Re: [PATCH 08/13] btrfs: Use alloc_ordered_workqueue() to create ordered workqueues

[David Sterba]

On Mon, May 08, 2023 at 03:50:27PM -1000, Tejun Heo wrote:
> BACKGROUND
> ==========
> 
> When multiple work items are queued to a workqueue, their execution order
> doesn't match the queueing order. They may get executed in any order and
> simultaneously. When fully serialized execution - one by one in the queueing
> order - is needed, an ordered workqueue should be used which can be created
> with alloc_ordered_workqueue().
> 
> However, alloc_ordered_workqueue() was a later addition. Before it, an
> ordered workqueue could be obtained by creating an UNBOUND workqueue with
> @max_active==1. This originally was an implementation side-effect which was
> broken by 4c16bd327c74 ("workqueue: restore WQ_UNBOUND/max_active==1 to be
> ordered"). Because there were users that depended on the ordered execution,
> 5c0338c68706 ("workqueue: restore WQ_UNBOUND/max_active==1 to be ordered")
> made workqueue allocation path to implicitly promote UNBOUND workqueues w/
> @max_active==1 to ordered workqueues.
> 
> While this has worked okay, overloading the UNBOUND allocation interface
> this way creates other issues. It's difficult to tell whether a given
> workqueue actually needs to be ordered and users that legitimately want a
> min concurrency level wq unexpectedly gets an ordered one instead. With
> planned UNBOUND workqueue updates to improve execution locality and more
> prevalence of chiplet designs which can benefit from such improvements, this
> isn't a state we wanna be in forever.
> 
> This patch series audits all callsites that create an UNBOUND workqueue w/
> @max_active==1 and converts them to alloc_ordered_workqueue() as necessary.
> 
> WHAT TO LOOK FOR
> ================
> 
> The conversions are from
> 
>   alloc_workqueue(WQ_UNBOUND | flags, 1, args..)
> 
> to
> 
>   alloc_ordered_workqueue(flags, args...)
> 
> which don't cause any functional changes. If you know that fully ordered
> execution is not ncessary, please let me know. I'll drop the conversion and
> instead add a comment noting the fact to reduce confusion while conversion
> is in progress.
> 
> If you aren't fully sure, it's completely fine to let the conversion
> through. The behavior will stay exactly the same and we can always
> reconsider later.
> 
> As there are follow-up workqueue core changes, I'd really appreciate if the
> patch can be routed through the workqueue tree w/ your acks. Thanks.
> 
> v3: btrfs_alloc_workqueue() excluded again. The concurrency level of a
>     workqueue allocated through btrfs_alloc_workqueue() may be dynamically
>     adjusted through thresh_exec_hook(), so they can't be ordered.
> 
> v2: btrfs_alloc_workqueue() updated too as suggested by Wang.
> 
> Signed-off-by: Tejun Heo <tj@xxxxxxxxxx>
> Cc: Wang Yugui <wangyugui@xxxxxxxxxxxx>
> Cc: Chris Mason <clm@xxxxxx>
> Cc: Josef Bacik <josef@xxxxxxxxxxxxxx>
> Cc: David Sterba <dsterba@xxxxxxxx>
> Cc: linux-btrfs@xxxxxxxxxxxxxxx
> ---
>  fs/btrfs/disk-io.c | 2 +-
>  fs/btrfs/scrub.c   | 6 ++++--
>  2 files changed, 5 insertions(+), 3 deletions(-)
> 
> diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c
> index 59ea049fe7ee..32d08aed88b6 100644
> --- a/fs/btrfs/disk-io.c
> +++ b/fs/btrfs/disk-io.c
> @@ -2217,7 +2217,7 @@ static int btrfs_init_workqueues(struct btrfs_fs_info *fs_info)
>  	fs_info->qgroup_rescan_workers =
>  		btrfs_alloc_workqueue(fs_info, "qgroup-rescan", flags, 1, 0);
>  	fs_info->discard_ctl.discard_workers =
> -		alloc_workqueue("btrfs_discard", WQ_UNBOUND | WQ_FREEZABLE, 1);
> +		alloc_ordered_workqueue("btrfs_discard", WQ_FREEZABLE);
>  
>  	if (!(fs_info->workers && fs_info->hipri_workers &&
>  	      fs_info->delalloc_workers && fs_info->flush_workers &&

I think there are a few more conversions missing. There's a local flags
variable in btrfs_init_workqueues

2175 static int btrfs_init_workqueues(struct btrfs_fs_info *fs_info)
2176 {
2177         u32 max_active = fs_info->thread_pool_size;
2178         unsigned int flags = WQ_MEM_RECLAIM | WQ_FREEZABLE | WQ_UNBOUND;

And used like

2194         fs_info->fixup_workers =
2195                 btrfs_alloc_workqueue(fs_info, "fixup", flags, 1, 0);

2213         fs_info->qgroup_rescan_workers =
2214                 btrfs_alloc_workqueue(fs_info, "qgroup-rescan", flags, 1, 0);

WQ_UNBOUND is not mentioned explicitliy like for the "btrfs_discard"
workqueue.  Patch v2 did the switch in btrfs_alloc_workqueue according
to the max_active/limit_active parameter but this would affect all
queues and not all of them require to be ordered.

In btrfs_resize_thread_pool the workqueue_set_max_active is called
directly or indirectly so this can set the max_active to a user-defined
mount option. Could this be a problem or trigger a warning? This would
lead to max_active==1 + WQ_UNBOUND.