From mboxrd@z Thu Jan 1 00:00:00 1970 From: Theodore Tso Subject: Re: [PATCH] ext4: slab caches set to SLAB_MEM_SPREAD flags. Date: Wed, 16 Nov 2011 10:51:59 -0500 Message-ID: <47270FB7-96AC-4F41-98D3-9AD469080DBE@mit.edu> References: <1321455848-1929-1-git-send-email-linkinjeon@gmail.com> Mime-Version: 1.0 (Apple Message framework v1251.1) Content-Type: text/plain; charset=windows-1252 Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: Theodore Tso , linux-kernel@vger.kernel.org, linux-ext4@vger.kernel.org, Amit Sahrawat To: Namjae Jeon Return-path: Received: from DMZ-MAILSEC-SCANNER-1.MIT.EDU ([18.9.25.12]:53346 "EHLO dmz-mailsec-scanner-1.mit.edu" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751517Ab1KPPwI convert rfc822-to-8bit (ORCPT ); Wed, 16 Nov 2011 10:52:08 -0500 In-Reply-To: <1321455848-1929-1-git-send-email-linkinjeon@gmail.com> Sender: linux-ext4-owner@vger.kernel.org List-ID: On Nov 16, 2011, at 10:04 AM, Namjae Jeon wrote: > If slab caches set to SLAB_MEM_SPREAD flags, The allocation is spread= evenly over all the memory nodes instead of favoring allocation on the= node local to current cpu. And why do you think this is a good thing? For mballoc in particular,= the data structures are used immediately and then freed immediately --= - on the local node, so using a non-local memory just makes things wors= e in a NUMA system. =46or the pageio, they are used immediately, and may end up getting fre= ed on the same CPU if the I/O interrupt (and hence competion callbacks)= happen on the same CPU that they are submitted on=85. Regards, -- Ted -- To unsubscribe from this list: send the line "unsubscribe linux-ext4" i= n the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html