Skip to content

Commit

Permalink
net: introduce helper sendpages_ok()
Browse files Browse the repository at this point in the history
Network drivers are using sendpage_ok() to check the first page of an
iterator in order to disable MSG_SPLICE_PAGES. The iterator can
represent list of contiguous pages.

When MSG_SPLICE_PAGES is enabled skb_splice_from_iter() is being used,
it requires all pages in the iterator to be sendable. Therefore it needs
to check that each page is sendable.

The patch introduces a helper sendpages_ok(), it returns true if all the
contiguous pages are sendable.

Drivers who want to send contiguous pages with MSG_SPLICE_PAGES may use
this helper to check whether the page list is OK. If the helper does not
return true, the driver should remove MSG_SPLICE_PAGES flag.

Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Signed-off-by: Ofir Gal <ofir.gal@volumez.com>
Acked-by: Jakub Kicinski <kuba@kernel.org>
Link: https://lore.kernel.org/r/20240718084515.3833733-2-ofir.gal@volumez.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
  • Loading branch information
Ofir Gal authored and Jens Axboe committed Jul 28, 2024
1 parent 79c6c60 commit 23a55f4
Showing 1 changed file with 19 additions and 0 deletions.
19 changes: 19 additions & 0 deletions include/linux/net.h
Original file line number Diff line number Diff line change
Expand Up @@ -322,6 +322,25 @@ static inline bool sendpage_ok(struct page *page)
return !PageSlab(page) && page_count(page) >= 1;
}

/*
* Check sendpage_ok on contiguous pages.
*/
static inline bool sendpages_ok(struct page *page, size_t len, size_t offset)
{
struct page *p = page + (offset >> PAGE_SHIFT);
size_t count = 0;

while (count < len) {
if (!sendpage_ok(p))
return false;

p++;
count += PAGE_SIZE;
}

return true;
}

int kernel_sendmsg(struct socket *sock, struct msghdr *msg, struct kvec *vec,
size_t num, size_t len);
int kernel_sendmsg_locked(struct sock *sk, struct msghdr *msg,
Expand Down

0 comments on commit 23a55f4

Please sign in to comment.