-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On 04/12/2011 02:53 PM, Karel Zak wrote: > Not true, > > $ cat /proc/mounts | grep mapper > /dev/mapper/kzak-home /home/kzak ext4 rw,noatime,barrier=1,data=ordered 0 0 I had gotten used to seeing the /dev/dm-N name in /proc/mounts in the past and I may have just assumed it was normal. Now I see that other dm mounts do use the /dev/mapper/ name as you say, but the ntfs-3g mount also shows up as /dev/dm-7 in /proc/mounts. I thought that the kernel used to just find the dev_t in the superblock and print the canonical name of that device as generated from the major/minor number, but it seems it now remembers the name that was passed to mount(), which ntfs-3g must be passing the canonicalized name, and shouldn't be. > It all seems like an extra unnecessary step, if you use everywhere > (including /proc/mounts) user-friendly /dev/mapper/<name> names then > you have to translate from dm-N very rarely. > > Don't forget that mtab and /proc/mounts could be really large > files. Our goal is minimize number of situation where we have to > translate any path. > > Anyway, I'm not sure what issue you're trying to resolve. The > mount(8) and umount(8) have to be able to work with both versions of > the name. The problem in umount(8) was fixed. Problem resolved. You're right; sticking with the friendly name is better, as long as you can still umount with the unfriendly name as well, which it appears you can with your patch. Nice work. -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.10 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/ iEYEARECAAYFAk2lFj4ACgkQJ4UciIs+XuKm2ACfTFTR3VoBsRiE3wpD5nkZvDDY aAAAniqpxpBAZl1Itusan8a2o9kyWHZQ =Zfzs -----END PGP SIGNATURE----- -- To unsubscribe from this list: send the line "unsubscribe util-linux" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
