From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mga14.intel.com ([192.55.52.115]:22815 "EHLO mga14.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753486AbdK1TXZ (ORCPT ); Tue, 28 Nov 2017 14:23:25 -0500 Message-ID: <1511897003.5796.2.camel@intel.com> Subject: Re: Patch "e1000e: fix buffer overrun while the I219 is processing DMA transactions" has been added to the 4.14-stable tree From: Jeff Kirsher To: Greg KH , "Neftin, Sasha" Cc: aaron.f.brown@intel.com, amit.pundir@linaro.org, dima.ruinskiy@intel.com, raanan.avargil@intel.com, stable@vger.kernel.org, stable-commits@vger.kernel.org Date: Tue, 28 Nov 2017 11:23:23 -0800 In-Reply-To: <20171128101847.GA16587@kroah.com> References: <1511861268206192@kroah.com> <20171128101847.GA16587@kroah.com> Content-Type: multipart/signed; micalg="pgp-sha512"; protocol="application/pgp-signature"; boundary="=-G/CuBNU4k4sP+GKxVGtL" Mime-Version: 1.0 Sender: stable-owner@vger.kernel.org List-ID: --=-G/CuBNU4k4sP+GKxVGtL Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable On Tue, 2017-11-28 at 11:18 +0100, Greg KH wrote: > On Tue, Nov 28, 2017 at 11:56:41AM +0200, Neftin, Sasha wrote: > > On 28/11/2017 11:27, gregkh@linuxfoundation.org wrote: > > > This is a note to let you know that I've just added the patch titled > > >=20 > > > e1000e: fix buffer overrun while the I219 is processing DMA > > > transactions > > >=20 > > > to the 4.14-stable tree which can be found at: > > > http://www.kernel.org/git/?p=3Dlinux/kernel/git/stable/stable-qu= eu > > > e.git;a=3Dsummary > > >=20 > > > The filename of the patch is: > > > e1000e-fix-buffer-overrun-while-the-i219-is-processing-dma- > > > transactions.patch > > > and it can be found in the queue-4.14 subdirectory. > > >=20 > > > If you, or anyone else, feels it should not be added to the stable > > > tree, > > > please let know about it. > > >=20 > > >=20 > > > From b10effb92e272051dd1ec0d7be56bf9ca85ab927 Mon Sep 17 00:00:00 > > > 2001 > > > From: Sasha Neftin > > > Date: Sun, 6 Aug 2017 16:49:18 +0300 > > > Subject: e1000e: fix buffer overrun while the I219 is processing DMA > > > transactions > > > MIME-Version: 1.0 > > > Content-Type: text/plain; charset=3DUTF-8 > > > Content-Transfer-Encoding: 8bit > > >=20 > > > From: Sasha Neftin > > >=20 > > > commit b10effb92e272051dd1ec0d7be56bf9ca85ab927 upstream. > > >=20 > > > Intel=C2=AE 100/200 Series Chipset platforms reduced the round-trip > > > latency for the LAN Controller DMA accesses, causing in some high > > > performance cases a buffer overrun while the I219 LAN Connected > > > Device is processing the DMA transactions. I219LM and I219V devices > > > can fall into unrecovered Tx hang under very stressfully UDP traffic > > > and multiple reconnection of Ethernet cable. This Tx hang of the LAN > > > Controller is only recovered if the system is rebooted. Slightly slow > > > down DMA access by reducing the number of outstanding requests. > > > This workaround could have an impact on TCP traffic performance > > > on the platform. Disabling TSO eliminates performance loss for TCP > > > traffic without a noticeable impact on CPU performance. > > >=20 > > > Please, refer to I218/I219 specification update: > > > https://www.intel.com/content/www/us/en/embedded/products/networking/ > > > ethernet-connection-i218-family-documentation.html > > >=20 > > > Signed-off-by: Sasha Neftin > > > Reviewed-by: Dima Ruinskiy > > > Reviewed-by: Raanan Avargil > > > Tested-by: Aaron Brown > > > Signed-off-by: Jeff Kirsher > > > Signed-off-by: Amit Pundir > > > Signed-off-by: Greg Kroah-Hartman > > >=20 > > > --- > > > drivers/net/ethernet/intel/e1000e/netdev.c | 8 +++++--- > > > 1 file changed, 5 insertions(+), 3 deletions(-) > > >=20 > > > --- a/drivers/net/ethernet/intel/e1000e/netdev.c > > > +++ b/drivers/net/ethernet/intel/e1000e/netdev.c > > > @@ -3021,8 +3021,8 @@ static void e1000_configure_tx(struct e1 > > > hw->mac.ops.config_collision_dist(hw); > > > - /* SPT and CNP Si errata workaround to avoid data corruption > > > */ > > > - if (hw->mac.type >=3D e1000_pch_spt) { > > > + /* SPT and KBL Si errata workaround to avoid data corruption > > > */ > > > + if (hw->mac.type =3D=3D e1000_pch_spt) { > > > u32 reg_val; > > > reg_val =3D er32(IOSFPC); > > > @@ -3030,7 +3030,9 @@ static void e1000_configure_tx(struct e1 > > > ew32(IOSFPC, reg_val); > > > reg_val =3D er32(TARC(0)); > > > - reg_val |=3D E1000_TARC0_CB_MULTIQ_3_REQ; > > > + /* SPT and KBL Si errata workaround to avoid Tx hang > > > */ > > > + reg_val &=3D ~BIT(28); > > > + reg_val |=3D BIT(29); > > > ew32(TARC(0), reg_val); > > > } > > > } > > >=20 > > >=20 > > > Patches currently in stable-queue which might be from sasha.neftin@in > > > tel.com are > > >=20 > > > queue-4.14/e1000e-fix-buffer-overrun-while-the-i219-is-processing- > > > dma-transactions.patch > >=20 > > This is good. To follow of this one we have sent (Jeff sent today) > > another > > one, same functionality but more fundamental and clear patch, please, > > refer > > to https://patchwork.ozlabs.org/patch/834472/. I believe you can take > > it too > > for stable build . >=20 > What is the git commit id of the patch in Linus's tree? >=20 > You all do know how to properly tag patches to be included automatically > in stable releases, right? If so, how come no one does this for this > driver? Sorry Greg, I did not realize the above commit was queued up for stable, which is why the follow on commit that Sasha was referring to was not CC'd to stable mailing list. I do and will usual add the stable mailing list to commits that need to be queued up for the stable trees. Anyways, here is the commit id for the follow on patch for the changes above. David Miller just accepted it into his net tree this morning, here is the commit id info: commit c0f4b163a03e73055dd734eaca64b9580e72e7fb Author: Sasha Neftin Date: Mon Nov 6 08:31:59 2017 +0200 e1000e: fix the use of magic numbers for buffer overrun issue =20 This is a follow on to commit b10effb92e27 ("fix buffer overrun while the I219 is processing DMA transactions") to address David Laights concerns about the use of "magic" numbers. So define masks as well as add additional code comments to give a better understanding of what needs to be done to avoid a buffer overrun. =20 Signed-off-by: Sasha Neftin Reviewed-by: Alexander H Duyck Reviewed-by: Dima Ruinskiy Reviewed-by: Raanan Avargil Tested-by: Aaron Brown Signed-off-by: Jeff Kirsher --=-G/CuBNU4k4sP+GKxVGtL Content-Type: application/pgp-signature; name="signature.asc" Content-Description: This is a digitally signed message part Content-Transfer-Encoding: 7bit -----BEGIN PGP SIGNATURE----- iQIzBAABCgAdFiEEiTyZWz+nnTrOJ1LZ5W/vlVpL7c4FAlodt6sACgkQ5W/vlVpL 7c73QA//VoQi8LbdbWNjG0ZoE1grJd1Q5wK4e18+JsRPAqUxAKSrwf3GSqgLzWLi Z7VAxsQ3NsdTqQ/4my9OMWSZQD5zny6S/1Tfe0hufUrQJUHmANhuwTjLcHDpic6m UlAN1nYYVB8rR2OHYaXDa71+JCvG/fvj40CPoL+kTubQHX39OMcqtEKiJbuiwhRT ZiKCSgQxaSRCUNL4nnuDQN1CrGIxt9NHBjnflNv6C5Kg5FFYAW9HnHxVLn47Zzlu NqjEEAnagvKcOGo8023Cn6YFxfr2vI3c0anU6TVQ9GS2fZig42x68LrtCf08oJZ2 lwd1Yun8hHM+1k19XFthUIU/uDndOPP5n8UMq94fS7pM9XXaKSg3xusWqaTaLtD5 ofjfujVjKq6SUHN6yIjT/VWEiWXtFSjBtWAP9p9xXpnQSy4U7kBpbxNcS8nvZZc4 CPnTKqGtPudaz68xEOu/LcR4OTjQWvZJ+CpOljQGcQvE8WdyqPUi6c5U5Q5csg7a Dp254snKMsoHzJFAjYZz/pClRQ+JrjwBAM4he65bn18K+Dfs90gSQXDO/qR+mk1u GVqlUkAJPE1zOKNQFEM5OnjdWkIT87vYMuDIs7KNLAXxAr8NqtrE8LIMzhGPWmCY 6zwR8XER3rMHgIeCZT6x3Dr67Hbkwm6+X/Xl5i6j7TedNqRMuU8= =ur0y -----END PGP SIGNATURE----- --=-G/CuBNU4k4sP+GKxVGtL--