[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <aTGbPIsEO94cFHZr@redhat.com>
Date: Thu, 4 Dec 2025 15:31:24 +0100
From: Oleg Nesterov <oleg@...hat.com>
To: Mateusz Guzik <mjguzik@...il.com>
Cc: brauner@...nel.org, linux-kernel@...r.kernel.org,
akpm@...ux-foundation.org, linux-mm@...ck.org, willy@...radead.org
Subject: Re: [PATCH v2 2/2] pid: only take pidmap_lock once on alloc
On 12/03, Mateusz Guzik wrote:
>
> It has an unintentional side effect of no longer issuing spurious
> idr_preload() around idr_replace().
^^^^^^^^^^^
Around pidfs_add_pid(). The extra idr_preload/end was added by commit
9698d5a483654 ("pidfs: rework inode number allocation") because
pidfs_add_pid() used to call idr_alloc_cyclic(GFP_ATOMIC) too.
The patch looks technically correct to me, so FWIW
Reviewed-by: Oleg Nesterov <oleg@...hat.com>
Powered by blists - more mailing lists