Re: [PATCH v2 0/7] Allow dynamic allocation of software IO TLB bounce buffers

[Mike Lothian]

On Tue, 11 Jul 2023 at 14:21, Petr Tesařík <petr@xxxxxxxxxxx> wrote:
>
> On Tue, 11 Jul 2023 12:29:44 +0100
> Mike Lothian <mike@xxxxxxxxxxxxxx> wrote:
>
> > On Tue, 11 Jul 2023 at 10:03, Petr Tesařík <petr@xxxxxxxxxxx> wrote:
> > >
> > > On Mon, 10 Jul 2023 23:23:45 +0100
> > > Mike Lothian <mike@xxxxxxxxxxxxxx> wrote:
> > >
> > > > Hi
> > > >
> > > > I was hoping this might land for 6.5-rc1, is there a new version that
> > > > might apply against 6.5?
> > >
> > > Yes, there is a v3, which is a complete rewrite based on feedback from
> > > various people on this mailing list:
> > >
> > > https://lore.kernel.org/linux-iommu/cover.1687859323.git.petr.tesarik.ext@xxxxxxxxxx/T/
> > >
> > > Petr T
> > >
> >
> >
> > Patch 2 doesn't apply cleanly for me on 6.5-rc1
>
> Ah, right. I'm going to rebase the series and include a few other
> suggested changes.
>
> I'm a bit worried that Christoph and all other maintainers (all taken
> back into Cc) have stayed silent about the v3 series.
>
> @Christoph: Are uncomfortable with something in the idea itself, or are
> you just busy with other things?
>
> Petr T

I imagine once 6.4 starts appearing in distros there will be more bugs
appearing when people's wifi disconnects

If this is being delayed until 6.6 would it be better to increase the
default number, perhaps based on how much memory the system has and
maybe back port the fix?