Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S932072Ab1FFNQN (ORCPT ); Mon, 6 Jun 2011 09:16:13 -0400 Received: from mail-wy0-f174.google.com ([74.125.82.174]:55303 "EHLO mail-wy0-f174.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1756763Ab1FFNQK (ORCPT ); Mon, 6 Jun 2011 09:16:10 -0400 DomainKey-Signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=subject:from:to:cc:in-reply-to:references:content-type:date :message-id:mime-version:x-mailer:content-transfer-encoding; b=ghy4gasDX8PReblAD06hFrLNgHlhGN1DkR7fKh5H7dhFAxsafdpkgM9ws/jUi5rOGe DL0O1aB425ZCCt30aQODLmMRVLm1ctJQVjhYX1KR7Q6btlI1fIc2/jwPYCXcrTWtJsQf gLaJaLAurMnb8fyZ7mDmCD+8JA6zxGejzDDp4= Subject: Re: bridge/netfilter: regression in 2.6.39.1 From: Eric Dumazet To: Alexander Holler Cc: Neil Horman , linux-kernel@vger.kernel.org, David Miller , Herbert Xu , netdev@vger.kernel.org In-Reply-To: <4DECCCC0.70905@ahsoftware.de> References: <4DE93422.3070000@ahsoftware.de> <20110606111507.GA1000@hmsreliant.think-freely.org> <4DECBEA3.6070408@ahsoftware.de> <1307362358.3098.6.camel@edumazet-laptop> <4DECCCC0.70905@ahsoftware.de> Content-Type: text/plain; charset="UTF-8" Date: Mon, 06 Jun 2011 15:16:06 +0200 Message-ID: <1307366166.3098.15.camel@edumazet-laptop> Mime-Version: 1.0 X-Mailer: Evolution 2.32.2 Content-Transfer-Encoding: 8bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 7606 Lines: 176 Le lundi 06 juin 2011 à 14:49 +0200, Alexander Holler a écrit : > Am 06.06.2011 14:12, schrieb Eric Dumazet: > > Le lundi 06 juin 2011 à 13:48 +0200, Alexander Holler a écrit : > >> Am 06.06.2011 13:15, schrieb Neil Horman: > >>> On Fri, Jun 03, 2011 at 09:21:06PM +0200, Alexander Holler wrote: > >>>> Hello, > >>>> > >>>> I'm getting a oops in the bridge code in br_change_mtu() with > >>>> 2.6.39.1. The patch below seems to fix that. > >>>> > >>>> I'm not sure about the usage of dst_cow_metrics_generic() in > >>>> fake_dst_ops, but after having a quick look at it seems to be ok to > >>>> use that here. > >>>> > >>>> Regards, > >>>> > >>>> Alexander > >>>> > >>> How did the flags of the dst entry on which we're callnig dst_entry_write_ptr > >>> wind up getting the READ_ONLY flag set on it? I don't see how we'er falling > >>> into that clause in which we call cow_metrics when we call dst_metric_set. It > >>> seems like that flag is set erroneously. perhaps we should just update > >>> fake_rtable.dst to have the correct flags? > >>> Neil > >> > >> It is set by that change: > >> > >> -------- > >> @@ -124,7 +128,7 @@ void br_netfilter_rtable_init(struct net_bridge *br) > >> atomic_set(&rt->dst.__refcnt, 1); > >> rt->dst.dev = br->dev; > >> rt->dst.path =&rt->dst; > >> - dst_metric_set(&rt->dst, RTAX_MTU, 1500); > >> + dst_init_metrics(&rt->dst, br_dst_default_metrics, true); > >> rt->dst.flags = DST_NOXFRM; > >> rt->dst.ops =&fake_dst_ops; > >> } > >> -------- > >> > >> The true in dst_init_metrics() is responsible for that flag. > >> > > > > You are aware this change fixed an oops ? > > No, I'm not aware of this. I know almost nothing about what all that > stuff is doing. For me that change above just introduced an oops through > an immediate NULL pointer dereference in br_change_mtu(). > > > read_only in this context means : In case this must be written, we make > > a COW first > > (allocate a piece of memory, copy the source in it before applying any > > change) > > > > It would be nice you send us the stack trace, so that we can have a clue > > of whats going on. > > Here is the text as found in my first mail: > ---- > Commit 42923465fb8d025a2b5153f2e7ab1e6e1058bf00 does here what it > should prevent, it introduces NULL a dereference. > Hmm, this commit uncovers a previous bug, introduced in commit 62fa8a846d7d. > The above commit uses dst_init_metrics() which sets the metrics as > read only. As result br_change_mtu() dies in dst_metric_set() > which calls dst_metrics_write_ptr() which calls > dst->ops->cow_metrics() if the metrics are read only. > > I don't have a system with 2.6.39.1 or 3.0-rcN ready which writes an > oops to somewhere else than the screen. Just set up a bridge and change > the MTU for the IF, that will