From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1757078Ab2AJXQB (ORCPT ); Tue, 10 Jan 2012 18:16:01 -0500 Received: from mail-iy0-f174.google.com ([209.85.210.174]:33630 "EHLO mail-iy0-f174.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1757184Ab2AJXP5 (ORCPT ); Tue, 10 Jan 2012 18:15:57 -0500 Date: Tue, 10 Jan 2012 15:15:52 -0800 From: Tejun Heo To: Konrad Rzeszutek Wilk Cc: linux-kernel@vger.kernel.org, rjw@sisk.pl, xen-devel@lists.xensource.com Subject: Re: Bootup regression introduced by 7bd0b0f0da3b1ec11cbcc798eb0ef747a1184077 ("memblock: Reimplement memblock allocation using reverse free area iterato") in v3.3-rc0 Message-ID: <20120110231552.GB26832@google.com> References: <20120110202838.GA10402@phenom.dumpdata.com> <20120110222625.GA26832@google.com> <20120110224537.GA6572@phenom.dumpdata.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20120110224537.GA6572@phenom.dumpdata.com> User-Agent: Mutt/1.5.20 (2009-06-14) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Hello, On Tue, Jan 10, 2012 at 05:45:37PM -0500, Konrad Rzeszutek Wilk wrote: > (early) [ 0.000000] memblock_find: [0x0, 0xfcdd000) size=8409088 align=4096 nid=1024 > (early) [ 0.000000] memblock_find: [0x805000, 0xfcdd000) - adjusted > (early) [ 0.000000] memblock_find: cand [0x10567000, 0x100000000) -> (early) [0xfcdd000, 0xfcdd000) (early) - rejected > (early) [ 0.000000] memblock_find: cand [0x1e03000, 0x220a000) -> (early) [0x1e03000, 0x220a000) (early) - rejected > (early) [ 0.000000] memblock_find: cand [0x100000, 0x1000000) -> (early) [0x805000, 0x1000000) (early) - rejected > (early) [ 0.000000] memblock_find: cand [0x10000, 0x9b000) -> (early) [0x805000, 0x805000) (early) - rejected > (early) [ 0.000000] Kernel panic - not syncing: Cannot find space for the kernel page tables So, it actually is a legitimate alloc failure. It seems I've tried a bit too hard at simplifying the allocator. Does the following fix the problem? Thanks. diff --git a/mm/memblock.c b/mm/memblock.c index 2f55f19..77b5f22 100644 --- a/mm/memblock.c +++ b/mm/memblock.c @@ -106,14 +106,17 @@ phys_addr_t __init_memblock memblock_find_in_range_node(phys_addr_t start, if (end == MEMBLOCK_ALLOC_ACCESSIBLE) end = memblock.current_limit; - /* adjust @start to avoid underflow and allocating the first page */ - start = max3(start, size, (phys_addr_t)PAGE_SIZE); + /* avoid allocating the first page */ + start = max_t(phys_addr_t, start, PAGE_SIZE); end = max(start, end); for_each_free_mem_range_reverse(i, nid, &this_start, &this_end, NULL) { this_start = clamp(this_start, start, end); this_end = clamp(this_end, start, end); + if (this_end < size) + continue; + cand = round_down(this_end - size, align); if (cand >= this_start) return cand;