On Mon, 2020-06-08 at 12:10 +0200, Roberto Bergantinos Corpas wrote:
> This can be useful for troubleshooting purposes and this
> information
> is not directly avalaible to userland
>
> Signed-off-by: Roberto Bergantinos Corpas <rbergant@xxxxxxxxxx>
> ---
> fs/autofs/waitq.c | 5 +++--
> 1 file changed, 3 insertions(+), 2 deletions(-)
>
> diff --git a/fs/autofs/waitq.c b/fs/autofs/waitq.c
> index b04c528b19d3..1a151a3b7d8b 100644
> --- a/fs/autofs/waitq.c
> +++ b/fs/autofs/waitq.c
> @@ -88,9 +88,10 @@ static void autofs_notify_daemon(struct
> autofs_sb_info *sbi,
> size_t pktsz;
> int ret;
>
> - pr_debug("wait id = 0x%08lx, name = %.*s, type=%d\n",
> + pr_debug("wait id = 0x%08lx, name = %.*s, type=%d, requested by
> (%d/%s)\n",
> (unsigned long) wq->wait_queue_token,
> - wq->name.len, wq->name.name, type);
> + wq->name.len, wq->name.name, type,
> + wq->pid, current->comm);
That wq->pid is a bit of a problem.
It could be a pid in a different pid name space.
It's a problem for the daemon too since it will log the pid from the
request packet it gets. The pid value used in the request packet is
already wrong if the request came from a container, not sure what
to do about that.
What will current->comm contain, just the process name or full path?
That would be useful and the daemon doesn't get this from the request.
Ian
