bio: return EINTR if copying to user space got interrupted

commit 2d99b55 upstream.

Commit 35dc248 introduced a check for
current->mm to see if we have a user space context and only copies data
if we do. Now if an IO gets interrupted by a signal data isn't copied
into user space any more (as we don't have a user space context) but
user space isn't notified about it.

This patch modifies the behaviour to return -EINTR from bio_uncopy_user()
to notify userland that a signal has interrupted the syscall, otherwise
it could lead to a situation where the caller may get a buffer with
no data returned.

This can be reproduced by issuing SG_IO ioctl()s in one thread while
constantly sending signals to it.

[js] backport to 3.12

Fixes: 35dc248 [SCSI] sg: Fix user memory corruption when SG_IO is interrupted by a signal
Signed-off-by: Johannes Thumshirn <jthumshirn@suse.de>
Signed-off-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Jens Axboe <axboe@fb.com>
Signed-off-by: Jiri Slaby <jslaby@suse.cz>

fs/bio.c

@@ -1051,15 +1051,19 @@ int bio_uncopy_user(struct bio *bio)
 	if (!bio_flagged(bio, BIO_NULL_MAPPED)) {
 		/*
 		 * if we're in a workqueue, the request is orphaned, so
-		 * don't copy into a random user address space, just free.
+		 * don't copy into a random user address space, just free
+		 * and return -EINTR so user space doesn't expect any data.
 		 */
 		if (current->mm)
 			ret = __bio_copy_iov(bio, bmd->iovecs, bmd->sgvecs,
 					     bmd->nr_sgvecs, bio_data_dir(bio) == READ,
 					     0, bmd->is_our_pages);
-		else if (bmd->is_our_pages)
-			bio_for_each_segment_all(bvec, bio, i)
-				__free_page(bvec->bv_page);
+		else {
+			ret = -EINTR;
+			if (bmd->is_our_pages)
+				bio_for_each_segment_all(bvec, bio, i)
+					__free_page(bvec->bv_page);
+		}
 	}
 	bio_free_map_data(bmd);
 	bio_put(bio);