Re: oh, btw...

1999-12-06 Thread Matthew Dillon


: How do we fix this problem?  I keeps from from
: ``dd if=/dev/da1 of=/dev/da2'' ??  I was very peeved at having to put the
: disks on a Solaris box to do such a normal Unix task.
:
:
:this will fix your dd
:
:Index: dd.c
:===
:RCS file: /home/ncvs/src/bin/dd/dd.c,v
:retrieving revision 1.27
:diff -u -r1.27 dd.c
:--- dd.c   1999/10/03 18:49:51 1.27
:+++ dd.c   1999/12/05 23:17:16
:@@ -216,6 +216,16 @@
:   if (ioctl(io-fd, FIODTYPE, type) == -1) {
:   err(1, "%s", io-name);
:   } else {
:+if (type  D_DISK) {
:+static int one = 1;
:+#ifndef   DIOCWLABEL
:+#define   DIOCWLABEL _IOW('d', 109, int) 
:+#endif
:+fprintf(stderr, "is a disk\n");
:+if (ioctl(io-fd, DIOCWLABEL, one)  0) {
:+perror("DIOCWLABEL");
:+}
:+}
:   if (type  D_TAPE)
:   io-flags |= ISTAPE;
:   else if (type  (D_DISK | D_MEM))
:

Woa, timeout!

dd if=/dev/da1 of=/dev/da2 had better work - the label area should not
be in any way write protected when you use the non-partition device.

If you were to use /dev/da2c, *then* it would be reasonable to 
write-protect the label area.  But if you use /dev/da2 it is not.

We should not go around hacking dd when the problem is in the kernel.

-Matt
Matthew Dillon 
[EMAIL PROTECTED]


To Unsubscribe: send mail to [EMAIL PROTECTED]
with "unsubscribe freebsd-current" in the body of the message



Re: oh, btw...

1999-12-06 Thread David O'Brien

On Mon, Dec 06, 1999 at 10:00:51AM -0800, Matthew Dillon wrote:
 dd if=/dev/da1 of=/dev/da2 had better work - the label area should not
 be in any way write protected when you use the non-partition device.

It doesn't on my Alpha.  I certainly would not mind a confermation on
this.

-- 
-- David([EMAIL PROTECTED])


To Unsubscribe: send mail to [EMAIL PROTECTED]
with "unsubscribe freebsd-current" in the body of the message