Question about read only flag

Question about read only flag

[Marc-André Hébert]

Hello,

I have on an OMAPL138 board running 2.6.33-rc4 (I would liked to try a
more recent kernel but I do not have any boards which supports a more
recent version). I noticed that I was unable to detect from sysfs if
the inserted card was set to read only. The block device's
/sys/class/block/mmcblk0/ro entry always reported 0. The read only
gpio is correctly setup as opening the device for writing fails as
expected.

If I compare the behavior with a USB card reader (which of course
doesn't initialize the block device in the same way), the
/sys/class/block/xxx/ro entry is correct.

I am unsure if this is the desired behavior (if so I would really like
to understand why) but here's what I did to correct it.

Regards
Marc

--- a/drivers/mmc/card/block.c     2011-01-31 08:54:56.908952200 -0500
+++ b/drivers/mmc/card/block.c    2011-01-31 08:56:44.175950200 -0500
@@ -621,6 +621,7 @@
        md->disk->private_data = md;
        md->disk->queue = md->queue.queue;
        md->disk->driverfs_dev = &card->dev;
+       set_disk_ro(md->disk, md->read_only);

        /*
         * As discussed on lkml, GENHD_FL_REMOVABLE should:

Re: Question about read only flag

[Chris Ball]

Hi Marc-André,

On Mon, Jan 31, 2011 at 09:01:02AM -0500, Marc-André Hébert wrote:
> Hello,
> 
> I have on an OMAPL138 board running 2.6.33-rc4 (I would liked to try a
> more recent kernel but I do not have any boards which supports a more
> recent version). I noticed that I was unable to detect from sysfs if
> the inserted card was set to read only. The block device's
> /sys/class/block/mmcblk0/ro entry always reported 0. The read only
> gpio is correctly setup as opening the device for writing fails as
> expected.

Interesting, thanks -- no, I don't imagine this is intentional, and I can
confirm the behavior on an x86 laptop/SDHCI running 2.6.38-rc2.  Your
patch looks correct to me.  I'm surprised no-one's noticed this before.

Could you re-send it as a Git patch with a Signed-off-by: line, please?

-- 
Chris Ball   <cjb@laptop.org>   <http://printf.net/>
One Laptop Per Child