Create interrupt hal unit under hal.gr.intr.
This holds the interrupt and exception related hals.
Move enable_exceptions and enable_gpc_exceptions hal functions to
hal.gr.init location.
Modify enable_exceptions hal to pass gr->config and enable or disable
parameters.
Modify enable_gpc_exceptions to pass gr->config parameter.
Add new hal function enable_interrupts with enable or disable parameter
This hal helps to enable and disable the gr interrupts as needed.
gr init calls that use these hals are modified to
g->ops.gr.intr.enable_exceptions
g->ops.gr.intr.enable_gpc_exceptions
JIRA NVGPU-3016
Change-Id: Ib62f8bf0b5289b815c8eff4d32a47387f24af51b
Signed-off-by: Vinod G <vinodg@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/2077857
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Move g->ops.gr.commit_global_bundle_cb() hal to hal.gr.init unit as
g->ops.gr.init.commit_global_bundle_cb()
Remove register header accessor from gr_gk20a_commit_global_ctx_buffers()
and move it to hal functions
Move hal definitions to gm20b/gp10b hal files appropriately
Jira NVGPU-2961
Change-Id: I6358dce963857402aa1d4d5606bf75398b9be83d
Signed-off-by: Deepak Nibade <dnibade@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/2077216
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Move g->ops.gr.get_global_ctx_cb_buffer_size() and
g->ops.gr.get_global_ctx_pagepool_buffer_size() hals to hal.gr.init
unit
Move corresponding hal definitions to hal.gr.init unit
Jira NVGPU-2961
Change-Id: Ifff3e2073f6d9bca5b37244f7e107bad885e7ca7
Signed-off-by: Deepak Nibade <dnibade@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/2077215
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Add new unit vgpu subctx under common/vgpu/gr to manage
GR subcontext. This unit provides interfaces to allocate
and free subctx header.
Rename vgpu_subctx_gv11b* files to subctx_vgpu* files
renaming functions vgpu_gv11b_alloc_subctx_header to
vgpu_alloc_subctx_header and vgpu_gv11b_free_subctx_header
to vgpu_free_subctx_header which are called only if
NVGPU_SUPPORT_TSG_SUBCONTEXTS is enabled or free_channel_ctx_header
HAL op is set which is set only for gv11b for virtualization.
Also assign fifo HAL op free_channel_ctx_header to
vgpu_channel_free_ctx_header for vgpu gv11b which in turn
calls vgpu_free_subctx_header.
Jira GVSCI-334
Change-Id: Ib46e7be911632eba01cd21881077683b795f8bad
Signed-off-by: Aparna Das <aparnad@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/2075872
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Moved following ltc related data to struct nvgpu_ltc
and has a reference to it from struct gk20a:
struct nvgpu_spinlock ltc_enabled_lock;
u32 max_ltc_count;
u32 ltc_count;
u32 slices_per_ltc;
u32 cacheline_size;
Added function remove_support for ltc and it is called
during nvgpu remove sequence.
Added following helper functions in ltc.h:
u32 nvgpu_ltc_get_ltc_count(struct gk20a *g);
u32 nvgpu_ltc_get_slices_per_ltc(struct gk20a *g);
u32 nvgpu_ltc_get_cacheline_size(struct gk20a *g);
Removed redudnant ltc.init_fs_state call from vgpu init
sequence since it is getting called from nvgpu_init_ltc_support.
