Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1756497Ab0FETLm (ORCPT ); Sat, 5 Jun 2010 15:11:42 -0400 Received: from mail-px0-f174.google.com ([209.85.212.174]:36360 "EHLO mail-px0-f174.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752877Ab0FETLk convert rfc822-to-8bit (ORCPT ); Sat, 5 Jun 2010 15:11:40 -0400 DomainKey-Signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=mime-version:sender:in-reply-to:references:date :x-google-sender-auth:message-id:subject:from:to:cc:content-type :content-transfer-encoding; b=mRrdGL7OsoSYEhjeZ7my/HK0QdVIGyG5JNT/eW98hCPoA/2bCwEmfhuDCcbi+OPI3o SgsBPHoZf3NV7X2kcmpBia9r3fwDpSEuvjC3fEPHeT7iINgjRAmQJKe9kZ+aTl10eGrC sTjZmrmucUOrTXM/OZ3odha7wKZMUeqKybjI8= MIME-Version: 1.0 In-Reply-To: <4C07C3B8.7020003@hp.com> References: <4C07C3B8.7020003@hp.com> Date: Sat, 5 Jun 2010 12:11:38 -0700 X-Google-Sender-Auth: 0rd61soRSQ3-l3A4VdQeJI6Cl3c Message-ID: Subject: Re: [PATCH 1/6] sctp multistream scheduling: declare sctp_sched_ops From: Yaogong Wang To: Vlad Yasevich Cc: linux-sctp@vger.kernel.org, Sridhar Samudrala , linux-kernel@vger.kernel.org Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8BIT Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 4265 Lines: 117 I agree that it's clearer to put sched_ops in sctp_outq. That way, the data structure and the operations on it are put together. I'll make the required changes. Yaogong On Thu, Jun 3, 2010 at 8:01 AM, Vlad Yasevich wrote: > > > Yaogong Wang wrote: >> Declare sctp_sched_ops structure and related functions >> >> Signed-off-by: Yaogong Wang >> --- >> diff -uprN -X linux-2.6.32.8/Documentation/dontdiff >> linux-2.6.32.8/include/net/sctp/structs.h >> p1/include/net/sctp/structs.h >> --- linux-2.6.32.8/include/net/sctp/structs.h 2010-02-09 >> 04:57:19.000000000 -0800 >> +++ p1/include/net/sctp/structs.h ? ? 2010-05-28 10:33:12.000000000 -0700 >> @@ -320,6 +320,9 @@ struct sctp_sock { >> ? ? ? /* Flags controlling Heartbeat, SACK delay, and Path MTU Discovery. */ >> ? ? ? __u32 param_flags; >> >> + ? ? /* Multistream scheduling ?*/ >> + ? ? const struct sctp_sched_ops *sched_ops; >> + >> ? ? ? struct sctp_initmsg initmsg; >> ? ? ? struct sctp_rtoinfo rtoinfo; >> ? ? ? struct sctp_paddrparams paddrparam; >> @@ -534,6 +537,36 @@ static inline void sctp_ssn_skip(struct >> ?} >> > > BTW, is there a reason you decided to through the sched_ops into the association > and not the queue? ?I seems to make more sense to me to put one into the other. > > -vlad > >> ?/* >> + * Interface for adding new SCTP scheduling handlers >> + * This is similar to the pluggable TCP congestion control >> + */ >> +struct sctp_sched_ops { >> + ? ? struct list_head ? ? ? ?list; >> + >> + ? ? /* initialize out_chunk_list (required) */ >> + ? ? int (*init)(struct sctp_outq *q, gfp_t gfp); >> + ? ? /* cleanup out_chunk_list (required) */ >> + ? ? void (*release)(struct sctp_outq *q); >> + ? ? /* enqueue head function (required) */ >> + ? ? void (*enqueue_head_data)(struct sctp_outq *q, struct sctp_chunk *ch); >> + ? ? /* enqueue tail function (required) */ >> + ? ? void (*enqueue_tail_data)(struct sctp_outq *q, struct sctp_chunk *ch); >> + ? ? /* dequeue function (required) */ >> + ? ? struct sctp_chunk* (*dequeue_data)(struct sctp_outq *q); >> + ? ? /* is out_chunk_list empty? (required) */ >> + ? ? int (*is_empty)(struct sctp_outq *q); >> + >> + ? ? char ? ? ? ? ? ?name[SCTP_SCHED_NAME_MAX]; >> + ? ? struct module ? *owner; >> +}; >> + >> +extern int sctp_register_sched(struct sctp_sched_ops *type); >> +extern void sctp_unregister_sched(struct sctp_sched_ops *type); >> + >> +extern void sctp_cleanup_sched(struct sock *sk); >> +extern int sctp_set_sched(struct sock *sk, const char *name); >> + >> +/* >> ? * Pointers to address related SCTP functions. >> ? * (i.e. things that depend on the address family.) >> ? */ >> @@ -1650,6 +1683,9 @@ struct sctp_association { >> ? ? ? /* The largest timeout or RTO value to use in attempting an INIT */ >> ? ? ? unsigned long max_init_timeo; >> >> + ? ? /* Multistream scheduling ?*/ >> + ? ? const struct sctp_sched_ops *sched_ops; >> + >> ? ? ? /* Heartbeat interval: The endpoint sends out a Heartbeat chunk to >> ? ? ? ?* the destination address every heartbeat interval. This value >> ? ? ? ?* will be inherited by all new transports. >> diff -uprN -X linux-2.6.32.8/Documentation/dontdiff >> linux-2.6.32.8/include/net/sctp/user.h p1/include/net/sctp/user.h >> --- linux-2.6.32.8/include/net/sctp/user.h ? ?2010-02-09 04:57:19.000000000 -0800 >> +++ p1/include/net/sctp/user.h ? ? ? ?2010-05-28 10:52:57.000000000 -0700 >> @@ -171,6 +171,8 @@ struct sctp_initmsg { >> ? ? ? __u16 sinit_max_init_timeo; >> ?}; >> >> +#define SCTP_SCHED_NAME_MAX ?16 >> + >> ?/* >> ? * 5.2.2 SCTP Header Information Structure (SCTP_SNDRCV) >> ? * >> -- >> To unsubscribe from this list: send the line "unsubscribe linux-sctp" in >> the body of a message to majordomo@vger.kernel.org >> More majordomo info at ?http://vger.kernel.org/majordomo-info.html >> > -- ======================== Yaogong Wang, PhD candidate Department of Computer Science North Carolina State University http://www4.ncsu.edu/~ywang15/ ======================== -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/