gpu: nvgpu: abstract out timeout rewinding

The channel timeout ends up in a strange state during timeout handling
for a brief moment; it can become stopped and started again, and the
timeout lock is released in the middle. Add a more explicit rewind
function to reset the timeout to start if it's active. The active check
allows to use this from gk20a_channel_timeout_restart_all_channels(), so
that's also modified.

Also replace the return statements with more readable control flow in
gk20a_channel_timeout_handler().

Bug 200484795

Change-Id: Ia7d67242dfc149ace1f4f841a837e90b6c985308
Signed-off-by: Konsta Holtta <kholtta@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1989327
Reviewed-by: Deepak Nibade <dnibade@nvidia.com>
Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com>
(cherry picked from commit 8979a97af3dbb65904f9db8a22a6a168a3f41447
in dev-kernel)
Reviewed-on: https://git-master.nvidia.com/r/2017922
Reviewed-by: Debarshi Dutta <ddutta@nvidia.com>
Tested-by: Debarshi Dutta <ddutta@nvidia.com>
Reviewed-by: Bibek Basu <bbasu@nvidia.com>
Reviewed-by: svc-mobile-coverity <svc-mobile-coverity@nvidia.com>
GVS: Gerrit_Virtual_Submit
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>

1 files changed, 33 insertions, 23 deletions

diff --git a/drivers/gpu/nvgpu/common/fifo/channel.c b/drivers/gpu/nvgpu/common/fifo/channel.c
index 0174e369..fc82748b 100644
--- a/drivers/gpu/nvgpu/common/fifo/channel.c
+++ b/drivers/gpu/nvgpu/common/fifo/channel.c
@@ -1474,6 +1474,25 @@ static void gk20a_channel_timeout_continue(struct channel_gk20a *ch)
 }
 
 /**
+ * Reset the counter of a timeout that is in effect.
+ *
+ * If this channel has an active timeout, act as if something happened on the
+ * channel right now.
+ *
+ * Rewinding a stopped counter is irrelevant; this is a no-op for non-running
+ * timeouts. Stopped timeouts can only be started (which is technically a
+ * rewind too) or continued (where the stop is actually pause).
+ */
+static void gk20a_channel_timeout_rewind(struct channel_gk20a *ch)
+{
+        nvgpu_raw_spinlock_acquire(&ch->timeout.lock);
+        if (ch->timeout.running) {
+                __gk20a_channel_timeout_start(ch);
+        }
+        nvgpu_raw_spinlock_release(&ch->timeout.lock);
+}
+
+/**
  * Rewind the timeout on each non-dormant channel.
  *
  * Reschedule the timeout of each active channel for which timeouts are running
@@ -1491,11 +1510,7 @@ void gk20a_channel_timeout_restart_all_channels(struct gk20a *g)
 
                 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_timeout_rewind(ch);
                         }
                         gk20a_channel_put(ch);
                 }
@@ -1538,28 +1553,23 @@ static void gk20a_channel_timeout_handler(struct channel_gk20a *ch)
         new_pb_get = g->ops.fifo.userd_pb_get(ch->g, ch);
 
         if (new_gp_get != gp_get || new_pb_get != pb_get) {
-                /* Channel has advanced, rewind timer */
-                gk20a_channel_timeout_stop(ch);
-                gk20a_channel_timeout_start(ch);
-                return;
-        }
-
-        if (nvgpu_timeout_peek_expired(&ch->timeout.timer) == 0) {
+                /* Channel has advanced, timer keeps going but resets */
+                gk20a_channel_timeout_rewind(ch);
+        } else if (nvgpu_timeout_peek_expired(&ch->timeout.timer) == 0) {
                 /* Seems stuck but waiting to time out */
-                return;
-        }
+        } else {
+                nvgpu_err(g, "Job on channel %d timed out",
+                          ch->chid);
 
-        nvgpu_err(g, "Job on channel %d timed out",
-                  ch->chid);
+                /* force reset calls gk20a_debug_dump but not this */
+                if (ch->timeout.debug_dump) {
+                        gk20a_gr_debug_dump(g);
+                }
 
-        /* force reset calls gk20a_debug_dump but not this */
-        if (ch->timeout.debug_dump) {
-                gk20a_gr_debug_dump(g);
+                g->ops.fifo.force_reset_ch(ch,
+                        NVGPU_ERR_NOTIFIER_FIFO_ERROR_IDLE_TIMEOUT,
+                        ch->timeout.debug_dump);
         }
-
-        g->ops.fifo.force_reset_ch(ch,
-                NVGPU_ERR_NOTIFIER_FIFO_ERROR_IDLE_TIMEOUT,
-                ch->timeout.debug_dump);
 }
 
 /**