list.
Update global ESR report mask to enable all exceptions.
Bug 3029888
Bug 2997718
Change-Id: Id7ad4cff5409982143f49695c95c5e1d1c9fdec9
Signed-off-by: Antony Clince Alex <aalex@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2367466
Reviewed-by: automaticguardword <automaticguardword@nvidia.com>
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: svc-mobile-misra <svc-mobile-misra@nvidia.com>
Reviewed-by: Vaibhav Kachore <vkachore@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
GVS: Gerrit_Virtual_Submit
While booting LS falcons, gr.falcon.bind_instblk gops is
used to bind WPR VA to gr falcon. Only FECS_METHOD must be
used to bind instblks. But at this point FECS falcon is not loaded
and running. Hence FECS_METHOD cannot be used to bind this instblk.
Besides that, this code is not required
for successful falcon boot and functioning of chips other
than gm20b.
JIRA NVGPU-5323
Change-Id: I148ccc77d65d5f01adbba6261369e7a292dccfc3
Signed-off-by: smadhavan <smadhavan@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2369736
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Add the following gr gops functions:
- enable_gpc_crop_hww
- enable_gpc_zrop_hww
- handle_gpc_crop_hww
- handle_gpc_zrop_hww
- handle_gpc_rrh_hww
These gr gops will be used in nvgpu-next.
Add function: nvgpu_gr_rop_offset to compute rop pri offsets.
Jira: NVGPU-5237
Change-Id: I9e2437c1d2893238b16ec7a134543e20c81b49f7
Signed-off-by: Antony Clince Alex <aalex@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2335687
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Currently, we are performing obj ctx alloction for bellow classes
1. VOLTA_COMPUTE_A
2. VOLTA_DMA_COPY_A
3. VOLTA_CHANNEL_GPFIFO_A
In safety, we use Async CE but not GRCE.
So allocating obj context only for COMPUTE_A and return success(0) for
all other valid classes, after setting class in the channel struct.
Jira NVGPU-4378
Change-Id: Ie99872e062cc66f9ddf699397a13df85c3d8d59e
Signed-off-by: sagar <skadamati@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2287486
Reviewed-by: svc-mobile-coverity <svc-mobile-coverity@nvidia.com>
Reviewed-by: Deepak Nibade <dnibade@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
GVS: Gerrit_Virtual_Submit
Currently, size of zbc index table is defined as a macro. This macro is
independent of the number of address bits in the ltc zbc index register.
Adding below hal will update zbc index table size as per number of
address bits.
Add hal to get gr_zbc_index_table_size:
u32 (*zbc_table_size)(struct gk20a *g);
ZBC index table address 0 is reserved. Logic to start zbc table index
from 1 is moved to corresponding hals.
JIRA NVGPU-4838
Change-Id: I700cadfdd1f3dc5f323055b8f44d769d6627920a
Signed-off-by: Vedashree Vidwans <vvidwans@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2288479
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Below functions in common.gr.obj_ctx subunit include unnecessary
asserts to ensure value is not truncated when parsing into U32 size.
nvgpu_gr_obj_ctx_gr_ctx_alloc()
Make use of nvgpu_safe_cast_u64_to_u32() and remove unnecessary
asserts
Jira NVGPU-4778
Change-Id: Ic06c2f4131b3bba35222f7de5441f82ecee6d83d
Signed-off-by: Deepak Nibade <dnibade@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2277158
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
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 proposes modifications
in the software to modify the virtual SM id to TPC mapping across
the mission and redundant contexts. The virtual SM identifier to TPC
mapping is done by nvgpu when setting up the patch context.
The recommendation for the redundant setting is to offset the
assignment by one TPC, and not by one GPC. This will ensure that both
GPC and TPC diversity. The SM and Quadrant diversity will happen
naturally. For kernels with few CTAs, the diversity is guaranteed
to be 100%. In case of completely random CTA allocation,
e.g. large number of CTAs in the waiting queue, the diversity is
1 - 1/#SM, or 87.5% for GV11B, 97.9% for TU104.
