gpu: nvgpu: check ch_timedout for poll/restart

poll_timeouts and timeout_restart_all_channels should
only handle channels that have not been recovered/aborted.
Check ch_timedout status of the channel to make sure
channel is still alive to be used. A channel reference
could still be available even if it is recovered but not
closed.

Bug 2404865

Change-Id: I016c8b9952ef1d4c349c2a2a2ca55cb81326d380
Signed-off-by: Seema Khowala <seemaj@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1929339
Signed-off-by: Debarshi Dutta <ddutta@nvidia.com>
(cherry picked from commit def687d4dfbc92dbc8a13781f8cee606eab4f3f3
in rel-32)
Reviewed-on: https://git-master.nvidia.com/r/2016995
GVS: Gerrit_Virtual_Submit
Reviewed-by: Alex Waterman <alexw@nvidia.com>
Reviewed-by: Bibek Basu <bbasu@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>

1 files changed, 23 insertions, 10 deletions

diff --git a/drivers/gpu/nvgpu/common/fifo/channel.c b/drivers/gpu/nvgpu/common/fifo/channel.c
index 2ada3911..0174e369 100644
--- a/drivers/gpu/nvgpu/common/fifo/channel.c
+++ b/drivers/gpu/nvgpu/common/fifo/channel.c
@@ -1387,6 +1387,11 @@ u32 nvgpu_get_gp_free_count(struct channel_gk20a *c)
 
 static void __gk20a_channel_timeout_start(struct channel_gk20a *ch)
 {
+        if (gk20a_channel_check_timedout(ch)) {
+                ch->timeout.running = false;
+                return;
+        }
+
         ch->timeout.gp_get = ch->g->ops.fifo.userd_gp_get(ch->g, ch);
         ch->timeout.pb_get = ch->g->ops.fifo.userd_pb_get(ch->g, ch);
         ch->timeout.running = true;
@@ -1484,16 +1489,16 @@ void gk20a_channel_timeout_restart_all_channels(struct gk20a *g)
         for (chid = 0; chid < f->num_channels; chid++) {
                 struct channel_gk20a *ch = gk20a_channel_from_id(g, chid);
 
-                if (ch == NULL)
-                        continue;
-
-                nvgpu_raw_spinlock_acquire(&ch->timeout.lock);
-                if (ch->timeout.running) {
-                        __gk20a_channel_timeout_start(ch);
+                if (ch != NULL) {
+                        if (!gk20a_channel_check_timedout(ch)) {
+                                nvgpu_raw_spinlock_acquire(&ch->timeout.lock);
+                                if (ch->timeout.running) {
+                                        __gk20a_channel_timeout_start(ch);
+                                }
+                                nvgpu_raw_spinlock_release(&ch->timeout.lock);
+                        }
+                        gk20a_channel_put(ch);
                 }
-                nvgpu_raw_spinlock_release(&ch->timeout.lock);
-
-                gk20a_channel_put(ch);
         }
 }
 
@@ -1517,6 +1522,12 @@ static void gk20a_channel_timeout_handler(struct channel_gk20a *ch)
 
         nvgpu_log_fn(g, " ");
 
+        if (gk20a_channel_check_timedout(ch)) {
+                /* channel is already recovered */
+                gk20a_channel_timeout_stop(ch);
+                return;
+        }
+
         /* Get status but keep timer running */
         nvgpu_raw_spinlock_acquire(&ch->timeout.lock);
         gp_get = ch->timeout.gp_get;
@@ -1587,7 +1598,9 @@ static void gk20a_channel_poll_timeouts(struct gk20a *g)
                 struct channel_gk20a *ch = gk20a_channel_from_id(g, chid);
 
                 if (ch != NULL) {
-                        gk20a_channel_timeout_check(ch);
+                        if (!gk20a_channel_check_timedout(ch)) {
+                                gk20a_channel_timeout_check(ch);
+                        }
                         gk20a_channel_put(ch);
                 }
         }