Re: [PATCH v1] fsnotify: Pass correct offset to fsnotify_mmap_perm()

[Ryan Roberts]

On 06/10/2025 15:55, Jan Kara wrote:
> On Fri 03-10-25 16:52:36, Ryan Roberts wrote:
>> fsnotify_mmap_perm() requires a byte offset for the file about to be
>> mmap'ed. But it is called from vm_mmap_pgoff(), which has a page offset.
>> Previously the conversion was done incorrectly so let's fix it, being
>> careful not to overflow on 32-bit platforms.
>>
>> Discovered during code review.
>>
>> Cc: <stable@xxxxxxxxxxxxxxx>
>> Fixes: 066e053fe208 ("fsnotify: add pre-content hooks on mmap()")
>> Signed-off-by: Ryan Roberts <ryan.roberts@xxxxxxx>
>> ---
>> Applies against today's mm-unstable (aa05a436eca8).
> 
> Thanks Ryan! I've added the patch to my tree. As a side note, I know the
> callsite is in mm/ but since this is clearly impacting fsnotify, it would
> be good to add to CC relevant people (I'm not following linux-mm nor
> linux-kernel) and discovered this only because of Kiryl's link...

Ahh good point... Sorry I was sleepwalking through the process on Friday
afternoon and blindly sent it to the maintainers and reviewers that
get_maintainer.pl spat out. It didn't even occur to me that this wasn't an mm
thing. :-|

> 
> 								Honza
> 
>>  mm/util.c | 3 ++-
>>  1 file changed, 2 insertions(+), 1 deletion(-)
>>
>> diff --git a/mm/util.c b/mm/util.c
>> index 6c1d64ed0221..8989d5767528 100644
>> --- a/mm/util.c
>> +++ b/mm/util.c
>> @@ -566,6 +566,7 @@ unsigned long vm_mmap_pgoff(struct file *file, unsigned long addr,
>>  	unsigned long len, unsigned long prot,
>>  	unsigned long flag, unsigned long pgoff)
>>  {
>> +	loff_t off = (loff_t)pgoff << PAGE_SHIFT;
>>  	unsigned long ret;
>>  	struct mm_struct *mm = current->mm;
>>  	unsigned long populate;
>> @@ -573,7 +574,7 @@ unsigned long vm_mmap_pgoff(struct file *file, unsigned long addr,
>>
>>  	ret = security_mmap_file(file, prot, flag);
>>  	if (!ret)
>> -		ret = fsnotify_mmap_perm(file, prot, pgoff >> PAGE_SHIFT, len);
>> +		ret = fsnotify_mmap_perm(file, prot, off, len);
>>  	if (!ret) {
>>  		if (mmap_write_lock_killable(mm))
>>  			return -EINTR;
>> --
>> 2.43.0
>>