netdev.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: marcelo.leitner@gmail.com
To: David Miller <davem@davemloft.net>
Cc: joe@perches.com, netdev@vger.kernel.org, nhorman@tuxdriver.com,
	vyasevich@gmail.com, linux-sctp@vger.kernel.org
Subject: Re: [PATCH v2 1/2] sctp: compress bit-wide flags to a bitfield on sctp_sock
Date: Wed, 6 Apr 2016 18:21:48 -0300	[thread overview]
Message-ID: <20160406212148.GB15005@localhost.localdomain> (raw)
In-Reply-To: <20160406.155735.1431363453832315669.davem@davemloft.net>

On Wed, Apr 06, 2016 at 03:57:35PM -0400, David Miller wrote:
> From: Joe Perches <joe@perches.com>
> Date: Wed, 06 Apr 2016 12:53:24 -0700
> 
> > On Wed, 2016-04-06 at 14:53 -0300, Marcelo Ricardo Leitner wrote:
> >> It wastes space and gets worse as we add new flags, so convert bit-wide
> >> flags to a bitfield.
> >> 
> >> Currently it already saves 4 bytes in sctp_sock, which are left as holes
> >> in it for now. The whole struct needs packing, which should be done in
> >> another patch.
> >> 
> >> Note that do_auto_asconf cannot be merged, as explained in the comment
> >> before it.
> >> 
> >> Signed-off-by: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com>
> >> ---
> > []
> >> diff --git a/include/net/sctp/structs.h b/include/net/sctp/structs.h
> > []
> >> @@ -210,14 +210,14 @@ struct sctp_sock {
> >>  	int user_frag;
> >>  
> >>  	__u32 autoclose;
> >> -	__u8 nodelay;
> >> -	__u8 disable_fragments;
> >> -	__u8 v4mapped;
> >> -	__u8 frag_interleave;
> >>  	__u32 adaptation_ind;
> >>  	__u32 pd_point;
> >> -	__u8 recvrcvinfo;
> >> -	__u8 recvnxtinfo;
> >> +	__u16	nodelay:1,
> >> +		disable_fragments:1,
> >> +		v4mapped:1,
> >> +		frag_interleave:1,
> >> +		recvrcvinfo:1,
> >> +		recvnxtinfo:1;
> > 
> > Might as well make this __u32 as the next field would be
> > aligned on an atomic_t

On changelog I meant that this (re-)ordering will happen in another
patch. The hole is already there today and there are others to be
considered/fixed too. Please consider this patch as a preparation for
the other one.

After this patch, pahole gives for this struct:

        /* size: 1272, cachelines: 20, members: 40 */
        /* sum members: 1250, holes: 7, sum holes: 18 */
        /* bit holes: 1, sum bit holes: 9 bits */
        /* padding: 4 */
        /* paddings: 1, sum paddings: 2 */
        /* last cacheline: 56 bytes */

Quite some work todo :-)

> > It might be better if these fields didn't use the __ prefix.
> 
> Indeed, this isn't in a UAPI file so __ prefixed types really aren't
> appropriate.

The whole file is using __ prefixed types. I can remove them in another
patch instead if that's okay with you.
It's also present in other files not under UAPI:
$ git grep -wl '\(__u32\|__u16\)' include/net/sctp/
include/net/sctp/auth.h
include/net/sctp/checksum.h
include/net/sctp/command.h
include/net/sctp/constants.h
include/net/sctp/sctp.h
include/net/sctp/sm.h
include/net/sctp/structs.h
include/net/sctp/tsnmap.h
include/net/sctp/ulpevent.h
include/net/sctp/ulpqueue.h

We can start changing it progressively but I think it would be better if
we replace them all at once.

  Marcelo

  reply	other threads:[~2016-04-06 21:21 UTC|newest]

Thread overview: 8+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-04-06 17:53 [PATCH v2 0/2] sctp: delay calls to sk_data_ready() as much as possible Marcelo Ricardo Leitner
2016-04-06 17:53 ` [PATCH v2 1/2] sctp: compress bit-wide flags to a bitfield on sctp_sock Marcelo Ricardo Leitner
2016-04-06 19:53   ` Joe Perches
2016-04-06 19:57     ` David Miller
2016-04-06 21:21       ` marcelo.leitner [this message]
2016-04-06 17:53 ` [PATCH v2 2/2] sctp: delay calls to sk_data_ready() as much as possible Marcelo Ricardo Leitner
2016-04-07  8:05   ` Jakub Sitnicki
2016-04-07 13:35     ` Marcelo Ricardo Leitner

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20160406212148.GB15005@localhost.localdomain \
    --to=marcelo.leitner@gmail.com \
    --cc=davem@davemloft.net \
    --cc=joe@perches.com \
    --cc=linux-sctp@vger.kernel.org \
    --cc=netdev@vger.kernel.org \
    --cc=nhorman@tuxdriver.com \
    --cc=vyasevich@gmail.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).