The PMA unit can only access GPU VAs within a 4GB window, hence both
the user allocated PMA buffer and the kernel allocated bytes available
buffer should lie in the same 4GB window. This is accomplished by
carving out and reserving a 4GB VA space in perbuf.vm and using fixed
GPU VAs to ensure that both buffers are bound within the same 4GB window.
In addition, update ALLOC_PMA_STREAM to use pma_buffer_offset,
pma_buffer_map_size fields correctly.
Bug 3503708
Change-Id: Ic5297a22c2db42b18ff5e676d565d3be3c1cd780
Signed-off-by: Antony Clince Alex <aalex@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2671637
Reviewed-by: svcacv <svcacv@nvidia.com>
Reviewed-by: Deepak Nibade <dnibade@nvidia.com>
Reviewed-by: svc-mobile-coverity <svc-mobile-coverity@nvidia.com>
Reviewed-by: svc-mobile-cert <svc-mobile-cert@nvidia.com>
Reviewed-by: Vaibhav Kachore <vkachore@nvidia.com>
GVS: Gerrit_Virtual_Submit
Add following HALs for Ga100 and Ga10b. These will
be used for calculating chiplet offsets corresponding
to GPC/FBP perf register.
get_pmmgpcrouter_per_chiplet_offset
get_pmmfbprouter_per_chiplet_offset
get_hwpm_fbp_perfmon_regs_base
get_hwpm_gpc_perfmon_regs_base
get_hwpm_fbprouter_perfmon_regs_base
get_hwpm_gpcrouter_perfmon_regs_base
Bug 200712091
Signed-off-by: Debarshi Dutta <ddutta@nvidia.com>
Change-Id: Iec1a16ef4a3c26dca054c30d95bef991983dc2b7
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2648832
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
nvgpu_timeout_init() returns an error code only when the flags parameter
is invalid. There are very few possible values for flags, so extract the
two most common cases - cpu clock based and a retry based timeout - to
functions that cannot fail and thus return nothing. Adjust all callers
to use those, simplfying error handling quite a bit.
Change-Id: I985fe7fa988ebbae25601d15cf57fd48eda0c677
Signed-off-by: Konsta Hölttä <kholtta@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2613833
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Query the num_perfmon requires golden context to be ready. Accessing
golden context might require gr_instance_id, specific to a GR engine.
On TOT, get_num_hwpm_perfmon() called from perfmon HAL which might
require to call nvgpu_gr_exec_with_err_for_instance().
It internally calls nvgpu_grmgr_config_gr_remap_window() to change
gr_window_remap register points to a current gr_instance_id for MIG.
This approach indirectly mandates to call
nvgpu_gr_exec_with_err_for_instance() which can be
completely avoided. get_num_hwpm_perfmon() is just a query call
which can be moved after the golden context creation.
Using this logic, we can avoid unnecessary invocation of
nvgpu_gr_exec_with_err_for_instance() during perform specific
HAL accesses.
1) Moved get_num_hwpm_perfmon() after golden context creation.
2) Added nvgpu_assert() if (g->num_sys_perfmon == 0U).
JIRA NVGPU-5656
Change-Id: I59a6ab4df93763adbc0765fa5e4d1712b2477521
Signed-off-by: Lakshmanan M <lm@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2542438
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Reviewed-by: Deepak Nibade <dnibade@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Implement below functions:
- nvgpu_profiler_quiesce_hwpm_streamout_resident
Teardown sequence when context is resident or in case profiling
session is a device level session.
- nvgpu_profiler_quiesce_hwpm_streamout_non_resident
Teardown sequence when context is non resident
- nvgpu_profiler_quiesce_hwpm_streamout
Generic sequence to call either of above API based on whether
context is resident or not.
Trigger HWPM streamout teardown sequence while unbinding resources
in nvgpu_profiler_unbind_hwpm_streamout()
Add a new HAL gops.gr.is_tsg_ctx_resident to call
gk20a_is_tsg_ctx_resident() from common code.
