Skip to contentSkip to conversation

nfsd: Cleanup nfsd/pnfsd headers and code placment

PATCH In review Boaz Harrosh 4 participants 31 messages Oct 21, 2009 – Nov 12, 2009 5 messages not in the archive

Subsystems (3)
lore
Share & tools
lore
#0 lore raw

cover nfsd: Cleanup nfsd/pnfsd headers and code placment

In motivation to better organize pnfs code and have minimal impact on the rest of the nfsd code-base, I've done some cleanups and code movements. The begging of which was done by Andy in the last round.

The first needed thing was to re-introduce some sense into nfsd headers. I've taken all include/linux/nfsd/*.h files, one by one, and included them as first header in a .c file. Then fixed any missing includes from that header.

Here is current state after this patchset:

// Still BAD

#include <linux/nfsd/const.h> 
	/* NEEDS #include <linux/types.h> in linux/sunrpc/msg_prot.h */
#include <linux/nfsd/export.h> 
	/* NEEDS #include <linux/nfsd/nfsfh.h> */

#include <linux/nfsd/syscall.h> /* Broken because of export.h above */

[Should I fix those too?]

// GOOD or fixed
#include <linux/nfsd/cache.h>
#include <linux/nfsd/debug.h>
#include <linux/nfsd/nfs4layoutxdr.h>
#include <linux/nfsd/nfs4pnfsdlm.h>
#include <linux/nfsd/nfsd4_pnfs.h>
#include <linux/nfsd/nfsd.h>
#include <linux/nfsd/nfsfh.h>
#include <linux/nfsd/pnfsd.h>
#include <linux/nfsd/state.h>
#include <linux/nfsd/stats.h>
#include <linux/nfsd/xdr.h>
#include <linux/nfsd/xdr3.h>
#include <linux/nfsd/xdr4.h>

I then proceeded to re-arrange some pnfs code/definitions into pnfsd specific files.

And finally all files thouched in this round where include-cleaned

Here are the list of patches:

[PATCH 1/5] sunrpc: Clean never used include files

This one is for Trond, include-cleaning

[PATCH 2/5] nfsd: Fix independence of a few nfsd related headers [PATCH 3/5] nfsd: Fix independence of linux/nfsd/ headers

These two fix nfsd headers independence. As explained above. (Should I squash these together)

[PATCH 4/5] SQUASHME pnfsd: Move pnfsd code out of nfs4state.c/h

Now it is much easier to do this one.

[PATCH 5/5] nfsd: Remove lots of un-needed includes

And some more include-cleaning

TODO: Some more of above magic done to more nfsd files

Thanks
Boaz

#1 lore raw

PATCH 1/5 sunrpc: Clean never used include files

Remove include of two headers never used by this file. I did not see any users dependency but if exist they should be fixed since these headers are totally irrelevant to here.

---
 include/linux/sunrpc/debug.h |    3 ---
 1 files changed, 0 insertions(+), 3 deletions(-)
include/linux/sunrpc/debug.h +0 -3
index 10709cb..c2786f2 100644
@@ -28,9 +28,6 @@
 
 #ifdef __KERNEL__
 
-#include <linux/timer.h>
-#include <linux/workqueue.h>
-
 /*
  * Enable RPC debugging/profiling.
  */
Signature · 1 line
1.6.2.1
#4 lore raw

PATCH 2/5 nfsd: Fix independence of a few nfsd related headers

In reply to Boaz Harrosh

An header should be compilation independent, .i.e pull in any header who's declarations are directly used by this header. And not let users re-include all it's dependencies all over again.

[At the end of the day what's the use of a header if it does
not have more then one user?]

---
 include/linux/nfs_xdr.h   |    1 +
 include/linux/nfsacl.h    |    1 +
 include/linux/posix_acl.h |    1 +
 3 files changed, 3 insertions(+), 0 deletions(-)
include/linux/nfs_xdr.h +1 -0
index 2848a26..c316ca8 100644
@@ -2,6 +2,7 @@
 #define _LINUX_NFS_XDR_H
 
 #include <linux/nfsacl.h>
+#include <linux/nfs3.h>
 
 /*
  * To change the maximum rsize and wsize supported by the NFS client, adjust
include/linux/nfsacl.h +1 -0
index 43011b6..f321b57 100644
@@ -29,6 +29,7 @@
 #ifdef __KERNEL__
 
 #include <linux/posix_acl.h>
+#include <linux/sunrpc/xdr.h>
 
 /* Maximum number of ACL entries over NFS */
 #define NFS_ACL_MAX_ENTRIES	1024
include/linux/posix_acl.h +1 -0
index 065a365..0dcf674 100644
@@ -9,6 +9,7 @@
 #define __LINUX_POSIX_ACL_H
 
 #include <linux/slab.h>
+#include <linux/fs.h>
 
 #define ACL_UNDEFINED_ID	(-1)
 
Signature · 1 line
1.6.2.1
#32 lore raw

PATCH 3/5 nfsd: Fix independence of linux/nfsd/ headers

In reply to Boaz Harrosh

An header should be compilation independent, .i.e pull in any header who's declarations are directly used by this header. And not let users re-include all it's dependencies all over again.

[At the end of the day what's the use of a header if it does
not have more then one user?]

---
 include/linux/nfsd/cache.h         |    1 +
 include/linux/nfsd/nfs4layoutxdr.h |    2 ++
 include/linux/nfsd/nfsfh.h         |    1 +
 include/linux/nfsd/xdr.h           |    1 +
 include/linux/nfsd/xdr4.h          |    1 +
 5 files changed, 6 insertions(+), 0 deletions(-)
include/linux/nfsd/cache.h +1 -0
index 3a3f589..3890284 100644
@@ -12,6 +12,7 @@
 
 #include <linux/in.h>
 #include <linux/uio.h>
+#include <linux/sunrpc/svc.h>
 
 /*
  * Representation of a reply cache entry.
include/linux/nfsd/nfs4layoutxdr.h +2 -0
index ad7f4cb..d4d05d7 100644
@@ -38,6 +38,8 @@
 
 #if defined(CONFIG_PNFSD)
 
+#include <linux/nfsd/nfsd4_pnfs.h>
+
 /* the nfsd4_pnfs_devlist dev_addr for the file layout type */
 struct pnfs_filelayout_devaddr {
 	struct xdr_netobj	r_netid;
include/linux/nfsd/nfsfh.h +1 -0
index b91760c..b4e024c 100644
@@ -18,6 +18,7 @@
 #ifdef __KERNEL__
 # include <linux/string.h>
 # include <linux/fs.h>
+# include <linux/sunrpc/svc.h>
 #endif
 #include <linux/nfsd/const.h>
 
include/linux/nfsd/xdr.h +1 -0
index a0132ef..84dc0f0 100644
@@ -10,6 +10,7 @@
 #include <linux/fs.h>
 #include <linux/vfs.h>
 #include <linux/nfs.h>
+#include <linux/nfsd/nfsd.h>
 
 struct nfsd_fhandle {
 	struct svc_fh		fh;
include/linux/nfsd/xdr4.h +1 -0
index bdd4beb..9243d1b 100644
@@ -40,6 +40,7 @@
 #define _LINUX_NFSD_XDR4_H
 
 #include <linux/nfs4.h>
+#include <linux/nfsd/nfsd.h>
 #include <linux/nfsd/nfsd4_pnfs.h>
 
 #define NFSD4_MAX_TAGLEN	128
Signature · 1 line
1.6.2.1
#33 lore raw

PATCH 4/5 SQUASHME pnfsd: Move pnfsd code out of nfs4state.c/h

In reply to Boaz Harrosh

* Introduce an fs/nfsd/nfsd_priv.h header that collects private nfsd definitions/declarations. Specially all none exported extern will move to that file.

* Include "nfsd_priv.h" in nfsd source files where needed.

* pnfs specific declarations are moved out of linux/nfsd/state.h and into linux/nfsd/nfsd4_pnfs.h, though removing inclusion of nfsd4_pnfs.h from state.h and all the #if defined(CONFIG_PNFSD)

* Finally moving remaining #if defined(CONFIG_PNFSD) code from nfs4state.c into nfs4pnfsd.c

TODO: I know that lots of other code from include/linux/... nfsd headers could/should be moved into "nfsd_priv.h"

---
 fs/nfsd/export.c                |    1 +
 fs/nfsd/nfs4callback.c          |    3 +
 fs/nfsd/nfs4pnfsd.c             |   83 ++++++++++++++++++++++++++++++
 fs/nfsd/nfs4proc.c              |    1 +
 fs/nfsd/nfs4state.c             |  101 +------------------------------------
 fs/nfsd/nfsd_priv.h             |  105 +++++++++++++++++++++++++++++++++++++++
 include/linux/nfsd/nfsd4_pnfs.h |   61 ++++++++++++++++++----
 include/linux/nfsd/pnfsd.h      |    8 ---
 include/linux/nfsd/state.h      |   85 +-------------------------------
 9 files changed, 246 insertions(+), 202 deletions(-)
 create mode 100644 fs/nfsd/nfsd_priv.h
fs/nfsd/export.c +1 -0
index 0eea1a6..f51c413 100644
@@ -32,6 +32,7 @@
 #include <linux/nfsd/nfsd.h>
 #include <linux/nfsd/nfsfh.h>
 #include <linux/nfsd/pnfsd.h>
+#include "nfsd_priv.h"
 #if defined(CONFIG_SPNFS)
 #include <linux/nfsd4_spnfs.h>
 #if defined(CONFIG_SPNFS_BLOCK)
fs/nfsd/nfs4callback.c +3 -0
index b53194c..3347749 100644
@@ -311,6 +311,9 @@ encode_cb_sequence(struct xdr_stream *xdr, struct nfsd4_cb_sequence *args,
 }
 
 #if defined(CONFIG_PNFSD)
+
+#include "nfsd_priv.h"
+
 static void
 encode_cb_layout(struct xdr_stream *xdr, struct nfs4_layoutrecall *clr,
 		 struct nfs4_cb_compound_hdr *hdr)
fs/nfsd/nfs4pnfsd.c +83 -0

93 lines not on this page · raw message

fs/nfsd/nfs4proc.c +1 -0
index 83acf8d..dd1da22 100644
@@ -52,6 +52,7 @@
 #include <linux/nfsd/pnfsd.h>
 #include <linux/exportfs.h>
 #include <linux/nfsd/nfs4layoutxdr.h>
+#include "nfsd_priv.h"
 #if defined(CONFIG_SPNFS)
 #include <linux/nfsd4_spnfs.h>
 #if defined(CONFIG_SPNFS_BLOCK)
fs/nfsd/nfs4state.c +2 -99

122 lines not on this page · raw message

fs/nfsd/nfsd_priv.h new file +105 -0

105 lines not on this page · raw message

include/linux/nfsd/nfsd4_pnfs.h +50 -11

81 lines not on this page · raw message

include/linux/nfsd/pnfsd.h +0 -8
index d1d7395..6aa3a87 100644
@@ -96,14 +96,6 @@ int nfs4_pnfs_return_layout(struct super_block *, struct svc_fh *,
 					struct nfsd4_pnfs_layoutreturn *);
 void pnfs_set_device_notify(clientid_t *clid, unsigned int types);
 void nfs4_pnfs_state_shutdown(void);
-
-#if defined(CONFIG_PNFSD_LOCAL_EXPORT)
-extern struct sockaddr pnfsd_lexp_addr;
-extern size_t pnfs_lexp_addr_len;
-
-void pnfsd_lexp_init(struct inode *inode);
-#endif /* CONFIG_PNFSD_LOCAL_EXPORT */
-
 #endif /* CONFIG_PNFSD */
 
