Hello!
On Tue, Sep 24, 2002 at 12:39:24PM +0200, Jakob Oestergaard wrote:
> > amount of data. If it writes these sectors from the start of the data flow,
> > what will it do on data transefer timeout?
> > So I still think that data is written on disk in 512 bytes atomic blocks
> > until I see IDE device that does otherwise (and then I will probably
> > dig some IDE datasheet and find out they are violating some spec ;) )
> That would be really interesting.
> I mean, my point still stands even though my proof was wrong, unless
> someone can come up with a "proof" (a spec would be close enough) that
> writes must be 512 bytes.
Ok, here's what I have quckly found, it seems to proove that my assumptions are
right:
http://www.repairfaq.org/filipg/LINK/F_IDE-tech.html
Section: "9. Command Descriptions":
Here the quote:
The recommended procedure for executing a command on the selected
drive is:
1. Wait for drive to clear BUSY.
2. Load required parameters in the Command Block Registers.
3. Activate the Interrupt Enable (nIEN) bit.
4. Wait for drive to set DRDY.
5. Write the command code to the Command Register.
Execution of the command begins as soon as the drive loads the Command
Block Register.
Commands:
E8h: Write Buffer
This command enables the host to overwrite the contents of the
drive's sector buffer with any data pattern. On receipt of this
comand the drive sets BUSY within 400 nsec, sets up the sector
buffer for a write operation, sets DRQ and clears BUSY. The
host then writes up to 512 bytes of data to the buffer. The
Read Buffer and Write Buffer are synchronized so that back to
back Read Buffer and Write Buffer commands access the same 512
bytes within the buffer.
30h: Write Sectors with Retry
31h: Write Sectors without Retry
This command writes from 1 to 256 sectors beginning at the
specified sector and as stated earlier, a sector count of 0 in
the Command Block Register will request 256 sectors. When the
drive accepts this command it sets DRQ and wait for the host to
fill the sector buffer with the data to be written to disk. No
interrupt is generated to start the first buffer fill operation
and once the buffer is full the drive clears the DRQ, sets BUSY
and begins execution of the command.
For single sector Write operations, the drive sets DRQ upon
receipt of the command and waits for the host to fill the
sector buffer. Once a sector has been transferred, the drive
sets BUSY and clears DRQ. If the drive is not on the requested
cylinder and head an inplied seek and/or head switch is
performed. Once the desired track is reached the drive searches
for the appropriate ID field.
....
End of quote
So no write should be even started until entire sector buffer is filled.
Bye,
Oleg
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
This archive was generated by hypermail 2b29 : Mon Sep 30 2002 - 22:00:18 EST