From mboxrd@z Thu Jan 1 00:00:00 1970 From: "Chen, Kenneth W" Date: Thu, 20 Apr 2006 17:48:09 +0000 Subject: RE: [RFC] - Kernel text replication on IA64 Message-Id: <4t153d$ofhg7@azsmga001.ch.intel.com> List-Id: In-Reply-To: <20060420164111.GA18770@agluck-lia64.sc.intel.com> References: <20060420135315.GA28021@sgi.com> In-Reply-To: <20060420135315.GA28021@sgi.com> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: "Luck, Tony" , Jack Steiner Cc: linux-ia64@vger.kernel.org, lee.schermerhorn@hp.com, clameter@sgi.com, linux-mm@kvack.org Luck, Tony wrote on Thursday, April 20, 2006 9:41 AM > On Thu, Apr 20, 2006 at 08:53:16AM -0500, Jack Steiner wrote: > > Enabling replication reserves 1 additional DTLB entry for kernel code. > > This reduces the number of DTLB entries that is available for user code. > > There is the potential that this could impact some applications. > > Additional measurements are still needed. > > Ken's recent patch to free up the DTLB that is currently used for per-cpu > data would mitigate this (though I'm sure he'll be unamused if I blow the > 1.6% gain he saw on his transaction processing benchmark on this :-) How much benefit is there to have readonly section replicated? Do you really have to use two DTRs - one to map the readonly and one to map rw? What about just replicate text so we don't need to burn an extra DTR? - Ken