[PATCH] nvmet-tcp: bound sgl->length check in nvmet_tcp_map_data()

[Shivam Kumar]

Hi all,

Gentle ping on this. Is there anything else needed from my side?
Or is a consensus forming around the 2 MiB default?

Happy to send an updated patch if that helps move things along.

Thanks,
Shivam

On Thu, Apr 9, 2026 at 2:08 AM Christoph Hellwig <hch at lst.de> wrote:
>
> On Wed, Apr 08, 2026 at 08:30:12AM +0200, Maurizio Lombardi wrote:
> > Personally, I would pick a default of 2 MiB, it should be large
> > enough to avoid excessive splitting while still being reasonable from a
> > memory perspective.
> >
> > Sagi? Any thought on that?
> >
> > I have seen a patch on this mailing list that makes mdts configurable
> > via configfs, so in case of need a user could still change it
> > to his preferred value.
>
> Right now that just allows adjusting down, so picking the larger
> value sounds sane.  We can still increase it later and/or allow
> adjusting above the default to an upper cap if needed.
>