 #endif /* LINUX_NFSD_PNFSD_H */
include/linux/nfsd/state.h +1 -84

103 lines not on this page · raw message

Signature · 1 line
1.6.2.1
#35 lore raw

PATCH 5/5 nfsd: Remove lots of un-needed includes

In reply to Boaz Harrosh

Once a few headers where fixed lots of includes are not needed any more. (OK probably a few where so even before)

---
 fs/exportfs/nfs4filelayoutxdr.c |    8 --------
 fs/nfsd/export.c                |   16 ----------------
 fs/nfsd/nfs4callback.c          |   14 --------------
 fs/nfsd/nfs4pnfsd.c             |    9 ---------
 fs/nfsd/nfs4proc.c              |    9 ---------
 fs/nfsd/nfs4state.c             |   18 ------------------
 fs/nfsd/pnfsd_lexp.c            |    5 -----
 fs/nfsd/spnfs_com.c             |   13 -------------
 8 files changed, 0 insertions(+), 92 deletions(-)
fs/exportfs/nfs4filelayoutxdr.c +0 -8
index 9ce2a9d..d07685d 100644
@@ -35,15 +35,7 @@
 */
 #if defined(CONFIG_EXPORTFS_FILE_LAYOUT)
 
-#include <linux/module.h>
-#include <linux/sunrpc/svc.h>
-#include <linux/nfsd/nfsd.h>
-#include <linux/nfsd/cache.h>
-#include <linux/nfs4.h>
-#include <linux/nfsd/state.h>
-#include <linux/nfsd/xdr4.h>
 #include <linux/nfsd/nfs4layoutxdr.h>
-#include <linux/nfsd/nfsd4_pnfs.h>
 
 /* We do our-own dprintk so filesystems are not dependent on sunrpc */
 #ifdef dprintk
fs/nfsd/export.c +0 -16
index f51c413..616980c 100644
@@ -14,23 +14,11 @@
  * Copyright (C) 1995, 1996 Olaf Kirch, <[email protected]>
  */
 
-#include <linux/unistd.h>
-#include <linux/slab.h>
-#include <linux/stat.h>
-#include <linux/in.h>
-#include <linux/seq_file.h>
 #include <linux/syscalls.h>
-#include <linux/rwsem.h>
-#include <linux/dcache.h>
 #include <linux/namei.h>
-#include <linux/mount.h>
-#include <linux/hash.h>
-#include <linux/module.h>
 #include <linux/exportfs.h>
 
-#include <linux/sunrpc/svc.h>
 #include <linux/nfsd/nfsd.h>
-#include <linux/nfsd/nfsfh.h>
 #include <linux/nfsd/pnfsd.h>
 #include "nfsd_priv.h"
 #if defined(CONFIG_SPNFS)
@@ -40,10 +28,6 @@
 #endif
 #endif
 #include <linux/nfsd/syscall.h>
-#include <linux/lockd/bind.h>
-#include <linux/sunrpc/msg_prot.h>
-#include <linux/sunrpc/gss_api.h>
-#include <net/ipv6.h>
 
 #define NFSDDBG_FACILITY	NFSDDBG_EXPORT
 
fs/nfsd/nfs4callback.c +0 -14
index 3347749..caf84a6 100644
@@ -33,22 +33,8 @@
  *  SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
  */
 
-#include <linux/module.h>
-#include <linux/list.h>
-#include <linux/inet.h>
-#include <linux/errno.h>
-#include <linux/delay.h>
-#include <linux/sched.h>
-#include <linux/kthread.h>
-#include <linux/sunrpc/xdr.h>
-#include <linux/sunrpc/svc.h>
-#include <linux/sunrpc/clnt.h>
-#include <linux/sunrpc/svcsock.h>
 #include <linux/nfsd/nfsd.h>
 #include <linux/nfsd/state.h>
-#include <linux/sunrpc/sched.h>
-#include <linux/nfs4.h>
-#include <linux/sunrpc/xprtsock.h>
 
 #define NFSDDBG_FACILITY                NFSDDBG_PROC
 
fs/nfsd/nfs4pnfsd.c +0 -9
index fdbc447..cfffead 100644
@@ -23,16 +23,7 @@
 
 #if defined(CONFIG_PNFSD)
 
-#include <linux/param.h>
-#include <linux/slab.h>
-#include <linux/sunrpc/svc.h>
-#include <linux/sunrpc/debug.h>
 #include <linux/nfsd/nfsd.h>
-#include <linux/nfs4.h>
-#include <linux/nfsd/state.h>
-#include <linux/nfsd/xdr4.h>
-#include <linux/exportfs.h>
-#include <linux/nfsd/pnfsd.h>
 #include "nfsd_priv.h"
 
 #define NFSDDBG_FACILITY                NFSDDBG_PROC
fs/nfsd/nfs4proc.c +0 -9
index dd1da22..e22a9e6 100644
@@ -35,23 +35,14 @@
  *  SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
  */
 
-#include <linux/param.h>
-#include <linux/major.h>
-#include <linux/slab.h>
 #include <linux/file.h>
 
-#include <linux/sunrpc/svc.h>
 #include <linux/nfsd/nfsd.h>
 #include <linux/nfsd/cache.h>
-#include <linux/nfs4.h>
 #include <linux/nfsd/state.h>
 #include <linux/nfsd/xdr4.h>
-#include <linux/nfs4_acl.h>
-#include <linux/sunrpc/gss_api.h>
 #if defined(CONFIG_PNFSD)
 #include <linux/nfsd/pnfsd.h>
-#include <linux/exportfs.h>
-#include <linux/nfsd/nfs4layoutxdr.h>
 #include "nfsd_priv.h"
 #if defined(CONFIG_SPNFS)
 #include <linux/nfsd4_spnfs.h>
fs/nfsd/nfs4state.c +0 -18
index 018292e..b635ed0 100644
@@ -34,31 +34,13 @@
 *
 */
 
-#include <linux/param.h>
-#include <linux/major.h>
-#include <linux/slab.h>
-
-#include <linux/sunrpc/svc.h>
-#include <linux/nfsd/nfsd.h>
-#include <linux/nfsd/cache.h>
 #include <linux/file.h>
-#include <linux/mount.h>
-#include <linux/workqueue.h>
 #include <linux/smp_lock.h>
-#include <linux/kthread.h>
-#include <linux/nfs4.h>
 #include <linux/nfsd/state.h>
 #include <linux/nfsd/xdr4.h>
 #include <linux/namei.h>
 #include <linux/swap.h>
-#include <linux/mutex.h>
-#include <linux/lockd/bind.h>
-#include <linux/module.h>
 #include <linux/sunrpc/svcauth_gss.h>
-#include <linux/sunrpc/clnt.h>
-#if defined(CONFIG_PNFSD)
-#include <linux/nfsd/pnfsd.h>
-#endif /* CONFIG_PNFSD */
 #include "nfsd_priv.h"
 
 #define NFSDDBG_FACILITY                NFSDDBG_PROC
fs/nfsd/pnfsd_lexp.c +0 -5
index 2bcb6b5..a09e3c1 100644
@@ -20,12 +20,7 @@
 
 #if defined(CONFIG_PNFSD_LOCAL_EXPORT)
 
-#include <linux/nfs_fs.h>
-#include <linux/sunrpc/svc.h>
 #include <linux/sunrpc/svc_xprt.h>
-#include <linux/nfsd/nfsfh.h>
-#include <linux/nfsd/state.h>
-#include <linux/nfsd/pnfsd.h>
 #include <linux/nfsd/nfs4layoutxdr.h>
 #include <linux/nfsd/debug.h>
  
fs/nfsd/spnfs_com.c +0 -13
index 7c7d48d..deab09b 100644
@@ -39,23 +39,10 @@
  */
 #if defined(CONFIG_SPNFS)
 
-#include <linux/module.h>
-#include <linux/mutex.h>
-#include <linux/init.h>
-#include <linux/types.h>
-#include <linux/slab.h>
-#include <linux/socket.h>
-#include <linux/in.h>
-#include <linux/sched.h>
 #include <linux/namei.h>
 #include <linux/mount.h>
-#include <linux/path.h>
 #include <linux/sunrpc/clnt.h>
-#include <linux/workqueue.h>
 #include <linux/sunrpc/rpc_pipe_fs.h>
-#include <linux/proc_fs.h>
-
-#include <linux/nfs_fs.h>
 
 #include <linux/nfsd4_spnfs.h>
 #include <linux/nfsd/debug.h>
Signature · 1 line
1.6.2.1
#2 lore raw

Re: [pnfs] [PATCH 1/5] sunrpc: Clean never used include files

In reply to Boaz Harrosh
Quoted 15 lines · Boaz Harrosh

Remove include of two headers never used by this file. I did not see any users dependency but if exist they should be fixed since these headers are totally irrelevant to here.

---
 include/linux/sunrpc/debug.h |    3 ---
 1 files changed, 0 insertions(+), 3 deletions(-)
include/linux/sunrpc/debug.h +0 -0
index 10709cb..c2786f2 100644

include/linux/sunrpc/debug.h line 33

Quoted 9 lines
include/linux/sunrpc/debug.h +0 -3
@@ -28,9 +28,6 @@
 
 #ifdef __KERNEL__
 
-#include <linux/timer.h>
-#include <linux/workqueue.h>
-
 /*
  * Enable RPC debugging/profiling.
  */

This patch exposed a missing "#include <linux/types.h>" in linux/sunrpc/rpc_rdma.h. I'll re-post version 2 soon

Boaz

#3 lore raw

PATCH 1/1 sunrpc: Clean never used include files

In reply to Boaz Harrosh

Remove include of two headers never used by this file. Doing so exposed a missing #include <linux/types.h> in include/linux/sunrpc/rpc_rdma.h.

I did not see any other users dependency but if exist they should be fixed since these headers are totally irrelevant to here.

---
 include/linux/sunrpc/debug.h    |    3 ---
 include/linux/sunrpc/rpc_rdma.h |    2 ++
 2 files changed, 2 insertions(+), 3 deletions(-)
include/linux/sunrpc/debug.h +0 -3
index 10709cb..c2786f2 100644
@@ -28,9 +28,6 @@
 
 #ifdef __KERNEL__
 
-#include <linux/timer.h>
-#include <linux/workqueue.h>
-
 /*
  * Enable RPC debugging/profiling.
  */
include/linux/sunrpc/rpc_rdma.h +2 -0
index 87b895d..b78f16b 100644
@@ -40,6 +40,8 @@
 #ifndef _LINUX_SUNRPC_RPC_RDMA_H
 #define _LINUX_SUNRPC_RPC_RDMA_H
 
+#include <linux/types.h>
+
 struct rpcrdma_segment {
 	__be32 rs_handle;	/* Registered memory handle */
 	__be32 rs_length;	/* Length of the chunk in bytes */
Signature · 1 line
1.6.2.1
#5 lore raw

Re: [PATCH 2/5] nfsd: Fix independence of a few nfsd related headers

In reply to Boaz Harrosh
Quoted 7 lines · Boaz Harrosh

An header should be compilation independent, .i.e pull in any header who's declarations are directly used by this header. And not let users re-include all it's dependencies all over again.

[At the end of the day what's the use of a header if it does
not have more then one user?]

The problem with this is that you quickly end up including the same header files over and over and over again as they get pulled in several times over by different headers.

Signature · 6 lines
Trond Myklebust
Linux NFS client maintainer

NetApp
[email protected]
www.netapp.com
#7 lore raw

Re: [PATCH 2/5] nfsd: Fix independence of a few nfsd related headers

In reply to a message not in the archive
Quoted 6 lines · Trond Myklebust
Quoted 5 lines · Boaz Harrosh

An header should be compilation independent, .i.e pull in any header who's declarations are directly used by this header. And not let users re-include all it's dependencies all over again.

