gpu: nvgpu: use gpfifo_mem via gk20a_mem_{rd,wr}

Use gk20a_mem_*() accessors for gpfifo memory in work submission instead
of direct cpu accesses in order to support other apertures than sysmem.

The gpfifo memory is still allocated from sysmem for dgpus too.

Split the copying of priv_cmds and the main gpfifo to be submitted in
gk20a_submit_channel_gpfifo() into separate functions.

JIRA DNVGPU-21

Change-Id: If271ca8e7e34235f00d31855dbccf77c0008e10b
Signed-off-by: Konsta Holtta <kholtta@nvidia.com>
Reviewed-on: http://git-master/r/1145923
Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com>
Tested-by: Terje Bergstrom <tbergstrom@nvidia.com>

1 files changed, 4 insertions, 0 deletions

diff --git a/drivers/gpu/nvgpu/gk20a/mm_gk20a.h b/drivers/gpu/nvgpu/gk20a/mm_gk20a.h
index f9c5477e..db74a5ca 100644
--- a/drivers/gpu/nvgpu/gk20a/mm_gk20a.h
+++ b/drivers/gpu/nvgpu/gk20a/mm_gk20a.h
@@ -69,6 +69,10 @@ struct gpfifo_desc {
         u32 put;
 
         bool wrap;
+
+        /* if gpfifo lives in vidmem or is forced to go via PRAMIN, first copy
+         * from userspace to pipe and then from pipe to gpu buffer */
+        void *pipe;
 };
 
 struct patch_desc {