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

Re: [Xen-devel] [XEN PATCH v2 1/2] Check zone before merging adjacent blocks in heap

> On 05.02.2020 12:23, David Woodhouse wrote:
>> On Wed, 2020-02-05 at 11:49 +0100, Jan Beulich wrote:
>>> Yet, as you say elsewhere, whether an MFN has an
>>> entry in frame_table[] is entirely unclear, so permitting boot-
>>> allocator pages to be freed via alloc_domheap_pages() nevertheless
>>> still doesn't look any better an idea to me.
>> Hm, I don't think I said that, did I? That would be a new and exciting
>> complication.
>> I think every MFN handed out by the boot allocator *should* have a
>> corresponding entry in the frame table. All the PDX setup is done for
>> those pages, just as it is for heap pages. In fact, some of that setup
>> is *only* done in init_boot_pages() right now, and if page ranges don't
>> go through the boot allocator and end up being transferred to the heap
>> in end_boot_allocator(), things (like badpages= on the command line)
>> don't work right.
> I guess I should have said "whether an MFN has a properly initialized
> entry in frame_table[] is entirely unclear".

Aha, right. Yes, I admit to having said that :)

I think we have a viable path to fixing that, by folding PGC_broken in to
the state bits so that we can disambiguate. Will experiment.


Xen-devel mailing list



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