Quoted 10 lines
Quoted 2 lines

[At the end of the day what's the use of a header if it does
not have more then one user?]

The problem with this is that you quickly end up including the same header files over and over and over again as they get pulled in several times over by different headers.

What is the problem with that?

two things:
1. it is unavoidable.
2. That's why we invented the #ifndef FOO_H #define FOO_H for.
3. Look at the current mess, to the point that you don't understand why the code does not compile, you end up just copy-past the include list of that other file, and now actually do end with extra un-needed includes. (Don't believe me look at the last patch as proof).
4. So I add to an header a use of a type that now needs a new include. All my users do not compile any more. What to do? OK I'll include it so not to change all existing users all over again. Now we get a double standard. All headers used before any users, must be carried around. The late comers are escaped.
5. The opposite of 4. An header is no longer needed. Extra header left at all users.

It used to be a problem before me and you have begun programing. Since then the cpp looks at it's internal structures and will not re-open. Note that the compiler never sees the second instance, ever.

That said we have no choice of the matter. It is a Kernel style guide. I should know because I was banged on the head with it a couple of times. And rightly so.

Come on that was a joke right
Boaz

#8 lore raw

Re: [PATCH 2/5] nfsd: Fix independence of a few nfsd related headers

Quoted 34 lines · Boaz Harrosh
Quoted 6 lines · Trond Myklebust
Quoted 5 lines · Boaz Harrosh

An header should be compilation independent, .i.e pull in any header who's declarations are directly used by this header. And not let users re-include all it's dependencies all over again.

Quoted 9 lines
Quoted 2 lines

[At the end of the day what's the use of a header if it does
not have more then one user?]

The problem with this is that you quickly end up including the same header files over and over and over again as they get pulled in several times over by different headers.

What is the problem with that?

two things:
1. it is unavoidable.
2. That's why we invented the #ifndef FOO_H #define FOO_H for.
3. Look at the current mess, to the point that you don't understand why the code does not compile, you end up just copy-past the include list of that other file, and now actually do end with extra un-needed includes. (Don't believe me look at the last patch as proof).
4. So I add to an header a use of a type that now needs a new include. All my users do not compile any more. What to do? OK I'll include it so not to change all existing users all over again. Now we get a double standard. All headers used before any users, must be carried around. The late comers are escaped.
5. The opposite of 4. An header is no longer needed. Extra header left at all users.

Quoted 3 lines

It used to be a problem before me and you have begun programing. Since then the cpp looks at it's internal structures and will not re-open. Note that the compiler never sees the second instance, ever.

For the fun of it, I took an average of 5 incremental build times before and after this patchset, average user and system times below:

before: 68.84 12.51
after:  68.74 12.48

This shows a minor improvement of about 0.14%
for user time and 0.27% for system time.

Benny

Quoted 7 lines

That said we have no choice of the matter. It is a Kernel style guide. I should know because I was banged on the head with it a couple of times. And rightly so.

Come on that was a joke right
Boaz

#9 lore raw

Re: [PATCH 2/5] nfsd: Fix independence of a few nfsd related headers

In reply to Boaz Harrosh
Quoted 39 lines · Boaz Harrosh
Quoted 6 lines · Trond Myklebust
Quoted 5 lines · Boaz Harrosh

An header should be compilation independent, .i.e pull in any header who's declarations are directly used by this header. And not let users re-include all it's dependencies all over again.

Quoted 10 lines
Quoted 2 lines

[At the end of the day what's the use of a header if it does
not have more then one user?]

The problem with this is that you quickly end up including the same header files over and over and over again as they get pulled in several times over by different headers.

What is the problem with that?

two things:
1. it is unavoidable.
2. That's why we invented the #ifndef FOO_H #define FOO_H for.
3. Look at the current mess, to the point that you don't understand why the code does not compile, you end up just copy-past the include list of that other file, and now actually do end with extra un-needed includes. (Don't believe me look at the last patch as proof).
4. So I add to an header a use of a type that now needs a new include. All my users do not compile any more. What to do? OK I'll include it so not to change all existing users all over again. Now we get a double standard. All headers used before any users, must be carried around. The late comers are escaped.
5. The opposite of 4. An header is no longer needed. Extra header left at all users.

It used to be a problem before me and you have begun programing. Since then the cpp looks at it's internal structures and will not re-open. Note that the compiler never sees the second instance, ever.

Quoted 6 lines

That said we have no choice of the matter. It is a Kernel style guide. I should know because I was banged on the head with it a couple of times. And rightly so.

Come on that was a joke right
Boaz

No. What I'm saying is that this doesn't have to be an absolute rule. The Kernel style guide assumes that everything in 'include/*' is going to be exported all around the kernel.
The problem is that we put a lot of stuff which is private to fs/nfs and fs/nfsd in there. Those header files do not have to absolutely follow the style guide rule, 'cos we know what is being included before and after them...

Cheers
Trond

Signature · 6 lines
Trond Myklebust
Linux NFS client maintainer

NetApp
[email protected]
www.netapp.com
#11 lore raw

Re: [PATCH 2/5] nfsd: Fix independence of a few nfsd related headers

In reply to a message not in the archive
Quoted 47 lines · Trond Myklebust
Quoted 39 lines · Boaz Harrosh
Quoted 6 lines · Trond Myklebust
Quoted 5 lines · Boaz Harrosh

An header should be compilation independent, .i.e pull in any header who's declarations are directly used by this header. And not let users re-include all it's dependencies all over again.

Quoted 10 lines
Quoted 2 lines

[At the end of the day what's the use of a header if it does
not have more then one user?]

The problem with this is that you quickly end up including the same header files over and over and over again as they get pulled in several times over by different headers.

What is the problem with that?

two things:
1. it is unavoidable.
2. That's why we invented the #ifndef FOO_H #define FOO_H for.
3. Look at the current mess, to the point that you don't understand why the code does not compile, you end up just copy-past the include list of that other file, and now actually do end with extra un-needed includes. (Don't believe me look at the last patch as proof).
4. So I add to an header a use of a type that now needs a new include. All my users do not compile any more. What to do? OK I'll include it so not to change all existing users all over again. Now we get a double standard. All headers used before any users, must be carried around. The late comers are escaped.
5. The opposite of 4. An header is no longer needed. Extra header left at all users.

It used to be a problem before me and you have begun programing. Since then the cpp looks at it's internal structures and will not re-open. Note that the compiler never sees the second instance, ever.

Quoted 6 lines

That said we have no choice of the matter. It is a Kernel style guide. I should know because I was banged on the head with it a couple of times. And rightly so.

Come on that was a joke right
Boaz

Quoted 8 lines

No. What I'm saying is that this doesn't have to be an absolute rule. The Kernel style guide assumes that everything in 'include/*' is going to be exported all around the kernel.
The problem is that we put a lot of stuff which is private to fs/nfs and fs/nfsd in there. Those header files do not have to absolutely follow the style guide rule, 'cos we know what is being included before and after them...

I'm not sure I understand
You are saying that the patches are very good, but only the rule I stated originally could be relaxed a little with private headers where we might get lazy, if the effects are very local?

Well, that's not a problem then, right? just that I can relax a bit if I want.

But I disagree: see 3, 4, 5 above and that last patch I submitted. That patch is only the beginning. 85% of all source files in nfs/nfsd could receive the same love. I only done these I touched. Code tends to stay much-much longer then we spend time on it. Better get it in shape the first time.

Quoted 3 lines

Cheers
Trond

I've now compiled this with x86 allmodeconfig. It should stay in linux-next for a while to make sure there's no side effects.

Boaz

#34 lore raw

Re: [PATCH 4/5] SQUASHME pnfsd: Move pnfsd code out of nfs4state.c/h

In reply to Boaz Harrosh

Boaz, I merged this patch into the pnfs tree.
Only it at this point, until the cleanup, which I like, gets acked.

I actually split it up for pnfsd-files and restricted the use of CONFIG_PNFSD in the private header file only for functions that are defined under the same config variable.

Benny

Quoted 625 lines · Boaz Harrosh

* Introduce an fs/nfsd/nfsd_priv.h header that collects private nfsd definitions/declarations. Specially all none exported extern will move to that file.

* Include "nfsd_priv.h" in nfsd source files where needed.

* pnfs specific declarations are moved out of linux/nfsd/state.h and into linux/nfsd/nfsd4_pnfs.h, though removing inclusion of nfsd4_pnfs.h from state.h and all the #if defined(CONFIG_PNFSD)

* Finally moving remaining #if defined(CONFIG_PNFSD) code from nfs4state.c into nfs4pnfsd.c

TODO: I know that lots of other code from include/linux/... nfsd headers could/should be moved into "nfsd_priv.h"

---
 fs/nfsd/export.c                |    1 +
 fs/nfsd/nfs4callback.c          |    3 +
 fs/nfsd/nfs4pnfsd.c             |   83 ++++++++++++++++++++++++++++++
 fs/nfsd/nfs4proc.c              |    1 +
 fs/nfsd/nfs4state.c             |  101 +------------------------------------
 fs/nfsd/nfsd_priv.h             |  105 +++++++++++++++++++++++++++++++++++++++
 include/linux/nfsd/nfsd4_pnfs.h |   61 ++++++++++++++++++----
 include/linux/nfsd/pnfsd.h      |    8 ---
 include/linux/nfsd/state.h      |   85 +-------------------------------
 9 files changed, 246 insertions(+), 202 deletions(-)
 create mode 100644 fs/nfsd/nfsd_priv.h
fs/nfsd/export.c +1 -0
index 0eea1a6..f51c413 100644
@@ -32,6 +32,7 @@
 #include <linux/nfsd/nfsd.h>
 #include <linux/nfsd/nfsfh.h>
 #include <linux/nfsd/pnfsd.h>
+#include "nfsd_priv.h"
 #if defined(CONFIG_SPNFS)
 #include <linux/nfsd4_spnfs.h>
 #if defined(CONFIG_SPNFS_BLOCK)
fs/nfsd/nfs4callback.c +3 -0
index b53194c..3347749 100644
@@ -311,6 +311,9 @@ encode_cb_sequence(struct xdr_stream *xdr, struct nfsd4_cb_sequence *args,
 }
 
 #if defined(CONFIG_PNFSD)
+
+#include "nfsd_priv.h"
+
 static void
 encode_cb_layout(struct xdr_stream *xdr, struct nfs4_layoutrecall *clr,
 		 struct nfs4_cb_compound_hdr *hdr)
fs/nfsd/nfs4pnfsd.c +83 -0
index d52a2e5..fdbc447 100644
@@ -33,6 +33,7 @@
 #include <linux/nfsd/xdr4.h>
 #include <linux/exportfs.h>
 #include <linux/nfsd/pnfsd.h>
+#include "nfsd_priv.h"
 
