> From: Jan Beulich <jbeul...@suse.com> > Sent: Monday, April 17, 2023 11:41 AM > To: Andrew Cooper <andrew.coop...@citrix.com> > Cc: Konrad Rzeszutek Wilk <konrad.w...@oracle.com>; Ross Lagerwall > <ross.lagerw...@citrix.com>; Xen-devel <xen-devel@lists.xenproject.org> > Subject: Re: [PATCH] xen/livepatch: Fix secure_payload() in non-debug builds > > On 17.04.2023 11:58, Andrew Cooper wrote: > > The ro_pages + rw_pages + text_pages != payload->pages check is not > > something > > which is reasonable to skip at runtime. Rewrite it to not be an ASSERT(). > > Isn't this merely a sanity check? IOW isn't returning -EINVAL in this case > misleading, as to calling "invalid input" what really is an internal error > in Xen? But anyway, I guess I'll leave this to the maintainers. >
Yes, it looks like it is just a sanity check of the payload->pages calculation in move_payload(). Since it is not dependent on the payload, I think ASSERT() is correct. Ross
