Re: [PATCH] hugetlbfs: Take read_lock on i_mmap for PMD sharing

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

On 11/7/19 1:49 PM, Mike Kravetz wrote:
> On 11/7/19 11:54 AM, Matthew Wilcox wrote:
>> Are there other current users of the write lock that could use a read lock?
>> At first blush, it would seem that unmap_ref_private() also only needs
>> a read lock on the i_mmap tree.  I don't think hugetlb_change_protection()
>> needs the write lock either.  Nor retract_page_tables().

Sorry, I missed retract_page_tables which is not part of hugetlb code.
The comments below do not apply to retract_page_tables.  Someone would
need to take a closer look to see if that really needs write mode.
-- 
Mike Kravetz

> 
> I believe that the semaphore still needs to be held in write mode while
> calling huge_pmd_unshare (as is done in the call sites above).  Why?
> There is this check for sharing in huge_pmd_unshare,
> 
> 	if (page_count(virt_to_page(ptep)) == 1)
> 		return 0;	// implies no sharing
> 
> Note that huge_pmd_share now increments the page count with the semaphore
> held just in read mode.  It is OK to do increments in parallel without
> synchronization.  However, we don't want anyone else changing the count
> while that check in huge_pmd_unshare is happening.  Hence, the need for
> taking the semaphore in write mode.
>