 #define NFSDDBG_FACILITY                NFSDDBG_PROC
 
@@ -1440,4 +1441,86 @@ int nfsd_device_notify_cb(struct super_block *sb,
 		__func__, status, notify_num);
 	return status;
 }
+
+/* Create a layoutrecall structure for each client based on the
+ * original structure. */
+int
+create_layout_recall_list(struct list_head *todolist, unsigned *todo_len,
+			  struct nfsd4_pnfs_cb_layout *cbl,
+			  struct nfs4_file *lrfile)
+{
+	struct nfs4_client *clp;
+	unsigned int i, len = 0;
+	int status = 0;
+
+	dprintk("%s: -->\n", __func__);
+
+	/* If client given by fs, just do single client */
+	if (cbl->cbl_seg.clientid) {
+		clp = find_confirmed_client(
+				(clientid_t *)&cbl->cbl_seg.clientid);
+		if (!clp) {
+			status = -ENOENT;
+			dprintk("%s: clientid %llx not found\n", __func__,
+				(unsigned long long)cbl->cbl_seg.clientid);
+			goto out;
+		}
+
+		status = lo_recall_per_client(clp, cbl, lrfile, todolist);
+		if (!status)
+			len++;
+		goto out;
+	}
+
+	/* Check all clients for layout matches */
+	for (i = 0; i < CLIENT_HASH_SIZE; i++)
+		list_for_each_entry(clp, &conf_str_hashtbl[i], cl_strhash) {
+			status = lo_recall_per_client(clp, cbl, lrfile,
+						      todolist);
+			if (!status)
+				len++;
+			else if (status != -ENOENT)
+				goto out;
+		}
+out:
+	*todo_len = len;
+	/* -ENOENT is a good thing don't return it if some recalls are needed */
+	if ((status == -ENOENT) && len)
+		status = 0;
+	dprintk("%s: <-- list len %u status %d\n", __func__, len, status);
+	return status;
+}
+
+/* Create a list of clients to send device notifications. */
+int
+create_device_notify_list(struct list_head *todolist,
+			  struct nfsd4_pnfs_cb_dev_list *ndl)
+{
+	int status = 0, i;
+	struct nfs4_client *clp = NULL;
+	struct nfs4_notify_device *cbnd;
+
+	nfs4_lock_state();
+
+	/* Create notify client list */
+	for (i = 0; i < CLIENT_HASH_SIZE; i++)
+		list_for_each_entry(clp, &conf_str_hashtbl[i], cl_strhash) {
+			if (atomic_read(&clp->cl_deviceref) <= 0)
+				continue;
+			cbnd = kmalloc(sizeof(*cbnd), GFP_KERNEL);
+			if (!cbnd) {
+				status = -ENOMEM;
+				goto out;
+			}
+			cbnd->nd_list = ndl;
+			cbnd->nd_client = clp;
+			list_add(&cbnd->nd_perclnt, todolist);
+			atomic_inc(&clp->cl_count);
+		}
+
+out:
+	nfs4_unlock_state();
+	return status;
+}
+
 #endif /* CONFIG_PNFSD */
fs/nfsd/nfs4proc.c +1 -0
index 83acf8d..dd1da22 100644
@@ -52,6 +52,7 @@
 #include <linux/nfsd/pnfsd.h>
 #include <linux/exportfs.h>
 #include <linux/nfsd/nfs4layoutxdr.h>
+#include "nfsd_priv.h"
 #if defined(CONFIG_SPNFS)
 #include <linux/nfsd4_spnfs.h>
 #if defined(CONFIG_SPNFS_BLOCK)
fs/nfsd/nfs4state.c +2 -99
index 7495df8..018292e 100644
@@ -59,6 +59,7 @@
 #if defined(CONFIG_PNFSD)
 #include <linux/nfsd/pnfsd.h>
 #endif /* CONFIG_PNFSD */
+#include "nfsd_priv.h"
 
 #define NFSDDBG_FACILITY                NFSDDBG_PROC
 
@@ -275,20 +276,6 @@ unhash_delegation(struct nfs4_delegation *dp)
 	nfs4_put_delegation(dp);
 }
 
-/* 
- * SETCLIENTID state 
- */
-
-/* Hash tables for nfs4_clientid state */
-#define CLIENT_HASH_BITS                 4
-#define CLIENT_HASH_SIZE                (1 << CLIENT_HASH_BITS)
-#define CLIENT_HASH_MASK                (CLIENT_HASH_SIZE - 1)
-
-#define clientid_hashval(id) \
-	((id) & CLIENT_HASH_MASK)
-#define clientstr_hashval(name) \
-	(opaque_hashval((name), 8) & CLIENT_HASH_MASK)
-
 /*
  * reclaim_str_hashtbl[] holds known client info from previous reset/reboot
  * used in reboot/reset lease grace period processing
@@ -308,7 +295,7 @@ unhash_delegation(struct nfs4_delegation *dp)
 static struct list_head reclaim_str_hashtbl[CLIENT_HASH_SIZE];
 static int reclaim_str_hashtbl_size = 0;
 static struct list_head	conf_id_hashtbl[CLIENT_HASH_SIZE];
-static struct list_head	conf_str_hashtbl[CLIENT_HASH_SIZE];
+struct list_head	conf_str_hashtbl[CLIENT_HASH_SIZE];
 static struct list_head	unconf_str_hashtbl[CLIENT_HASH_SIZE];
 static struct list_head	unconf_id_hashtbl[CLIENT_HASH_SIZE];
 static struct list_head client_lru;
@@ -4245,87 +4232,3 @@ nfs4_reset_lease(time_t leasetime)
 	user_lease_time = leasetime;
 }
 
-#if defined(CONFIG_PNFSD)
-
-/* Create a layoutrecall structure for each client based on the
- * original structure. */
-int
-create_layout_recall_list(struct list_head *todolist, unsigned *todo_len,
-			  struct nfsd4_pnfs_cb_layout *cbl,
-			  struct nfs4_file *lrfile)
-{
-	struct nfs4_client *clp;
-	unsigned int i, len = 0;
-	int status = 0;
-
-	dprintk("%s: -->\n", __func__);
-
-	/* If client given by fs, just do single client */
-	if (cbl->cbl_seg.clientid) {
-		clp = find_confirmed_client(
-				(clientid_t *)&cbl->cbl_seg.clientid);
-		if (!clp) {
-			status = -ENOENT;
-			dprintk("%s: clientid %llx not found\n", __func__,
-				(unsigned long long)cbl->cbl_seg.clientid);
-			goto out;
-		}
-
-		status = lo_recall_per_client(clp, cbl, lrfile, todolist);
-		if (!status)
-			len++;
-		goto out;
-	}
-
-	/* Check all clients for layout matches */
-	for (i = 0; i < CLIENT_HASH_SIZE; i++)
-		list_for_each_entry(clp, &conf_str_hashtbl[i], cl_strhash) {
-			status = lo_recall_per_client(clp, cbl, lrfile,
-						      todolist);
-			if (!status)
-				len++;
-			else if (status != -ENOENT)
-				goto out;
-		}
-out:
-	*todo_len = len;
-	/* -ENOENT is a good thing don't return it if some recalls are needed */
-	if ((status == -ENOENT) && len)
-		status = 0;
-	dprintk("%s: <-- list len %u status %d\n", __func__, len, status);
-	return status;
-}
-
-/* Create a list of clients to send device notifications. */
-int
-create_device_notify_list(struct list_head *todolist,
-			  struct nfsd4_pnfs_cb_dev_list *ndl)
-{
-	int status = 0, i;
-	struct nfs4_client *clp = NULL;
-	struct nfs4_notify_device *cbnd;
-
-	nfs4_lock_state();
-
-	/* Create notify client list */
-	for (i = 0; i < CLIENT_HASH_SIZE; i++)
-		list_for_each_entry(clp, &conf_str_hashtbl[i], cl_strhash) {
-			if (atomic_read(&clp->cl_deviceref) <= 0)
-				continue;
-			cbnd = kmalloc(sizeof(*cbnd), GFP_KERNEL);
-			if (!cbnd) {
-				status = -ENOMEM;
-				goto out;
-			}
-			cbnd->nd_list = ndl;
-			cbnd->nd_client = clp;
-			list_add(&cbnd->nd_perclnt, todolist);
-			atomic_inc(&clp->cl_count);
-		}
-
-out:
-	nfs4_unlock_state();
-	return status;
-}
-
-#endif  /* CONFIG_PNFSD */
fs/nfsd/nfsd_priv.h new file +105 -0
new file mode 100644
index 0000000..814fe8f
@@ -0,0 +1,105 @@
+/*
+ *  nfsd_priv.h
+ *
+ *  nfsd declarations shared by a few .c files
+ *
+ *  Copyright (c) 2002 The Regents of the University of Michigan.
+ *  All rights reserved.
+ *
+ *  Redistribution and use in source and binary forms, with or without
+ *  modification, are permitted provided that the following conditions
+ *  are met:
+ *
+ *  1. Redistributions of source code must retain the above copyright
+ *     notice, this list of conditions and the following disclaimer.
+ *  2. Redistributions in binary form must reproduce the above copyright
+ *     notice, this list of conditions and the following disclaimer in the
+ *     documentation and/or other materials provided with the distribution.
+ *  3. Neither the name of the University nor the names of its
+ *     contributors may be used to endorse or promote products derived
+ *     from this software without specific prior written permission.
+ *
+ *  THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESS OR IMPLIED
+ *  WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF
+ *  MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
+ *  DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
+ *  FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
+ *  CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
+ *  SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR
+ *  BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
+ *  LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
+ *  NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
+ *  SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ *
+ */
+
+#ifndef _NFSD_PRIV_H_
+#define _NFSD_PRIV_H_
+
+/* 
+ * SETCLIENTID state 
+ */
+
+/* Hash tables for nfs4_clientid state */
+#define CLIENT_HASH_BITS                 4
+#define CLIENT_HASH_SIZE                (1 << CLIENT_HASH_BITS)
+#define CLIENT_HASH_MASK                (CLIENT_HASH_SIZE - 1)
+
+#define clientid_hashval(id) \
+	((id) & CLIENT_HASH_MASK)
+#define clientstr_hashval(name) \
+	(opaque_hashval((name), 8) & CLIENT_HASH_MASK)
+
+extern struct list_head	conf_str_hashtbl[CLIENT_HASH_SIZE];
+
+#if defined(CONFIG_PNFSD)
+
+#include <linux/nfsd/nfsd4_pnfs.h>
+
+#if defined(CONFIG_PNFSD_LOCAL_EXPORT)
+extern struct sockaddr pnfsd_lexp_addr;
+extern size_t pnfs_lexp_addr_len;
+
+void pnfsd_lexp_init(struct inode *inode);
+#endif /* CONFIG_PNFSD_LOCAL_EXPORT */
+
+extern int nfsd4_init_pnfs_slabs(void);
+extern void nfsd4_free_pnfs_slabs(void);
+extern void release_pnfs_ds_dev_list(struct nfs4_stateid *stp);
+extern void pnfs_expire_client(struct nfs4_client *clp);
+
+extern void release_pnfs_ds_dev_list(struct nfs4_stateid *stp);
+extern void nfs4_pnfs_state_init(void);
+extern int put_layoutrecall(struct nfs4_layoutrecall *);
+extern void nomatching_layout(struct nfs4_layoutrecall *);
+extern void *layoutrecall_done(struct nfs4_layoutrecall *);
+extern int nfsd4_cb_layout(struct nfs4_layoutrecall *lp);
+extern void nfsd4_free_slab(struct kmem_cache **slab);
+extern struct nfs4_file * find_file(struct inode *ino);
+extern struct nfs4_file * alloc_init_file(struct inode *ino, struct svc_fh *current_fh);
+extern void put_nfs4_file(struct nfs4_file *fi);
+extern void get_nfs4_file(struct nfs4_file *fi);
+extern struct nfs4_client * find_confirmed_client(clientid_t *clid);
+extern void nfs4_bug_on_unlocked_state(void);
+extern void nfs4_ds_get_verifier(stateid_t *stateid,
+		struct super_block *sb, u32 *p);
+extern int nfs4_preprocess_pnfs_ds_stateid(struct svc_fh *, stateid_t *);
+extern void nfs4_pnfs_state_shutdown(void);
+extern int lo_recall_per_client(struct nfs4_client *clp,
+		struct nfsd4_pnfs_cb_layout *cbl, struct nfs4_file *lrfile,
+		struct list_head *todolist);
+extern int create_layout_recall_list(struct list_head *todolist,
+		unsigned *todo_len, struct nfsd4_pnfs_cb_layout *cbl,
+		struct nfs4_file *lrfile);
+extern int nfsd4_cb_notify_device(struct nfs4_notify_device *cbnd);
+extern void pnfs_clear_device_notify(struct nfs4_client *clp);
+extern int create_device_notify_list(struct list_head *todolist,
+		struct nfsd4_pnfs_cb_dev_list *ndl);
+#else /* CONFIG_PNFSD */
+static inline void nfsd4_free_pnfs_slabs(void) {}
+static inline int nfsd4_init_pnfs_slabs(void) { return 0; }
+static inline void release_pnfs_ds_dev_list(struct nfs4_stateid *stp) {}
+static inline void pnfs_expire_client(struct nfs4_client *clp) {}
+#endif /* CONFIG_PNFSD */
+
+#endif /* _NFSD_PRIV_H_ */
include/linux/nfsd/nfsd4_pnfs.h +50 -11
index 410c459..70ca179 100644
@@ -36,10 +36,9 @@
 #ifndef _LINUX_NFSD_NFSD4_PNFS_H
 #define _LINUX_NFSD_NFSD4_PNFS_H
 
