From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:58092) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1atKc8-000823-S0 for qemu-devel@nongnu.org; Thu, 21 Apr 2016 15:50:13 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1atKc5-0007dM-Jr for qemu-devel@nongnu.org; Thu, 21 Apr 2016 15:50:12 -0400 Received: from mx2.suse.de ([195.135.220.15]:48363) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1atKc5-0007cm-B7 for qemu-devel@nongnu.org; Thu, 21 Apr 2016 15:50:09 -0400 References: <20160421162348.GA24178@cbox> From: Alexander Graf Message-ID: <57192EED.2040501@suse.de> Date: Thu, 21 Apr 2016 21:50:05 +0200 MIME-Version: 1.0 In-Reply-To: <20160421162348.GA24178@cbox> Content-Type: text/plain; charset=windows-1252 Content-Transfer-Encoding: 7bit Subject: Re: [Qemu-devel] Performance regression using KVM/ARM List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Christoffer Dall , Peter Maydell , "Michael S. Tsirkin" , qemu-devel@nongnu.org Cc: Marc Zyngier On 21.04.16 18:23, Christoffer Dall wrote: > Hi, > > Commit 9fac18f (oslib: allocate PROT_NONE pages on top of RAM, > 2015-09-10) had the unfortunate side effect that memory slots registered > with KVM no longer contain a userspace address that is aligned to a 2M > boundary, causing the use of THP to fail in the kernel. > > I fail to see where in the QEMU code we should be asking for a 2M > alignment of our memory region. Can someone help pointing me to the > right place to fix this or suggest a patch? > > This causes a performance regssion of hackbench on KVM/ARM of about 62% > compared to the workload running with THP. > > We have verified that this is indeed the cause of the failure by adding > various prints to QEMU and the kernel, but unfortunatley my QEMU > knowledge is not sufficient for me to fix it myself. > > Any help would be much appreciated! The code changed quite heavily since I last looked at it, but could you please try whether the (untested) patch below makes a difference? Alex diff --git a/util/mmap-alloc.c b/util/mmap-alloc.c index 0b4cc7f..24e73b1 100644 --- a/util/mmap-alloc.c +++ b/util/mmap-alloc.c @@ -36,7 +36,7 @@ size_t qemu_fd_getpagesize(int fd) } #endif - return getpagesize(); + return 2 * 1024 * 1024; } void *qemu_ram_mmap(int fd, size_t size, size_t align, bool shared)