On Fri, 08 Dec 2006 19:56:20 +0100 Richard Knutsson wrote:Oh, an error of thought when I read the patch 7dfb71030f7636a0d65200158113c37764552f93 made that statement. After more checking, sched.h is apperarently included in suspend.h from swap.h so the direct include of sched.h in most drivers was/is not nessecary.
CC [M] fs/jfs/jfs_txnmgr.o
In file included from fs/jfs/jfs_txnmgr.c:49:
include/linux/freezer.h: In function âfrozenâ:
include/linux/freezer.h:9: error: dereferencing pointer to incomplete type
include/linux/freezer.h:9: error: âPF_FROZENâ undeclared (first use in this function)
<snip>
fs/jfs/jfs_txnmgr.c: In function âfreezingâ:
include/linux/freezer.h:18: warning: control reaches end of non-void function
make[2]: *** [fs/jfs/jfs_txnmgr.o] Error 1
make[1]: *** [fs/jfs] Error 2
make: *** [fs] Error 2
Signed-off-by: Richard Knutsson <ricknu-0@xxxxxxxxxxxxxx>
---
Guess this is the desired fix, since including linux/sched.h in linux/freezer.h
make little sense.
Why do you say that? freezer.h is what uses those #defined values,
and freezer.h is what uses struct task_struct fields as well,
so it needs sched.h.
CC [M] fs/jfs/jfs_txnmgr.odiff --git a/fs/jfs/jfs_txnmgr.c b/fs/jfs/jfs_txnmgr.c
index d558e51..2aee0a8 100644
--- a/fs/jfs/jfs_txnmgr.c
+++ b/fs/jfs/jfs_txnmgr.c
@@ -46,6 +46,7 @@ #include <linux/fs.h>
#include <linux/vmalloc.h>
#include <linux/smp_lock.h>
#include <linux/completion.h>
+#include <linux/sched.h>
#include <linux/freezer.h>
#include <linux/module.h>
#include <linux/moduleparam.h>