NetBSD Problem Report #51756

From martin@aprisoft.de  Sat Dec 31 11:24:59 2016
Return-Path: <martin@aprisoft.de>
Received: from mail.netbsd.org (mail.netbsd.org [199.233.217.200])
	(using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits))
	(Client CN "mail.netbsd.org", Issuer "Postmaster NetBSD.org" (verified OK))
	by mollari.NetBSD.org (Postfix) with ESMTPS id 807D97A266
	for <gnats-bugs@gnats.NetBSD.org>; Sat, 31 Dec 2016 11:24:59 +0000 (UTC)
Message-Id: <20161231112448.BA2095CC761@emmas.aprisoft.de>
Date: Sat, 31 Dec 2016 12:24:48 +0100 (CET)
From: martin@NetBSD.org
Reply-To: martin@NetBSD.org
To: gnats-bugs@NetBSD.org
Subject: wd_trim() fails
X-Send-Pr-Version: 3.95

>Number:         51756
>Category:       kern
>Synopsis:       wd_trim() fails
>Confidential:   no
>Severity:       serious
>Priority:       high
>Responsible:    kern-bug-people
>State:          closed
>Class:          sw-bug
>Submitter-Id:   net
>Arrival-Date:   Sat Dec 31 11:25:00 +0000 2016
>Closed-Date:    Tue Jan 03 11:11:26 +0000 2017
>Last-Modified:  Tue Jan 03 11:11:26 +0000 2017
>Originator:     Martin Husemann
>Release:        NetBSD 7.99.53
>Organization:
The NetBSD Foundation, Inc.
>Environment:
System: NetBSD whoever-brings-the-night.aprisoft.de 7.99.53 NetBSD 7.99.53 (WHOEVER) #69: Wed Dec 28 20:52:33 CET 2016 martin@martins.aprisoft.de:/ssd/src/sys/arch/sparc64/compile/WHOEVER sparc64
Architecture: sparc64
Machine: sparc64
>Description:

Mounting a SSD device connected to a siisata(4) controller on sparc64 with
option "discard" causes failures from wd_trim:

wd0: wd_trim: status=0xaa<ERROR>
siisata0: fatal error 5 on channel 3 (ctx 0x400c361b), resetting

I have tested different SSD drives and they all behave the same.
All wd_trim() calls fail.

>How-To-Repeat:
s/a

>Fix:
n/a

>Release-Note:

>Audit-Trail:
From: Martin Husemann <martin@duskware.de>
To: gnats-bugs@NetBSD.org
Cc: 
Subject: Re: kern/51756: wd_trim failure with siisata
Date: Sun, 1 Jan 2017 15:36:24 +0100

 Here is a siisata_debug log showing the failure.

 Martin

 siisata0: siisata_ata_bio.
 siisata0: siisata_bio_start port 3, slot 27
 siisata_dma_setup: 1 segs, 4096 count
 siisata0: siisata_bio_start: done
 siisata0: siisata_intr: GR_GIS: 0x00000008
 siisata0: siisata_intr_port port 3, pis 0x10001 pss 0x0
 siisata0: siisata_bio_complete bcount: 4096 now 0
 siisata0: siisata_ata_bio.
 siisata0: siisata_bio_start port 3, slot 27
 siisata_dma_setup: 1 segs, 32768 count
 siisata0: siisata_bio_start: done
 siisata0: siisata_intr: GR_GIS: 0x00000008
 siisata0: siisata_intr_port port 3, pis 0x10001 pss 0x0
 siisata0: siisata_bio_complete bcount: 32768 now 0
 siisata0: siisata_ata_bio.
 siisata0: siisata_bio_start port 3, slot 27
 siisata_dma_setup: 1 segs, 32768 count
 siisata0: siisata_bio_start: done
 siisata0: siisata_intr: GR_GIS: 0x00000008
 siisata0: siisata_intr_port port 3, pis 0x10001 pss 0x0
 siisata0: siisata_bio_complete bcount: 32768 now 0
 siisata0: siisata_ata_bio.
 siisata0: siisata_bio_start port 3, slot 27
 siisata_dma_setup: 1 segs, 32768 count
 siisata0: siisata_bio_start: done
 siisata0: siisata_intr: GR_GIS: 0x00000008
 siisata0: siisata_intr_port port 3, pis 0x10001 pss 0x0
 siisata0: siisata_bio_complete bcount: 32768 now 0
 siisata0: siisata_ata_bio.
 siisata0: siisata_bio_start port 3, slot 27
 siisata_dma_setup: 1 segs, 32768 count
 siisata0: siisata_bio_start: done
 siisata0: siisata_intr: GR_GIS: 0x00000008
 siisata0: siisata_intr_port port 3, pis 0x10001 pss 0x0
 siisata0: siisata_bio_complete bcount: 32768 now 0
 siisata0: siisata_ata_bio.
 siisata0: siisata_bio_start port 3, slot 27
 siisata_dma_setup: 1 segs, 32768 count
 siisata0: siisata_bio_start: done
 siisata0: siisata_intr: GR_GIS: 0x00000008
 siisata0: siisata_intr_port port 3, pis 0x10001 pss 0x0
 siisata0: siisata_bio_complete bcount: 32768 now 0
 siisata0: siisata_exec_command begins
 siisata0: siisata_cmd_start port 3 drive 0 command 0x6, slot 27
 siisata_dma_setup: 1 segs, 512 count
 siisata0: siisata_cmd_start: done
 siisata0: siisata_exec_command: sleeping
 siisata0: siisata_intr: GR_GIS: 0x00000008
 siisata0: siisata_intr_port port 3, pis 0x20002 ec 5
 siisata0: fatal error 5 on channel 3 (ctx 0x500c361b), resetting
 siisata0: siisata_cmd_complete
 siisata0: siisata_cmd_done flags 0x8a error 0x80
 siisata0: siisata_exec_command ends
 wd0: wd_trim: status=0xaa<ERROR>

