[RFC PATCH 36/61] cachefiles: Implement read and write parts of new I/O API

From: David Howells
Date: Mon May 04 2020 - 13:13:14 EST


Implement writing into the cache and reading back from the cache inside
cachefiles using asynchronous direct I/O from the specified iterator. The
size and position of the request should be aligned to the reported
dio_block_size.

Errors and completion are reported by callback.

Signed-off-by: David Howells <dhowells@xxxxxxxxxx>
---

fs/cachefiles/io.c | 208 +++++++++++++++++++++++++++++++++++++++++++++++++++-
1 file changed, 202 insertions(+), 6 deletions(-)

diff --git a/fs/cachefiles/io.c b/fs/cachefiles/io.c
index ddb44ec5a199..42e0d620d778 100644
--- a/fs/cachefiles/io.c
+++ b/fs/cachefiles/io.c
@@ -12,30 +12,226 @@
#include <linux/xattr.h>
#include "internal.h"

+struct cachefiles_kiocb {
+ struct kiocb iocb;
+ struct fscache_io_request *req;
+ refcount_t ki_refcnt;
+};
+
+static inline void cachefiles_put_kiocb(struct cachefiles_kiocb *ki)
+{
+ if (refcount_dec_and_test(&ki->ki_refcnt)) {
+ fscache_put_io_request(ki->req);
+ fput(ki->iocb.ki_filp);
+ kfree(ki);
+ }
+}
+
+/*
+ * Handle completion of a read from the cache.
+ */
+static void cachefiles_read_complete(struct kiocb *iocb, long ret, long ret2)
+{
+ struct cachefiles_kiocb *ki = container_of(iocb, struct cachefiles_kiocb, iocb);
+ struct fscache_io_request *req = ki->req;
+
+ _enter("%llx,%ld,%ld", req->len, ret, ret2);
+
+ fscache_end_io_operation(req->cookie);
+
+ if (ret < 0) {
+ req->error = ret;
+ } else if (ret != req->len) {
+ req->error = -ENODATA;
+ } else {
+ req->transferred = ret;
+ set_bit(FSCACHE_IO_DATA_FROM_CACHE, &req->flags);
+ }
+ if (req->io_done)
+ req->io_done(req);
+ cachefiles_put_kiocb(ki);
+}
+
/*
* Initiate a read from the cache.
*/
-int cachefiles_read(struct fscache_object *object,
+int cachefiles_read(struct fscache_object *obj,
struct fscache_io_request *req,
struct iov_iter *iter)
{
- req->error = -ENODATA;
+ struct cachefiles_object *object =
+ container_of(obj, struct cachefiles_object, fscache);
+ struct cachefiles_kiocb *ki;
+ struct file *file = object->backing_file;
+ ssize_t ret = -ENOBUFS;
+
+ _enter("%pD,%li,%llx,%llx/%llx",
+ file, file_inode(file)->i_ino, req->pos, req->len, i_size_read(file->f_inode));
+
+ ki = kzalloc(sizeof(struct cachefiles_kiocb), GFP_KERNEL);
+ if (!ki)
+ goto presubmission_error;
+
+ refcount_set(&ki->ki_refcnt, 2);
+ ki->iocb.ki_filp = get_file(file);
+ ki->iocb.ki_pos = req->pos;
+ ki->iocb.ki_flags = IOCB_DIRECT;
+ ki->iocb.ki_hint = ki_hint_validate(file_write_hint(file));
+ ki->iocb.ki_ioprio = get_current_ioprio();
+ ki->req = req;
+
+ if (req->io_done)
+ ki->iocb.ki_complete = cachefiles_read_complete;
+
+ ret = rw_verify_area(READ, file, &ki->iocb.ki_pos, iov_iter_count(iter));
+ if (ret < 0)
+ goto presubmission_error_free;
+
+ fscache_get_io_request(req);
+ ret = call_read_iter(file, &ki->iocb, iter);
+ switch (ret) {
+ case -EIOCBQUEUED:
+ goto in_progress;
+
+ case -ERESTARTSYS:
+ case -ERESTARTNOINTR:
+ case -ERESTARTNOHAND:
+ case -ERESTART_RESTARTBLOCK:
+ /* There's no easy way to restart the syscall since other AIO's
+ * may be already running. Just fail this IO with EINTR.
+ */
+ ret = -EINTR;
+ /* Fall through */
+ default:
+ cachefiles_read_complete(&ki->iocb, ret, 0);
+ if (ret > 0)
+ ret = 0;
+ break;
+ }
+
+in_progress:
+ cachefiles_put_kiocb(ki);
+ _leave(" = %zd", ret);
+ return ret;
+
+presubmission_error_free:
+ fput(file);
+ kfree(ki);
+presubmission_error:
+ req->error = -ENOMEM;
+ if (req->io_done)
+ req->io_done(req);
+ return -ENOMEM;
+}
+
+/*
+ * Handle completion of a write to the cache.
+ */
+static void cachefiles_write_complete(struct kiocb *iocb, long ret, long ret2)
+{
+ struct cachefiles_kiocb *ki = container_of(iocb, struct cachefiles_kiocb, iocb);
+ struct fscache_io_request *req = ki->req;
+ struct inode *inode = file_inode(ki->iocb.ki_filp);
+
+ _enter("%llx,%ld,%ld", req->len, ret, ret2);
+
+ /* Tell lockdep we inherited freeze protection from submission thread */
+ __sb_writers_acquired(inode->i_sb, SB_FREEZE_WRITE);
+ __sb_end_write(inode->i_sb, SB_FREEZE_WRITE);
+
+ fscache_end_io_operation(req->cookie);
+
+ if (ret < 0)
+ req->error = ret;
+ else if (ret != req->len)
+ req->error = -ENOBUFS;
+ else
+ cachefiles_mark_content_map(req);
if (req->io_done)
req->io_done(req);
- return -ENODATA;
+ cachefiles_put_kiocb(ki);
}

