gpu: nvgpu: use inplace allocation in sync framework

This change is the first of a series of changes to
support the usage of pre-allocated job tracking resources
in the submit path. With this change, we still maintain a
dynamically-allocated joblist, but make the necessary changes
in the channel_sync & fence framework to use in-place
allocations. Specifically, we:

1) Update channel sync framework routines to take in
pre-allocated priv_cmd_entry(s) & gk20a_fence(s) rather
than dynamically allocating themselves

2) Move allocation of priv_cmd_entry(s) & gk20a_fence(s)
to gk20a_submit_prepare_syncs

3) Modify fence framework to have seperate allocation
and init APIs. We expose allocation as a seperate API, so
the client can allocate the object before passing it into
the channel sync framework.

4) Fix clean_up logic in channel sync framework

Bug 1795076

Change-Id: I96db457683cd207fd029c31c45f548f98055e844
Signed-off-by: Sachit Kadle <skadle@nvidia.com>
Reviewed-on: http://git-master/r/1206725
(cherry picked from commit 9d196fd10db6c2f934c2a53b1fc0500eb4626624)
Reviewed-on: http://git-master/r/1223933
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>

1 files changed, 10 insertions, 10 deletions

diff --git a/drivers/gpu/nvgpu/gk20a/channel_sync_gk20a.h b/drivers/gpu/nvgpu/gk20a/channel_sync_gk20a.h
index 4b0918de..c3a92ad2 100644
--- a/drivers/gpu/nvgpu/gk20a/channel_sync_gk20a.h
+++ b/drivers/gpu/nvgpu/gk20a/channel_sync_gk20a.h
@@ -36,8 +36,8 @@ struct gk20a_channel_sync {
          *    cmdbuf is executed.
          */
         int (*wait_syncpt)(struct gk20a_channel_sync *s, u32 id, u32 thresh,
-                           struct priv_cmd_entry **entry,
-                           struct gk20a_fence **fence);
+                           struct priv_cmd_entry *entry,
+                           struct gk20a_fence *fence);
 
         /* Generate a gpu wait cmdbuf from sync fd.
          * Returns
@@ -46,8 +46,8 @@ struct gk20a_channel_sync {
          *    cmdbuf is executed.
          */
         int (*wait_fd)(struct gk20a_channel_sync *s, int fd,
-                       struct priv_cmd_entry **entry,
-                       struct gk20a_fence **fence);
+                       struct priv_cmd_entry *entry,
+                       struct gk20a_fence *fence);
 
         /* Increment syncpoint/semaphore.
          * Returns
@@ -55,8 +55,8 @@ struct gk20a_channel_sync {
          *  - a fence that can be passed to wait_cpu() and is_expired().
          */
         int (*incr)(struct gk20a_channel_sync *s,
-                    struct priv_cmd_entry **entry,
-                    struct gk20a_fence **fence,
+                    struct priv_cmd_entry *entry,
+                    struct gk20a_fence *fence,
                     bool need_sync_fence);
 
         /* Increment syncpoint/semaphore, preceded by a wfi.
@@ -65,8 +65,8 @@ struct gk20a_channel_sync {
          *  - a fence that can be passed to wait_cpu() and is_expired().
          */
         int (*incr_wfi)(struct gk20a_channel_sync *s,
-                        struct priv_cmd_entry **entry,
-                        struct gk20a_fence **fence);
+                        struct priv_cmd_entry *entry,
+                        struct gk20a_fence *fence);
 
         /* Increment syncpoint/semaphore, so that the returned fence represents
          * work completion (may need wfi) and can be returned to user space.
@@ -77,8 +77,8 @@ struct gk20a_channel_sync {
          */
         int (*incr_user)(struct gk20a_channel_sync *s,
                          int wait_fence_fd,
-                         struct priv_cmd_entry **entry,
-                         struct gk20a_fence **fence,
+                         struct priv_cmd_entry *entry,
+                         struct gk20a_fence *fence,
                          bool wfi,
                          bool need_sync_fence);