Re: [PATCH v4 1/2] hugetlb: use f_mode & FMODE_HUGETLBFS to identify hugetlbfs files

[Mike Kravetz <mike.kravetz@oracle.com>]

On 6/11/20 6:58 PM, Al Viro wrote:
> On Thu, Jun 11, 2020 at 05:46:43PM -0700, Mike Kravetz wrote:
>> The routine is_file_hugepages() checks f_op == hugetlbfs_file_operations
>> to determine if the file resides in hugetlbfs.  This is problematic when
>> the file is on a union or overlay.  Instead, define a new file mode
>> FMODE_HUGETLBFS which is set when a hugetlbfs file is opened.  The mode
>> can easily be copied to other 'files' derived from the original hugetlbfs
>> file.
>>
>> With this change hugetlbfs_file_operations can be static as it should be.
>>
>> There is also a (duplicate) set of shm file operations used for the routine
>> is_file_shm_hugepages().  Instead of setting/using special f_op's, just
>> propagate the FMODE_HUGETLBFS mode.  This means is_file_shm_hugepages() and
>> the duplicate f_ops can be removed.
> 
> s/HUGETLBFS/HUGEPAGES/, please.
> 
>> While cleaning things up, change the name of is_file_hugepages() to
>> is_file_hugetlbfs().  The term hugepages is a bit ambiguous.
> 
> Don't, especially since the very next patch adds such on overlayfs...

Ok. This is just something I thought might clarify things.  I seem to
recall questions about 'huge page' routines such as "is that for THP or
hugetlb huge pages"?  That was my motivation for the change.  Since this
is only about hugetlbfs, make it explicit.

> Incidentally, can a hugetlbfs be a lower layer, while the upper one
> is a normal filesystem?  What should happen on copyup?

Yes, that seems to work as expected.  When accessed for write the hugetlb
file is copied to the normal filesystem.

The BUG found by syzbot actually has a single hugetlbfs as both lower and
upper.  With the BUG 'fixed', I am not exactly sure what the expected
behavior is in this case.  I may be wrong, but I would expect any operations
that can be performed on a stand alone hugetlbfs to also be performed on
the overlay.  However, mmap() still fails.  I will look into it.

I also looked at normal filesystem lower and hugetlbfs upper.  Yes, overlayfs
allows this.  This is somewhat 'interesting' as write() is not supported in
hugetlbfs.  Writing to files in the overlay actually ended up writing to
files in the lower filesystem.  That seems wrong, but overlayfs is new to me.

Earlier in the discussion of these issues, Colin Walters asked "Is there any
actual valid use case for mounting an overlayfs on top of hugetlbfs?"  I can
not think of one.  Perhaps we should consider limiting the ways in which
hugetlbfs can be used in overlayfs?  Preventing it from being an upper
filesystem might be a good start?  Or, do people think making hugetlbfs and
overlayfs play nice together is useful?
-- 
Mike Kravetz