io_uring: fix manual setup of iov_iter for fixed buffers

Commit bd11b3a3 ("io_uring: don't use iov_iter_advance() for fixed
buffers") introduced an optimization to avoid using the slow
iov_iter_advance by manually populating the iov_iter iterator in some
cases.

However, the computation of the iterator count field was erroneous: The
first bvec was always accounted for an extent of page size even if the
bvec length was smaller.

In consequence, some I/O operations on fixed buffers were unable to
operate on the full extent of the buffer, consistently skipping some
bytes at the end of it.

Fixes: bd11b3a3 ("io_uring: don't use iov_iter_advance() for fixed buffers")
Cc: stable@vger.kernel.org
Signed-off-by: Aleix Roca Nonell <aleix.rocanonell@bsc.es>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

fs/io_uring.c

@@ -1097,10 +1097,8 @@ static int io_import_fixed(struct io_ring_ctx *ctx, int rw,
 
 			iter->bvec = bvec + seg_skip;
 			iter->nr_segs -= seg_skip;
-			iter->count -= (seg_skip << PAGE_SHIFT);
+			iter->count -= bvec->bv_len + offset;
 			iter->iov_offset = offset & ~PAGE_MASK;
-			if (iter->iov_offset)
-				iter->count -= iter->iov_offset;
 		}
 	}