From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752729AbcEIT5n (ORCPT ); Mon, 9 May 2016 15:57:43 -0400 Received: from youngberry.canonical.com ([91.189.89.112]:54987 "EHLO youngberry.canonical.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752657AbcEIT5k (ORCPT ); Mon, 9 May 2016 15:57:40 -0400 From: Kamal Mostafa To: linux-kernel@vger.kernel.org, stable@vger.kernel.org, kernel-team@lists.ubuntu.com Cc: Jason Baron , Andrew Morton , Linus Torvalds , Kamal Mostafa Subject: [PATCH 4.2.y-ckt 55/59] mm: update min_free_kbytes from khugepaged after core initialization Date: Mon, 9 May 2016 12:56:13 -0700 Message-Id: <1462823777-8384-56-git-send-email-kamal@canonical.com> X-Mailer: git-send-email 2.7.4 In-Reply-To: <1462823777-8384-1-git-send-email-kamal@canonical.com> References: <1462823777-8384-1-git-send-email-kamal@canonical.com> X-Extended-Stable: 4.2 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org 4.2.8-ckt10 -stable review patch. If anyone has any objections, please let me know. ---8<------------------------------------------------------------ From: Jason Baron commit bc22af74f271ef76b2e6f72f3941f91f0da3f5f8 upstream. Khugepaged attempts to raise min_free_kbytes if its set too low. However, on boot khugepaged sets min_free_kbytes first from subsys_initcall(), and then the mm 'core' over-rides min_free_kbytes after from init_per_zone_wmark_min(), via a module_init() call. Khugepaged used to use a late_initcall() to set min_free_kbytes (such that it occurred after the core initialization), however this was removed when the initialization of min_free_kbytes was integrated into the starting of the khugepaged thread. The fix here is simply to invoke the core initialization using a core_initcall() instead of module_init(), such that the previous initialization ordering is restored. I didn't restore the late_initcall() since start_stop_khugepaged() already sets min_free_kbytes via set_recommended_min_free_kbytes(). This was noticed when we had a number of page allocation failures when moving a workload to a kernel with this new initialization ordering. On an 8GB system this restores min_free_kbytes back to 67584 from 11365 when CONFIG_TRANSPARENT_HUGEPAGE=y is set and either CONFIG_TRANSPARENT_HUGEPAGE_ALWAYS=y or CONFIG_TRANSPARENT_HUGEPAGE_MADVISE=y. Fixes: 79553da293d3 ("thp: cleanup khugepaged startup") Signed-off-by: Jason Baron Acked-by: Kirill A. Shutemov Acked-by: David Rientjes Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds Signed-off-by: Kamal Mostafa --- mm/page_alloc.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/mm/page_alloc.c b/mm/page_alloc.c index 240c22d..6d95dea 100644 --- a/mm/page_alloc.c +++ b/mm/page_alloc.c @@ -6319,7 +6319,7 @@ int __meminit init_per_zone_wmark_min(void) setup_per_zone_inactive_ratio(); return 0; } -module_init(init_per_zone_wmark_min) +core_initcall(init_per_zone_wmark_min) /* * min_free_kbytes_sysctl_handler - just a wrapper around proc_dointvec() so -- 2.7.4