Accessing ram drive as raw disk volume fails?

Fri Dec 1 09:36:00 GMT 2006

On Dec 1 02:31, Dave Korn wrote:
> On 30 November 2006 19:56, Igor Peshansky wrote:
> > On 2006年11月30日, Dave Korn wrote:
> > 
> >> Hi all,
> >> 
> >> I'm trying to use dd to dump stuff to a usb flash drive (i.e. mass
> >> [...]
> It was. The strace showed:
>> 91 15071 [main] dd 3468 fhandler_base::read: returning 100, binary mode
> 29 15100 [main] dd 3468 readv: 100 = readv (0, 0x22EE30, 1), errno 0
> 31 15131 [main] dd 3468 writev: writev (1, 0x22EE20, 1)
> 29 15160 [main] dd 3468 fhandler_base::write: binary write
> 53 15213 [main] dd 3468 fhandler_dev_raw::write_file: 0 (err 87) =
 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
			 Tsk, tsk, tsk. You're using an old
			 Cygwin version.
> WriteFile (1756, 268607488, write 100, written 0, 0)
> 34 15247 [main] dd 3468 fhandler_dev_floppy::is_eom: end of medium
>> which is "ERROR_INVALID_PARAMETER 87 The parameter is incorrect."
>> Looking at a comment in fhandler_dev_floppy::ioctl, it looks to me like the
> issue is that writes to a raw block device have to be integer multiples of the
> blocksize:
>> case RDSETBLK:
> /* Just check the restriction that blocksize must be a multiple
> 	 of the sector size of the underlying volume sector size,
> 	 then fall through to fhandler_dev_raw::ioctl. */
>> which I imagine I'll find documented even in MSDN somewhere if I look hard
> enough. So, the answers are:
>> >> Is this expected behaviour under cygwin? 
>> Yes, as indeed under windows in general.
>> >> What does "no space left on
> >> device" even mean for a raw disk volume as opposed to a logical volume /
> >> partition?
>> It means 'invalid parameter' in this particular instance. 

The reason you're seeing this is that ERROR_INVALID_PARAMETER is
explicitely treated as EOM, see fhandler_floppy.cc, right at the start.
There was a time way back when I did that with a purpose. The problem
is, I don't know what this purpose was. I *think* it had something
to do with some device drivers (floppy?!?) actually returning
ERROR_INVALID_PARAMETER when hitting EOM on reading or writing but
I'm not sure anymore.
If you have fun and a chance to test on various different raw disk
devices, I'd be glad to see a patch which handles EOM better, maybe
even differently on reading and writing.
Corinna
-- 
Corinna Vinschen Please, send mails regarding Cygwin to
Cygwin Project Co-Leader cygwin AT cygwin DOT com
Red Hat
--
Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple
Problem reports: http://cygwin.com/problems.html
Documentation: http://cygwin.com/docs.html
FAQ: http://cygwin.com/faq/