NVGPU-2044
Change-Id: I3c256dc3866f894c38715aa2609e85bd2e5cfe5a
Signed-off-by: Seshendra Gadagottu <sgadagottu@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/2073417
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Remove below variables from struct gr_gk20a
u32 bundle_cb_default_size;
u32 min_gpm_fifo_depth;
u32 bundle_cb_token_limit;
u32 attrib_cb_default_size;
u32 alpha_cb_default_size;
u32 attrib_cb_gfxp_default_size;
u32 attrib_cb_gfxp_size;
u32 attrib_cb_size;
u32 alpha_cb_size;
Instead add below hals in hal.gr.init unit to get all of above sizes
u32 (*get_bundle_cb_default_size)(struct gk20a *g);
u32 (*get_min_gpm_fifo_depth)(struct gk20a *g);
u32 (*get_bundle_cb_token_limit)(struct gk20a *g);
u32 (*get_attrib_cb_default_size)(struct gk20a *g);
u32 (*get_alpha_cb_default_size)(struct gk20a *g);
u32 (*get_attrib_cb_gfxp_default_size)(struct gk20a *g);
u32 (*get_attrib_cb_gfxp_size)(struct gk20a *g);
u32 (*get_attrib_cb_size)(struct gk20a *g, u32 tpc_count);
u32 (*get_alpha_cb_size)(struct gk20a *g, u32 tpc_count);
u32 (*get_global_attr_cb_size)(struct gk20a *g, u32 max_tpc);
Define these hals for all gm20b/gp10b/gv11b/gv100/tu104 chips
Also add hal.gr.init support for gv100 chip
Remove all accesses to variables from struct gr_gk20a and start using
newly defined hals
Remove below hals to initialize sizes since they are no more required
g->ops.gr.bundle_cb_defaults(g);
g->ops.gr.cb_size_default(g);
g->ops.gr.calc_global_ctx_buffer_size(g);
Also remove definitions of above hals from all the chip files
Jira NVGPU-2961
Change-Id: I130b578ababf22328d68fe19df581e46aebeccc9
Signed-off-by: Deepak Nibade <dnibade@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/2077214
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Change the global_ctx_buffer_index member in the nvgpu_gr_ctx struct
to be an enum nvgpu_gr_global_ctx_index. global_ctx_buffer_index is
used as an array of these indicies, but had been declared as an int.
This change resolves a number of MISRA Rule 10.3 violations for implicit
assignment of objects of different essential or narrower type.
In order to use this enum, it is moved out of global_ctx.h into a new
header file ctx_common.h that can be used by both ctx.h and global_ctx.h.
JIRA NVGPU-2955
Change-Id: I5e399ba3b0821d696aa0b9909d3bc6bbe99d274c
Signed-off-by: Philip Elcan <pelcan@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/2075753
Reviewed-by: svc-mobile-coverity <svc-mobile-coverity@nvidia.com>
Reviewed-by: svc-mobile-misra <svc-mobile-misra@nvidia.com>
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>
While resolving MISRA 10.8 violations in our definitions of U8_MAX,
U16_MAX, etc, some of the variants tried that looked valid created
invalid values due to integer promotions. So, for sanity, we should
validate these macros in the unit test for posix-env.
JIRA NVGPU-2955
Change-Id: I9c4b66771a875728c2bf79036e9a510bbc871bfb
Signed-off-by: Philip Elcan <pelcan@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/2078365
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>
The POSIX definitions of U8_MAX, U16_MAX, U32_MAX, and U64_MAX in use
was causing a MISRA 10.8 violation because the value 0U was being used
as an 8-bit value per the spec. This updates the definitions to cast 0
to the appropriate bit width instead of using 0U. This eliminates the
MISRA violation.
JIRA NVGPU-2955
Change-Id: Ib3220df8c08566b4594136f8a7deb0dec8b01ab3
Signed-off-by: Philip Elcan <pelcan@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/2078364
Reviewed-by: Automatic_Commit_Validation_User
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>
As part of ACR bootstrap, falcon bootstrap request is sent to engine HAL
functions along with bootloader structure & perform falcon boot, but
this adds constraint to HAL separation due to struct parameter, so
made ACR to handle falcon boot by using falcon interfaces along with
new HAL ops to setup engine falcon setup. This also helps to reduce
code duplication too.
JIRA NVGPU-2039
Change-Id: I6ca29390b74d75bad0467a3c17623a395ec9bc25
Signed-off-by: Mahantesh Kumbar <mkumbar@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/2072940
Reviewed-by: svc-misra-checker <svc-misra-checker@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>
- Make a new structure: nvgpu_pmu_pg for PG unit
- This new struct combines all PG unit variables like
elpg_stat, elpg_refcnt, pg_init etc. into one
structure as a part of PG unit refactoring.
- Use pmu_pg struct to access all PG variables.
- Eg: &pmu->pmu_pg.elpg_mutex, &pmu->pmu_pg.pg_mutex
and so on.
NVGPU-1973
Change-Id: I9973072826f4089f6315827bce49fa30dbcbcdda
Signed-off-by: Divya Singhatwaria <dsinghatwari@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/2071306
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Add nvgpu_tsg_set_error_notifier function for setting error_notifier
for all channels of a tsg.
Add nvgpu_tsg_timeout_debug_dump_state function for finding if
timeout_debug_dump is set for any of the channels of a tsg.
Add nvgpu_tsg_set_timeout_accumulated_ms to set
timeout_accumulated_ms for all the channels of a tsg.
