linux-next: manual merge of the folio-iomap tree with the nvdimm tree

Stephen Rothwell posted 1 patch 4 years, 6 months ago
linux-next: manual merge of the folio-iomap tree with the nvdimm tree
Posted by Stephen Rothwell 4 years, 6 months ago
Hi all,

FIXME: Add owner of second tree to To:
       Add author(s)/SOB of conflicting commits.

Today's linux-next merge of the folio-iomap tree got a conflict in:

  fs/iomap/buffered-io.c

between commit:

  de291b590286 ("iomap: turn the byte variable in iomap_zero_iter into a ssize_t")

from the nvdimm tree and commits:

  a25def1fe568 ("iomap: Convert __iomap_zero_iter to use a folio")
  4d7bd0eb72e5 ("iomap: Inline __iomap_zero_iter into its caller")

from the folio-iomap tree.

I fixed it up (following Matthew's suggestion, I ended up with the
below) and can carry the fix as necessary. This is now fixed as far as
linux-next is concerned, but any non trivial conflicts should be
mentioned to your upstream maintainer when your tree is submitted for
merging.  You may also want to consider cooperating with the maintainer
of the conflicting tree to minimise any particularly complex conflicts.

-- 
Cheers,
Stephen Rothwell

diff --cc fs/iomap/buffered-io.c
index ff541190fc73,c6b3a148e898..c938bbad075e
--- a/fs/iomap/buffered-io.c
+++ b/fs/iomap/buffered-io.c
@@@ -888,21 -908,34 +907,25 @@@ static loff_t iomap_zero_iter(struct io
  		return length;
  
  	do {
- 		unsigned offset = offset_in_page(pos);
- 		ssize_t bytes = min_t(u64, PAGE_SIZE - offset, length);
- 		struct page *page;
+ 		struct folio *folio;
  		int status;
+ 		size_t offset;
+ 		size_t bytes = min_t(u64, SIZE_MAX, length);
  
- 		status = iomap_write_begin(iter, pos, bytes, &page);
 -		if (IS_DAX(iter->inode)) {
 -			s64 tmp = dax_iomap_zero(pos, bytes, iomap);
 -			if (tmp < 0)
 -				return tmp;
 -			bytes = tmp;
 -			goto good;
 -		}
 -
+ 		status = iomap_write_begin(iter, pos, bytes, &folio);
  		if (status)
  			return status;
  
- 		zero_user(page, offset, bytes);
- 		mark_page_accessed(page);
+ 		offset = offset_in_folio(folio, pos);
+ 		if (bytes > folio_size(folio) - offset)
+ 			bytes = folio_size(folio) - offset;
  
- 		bytes = iomap_write_end(iter, pos, bytes, bytes, page);
- 		if (bytes < 0)
- 			return bytes;
+ 		folio_zero_range(folio, offset, bytes);
+ 		folio_mark_accessed(folio);
+ 
+ 		bytes = iomap_write_end(iter, pos, bytes, bytes, folio);
 -good:
+ 		if (WARN_ON_ONCE(bytes == 0))
+ 			return -EIO;
  
  		pos += bytes;
  		length -= bytes;
Re: linux-next: manual merge of the folio-iomap tree with the nvdimm tree
Posted by Stephen Rothwell 4 years, 6 months ago
Hi all,

On Wed, 22 Dec 2021 21:15:36 +1100 Stephen Rothwell <sfr@canb.auug.org.au> wrote:
>
> Today's linux-next merge of the folio-iomap tree got a conflict in:
> 
>   fs/iomap/buffered-io.c
> 
> between commit:
> 
>   de291b590286 ("iomap: turn the byte variable in iomap_zero_iter into a ssize_t")
> 
> from the nvdimm tree and commits:
> 
>   a25def1fe568 ("iomap: Convert __iomap_zero_iter to use a folio")
>   4d7bd0eb72e5 ("iomap: Inline __iomap_zero_iter into its caller")
> 
> from the folio-iomap tree.

Thanks to the addition of commit

  9e05e95ca8da ("iomap: Fix error handling in iomap_zero_iter()")

to the nvdimm tree, the resolutions is now as below.

-- 
Cheers,
Stephen Rothwell

diff --cc fs/iomap/buffered-io.c
index d3b1169602fa,c6b3a148e898..000000000000
--- a/fs/iomap/buffered-io.c
+++ b/fs/iomap/buffered-io.c
@@@ -888,19 -908,32 +907,23 @@@ static loff_t iomap_zero_iter(struct io
  		return length;
  
  	do {
- 		unsigned offset = offset_in_page(pos);
- 		ssize_t bytes = min_t(u64, PAGE_SIZE - offset, length);
- 		struct page *page;
+ 		struct folio *folio;
  		int status;
+ 		size_t offset;
+ 		size_t bytes = min_t(u64, SIZE_MAX, length);
  
- 		status = iomap_write_begin(iter, pos, bytes, &page);
 -		if (IS_DAX(iter->inode)) {
 -			s64 tmp = dax_iomap_zero(pos, bytes, iomap);
 -			if (tmp < 0)
 -				return tmp;
 -			bytes = tmp;
 -			goto good;
 -		}
 -
+ 		status = iomap_write_begin(iter, pos, bytes, &folio);
  		if (status)
  			return status;
  
- 		zero_user(page, offset, bytes);
- 		mark_page_accessed(page);
+ 		offset = offset_in_folio(folio, pos);
+ 		if (bytes > folio_size(folio) - offset)
+ 			bytes = folio_size(folio) - offset;
+ 
+ 		folio_zero_range(folio, offset, bytes);
+ 		folio_mark_accessed(folio);
  
- 		bytes = iomap_write_end(iter, pos, bytes, bytes, page);
+ 		bytes = iomap_write_end(iter, pos, bytes, bytes, folio);
 -good:
  		if (WARN_ON_ONCE(bytes == 0))
  			return -EIO;