Re: [PATCH 0/5] vmsplice: fix some problems in my previous vmsplice patchset

[David Hildenbrand (Arm)]

On 6/6/26 08:10, Askar Safin wrote:
> This patchset is for VFS. Of course, it depends on my previous vmsplice
> patchset.
> 
> I fix some problems in my previous patchset.
> 
> 1. Fix problem with CLASS(fd, f)(fd). See first patch for details.
> This is probably not so important, but I fix it anyway.
> 
> 2. Change "unsigned long" back to "int". See second patch for details.
> Again, this is probably not important, but I want to fix this anyway.
> 
> 3. Fix that LTP vmsplice01 bug.
> 
> See patches for details.
> 
> Please, run that LTP vmsplice01 test again.

How to proceed with this patch set given that, as Joanne explains [1], libfuse
seems to modify the data in the page after splicing, so really depending on
vmsplice() reading data that we update through the page tables -- the page
actually being pinned.

IIUC, reverting the patch might be required. Or maybe Christian can still simply
drop the topic branch entirely.

[1]
https://lore.kernel.org/r/CAJnrk1Y9egYizkx1H9K0cqxSYuB+7vLvQbV7Tf4C5eHFqnnC-A@xxxxxxxxxxxxxx


-- 
Cheers,

David