From: Markus Stockhausen <stockhausen@collogia.de>
To: Maxime Ripard <maxime.ripard@free-electrons.com>,
        Neil Brown <neilb@suse.de>, Shaohua Li <shli@fusionio.com>
CC: "linux-raid@vger.kernel.org" <linux-raid@vger.kernel.org>,
        "linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
        Lior Amsalem <alior@marvell.com>,
        Thomas Petazzoni <thomas@free-electrons.com>,
        "Gregory Clement" <gregory.clement@free-electrons.com>,
        Boris Brezillon <boris@free-electrons.com>
Subject: AW: Possible RAID6 regression with ASYNC_TX_DMA enabled in 4.1
Thread-Topic: Possible RAID6 regression with ASYNC_TX_DMA enabled in 4.1
Thread-Index: AQHQiMXKe6lJKtRy0EWiBQ2hzchg251wjuTa
Date: Thu, 7 May 2015 14:39:07 +0000
Message-ID: <12EF8D94C6F8734FB2FF37B9FBEDD1735FC9B959@EXCHANGE.collogia.de>
References: <20150507125702.GI11057@lukather>
In-Reply-To: <20150507125702.GI11057@lukather>
Accept-Language: de-DE, en-US
Content-Language: de-DE
Content-Type: multipart/mixed;
	boundary="----=_NextPartTM-000-cddc7804-b1a2-4d95-8d73-ca17e781a1c2"
MIME-Version: 1.0
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 5228
Lines: 146


This is a multi-part message in MIME format.

------=_NextPartTM-000-cddc7804-b1a2-4d95-8d73-ca17e781a1c2
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: quoted-printable

Hi Maxime,=0A=
=0A=
> Von: linux-raid-owner@vger.kernel.org [linux-raid-owner@vger.kernel.org]&=
quot; im Auftrag von &quot;Maxime Ripard [maxime.ripard@free-electrons.com]=
=0A=
> Gesendet: Donnerstag, 7. Mai 2015 14:57=0A=
> An: Neil Brown; Shaohua Li=0A=
> Cc: linux-raid@vger.kernel.org; linux-kernel@vger.kernel.org; Lior Amsale=
m; Thomas Petazzoni; Gregory Clement; Boris Brezillon=0A=
> Betreff: Possible RAID6 regression with ASYNC_TX_DMA enabled in 4.1=0A=
> =0A=
> Hi,=0A=
> =0A=
> I'm currently trying to add support for the PQ operations on the=0A=
> marvell XOR engine, in dmaengine, obviously to be able to use async_tx=0A=
> to offload these operations.=0A=
> =0A=
> I'm testing these patches with a RAID6 array with 4 disks.=0A=
> =0A=
> However, since the commit 59fc630b8b5f ("RAID5: batch adjacent full=0A=
> stripe write", every write to that array fails with the following=0A=
> stacktrace.=0A=
> =0A=
> http://code.bulix.org/eh8iew-88342?raw=0A=
=0A=
I don't know if it might be related. I added support for RAID6 Read-Modify-=
Write=0A=
in software XOR with some patches. The following commit mangles some lines =
in =0A=
async_pq.c:=0A=
=0A=
https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/commit/?=0A=
id=3D584acdd49cd2472ca0f5a06adbe979db82d0b4af=0A=
=0A=
I introduced a new flag ASYNC_TX_PQ_XOR_DST that notifies the async layer=
=0A=
that we want to do a XOR syndrome operation instead of a full calculation.=
=0A=
This will enforce the software path because I guessed that hardware does no=
t=0A=
support that case. Without hardware to check I might have missed some =0A=
checks in the async layer.=0A=
=0A=
In the upper layer ops_run_reconstruct6 will set the flag if we determined=
=0A=
that rmw is faster than rcw.=0A=
=0A=
Can you check if rmw_level=3D0 fixes the issue. See:=0A=
=0A=
https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/commit/?=0A=
id=3Dd06f191f8ecaef4d524e765fdb455f96392fbd42=0A=
=0A=
> It seems to be generated by that warning here:=0A=
> =0A=
> http://lxr.free-electrons.com/source/crypto/async_tx/async_tx.c#L173=0A=
> =0A=
> And indeed, if we dump the status of depend_tx here, it's already been=0A=
> acked.=0A=
> =0A=
> That doesn't happen if ASYNC_TX_DMA is disabled, hence using the=0A=
> software version of it, instead of relying on our XOR engine. It=0A=
> doesn't happen on any commit prior to the one mentionned above, with=0A=
> the exact same changes applied. These changes are meant to be=0A=
> contributed, so I can definitely push them somewhere if needed.=0A=
> =0A=
> I don't really know where to look for though, the change that is=0A=
> causing this is probably the change in ops_run_reconstruct6, but I'm=0A=
> not sure that this partial revert alone would work with regard to the=0A=
> rest of the patch.=0A=
> =0A=
> Maxime=0A=
=0A=
Markus=0A=
------=_NextPartTM-000-cddc7804-b1a2-4d95-8d73-ca17e781a1c2
Content-Type: text/plain;
	name="InterScan_Disclaimer.txt"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
	filename="InterScan_Disclaimer.txt"

****************************************************************************
Diese E-Mail enthält vertrauliche und/oder rechtlich geschützte
Informationen. Wenn Sie nicht der richtige Adressat sind oder diese E-Mail
irrtümlich erhalten haben, informieren Sie bitte sofort den Absender und
vernichten Sie diese Mail. Das unerlaubte Kopieren sowie die unbefugte
Weitergabe dieser Mail ist nicht gestattet.

Über das Internet versandte E-Mails können unter fremden Namen erstellt oder
manipuliert werden. Deshalb ist diese als E-Mail verschickte Nachricht keine
rechtsverbindliche Willenserklärung.

Collogia
Unternehmensberatung AG
Ubierring 11
D-50678 Köln

Vorstand:
Kadir Akin
Dr. Michael Höhnerbach

Vorsitzender des Aufsichtsrates:
Hans Kristian Langva

Registergericht: Amtsgericht Köln
Registernummer: HRB 52 497

This e-mail may contain confidential and/or privileged information. If you
are not the intended recipient (or have received this e-mail in error)
please notify the sender immediately and destroy this e-mail. Any
unauthorized copying, disclosure or distribution of the material in this
e-mail is strictly forbidden.

e-mails sent over the internet may have been written under a wrong name or
been manipulated. That is why this message sent as an e-mail is not a
legally binding declaration of intention.

Collogia
Unternehmensberatung AG
Ubierring 11
D-50678 Köln

executive board:
Kadir Akin
Dr. Michael Höhnerbach

President of the supervisory board:
Hans Kristian Langva

Registry office: district court Cologne
Register number: HRB 52 497

****************************************************************************

------=_NextPartTM-000-cddc7804-b1a2-4d95-8d73-ca17e781a1c2--
--
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/