All of lore.kernel.org
 help / color / mirror / Atom feed
* [PATCH net-next v2 0/3] genetlink: remove linux/genetlink.h
@ 2024-03-25 17:37 Jakub Kicinski
  2024-03-25 17:37 ` [PATCH net-next v2 1/3] netlink: create a new header for internal genetlink symbols Jakub Kicinski
                   ` (2 more replies)
  0 siblings, 3 replies; 6+ messages in thread
From: Jakub Kicinski @ 2024-03-25 17:37 UTC (permalink / raw)
  To: davem
  Cc: netdev, edumazet, pabeni, dw, jiri, andriy.shevchenko, Jakub Kicinski

There are two genetlink headers net/genetlink.h and linux/genetlink.h
This is similar to netlink.h, but for netlink.h both contain good
amount of code. For genetlink.h the linux/ version is leftover
from before uAPI headers were split out, it has 10 lines of code.
Move those 10 lines into other appropriate headers and delete
linux/genetlink.h.

I occasionally open the wrong header in the editor when coding,
I guess I'm not the only one.

v2:
 - fix Andy's nits on patch 3
v1: https://lore.kernel.org/all/20240309183458.3014713-1-kuba@kernel.org

Jakub Kicinski (3):
  netlink: create a new header for internal genetlink symbols
  net: openvswitch: remove unnecessary linux/genetlink.h include
  genetlink: remove linux/genetlink.h

 drivers/net/wireguard/main.c      |  2 +-
 include/linux/genetlink.h         | 19 -------------------
 include/linux/genl_magic_struct.h |  2 +-
 include/net/genetlink.h           | 10 +++++++++-
 net/batman-adv/main.c             |  2 +-
 net/batman-adv/netlink.c          |  1 -
 net/netlink/af_netlink.c          |  2 +-
 net/netlink/genetlink.c           |  2 ++
 net/netlink/genetlink.h           | 11 +++++++++++
 net/openvswitch/datapath.c        |  1 -
 net/openvswitch/meter.h           |  1 -
 11 files changed, 26 insertions(+), 27 deletions(-)
 delete mode 100644 include/linux/genetlink.h
 create mode 100644 net/netlink/genetlink.h

-- 
2.44.0


^ permalink raw reply	[flat|nested] 6+ messages in thread

* [PATCH net-next v2 1/3] netlink: create a new header for internal genetlink symbols
  2024-03-25 17:37 [PATCH net-next v2 0/3] genetlink: remove linux/genetlink.h Jakub Kicinski
@ 2024-03-25 17:37 ` Jakub Kicinski
  2024-03-25 17:37 ` [PATCH net-next v2 2/3] net: openvswitch: remove unnecessary linux/genetlink.h include Jakub Kicinski
  2024-03-25 17:37 ` [PATCH net-next v2 3/3] genetlink: remove linux/genetlink.h Jakub Kicinski
  2 siblings, 0 replies; 6+ messages in thread
From: Jakub Kicinski @ 2024-03-25 17:37 UTC (permalink / raw)
  To: davem
  Cc: netdev, edumazet, pabeni, dw, jiri, andriy.shevchenko,
	Jakub Kicinski, kuniyu

There are things in linux/genetlink.h which are only used
under net/netlink/. Move them to a new local header.
A new header with just 2 externs isn't great, but alternative
would be to include af_netlink.h in genetlink.c which feels
even worse.

Signed-off-by: Jakub Kicinski <kuba@kernel.org>
---
CC: kuniyu@amazon.com
CC: jiri@resnulli.us
---
 include/linux/genetlink.h |  5 -----
 net/netlink/af_netlink.c  |  2 +-
 net/netlink/genetlink.c   |  2 ++
 net/netlink/genetlink.h   | 11 +++++++++++
 4 files changed, 14 insertions(+), 6 deletions(-)
 create mode 100644 net/netlink/genetlink.h

diff --git a/include/linux/genetlink.h b/include/linux/genetlink.h
index c285968e437a..9dbd7ba9b858 100644
--- a/include/linux/genetlink.h
+++ b/include/linux/genetlink.h
@@ -4,15 +4,10 @@
 
