Skip to content

Commit

Permalink
iov_iter: add helper to save iov_iter state
Browse files Browse the repository at this point in the history
In an ideal world, when someone is passed an iov_iter and returns X bytes,
then X bytes would have been consumed/advanced from the iov_iter. But we
have use cases that always consume the entire iterator, a few examples
of that are iomap and bdev O_DIRECT. This means we cannot rely on the
state of the iov_iter once we've called ->read_iter() or ->write_iter().

This would be easier if we didn't always have to deal with truncate of
the iov_iter, as rewinding would be trivial without that. We recently
added a commit to track the truncate state, but that grew the iov_iter
by 8 bytes and wasn't the best solution.

Implement a helper to save enough of the iov_iter state to sanely restore
it after we've called the read/write iterator helpers. This currently
only works for IOVEC/BVEC/KVEC as that's all we need, support for other
iterator types are left as an exercise for the reader.

Link: https://lore.kernel.org/linux-fsdevel/CAHk-=wiacKV4Gh-MYjteU0LwNBSGpWrK-Ov25HdqB1ewinrFPg@mail.gmail.com/
Signed-off-by: Jens Axboe <[email protected]>
  • Loading branch information
axboe committed Sep 14, 2021
1 parent d6c338a commit 8fb0f47
Show file tree
Hide file tree
Showing 2 changed files with 51 additions and 0 deletions.
15 changes: 15 additions & 0 deletions include/linux/uio.h
Original file line number Diff line number Diff line change
Expand Up @@ -27,6 +27,12 @@ enum iter_type {
ITER_DISCARD,
};

struct iov_iter_state {
size_t iov_offset;
size_t count;
unsigned long nr_segs;
};

struct iov_iter {
u8 iter_type;
bool data_source;
Expand Down Expand Up @@ -55,6 +61,14 @@ static inline enum iter_type iov_iter_type(const struct iov_iter *i)
return i->iter_type;
}

static inline void iov_iter_save_state(struct iov_iter *iter,
struct iov_iter_state *state)
{
state->iov_offset = iter->iov_offset;
state->count = iter->count;
state->nr_segs = iter->nr_segs;
}

static inline bool iter_is_iovec(const struct iov_iter *i)
{
return iov_iter_type(i) == ITER_IOVEC;
Expand Down Expand Up @@ -233,6 +247,7 @@ ssize_t iov_iter_get_pages(struct iov_iter *i, struct page **pages,
ssize_t iov_iter_get_pages_alloc(struct iov_iter *i, struct page ***pages,
size_t maxsize, size_t *start);
int iov_iter_npages(const struct iov_iter *i, int maxpages);
void iov_iter_restore(struct iov_iter *i, struct iov_iter_state *state);

const void *dup_iter(struct iov_iter *new, struct iov_iter *old, gfp_t flags);

Expand Down
36 changes: 36 additions & 0 deletions lib/iov_iter.c
Original file line number Diff line number Diff line change
Expand Up @@ -1972,3 +1972,39 @@ int import_single_range(int rw, void __user *buf, size_t len,
return 0;
}
EXPORT_SYMBOL(import_single_range);

/**
* iov_iter_restore() - Restore a &struct iov_iter to the same state as when
* iov_iter_save_state() was called.
*
* @i: &struct iov_iter to restore
* @state: state to restore from
*
* Used after iov_iter_save_state() to bring restore @i, if operations may
* have advanced it.
*
* Note: only works on ITER_IOVEC, ITER_BVEC, and ITER_KVEC
*/
void iov_iter_restore(struct iov_iter *i, struct iov_iter_state *state)
{
if (WARN_ON_ONCE(!iov_iter_is_bvec(i) && !iter_is_iovec(i)) &&
!iov_iter_is_kvec(i))
return;
i->iov_offset = state->iov_offset;
i->count = state->count;
/*
* For the *vec iters, nr_segs + iov is constant - if we increment
* the vec, then we also decrement the nr_segs count. Hence we don't
* need to track both of these, just one is enough and we can deduct
* the other from that. ITER_KVEC and ITER_IOVEC are the same struct
* size, so we can just increment the iov pointer as they are unionzed.
* ITER_BVEC _may_ be the same size on some archs, but on others it is
* not. Be safe and handle it separately.
*/
BUILD_BUG_ON(sizeof(struct iovec) != sizeof(struct kvec));
if (iov_iter_is_bvec(i))
i->bvec -= state->nr_segs - i->nr_segs;
else
i->iov -= state->nr_segs - i->nr_segs;
i->nr_segs = state->nr_segs;
}

0 comments on commit 8fb0f47

Please sign in to comment.