Skip to contentSkip to conversation

: MIGRATE support (take 2)

PATCH In review Shinta Sugimoto 3 participants 6 messages Feb 8, 2007 – Feb 10, 2007

Subsystems (2)
lore
Share & tools
lore
#0 lore raw

cover : MIGRATE support (take 2)

Dear David,

I noticed that I sent wrong version (not the newest one) of the patch set by mistake. I will re-post the patch set soon, so could you please wait for that? I am sorry for the confusion.

Regards,
Shinta

Forwarded by Shinta Sugimoto <[email protected]> ----------------------- Original Message -----------------------

 From:    Shinta Sugimoto <[email protected]>
 To:      [email protected]
 Cc:      Francis Dupont <[email protected]>,

Masahide Nakamura <[email protected]>, [email protected], [email protected]
Date: Wed, 07 Feb 2007 22:17:41 +0900
Subject: [PATCH 0/5] [XFRM,PFKEYV2]: MIGRATE support (take 2)
----

Hello,

Let me send an updated version of patch set of MIGRATE support for XFRM/PFKEYv2. About a week ago, I sent a request for comments for this patch set. All the comments received were taken into account. In summary, the changes are as follow:

- replace alloc & memcpy with kmemdup
- remove unnecessary null-check
- fix bugs in handling protocol value for selector
- add sanity check for handling MIGRATE message from PF_KEY socket
- some other minor changes

Please note that the patch set is against current net-2.6.21 tree and these changes are available in the branch named "migrate-20070207-2131" at:

git://git.linux-ipv6.org/gitroot/shinta/net-2.6.21-migrate.git

Regards,
Shinta
-
To unsubscribe from this list: send the line "unsubscribe netdev" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
--------------------- Original Message Ends --------------------

maintainer
#1 lore raw

From: Shinta Sugimoto <[email protected]>
Date: Thu, 08 Feb 2007 09:42:00 +0900

Quoted 3 lines

I noticed that I sent wrong version (not the newest one) of the patch set by mistake. I will re-post the patch set soon, so could you please wait for that? I am sorry for the confusion.

Ok, no problem.

#2 lore raw

cover NET Improve layering of structs dst_entry/rtable/rt6_info/dn_route to get less dcache footprint

Hi David

Large routers/servers spend lot of CPU time in route cache lookups. This is because each item lookup uses two cache lines : the next pointer is at the begining of item, but keys are far away from this pointer.

Furthermore, the u unions declared in struct rtable, rt6_info and dn_route are imho quite ugly (forcing next pointer be at null offset)

This patch try to cleanup the layering and try to put the 'next' pointer near the fields used at lookup time (in ip_route_input() for example)

I prepared the following five patches :

[PATCH 1/5] : Introduce union in struct dst_entry, to prepare patches 2,3,4 [PATCH 2/5] : Convert ipv4 route to use the new dst_entry 'next' pointer [PATCH 3/5] : Convert ipv6 route to use the new dst_entry 'next' pointer [PATCH 4/5] : Convert decnet route to use the new dst_entry 'next' pointer [PATCH 5/5] : Reorder fields of struct dst_entry

Thank you
Eric Dumazet

maintainer
#3 lore raw

Re: [PATCH 0/5] NET Improve layering of structs dst_entry/rtable/rt6_info/dn_route to get less dcache footprint

From: Eric Dumazet <[email protected]>
Date: Thu, 08 Feb 2007 22:32:10 +0100

Quoted 7 lines

I prepared the following five patches :

[PATCH 1/5] : Introduce union in struct dst_entry, to prepare patches 2,3,4 [PATCH 2/5] : Convert ipv4 route to use the new dst_entry 'next' pointer [PATCH 3/5] : Convert ipv6 route to use the new dst_entry 'next' pointer [PATCH 4/5] : Convert decnet route to use the new dst_entry 'next' pointer [PATCH 5/5] : Reorder fields of struct dst_entry

I've applied all of this, and fixed up the changelog description of patch #4 as-needed :-)

Thanks a lot Eric.

#4 lore raw

Re: [PATCH 0/5] NET Improve layering of structs dst_entry/rtable/rt6_info/dn_route to get less dcache footprint

Quoted 11 lines · David Miller

From: Eric Dumazet <[email protected]>
Date: Thu, 08 Feb 2007 22:32:10 +0100

Quoted 7 lines

I prepared the following five patches :

[PATCH 1/5] : Introduce union in struct dst_entry, to prepare patches 2,3,4 [PATCH 2/5] : Convert ipv4 route to use the new dst_entry 'next' pointer [PATCH 3/5] : Convert ipv6 route to use the new dst_entry 'next' pointer [PATCH 4/5] : Convert decnet route to use the new dst_entry 'next' pointer [PATCH 5/5] : Reorder fields of struct dst_entry

Quoted 4 lines

I've applied all of this, and fixed up the changelog description of patch #4 as-needed :-)

Thanks a lot Eric.

I was about to correct this stuff this saturday morning, so *many* thanks to you David, I can start my week end like normal guys :)

#5 lore raw

PATCH 1/5 NET : Introduce union in struct dst_entry to hold 'next' pointer

In reply to David Miller

This patch introduces an anonymous union to nicely express the fact that all objects inherited from struct dst_entry should access to the generic 'next' pointer but with appropriate type verification.

This patch is a prereq before following patches.

include/net/dst.h +6 -1
@@ -37,7 +37,12 @@ struct sk_buff;
 
 struct dst_entry
 {
-	struct dst_entry        *next;
+	union {
+		struct dst_entry *next;
+		struct rtable    *rt_next;
+		struct rt6_info   *rt6_next;
+		struct dn_route  *dn_next;
+	};
 	atomic_t		__refcnt;	/* client references	*/
 	int			__use;
 	struct dst_entry	*child;