Implement below supporting HALs for resident teardown sequence:
- gops.perf.pma_stream_enable()
- gops.perf.disable_all_perfmons()
- gops.perf.wait_for_idle_pmm_routers()
- gops.perf.wait_for_idle_pma()
- gops.gr.disable_cau()
- gops.gr.disable_smpc()
Jira NVGPU-5360
Change-Id: I304ea25d296fae0146937b15228ea21edc091e16
Signed-off-by: Deepak Nibade <dnibade@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2461333
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: Antony Clince Alex <aalex@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
GVS: Gerrit_Virtual_Submit
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
gr_gv100_reset_hwpm_pmm_registers() writes a bunch of registers in
sys/gpc/fbp chiplets to reset perfmons. To ensure all the writes have
completed it is necessary to readback each chiplet's PRI fence register.
Add and use new HAL g->ops.priv_ring.read_pri_fence() to achieve this.
Implement the HAL for gv11b in new source code file
hal/priv_ring/priv_ring_gv11b.c.
Bug 2510974
Jira NVGPU-5360
Change-Id: If4dd61cb4265422e8c2d16884790eb0fe7f2c103
Signed-off-by: Deepak Nibade <dnibade@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2453631
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Below HALs are implemented in common.gr unit, but they really belong
to common.perf unit since they access registers from perf unit.
gops.gr.init_hwpm_pmm_register()
gops.gr.get_num_hwpm_perfmon()
gops.gr.set_pmm_register()
gops.gr.reset_hwpm_pmm_registers()
Move them to common.perf unit, and update all the code accordingly
gops.perf.init_hwpm_pmm_register()
gops.perf.get_num_hwpm_perfmon()
gops.perf.set_pmm_register()
gops.perf.reset_hwpm_pmm_registers()
Add new HAL gops.gr.get_pm_ctx_buffer_offsets() and set it to
gr_gk20a_get_pm_ctx_buffer_offsets() for all chips.
Bug 2510974
Jira NVGPU-5360
Change-Id: Ib5e84ed5c8b6e72cc6923161e55fc2c3a6a4070e
Signed-off-by: Deepak Nibade <dnibade@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2418306
Reviewed-by: automaticguardword <automaticguardword@nvidia.com>
Reviewed-by: svc-mobile-coverity <svc-mobile-coverity@nvidia.com>
Reviewed-by: Alex Waterman <alexw@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
GVS: Gerrit_Virtual_Submit
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Add new HAL g->ops.gr.reset_hwpm_pmm_registers() to reset all HWPM regs
while binding HWPM in global mode in nvgpu_profiler_bind_hwpm()
Add below new HALs to get sys/gpc/fbp register list and count
g->ops.perf.get_hwpm_sys_perfmon_regs()
g->ops.perf.get_hwpm_gpc_perfmon_regs()
g->ops.perf.get_hwpm_fbp_perfmon_regs()
Auto generate all the HWPM regs in below arrays for gv11b/tu104
static const u32 hwpm_sys_perfmon_regs[]
static const u32 hwpm_gpc_perfmon_regs[]
static const u32 hwpm_fbp_perfmon_regs[]
Bug 2510974
Jira NVGPU-5360
Change-Id: I2ca5c04ed75c7b30ae942807bf018a24551d7ba0
Signed-off-by: Deepak Nibade <dnibade@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2414934
Reviewed-by: automaticguardword <automaticguardword@nvidia.com>
Reviewed-by: Antony Clince Alex <aalex@nvidia.com>
Reviewed-by: Alex Waterman <alexw@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
GVS: Gerrit_Virtual_Submit
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Support new IOCTL to manage PMA stream meta data by adding below API
nvgpu_prof_ioctl_pma_stream_update_get_put()
Add nvgpu_perfbuf_update_get_put() to handle all the updates coming
from userspace and to pass all required information.
Add gops.perf.update_get_put() to handle all HW accesses required in
perf HW unit.
Add gops.perf.bind_mem_bytes_buffer_addr() to bind the available bytes
buffer while binding HWPM streamout.
