On 21-02-04 21:53:29, John Groves (jgroves) wrote:
> Micron Confidential
>
>
>
> From: Dan Williams <[email protected]>
> Date: Monday, February 1, 2021 at 1:28 PM
> To: Ben Widawsky <[email protected]>
> Cc: Konrad Rzeszutek Wilk <[email protected]>,
> [email protected] <[email protected]>, Linux ACPI
> <[email protected]>, Linux Kernel Mailing List
> <[email protected]>, linux-nvdimm
> <[email protected]>, Linux PCI <[email protected]>,
> Bjorn Helgaas <[email protected]>, Chris Browy
> <[email protected]>, Christoph Hellwig <[email protected]>, Ira
> Weiny <[email protected]>, Jon Masters <[email protected]>, Jonathan
> Cameron <[email protected]>, Rafael Wysocki
> <[email protected]>, Randy Dunlap <[email protected]>,
> Vishal Verma <[email protected]>, [email protected]
> <[email protected]>, John Groves (jgroves)
> <[email protected]>, Kelley, Sean V <[email protected]>
> Subject: [EXT] Re: [PATCH 04/14] cxl/mem: Implement polled mode mailbox
>
> CAUTION: EXTERNAL EMAIL. Do not click links or open attachments unless
> you recognize the sender and were expecting this message.
> On Mon, Feb 1, 2021 at 11:13 AM Ben Widawsky <[email protected]>
> wrote:
> >
> > On 21-02-01 12:54:00, Konrad Rzeszutek Wilk wrote:
> > > > +#define
> cxl_doorbell_busy(cxlm)
> \
> > > > + (cxl_read_mbox_reg32(cxlm, CXLDEV_MB_CTRL_OFFSET)
> & \
> > > > + CXLDEV_MB_CTRL_DOORBELL)
> > > > +
> > > > +#define CXL_MAILBOX_TIMEOUT_US 2000
> > >
> > > You been using the spec for the values. Is that number also from it
> ?
> > >
> >
> > Yes it is. I'll add a comment with the spec reference.
>
>
>
> From section 8.2.8.4 in the CXL 2.0 spec: “The mailbox command timeout
> is 2 seconds.” So this should be:
>
>
> #define CXL_MAILBOX_TIMEOUT_US 2000000
>
>
> …right? 2000us is 2ms…
>
>
Thanks. This was caught already in earlier review by David Rientjes
<[email protected]>
It's renamed CXL_MAILBOX_TIMEOUT_MS 2000