[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Xen-devel] [PATCH] xen/tmem: Don't use map_domain_page for long-life-time pages.

On Thu, 2013-06-13 at 15:04 +0100, Jan Beulich wrote:
> >>> On 13.06.13 at 15:55, Ian Campbell <Ian.Campbell@xxxxxxxxxx> wrote:
> > On Thu, 2013-06-13 at 14:49 +0100, Jan Beulich wrote:
> >> >>> On 13.06.13 at 15:24, George Dunlap <george.dunlap@xxxxxxxxxxxxx> 
> >> >>> wrote:
> >> > Did I understand correctly that the map_domain_page() was required on 
> >> >  >5TiB systems, presumably because of limited virtual address space?  In 
> >> > which case this code will fail on those systems?
> >> > 
> >> > If that is the case, then we need to have a way to make sure tmem cannot 
> >> > be enabled on such systems.
> >> 
> >> Tmem is full of assumptions that all memory is mapped, and hence
> >> one of the requirements here was to disable tmem when that
> >> assumption is not true (until eventually tmem would get fixed).
> > 
> > Isn't this change moving it further away from that fixed state though?
> In a way, yes. But what do you do? The code already assumes it
> can use mfn_to_virt() or alike on memory taken from
> alloc_domheap_pages(), so it's just being made consistently
> wrong in doing it one more time. Any alternative fix would be way
> more involved, as it would require adding the (un-)mapping around
> the actual use points of the memory.

Yes, I'd temporarily forgotten that we needed a quick fix for 4.3. With
that in mind I suppose this is the best we can do.


Xen-devel mailing list



Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.