 #include <uapi/linux/genetlink.h>
 
-
 /* All generic netlink requests are serialized by a global lock.  */
 extern void genl_lock(void);
 extern void genl_unlock(void);
 
-/* for synchronisation between af_netlink and genetlink */
-extern atomic_t genl_sk_destructing_cnt;
-extern wait_queue_head_t genl_sk_destructing_waitq;
-
 #define MODULE_ALIAS_GENL_FAMILY(family)\
  MODULE_ALIAS_NET_PF_PROTO_NAME(PF_NETLINK, NETLINK_GENERIC, "-family-" family)
 
diff --git a/net/netlink/af_netlink.c b/net/netlink/af_netlink.c
index 7554803218a2..dc8c3c01d51b 100644
--- a/net/netlink/af_netlink.c
+++ b/net/netlink/af_netlink.c
@@ -59,7 +59,6 @@
 #include <linux/rhashtable.h>
 #include <asm/cacheflush.h>
 #include <linux/hash.h>
-#include <linux/genetlink.h>
 #include <linux/net_namespace.h>
 #include <linux/nospec.h>
 #include <linux/btf_ids.h>
@@ -73,6 +72,7 @@
 #include <trace/events/netlink.h>
 
 #include "af_netlink.h"
+#include "genetlink.h"
 
 struct listeners {
 	struct rcu_head		rcu;
diff --git a/net/netlink/genetlink.c b/net/netlink/genetlink.c
index 3b7666944b11..feb54c63a116 100644
--- a/net/netlink/genetlink.c
+++ b/net/netlink/genetlink.c
@@ -22,6 +22,8 @@
 #include <net/sock.h>
 #include <net/genetlink.h>
 
+#include "genetlink.h"
+
 static DEFINE_MUTEX(genl_mutex); /* serialization of message processing */
 static DECLARE_RWSEM(cb_lock);
 
diff --git a/net/netlink/genetlink.h b/net/netlink/genetlink.h
new file mode 100644
index 000000000000..89bd9d2631c3
--- /dev/null
+++ b/net/netlink/genetlink.h
@@ -0,0 +1,11 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+#ifndef __NET_GENETLINK_H
+#define __NET_GENETLINK_H
+
+#include <linux/wait.h>
+
+/* for synchronisation between af_netlink and genetlink */
+extern atomic_t genl_sk_destructing_cnt;
+extern wait_queue_head_t genl_sk_destructing_waitq;
+
+#endif	/* __LINUX_GENERIC_NETLINK_H */
-- 
2.44.0


^ permalink raw reply related	[flat|nested] 6+ messages in thread

* [PATCH net-next v2 2/3] net: openvswitch: remove unnecessary linux/genetlink.h include
  2024-03-25 17:37 [PATCH net-next v2 0/3] genetlink: remove linux/genetlink.h Jakub Kicinski
  2024-03-25 17:37 ` [PATCH net-next v2 1/3] netlink: create a new header for internal genetlink symbols Jakub Kicinski
@ 2024-03-25 17:37 ` Jakub Kicinski
  2024-03-25 17:37 ` [PATCH net-next v2 3/3] genetlink: remove linux/genetlink.h Jakub Kicinski
  2 siblings, 0 replies; 6+ messages in thread
From: Jakub Kicinski @ 2024-03-25 17:37 UTC (permalink / raw)
  To: davem
  Cc: netdev, edumazet, pabeni, dw, jiri, andriy.shevchenko,
	Jakub Kicinski, pshelar, dev

The only legit reason I could think of for net/genetlink.h
and linux/genetlink.h to be separate would be if one was
included by other headers and we wanted to keep it lightweight.
That is not the case, net/openvswitch/meter.h includes
linux/genetlink.h but for no apparent reason (for struct genl_family
perhaps? it's not necessary, types of externs do not need
to be known).

