On 01/15/2016 01:53 PM, Linus Torvalds wrote:I have just sent out the latest version of the patch in the email with the subject
On Fri, Jan 15, 2016 at 1:46 PM, Andy Lutomirski <luto@xxxxxxxxxxxxxx>
wrote:
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?
Why would they care?
Also, I don't think you actually need to change vm_file - we have this
whole notion of "inode->i_data" vs "inode->i_mapping".
So I think you could set the "i_mapping" of the fuse inode to be the
i_mapping of the passed-through-to inode, and it should just work.
And no, I didn't look into this very deeply, and maybe there is some
annoying detail that would make that not work well. But that's part of
the whole point of the i_mapping indirection: so that you can share
the page cache when you have two separate anchor points. I think coda
uses it for the local caching, and block devices use it to not have
mapping aliases between different inodex that all are the same block
device.
Linus
Hi
Thanks for your support.
I am looking into adding the mmap by change the i_mmaping but seems to
be getting stuck for some stress tests.
Once i get that debugged and working i will definitely send out a new
patch with that functionality.
( Sending it as a second patch will help make it easier to go through
the legal redtape and procedures i am forced to follow).
For now i am going to update the current patch and call it "passthrough"
as you suggested and also update the commit message giving a clearer
explanation of the motivation.
Once i get it tested I can send it out for consideration and review
Please do let me know if you have any questions or concerns on this and
i will try my best to address them all.