JIRA NVGPU-1312
Change-Id: Ib2daf2d462c2cf767f5a6e6fd3436abf6860091d
Signed-off-by: Seema Khowala <seemaj@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/2077626
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Add new hal fifo_access to control the gr_gpfifo_ctl_r register to
enable or disable the access bit and semaphore access bit.
g->ops.gr.init.fifo_access function call with true or false
parameter to enable or disable the fifo_access.
JIRA NVGPU-2951
Change-Id: I67ad7ce9f176d7ce347e8acb425f7a4bb9e088ca
Signed-off-by: Vinod G <vinodg@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/2077705
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
This patch does the following.
1. Remove unused functions from volt_pmu.c.
2. Append public functions with nvgpu.
3. Remove GP106 functions and rename TU104 to generic functions.
4. Rename volt struct from gpu_ops.
5. Remove the unused volt.h header file.
6. Make local functions as static and put in order.
7. Remove unused inclusion on header files.
8. After 4, generic functions can be called directly instead of g->ops.
Jira NVGPU-1956
Change-Id: Icaea0ca817d37cccfc09241baa2f047ec2688169
Signed-off-by: Abdul Salam <absalam@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/2076535
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
init_pbdma_intr_descs HAL ops is used to update the internal values of
the struct intr within struct fifo_gk20a. Three kinds of
intr_descriptors are filled i.e. device_fatal_0, channel_fatal_0 and
restartable_0. Breaking them into separate HALs has the advantage of
reusing the h/w headers corresponding to the device_fatal_0 as they are
same across all the architectures while those of channel_fatal_0 varies.
Another advantage is to now decouple pbdma from filling in values
within the fifo_gk20a struct. A new method gk20a_fifo_init_pbdma_descs
is constructed that initializes the above intr struct by calling the
separate HAL ops for these.
Jira NVGPU-2950
Change-Id: I78ddc61a5d9b2088d34259af90f8b85817bf19d9
Signed-off-by: Debarshi Dutta <ddutta@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/2072741
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
The following HAL pointers are moved to a separate HAL unit named pbdma.
pbdma_acquire_val
get_pbdma_signature
dump_pbdma_status
handle_pbdma_intr_0
handle_pbdma_intr_1
read_pbdma_data
reset_pbdma_header
The functions corresponding to these HAL units are also moved to
pbdma_{arch} files under hal/fifo correspondinging to arch gm20b,
gp10b, gv11b and tu104. Any calls to gk20a_readl and gk20a_writel
are replaced by nvgpu_readl and nvgpu_writel respectively.
Jira NVGPU-2950
Change-Id: I9723f30ddf6582df02c03fceb1fba26a206e1230
Signed-off-by: Debarshi Dutta <ddutta@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/2071782
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Move load_smid_config and program_sm_id_numbering hal functions to
corresponding hal files in hal.gr.init.
Add new hal for get_sm_id_size and new static function in common.gr
init code for gr_load_sm_id_config.
JIRA NVGPU-2951
Change-Id: I27755306d278db9fcd5ab60169a64057b2989ea8
Signed-off-by: Vinod G <vinodg@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/2075875
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Move load_tpc_mask and setup_rop_mapping hal functions to hal.gr.init.
Existing load_tpc_mask hal code is split to two parts, one as a common
code in gr_load_tpc_mask and register write to init.tpc_mask hal
functions.
Modify pd_tpc_per_gpc and pd_skip_table_gpc hals in the
hal.gr.init to pass struct nvgpu_gr_config as a parameter.
JIRA NVGPU-2951
Change-Id: I52e26d0f023afa511a8cf8c3e4c54f45350be4ae
Signed-off-by: Vinod G <vinodg@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/2074892
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Currently ACR header files are part of "include/nvgpu/acr/" folder &
ACR interfaces are not used by any other UNIT which allows headers to
keep restricted to ACR unit, as ACR can be divided into two stage
process like blob preparation & bootstrap, so moved header files from
of "include/nvgpu/acr/" to "nvgpu/common/acr/" to respective blob/
bootstrap/acr header files along with its dependent interfaces, this
allows interfaces restricted to header file based on operation it does.
With this any access to ACR must go through provided public functions,
this header move change caused large code modification & required to
make it with big single CL to avoid build break.
