On Thu, 12 Jun 2014, Heiko Carstens wrote:

Yes, that doesn't make any sense. I wrote the patch together in a hurry and
didn't think much about it.
So below is the what I think most simple conversion to a vmalloc fallback
approach for seq files. However the question remains if this seems to be an
acceptable approach at all...

I think the approach is fine.  We do now have a generic kvfree() function 
defined in mm.h that will handle the freeing, though, determined by 
is_vmalloc_addr() so seq_free() is no longer necessary.