From mboxrd@z Thu Jan 1 00:00:00 1970 From: "Kirill A. Shutemov" Subject: RE: [PATCH] mm: create a separate slab for page->ptl allocation Date: Mon, 4 Nov 2013 12:42:59 +0200 (EET) Message-ID: <20131104104259.86AFEE0090@blue.fi.intel.com> References: <1382442839-7458-1-git-send-email-kirill.shutemov@linux.intel.com> Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <1382442839-7458-1-git-send-email-kirill.shutemov@linux.intel.com> Sender: linux-kernel-owner@vger.kernel.org To: Andrew Morton Cc: Peter Zijlstra , Ingo Molnar , linux-kernel@vger.kernel.org, linux-mm@kvack.org, linux-arch@vger.kernel.org, "Kirill A. Shutemov" List-Id: linux-arch.vger.kernel.org Kirill A. Shutemov wrote: > If DEBUG_SPINLOCK and DEBUG_LOCK_ALLOC are enabled spinlock_t on x86_64 > is 72 bytes. For page->ptl they will be allocated from kmalloc-96 slab, > so we loose 24 on each. An average system can easily allocate few tens > thousands of page->ptl and overhead is significant. > > Let's create a separate slab for page->ptl allocation to solve this. > > Signed-off-by: Kirill A. Shutemov ping? -- Kirill A. Shutemov From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mga02.intel.com ([134.134.136.20]:51485 "EHLO mga02.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751821Ab3KDKnD (ORCPT ); Mon, 4 Nov 2013 05:43:03 -0500 From: "Kirill A. Shutemov" In-Reply-To: <1382442839-7458-1-git-send-email-kirill.shutemov@linux.intel.com> References: <1382442839-7458-1-git-send-email-kirill.shutemov@linux.intel.com> Subject: RE: [PATCH] mm: create a separate slab for page->ptl allocation Content-Transfer-Encoding: 7bit Message-ID: <20131104104259.86AFEE0090@blue.fi.intel.com> Date: Mon, 4 Nov 2013 12:42:59 +0200 (EET) Sender: linux-arch-owner@vger.kernel.org List-ID: To: Andrew Morton Cc: Peter Zijlstra , Ingo Molnar , linux-kernel@vger.kernel.org, linux-mm@kvack.org, linux-arch@vger.kernel.org, "Kirill A. Shutemov" Message-ID: <20131104104259.oE8gw2B-8aA37gKfiM2bn9vEmOAMDd-LEtzhkrha-hU@z> Kirill A. Shutemov wrote: > If DEBUG_SPINLOCK and DEBUG_LOCK_ALLOC are enabled spinlock_t on x86_64 > is 72 bytes. For page->ptl they will be allocated from kmalloc-96 slab, > so we loose 24 on each. An average system can easily allocate few tens > thousands of page->ptl and overhead is significant. > > Let's create a separate slab for page->ptl allocation to solve this. > > Signed-off-by: Kirill A. Shutemov ping? -- Kirill A. Shutemov