Julien Grall writes ("Re: [PATCH for-4.16] xen/arm: allocate_bank_memory: don't
create memory banks of size zero"):
> Instead, it looks like the issue is some part of Xen may fall over if
> one of the bank is zero-sized. But from the earlier discussion [2], this
> is just latent. So I think this should be clarified in the commit message.
Yes, please.
Indeed, this is part of:
> Please explain why you would like to include it in 4.16. In particular
> that as I wrote above, Linux is able to cope with zero-size memory bank.
How is 4.16 bad without this patch ? And, what might it break ?
What mistakes could have been made in analysis and preparation ?
(In this case the preparation of the patch seems very simple but the
analysis much less so.)
Ian.
