Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1754189AbYJ1MKA (ORCPT ); Tue, 28 Oct 2008 08:10:00 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1754059AbYJ1MJp (ORCPT ); Tue, 28 Oct 2008 08:09:45 -0400 Received: from TYO202.gate.nec.co.jp ([202.32.8.206]:63219 "EHLO tyo202.gate.nec.co.jp" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753922AbYJ1MJo (ORCPT ); Tue, 28 Oct 2008 08:09:44 -0400 Date: Tue, 28 Oct 2008 21:07:31 +0900 From: Daisuke Nishimura To: KAMEZAWA Hiroyuki Cc: nishimura@mxp.nes.nec.co.jp, "linux-kernel@vger.kernel.org" , "linux-mm@kvack.org" , "balbir@linux.vnet.ibm.com" , "akpm@linux-foundation.org" , "menage@google.com" , "xemul@openvz.org" Subject: Re: [PATCH 3/4][mmotm] memcg: fix gfp_mask of callers of charge Message-Id: <20081028210731.4453bbc0.nishimura@mxp.nes.nec.co.jp> In-Reply-To: <20081028191449.22a79033.kamezawa.hiroyu@jp.fujitsu.com> References: <20081028190911.6857b0a6.kamezawa.hiroyu@jp.fujitsu.com> <20081028191449.22a79033.kamezawa.hiroyu@jp.fujitsu.com> Organization: NEC Soft, Ltd. X-Mailer: Sylpheed 2.4.8 (GTK+ 2.10.14; i686-pc-mingw32) Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 5122 Lines: 142 On Tue, 28 Oct 2008 19:14:49 +0900, KAMEZAWA Hiroyuki wrote: > fix misuse of gfp_kernel. > > Now, most of callers of mem_cgroup_charge_xxx functions uses GFP_KERNEL. > > I think that this is from the fact that page_cgroup *was* dynamically allocated. > > But now, we allocate all page_cgroup at boot. And mem_cgroup_try_to_free_pages() > reclaim memory from GFP_HIGHUSER_MOVABLE + specified GFP_RECLAIM_MASK. > * This is because we just want to reduce memory usage. > "Where we should reclaim from ?" is not a problem in memcg. > > This patch modifies gfp masks to be GFP_HIGUSER_MOVABLE if possible. > Note: This patch is not for fixing behavior but for showing sane information > in source code. > > Signed-off-by: KAMEZAWA Hiroyuki > > > mm/memcontrol.c | 8 +++++--- > mm/memory.c | 9 +++++---- > mm/shmem.c | 6 +++--- > mm/swapfile.c | 2 +- > 4 files changed, 14 insertions(+), 11 deletions(-) > > Index: mmotm-2.6.28rc2+/mm/memcontrol.c > =================================================================== > --- mmotm-2.6.28rc2+.orig/mm/memcontrol.c > +++ mmotm-2.6.28rc2+/mm/memcontrol.c > @@ -808,8 +808,9 @@ int mem_cgroup_prepare_migration(struct > } > unlock_page_cgroup(pc); > if (mem) { > - ret = mem_cgroup_charge_common(newpage, NULL, GFP_KERNEL, > - ctype, mem); > + ret = mem_cgroup_charge_common(newpage, NULL, > + GFP_HIGHUSER_MOVABLE, > + ctype, mem); > css_put(&mem->css); > } > return ret; > @@ -888,7 +889,8 @@ int mem_cgroup_resize_limit(struct mem_c > ret = -EBUSY; > break; > } > - progress = try_to_free_mem_cgroup_pages(memcg, GFP_KERNEL); > + progress = try_to_free_mem_cgroup_pages(memcg, > + GFP_HIGHUSER_MOVABLE); > if (!progress) > retry_count--; > } > Index: mmotm-2.6.28rc2+/mm/memory.c > =================================================================== > --- mmotm-2.6.28rc2+.orig/mm/memory.c > +++ mmotm-2.6.28rc2+/mm/memory.c > @@ -1889,7 +1889,7 @@ gotten: > cow_user_page(new_page, old_page, address, vma); > __SetPageUptodate(new_page); > > - if (mem_cgroup_newpage_charge(new_page, mm, GFP_KERNEL)) > + if (mem_cgroup_newpage_charge(new_page, mm, GFP_HIGHUSER_MOVABLE)) > goto oom_free_new; > > /* > @@ -2324,7 +2324,7 @@ static int do_swap_page(struct mm_struct > lock_page(page); > delayacct_clear_flag(DELAYACCT_PF_SWAPIN); > > - if (mem_cgroup_try_charge(mm, GFP_KERNEL, &ptr) == -ENOMEM) { > + if (mem_cgroup_try_charge(mm, GFP_HIGHUSER_MOVABLE, &ptr) == -ENOMEM) { > ret = VM_FAULT_OOM; > unlock_page(page); > goto out; > @@ -2405,7 +2405,7 @@ static int do_anonymous_page(struct mm_s > goto oom; > __SetPageUptodate(page); > > - if (mem_cgroup_newpage_charge(page, mm, GFP_KERNEL)) > + if (mem_cgroup_newpage_charge(page, mm, GFP_HIGHUSER_MOVABLE)) > goto oom_free_page; > > entry = mk_pte(page, vma->vm_page_prot); > @@ -2498,7 +2498,8 @@ static int __do_fault(struct mm_struct * > ret = VM_FAULT_OOM; > goto out; > } > - if (mem_cgroup_newpage_charge(page, mm, GFP_KERNEL)) { > + if (mem_cgroup_newpage_charge(page, > + mm, GFP_HIGHUSER_MOVABLE)) { > ret = VM_FAULT_OOM; > page_cache_release(page); > goto out; > Index: mmotm-2.6.28rc2+/mm/swapfile.c > =================================================================== > --- mmotm-2.6.28rc2+.orig/mm/swapfile.c > +++ mmotm-2.6.28rc2+/mm/swapfile.c > @@ -535,7 +535,7 @@ static int unuse_pte(struct vm_area_stru > pte_t *pte; > int ret = 1; > > - if (mem_cgroup_try_charge(vma->vm_mm, GFP_KERNEL, &ptr)) > + if (mem_cgroup_try_charge(vma->vm_mm, GFP_HIGHUSER_MOVABLE, &ptr)) > ret = -ENOMEM; > > pte = pte_offset_map_lock(vma->vm_mm, pmd, addr, &ptl); > Index: mmotm-2.6.28rc2+/mm/shmem.c > =================================================================== > --- mmotm-2.6.28rc2+.orig/mm/shmem.c > +++ mmotm-2.6.28rc2+/mm/shmem.c > @@ -920,8 +920,8 @@ found: > error = 1; > if (!inode) > goto out; > - /* Precharge page using GFP_KERNEL while we can wait */ > - error = mem_cgroup_cache_charge(page, current->mm, GFP_KERNEL); > + /* Charge page using GFP_HIGHUSER_MOVABLE while we can wait */ > + error = mem_cgroup_cache_charge(page, current->mm, GFP_HIGHUSER_MOVABLE); This line exceeds 80 characters. Looks good to me except for that. Reviewed-by: Daisuke Nishimura > if (error) > goto out; > error = radix_tree_preload(GFP_KERNEL); > @@ -1371,7 +1371,7 @@ repeat: > > /* Precharge page while we can wait, compensate after */ > error = mem_cgroup_cache_charge(filepage, current->mm, > - gfp & ~__GFP_HIGHMEM); > + GFP_HIGHUSER_MOVABLE); > if (error) { > page_cache_release(filepage); > shmem_unacct_blocks(info->flags, 1); > -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/