[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: Using LOCK# to prevent deadlock
- To: Mailing List Recipients <pci-sig-request@znyx.com>
- Subject: Re: Using LOCK# to prevent deadlock
- From: "Monish Shah" <monish@mcsy2.fc.hp.com>
- Date: Wed, 28 Aug 1996 13:06:18 -0600
- In-Reply-To: frank.story@TEMPE.VLSI.COM "Unidentified subject!" (Aug 28, 9:21am)
- References: <9608281621.AA20173@znyx.com>
- Resent-Date: Wed, 28 Aug 1996 13:06:18 -0600
- Resent-From: pci-sig-request@znyx.com
- Resent-Message-Id: <"ioXKn.0.p87.8b99o"@dart>
- Resent-Sender: pci-sig-request@znyx.com
On Aug 28, 9:21am, frank.story@TEMPE.VLSI.COM wrote:
> Subject: Unidentified subject!
> LOCK# is also used to cover a potential deadlock scenario for PCI to
> PCI bridges. See page 115 of the 2.1 spec for some discussion of the
> problem.
Yes, a host bridge designer may choose to solve that deadlock using LOCK#.
So, let me clarify my original recommendation: PCI add-in devices should
avoid the use of LOCK# if at all possible.
> Frank
Monish Shah
Hewlett Packard
§ $