+#include <linux/nfsd/state.h>
 #include <linux/nfs_xdr.h>
 #include <linux/exportfs.h>
-#include <linux/exp_xdr.h>
-#include <linux/nfsd/nfsfh.h>
 
 typedef struct {
 	uint64_t	pnfs_fsid;	/* fsid */
@@ -83,6 +82,55 @@ struct nfsd4_layout_seg {
 	u64	length;
 };
 
+/* outstanding layout stateid */
+struct nfs4_layout_state {
+	struct list_head	ls_perfile;
+	struct list_head	ls_layouts; /* list of nfs4_layouts */
+	struct kref		ls_ref;
+	struct nfs4_client	*ls_client;
+	struct nfs4_file	*ls_file;
+	stateid_t		ls_stateid;
+};
+
+/* outstanding layout */
+struct nfs4_layout {
+	struct list_head		lo_perfile;	/* hash by f_id */
+	struct list_head		lo_perclnt;	/* hash by clientid */
+	struct list_head		lo_perstate;
+	struct nfs4_file		*lo_file;	/* backpointer */
+	struct nfs4_client		*lo_client;
+	struct nfs4_layout_state	*lo_state;
+	struct nfsd4_layout_seg 	lo_seg;
+};
+
+struct nfsd4_pnfs_cb_layout {
+	u32			cbl_recall_type;	/* request */
+	struct nfsd4_layout_seg cbl_seg;		/* request */
+	u32			cbl_layoutchanged;	/* request */
+	stateid_t		cbl_sid;		/* request */
+	struct nfs4_fsid	cbl_fsid;
+	void			*cbl_cookie;		/* fs private */
+};
+
+/* layoutrecall request (from exported filesystem) */
+struct nfs4_layoutrecall {
+	struct kref			clr_ref;
+	struct nfsd4_pnfs_cb_layout	cb;	/* request */
+	struct list_head		clr_perclnt; /* on cl_layoutrecalls */
+	struct nfs4_client	       *clr_client;
+	struct nfs4_file	       *clr_file;
+	struct timespec			clr_time;	/* last activity */
+	struct super_block 		*clr_sb; /* We might not have a file */
+	struct nfs4_layoutrecall	*parent; /* The initiating recall */
+};
+
+/* notify device request (from exported filesystem) */
+struct nfs4_notify_device {
+	struct nfsd4_pnfs_cb_dev_list  *nd_list;
+	struct nfs4_client	       *nd_client;
+	struct list_head	        nd_perclnt;
+};
+
 /* Used by layout_get to encode layout (loc_body var in spec)
  * Args:
  * minlength - min number of accessible bytes given by layout
@@ -168,15 +216,6 @@ struct nfsd4_pnfs_open {
 	int			op_truncate;
 };
 
-struct nfsd4_pnfs_cb_layout {
-	u32			cbl_recall_type;	/* request */
-	struct nfsd4_layout_seg cbl_seg;		/* request */
-	u32			cbl_layoutchanged;	/* request */
-	stateid_t		cbl_sid;		/* request */
-	struct nfs4_fsid	cbl_fsid;
-	void			*cbl_cookie;		/* fs private */
-};
-
 struct nfsd4_pnfs_cb_dev_item {
 	u32			cbd_notify_type;	/* request */
 	u32			cbd_layout_type;	/* request */
include/linux/nfsd/pnfsd.h +0 -8
index d1d7395..6aa3a87 100644
@@ -96,14 +96,6 @@ int nfs4_pnfs_return_layout(struct super_block *, struct svc_fh *,
 					struct nfsd4_pnfs_layoutreturn *);
 void pnfs_set_device_notify(clientid_t *clid, unsigned int types);
 void nfs4_pnfs_state_shutdown(void);
-
-#if defined(CONFIG_PNFSD_LOCAL_EXPORT)
-extern struct sockaddr pnfsd_lexp_addr;
-extern size_t pnfs_lexp_addr_len;
-
-void pnfsd_lexp_init(struct inode *inode);
-#endif /* CONFIG_PNFSD_LOCAL_EXPORT */
-
 #endif /* CONFIG_PNFSD */
 
 #endif /* LINUX_NFSD_PNFSD_H */
include/linux/nfsd/state.h +1 -84
index ffd5a41..7367696 100644
@@ -40,6 +40,7 @@
 #include <linux/list.h>
 #include <linux/kref.h>
 #include <linux/sunrpc/clnt.h>
+#include <linux/nfsd/nfsfh.h>
 
 typedef struct {
 	u32             cl_boot;
@@ -245,52 +246,6 @@ struct nfs4_fsid {
 	u64	minor;
 };
 
-#if defined(CONFIG_PNFSD)
-
-#include <linux/nfsd/nfsd4_pnfs.h>
-
-/* outstanding layout stateid */
-struct nfs4_layout_state {
-	struct list_head	ls_perfile;
-	struct list_head	ls_layouts; /* list of nfs4_layouts */
-	struct kref		ls_ref;
-	struct nfs4_client	*ls_client;
-	struct nfs4_file	*ls_file;
-	stateid_t		ls_stateid;
-};
-
-/* outstanding layout */
-struct nfs4_layout {
-	struct list_head		lo_perfile;	/* hash by f_id */
-	struct list_head		lo_perclnt;	/* hash by clientid */
-	struct list_head		lo_perstate;
-	struct nfs4_file		*lo_file;	/* backpointer */
-	struct nfs4_client		*lo_client;
-	struct nfs4_layout_state	*lo_state;
-	struct nfsd4_layout_seg 	lo_seg;
-};
-
-/* layoutrecall request (from exported filesystem) */
-struct nfs4_layoutrecall {
-	struct kref			clr_ref;
-	struct nfsd4_pnfs_cb_layout	cb;	/* request */
-	struct list_head		clr_perclnt; /* on cl_layoutrecalls */
-	struct nfs4_client	       *clr_client;
-	struct nfs4_file	       *clr_file;
-	struct timespec			clr_time;	/* last activity */
-	struct super_block 		*clr_sb; /* We might not have a file */
-	struct nfs4_layoutrecall	*parent; /* The initiating recall */
-};
-
-/* notify device request (from exported filesystem) */
-struct nfs4_notify_device {
-	struct nfsd4_pnfs_cb_dev_list  *nd_list;
-	struct nfs4_client	       *nd_client;
-	struct list_head	        nd_perclnt;
-};
-
-#endif /* CONFIG_PNFSD */
-
 /* struct nfs4_client_reset
  * one per old client. Populates reset_str_hashtbl. Filled from conf_id_hashtbl
  * upon lease reset, or from upcall to state_daemon (to read in state
@@ -479,44 +434,6 @@ extern struct nfs4_delegation * find_delegation_stateid(struct inode *ino,
 		stateid_t *stid);
 extern __be32 nfs4_check_stateid(stateid_t *stateid);
 extern void expire_client_lock(struct nfs4_client *clp);
-#if defined(CONFIG_PNFSD)
-extern void release_pnfs_ds_dev_list(struct nfs4_stateid *stp);
-extern void nfs4_pnfs_state_init(void);
-extern int put_layoutrecall(struct nfs4_layoutrecall *);
-extern void nomatching_layout(struct nfs4_layoutrecall *);
-extern void *layoutrecall_done(struct nfs4_layoutrecall *);
-extern int nfsd4_cb_layout(struct nfs4_layoutrecall *lp);
-extern void nfsd4_free_pnfs_slabs(void);
-extern void nfsd4_free_slab(struct kmem_cache **slab);
-extern int nfsd4_init_pnfs_slabs(void);
-extern struct nfs4_file * find_file(struct inode *ino);
-extern struct nfs4_file * alloc_init_file(struct inode *ino, struct svc_fh *current_fh);
-extern void put_nfs4_file(struct nfs4_file *fi);
-extern void get_nfs4_file(struct nfs4_file *fi);
-extern struct nfs4_client * find_confirmed_client(clientid_t *clid);
-extern void nfs4_bug_on_unlocked_state(void);
-extern void release_pnfs_ds_dev_list(struct nfs4_stateid *stp);
-extern void nfs4_ds_get_verifier(stateid_t *stateid,
-		struct super_block *sb, u32 *p);
-extern int nfs4_preprocess_pnfs_ds_stateid(struct svc_fh *, stateid_t *);
-extern void nfs4_pnfs_state_shutdown(void);
-extern int lo_recall_per_client(struct nfs4_client *clp,
-		struct nfsd4_pnfs_cb_layout *cbl, struct nfs4_file *lrfile,
-		struct list_head *todolist);
-extern int create_layout_recall_list(struct list_head *todolist,
-		unsigned *todo_len, struct nfsd4_pnfs_cb_layout *cbl,
-		struct nfs4_file *lrfile);
-extern int nfsd4_cb_notify_device(struct nfs4_notify_device *cbnd);
-extern void pnfs_clear_device_notify(struct nfs4_client *clp);
-extern int create_device_notify_list(struct list_head *todolist,
-		struct nfsd4_pnfs_cb_dev_list *ndl);
-extern void pnfs_expire_client(struct nfs4_client *clp);
-#else /* CONFIG_PNFSD */
-static inline void nfsd4_free_pnfs_slabs(void) {}
-static inline int nfsd4_init_pnfs_slabs(void) { return 0; }
-static inline void release_pnfs_ds_dev_list(struct nfs4_stateid *stp) {}
-static inline void pnfs_expire_client(struct nfs4_client *clp) {}
-#endif /* CONFIG_PNFSD */
 
 static inline void
 nfs4_put_stateowner(struct nfs4_stateowner *so)
#12 lore raw

Re: [PATCH 2/5] nfsd: Fix independence of a few nfsd related headers

In reply to Boaz Harrosh
Quoted 15 lines · Boaz Harrosh
Quoted 8 lines · Trond Myklebust

No. What I'm saying is that this doesn't have to be an absolute rule. The Kernel style guide assumes that everything in 'include/*' is going to be exported all around the kernel.
The problem is that we put a lot of stuff which is private to fs/nfs and fs/nfsd in there. Those header files do not have to absolutely follow the style guide rule, 'cos we know what is being included before and after them...

I'm not sure I understand
You are saying that the patches are very good, but only the rule I stated originally could be relaxed a little with private headers where we might get lazy, if the effects are very local?

Quoted 7 lines

Well, that's not a problem then, right? just that I can relax a bit if I want.

But I disagree: see 3, 4, 5 above and that last patch I submitted. That patch is only the beginning. 85% of all source files in nfs/nfsd could receive the same love. I only done these I touched. Code tends to stay much-much longer then we spend time on it. Better get it in shape the first time.

