include/net/mctp.h | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
Group some variables based on their sizes to reduce hole and avoid padding.
On x86_64, this shrinks the size of 'struct mctp_route'
from 72 to 64 bytes.
It saves a few bytes of memory and is more cache-line friendly.
Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
---
Using pahole
Before:
======
struct mctp_route {
mctp_eid_t min; /* 0 1 */
mctp_eid_t max; /* 1 1 */
/* XXX 6 bytes hole, try to pack */
struct mctp_dev * dev; /* 8 8 */
unsigned int mtu; /* 16 4 */
unsigned char type; /* 20 1 */
/* XXX 3 bytes hole, try to pack */
int (*output)(struct mctp_route *, struct sk_buff *); /* 24 8 */
struct list_head list; /* 32 16 */
refcount_t refs; /* 48 4 */
/* XXX 4 bytes hole, try to pack */
struct callback_head rcu __attribute__((__aligned__(8))); /* 56 16 */
/* size: 72, cachelines: 2, members: 9 */
/* sum members: 59, holes: 3, sum holes: 13 */
/* forced alignments: 1, forced holes: 1, sum forced holes: 4 */
/* last cacheline: 8 bytes */
} __attribute__((__aligned__(8)));
After:
=====
struct mctp_route {
mctp_eid_t min; /* 0 1 */
mctp_eid_t max; /* 1 1 */
unsigned char type; /* 2 1 */
/* XXX 1 byte hole, try to pack */
unsigned int mtu; /* 4 4 */
struct mctp_dev * dev; /* 8 8 */
int (*output)(struct mctp_route *, struct sk_buff *); /* 16 8 */
struct list_head list; /* 24 16 */
refcount_t refs; /* 40 4 */
/* XXX 4 bytes hole, try to pack */
struct callback_head rcu __attribute__((__aligned__(8))); /* 48 16 */
/* size: 64, cachelines: 1, members: 9 */
/* sum members: 59, holes: 2, sum holes: 5 */
/* forced alignments: 1, forced holes: 1, sum forced holes: 4 */
} __attribute__((__aligned__(8)));
---
include/net/mctp.h | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/include/net/mctp.h b/include/net/mctp.h
index 82800d521c3d..da86e106c91d 100644
--- a/include/net/mctp.h
+++ b/include/net/mctp.h
@@ -234,9 +234,9 @@ struct mctp_flow {
struct mctp_route {
mctp_eid_t min, max;
- struct mctp_dev *dev;
- unsigned int mtu;
unsigned char type;
+ unsigned int mtu;
+ struct mctp_dev *dev;
int (*output)(struct mctp_route *route,
struct sk_buff *skb);
--
2.34.1
Hello:
This patch was applied to netdev/net-next.git (main)
by Jakub Kicinski <kuba@kernel.org>:
On Sun, 18 Jun 2023 11:33:55 +0200 you wrote:
> Group some variables based on their sizes to reduce hole and avoid padding.
> On x86_64, this shrinks the size of 'struct mctp_route'
> from 72 to 64 bytes.
>
> It saves a few bytes of memory and is more cache-line friendly.
>
> Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
>
> [...]
Here is the summary with links:
- [net-next] mctp: Reorder fields in 'struct mctp_route'
https://git.kernel.org/netdev/net-next/c/066768b7305b
You are awesome, thank you!
--
Deet-doot-dot, I am a bot.
https://korg.docs.kernel.org/patchwork/pwbot.html
Sun, Jun 18, 2023 at 11:33:55AM CEST, christophe.jaillet@wanadoo.fr wrote: >Group some variables based on their sizes to reduce hole and avoid padding. >On x86_64, this shrinks the size of 'struct mctp_route' >from 72 to 64 bytes. > >It saves a few bytes of memory and is more cache-line friendly. > >Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr> Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Hi Christoph, > Group some variables based on their sizes to reduce hole and avoid > padding. On x86_64, this shrinks the size of 'struct mctp_route' from > 72 to 64 bytes. > > It saves a few bytes of memory and is more cache-line friendly. The savings will be fairly minimal, but this doesn't affect readability for the route struct. LGTM. Acked-by: Jeremy Kerr <jk@codeconstruct.com.au> Thanks! Jeremy
On Sun, Jun 18, 2023 at 11:33:55AM +0200, Christophe JAILLET wrote: > Group some variables based on their sizes to reduce hole and avoid padding. > On x86_64, this shrinks the size of 'struct mctp_route' > from 72 to 64 bytes. > > It saves a few bytes of memory and is more cache-line friendly. > > Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr> Reviewed-by: Simon Horman <simon.horman@corigine.com>
© 2016 - 2026 Red Hat, Inc.