[8920] in linux-scsi channel archive
Re: NCR5380.c: IO request dead lock within the _main() function?
daemon@ATHENA.MIT.EDU (Alan Cox)
Sun May 28 05:09:07 2000
To: christian@montanari.f9.co.uk (christian Montanari)
Date: Sun, 28 May 2000 10:05:01 +0100 (BST)
Cc: linux-scsi@vger.rutgers.edu
In-Reply-To: <14640.35038.648776.519709@gedeon.ailsa.co.uk> from "christian Montanari" at May 28, 2000 04:50:37 AM
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Message-Id: <E12vz0J-0005Hg-00@the-village.bc.nu>
From: Alan Cox <alan@lxorguk.ukuu.org.uk>
> in the SELECTION PHASE:at the last before statement in NCR5380_main()
> after a "won arbitration" in the (second) return of
> NCR5380_select(), the spin_lock_irq(&io_request_lock) hangs
> (it looks like a dead lock), and so does the system (too bad).
The NCR5380 driver is not SMP safe. The core processing loop needs rewriting
to fix this. Unfortunately the code is so bad nobody has had time to figure
what it is meant to be doing, let alone fix it.
-
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majordomo@vger.rutgers.edu