/*
* Initiate a write to the cache.
*/
-int cachefiles_write(struct fscache_object *object,
+int cachefiles_write(struct fscache_object *obj,
struct fscache_io_request *req,
struct iov_iter *iter)
{
- req->error = -ENOBUFS;
+ struct cachefiles_object *object =
+ container_of(obj, struct cachefiles_object, fscache);
+ struct cachefiles_kiocb *ki;
+ struct inode *inode;
+ struct file *file = object->backing_file;
+ ssize_t ret = -ENOBUFS;
+
+ _enter("%pD,%li,%llx,%llx/%llx",
+ file, file_inode(file)->i_ino, req->pos, req->len, i_size_read(file->f_inode));
+
+ ki = kzalloc(sizeof(struct cachefiles_kiocb), GFP_KERNEL);
+ if (!ki)
+ goto presubmission_error;
+
+ refcount_set(&ki->ki_refcnt, 2);
+ ki->iocb.ki_filp = get_file(file);
+ ki->iocb.ki_pos = req->pos;
+ ki->iocb.ki_flags = IOCB_DIRECT | IOCB_WRITE;
+ ki->iocb.ki_hint = ki_hint_validate(file_write_hint(file));
+ ki->iocb.ki_ioprio = get_current_ioprio();
+ ki->req = req;
+
+ if (req->io_done)
+ ki->iocb.ki_complete = cachefiles_write_complete;
+
+ ret = rw_verify_area(WRITE, file, &ki->iocb.ki_pos, iov_iter_count(iter));
+ if (ret < 0)
+ goto presubmission_error_free;
+
+ /* Open-code file_start_write here to grab freeze protection, which
+ * will be released by another thread in aio_complete_rw(). Fool
+ * lockdep by telling it the lock got released so that it doesn't
+ * complain about the held lock when we return to userspace.
+ */
+ inode = file_inode(file);
+ __sb_start_write(inode->i_sb, SB_FREEZE_WRITE, true);
+ __sb_writers_release(inode->i_sb, SB_FREEZE_WRITE);
+
+ fscache_get_io_request(req);
+ ret = call_write_iter(file, &ki->iocb, iter);
+ switch (ret) {
+ case -EIOCBQUEUED:
+ goto in_progress;
+
+ case -ERESTARTSYS:
+ case -ERESTARTNOINTR:
+ case -ERESTARTNOHAND:
+ case -ERESTART_RESTARTBLOCK:
+ /* There's no easy way to restart the syscall since other AIO's
+ * may be already running. Just fail this IO with EINTR.
+ */
+ ret = -EINTR;
+ /* Fall through */
+ default:
+ cachefiles_write_complete(&ki->iocb, ret, 0);
+ if (ret > 0)
+ ret = 0;
+ break;
+ }
+
+in_progress:
+ cachefiles_put_kiocb(ki);
+ _leave(" = %zd", ret);
+ return ret;
+
+presubmission_error_free:
+ fput(file);
+ kfree(ki);
+presubmission_error:
+ req->error = -ENOMEM;
if (req->io_done)
req->io_done(req);
- return -ENOBUFS;
+ return -ENOMEM;
}

/*