Re: [PATCH V1 vfio 6/6] vfio/mlx5: Add REINIT support to VFIO_MIG_GET_PRECOPY_INFO

[Peter Xu <peterx@redhat.com>]

Hi, Avihai,

On Tue, Mar 17, 2026 at 11:58:28AM +0200, Avihai Horon wrote:
> Yes, this is because the VFIO device stop_copy_size may hold data that can
> be transferred only when the device is stopped, i.e., during switchover (as
> opposed to RAM which is fully precopy-able).
> Reporting it as part of estimate() didn't seem right, as precopy iterations
> will not reduce it -- if it's big enough (above the threshold), it may block
> future exact() calls as we can't migrate this data during pre-copy and reach
> below threshold again.

This doesn't look right either. I think we should stick with the exact()
being only the accurate version of estimate() only.  The problem might be
that the current interfacing for both may not suite well enough with VFIO,
because VFIO is so far the only module that may contain stop-only data.

The other known issue (not directly relevant here, but we will need to look
into it very soon) is when VFIO is involved the query-migrate results are
sometimes not containing VFIO data, at least "total" and "remaining" fields.

I'll see if I can prepare some quick patch this week to address them, or at
least raise these problem explicitly so you can have a closer look if possible.

[...]

> Yes, I agree, but it's a delicate flow in QEMU and need to get to the
> details.
> Anyway, this case should be rare and we can further discuss these details
> when I send the QEMU series.

Yep, that's fine.

Thanks,

-- 
Peter Xu