[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20201005054552.GN2968@vkoul-mobl>
Date: Mon, 5 Oct 2020 11:15:52 +0530
From: Vinod Koul <vkoul@...nel.org>
To: Dave Jiang <dave.jiang@...el.com>
Cc: tglx@...utronix.de, mingo@...hat.com, bp@...en8.de,
dan.j.williams@...el.com, tony.luck@...el.com, jing.lin@...el.com,
ashok.raj@...el.com, sanjay.k.kumar@...el.com,
fenghua.yu@...el.com, kevin.tian@...el.com,
David.Laight@...lab.com, dmaengine@...r.kernel.org,
linux-kernel@...r.kernel.org
Subject: Re: [PATCH v6 4/5] dmaengine: idxd: Clean up descriptors with fault
error
On 04-10-20, 21:55, Dave Jiang wrote:
> > > +static bool process_fault(struct idxd_desc *desc, u64 fault_addr)
> > > +{
> > > + if ((u64)desc->hw == fault_addr ||
> > > + (u64)desc->completion == fault_addr) {
> >
> > you are casting descriptor address and completion, I can understand
> > former, but later..? Can you explain this please
> >
>
> It is possible to fail on the completion writeback address if the completion
> address programmed into the descriptor is bad.
Okay thanks for clarifying, maybe would help to add a comment.
--
~Vinod
Powered by blists - more mailing lists