Signed-off-by: Jakub Kicinski <kuba@kernel.org>
---
CC: pshelar@ovn.org
CC: dev@openvswitch.org
---
 net/openvswitch/meter.h | 1 -
 1 file changed, 1 deletion(-)

diff --git a/net/openvswitch/meter.h b/net/openvswitch/meter.h
index ed11cd12b512..8bbf983cd244 100644
--- a/net/openvswitch/meter.h
+++ b/net/openvswitch/meter.h
@@ -11,7 +11,6 @@
 #include <linux/kernel.h>
 #include <linux/netlink.h>
 #include <linux/openvswitch.h>
-#include <linux/genetlink.h>
 #include <linux/skbuff.h>
 #include <linux/bits.h>
 
-- 
2.44.0


^ permalink raw reply related	[flat|nested] 6+ messages in thread

* [PATCH net-next v2 3/3] genetlink: remove linux/genetlink.h
  2024-03-25 17:37 [PATCH net-next v2 0/3] genetlink: remove linux/genetlink.h Jakub Kicinski
  2024-03-25 17:37 ` [PATCH net-next v2 1/3] netlink: create a new header for internal genetlink symbols Jakub Kicinski
  2024-03-25 17:37 ` [PATCH net-next v2 2/3] net: openvswitch: remove unnecessary linux/genetlink.h include Jakub Kicinski
@ 2024-03-25 17:37 ` Jakub Kicinski
  2024-03-25 17:45   ` Andy Shevchenko
  2 siblings, 1 reply; 6+ messages in thread
From: Jakub Kicinski @ 2024-03-25 17:37 UTC (permalink / raw)
  To: davem
  Cc: netdev, edumazet, pabeni, dw, jiri, andriy.shevchenko,
	Jakub Kicinski, Sven Eckelmann, Jason, mareklindner, sw, a,
	pshelar, wireguard, dev

genetlink.h is a shell of what used to be a combined uAPI
and kernel header over a decade ago. It has fewer than
10 lines of code. Merge it into net/genetlink.h.
In some ways it'd be better to keep the combined header
under linux/ but it would make looking through git history
harder.

Acked-by: Sven Eckelmann <sven@narfation.org>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
---
v2:
 - remove extern
 - include linux/net.h
 - improve the comment, not "all" requests are serialized

CC: Jason@zx2c4.com
CC: mareklindner@neomailbox.ch
CC: sw@simonwunderlich.de
CC: a@unstable.cc
CC: pshelar@ovn.org
CC: andriy.shevchenko@linux.intel.com
CC: wireguard@lists.zx2c4.com
CC: dev@openvswitch.org
---
 drivers/net/wireguard/main.c      |  2 +-
 include/linux/genetlink.h         | 14 --------------
 include/linux/genl_magic_struct.h |  2 +-
 include/net/genetlink.h           | 10 +++++++++-
 net/batman-adv/main.c             |  2 +-
 net/batman-adv/netlink.c          |  1 -
 net/openvswitch/datapath.c        |  1 -
 7 files changed, 12 insertions(+), 20 deletions(-)
 delete mode 100644 include/linux/genetlink.h

diff --git a/drivers/net/wireguard/main.c b/drivers/net/wireguard/main.c
index ee4da9ab8013..a00671b58701 100644
--- a/drivers/net/wireguard/main.c
+++ b/drivers/net/wireguard/main.c
@@ -14,7 +14,7 @@
 
 #include <linux/init.h>
 #include <linux/module.h>
-#include <linux/genetlink.h>
+#include <net/genetlink.h>
 #include <net/rtnetlink.h>
 