I'm assuming Trond's objection is just to the patch changelog (specifically, to the statement that any header "should be compilation independent"), not to these specific changes.

--b.

#13 lore raw

Re: [PATCH 2/5] nfsd: Fix independence of a few nfsd related headers

Quoted 24 lines · J. Bruce Fields
Quoted 15 lines · Boaz Harrosh
Quoted 8 lines · Trond Myklebust

No. What I'm saying is that this doesn't have to be an absolute rule. The Kernel style guide assumes that everything in 'include/*' is going to be exported all around the kernel.
The problem is that we put a lot of stuff which is private to fs/nfs and fs/nfsd in there. Those header files do not have to absolutely follow the style guide rule, 'cos we know what is being included before and after them...

I'm not sure I understand
You are saying that the patches are very good, but only the rule I stated originally could be relaxed a little with private headers where we might get lazy, if the effects are very local?

Quoted 7 lines

Well, that's not a problem then, right? just that I can relax a bit if I want.

But I disagree: see 3, 4, 5 above and that last patch I submitted. That patch is only the beginning. 85% of all source files in nfs/nfsd could receive the same love. I only done these I touched. Code tends to stay much-much longer then we spend time on it. Better get it in shape the first time.

Quoted 5 lines

I'm assuming Trond's objection is just to the patch changelog (specifically, to the statement that any header "should be compilation independent"), not to these specific changes.

--b.

Speaking of which, Bruce I have a question.

There are a few files in include/linux/ that define xdr definitions these are used by exportfs nfs and nfsd. Some of it gets exposed to filesystems. With the pNFS tree we are adding lots more of these, specifically I'm now to move the pnfs_osd_xdr stuff as well, and blocks.

Could I open up a new include/linux/exportfs/ folder and put there any thing xdr and exportfs related?

What should be the scope of the move, should any include/linux/ common files used both by nfs & nfsd be moved there?

Thanks
Boaz

#14 lore raw

Re: [PATCH 2/5] nfsd: Fix independence of a few nfsd related headers

Quoted 29 lines · Boaz Harrosh
Quoted 28 lines · J. Bruce Fields
Quoted 15 lines · Boaz Harrosh
Quoted 8 lines · Trond Myklebust

No. What I'm saying is that this doesn't have to be an absolute rule. The Kernel style guide assumes that everything in 'include/*' is going to be exported all around the kernel.
The problem is that we put a lot of stuff which is private to fs/nfs and fs/nfsd in there. Those header files do not have to absolutely follow the style guide rule, 'cos we know what is being included before and after them...

I'm not sure I understand
You are saying that the patches are very good, but only the rule I stated originally could be relaxed a little with private headers where we might get lazy, if the effects are very local?

Quoted 7 lines

Well, that's not a problem then, right? just that I can relax a bit if I want.

But I disagree: see 3, 4, 5 above and that last patch I submitted. That patch is only the beginning. 85% of all source files in nfs/nfsd could receive the same love. I only done these I touched. Code tends to stay much-much longer then we spend time on it. Better get it in shape the first time.

I'm assuming Trond's objection is just to the patch changelog (specifically, to the statement that any header "should be compilation independent"), not to these specific changes.

Quoted 7 lines
Quoted 1 line

--b.

Speaking of which, Bruce I have a question.

There are a few files in include/linux/ that define xdr definitions these are used by exportfs nfs and nfsd. Some of it gets exposed to filesystems.

Which specifically?

Quoted 5 lines

With the pNFS tree we are adding lots more of these, specifically I'm now to move the pnfs_osd_xdr stuff as well, and blocks.

Could I open up a new include/linux/exportfs/ folder and put there any thing xdr and exportfs related?

Well, if it's just a few things, maybe include/linux/exportfs.h?

But, sure, makes sense if there's includes that are specific to pNFS and that wouldn't be needed by any filesystem that included exportfs.h.

Quoted 2 lines

What should be the scope of the move, should any include/linux/ common files used both by nfs & nfsd be moved there?

I don't see why we'd want to do that, but maybe I'm missing something.

--b.

#15 lore raw

Re: [PATCH 2/5] nfsd: Fix independence of a few nfsd related headers

By the way, along the lines of your .h moves, while looking at steved's export stuff I was reminded again of how much in include/linux/nfsd is really internal to nfsd, so I'm thinking of committing this.

--b.

Quoted 1 line

From 0142013a3f49ca296f8880829c67f2afd85b4bfe Mon Sep 17 00:00:00 2001

From: J. Bruce Fields <[email protected]>
Date: Wed, 4 Nov 2009 18:12:35 -0500
Subject: [PATCH] nfsd: make nfsd/vfs.h for common includes

None of this stuff is used outside nfsd, so move it out of the common linux include directory.

Actually, probably none of the stuff in include/linux/nfsd/nfsd.h really belongs there, so later we may remove that file entirely.

---
 fs/nfsd/lockd.c           |    1 +
 fs/nfsd/nfs2acl.c         |    1 +
 fs/nfsd/nfs3acl.c         |    1 +
 fs/nfsd/nfs3proc.c        |    1 +
 fs/nfsd/nfs4proc.c        |    1 +
 fs/nfsd/nfs4recover.c     |    1 +
 fs/nfsd/nfs4state.c       |    1 +
 fs/nfsd/nfs4xdr.c         |    1 +
 fs/nfsd/nfsfh.c           |    1 +
 fs/nfsd/nfsproc.c         |    1 +
 fs/nfsd/nfssvc.c          |    1 +
 fs/nfsd/vfs.c             |    1 +
 fs/nfsd/vfs.h             |   98 +++++++++++++++++++++++++++++++++++++++++++++
 include/linux/nfsd/nfsd.h |   87 +---------------------------------------
 14 files changed, 111 insertions(+), 86 deletions(-)
 create mode 100644 fs/nfsd/vfs.h
fs/nfsd/lockd.c +1 -0
index b2786a5..812bc64 100644
@@ -16,6 +16,7 @@
 #include <linux/sunrpc/svc.h>
 #include <linux/nfsd/nfsd.h>
 #include <linux/lockd/bind.h>
+#include "vfs.h"
 
 #define NFSDDBG_FACILITY		NFSDDBG_LOCKD
 
fs/nfsd/nfs2acl.c +1 -0
index 4e3219e..5e8ef72 100644
@@ -14,6 +14,7 @@
 #include <linux/nfsd/xdr3.h>
 #include <linux/posix_acl.h>
 #include <linux/nfsacl.h>
+#include "vfs.h"
 
 #define NFSDDBG_FACILITY		NFSDDBG_PROC
 #define RETURN_STATUS(st)	{ resp->status = (st); return (st); }
fs/nfsd/nfs3acl.c +1 -0
index 9981dbb..354bf0f 100644
@@ -13,6 +13,7 @@
 #include <linux/nfsd/xdr3.h>
 #include <linux/posix_acl.h>
 #include <linux/nfsacl.h>
+#include "vfs.h"
 
 #define RETURN_STATUS(st)	{ resp->status = (st); return (st); }
 
fs/nfsd/nfs3proc.c +1 -0
index a713c41..1a259d3 100644
@@ -25,6 +25,7 @@
 #include <linux/nfsd/cache.h>
 #include <linux/nfsd/xdr3.h>
 #include <linux/nfs3.h>
+#include "vfs.h"
 
 #define NFSDDBG_FACILITY		NFSDDBG_PROC
 
fs/nfsd/nfs4proc.c +1 -0
index bebc0c2..60a93cd 100644
@@ -48,6 +48,7 @@
 #include <linux/nfsd/xdr4.h>
 #include <linux/nfs4_acl.h>
 #include <linux/sunrpc/gss_api.h>
+#include "vfs.h"
 
 #define NFSDDBG_FACILITY		NFSDDBG_PROC
 
fs/nfsd/nfs4recover.c +1 -0
index b534840..c7a6b24 100644
@@ -47,6 +47,7 @@
 #include <linux/crypto.h>
 #include <linux/sched.h>
 #include <linux/mount.h>
+#include "vfs.h"
 
 #define NFSDDBG_FACILITY                NFSDDBG_PROC
 
fs/nfsd/nfs4state.c +1 -0
index fcb9817..d3fad54 100644
@@ -56,6 +56,7 @@
 #include <linux/module.h>
 #include <linux/sunrpc/svcauth_gss.h>
 #include <linux/sunrpc/clnt.h>
+#include "vfs.h"
 
 #define NFSDDBG_FACILITY                NFSDDBG_PROC
 
fs/nfsd/nfs4xdr.c +1 -0
index 0fbd50c..db0fc55 100644
@@ -57,6 +57,7 @@
 #include <linux/nfs4_acl.h>
 #include <linux/sunrpc/gss_api.h>
 #include <linux/sunrpc/svcauth_gss.h>
+#include "vfs.h"
 
 #define NFSDDBG_FACILITY		NFSDDBG_XDR
 
fs/nfsd/nfsfh.c +1 -0
index 01965b2..d0d8a21 100644
@@ -22,6 +22,7 @@
 #include <linux/sunrpc/svc.h>
 #include <linux/sunrpc/svcauth_gss.h>
 #include <linux/nfsd/nfsd.h>
+#include "vfs.h"
 #include "auth.h"
 
 #define NFSDDBG_FACILITY		NFSDDBG_FH
fs/nfsd/nfsproc.c +1 -0
index c5393d1..6c967e1 100644
@@ -24,6 +24,7 @@
 #include <linux/nfsd/nfsd.h>
 #include <linux/nfsd/cache.h>
 #include <linux/nfsd/xdr.h>
+#include "vfs.h"
 
 typedef struct svc_rqst	svc_rqst;
 typedef struct svc_buf	svc_buf;
fs/nfsd/nfssvc.c +1 -0
index 67ea83e..2944b31 100644
@@ -35,6 +35,7 @@
 #include <linux/lockd/bind.h>
 #include <linux/nfsacl.h>
 #include <linux/seq_file.h>
+#include "vfs.h"
 
 #define NFSDDBG_FACILITY	NFSDDBG_SVC
 
fs/nfsd/vfs.c +1 -0
index 6385739..a7038ed 100644
@@ -56,6 +56,7 @@
 #endif /* CONFIG_NFSD_V4 */
 #include <linux/jhash.h>
 #include <linux/ima.h>
+#include "vfs.h"
 
 #include <asm/uaccess.h>
 
fs/nfsd/vfs.h new file +98 -0

98 lines not on this page · raw message

include/linux/nfsd/nfsd.h +1 -86

108 lines not on this page · raw message

Signature · 1 line
1.6.3.3
#16 lore raw

Re: [PATCH 2/5] nfsd: Fix independence of a few nfsd related headers

Quoted 3 lines · J. Bruce Fields

By the way, along the lines of your .h moves, while looking at steved's export stuff I was reminded again of how much in include/linux/nfsd is really internal to nfsd, so I'm thinking of committing this.

Right. Most of the NFS client include files should really be moved into fs/nfs rather than into a new directory in include/linux.

Signature · 6 lines
Trond Myklebust
Linux NFS client maintainer

NetApp
[email protected]
www.netapp.com
#18 lore raw

Re: [PATCH 2/5] nfsd: Fix independence of a few nfsd related headers

In reply to a message not in the archive
Quoted 7 lines · Trond Myklebust
Quoted 3 lines · J. Bruce Fields

By the way, along the lines of your .h moves, while looking at steved's export stuff I was reminded again of how much in include/linux/nfsd is really internal to nfsd, so I'm thinking of committing this.

Right. Most of the NFS client include files should really be moved into fs/nfs rather than into a new directory in include/linux.

On the server side my greatest source of confusion here is some includes of nfsd headers in arch-specific code. Was it needed for some sort of compatibility glue for the nfsservctl system call, maybe? I've never figured it out.

