From owner-freebsd-net@freebsd.org Thu Oct 17 19:31:02 2019 Return-Path: Delivered-To: freebsd-net@mailman.nyi.freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2610:1c1:1:606c::19:1]) by mailman.nyi.freebsd.org (Postfix) with ESMTP id DDCBE15BCF8 for ; Thu, 17 Oct 2019 19:31:02 +0000 (UTC) (envelope-from jacob.e.keller@intel.com) Received: from mga04.intel.com (mga04.intel.com [192.55.52.120]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (Client CN "fmsmga104.fm.intel.com", Issuer "Sectigo RSA Organization Validation Secure Server CA" (verified OK)) by mx1.freebsd.org (Postfix) with ESMTPS id 46vK5d1d7xz4cLH; Thu, 17 Oct 2019 19:31:00 +0000 (UTC) (envelope-from jacob.e.keller@intel.com) X-Amp-Result: SKIPPED(no attachment in message) X-Amp-File-Uploaded: False Received: from orsmga007.jf.intel.com ([10.7.209.58]) by fmsmga104.fm.intel.com with ESMTP/TLS/DHE-RSA-AES256-GCM-SHA384; 17 Oct 2019 12:30:50 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="5.67,308,1566889200"; d="scan'208";a="186591176" Received: from orsmsx108.amr.corp.intel.com ([10.22.240.6]) by orsmga007.jf.intel.com with ESMTP; 17 Oct 2019 12:30:49 -0700 Received: from orsmsx154.amr.corp.intel.com (10.22.226.12) by ORSMSX108.amr.corp.intel.com (10.22.240.6) with Microsoft SMTP Server (TLS) id 14.3.439.0; Thu, 17 Oct 2019 12:30:49 -0700 Received: from orsmsx121.amr.corp.intel.com ([169.254.10.88]) by ORSMSX154.amr.corp.intel.com ([169.254.11.174]) with mapi id 14.03.0439.000; Thu, 17 Oct 2019 12:30:49 -0700 From: "Keller, Jacob E" To: John Baldwin , "freebsd-net@freebsd.org" CC: "shurd@llnw.com" , "Joyner, Eric" Subject: RE: panic on invalid ifp pointer in iflib drivers Thread-Topic: panic on invalid ifp pointer in iflib drivers Thread-Index: AdWEZQwvUgSbd6eoRs2vy4mdxXWSPwA3ezeAAAiyLPA= Date: Thu, 17 Oct 2019 19:30:48 +0000 Message-ID: <02874ECE860811409154E81DA85FBB5896931107@ORSMSX121.amr.corp.intel.com> References: <02874ECE860811409154E81DA85FBB589692E0D4@ORSMSX121.amr.corp.intel.com> <23f1e835-5dbb-055b-3768-f361311a9387@FreeBSD.org> In-Reply-To: <23f1e835-5dbb-055b-3768-f361311a9387@FreeBSD.org> Accept-Language: en-US Content-Language: en-US X-MS-Has-Attach: X-MS-TNEF-Correlator: x-titus-metadata-40: eyJDYXRlZ29yeUxhYmVscyI6IiIsIk1ldGFkYXRhIjp7Im5zIjoiaHR0cDpcL1wvd3d3LnRpdHVzLmNvbVwvbnNcL0ludGVsMyIsImlkIjoiOTJkZDBkMjctM2UxMy00MjgxLTk1N2EtMGQ3NzkyNjczYmUxIiwicHJvcHMiOlt7Im4iOiJDVFBDbGFzc2lmaWNhdGlvbiIsInZhbHMiOlt7InZhbHVlIjoiQ1RQX05UIn1dfV19LCJTdWJqZWN0TGFiZWxzIjpbXSwiVE1DVmVyc2lvbiI6IjE3LjEwLjE4MDQuNDkiLCJUcnVzdGVkTGFiZWxIYXNoIjoiSjZEZWxlbHdDVlBsOTM0cFZDZHJPaWtGQVB0azIwRER2RDk3MEgrOTg0VE9mYm5kT3Q1R1JnckZEZGZiOXplRCJ9 x-ctpclassification: CTP_NT dlp-product: dlpe-windows dlp-version: 11.2.0.6 dlp-reaction: no-action x-originating-ip: [10.22.254.140] Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: quoted-printable MIME-Version: 1.0 X-Rspamd-Queue-Id: 46vK5d1d7xz4cLH X-Spamd-Bar: --------- Authentication-Results: mx1.freebsd.org; dkim=none; dmarc=pass (policy=none) header.from=intel.com; spf=pass (mx1.freebsd.org: domain of jacob.e.keller@intel.com designates 192.55.52.120 as permitted