Commit Graph

5 Commits

Author SHA1 Message Date
seshendra Gadagottu
207e2ac7d1 gpu: nvgpu: gv11b: reading max veid number
To get maximum number of subctx, sw should read
NV_PGRAPH_PRI_FE_CHIP_DEF_INFO_MAX_VEID_COUNT instead of
LITTER_NUM_SUBCTX.

JIRA GV11B-72

Change-Id: I4d675ba49d8a600da77e7b60da449d9e5ba48971
Signed-off-by: seshendra Gadagottu <sgadagottu@nvidia.com>
Reviewed-on: http://git-master/r/1309591
Reviewed-by: Automatic_Commit_Validation_User
Reviewed-by: Seema Khowala <seemaj@nvidia.com>
GVS: Gerrit_Virtual_Submit
Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com>
2017-02-27 10:03:23 -08:00
Seema Khowala
8497f45a2e nvgpu: gpu: gv11b: Remove syncpt protection support
In gv11b sync point support is moved to a shim outside of GPU,
and gv11b does not support sync points anymore. Remove use of
the sync point protection.

JIRA GV11B-47
JIRA GV11B-2

Change-Id: I70f3d2ce0cfe016453efe03f2bbf64c59baeb154
Signed-off-by: Seema Khowala <seemaj@nvidia.com>
Reviewed-on: http://git-master/r/1300964
Reviewed-by: Seshendra Gadagottu <sgadagottu@nvidia.com>
Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com>
2017-02-13 17:54:28 -08:00
seshendra Gadagottu
f04a84b7ce gpu: nvgpu: gv11b: chip specific init_elcg_mode
Added thermal registers for gv11b. Implemented chip specific
init_elcg_mode. In thermal control register, engine power auto
control config is removed and added new field for engine holdoff
enable signal.

JIRA GV11B-58

Change-Id: I412d9a232800d25efbdb0a40f14949d3f085fb0e
Signed-off-by: seshendra Gadagottu <sgadagottu@nvidia.com>
Reviewed-on: http://git-master/r/1300119
Reviewed-by: Automatic_Commit_Validation_User
Reviewed-by: svccoveritychecker <svccoveritychecker@nvidia.com>
GVS: Gerrit_Virtual_Submit
Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com>
2017-02-07 15:16:53 -08:00
seshendra Gadagottu
d00b2000b5 gpu: nvgpu: gv11b: update zcull and pm pointers
Update zcull and perfmon buffer pointers in context header.
For gv11b maximum 49 bits gpu va possible. But,
zcull and perfmon buffer pointers uses maximum 41 bit
va address (258 bytes aligned). To accommodate this, high pointer
registers needs to be updated in context header.

JIRA GV11B-48

Change-Id: Ibe62b6bfedd32c4f3721e4d19d96cce58ef0f366
Signed-off-by: seshendra Gadagottu <sgadagottu@nvidia.com>
Reviewed-on: http://git-master/r/1291852
Reviewed-by: Automatic_Commit_Validation_User
GVS: Gerrit_Virtual_Submit
Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com>
Reviewed-by: svccoveritychecker <svccoveritychecker@nvidia.com>
2017-01-27 13:54:40 -08:00
Alex Waterman
4b09997772 nvgpu: gpu: HW header update for Volta
Similar HW header update as has been done for all the other chips.
HW header files are located under:

  drivers/gpu/nvgpu/include/nvgpu/hw/gv11b/

And can be included like so:

  #include <nvgpu/hw/gv11b/hw_gr_gv11b.h>

Bug 1799159

Change-Id: If39bd71480a34f85bf25f4c36aec0f8f6de4dc9f
Signed-off-by: Alex Waterman <alexw@nvidia.com>
Reviewed-on: http://git-master/r/1284433
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
2017-01-24 15:15:16 -08:00