Received-SPF: pass (google.com: domain of linux-nfs+bounces-3682-linux.lists.archive=gmail.com@vger.kernel.org designates 147.75.199.223 as permitted sender) client-ip=147.75.199.223;
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: 7bit
Precedence: bulk
MIME-Version: 1.0
From: "NeilBrown" <neilb@suse.de>
To: "Mike Snitzer" <snitzer@kernel.org>
Cc: linux-nfs@vger.kernel.org, "Jeff Layton" <jlayton@kernel.org>,
 "Chuck Lever" <chuck.lever@oracle.com>,
 "Trond Myklebust" <trondmy@hammerspace.com>, snitzer@hammerspace.com
Subject: Re: [RFC PATCH v2 08/15] NFS: for localio don't call filesystem
 read() and write() routines directly
In-reply-to: <20240612030752.31754-9-snitzer@kernel.org>
References: <20240612030752.31754-1-snitzer@kernel.org>,
 <20240612030752.31754-9-snitzer@kernel.org>
Date: Wed, 12 Jun 2024 14:39:15 +1000
Message-id: <171816715571.14261.10325425432606108026@noble.neil.brown.name>

On Wed, 12 Jun 2024, Mike Snitzer wrote:
> From: Trond Myklebust <trond.myklebust@hammerspace.com>
> 
> Some filesystem writeback routines can end up taking up a lot of stack
> space (particularly xfs). Instead of risking running over due to the
> extra overhead from the NFS stack, we should just call these routines
> from a workqueue job.
> 

Do we really need this today?  Kernel stack is a lot bigger than it once
was.

If we do need it, do we need a dedicated workqueue?  Why not just use
system_unbound_wq ??

NeilBrown