 static int __init wg_mod_init(void)
diff --git a/include/linux/genetlink.h b/include/linux/genetlink.h
deleted file mode 100644
index 9dbd7ba9b858..000000000000
--- a/include/linux/genetlink.h
+++ /dev/null
@@ -1,14 +0,0 @@
-/* SPDX-License-Identifier: GPL-2.0 */
-#ifndef __LINUX_GENERIC_NETLINK_H
-#define __LINUX_GENERIC_NETLINK_H
-
-#include <uapi/linux/genetlink.h>
-
-/* All generic netlink requests are serialized by a global lock.  */
-extern void genl_lock(void);
-extern void genl_unlock(void);
-
-#define MODULE_ALIAS_GENL_FAMILY(family)\
- MODULE_ALIAS_NET_PF_PROTO_NAME(PF_NETLINK, NETLINK_GENERIC, "-family-" family)
-
-#endif	/* __LINUX_GENERIC_NETLINK_H */
diff --git a/include/linux/genl_magic_struct.h b/include/linux/genl_magic_struct.h
index a419d93789ff..621b87a87d74 100644
--- a/include/linux/genl_magic_struct.h
+++ b/include/linux/genl_magic_struct.h
@@ -15,8 +15,8 @@
 #endif
 
 #include <linux/args.h>
-#include <linux/genetlink.h>
 #include <linux/types.h>
+#include <net/genetlink.h>
 
 extern int CONCATENATE(GENL_MAGIC_FAMILY, _genl_register)(void);
 extern void CONCATENATE(GENL_MAGIC_FAMILY, _genl_unregister)(void);
diff --git a/include/net/genetlink.h b/include/net/genetlink.h
index 9ece6e5a3ea8..7648dd6b8754 100644
--- a/include/net/genetlink.h
+++ b/include/net/genetlink.h
@@ -2,12 +2,20 @@
 #ifndef __NET_GENERIC_NETLINK_H
 #define __NET_GENERIC_NETLINK_H
 
-#include <linux/genetlink.h>
+#include <linux/net.h>
 #include <net/netlink.h>
 #include <net/net_namespace.h>
+#include <uapi/linux/genetlink.h>
 
 #define GENLMSG_DEFAULT_SIZE (NLMSG_DEFAULT_SIZE - GENL_HDRLEN)
 
+/* Non-parallel generic netlink requests are serialized by a global lock.  */
+void genl_lock(void);
+void genl_unlock(void);
+
+#define MODULE_ALIAS_GENL_FAMILY(family) \
+ MODULE_ALIAS_NET_PF_PROTO_NAME(PF_NETLINK, NETLINK_GENERIC, "-family-" family)
+
 /* Binding to multicast group requires %CAP_NET_ADMIN */
 #define GENL_MCAST_CAP_NET_ADMIN	BIT(0)
 /* Binding to multicast group requires %CAP_SYS_ADMIN */
diff --git a/net/batman-adv/main.c b/net/batman-adv/main.c
index 75119f1ffccc..8e0f44c71696 100644
--- a/net/batman-adv/main.c
+++ b/net/batman-adv/main.c
@@ -14,7 +14,6 @@
 #include <linux/crc32c.h>
 #include <linux/device.h>
 #include <linux/errno.h>
-#include <linux/genetlink.h>
 #include <linux/gfp.h>
 #include <linux/if_ether.h>
 #include <linux/if_vlan.h>
@@ -38,6 +37,7 @@
 #include <linux/string.h>
 #include <linux/workqueue.h>
 #include <net/dsfield.h>
+#include <net/genetlink.h>
 #include <net/rtnetlink.h>
 #include <uapi/linux/batadv_packet.h>
 #include <uapi/linux/batman_adv.h>
diff --git a/net/batman-adv/netlink.c b/net/batman-adv/netlink.c
index 0954757f0b8b..9362cd9d6f3d 100644
--- a/net/batman-adv/netlink.c
+++ b/net/batman-adv/netlink.c
@@ -15,7 +15,6 @@
 #include <linux/cache.h>
 #include <linux/err.h>
 #include <linux/errno.h>
-#include <linux/genetlink.h>
 #include <linux/gfp.h>
 #include <linux/if_ether.h>
 #include <linux/if_vlan.h>
diff --git a/net/openvswitch/datapath.c b/net/openvswitch/datapath.c
index 11c69415c605..99d72543abd3 100644
--- a/net/openvswitch/datapath.c
+++ b/net/openvswitch/datapath.c
@@ -15,7 +15,6 @@
 #include <linux/delay.h>
 #include <linux/time.h>
 #include <linux/etherdevice.h>
-#include <linux/genetlink.h>
 #include <linux/kernel.h>
 #include <linux/kthread.h>
 #include <linux/mutex.h>
-- 
2.44.0


