Commit Graph

5 Commits

Author SHA1 Message Date
Nicolas Benech
26d9b79162 gpu: nvgpu: change _suspend_clk_support to void
The HAL _suspend_clk_support was returning int but it
was always 0, so this patch changes it to void which
fixes a number of MISRA 17.7 violations.

JIRA NVGPU-677

Change-Id: Ia107da7ff043147ee00f2580390420914029daeb
Signed-off-by: Nicolas Benech <nbenech@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/2004120
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
2019-01-29 10:54:41 -08:00
Abdul Salam
146d8d3ce5 gpu: nvgpu: Add clk_arb for TU104
Add clk arbiter support for tu104
setup clk_arb for supporting functions in hal_tu04
TU104 supports GPCCLK and not GPC2CLK
Remove multiplication and division by 2 to convert gpcclk to gpc2clk
Provide support for following features
*Domains: Currently GPCCLK is supported
*clk Range: From P0 min to P0 max
*Freq Points: Gives the VF curve from PMU
*Default: Default value(P0 Max)
*Current Pstate: P0 is supported

All request for change is freq is validated against P0 value
Out of bound values are trimmed to match the Pstate limits
Multiple requests are supported and max of that will be set
Requests are sent to PMU via change sequencer

Bug 200454682
JIRA NVGPU-1653

Change-Id: I36735fa50c7963830ebc569a2ea2a2d7aafcf2ab
Signed-off-by: Abdul Salam <absalam@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1982078
Reviewed-by: Vijayakumar Subbu <vsubbu@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
2019-01-08 08:24:38 -08:00
Terje Bergstrom
bca27e31e3 gpu: nvgpu: Fix clk_gp106.h and clk_gv100.h headers
clk_gp106.h and clk_gv100.h define conflicting symbols, which prevent
including them both at the same time. One of the conflicting structs
is namemap_cfg, which has different definitions in clk_gp106.h and
include/nvgpu/clk.h.

Move all constants used only by clk_*.c to be defined there, delete
the extra namemap_cfg structure definition, and modify code to cope
with the unified namemap_cfg.

JIRa NVGPU-596

Change-Id: Id68919da4567ec1507eda0cfaa19bf047a7bfc59
Signed-off-by: Terje Bergstrom <tbergstrom@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1945285
Reviewed-by: svc-misra-checker <svc-misra-checker@nvidia.com>
GVS: Gerrit_Virtual_Submit
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
2018-11-14 13:33:29 -08:00
Debarshi Dutta
6c8be7cfe2 gpu: nvgpu: move header location of gk20a.h
Change path corresponding to gk20a.h to <nvgpu/gk20a.h> corresponding
to files in the following directories.

gk20a/
vgpu/
gv100/
tu104/
common/bus/
common/fb/
common/ltc/
common/mc/
common/perf/

Jira NVGPU-597

Change-Id: I7b4f5e5ea3d13a4d1810c5db35fbc26fe5da443e
Signed-off-by: Debarshi Dutta <ddutta@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1846826
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
2018-10-24 23:16:10 -07:00
absalam
d6424aec6e gpu: nvgpu:Add sysfs node for GV100 clocks
Creates sysfs nodes to read clk freq on GV100
Following sysfs nodes are created: gpcclk,xbarclk,sysclk
Uses default clock source and counters for measurement

Bug 200446261

Change-Id: I6903ba77fbe34e3f486f4b663e70eab4e7c5d662
Signed-off-by: absalam <absalam@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1828030
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
2018-10-12 17:35:06 +05:30