WQ_UNBOUND causes significant scheduler latency on ARM64/Android.  This
is problematic for latency sensitive workloads like I/O post-processing.

Removing WQ_UNBOUND gives a 96% reduction in fsverity workqueue related
scheduler latency and improves app cold startup times by ~30ms.

This code was tested by running Android app startup benchmarks and
measuring how long the fsverity workqueue spent in the ready queue.

Before
Total workqueue scheduler latency: 553800us
After
Total workqueue scheduler latency: 18962us

Change-Id: I693efee541757851ed6d229430111cd763d39067
Signed-off-by: Nathan Huckleberry <nhuck@google.com>
---
 fs/verity/verify.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/fs/verity/verify.c b/fs/verity/verify.c
index f50e3b5b52c9..e8ec37774d63 100644
--- a/fs/verity/verify.c
+++ b/fs/verity/verify.c
@@ -395,7 +395,7 @@ int __init fsverity_init_workqueue(void)
 	 * which blocks reads from completing, over regular application tasks.
 	 */
 	fsverity_read_workqueue = alloc_workqueue("fsverity_read_queue",
-						  WQ_UNBOUND | WQ_HIGHPRI,
+						  WQ_HIGHPRI,
 						  num_online_cpus());
 	if (!fsverity_read_workqueue)
 		return -ENOMEM;
-- 
2.40.0.rc1.284.g88254d51c5-goog

On Thu, Mar 09, 2023 at 01:37:41PM -0800, Nathan Huckleberry wrote:
> WQ_UNBOUND causes significant scheduler latency on ARM64/Android.  This
> is problematic for latency sensitive workloads like I/O post-processing.
> 
> Removing WQ_UNBOUND gives a 96% reduction in fsverity workqueue related
> scheduler latency and improves app cold startup times by ~30ms.

Maybe mention that WQ_UNBOUND was recently removed from the dm-verity workqueue
too, for the same reason?

I'm still amazed that it's such a big improvement!  I don't really need it to
apply this patch, but it would be very interesting to know exactly why the
latency is so bad with WQ_UNBOUND.

> 
> This code was tested by running Android app startup benchmarks and
> measuring how long the fsverity workqueue spent in the ready queue.
> 
> Before
> Total workqueue scheduler latency: 553800us
> After
> Total workqueue scheduler latency: 18962us
> 
> Change-Id: I693efee541757851ed6d229430111cd763d39067

No Change-Id in upstream patches, please.

> diff --git a/fs/verity/verify.c b/fs/verity/verify.c
> index f50e3b5b52c9..e8ec37774d63 100644
> --- a/fs/verity/verify.c
> +++ b/fs/verity/verify.c
> @@ -395,7 +395,7 @@ int __init fsverity_init_workqueue(void)
>  	 * which blocks reads from completing, over regular application tasks.
>  	 */
>  	fsverity_read_workqueue = alloc_workqueue("fsverity_read_queue",
> -						  WQ_UNBOUND | WQ_HIGHPRI,
> +						  WQ_HIGHPRI,
>  						  num_online_cpus());

There's a comment just above here that explains why WQ_UNBOUND is being used.
It needs to be updated to explain why WQ_UNBOUND is *not* being used.

- Eric