Re: [PATCH] poweroff: fix bug in orderly_poweroff

[Eric W. Biederman]

hongfeng <hongfeng@xxxxxxxxxxx> writes:

> orderly_poweroff is trying to poweroff platform by two steps:
> step 1: Call userspace application to poweroff
> step 2: If userspace poweroff fail, then do a force power off if force param is set.
>
> The bug here is, step 1 is always successful with param UMH_NO_WAIT,

This code has existed for 5 years.  Is this a recent regression?  Why
has no one complained before?

It looks to me that step 2 is:
step 2: If we can not launch the userspace poweroff fail.

> should change to UMH_WAIT_PROC which will monitor the return value
> ofuserspace application.

Is it safe to block indefinitely in the callers waiting for userspace?

If the caller is not running in a kernel thread then we can easily get
into a case where the userspace caller will block waiting for us when we
are waiting for the userspace caller.

I don't want to impeded progress but I don't see the evidence that this
change is good enough.


> Change-Id: I2f9ebbb90c0c2443780080ec9507c8d004e5da74
> Signed-off-by: Feng Hong <hongfeng@xxxxxxxxxxx>
> ---
>  kernel/sys.c |    2 +-
>  1 files changed, 1 insertions(+), 1 deletions(-)
>
> diff --git a/kernel/sys.c b/kernel/sys.c
> index 241507f..1b30b30 100644
> --- a/kernel/sys.c
> +++ b/kernel/sys.c
> @@ -2204,7 +2204,7 @@ static int __orderly_poweroff(void)
>  		return -ENOMEM;
>  	}
>  
> -	ret = call_usermodehelper_fns(argv[0], argv, envp, UMH_NO_WAIT,
> +	ret = call_usermodehelper_fns(argv[0], argv, envp, UMH_WAIT_PROC,
>  				      NULL, argv_cleanup, NULL);
>  	if (ret == -ENOMEM)
>  		argv_free(argv);
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/