From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751591AbaEUDgz (ORCPT ); Tue, 20 May 2014 23:36:55 -0400 Received: from mail.kernel.org ([198.145.19.201]:46316 "EHLO mail.kernel.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750737AbaEUDgx (ORCPT ); Tue, 20 May 2014 23:36:53 -0400 Message-ID: <1400643406.12240.6.camel@kjgkr> Subject: Re: [f2fs-dev] [PATCH] f2fs: avoid crash when trace f2fs_submit_page_mbio event in ra_sum_pages From: Jaegeuk Kim To: Chao Yu Cc: Jaegeuk Kim , linux-fsdevel@vger.kernel.org, linux-kernel@vger.kernel.org, linux-f2fs-devel@lists.sourceforge.net Date: Wed, 21 May 2014 12:36:46 +0900 In-Reply-To: <005801cf70e7$66cb0090$346101b0$@samsung.com> References: <005801cf70e7$66cb0090$346101b0$@samsung.com> Content-Type: multipart/signed; micalg="pgp-sha1"; protocol="application/pgp-signature"; boundary="=-KRDYOVaXrfDKNEFiWIQU" X-Mailer: Evolution 3.2.3-0ubuntu6 Mime-Version: 1.0 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org --=-KRDYOVaXrfDKNEFiWIQU Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable Hi Chao, 2014-05-16 (=EA=B8=88), 17:14 +0800, Chao Yu: > Previously we allocate pages with no mapping in ra_sum_pages(), so we may > encounter a crash in event trace of f2fs_submit_page_mbio where we access > mapping data of the page. >=20 > We'd better allocate pages in bd_inode mapping and invalidate these pages= after > we restore data from pages. It could avoid crash in above scenario. >=20 > Call Trace: > [] ? ftrace_raw_event_f2fs_write_checkpoint+0x80/0x80 [f2fs] > [] f2fs_submit_page_mbio+0x1cb/0x200 [f2fs] > [] restore_node_summary+0x13a/0x280 [f2fs] > [] build_curseg+0x2bd/0x620 [f2fs] > [] build_segment_manager+0x1cb/0x920 [f2fs] > [] f2fs_fill_super+0x535/0x8e0 [f2fs] > [] mount_bdev+0x16a/0x1a0 > [] f2fs_mount+0x1f/0x30 [f2fs] > [] mount_fs+0x36/0x170 > [] vfs_kern_mount+0x55/0xe0 > [] do_mount+0x1e8/0x900 > [] SyS_mount+0x82/0xc0 > [] sysenter_do_call+0x12/0x22 >=20 > Signed-off-by: Chao Yu > --- > fs/f2fs/node.c | 49 ++++++++++++++++++++++++++++--------------------- > 1 file changed, 28 insertions(+), 21 deletions(-) >=20 > diff --git a/fs/f2fs/node.c b/fs/f2fs/node.c > index 3d60d3d..b5cd814 100644 > --- a/fs/f2fs/node.c > +++ b/fs/f2fs/node.c > @@ -1658,13 +1658,16 @@ int recover_inode_page(struct f2fs_sb_info *sbi, = struct page *page) > =20 > /* > * ra_sum_pages() merge contiguous pages into one bio and submit. > - * these pre-readed pages are linked in pages list. > + * these pre-readed pages are alloced in bd_inode's mapping tree. > */ > -static int ra_sum_pages(struct f2fs_sb_info *sbi, struct list_head *page= s, > +static int ra_sum_pages(struct f2fs_sb_info *sbi, struct page **pages, > int start, int nrpages) > { > struct page *page; > + struct inode *inode =3D sbi->sb->s_bdev->bd_inode; > + struct address_space *mapping =3D inode->i_mapping; > int page_idx =3D start; > + int alloced, readed; > struct f2fs_io_info fio =3D { > .type =3D META, > .rw =3D READ_SYNC | REQ_META | REQ_PRIO > @@ -1672,21 +1675,23 @@ static int ra_sum_pages(struct f2fs_sb_info *sbi,= struct list_head *pages, > =20 > for (; page_idx < start + nrpages; page_idx++) { > /* alloc temporal page for read node summary info*/ > - page =3D alloc_page(GFP_F2FS_ZERO); > + page =3D grab_cache_page(mapping, page_idx); > if (!page) > break; > - > - lock_page(page); > - page->index =3D page_idx; > - list_add_tail(&page->lru, pages); > + page_cache_release(page); IMO, we don't need to do like this. Instead, for() { page =3D grab_cache_page(); if (!page) break; page[page_idx] =3D page; f2fs_submit_page_mbio(sbi, page, &fio); } f2fs_submit_merged_bio(sbi, META, READ); return page_idx - start; Afterwards, in restore_node_summry(), lock_page() will wait the end_io for read. ... f2fs_put_page(pages[index], 1); Thanks, > } > =20 > - list_for_each_entry(page, pages, lru) > - f2fs_submit_page_mbio(sbi, page, page->index, &fio); > + alloced =3D page_idx - start; > + readed =3D find_get_pages_contig(mapping, start, alloced, pages); > + BUG_ON(alloced !=3D readed); > + > + for (page_idx =3D 0; page_idx < readed; page_idx++) > + f2fs_submit_page_mbio(sbi, pages[page_idx], > + pages[page_idx]->index, &fio); > =20 > f2fs_submit_merged_bio(sbi, META, READ); > =20 > - return page_idx - start; > + return readed; > } > =20 > int restore_node_summary(struct f2fs_sb_info *sbi, > @@ -1694,11 +1699,11 @@ int restore_node_summary(struct f2fs_sb_info *sbi= , > { > struct f2fs_node *rn; > struct f2fs_summary *sum_entry; > - struct page *page, *tmp; > + struct inode *inode =3D sbi->sb->s_bdev->bd_inode; > block_t addr; > int bio_blocks =3D MAX_BIO_BLOCKS(max_hw_blocks(sbi)); > - int i, last_offset, nrpages, err =3D 0; > - LIST_HEAD(page_list); > + struct page *pages[bio_blocks]; > + int i, index, last_offset, nrpages, err =3D 0; > =20 > /* scan the node segment */ > last_offset =3D sbi->blocks_per_seg; > @@ -1709,29 +1714,31 @@ int restore_node_summary(struct f2fs_sb_info *sbi= , > nrpages =3D min(last_offset - i, bio_blocks); > =20 > /* read ahead node pages */ > - nrpages =3D ra_sum_pages(sbi, &page_list, addr, nrpages); > + nrpages =3D ra_sum_pages(sbi, pages, addr, nrpages); > if (!nrpages) > return -ENOMEM; > =20 > - list_for_each_entry_safe(page, tmp, &page_list, lru) { > + for (index =3D 0; index < nrpages; index++) { > if (err) > goto skip; > =20 > - lock_page(page); > - if (unlikely(!PageUptodate(page))) { > + lock_page(pages[index]); > + if (unlikely(!PageUptodate(pages[index]))) { > err =3D -EIO; > } else { > - rn =3D F2FS_NODE(page); > + rn =3D F2FS_NODE(pages[index]); > sum_entry->nid =3D rn->footer.nid; > sum_entry->version =3D 0; > sum_entry->ofs_in_node =3D 0; > sum_entry++; > } > - unlock_page(page); > + unlock_page(pages[index]); > skip: > - list_del(&page->lru); > - __free_pages(page, 0); > + page_cache_release(pages[index]); > } > + > + invalidate_mapping_pages(inode->i_mapping, addr, > + addr + nrpages); > } > return err; > } --=20 Jaegeuk Kim --=-KRDYOVaXrfDKNEFiWIQU Content-Type: application/pgp-signature; name="signature.asc" Content-Description: This is a digitally signed message part Content-Transfer-Encoding: 7bit -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.11 (GNU/Linux) iQIcBAABAgAGBQJTfB9OAAoJEEAUqH6CSFDSPu8P/iPyD5Oaye5bMapVkyaHnGQ7 NwoJXgwaz2Bc5Y6SBkNB2gv5Ubx2gGV32NIqiRgVRQOj7LESUNT128/ZVvgbMOgk M3Y2n340YhyZJDK39psHxhRvhbe6AblCn3cSiyrDOchKYbeIRaE1od2AGKHoYgbZ aQLaX1MucaV4bNBqII4VXolHykLaYofcJENCfZjE+CLFiO9dCtztjwvqO8RTXRkS Y8zEeSnU0FY2iFAYvVcriVYOiQJ9LSo6Hb+PonwHEschtZjIgJFKlqvjBjJzCeHo GEmEOQmZgwzUuSZ6wnqpsb5qoiVyN1uS2vE4jIJvP2i8lve1o/0WfhK0RCcaXrCw TVJi5OyGS1bQc0YWR+gMG/7aijuh3vy7l5gWsNkM/nY+eWxFFXTi59BWIYzwQVHb j9RIbEbVqCwxpaQW2vvPIyrG1XlLXss/E+bg9sUow5sKeuYKIN8PhYjCGBYS6QIB MCZvHg0WWQGnirT1TgTBAp2vrkPnO2GLM3soUIPzuGV+KWbbGgQ5WjIYlwglXgxC YkTpovdvBg353VCmh3QqAbsnxzIg9utA7KTHhiREsVSjKtJ1AFV7E3pCy0EW24Qz Vbr7IsiMg+27RRAc+oRxhL1Jyz/y0Lh0+RL/jLdtDm7CFqIZ/upvzZ4tfdLRz+sX q1TjpQbPEeB31eQbOqyv =VsdX -----END PGP SIGNATURE----- --=-KRDYOVaXrfDKNEFiWIQU--