sender) smtp.mailfrom=jacob.e.keller@intel.com X-Spamd-Result: default: False [-9.78 / 15.00]; ARC_NA(0.00)[]; TO_DN_EQ_ADDR_SOME(0.00)[]; RCVD_COUNT_FIVE(0.00)[5]; HAS_XOIP(0.00)[]; FROM_HAS_DN(0.00)[]; RCPT_COUNT_THREE(0.00)[4]; R_SPF_ALLOW(-0.20)[+ip4:192.55.52.120/32]; NEURAL_HAM_LONG(-1.00)[-1.000,0]; MIME_GOOD(-0.10)[text/plain]; IP_SCORE(-3.78)[ip: (-9.91), ipnet: 192.55.52.0/24(-4.96), asn: 4983(-3.96), country: US(-0.05)]; NEURAL_HAM_MEDIUM(-1.00)[-1.000,0]; TO_DN_SOME(0.00)[]; TO_MATCH_ENVRCPT_SOME(0.00)[]; RCVD_IN_DNSWL_MED(-0.20)[120.52.55.192.list.dnswl.org : 127.0.9.2]; DMARC_POLICY_ALLOW(-0.50)[intel.com,none]; FROM_EQ_ENVFROM(0.00)[]; R_DKIM_NA(0.00)[]; MIME_TRACE(0.00)[0:+]; ASN(0.00)[asn:4983, ipnet:192.55.52.0/24, country:US]; RCVD_TLS_LAST(0.00)[]; WHITELIST_SPF_DKIM(-3.00)[intel.com:s:+] X-BeenThere: freebsd-net@freebsd.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: Networking and TCP/IP with FreeBSD List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Thu, 17 Oct 2019 19:31:02 -0000 > -----Original Message----- > From: John Baldwin > Sent: Thursday, October 17, 2019 9:31 AM > To: Keller, Jacob E ; freebsd-net@freebsd.org > Cc: shurd@llnw.com; Joyner, Eric > Subject: Re: panic on invalid ifp pointer in iflib drivers >=20 > On 10/16/19 2:16 PM, Keller, Jacob E wrote: > > Hi, > > > > I'm investigating an issue on the iflib ixl driver in 11.3-RELEASE as w= ell as 12- > RELEASE. We found a panic in that occurs if SCTP/IPv6 traffic is being tr= ansmitted > while the device is detached: > > > > Fatal trap 12: page fault while in kernel mode > > cpuid =3D 0; apic id =3D 00 > > fault virtual address =3D 0xfffffe0000411e38 > > fault code =3D supervisor read data, page not present > > instruction pointer =3D 0x20:0xffffffff80c84700 > > stack pointer =3D 0x28:0xfffffe2f4351b600 > > frame pointer =3D 0x28:0xfffffe2f4351b650 > > code segment =3D base 0x0, limit 0xfffff, type 0x1b > > =3D DPL 0, pres 1, long 1, def32 0, gran 1 > > processor eflags =3D interrupt enabled, resume, IOPL =3D 0 > > current process =3D 12 (swi4: clock (0)) > > trap number =3D 12 > > panic: page fault > > cpuid =3D 0 > > KDB: stack backtrace: > > db_trace_self_wrapper() at db_trace_self_wrapper+0x2b/frame > 0xfffffe2f4351b2c0 > > vpanic() at vpanic+0x17e/frame 0xfffffe2f4351b320 > > panic() at panic+0x43/frame 0xfffffe2f4351b380 > > trap_fatal() at trap_fatal+0x369/frame 0xfffffe2f4351b3d0 > > trap_pfault() at trap_pfault+0x62/frame 0xfffffe2f4351b420 > > trap() at trap+0x2b3/frame 0xfffffe2f4351b530 > > calltrap() at calltrap+0x8/frame 0xfffffe2f4351b530 > > --- trap 0xc, rip =3D 0xffffffff80c84700, rsp =3D 0xfffffe2f4351b600, r= bp =3D > 0xfffffe2f4351b650 --- > > in6_selecthlim() at in6_selecthlim+0x20/frame 0xfffffe2f4351b650 > > sctp_lowlevel_chunk_output() at sctp_lowlevel_chunk_output+0xeb2/frame > 0xfffffe2f4351b790 > > sctp_chunk_output() at sctp_chunk_output+0x68c/frame 0xfffffe2f4351c110 > > sctp_timeout_handler() at sctp_timeout_handler+0x2d8/frame > 0xfffffe2f4351c180 > > softclock_call_cc() at softclock_call_cc+0x15b/frame 0xfffffe2f4351c230 > > softclock() at softclock+0x7c/frame 0xfffffe2f4351c260 > > intr_event_execute_handlers() at intr_event_execute_handlers+0x9a/frame > 0xfffffe2f4351c2a0 > > ithread_loop() at ithread_loop+0xb7/frame 0xfffffe2f4351c2f0 > > fork_exit() at fork_exit+0x84/frame 0xfffffe2f4351c330 > > fork_trampoline() at fork_trampoline+0xe/frame 0xfffffe2f4351c330 > > --- trap 0, rip =3D 0, rsp =3D 0, rbp =3D 0 --- > > KDB: enter: panic > > > > > > From what I've gathered so far, it appears that the issue is a use-afte= r-free > where the SCTP stack gets an ifp pointer that's no longer valid. We've re= produced > this issue on multiple iflib-based drivers, including ixl and the recentl= y published > ice driver code (available on phabricator). > > > > Additionally, we cannot reproduce it on legacy-stack drivers for ixl, o= r a > mellanox 100G board we have. This leads me to believe that it's an issue = in iflib > rather than in the specific device drivers. > > > > I am not sure exactly what's going wrong here... anyone have suggestion= s? I > thought it might be an issue of when ether_ifdetach is called. That funct= ion is > supposed to clear all of the pre-existing routes from the route entry lis= t. I'm > thinking maybe somehow a route gets added after ether_ifdetach is called. > > > > In the iflib_device_deregister function, ether_ifdetach is called just = after > iflib_stop, (which would call a device's if_stop routine), and then the t= ask queues > are shutdown, a driver's ifdi_detach handler is called, and the ifp is fr= ee'd at the > end. In the ixl legacy driver, ether_ifdetach is called prior to the stop= routine. > However, in the mlx5 driver, it's called after a call to close_locked()..= . > > > > So I'm really not sure exactly what could cause a stale ifp pointer to = get into the > route entry list. >=20 > Nominally, ifnet drivers should call ether_ifdetach first to remove publi= c > references to the ifnet and only call their stop routine after that has r= eturned. > This ensures any open if_ioctl invocations have completed, etc. before th= e > stop routine is invoked. Otherwise you are open to a race where the inte= face > can be upped via an ioctl after you have stopped the hardware. >=20 So we should (a) move ether_ifdetach before the stop, and... > Any other references to the ifnet via eventhandlers, etc. should also be > deregistered before calling the stop routine. >=20 (b) make sure all of the event handlers are moved before the stop too. > After the hardware is stopped, interrupt handlers should be torn down and > callouts > and tasks drained to ensure there are no other references to the ifp outs= ide of > the thread running detach. >=20 > After that you can release device resources, destroy mutexes, free the if= p, etc. > Note that drivers have to be prepared for ether_ifdetach to invoke if_ioc= tl (e.g. > when detaching bpf), but of the drivers I've looked at this has generally= been a > non-issue. >=20 > It sounds like iflib should be doing the detach before calling iflib_stop= . >=20 Right. I think also the VLAN event handlers need to be moved too. Let me give that a try out to see if it fixes things. Thanks, Jake > -- > John Baldwin