Re: [PATCH] fuse: Add support for fuse stacked I/O
From: Andy Lutomirski
Date: Fri Jan 15 2016 - 16:46:47 EST
On Fri, Jan 15, 2016 at 1:43 PM, Linus Torvalds
<torvalds@xxxxxxxxxxxxxxxxxxxx> wrote:
> On Fri, Jan 15, 2016 at 11:29 AM, Nikhilesh Reddy <reddyn@xxxxxxxxxxxxxx> wrote:
>>
>> FUSE allows users to implement extensions to filesystems ..such as enforcing policy or permissions without having to modify the kernel or maintain the policy in the kernel.
>>
>> One such example is what was quoted by Antonio above ..
>> Another example is a fuse based filesystem that tries to enforce additional permissions on a FAT based mount point.
>>
>> From what i could google there are many FUSE based filesystems out there that do things during the open call but simply pass through the read/and write I/O calls to the local "lower" filesystem where they actually store the data.
>
> So I think these are valid use-cases, and I just think that they should
>
> (a) be documented in the commit message as explanations of why people
> would do this/
>
> (b) not be called "stacked", because that tends to have some other
> connotations to fs people.
>
> I don't know what a better term would be, but you yourself used "pass
> through". Maybe that (perhaps together with a clarification that it's
> a per-file thing) might work fine.
>
> Btw, why is mmap not passed through? That sounds fairly simple and
> straightforward, I'm not seeing why it would be missing.
>
If mmap sets vm_file to the underlying thing, wouldn't CRIU and
anything else that uses map_files get confused? Or did you have
something else in mind?
--Andy