--b.

#19 lore raw

Re: [PATCH 2/5] nfsd: Fix independence of a few nfsd related headers

Quoted 6 lines · J. Bruce Fields
Quoted 5 lines · Trond Myklebust
Quoted 3 lines · J. Bruce Fields

By the way, along the lines of your .h moves, while looking at steved's export stuff I was reminded again of how much in include/linux/nfsd is really internal to nfsd, so I'm thinking of committing this.

Quoted 7 lines
Quoted 2 lines

Right. Most of the NFS client include files should really be moved into fs/nfs rather than into a new directory in include/linux.

On the server side my greatest source of confusion here is some includes of nfsd headers in arch-specific code. Was it needed for some sort of compatibility glue for the nfsservctl system call, maybe? I've never figured it out.

Possibly for the 32-bit emulation code on 64-bit systems?

Signature · 6 lines
Trond Myklebust
Linux NFS client maintainer

NetApp
[email protected]
www.netapp.com
#20 lore raw

Re: [PATCH 2/5] nfsd: Fix independence of a few nfsd related headers

In reply to J. Bruce Fields
Quoted 7 lines · J. Bruce Fields

On Thu, Oct 22, 2009 at 05:59:33PM +0200, Boaz Harrosh wrote:

I'm assuming Trond's objection is just to the patch changelog (specifically, to the statement that any header "should be compilation independent"), not to these specific changes.

--b.

Ping

Bruce? Trond? whatsup?

Can Benny put these patches in his tree? He said he would be happy to hold them for a while, but only if they will be eventually accepted into the tree as a pnfs pre-requisite. Please ACK on these patches?

I have to make all these put-the-includes-back patches to just make the tree compile.

Boaz

#21 lore raw

Re: [PATCH 2/5] nfsd: Fix independence of a few nfsd related headers

Quoted 11 lines · Boaz Harrosh
Quoted 7 lines · J. Bruce Fields

On Thu, Oct 22, 2009 at 05:59:33PM +0200, Boaz Harrosh wrote:

I'm assuming Trond's objection is just to the patch changelog (specifically, to the statement that any header "should be compilation independent"), not to these specific changes.

--b.

Ping

Quoted 8 lines

Bruce? Trond? whatsup?

Can Benny put these patches in his tree? He said he would be happy to hold them for a while, but only if they will be eventually accepted into the tree as a pnfs pre-requisite. Please ACK on these patches?

I have to make all these put-the-includes-back patches to just make the tree compile.

They're fine by me.

(Can't speak for Trond, but maybe his initial objection would be met just editing the changelog to replace the absolute "Any header should be compilation independent" by the particular advantages you saw in this case.)

--b.

Quoted 3 lines

Boaz

#22 lore raw

Re: [PATCH 2/5] nfsd: Fix independence of a few nfsd related headers

Quoted 21 lines · J. Bruce Fields
Quoted 11 lines · Boaz Harrosh
Quoted 7 lines · J. Bruce Fields

On Thu, Oct 22, 2009 at 05:59:33PM +0200, Boaz Harrosh wrote:

I'm assuming Trond's objection is just to the patch changelog (specifically, to the statement that any header "should be compilation independent"), not to these specific changes.

--b.

Ping

Quoted 8 lines

Bruce? Trond? whatsup?

Can Benny put these patches in his tree? He said he would be happy to hold them for a while, but only if they will be eventually accepted into the tree as a pnfs pre-requisite. Please ACK on these patches?

I have to make all these put-the-includes-back patches to just make the tree compile.

Quoted 7 lines

They're fine by me.

(Can't speak for Trond, but maybe his initial objection would be met just editing the changelog to replace the absolute "Any header should be compilation independent" by the particular advantages you saw in this case.)

I don't see why. Please advise?

"should be compilation independent", from what I understand of the English language, is suggestive and advisory only. Now, if I was using "must" or "shall" like the standard do then that would mean a mandatory directive. But I'm only saying "should" which is like saying: "I suggest", or "it is recommended". Am I misunderstanding the language?

Any way the commit log is just my saying so, my sign-off it's not the word of Linux-god, is it?

Quoted 5 lines

--b.

Quoted 3 lines

Boaz

Thanks
Boaz

#23 lore raw

Re: [PATCH 2/5] nfsd: Fix independence of a few nfsd related headers

Quoted 32 lines · Boaz Harrosh
Quoted 21 lines · J. Bruce Fields
Quoted 11 lines · Boaz Harrosh
Quoted 7 lines · J. Bruce Fields

On Thu, Oct 22, 2009 at 05:59:33PM +0200, Boaz Harrosh wrote:

I'm assuming Trond's objection is just to the patch changelog (specifically, to the statement that any header "should be compilation independent"), not to these specific changes.

--b.

Ping

Quoted 8 lines

Bruce? Trond? whatsup?

Can Benny put these patches in his tree? He said he would be happy to hold them for a while, but only if they will be eventually accepted into the tree as a pnfs pre-requisite. Please ACK on these patches?

I have to make all these put-the-includes-back patches to just make the tree compile.

Quoted 7 lines

They're fine by me.

(Can't speak for Trond, but maybe his initial objection would be met just editing the changelog to replace the absolute "Any header should be compilation independent" by the particular advantages you saw in this case.)

I don't see why. Please advise?

Quoted 8 lines

"should be compilation independent", from what I understand of the English language, is suggestive and advisory only. Now, if I was using "must" or "shall" like the standard do then that would mean a mandatory directive. But I'm only saying "should" which is like saying: "I suggest", or "it is recommended". Am I misunderstanding the language?

Any way the commit log is just my saying so, my sign-off it's not the word of Linux-god, is it?

I really don't care. Do what you think best.

--b.

#24 lore raw

Re: [pnfs] [PATCH 2/5] nfsd: Fix independence of a few nfsd related headers

In reply to Boaz Harrosh
Quoted 9 lines · Boaz Harrosh

An header should be compilation independent, .i.e pull in any header who's declarations are directly used by this header. And not let users re-include all it's dependencies all over again.

[At the end of the day what's the use of a header if it does
not have more then one user?]

Trond do I have an ACK on this patch.
If not, then what should be changed to get it accepted?

Quoted 2 lines
---
 include/linux/nfs_xdr.h   |    1 +

This header is used exclusively by fs/nfs/... files and could just be moved there. The include must be fixed as below though.

Quoted 1 line
 include/linux/nfsacl.h    |    1 +

This file is used mixed between fs/nfs && fs/nfsd

Quoted 1 line
 include/linux/posix_acl.h |    1 +

Used by nfsd and filesystems

Quoted 38 lines
 3 files changed, 3 insertions(+), 0 deletions(-)
include/linux/nfs_xdr.h +1 -0
index 2848a26..c316ca8 100644
@@ -2,6 +2,7 @@
 #define _LINUX_NFS_XDR_H
 
 #include <linux/nfsacl.h>
+#include <linux/nfs3.h>
 
 /*
  * To change the maximum rsize and wsize supported by the NFS client, adjust
include/linux/nfsacl.h +1 -0
index 43011b6..f321b57 100644
@@ -29,6 +29,7 @@
 #ifdef __KERNEL__
 
 #include <linux/posix_acl.h>
+#include <linux/sunrpc/xdr.h>
 
 /* Maximum number of ACL entries over NFS */
 #define NFS_ACL_MAX_ENTRIES	1024
include/linux/posix_acl.h +1 -0
index 065a365..0dcf674 100644
@@ -9,6 +9,7 @@
 #define __LINUX_POSIX_ACL_H
 
 #include <linux/slab.h>
+#include <linux/fs.h>
 
 #define ACL_UNDEFINED_ID	(-1)
 
#25 lore raw

Re: [pnfs] [PATCH 2/5] nfsd: Fix independence of a few nfsd related headers

Quoted 61 lines · Boaz Harrosh
Quoted 9 lines · Boaz Harrosh

An header should be compilation independent, .i.e pull in any header who's declarations are directly used by this header. And not let users re-include all it's dependencies all over again.

[At the end of the day what's the use of a header if it does
not have more then one user?]

Trond do I have an ACK on this patch.
If not, then what should be changed to get it accepted?

Quoted 2 lines
---
 include/linux/nfs_xdr.h   |    1 +

This header is used exclusively by fs/nfs/... files and could just be moved there. The include must be fixed as below though.

Quoted 1 line
 include/linux/nfsacl.h    |    1 +

This file is used mixed between fs/nfs && fs/nfsd

Quoted 1 line
 include/linux/posix_acl.h |    1 +

Used by nfsd and filesystems

Quoted 33 lines
 3 files changed, 3 insertions(+), 0 deletions(-)
include/linux/nfs_xdr.h +1 -0
index 2848a26..c316ca8 100644
@@ -2,6 +2,7 @@
 #define _LINUX_NFS_XDR_H
 
 #include <linux/nfsacl.h>
+#include <linux/nfs3.h>
 
 /*
  * To change the maximum rsize and wsize supported by the NFS client, adjust
include/linux/nfsacl.h +1 -0
index 43011b6..f321b57 100644
@@ -29,6 +29,7 @@
 #ifdef __KERNEL__
 
 #include <linux/posix_acl.h>
+#include <linux/sunrpc/xdr.h>
 
 /* Maximum number of ACL entries over NFS */
 #define NFS_ACL_MAX_ENTRIES	1024
include/linux/posix_acl.h +0 -0
index 065a365..0dcf674 100644
@@ -9,6 +9,7 @@
 #define __LINUX_POSIX_ACL_H
 
Quoted 2 lines

include/linux/posix_acl.h line 12

Quoted 2 lines
include/linux/posix_acl.h +1 -0
 #include <linux/slab.h>
+#include <linux/fs.h>

NACK to this. Pretty much _all_ filesystems already include linux/fs.h somewhere in their include chains. There should be no need to include it in posix_acl.h too.

Quoted 4 lines

include/linux/posix_acl.h line 15

Quoted 3 lines
include/linux/posix_acl.h +0 -0
 #define ACL_UNDEFINED_ID	(-1)
 

So, what is the motivation for all this? We have no dependency problems here today. What is changing in the pNFS tree that makes this so necessary?

Trond

Signature · 6 lines
Trond Myklebust
Linux NFS client maintainer

NetApp
[email protected]
www.netapp.com
#27 lore raw

Re: [pnfs] [PATCH 2/5] nfsd: Fix independence of a few nfsd related headers

In reply to a message not in the archive
Quoted 62 lines · Trond Myklebust
Quoted 61 lines · Boaz Harrosh
Quoted 9 lines · Boaz Harrosh

An header should be compilation independent, .i.e pull in any header who's declarations are directly used by this header. And not let users re-include all it's dependencies all over again.

[At the end of the day what's the use of a header if it does
not have more then one user?]

Trond do I have an ACK on this patch.
If not, then what should be changed to get it accepted?

Quoted 2 lines
---
 include/linux/nfs_xdr.h   |    1 +

This header is used exclusively by fs/nfs/... files and could just be moved there. The include must be fixed as below though.