Added NvGpu CFLAGS to enable/disable the SM diversity support
"CONFIG_NVGPU_SM_DIVERSITY".
This support is only enabled on gv11b and tu104 QNX non safety build.
JIRA NVGPU-4685
Change-Id: I8e3eaa72d8cf7aff97f61e4c2abd10b2afe0fe8b
Signed-off-by: Lakshmanan M <lm@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/2268026
Reviewed-by: svc-mobile-coverity <svc-mobile-coverity@nvidia.com>
Reviewed-by: svc-mobile-misra <svc-mobile-misra@nvidia.com>
Reviewed-by: Seshendra Gadagottu <sgadagottu@nvidia.com>
Reviewed-by: Shashank Singh <shashsingh@nvidia.com>
Reviewed-by: svc-mobile-cert <svc-mobile-cert@nvidia.com>
Reviewed-by: Automatic_Commit_Validation_User
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>
Add new unit test to cover gops.gr.init.ecc_scrub_reg HAL function
gops.gr.init.ecc_scrub_reg HAL can generate TIMEOUT errors which are
not returned to caller currently. Update this HAL to return int value
for error propagation.
Jira NVGPU-4458
Change-Id: I98f4d5af2ef17cc4301951fec4d660638c8ef72c
Signed-off-by: dnibade <dnibade@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/2265456
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Add negative tests that inject memory allocation failures and
HAL function call errors to verify error handling path in
common.gr.obj_ctx unit.
Update common.gr.setup test to cover invalid class input while
setting preemption mode.
Jira NVGPU-4457
Change-Id: I74d1ba63ba8aace6087b51fd50e2c136822d3a00
Signed-off-by: Deepak Nibade <dnibade@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/2260939
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
- nvgpu_gr_ctx_load_golden_ctx_image() does not return any error, change
the return type to void
- Check for preemption modes greater than CILP in
nvgpu_gr_ctx_check_valid_preemption_mode
- Check if received class is valid or not in
nvgpu_gr_setup_set_preemption_mode
- Compile out entire nvgpu_gr_obj_ctx_init_ctxsw_preemption_mode since
it is really not doing anything in safety
- Remove the switch statement in nvgpu_gr_obj_ctx_set_compute_preemption_mode
since it is not possible to receive any other value than supported.
Previous function calls ensure that input values are validated.
- nvgpu_gr_obj_ctx_commit_global_ctx_buffers() does not return any
error, change the return type to void
- gops.gr.init.preemption_state HAL is not needed in safety since it
only configures gfxp related timeout
- remove redundant call to gops.gr.init.wait_idle in
nvgpu_gr_obj_ctx_commit_hw_state. We trigger wait despite earlier
failure in same function call.
Jira NVGPU-4457
Change-Id: I06a474ef7cc1b16fbc3846e0cad1cda6bb2bf2af
Signed-off-by: Deepak Nibade <dnibade@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/2260938
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Split GR ECC initialization into GPC/TPC and FECS ECC init as FECS ECC
errors during acr_construct_execute need to be reported and handled
hence FECS ECC counters are required to be initialized before
acr_construct_execute.
GPC/TPC ECC counters are dependent on the GR config that will be
initialized only after acr_construct_execute.
nvgpu_gr_intr_init_support is moved to nvgpu_gr_prepare_sw.
FECS ECC interrupt is enabled by default hence interrupt is not
enabled through gr_fecs_host_int_enable_r in nvgpu_gr_prepare_sw.
JIRA NVGPU-4439
Change-Id: Ifc9912f0578015a6ba1e9d38765c42633632b15f
Signed-off-by: Sagar Kamble <skamble@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/2261987
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: Vinod Gopalakrishnakurup <vinodg@nvidia.com>
GVS: Gerrit_Virtual_Submit
Reviewed-by: Vaibhav Kachore <vkachore@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
get_gpc_mask hal is set only for tu104. Add CONFIG_NVGPU_DGPU check
in the code for using that hal.
gr_config_alloc_struct_mem function is called from nvgpu_gr_config_init
gr_config_free_mem is called gr_config_alloc_struct_mem on failure.
