[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20181108152642.03178c6c@bbrezillon>
Date: Thu, 8 Nov 2018 15:26:42 +0100
From: Boris Brezillon <boris.brezillon@...tlin.com>
To: Richard Weinberger <richard@....at>
Cc: linux-mtd@...ts.infradead.org, juergen.lachmann@...man.com,
linux-kernel@...r.kernel.org, thomas.roeder@...man.com
Subject: Re: [PATCH 2/2] ubi: Expose the bitrot interface
On Wed, 7 Nov 2018 23:16:19 +0100
Richard Weinberger <richard@....at> wrote:
> +/**
> + * ubi_bitflip_check - Check an eraseblock for bitflips and scrub it if needed.
> + * @ubi: UBI device description object
> + * @pnum: the physical eraseblock to schedule
> + * @force_scrub: force scrubbing if non-zero, schedule erase otherwise
Are you sure about the "schedule erase otherwise"? I'd say force_scrub
only influence when the scrub operation is done: either unconditionally
or depending on the result of ubi_io_read().
> + *
> + * This function reads the given eraseblock and checks if bitflips occured.
> + * In case of bitflips, the eraseblock is scheduled for scrubbing.
> + * If scrubbing is forced with @force_scrub, the eraseblock is not read,
> + * but scheduled for scrubbing right away.
> + *
> + * Returns:
> + * %EINVAL, PEB is out of range
> + * %ENOENT, PEB is no longer used by UBI
> + * %EBUSY, PEB cannot be checked now or a check is currently running on it
> + * %EAGAIN, bit flips happened but scrubbing is currently not possible
> + * %EUCLEAN, bit flips happened and PEB is scheduled for scrubbing
> + * %0, no bit flips detected
> + */
> +int ubi_bitflip_check(struct ubi_device *ubi, int pnum, int force_scrub)
Powered by blists - more mailing lists