Hi Martin

Perfect.
Now it worked flawlessly.:-))
Thanks.
   Peter


Martin Dickopp wrote:

On Mon, May 02, 2005 at 11:13:13AM +0200, Peter Stelmachovic wrote:


I am sending the output of user.log after applying the patch.



Hi,

Thank you for the quick reply. I have once again a patch for you. :-)
This should fix the issue, so it is hopefully the last patch I have to
ask you to try. Please let me know if it really fixes the issue.



P.S. I have noticed that the same problem occurs if the device is plugged into USB1 port
(different pc with debian stable - no usbmount, no udev, kernel 2.4 but a simple script to automatically mount the flashdisk when plugged in. Again the disk cannot be mounted immediately but the script needs to wait for some time before mounting a partition on the disk).



It appears that the device reports itself ready (i.e. 'cat /sys/block/sdc/device/state' returns 'running') before it really is. My proposed fix therefore tries to read from the device and only continues if this succeeds (or gives up after 20 seconds).

Cheers,
Martin


--- usbmount.orig 2005-05-02 12:02:25.000000000 +0200 +++ usbmount 2005-05-02 12:02:40.000000000 +0200 @@ -62,6 +62,25 @@ trap '( lockfile-remove /var/run/usbmount/.mount )' 0 log debug "acquired lock /var/run/usbmount/.mount.lock"

+ # Try to read from the device. Some devices need a few seconds
+ # initialization time before they can be accessed. Give up after
+ # 20 seconds. Thanks to Peter Stelmachovic for his help with
+ # debugging this.
+ log debug "testing whether $DEVNAME is readable"
+ read_success=no
+ for t in 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19; do
+ if dd if="$DEVNAME" of=/dev/null bs=512 count=1; then
+ read_success=yes
+ break
+ fi
+ log debug "attempt $t to read from $DEVNAME failed"
+ sleep 1
+ done
+ if test "$read_success" != yes; then
+ log err "cannot read from $DEVNAME"
+ exit 1
+ fi
+
# Test if the device contains a filesystem. If it doesn't, no
# further action is required, but calling udev_volume_id has the
# side effect that the partition table is read and partition




--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]



Reply via email to