Re: [PATCH v2 3/5] io_uring/rsrc: allow buffer release callback to be optional

[Joanne Koong <joannelkoong@gmail.com>]

On Tue, Mar 24, 2026 at 11:22 AM Joanne Koong <joannelkoong@gmail.com> wrote:
>
> This is a preparatory patch for supporting kernel-populated buffers in
> fuse io-uring, which does not need a release callback.
>
> Signed-off-by: Joanne Koong <joannelkoong@gmail.com>
> Reviewed-by: Caleb Sander Mateos <csander@purestorage.com>
> ---
>  io_uring/rsrc.c | 3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)
>
> diff --git a/io_uring/rsrc.c b/io_uring/rsrc.c
> index 01c3619e5f07..73ee03f85509 100644
> --- a/io_uring/rsrc.c
> +++ b/io_uring/rsrc.c
> @@ -133,7 +133,8 @@ static void io_buffer_unmap(struct io_ring_ctx *ctx, struct io_mapped_ubuf *imu)
>
>         if (imu->acct_pages)
>                 io_unaccount_mem(ctx->user, ctx->mm_account, imu->acct_pages);
> -       imu->release(imu->priv);
> +       if (imu->release)
> +               imu->release(imu->priv);
>         io_free_imu(ctx, imu);
>  }

I'm going to drop this patch. While going over the fuse side of
things, I realized there's the possibiltiy of spontaneous unmounts
(which even with privileged servers can happen while non-fuse io-uring
requests are inflight). fuse will need to grab and release refcounts
on the underlying pages.

I'll submit a new v3 with this change (and with an added patch that
exports IO_IMU_DEST/IO_IMU_SOURCE), sorry for the noise. The fuse
changes that depend on this series will be submitted shortly
thereafter.

Thanks,
Joanne



>
> --
> 2.52.0
>