On Wed, Aug 1, 2018 at 6:43 PM, Matthew Wilcox [off-list ref] wrote:

On Wed, Aug 01, 2018 at 06:34:45PM +0530, Souptick Joarder wrote:

quoted

quoted

The fact that ext4_page_mkwrite() returns a vm_fault_t, while
block_page_mkwrite() returns an int which then has to get translated
into a vm_fault_t via block_page_mkwrite_return() is I suspect going
to confuse an awful lot of callers.

We have also changed block_page_mkwrite() to return vm_fault_t, but in
a different patch. Hopefully that patch will be in linux-next tree soon.

I didn't sign off on that, so that's not "we", but "I".  And this is
completely against everything I've been telling you for this whole effort.
Patches should each make sense individually.  You can't make this patch
dependent on another patch without putting that in writing.

It was mistake form my side. Sorry about it.

Leave block_page_mkwrite() alone for now.  Eventually it should return
a vm_fault_t, probably.  But that patch needs to be delayed at least
one kernel cycle.

As caller of block_page_mkwrite() are -
fs/ext4/inode.c
fs/nilfs2/file.c

I will merge both changes in a single patch and send it.