From mboxrd@z Thu Jan 1 00:00:00 1970 From: Keir Fraser Subject: Re: [PATCH] only set scheduler timer for non-idle CPU Date: Thu, 02 Apr 2009 14:06:27 +0100 Message-ID: References: <49D4B41C.9030901@mytum.de> Mime-Version: 1.0 Content-Type: multipart/mixed; boundary="===============0831733790==" Return-path: In-Reply-To: <49D4B41C.9030901@mytum.de> List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Sender: xen-devel-bounces@lists.xensource.com Errors-To: xen-devel-bounces@lists.xensource.com To: Thomas Pfeuffer , "Yu, Ke" , "xen-devel@lists.xensource.com" List-Id: xen-devel@lists.xenproject.org > This message is in MIME format. Since your mail reader does not understand this format, some or all of this message may not be legible. --===============0831733790== Content-type: multipart/alternative; boundary="B_3321525996_3447108" > This message is in MIME format. Since your mail reader does not understand this format, some or all of this message may not be legible. --B_3321525996_3447108 Content-type: text/plain; charset="ISO-8859-1" Content-transfer-encoding: quoted-printable On 02/04/2009 13:48, "Thomas Pfeuffer" wrote: > Hello Ke,=20 >> =20 >> It is not necessary to set scheduler timer for idle CPU. so this patch a= dd >> conditional check for idle CPU. >> =20 > I think your patch is not good in case sedf-scheduler is used. If idle VC= PU is > the current "running" VCPU, the scheduler timer is set to the next "perio= d > begin" of the first VCPU in the wait queue. > Your patch prevents sedf from taking the VCPUs waiting for their next per= iod > into the runnable queue again. It=B9s probably cleaner always to respect the specific scheduler=B9s scheduling quantum in schedule.c, but then have sched_credit.c return a very large quantum (large as possible) for the idle VCPU. -- Keir --B_3321525996_3447108 Content-type: text/html; charset="ISO-8859-1" Content-transfer-encoding: quoted-printable Re: [Xen-devel] [PATCH] only set scheduler timer for non-idle CPU</T= ITLE> </HEAD> <BODY> <FONT FACE=3D"Calibri, Verdana, Helvetica, Arial"><SPAN STYLE=3D'font-size:11pt= '>On 02/04/2009 13:48, "Thomas Pfeuffer" <<a href=3D"thomas.pfeuf= fer@mytum.de">thomas.pfeuffer@mytum.de</a>> wrote:<BR> <BR> </SPAN></FONT><BLOCKQUOTE><FONT FACE=3D"Calibri, Verdana, Helvetica, Arial"><= SPAN STYLE=3D'font-size:11pt'>Hello Ke, <BR> </SPAN></FONT><BLOCKQUOTE><FONT FACE=3D"Calibri, Verdana, Helvetica, Arial"><= SPAN STYLE=3D'font-size:11pt'> <BR> It is not necessary to set scheduler timer for idle CPU. so this patch add = conditional check for idle CPU.<BR>   <BR> </SPAN></FONT></BLOCKQUOTE><FONT FACE=3D"Calibri, Verdana, Helvetica, Arial">= <SPAN STYLE=3D'font-size:11pt'>I think your patch is not good in case sedf-sch= eduler is used. If idle VCPU is the current "running" VCPU, the sc= heduler timer is set to the next "period begin" of the first VCPU = in the wait queue.<BR> Your patch prevents sedf from taking the VCPUs waiting for their next perio= d into the runnable queue again.<BR> </SPAN></FONT></BLOCKQUOTE><FONT FACE=3D"Calibri, Verdana, Helvetica, Arial">= <SPAN STYLE=3D'font-size:11pt'><BR> It’s probably cleaner always to respect the specific scheduler’= s scheduling quantum in schedule.c, but then have sched_credit.c return a ve= ry large quantum (large as possible) for the idle VCPU.<BR> <BR>  -- Keir<BR> </SPAN></FONT> </BODY> </HTML> --B_3321525996_3447108-- --===============0831733790== Content-Type: text/plain; charset="us-ascii" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit Content-Disposition: inline _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel --===============0831733790==--