Re: [PATCH -mm] fs.h: fix build error with !CONFIG_FILE_LOCKING

Previous thread: none

Next thread: [PATCH 2.6.26.2] ata_piix: IDE mode SATA patch for Intel Ibex Peak DeviceIDs by Seth Heasley on Friday, August 8, 2008 - 3:47 pm. (1 message)
From: Alexander Beregalov
Date: Friday, August 8, 2008 - 3:32 pm

I am not sure it is a right way to fix it, please review.
Commit 4a70917f881a38ae688c98be9396f0a38f559094
(configure-out-file-locking-features) changed behaviour.

From: Alexander Beregalov <a.beregalov@gmail.com>

fs.h: fix build error with !CONFIG_FILE_LOCKING

fs/gfs2/ops_file.c:585: error: implicit declaration of function 'generic_setlease'
fs/gfs2/ops_file.c:637: error: implicit declaration of function '__mandatory_lock'


Signed-off-by: Alexander Beregalov <a.beregalov@gmail.com>
---

 include/linux/fs.h |    3 +++
 1 files changed, 3 insertions(+), 0 deletions(-)

diff --git a/include/linux/fs.h b/include/linux/fs.h
index e9a7a7d..42d9ff9 100644
--- a/include/linux/fs.h
+++ b/include/linux/fs.h
@@ -1051,6 +1051,8 @@ extern struct seq_operations locks_seq_operations;
 #define flock_lock_file_wait(a, b) (-ENOLCK)
 #define __break_lease(a, b) (0)
 #define lease_get_mtime(a, b)
+#define generic_setlease(a, b, c) (0)
+#define vfs_setlease(a, b, c) (0)
 #define lock_may_read(a, b, c) (1)
 #define lock_may_write(a, b, c) (1)
 #endif /* !CONFIG_FILE_LOCKING */
@@ -1647,6 +1649,7 @@ static inline int break_lease(struct inode *inode, unsigned int mode)
 #else /* !CONFIG_FILE_LOCKING */
 #define locks_mandatory_locked(a) (0)
 #define locks_mandatory_area(a, b, c, d, e) (0)
+#define __mandatory_lock(a) (0)
 #define mandatory_lock(a) (0)
 #define locks_verify_locked(a) (0)
 #define locks_verify_truncate(a, b, c) (0)
--

From: Thomas Petazzoni
Date: Thursday, August 14, 2008 - 2:19 am

Le Sat, 9 Aug 2008 02:32:28 +0400,

I must admit that I'm quite confused by the commit you're referring to.
In which tree is it ?

Anyway, this problem has already been fixed in the latest version of
the patch I sent to LKML on August, 4th:
http://marc.info/?l=linux-kernel&m=121802891417797&w=2

However, I just sent another fixup patch to LKML, required to get NFS
server to compile with CONFIG_FILE_LOCKING=n. For reference, I copy it
below.

Sincerly,

Thomas

---

Fixup configure out fs locks support

This patch adds a few more stubs required to get NFS server to compile
properly with CONFIG_FILE_LOCKING=n.

Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>

---
 include/linux/fs.h |    2 ++
 1 file changed, 2 insertions(+)

Index: linuxdev/include/linux/fs.h
===================================================================
--- linuxdev.orig/include/linux/fs.h
+++ linuxdev/include/linux/fs.h
@@ -1050,6 +1050,8 @@
 #define __break_lease(a, b) ({ 0; })
 #define lease_get_mtime(a, b) ({ })
 #define generic_setlease(a, b, c) ({ -EINVAL; })
+#define vfs_setlease(a, b, c) ({ -EINVAL; })
+#define lease_modify(a, b) ({ -EINVAL; })
 #define lock_may_read(a, b, c) ({ 1; })
 #define lock_may_write(a, b, c) ({ 1; })
 #endif /* !CONFIG_FILE_LOCKING */


-- 
Thomas Petazzoni, Free Electrons
Kernel, drivers and embedded Linux development,
consulting, training and support.
http://free-electrons.com
--

From: Alexander Beregalov
Date: Thursday, August 14, 2008 - 3:51 am

It was in rc1-mm1.

Thanks
Previous thread: none

Next thread: [PATCH 2.6.26.2] ata_piix: IDE mode SATA patch for Intel Ibex Peak DeviceIDs by Seth Heasley on Friday, August 8, 2008 - 3:47 pm. (1 message)