Re: [EXTERNAL]Re: MIPS Cache Coherency Issue

From: Chris Packham
Date: Fri Aug 09 2019 - 00:26:59 EST


On Tue, 2019-08-06 at 01:43 +0000, Tommy Jin wrote:
> Hi Chris,
>
> If you're busy with other things, maybe I can give you a hand, so how
> can I replicate this issue locally?

Thanks for the offer but I'm not sure you'll be able to do much without
our specific hardware.

>
> From your log, I can get the following information, please correct me
> if I miss anything.
> kernel version: 5.1.0

That's what I was using at the time. I imagine 5.2 or 5.3 would behave
the same.

> hardware platform: broadcom BMIPS4350

The specific SoC isÂBCM6818GR which I believe is the same core as the
BMIPS4355.

> I probably couldn't get a broadcom BMIPS4350, is there any simulator
> for this board(e.g. QEMU)?

Yeah it's a pretty old chip. I'm not aware of any QEMU support, but I
haven't looked.

>
> Best regards,
> Tommy
> From: linux-mips-owner@xxxxxxxxxxxxxxx <linux-mips-owner@xxxxxxxxxxxx
> org> on behalf of Chris Packham <Chris.Packham@xxxxxxxxxxxxxxxxxxx>
> Sent: Tuesday, August 6, 2019 5:13 AM
> To: Tommy Jin <tjin@xxxxxxxxxxxx>
> Cc: linux-kernel@xxxxxxxxxxxxxxx <linux-kernel@xxxxxxxxxxxxxxx>; linu
> x-mips@xxxxxxxxxxxxxxx <linux-mips@xxxxxxxxxxxxxxx>
> Subject: [EXTERNAL]Re: MIPS Cache Coherency Issue
> Â
> On Mon, 2019-08-05 at 14:02 +0000, Tommy Jin wrote:
> > Hi Chris,
>
> > My name is Tommy, from wave computing Co,Ltd, our team is working
> on
> > the maintenance of the MIPS kernel.
>
> > You raised a MIPS cache coherency patch which can be found in the
> > following links
> > https://lore.kernel.org/linux-mips/20190528221255.22460-1-chris.pac
> kh
> > am@xxxxxxxxxxxxxxxxxxx/T/#u
>
> > With Paul's patch, It seems you still get "other bad behaviour", it
> > doesn't work for you. Has this issue been resolved?Â
>
> In short no it hasn't been resolved. I haven't been able to spend
> much
> time looking at the issue(s) so I haven't been able to tell if Paul's
> patch uncovered pre-existing issues or caused new ones.
>
> Unfortunately I probably won't get onto it any time soon. I was
> hoping
> to get some of my other team members to pick up the issue but they're
> all busy as well.