Quoted 1 line
 include/linux/nfsacl.h    |    1 +

This file is used mixed between fs/nfs && fs/nfsd

Quoted 1 line
 include/linux/posix_acl.h |    1 +

Used by nfsd and filesystems

Quoted 33 lines
 3 files changed, 3 insertions(+), 0 deletions(-)
include/linux/nfs_xdr.h +1 -0
index 2848a26..c316ca8 100644
@@ -2,6 +2,7 @@
 #define _LINUX_NFS_XDR_H
 
 #include <linux/nfsacl.h>
+#include <linux/nfs3.h>
 
 /*
  * To change the maximum rsize and wsize supported by the NFS client, adjust
include/linux/nfsacl.h +1 -0
index 43011b6..f321b57 100644
@@ -29,6 +29,7 @@
 #ifdef __KERNEL__
 
 #include <linux/posix_acl.h>
+#include <linux/sunrpc/xdr.h>
 
 /* Maximum number of ACL entries over NFS */
 #define NFS_ACL_MAX_ENTRIES	1024
include/linux/posix_acl.h +0 -0
index 065a365..0dcf674 100644
@@ -9,6 +9,7 @@
 #define __LINUX_POSIX_ACL_H
 
Quoted 7 lines
Quoted 2 lines

include/linux/posix_acl.h line 12

Quoted 2 lines
include/linux/posix_acl.h +1 -0
 #include <linux/slab.h>
+#include <linux/fs.h>

NACK to this. Pretty much _all_ filesystems already include linux/fs.h somewhere in their include chains. There should be no need to include it in posix_acl.h too.

from posix_acl.h:
extern int posix_acl_permission(struct inode *, const struct posix_acl *, int);

stuct inode is defined in fs.h. hence the direct dependency.

Again a double inclusion is not a bad thing, it costs absolutely *nothing*.

A miss-inclusion on the other hand is a bad thing. it causes a miss-compilation.

It does not matter that filesystems include fs.h or not. What matters is that they now have to do this headers ordering magic. One places code compiles fine, another place it does not. When it does not, people *never* analyze the missing dependency what they do is copy-paste an include list from another file that works.

Proof of the matter the last patches in this patchset.

Quoted 9 lines
Quoted 4 lines

include/linux/posix_acl.h line 15

Quoted 3 lines
include/linux/posix_acl.h +0 -0
 #define ACL_UNDEFINED_ID	(-1)
 

So, what is the motivation for all this? We have no dependency problems here today. What is changing in the pNFS tree that makes this so necessary?

We do have a dependency problem today!! look at the patchset. It is a grate cleanup and improvement of code today. And a much smoother ride for the future.

What changed is that all this code was touched today. I have not done the cleanup for any files not touched by pnfsd. Though I could and should, because they will greatly improve just like these I did touch. Should I go head and do the reset of the files?

And please note that this particular file is an NFSD and vfs related file only, it has nothing to do with nfs.

Quoted 2 lines

Trond

Boaz

#28 lore raw

Re: [pnfs] [PATCH 2/5] nfsd: Fix independence of a few nfsd related headers

Quoted 66 lines · Boaz Harrosh
Quoted 58 lines · Trond Myklebust
Quoted 57 lines · Boaz Harrosh
Quoted 9 lines · Boaz Harrosh

An header should be compilation independent, .i.e pull in any header who's declarations are directly used by this header. And not let users re-include all it's dependencies all over again.

[At the end of the day what's the use of a header if it does
not have more then one user?]

Trond do I have an ACK on this patch.
If not, then what should be changed to get it accepted?

Quoted 2 lines
---
 include/linux/nfs_xdr.h   |    1 +

This header is used exclusively by fs/nfs/... files and could just be moved there. The include must be fixed as below though.

Quoted 1 line
 include/linux/nfsacl.h    |    1 +

This file is used mixed between fs/nfs && fs/nfsd

Quoted 1 line
 include/linux/posix_acl.h |    1 +

Used by nfsd and filesystems

Quoted 33 lines
 3 files changed, 3 insertions(+), 0 deletions(-)
include/linux/nfs_xdr.h +1 -0
index 2848a26..c316ca8 100644
@@ -2,6 +2,7 @@
 #define _LINUX_NFS_XDR_H
 
 #include <linux/nfsacl.h>
+#include <linux/nfs3.h>
 
 /*
  * To change the maximum rsize and wsize supported by the NFS client, adjust
include/linux/nfsacl.h +1 -0
index 43011b6..f321b57 100644
@@ -29,6 +29,7 @@
 #ifdef __KERNEL__
 
 #include <linux/posix_acl.h>
+#include <linux/sunrpc/xdr.h>
 
 /* Maximum number of ACL entries over NFS */
 #define NFS_ACL_MAX_ENTRIES	1024
include/linux/posix_acl.h +0 -0
index 065a365..0dcf674 100644
@@ -9,6 +9,7 @@
 #define __LINUX_POSIX_ACL_H
 
Quoted 6 lines
Quoted 2 lines

include/linux/posix_acl.h line 12

Quoted 2 lines
include/linux/posix_acl.h +1 -0
 #include <linux/slab.h>
+#include <linux/fs.h>

NACK to this. Pretty much _all_ filesystems already include linux/fs.h somewhere in their include chains. There should be no need to include it in posix_acl.h too.

Quoted 2 lines

from posix_acl.h:
extern int posix_acl_permission(struct inode *, const struct posix_acl *, int);

get_cached_acl's inline definition would make a better example as it actually needs the formal definition of struct inode;

Quoted 5 lines

stuct inode is defined in fs.h. hence the direct dependency.

Again a double inclusion is not a bad thing, it costs absolutely *nothing*.

Quoted 16 lines

A miss-inclusion on the other hand is a bad thing. it causes a miss-compilation.

It does not matter that filesystems include fs.h or not. What matters is that they now have to do this headers ordering magic. One places code compiles fine, another place it does not. When it does not, people *never* analyze the missing dependency what they do is copy-paste an include list from another file that works.

Proof of the matter the last patches in this patchset.

Quoted 7 lines
Quoted 3 lines

include/linux/posix_acl.h line 15

Quoted 3 lines
include/linux/posix_acl.h +0 -0
 #define ACL_UNDEFINED_ID	(-1)
 

So, what is the motivation for all this? We have no dependency problems here today. What is changing in the pNFS tree that makes this so necessary?

The motivation for this patchset was mainly patch 4/5
pnfsd: Move pnfsd code out of nfs4state.c/h
Boaz meant to amortize the cleanup effort on this mini-project. Otherwise, by itself, I don't think we'd have started doing it...

Benny

Quoted 15 lines

We do have a dependency problem today!! look at the patchset. It is a grate cleanup and improvement of code today. And a much smoother ride for the future.

What changed is that all this code was touched today. I have not done the cleanup for any files not touched by pnfsd. Though I could and should, because they will greatly improve just like these I did touch. Should I go head and do the reset of the files?

And please note that this particular file is an NFSD and vfs related file only, it has nothing to do with nfs.

Quoted 2 lines

Trond

Boaz

#29 lore raw

Re: [pnfs] [PATCH 2/5] nfsd: Fix independence of a few nfsd related headers

In reply to Boaz Harrosh
Quoted 71 lines · Boaz Harrosh
Quoted 62 lines · Trond Myklebust
Quoted 61 lines · Boaz Harrosh
Quoted 9 lines · Boaz Harrosh

An header should be compilation independent, .i.e pull in any header who's declarations are directly used by this header. And not let users re-include all it's dependencies all over again.

[At the end of the day what's the use of a header if it does
not have more then one user?]

Trond do I have an ACK on this patch.
If not, then what should be changed to get it accepted?

Quoted 2 lines
---
 include/linux/nfs_xdr.h   |    1 +

This header is used exclusively by fs/nfs/... files and could just be moved there. The include must be fixed as below though.

Quoted 1 line
 include/linux/nfsacl.h    |    1 +

This file is used mixed between fs/nfs && fs/nfsd

Quoted 1 line
 include/linux/posix_acl.h |    1 +

Used by nfsd and filesystems

Quoted 33 lines
 3 files changed, 3 insertions(+), 0 deletions(-)
include/linux/nfs_xdr.h +1 -0
index 2848a26..c316ca8 100644
@@ -2,6 +2,7 @@
 #define _LINUX_NFS_XDR_H
 
 #include <linux/nfsacl.h>
+#include <linux/nfs3.h>
 
 /*
  * To change the maximum rsize and wsize supported by the NFS client, adjust
include/linux/nfsacl.h +1 -0
index 43011b6..f321b57 100644
@@ -29,6 +29,7 @@
 #ifdef __KERNEL__
 
 #include <linux/posix_acl.h>
+#include <linux/sunrpc/xdr.h>
 
 /* Maximum number of ACL entries over NFS */
 #define NFS_ACL_MAX_ENTRIES	1024
include/linux/posix_acl.h +0 -0
index 065a365..0dcf674 100644
@@ -9,6 +9,7 @@
 #define __LINUX_POSIX_ACL_H
 
Quoted 7 lines
Quoted 2 lines

include/linux/posix_acl.h line 12

Quoted 2 lines
include/linux/posix_acl.h +1 -0
 #include <linux/slab.h>
+#include <linux/fs.h>

NACK to this. Pretty much _all_ filesystems already include linux/fs.h somewhere in their include chains. There should be no need to include it in posix_acl.h too.

Quoted 4 lines

from posix_acl.h:
extern int posix_acl_permission(struct inode *, const struct posix_acl *, int);

stuct inode is defined in fs.h. hence the direct dependency.

Add a forward declaration of the form

struct inode;

posix_acl.h, and you're done. No need to read the entire contents of fs.h plus all the crap it drags in with it, in order to satisfy a single pointer dependence.

Quoted 3 lines

Again a double inclusion is not a bad thing, it costs absolutely *nothing*.

A miss-inclusion on the other hand is a bad thing. it causes a miss-compilation.

Really?

Quoted 20 lines

It does not matter that filesystems include fs.h or not. What matters is that they now have to do this headers ordering magic. One places code compiles fine, another place it does not. When it does not, people *never* analyze the missing dependency what they do is copy-paste an include list from another file that works.

Proof of the matter the last patches in this patchset.

Quoted 9 lines
Quoted 4 lines

include/linux/posix_acl.h line 15

Quoted 3 lines
include/linux/posix_acl.h +0 -0
 #define ACL_UNDEFINED_ID	(-1)
 

So, what is the motivation for all this? We have no dependency problems here today. What is changing in the pNFS tree that makes this so necessary?

We do have a dependency problem today!! look at the patchset. It is a grate cleanup and improvement of code today. And a much smoother ride for the future.

Quoted 6 lines

What changed is that all this code was touched today. I have not done the cleanup for any files not touched by pnfsd. Though I could and should, because they will greatly improve just like these I did touch. Should I go head and do the reset of the files?

And please note that this particular file is an NFSD and vfs related file only, it has nothing to do with nfs.

...and hence you need to send it to the VFS maintainer. I'm not touching that patch, or the patchset that "requires" it.

Signature · 6 lines
Trond Myklebust
Linux NFS client maintainer

NetApp
[email protected]
www.netapp.com
#31 lore raw

Re: [pnfs] [PATCH 2/5] nfsd: Fix independence of a few nfsd related headers

In reply to a message not in the archive
Quoted 11 lines · Trond Myklebust
Quoted 5 lines · Boaz Harrosh

from posix_acl.h:
extern int posix_acl_permission(struct inode *, const struct posix_acl *, int);

stuct inode is defined in fs.h. hence the direct dependency.

Add a forward declaration of the form

struct inode;

This will not work.

<quote>
...

static inline struct posix_acl *get_cached_acl(struct inode *inode, int type)

{
	struct posix_acl **p, *acl;
	switch (type) {
	case ACL_TYPE_ACCESS:
		p = &inode->i_acl;
		break;
	case ACL_TYPE_DEFAULT:
		p = &inode->i_default_acl;
		break;
	default:
		return ERR_PTR(-EINVAL);
	}

...
</quote>

the header has inlines that directly manipulate inode members a forward declaration is not enough.

<snip>

Quoted 6 lines
Quoted 2 lines

And please note that this particular file is an NFSD and vfs related file only, it has nothing to do with nfs.

...and hence you need to send it to the VFS maintainer. I'm not touching that patch, or the patchset that "requires" it.

OK This particular hunk I will send to fsdevel, anything else?

Boaz