^ permalink raw reply related	[flat|nested] 6+ messages in thread

* Re: [PATCH net-next v2 3/3] genetlink: remove linux/genetlink.h
  2024-03-25 17:37 ` [PATCH net-next v2 3/3] genetlink: remove linux/genetlink.h Jakub Kicinski
@ 2024-03-25 17:45   ` Andy Shevchenko
  2024-03-26  1:58     ` Jakub Kicinski
  0 siblings, 1 reply; 6+ messages in thread
From: Andy Shevchenko @ 2024-03-25 17:45 UTC (permalink / raw)
  To: Jakub Kicinski
  Cc: davem, netdev, edumazet, pabeni, dw, jiri, Sven Eckelmann, Jason,
	mareklindner, sw, a, pshelar, wireguard, dev

On Mon, Mar 25, 2024 at 10:37:16AM -0700, Jakub Kicinski wrote:
> genetlink.h is a shell of what used to be a combined uAPI
> and kernel header over a decade ago. It has fewer than
> 10 lines of code. Merge it into net/genetlink.h.
> In some ways it'd be better to keep the combined header
> under linux/ but it would make looking through git history
> harder.

Some last moment minor comments (no need to send a new version, JFYI,
maybe you tide this up when applying).

...

> +/* Non-parallel generic netlink requests are serialized by a global lock.  */

While at it, maybe drop extra space? (I noticed it was originally like this.

...

> +#define MODULE_ALIAS_GENL_FAMILY(family) \
> + MODULE_ALIAS_NET_PF_PROTO_NAME(PF_NETLINK, NETLINK_GENERIC, "-family-" family)

I would still make this TAB indented.

-- 
With Best Regards,
Andy Shevchenko



^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: [PATCH net-next v2 3/3] genetlink: remove linux/genetlink.h
  2024-03-25 17:45   ` Andy Shevchenko
@ 2024-03-26  1:58     ` Jakub Kicinski
  0 siblings, 0 replies; 6+ messages in thread
From: Jakub Kicinski @ 2024-03-26  1:58 UTC (permalink / raw)
  To: Andy Shevchenko
  Cc: davem, netdev, edumazet, pabeni, dw, jiri, Sven Eckelmann, Jason,
	mareklindner, sw, a, pshelar, wireguard, dev

On Mon, 25 Mar 2024 19:45:05 +0200 Andy Shevchenko wrote:
> > +/* Non-parallel generic netlink requests are serialized by a global lock.  */  
> 
> While at it, maybe drop extra space? (I noticed it was originally like this.

Fair, I'll post v3.

> > +#define MODULE_ALIAS_GENL_FAMILY(family) \
> > + MODULE_ALIAS_NET_PF_PROTO_NAME(PF_NETLINK, NETLINK_GENERIC, "-family-" family)  
> 
> I would still make this TAB indented.

That one I'd prefer to keep. Coin toss between going over 80 chars
and having someone atypical indent.
-- 
pw-bot: cr

^ permalink raw reply	[flat|nested] 6+ messages in thread

end of thread, other threads:[~2024-03-26  1:58 UTC | newest]

Thread overview: 6+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2024-03-25 17:37 [PATCH net-next v2 0/3] genetlink: remove linux/genetlink.h Jakub Kicinski
2024-03-25 17:37 ` [PATCH net-next v2 1/3] netlink: create a new header for internal genetlink symbols Jakub Kicinski
2024-03-25 17:37 ` [PATCH net-next v2 2/3] net: openvswitch: remove unnecessary linux/genetlink.h include Jakub Kicinski
2024-03-25 17:37 ` [PATCH net-next v2 3/3] genetlink: remove linux/genetlink.h Jakub Kicinski
2024-03-25 17:45   ` Andy Shevchenko
2024-03-26  1:58     ` Jakub Kicinski

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.