>Recent discussion triggered due to a patch linked below, from Qiang, >shed light on the need to accelerate from QS reporting paths. > >Update the comments to capture this piece of knowledge. > >Link: https://lore.kernel.org/all/20230118073014.2020743-1-qiang1.zhang@xxxxxxxxx/ >Cc: Qiang Zhang <Qiang1.zhang@xxxxxxxxx> >Cc: Frederic Weisbecker <frederic@xxxxxxxxxx> >Signed-off-by: Joel Fernandes (Google) <joel@xxxxxxxxxxxxxxxxx> > >--- > kernel/rcu/tree.c | 13 ++++++++++++- > 1 file changed, 12 insertions(+), 1 deletion(-) > >diff --git a/kernel/rcu/tree.c b/kernel/rcu/tree.c >index 93eb03f8ed99..713eb6ca6902 100644 >--- a/kernel/rcu/tree.c >+++ b/kernel/rcu/tree.c >@@ -1983,7 +1983,12 @@ rcu_report_qs_rdp(struct rcu_data *rdp) > } else { > /* > * This GP can't end until cpu checks in, so all of our >- * callbacks can be processed during the next GP. >+ * callbacks can be processed during the next GP. Do >+ * the acceleration from here otherwise there may be extra >+ * grace period delays, as any accelerations from rcu_core() Does the extra grace period delays means that if not accelerate callback, the grace period will take more time to end ? or refers to a delay in the start time of a new grace period? Thanks Zqiang >+ * or note_gp_changes() may happen only after the GP after the >+ * current one has already started. Further, rcu_core() >+ * only accelerates if RCU is idle (no GP in progress). > * > * NOCB kthreads have their own way to deal with that... > */ >@@ -1993,6 +1998,12 @@ rcu_report_qs_rdp(struct rcu_data *rdp) > /* > * ...but NOCB kthreads may miss or delay callbacks acceleration > * if in the middle of a (de-)offloading process. >+ * >+ * Such missed acceleration may cause the callbacks to >+ * be stranded until RCU is fully de-offloaded, as >+ * acceleration from rcu_core() and note_gp_changes() >+ * cannot happen for fully/partially offloaded mode due >+ * to ordering dependency between rnp lock and nocb_lock. > */ > needacc = true; > } >-- >2.39.1.519.gcb327c4b5f-goog >
