Re: GNTTABOP_setup_table yields -1 PFNs

[Jan Beulich <jbeulich@xxxxxxxx>]

On 08.07.2024 14:50, Taylor R Campbell wrote:
>> Date: Mon, 8 Jul 2024 11:09:21 +0200
>> From: Jan Beulich <jbeulich@xxxxxxxx>
>>
>> On 06.07.2024 04:22, Taylor R Campbell wrote:
>>> On a Xen 4.14 host (with extraversion=.0.88.g1d1d1f53), with version 1
>>> grant tables, where GNTTABOP_query_size initially returns nr_frames=32
>>> max_nr_frames=64, a NetBSD guest repeatedly queries
>>> GNTTABOP_setup_table for successively larger nr_frames from 1 up.
>>
>> First question: Is there some earlier GNTTABOP_setup_table that you invoke?
>> I'd expect (and also observe) nr_frames=1 initially.
> 
> Not that the guest OS invokes.  Perhaps the bootloader, pv-grub 0.97,
> might invoke GNTTABOP_setup_table?  I looked around but couldn't find
> an obvious source for pv-grub 0.97.
> 
>> Second: The version you name is pretty unclear from an upstream perspective.
>> Leaving aside that 4.14 is out of support, it's entirely unclear whether you
>> at least have all bug fixes in place that we have upstream (4.14.6). Without
>> that it's hard to see what you're asking for.
> 
> It looks like it's based on commit:
> 
> commit 1d1d1f5391976456a79daac0dcfe7157da1e54f7
> Author: Roger Pau Monné <roge.rpau@xxxxxxxxxx>
> Date:   Tue Dec 1 15:34:55 2020 +0100
> 
>     x86/vioapic: fix usage of index in place of GSI in vioapic_write_redirent
> 
> Does that help to narrow it down?

Only in so far as telling us that this looks to be far behind.

>  (I don't know if I can rule out local changes, though.)

Which, however, would be relevant.

Any chance you could try, against the version you presently run, the XTF patch I
sent earlier? Assumes of course you're at least lightly familiar with XTF
(xenbits.xen.org/xtf.git).

Jan