The amount of GC I/O on the write buffer is managed by the rate-limiter, which is calculated as a function of the number of available free blocks. When reaching the stable point, we risk having scheduled more I/Os for GC than are allowed on the write buffer. This would result on the GC semaphore balancing the outstanding read GC I/Os to be reported as "hung", though the behavior is normal. Solve this by allowing to schedule when we detect that the read GC path is not moving forward. Signed-off-by: Javier González <javier@xxxxxxxxxxxx> --- drivers/lightnvm/pblk-gc.c | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-) diff --git a/drivers/lightnvm/pblk-gc.c b/drivers/lightnvm/pblk-gc.c index b8323e34d1bc..00d5698d64a9 100644 --- a/drivers/lightnvm/pblk-gc.c +++ b/drivers/lightnvm/pblk-gc.c @@ -218,7 +218,13 @@ static void pblk_gc_line_prepare_ws(struct work_struct *work) gc_rq_ws->line = line; gc_rq_ws->priv = gc_rq; - down(&gc->gc_sem); + /* The write GC path can be much slower than the read GC one due to + * the budget imposed by the rate-limiter. Balance in case that we get + * back pressure from the write GC path. + */ + while (down_timeout(&gc->gc_sem, msecs_to_jiffies(30000))) + io_schedule(); + kref_get(&line->ref); INIT_WORK(&gc_rq_ws->ws, pblk_gc_line_ws); -- 2.7.4