No need to call gr_config_free_mem from nvgpu_gr_config_init again
for failure.
nvgpu_gr_config_init allocate nvgpu_gr_config struct.
config->sm_to_cluster will never get allocated before.So no need to
check for config->sm_to_cluster and do a memset.
Jira NVGPU-4531
Change-Id: I928041c110019bec885f9d5b6978db3032bc493c
Signed-off-by: vinodg <vinodg@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/2262229
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: Deepak Nibade <dnibade@nvidia.com>
GVS: Gerrit_Virtual_Submit
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
handle_tex_exception hal is not set for safety build. Add
CONFIG_NVGPU_HAL_NON_FUSA checking for that hal.
log_mme_exception hal is supported only for turing. Add
CONFIG_NVGPU_DGPU checking for that hal.
gr_intr_handle_class_error always return -EINVAL. Change the
return as void to avoid unwanted error checking.
nvgpu_gr_intr_get_channel_from_ctx function parameter curr_tsgid will
never be NULL based on the current call. Remove unwanted
(curr_tsgid != NULL) check from this function.
Jira NVGPU-4454
Change-Id: I165d1cc5f9e308dfb11d905b59151b44f63a31bb
Signed-off-by: vinodg <vinodg@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/2259763
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
- Set ENOMEM error if GPU mapping fails in nvgpu_gr_ctx_alloc().
- In nvgpu_gr_ctx_free_patch_ctx(), it is not possible to have a
valid patch buffer without GPU virtual address space. Hence instead
of checking for gpu_va, use nvgpu_mem_is_valid() to check if GPU
mappings can be removed.
- Check if RTV circular buffer is ready within DGPU config.
nvgpu_gr_global_ctx_buffer_map() already checks if buffer is ready
or not and returns error if buffer is not ready.
- g->ops.gr.ctxsw_prog.init_ctxsw_hdr_data() will always be set for
each chip. Remove unnecessary NULL check.
Jira NVGPU-4373
Change-Id: Ib490f81f8b8299f87cffbb8a33fde8cf98e6c288
Signed-off-by: Deepak Nibade <dnibade@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/2259073
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
Reviewed-by: Vinod Gopalakrishnakurup <vinodg@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>
In nvgpu_gr_setup_set_preemption_mode function, nvgpu_channel_enable_tsg
and nvgpu_channel_disable_tsg calls are changed to gops calls.
In this function beginning nvgpu_tsg_from_ch is checked for NULL pointer
so nvgpu_channel_enable_tsg and nvgpu_channel_disable_tsg functions
never fail with NULL tsg pointer for branch coverage.
Jira NVGPU-3968
Change-Id: If2e1fee493426e56d62865b015dc8b21bad494b6
Signed-off-by: vinodg <vinodg@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/2258788
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: Automatic_Commit_Validation_User
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>
* Updated types and added error checks
* Modified GR condition for ctxsw disable count
CERT-C error check was added to detect error on integer overflow
But below logic couldn't detect first overflow, so updated condition
INT_MAX < gr->ctxsw_disable_count --> it became true after overflow
So, we didn't detected in first overflow and lead to assert on enable
JIRA NVGPU-3400
Change-Id: I6b0265a464f8f19efa7b0761612c6e9ffb3bd2bd
Signed-off-by: Sagar Kadamati <skadamati@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/2206282
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
HAL gops.gr.config.init_sm_id_table() initializes SM count in struct
nvgpu_gr_config. It is almost impossible that SM count is detected as
zero.
Hence remove the error check and add an assert instead.
This also helps with code coverage tests since it is difficult to
simulate error condition of having zero SMs detected.
Also, HAL gops.gr.config.init_sm_id_table() should always be defined
for each platform. Hence remove unnecessary check.
Jira NVGPU-4373
Change-Id: Ibd9b301b28d5bd2952367346a8f12fabcee2abd9
Signed-off-by: Deepak Nibade <dnibade@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/2247845
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: Seshendra Gadagottu <sgadagottu@nvidia.com>
GVS: Gerrit_Virtual_Submit
Reviewed-by: Vinod Gopalakrishnakurup <vinodg@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
