Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751730AbbKOM7X (ORCPT ); Sun, 15 Nov 2015 07:59:23 -0500 Received: from mail.parknet.co.jp ([210.171.160.6]:52325 "EHLO mail.parknet.co.jp" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751379AbbKOM7U (ORCPT ); Sun, 15 Nov 2015 07:59:20 -0500 From: OGAWA Hirofumi To: Andrew Morton Cc: Vegard Nossum , LKML , Richard Weinberger Subject: Re: Endless getdents() in vfat filesystem References: <56468C1E.2060401@oracle.com> <56472C2F.5090207@oracle.com> <87r3js3baj.fsf@mail.parknet.co.jp> <56474DDB.7020904@nod.at> <87egfs30l8.fsf@mail.parknet.co.jp> <56486709.7000004@oracle.com> <56486B59.9010400@nod.at> Date: Sun, 15 Nov 2015 21:59:15 +0900 In-Reply-To: <56486B59.9010400@nod.at> (Richard Weinberger's message of "Sun, 15 Nov 2015 12:24:09 +0100") Message-ID: <87a8qf2zbw.fsf@mail.parknet.co.jp> User-Agent: Gnus/5.13 (Gnus v5.13) Emacs/25.0.50 (gnu/linux) MIME-Version: 1.0 Content-Type: text/plain Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 3171 Lines: 102 Richard Weinberger writes: >> It would be nice to have a proper patch description too. How about this? >> >> """ >> For the root directory, . and .. are faked (using dir_emit_dots()) and >> ctx->pos is reset from 2 to 0. >> >> A corrupted root directory could cause fat_get_entry() to fail, but >> ->iterate() (fat_readdir()) reports progress to the VFS (with ctx->pos >> rewound to 0), so any following calls to ->iterate() continue to return >> the same entries again and again. >> >> The result is that userspace will never see the end of the directory, >> causing e.g. 'ls' to hang in a getdents() loop. >> """ > > Agreed. And you deserve also a Reported-and-tested-by. :-) Sounds good, updated patch here. -- OGAWA Hirofumi [PATCH v3] fat: Fix fake_offset handling on error path For the root directory, . and .. are faked (using dir_emit_dots()) and ctx->pos is reset from 2 to 0. A corrupted root directory could cause fat_get_entry() to fail, but ->iterate() (fat_readdir()) reports progress to the VFS (with ctx->pos rewound to 0), so any following calls to ->iterate() continue to return the same entries again and again. The result is that userspace will never see the end of the directory, causing e.g. 'ls' to hang in a getdents() loop. [hirofumi@mail.parknet.co.jp: cleanup and make sure to correct fake_offset] Cc: stable@vger.kernel.org Reported-and-tested-by: Vegard Nossum Signed-off-by: Richard Weinberger Signed-off-by: OGAWA Hirofumi --- fs/fat/dir.c | 16 +++++++++++----- 1 file changed, 11 insertions(+), 5 deletions(-) diff -puN fs/fat/dir.c~fat-fake_offset-fix fs/fat/dir.c --- linux/fs/fat/dir.c~fat-fake_offset-fix 2015-11-14 23:31:45.904700155 +0900 +++ linux-hirofumi/fs/fat/dir.c 2015-11-15 02:45:13.861256766 +0900 @@ -610,9 +610,9 @@ parse_record: int status = fat_parse_long(inode, &cpos, &bh, &de, &unicode, &nr_slots); if (status < 0) { - ctx->pos = cpos; + bh = NULL; ret = status; - goto out; + goto end_of_dir; } else if (status == PARSE_INVALID) goto record_end; else if (status == PARSE_NOT_LONGNAME) @@ -654,8 +654,9 @@ parse_record: fill_len = short_len; start_filldir: - if (!fake_offset) - ctx->pos = cpos - (nr_slots + 1) * sizeof(struct msdos_dir_entry); + ctx->pos = cpos - (nr_slots + 1) * sizeof(struct msdos_dir_entry); + if (fake_offset && ctx->pos < 2) + ctx->pos = 2; if (!memcmp(de->name, MSDOS_DOT, MSDOS_NAME)) { if (!dir_emit_dot(file, ctx)) @@ -681,14 +682,19 @@ record_end: fake_offset = 0; ctx->pos = cpos; goto get_new; + end_of_dir: - ctx->pos = cpos; + if (fake_offset && cpos < 2) + ctx->pos = 2; + else + ctx->pos = cpos; fill_failed: brelse(bh); if (unicode) __putname(unicode); out: mutex_unlock(&sbi->s_lock); + return ret; } _ -- 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/