Re: [PATCH v9 08/15] hugetlb: Use the has_hwpoisoned flag

["David Hildenbrand (Arm)" <david@kernel.org>]

I'm a terrible person and it took me way too long to get back to this :(

> 
> You make it sound so easy ;-)

Heh, as long as you hold a folio reference it really is :)

> 
> If we don't hold a reference on the folio, then the folio (whether it's
> hugetlb or not) can be split.  And then folio_test_hwpoison() can hit
> the assertion that it's now a tail page.

Right. I am/was missing the connection to "generic_file_read_iter() in hugetlbfs".

> 
>> Then you write:
>>
>> "This closes a gap where a page in a previously-poisoned hugetlb folio
>> could be observed to not have the hwpoison bit set."
>>
>> When does the race you describe become relevant? When freeing the hugetlb folio
>> or when demoting it? in both cases, we should be holding the hugetlb lock that
>> nasty hwpoison code could sync with (if it doesn't already do that).
> 
> Before this patch series, we don't always hold the hugetlb lock when
> splitting a hugetlb folio that contains hwpoison.  And we don't want
> to have to grab the hugetlb lock if we can avoid it -- unprivileged
> userspace can hammer these paths hard, and I don't want to see this
> lock be contended.

Right. But doesn't at least the pagecache always hold a folio reference when
testing for hwpoison?

I mean, there is other code that might need care (PFN walkers like memory
offlining), but I was surprised to see pagecache code require a rework of
lockless hwpoison checking.

I'm sure I am missing something.

> 
>> I can understand that we want something that covers both scenarios: an
>> indication whether any large folio has any hwpoisoned page. Such a unification
>> would be good.
> 
> Right, and after this patch, we now have that.  Any unknowing part of
> the kernel can test folio_has_hwpoisoned_page() (assuming it has a
> refcount on the folio).
> 
>> But then, special-casing hugetlb *again* through folio_test_huge_poison(),
>> that's rather odd, no?
> 
> Oh, I'm not happy about it.  But we need an atomic way to determine if
> a page belongs to a hugetlb folio with hwpoison detected.  Short of a
> complete rearchitecture of how we handle hwpoison, this is the best I've
> come up with.

(did I express how much I hate the hwpoison infrastructure and how it's racy
left and right? :) )

> 
>> (what on earth is "huge_poison" is this supposed to be "hugetlb_poison" ? Why
>> "poison" and not "hwpoison"? Really odd)
> 
> We're inconsistent in our naming on both of these things.  It doesn't
> help that somebody decided to reuse the term "poison" to mean
> "uninitialised struct page".

Right, but let's be consistent with hugetlb and with hwpoison. ;)

> 
>> I'd expect that we actually get rid of folio_test_hwpoison entirely and
>> exclusively work on per-page state and has_hwpoison. But IIUC, now it's some
>> mixture of folio checks, page checks, folio_has, mixed with some hugetlb oddity.
> 
> If we could get rid of HVO, we could do it entirely on struct page.

That's what I hope we will achieve at some point.

> Or, as above, entirely rearchitecture hwpoison handling to not be based
> around pages or folios any more.
> 
>> I am not quite clear whether the change you propose here is actually required
>> for the remainder of this series?
>>
>> 	[PATCH v9 00/15] Use generic_file_read_iter() in hugetlbfs
>>
>> IOW, do we really need all this hugetlb hwpoison handling just to accomplish
>> that, or could some of that (bigger hwpoison rework) be done separately?
> 
> Blame Sashiko!  I'm going to ignore it in future, but it's really good
> at nerd-sniping "hey all of this is already broken and you could fix it
> as part of this series".

Right, and this is only the tip of the iceberg, because the entire hwpoison
infrastructure is a complete hacked-on racy piece of ... engineering excellence.

To summarize my question: is it possible to separate for this series the
pagecache part (Use generic_file_read_iter() in hugetlbfs) from all the hugetlb
hwpoison rework?

OTOH, if it's really not avoidable, please let me know.

> 
>>
>>> +/*
>>> + * Since we cannot split a hugetlb folio, we want to insert a poison
>>> + * entry into the page table for the whole folio even if only one page
>>> + * is poisoned.  Otherwise, we've split down to the PTE level and we only
>>> + * want to poison the precise page
>>> + */
>>> +static bool ttu_create_hwpoison(const struct folio *folio,
>>> +		const struct page *page, enum ttu_flags flags)
>>> +{
>>> +	if (!(flags & TTU_HWPOISON))
>>> +		return false;
>>
>> Why not keep the flag in the caller? Really looks weird given that we have two
>> users and one of them doesn't even need the check and must fake it.
> 
> try_to_unmap_one() is already 200 lines and contains some tricky code flow.
> I'm trying to at least make the problem not-worse.  I think we should
> pull out even more code into helpers, but not in this patch series.

Note that this code now was partly reworked such that this helper will soon no
longer be required.

[...]

>>> @@ -2423,7 +2436,6 @@ static bool try_to_migrate_one(struct folio *folio, struct vm_area_struct *vma,
>>>  	struct mmu_notifier_range range;
>>>  	enum ttu_flags flags = (enum ttu_flags)(long)arg;
>>>  	unsigned long pfn;
>>> -	unsigned long hsz = 0;
>>
>> While these size changes are reasonable, they make this patch harder to review
>> and are not even described in the patch desription.
> 
> I suppose.  Do you want me to take them out?

Cleaner, yes.


Again, sorry for being so terribly slow.

-- 
Cheers,

David