[PATCH v3] io_uring: fix incorrect __kernel_rwf_t cast

[Vasily Averin]

Currently 'make C=1 fs/io_uring.o' generates sparse warning:

  CHECK   fs/io_uring.c
fs/io_uring.c: note: in included file (through
include/trace/trace_events.h, include/trace/define_trace.h, i
nclude/trace/events/io_uring.h):
./include/trace/events/io_uring.h:488:1:
 warning: incorrect type in assignment (different base types)
    expected unsigned int [usertype] op_flags
    got restricted __kernel_rwf_t const [usertype] rw_flags

This happen on cast of sqe->rw_flags which is defined as __kernel_rwf_t,
this type is bitwise and requires __force attribute for any casts.
However rw_flags is a member of the union, and its access can be safely
replaced by using of its neighbours, so let's use poll32_events to fix
the sparse warning.

Signed-off-by: Vasily Averin <vvs@xxxxxxxxxx>
---
v3:
  1) fix only hunk in TRACE_EVENT(io_uring_req_failed),
     rest ones was fixed by Christoph Hellwig already.
  2) updated patch description

v2: updated according to comments by Christoph Hellwig

---
 include/trace/events/io_uring.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/include/trace/events/io_uring.h b/include/trace/events/io_uring.h
index 80d2588a090c..6ba87a290a24 100644
--- a/include/trace/events/io_uring.h
+++ b/include/trace/events/io_uring.h
@@ -520,7 +520,7 @@ TRACE_EVENT(io_uring_req_failed,
 		__entry->off		= sqe->off;
 		__entry->addr		= sqe->addr;
 		__entry->len		= sqe->len;
-		__entry->op_flags	= sqe->rw_flags;
+		__entry->op_flags	= sqe->poll32_events;
 		__entry->buf_index	= sqe->buf_index;
 		__entry->personality	= sqe->personality;
 		__entry->file_index	= sqe->file_index;
-- 
2.31.1