Background:
In Hypervisor mode dGPU device is configured in pass through mode for
the Guest (QNX/Linux). GMMU programming is handled by the guest which
converts a mapped buffer's GVA into SGLes in IPA (Intermediate/Guest
Physical address) which is then translated into PA (Acutual Physical
address) and programs the GMMU PTEes with correct GVA to PA mapping.
Incase of the vgpu this work is delegated to the RM server which takes care
of the GMMU programming and IPA to PA conversion.
Problem:
The current GMMU mapping logic in the guest assumes that PA range is
continuous over a given IPA range. Hence, it doesn't account for holes being
present in the PA range. But this is not the case, a continous IPA range
can be mapped to dis-contiguous PA ranges. In this situation the mapping
logic sets up GMMU PTEes ignoring the holes in physical memory and
creates GVA => PA mapping which intrudes into the PA ranges which are
reserved. This results in memory being corrupted.
This change takes into account holes being present in a given PA range and
for a given IPA range it also identifies the discontiguous PA ranges and
sets up the PTE's appropriately.
Bug 200451447
Jira VQRM-5069
Change-Id: I354d984f6c44482e4576a173fce1e90ab52283ac
Signed-off-by: aalex <aalex@nvidia.com>
Signed-off-by: Antony Clince Alex <aalex@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1850972
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
The power rail of dGPU is managed with help of a set of
GPIOs. Using those GPIOs add an interface to power off and
power on dGPU.
Before dGPU is powered off, new work is blocked by setting
NVGPU_DRIVER_IS_DYING and current jobs are allowed to finish
by waiting for gpu to be idle.
The tegra PCIe controller driver provided APIs
tegra_pcie_attach_controller() and tegra_pcie_detach_controller()
are used to manage PCIe link shutdown, PCIe refclk management
and PCIe rescan.
JIRA NVGPU-1100
Change-Id: Ifae5b81535f40dceca5292a987d3daf6984f3210
Signed-off-by: Nitin Kumbhar <nkumbhar@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1749847
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
- TPC powergating should be done before
calling gk20a_enable_gr_hw.
gk20a_enable_gr_hw() issues a GR engine reset.
Without this fix, enabling 1 TPC from each PES
causes ctxsw timeout error while running GFX Benchmark.
- Adds valid tpc-pg mask for 1/2/3/4 active TPC configs.
TPC Config - TPC-MASK
4 TPC configuration - 0x0
3 TPC configuration - 0x1/0x2/0x4/0x8
2 TPC configuration - 0x5/0x9/0x6/0xa
- We should not write to gr_fe_tpc_pesmask_r()
as part of TPC-PG sequence. This register is for
debug purpose only.
Bug 200442360
Change-Id: I6fbe1ad8fbc836ace8cbaf00ec3d21a12c73e0bd
Signed-off-by: Deepak Goyal <dgoyal@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1809772
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
1) Update header path of gk20a.h files present in os/
to <nvgpu/gk20a.h>
2) os_fence_android_sema.c indirectly was dependent on gk20a.h via
semaphore.h. So, added #include <nvgpu/gk20a.h> in
os_fence_android_sema.c and replaced the header with forward
declaration of struct gk20a in semaphore.h
Jira NVGPU-597
Change-Id: I96e23befeb80713f3a399071eb5498f6f580211d
Signed-off-by: Debarshi Dutta <ddutta@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1842868
Reviewed-by: svc-misra-checker <svc-misra-checker@nvidia.com>
GVS: Gerrit_Virtual_Submit
Reviewed-by: Konsta Holtta <kholtta@nvidia.com>
Reviewed-by: Automatic_Commit_Validation_User
Reviewed-by: Alex Waterman <alexw@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Expose the linux specific clock implementations via the HAL
interface to allow nvgpu to use the controls globally. This patch
does the following.
1) Implement a new ops interface and a corresponding linux specific
implementation for allowing nvgpu to iterate through a list of
available clock frequencies via nvgpu_linux_clk_get_f_points().
2) Implement nvgpu_linux_clk_get_range().
Bug 2061372
Change-Id: I7ce9a999dbdcd9fafcc84301af148545f6ca97a9
Signed-off-by: Debarshi Dutta <ddutta@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1774280
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Adding a comptag_mem_deduct in the platform_gk20a has certain problems:
1) It's not really convenient for platform users to configure it.
2) All products using the same GPU have to share the same configuration.
So this patch moves this comptag_mem_deduct from struct platform_gk20a
to struct gr_gk20a (per device). And it adds an sysfs node for products
or platform users to easily configure from user space.
Note: The comptag memory will not be allocated until the GPU driver goes
through the final poweron routine. So the user space has a small window
to configure this sysfs node.
Bug 2327574
Bug 2284925
Change-Id: Ie7d00b082704e422645c0ea254b59e22f9fc3b7f
Signed-off-by: Nicolin Chen <nicolinc@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1810334
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
The comptag allocates memory based on the available total RAM, which
theoretically should be the MAX physical RAM size however practically
should deduct the part being taken by the running system. Otherwise,
the taken memory part will never get used and wasted.
This change adds a comptag_mem_deduct to each platform and to allow
them to assign the deductible value based on their own use cases so
as to save memory.
Bug 2327574
Bug 2284925
Change-Id: I124e20a66183c22723c34a7ec6ce34832c12f02e
Signed-off-by: Nicolin Chen <nicolinc@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1804157
Reviewed-by: svc-misra-checker <svc-misra-checker@nvidia.com>
GVS: Gerrit_Virtual_Submit
Reviewed-by: Sachin Nikam <snikam@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
Add a WAR for gm20b that allows us to force the PMU VM to use
128K large pages. For some reason setting the small page size
to 64K breaks the PMU boot. Unclear why. Bug needs to be filed
and fixed. Once fixed this patch can and should be reverted.
Bug 200105199
Change-Id: I2b4c9e214e2a6dff33bea18bd2359c33364ba03f
Signed-off-by: Alex Waterman <alexw@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1782769
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
- adds static tpc-powergating through sysfs.
- active tpc count will remain till the GPU/systems is not booted again.
- tpc_pg_mask can be written only after GPU probe finishes and
GPU boot is triggered.
Note:
To be able to use this feature, we need to change boot/init
scripts of the OS(used with nvgpu driver) to write to sysfs nodes before
posting discover image size query to FECS.
Bug 200406784
Change-Id: Id749c7a617422c625f77d0c1a9aada2eb960c4d0
Signed-off-by: Deepak Goyal <dgoyal@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1742422
Reviewed-by: svc-misra-checker <svc-misra-checker@nvidia.com>
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>
