On 3/12/20 12:31 AM, Stephen Rothwell wrote:
> Hi all,
>
> After merging the akpm-current tree, today's linux-next build (powerpc
> ppc64_defconfig) produced this warning:
>
> fs/hugetlbfs/inode.c: In function 'remove_inode_hugepages':
> fs/hugetlbfs/inode.c:460:44: warning: 'hash' may be used uninitialized in this function [-Wmaybe-uninitialized]
> 460 | mutex_unlock(&hugetlb_fault_mutex_table[hash]);
> | ^
> fs/hugetlbfs/inode.c:463:5: warning: 'index' may be used uninitialized in this function [-Wmaybe-uninitialized]
> 463 | hugetlb_vmdelete_list(&mapping->i_mmap,
> | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> 464 | index * pages_per_huge_page(h),
> | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> 465 | (index + 1) * pages_per_huge_page(h));
> | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
>
> Introduced by commit
>
> 6fdc8f8d1781 ("hugetlbfs: use i_mmap_rwsem to address page fault/truncate race")
>
This is a false positive. However, there are more serious issues with this
patch series as reported here:
https://lore.kernel.org/linux-mm/1584028670.7365.182.camel@xxxxxx/
I'm working on the issue, but these may need to be reverted if I can not come
up with a solution quickly. So, I am ignoring the false positive warning
until the more serious issue is resolved.
--
Mike Kravetz