From: "Jonathan A. Kollasch" <jakllsch@kollasch.net>
To: gnats-bugs@NetBSD.org
Cc: 
Subject: Re: kern/51756: wd_trim() fails
Date: Sun, 1 Jan 2017 19:53:21 -0600

 ATA's DATA SET MANAGEMENT command didn't exist when the siisata(4) chips
 were implemented.  Could you try the following patch?

 Index: src/sys/dev/ic/siisata.c
 ===================================================================
 RCS file: /cvsroot/src/sys/dev/ic/siisata.c,v
 retrieving revision 1.29
 diff -d -u -a -p -r1.29 siisata.c
 --- src/sys/dev/ic/siisata.c	19 Sep 2016 19:07:53 -0000	1.29
 +++ src/sys/dev/ic/siisata.c	2 Jan 2017 01:43:26 -0000
 @@ -861,6 +861,11 @@ siisata_cmd_start(struct ata_channel *ch
  		siisata_disable_port_interrupt(chp);
  	}

 +	if (ata_c->r_command == ATA_DATA_SET_MANAGEMENT) {
 +		prb->prb_control |= htole16(PRB_CF_PROTOCOL_OVERRIDE);
 +		prb->prb_protocol_override |= htole16(PRB_PO_WRITE);
 +	}
 +
  	/* go for it */
  	siisata_activate_prb(schp, slot);

 Thanks.

From: Martin Husemann <martin@duskware.de>
To: gnats-bugs@NetBSD.org
Cc: 
Subject: Re: kern/51756: wd_trim() fails
Date: Mon, 2 Jan 2017 20:06:17 +0100

 On Mon, Jan 02, 2017 at 01:55:01AM +0000, Jonathan A. Kollasch wrote:
 >  ATA's DATA SET MANAGEMENT command didn't exist when the siisata(4) chips
 >  were implemented.  Could you try the following patch?

 This works fine!

 Martin

From: "Jonathan A. Kollasch" <jakllsch@netbsd.org>
To: gnats-bugs@gnats.NetBSD.org
Cc: 
Subject: PR/51756 CVS commit: src/sys/dev/ic
Date: Tue, 3 Jan 2017 01:30:16 +0000

 Module Name:	src
 Committed By:	jakllsch
 Date:		Tue Jan  3 01:30:16 UTC 2017

 Modified Files:
 	src/sys/dev/ic: siisata.c

 Log Message:
 The SiI3124 was implemented at a time when the ATA/ACS2
 DATA SET MANAGEMENT/TRIM command did not exist.  As such, the (presumably
 immutable) lookup table in the chip does not know that this command
 indicates a data write to follow.  Use the PRB Protocol Override
 functionality to tell the chip the protocol we need for this command.

 For PR kern/51756.


 To generate a diff of this commit:
 cvs rdiff -u -r1.29 -r1.30 src/sys/dev/ic/siisata.c

 Please note that diffs are not public domain; they are subject to the
 copyright notices on the relevant files.

State-Changed-From-To: open->closed
State-Changed-By: martin@NetBSD.org
State-Changed-When: Tue, 03 Jan 2017 11:11:26 +0000
State-Changed-Why:
Fixed, thanks!


>Unformatted:

Home	PR Database Search