Bug 2510974
Jira NVGPU-5360
Change-Id: Ibacc2299b845e47776babc081759dfc4afde34fe
Signed-off-by: Deepak Nibade <dnibade@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2406484
Reviewed-by: automaticguardword <automaticguardword@nvidia.com>
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: Antony Clince Alex <aalex@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
GVS: Gerrit_Virtual_Submit
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
gr_gv100_init_hwpm_pmm_register() and gr_gv100_set_pmm_register() right
now assume common chiplet stride for all sys/fbp/gpc and use common API
g->ops.perf.get_pmm_per_chiplet_offset() to get the stride.
Chiplet strides are same for all partitions only by chance, and future
chip might change that.
Hence add and use below 3 separate HALs to get appropriate strides.
g->ops.perf.get_pmmsys_per_chiplet_offset()
g->ops.perf.get_pmmgpc_per_chiplet_offset()
g->ops.perf.get_pmmfbp_per_chiplet_offset()
Also store sys/fbp/gpc perfmon count in struct gk20a after first query
instead of querying them again and again. Querying the counts from HW
is time consuming.
Bug 2510974
Jira NVGPU-5360
Change-Id: I186009221009780d561617c0cd6f535854db585f
Signed-off-by: Deepak Nibade <dnibade@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2413108
Reviewed-by: automaticguardword <automaticguardword@nvidia.com>
Reviewed-by: svc-mobile-coverity <svc-mobile-coverity@nvidia.com>
Reviewed-by: svc-mobile-misra <svc-mobile-misra@nvidia.com>
Reviewed-by: Antony Clince Alex <aalex@nvidia.com>
Reviewed-by: Alex Waterman <alexw@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
GVS: Gerrit_Virtual_Submit
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Legacy profiler does not reserve PMA stream resource with PM
reservation system. Also, HWPM system reset is separately implemented
in membuf disable path. And it does not even restore perf unit SLCG
prod values.
Allcoate a dummy profiler object for debug session in perfbuf map
path. Free it in perfbuf unmap path.
This has advantage of synchronizing PMA stream reservation with new
profiler stack. And this also leverages HWPM system reset and SLCG
handling code during resource reservation.
Remove explicit HWPM reset from gops.perf.membuf_reset_streaming()
HALs
Bug 2510974
Jira NVGPU-5360
Change-Id: I54c5202b6251dea3d80a4dfc011e8a296339e07f
Signed-off-by: Deepak Nibade <dnibade@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2399595
Reviewed-by: automaticguardword <automaticguardword@nvidia.com>
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: mobile promotions <svcmobile_promotions@nvidia.com>
GVS: Gerrit_Virtual_Submit
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
gk20a_perfbuf_map() allocates perfbuf VM, maps the user buffer into new
VM, and then triggers gops.perfbuf.perfbuf_enable(). This HAL then does
following :
- Allocate perfbuf instance block
- Initialize perfbuf instance block
- Reset stream buffer
- Program instance block address in PMA registers
- Program user buffer address into PMA registers
New profiler interface will have it's own API to setup PMA strem, and
it requires above setup to be done in two phases of perfbuf
initialization and then user buffer setup.
Split above functionalities into below functions
- nvgpu_perfbuf_init_vm()
- Allocate perfbuf VM
- Call gops.perfbuf.init_inst_block() to initialize perfbuf instance
block
- gops.perfbuf.init_inst_block()
- Allocate perfbuf instance block
- Initialize perfbuf instance block
- Program instance block address in PMA registers using
gops.perf.init_inst_block()
- In case of vGPU, trigger TEGRA_VGPU_CMD_PERFBUF_INST_BLOCK_MGT
command to gpu server
- gops.perf.init_inst_block()
- Reset stream buffer
- Program user buffer address into PMA registers
Also add corresponding cleanup functions as below :
gops.perf.deinit_inst_block()
gops.perfbuf.deinit_inst_block()
nvgpu_perfbuf_deinit_vm()
Bug 2510974
Jira NVGPU-5360
Change-Id: I486370f21012cbb7fea84fe46fb16db95bc16790
Signed-off-by: Deepak Nibade <dnibade@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2372984
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Reviewed-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>
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>
