On Tue, Nov 14, 2006 at 12:13:53AM -0500, A. Costa wrote: > This looks like one of those bugs that are so old and people are so > accustomed to that they've ceased to think of it as bug, it just seems > normal to them. I didn't find a report on it in the BTS, and it seems > there should be one...
And it's a kernel bug... Or not, depending on how you think about it... > 2) So I mount it as 'write': > % mount -w -o loop /tmp/di/I4000A23.iso /mnt/tmp ; echo $? > 0 The kernel returned success on a read/write mount request. It's possible that the kernel should return a failure for the mount in this case. > 4) Copy a file over: > % cp /tmp/BIOS/I4100A06.EXE /mnt/tmp ; echo $? > cp: cannot create regular file `/mnt/tmp/I4100A06.EXE': Read-only file > system > 1 The underlying filesystem in the kernel returns read-only for all writes, since it doesn't support doing them. > Questions: > Why does step '3)' show "rw" if it's not? It is mounted rw - but the underlying file system doesn't support writes. > Suggested easy kludges, nothing low-level: > Parse for '-w' and an ISO9660 mount, and return an error. > Don't show ISO9660 as 'rw'. That's a kludge - not a fix. having the filesystem return an error is a correct solution. lamont -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
