netdev - Re: 2.6.23-mm1 s390 driver problem

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [thread-next>] [day] [month] [year] [list]
Message-Id: <20071019022719.537ee71f.akpm@linux-foundation.org>
Date:	Fri, 19 Oct 2007 02:27:19 -0700
From:	Andrew Morton <akpm@...ux-foundation.org>
To:	Cedric Le Goater <clg@...ibm.com>
Cc:	schwidefsky@...ibm.com, "Serge E. Hallyn" <serge@...lyn.com>,
	Christian Borntraeger <borntraeger@...ibm.com>,
	linux-kernel@...r.kernel.org, cornelia.huck@...ibm.com,
	heiko.carstens@...ibm.com, sam@...nborg.org,
	Linux Containers <containers@...ts.osdl.org>,
	netdev@...r.kernel.org
Subject: Re: 2.6.23-mm1 s390 driver problem

On Fri, 19 Oct 2007 11:16:16 +0200 Cedric Le Goater <clg@...ibm.com> wrote:

> Martin Schwidefsky wrote:
> > On Thu, 2007-10-18 at 15:31 -0500, Serge E. Hallyn wrote:
> >> Quoting Christian Borntraeger (borntraeger@...ibm.com):
> >>> Am Donnerstag, 18. Oktober 2007 schrieb Serge E. Hallyn:
> >>>> Sigh, well this turned out less informative than I'd liked.
> >>>> After bisecting 2.6.23 to 2.6.23-mm1, I found that
> >>>> git-s390.patch is the one breaking my s390 boot :(
> >>>> (Frown bc it's a conglomeration of patches0
> >>>>
> >>>> Symptom is:
> >>>> 	"Cannot open root device "dasdd2" or unknown-block(94,14)"
> >>>> even though dasdd2 appeared to be found earlier in the boot.  I also
> >>>> get
> >>> Can you post the full console output from IPL to the unsuccessful end?
> >> Yeah, sorry, appended below.
> >>
> >> I had thought that the line
> >> 	sysctl table check failed: /sunrpc/transports .7249.14 Missing strategy
> >> meant that the fix referenced in http://lkml.org/lkml/2007/10/11/48
> >> would fix it, but it appeared to have no effect.
> > 
> > This is the vmlinux.lds.S problem. The cleanup patch from Sam Ravnborg
> > moved the __initramfs_start and __initramfs_end symbols into
> > the .init.ramfs section. This is in itself not a problem, but it
> > surfaced a bug: there is no *(.init.initramfs), that needs to be
> > *(init.ramfs). I corrected this in the upstream patch but 2.6.23-mm1 has
> > the older one that still causes the "Cannot open root device". For
> > 2.6.23-mm1 use the patch below.
> > 
> 
> thanks martin, 
> 
> that helped going a little further in the boot process but we then have 
> a network issue when bringing the network interface up :

please cc netdev on network issues.

> Bringing up interface eth0:  ------------Ý cut here ¨------------ 
> Kernel BUG at 0000000000000002 Ýverbose debug info unavailable¨ 
> illegal operation: 0001 Ý#1¨ 
> Modules linked in: 
> CPU:    0    Not tainted 
> Process ip (pid: 1167, task: 0000000001d46038, ksp: 00000000025efb28) 
> Krnl PSW : 0704200180000000 0000000000000002 (0x2) 
>            R:0 T:1 IO:1 EX:1 Key:0 M:1 W:0 P:0 AS:0 CC:2 PM:0 EA:3 
> Krnl GPRS: 0000000000000000 0000000000000000 000000000241f600 0000000001c8d
>            00000000000086dd 0000000001eb6d70 0000000000000000 0000000001c8d
>            0000000001eb6d40 00000000003abc28 0000000001eb6d00 00000000025ef
>            000000000241f600 00000000003b6d18 00000000002b33d2 00000000025ef
> Krnl Code:>0000000000000002: 0000               unknown 
>            0000000000000004: 0000               unknown 
>            0000000000000006: 0000               unknown 
>            0000000000000008: 0000               unknown 
>            000000000000000a: 0000               unknown 
>            000000000000000c: 0000               unknown 
>            000000000000000e: 0000               unknown 
>            0000000000000010: 0000               unknown 
> Call Trace: 
> (Ý<00000000002b3352>¨ neigh_connected_output+0x76/0x138) 
>  Ý<0000000000325402>¨ ip6_output2+0x2da/0x470 
>  Ý<0000000000326ea6>¨ ip6_output+0x816/0x1064 
>  Ý<0000000000338e46>¨ __ndisc_send+0x416/0x6a8 
>  Ý<0000000000339330>¨ ndisc_send_rs+0x58/0x68 
>  Ý<000000000032cbf4>¨ addrconf_dad_completed+0xbc/0x100 
>  Ý<000000000032d2de>¨ addrconf_dad_start+0xa2/0x14c 
>  Ý<000000000032d408>¨ addrconf_add_linklocal+0x80/0xa8 
>  Ý<000000000032fa7e>¨ addrconf_notify+0x2de/0x8d4 
>  Ý<0000000000383990>¨ notifier_call_chain+0x5c/0x98 
>  Ý<0000000000063bca>¨ __raw_notifier_call_chain+0x26/0x34 
>  Ý<0000000000063c06>¨ raw_notifier_call_chain+0x2e/0x3c 
>  Ý<00000000002aa54c>¨ call_netdevice_notifiers+0x34/0x44 
>  Ý<00000000002ad1aa>¨ dev_open+0x9e/0xe0 
>  Ý<00000000002ad80a>¨ dev_change_flags+0x9e/0x1cc 
>  Ý<0000000000302c74>¨ devinet_ioctl+0x650/0x73c 
>  Ý<00000000003050ba>¨ inet_ioctl+0xde/0xf4 
>  Ý<000000000029a8d0>¨ sock_ioctl+0x1cc/0x2dc 
>  Ý<00000000000cb844>¨ do_ioctl+0xb8/0xcc 
>  Ý<00000000000cb8f2>¨ vfs_ioctl+0x9a/0x3ec 
>  Ý<00000000000cbc96>¨ sys_ioctl+0x52/0x7c 
>  Ý<0000000000022484>¨ sysc_noemu+0x10/0x16 
>  Ý<000002000010df12>¨ 0x2000010df12 

that's a network issue ;)

> 00000000002b3352 gives : include/linux/netdevice.h:819
> 

I have a feeling that we fixed this.  But there's no BUG at 2.6.23-mm1's
include/linux/netdevice.h:819.  How about setting
CONFIG_DEBUG_BUGVERBOSE=y?


-
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html