JIRA NVGPU-2907
Change-Id: Idb24b17a35f7c7a85efe923c4e26edfd42b028e3
Signed-off-by: Mahantesh Kumbar <mkumbar@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/2071393
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
nvgpu_gr_ctx_load_golden_ctx_image() in common.gr.ctx unit programs
initial pm_mode in context. gk20a_alloc_obj_ctx() then disables pm_mode
ctxsw by default.
Fix this by disabling pm_mode ctxsw by default in
nvgpu_gr_ctx_load_golden_ctx_image() itself. Remove corresponding code
from gk20a_alloc_obj_ctx()
Jira NVGPU-1887
Change-Id: I6e1f83cefcb6229394da353e4cd87f1f5a0b10d4
Signed-off-by: Deepak Nibade <dnibade@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/2076273
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Add below two new APIs to set preemption buffer in graphics context or
subcontext respectively
nvgpu_gr_ctx_set_preemption_buffer_va()
nvgpu_gr_subctx_set_preemption_buffer_va()
Remove g->ops.gr.set_preemption_buffer_va() hal and use above APIs to
set preemption buffer VA.
Jira NVGPU-1887
Change-Id: I38fb76eaf01d3fc73fd8104f30bcd89be9fa45b6
Signed-off-by: Deepak Nibade <dnibade@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/2076272
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Add new hals in unit hal.gr.init to commit RTV circular buffer
g->ops.gr.init.commit_rtv_cb()
g->ops.gr.init.commit_gfxp_rtv_cb()
Remove tu104 hal to commit global ctx buffers
gr_tu104_commit_global_ctx_buffers() since we have specific hals to
commit RTB circular buffer
Update gr_gk20a_commit_global_ctx_buffers() to directly call
hal.gr.init hals to commit RTV buffers
Jira NVGPU-2961
Change-Id: I12a53386654ebfeb98bf187385bb8b839070d569
Signed-off-by: Deepak Nibade <dnibade@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/2075230
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Add a new hal.gr.init unit hal g->ops.gr.init.get_rtv_cb_size() to
retrieve RTV buffer size
Update gr_gk20a_alloc_global_ctx_buffers() to initialize RTV buffer
size if g->ops.gr.init.get_rtv_cb_size hal is present
Remove gr_tu104_alloc_global_ctx_buffers() since it is no longer
required
Jira NVGPU-2961
Change-Id: I44be8dfdda5c813eac445192635a3a6c2b867b3a
Signed-off-by: Deepak Nibade <dnibade@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/2075229
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Move g->ops.gr.commit_global_timeslice() hal operation to hal.gr.init
unit as g->ops.gr.init.commit_global_timeslice()
Drop channel pointer in parameter list since it was unused
Also change return type to void since it never returns error
Move corresponding gm20b and gv11b hal operations to hal.gr.init unit
Jira NVGPU-2961
Change-Id: I68deef45af1d52149eb354a1478cc2b5f2e4ec2a
Signed-off-by: Deepak Nibade <dnibade@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/2075228
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Add a new hal operation g->ops.gr.init.load_method_init() in hal.gr.init
unit that reads method init netlist bundle and writes those values to
h/w appropriately
Use new hal in gr_gk20a_init_golden_ctx_image() instead of direct
register accesses
Jira NVGPU-2961
Change-Id: If1edd09445e55b5ad9cb1ec7b0f32cab9bfd6f05
Signed-off-by: Deepak Nibade <dnibade@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/2075227
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
nvrm_gpu_tests uses dGPU detach/attach APIs to test
GCOFF sequence. This results in a timeout on tu104 while
detaching dGPU.
...snip...
[start: NvRmGpuTest_Lib_AttachDevice]
Detaching dGPU at index 2001...
[ 89.933604] -- nvrm_gpu_tests: Ended subtest
NvRmGpuTest_Device_ChooseKind_Color_Incompressible
[ 91.473919] nvgpu: 0001:01:00.0 __nvgpu_timeout_expired_msg_retry:118 [ERR]
No more retries @ gk20a_mm_fb_flush+0xd4/0x328 [nvgpu]
[ 91.474280] nvgpu: 0001:01:00.0 gm20b_fb_dump_wpr_info:240 [ERR]
WPR: 08000080 08000081 01c00002 01c01e03 1ffffe04 00000005
...snip...
Increase retries from 2000 to 2500 which seems to be sufficient.
Bug 200491474
Change-Id: I7bab4e39c677361d3642e034f5bdb9ba75d4c450
Signed-off-by: Nitin Kumbhar <nkumbhar@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/2019432
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