trigger the oops. > ---- > > Here is the oops: > > ---- > [ 136.546023] BUG: unable to handle kernel NULL pointer dereference at > (null) > [ 136.546038] IP: [< (null)>] (null) > [ 136.546046] *pde = 00000000 > [ 136.546052] Oops: 0000 [#1] SMP > [ 136.546059] last sysfs file: /sys/devices/virtual/net/mybridge/uevent > [ 136.546066] Modules linked in: sit tunnel4 tun iwlagn sch_sfq bridge > stp llc nfs lockd auth_rpcgss nfs_acl sunrpc ipt_LOG xt_recent xt_state > iptable_filter iptable_nat nf_nat nf_conntrack_ipv4 nf_defrag_ipv4 > xt_addrtype xt_dscp xt_iprange xt_DSCP xt_set ip_set nfnetlink ip6t_LOG > xt_limit ip6table_filter xt_string xt_owner xt_multiport xt_hashlimit > xt_conntrack xt_NFQUEUE xt_mark xt_connmark nf_conntrack ip6_tables > snd_seq_oss snd_seq_midi_event snd_seq snd_seq_device snd_pcm_oss > snd_mixer_oss uinput ipv6 dm_mod nvidiafb fb_ddc i2c_algo_bit vgastate > nvidia(P) arc4 uvcvideo videodev ecb snd_hda_codec_realtek iwl4965 > snd_hda_intel iwl_legacy snd_hda_codec uhci_hcd firewire_ohci mac80211 > sdhci_pci snd_hwdep snd_pcm sdhci ehci_hcd intel_agp firewire_core > cfg80211 snd_timer asus_laptop snd e1000e usbcore sparse_keymap rfkill > tpm_infineon crc_itu_t intel_gtt mmc_core iTCO_wdt tpm_tis ata_piix > agpgart tpm video soundcore sg tpm_bios joydev snd_page_alloc [last > unloaded: microcode] > [ 136.546235] > [ 136.546243] Pid: 8415, comm: ip Tainted: P > 2.6.39.1-00006-g40545b7 #103 ASUSTeK Computer Inc. V1Sn > /V1Sn > [ 136.546256] EIP: 0060:[<00000000>] EFLAGS: 00010202 CPU: 0 > [ 136.546268] EIP is at 0x0 > [ 136.546273] EAX: f14a389c EBX: 000005d4 ECX: f80d32c0 EDX: f80d1da1 > [ 136.546279] ESI: f14a3000 EDI: f255bf10 EBP: f15c3b54 ESP: f15c3b48 > [ 136.546285] DS: 007b ES: 007b FS: 00d8 GS: 0033 SS: 0068 > [ 136.546293] Process ip (pid: 8415, ti=f15c2000 task=f4741f80 > task.ti=f15c2000) > [ 136.546297] Stack: > [ 136.546301] f80c658f f14a3000 ffffffed f15c3b64 c12cb9c8 f80d1b80 > ffffffa1 f15c3bbc > [ 136.546315] c12da347 c12d9c7d 00000000 f7670b00 00000000 f80d1b80 > ffffffa6 f15c3be4 > [ 136.546329] 00000004 f14a3000 f255bf20 00000008 f15c3bbc c11d6cae > 00000000 00000000 > [ 136.546343] Call Trace: > [ 136.546359] [] ? br_change_mtu+0x5f/0x80 [bridge] > [ 136.546372] [] dev_set_mtu+0x38/0x80 > [ 136.546381] [] do_setlink+0x1a7/0x860 > [ 136.546390] [] ? rtnl_fill_ifinfo+0x9bd/0xc70 > [ 136.546400] [] ? nla_parse+0x6e/0xb0 > [ 136.546409] [] rtnl_newlink+0x361/0x510 > [ 136.546420] [] ? vmalloc_sync_all+0x100/0x100 > [ 136.546429] [] ? error_code+0x5a/0x60 > [ 136.546438] [] ? rtnl_configure_link+0x80/0x80 > [ 136.546446] [] rtnetlink_rcv_msg+0xfa/0x210 > [ 136.546454] [] ? __rtnl_unlock+0x20/0x20 > [ 136.546463] [] netlink_rcv_skb+0x8e/0xb0 > [ 136.546471] [] rtnetlink_rcv+0x1c/0x30 > [ 136.546479] [] netlink_unicast+0x23a/0x280 > [ 136.546487] [] netlink_sendmsg+0x26b/0x2f0 > [ 136.546497] [] sock_sendmsg+0xc8/0x100 > [ 136.546508] [] ? __alloc_pages_nodemask+0xe1/0x750 > [ 136.546517] [] ? _copy_from_user+0x42/0x60 > [ 136.546525] [] ? verify_iovec+0x4c/0xc0 > [ 136.546534] [] sys_sendmsg+0x1c5/0x200 > [ 136.546542] [] ? __do_fault+0x310/0x410 > [ 136.546549] [] ? do_wp_page+0x1d6/0x6b0 > [ 136.546557] [] ? handle_pte_fault+0xe1/0x720 > [ 136.546565] [] ? sys_getsockname+0x7f/0x90 > [ 136.546574] [] ? handle_mm_fault+0xb1/0x180 > [ 136.546582] [] ? vmalloc_sync_all+0x100/0x100 > [ 136.546589] [] ? do_page_fault+0x173/0x3d0 > [ 136.546596] [] ? sys_recvmsg+0x3b/0x60 > [ 136.546605] [] sys_socketcall+0x293/0x2d0 > [ 136.546614] [] sysenter_do_call+0x12/0x26 > [ 136.546619] Code: Bad EIP value. > [ 136.546627] EIP: [<00000000>] 0x0 SS:ESP 0068:f15c3b48 > [ 136.546645] CR2: 0000000000000000 > [ 136.546652] ---[ end trace 6909b560e78934fa ]--- > ---- > > And if you want the oops on your machine (uisng 2.6.39.1 or 3.0-rcN): > > ---- > brctl addbr mybridge > ip link set mybridge mtu 1234 > oops > ---- Nice, now please submit a patch with 0972ddb237 as a guideline. BTW, you could also check other struct dst_ops methods for bridge/netfilter: - What about ADVMSS or things like that, see commit 214f45c9 -- 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/