diff options
| author | Puneet Saxena <puneets@nvidia.com> | 2018-07-16 02:06:23 -0400 |
|---|---|---|
| committer | mobile promotions <svcmobile_promotions@nvidia.com> | 2018-07-18 08:53:59 -0400 |
| commit | e04e2233145933ebbd68278cf930d04c403863ba (patch) | |
| tree | f4e971aad2252188fa8cc2da09685dde11147d04 /include | |
| parent | 54702fe6b41a0124b9204b1a4ec2e181266756c6 (diff) | |
video: tegra: nvmap: alloc 6, 4K pages for 6 cv devs
gos mem requirement is to allocate 6, 4K pages from
kernel common heap.
In K4.9, dma_alloc_coherent has legacy support of
allocating from iomem heap even if "device pointer"
is passed as NULL. Gos mem code was passing device pointer
NULL to allocate 6, 4K pages.
On K4.14, the support is removed. dma_alloc_coherent()
will succeed only if passed, not NULL device pointer.
As there is no requirement of creating gosmem device and
each 4K pages for CV devices may be discontiguous,
using Vmalloc to allocate 6, 4K pages.
Bug 200430056
Change-Id: Idc43b1ee3565fe1d66470035a2a68a681e81abce
Signed-off-by: Puneet Saxena <puneets@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1778946
Reviewed-by: svc-mobile-coverity <svc-mobile-coverity@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>
Diffstat (limited to 'include')
0 files changed, 0 insertions, 0 deletions