Bug#480310: [ixp4xx] armel NSLU2 fails to boot with 2.6.25
Martin Michlmayr
> Package: linux-image-2.6.25-1-ixp4xx
> Version: 2.6.25-1
>
> After the usual initial lightshow, the NSLU2 sits there with only the
> Ethernet LED illuminated. I have initramfs-tools 0.92a and the initrd
> is 1.4MB, so I think that should be mustard. I'm running armel. I have
> no serial console, so I guess I can't be much help with debugging.
I just tried (on arm) and it works for me. CCing debian-arm. Anyone
else seeing problems with 2.6.25?
--
Martin Michlmayr
http://www.cyrius.com/
--
To UNSUBSCRIBE, email to debian-bugs-...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listm...@lists.debian.org
Kevin Price
Martin Michlmayr schrieb:
> I just tried (on arm) and it works for me. CCing debian-arm. Anyone
> else seeing problems with 2.6.25?
Yes, I can confirm this bug on armel. Thanks Martin fo CCing debian-arm.
In order to reproduce this, I installed 2.6.25-1-ixp4xx_2.6.25-1_armel
onto my otherwise complete lenny/armel slug. Since then it fails to boot.
My slug does have a serial port, and for debugging purposes I will try
to find something to attach to it today, and then send you the console
output. Someone forgot to equip my ThinkPad with an RS232 ;-)
@Riku: flash-kernel only sounds scary if you don't have a
/proc/mtdblock* backup ;-)
cheers
Kevin
Martin Michlmayr
> Yes, I can confirm this bug on armel. Thanks Martin fo CCing debian-arm.
> In order to reproduce this, I installed 2.6.25-1-ixp4xx_2.6.25-1_armel
> onto my otherwise complete lenny/armel slug. Since then it fails to boot.
>
> My slug does have a serial port, and for debugging purposes I will try
> to find something to attach to it today, and then send you the console
Cool, that'd be great!
Kevin Price
> Cool, that'd be great!
Here we go:
--------------------------------
Ethernet eth0: MAC address 00:14:bf:65:db:3d
IP: 192.168.0.1/255.255.255.0, Gateway: 192.168.0.1
Default server: 0.0.0.0, DNS server IP: 0.0.0.0
RedBoot(tm) bootstrap and debug environment [ROMRAM]
Red Hat certified release, version 1.92 - built 15:16:07, Feb 3 2004
Platform: IXDP425 Development Platform (XScale)
Copyright (C) 2000, 2001, 2002, Red Hat, Inc.
RAM: 0x00000000-0x02000000, 0x000723a0-0x01ff3000 available
FLASH: 0x50000000 - 0x50800000, 64 blocks of 0x00020000 bytes each.
== Executing boot script in 2.000 seconds - enter ^C to abort
RedBoot> boot;exec 0x01d00000
have eRcOmM
copy kernel code from flash to RAM
copy ramdisk file from flash to RAM
run kernel
Using base address 0x01d00000 and length 0x0000a85c
APEX Boot Loader 1.4.15 -- Copyright (c) 2004,2005,2006,2007 Marc Singer
APEX comes with ABSOLUTELY NO WARRANTY. It is free software and you
are welcome to redistribute it under certain circumstances.
For details, refer to the file COPYING in the program source.
apex => mem:0x00200000+0xa84c (43084 bytes)
env => nor:0x7c000+15k (empty)
Use the command 'help help' to get started.
# copy -s $kernelsrc $bootaddr
# copy -s fis://kernel 0x00008000
1441760 bytes transferred
# copy -s $ramdisksrc $ramdiskaddr
# copy -s fis://ramdisk 0x01000000
6291440 bytes transferred
# wait 10 Type ^C key to cancel autoboot.
Type ^C key to cancel autoboot.
# boot
ATAG_HEADER
ATAG_MEM: start 0x00000000 size 0x02000000
ATAG_CMDLINE: (55 bytes) 'console=ttyS0,115200 rtc-x1205.probe=0,0x6f noirqdebug'
ATAG_INITRD2: start 0x01000000 size 0x00400000
ATAG_END
Booting kernel at 0x00008000...
Uncompressing Linux.......................................................................................... done, booting the kernel.
[ 0.000000] Linux version 2.6.25-1-ixp4xx (Debian 2.6.25-1) (wa...@debian.org) (gcc version 4.1.3 20080308 (prerelease) (Debian 4.1.2-21)) #2 Wed Apr 38
[ 0.000000] CPU: XScale-IXP42x Family [690541f1] revision 1 (ARMv5TE), cr=0000397f
[ 0.000000] Machine: Linksys NSLU2
[ 0.000000] Memory policy: ECC disabled, Data cache writeback
[ 0.000000] CPU0: D VIVT undefined 5 cache
[ 0.000000] CPU0: I cache: 32768 bytes, associativity 32, 32 byte lines, 32 sets
[ 0.000000] CPU0: D cache: 32768 bytes, associativity 32, 32 byte lines, 32 sets
[ 0.010000] Built 1 zonelists in Zone order, mobility grouping on. Total pages: 8128
[ 0.010000] Kernel command line: console=ttyS0,115200 rtc-x1205.probe=0,0x6f noirqdebug
[ 0.010000] Unknown boot option `rtc-x1205.probe=0,0x6f': ignoring
[ 0.010000] IRQ lockup detection disabled
[ 0.010000] PID hash table entries: 128 (order: 7, 512 bytes)
[ 0.010000] Console: colour dummy device 80x30
[ 0.010000] Dentry cache hash table entries: 4096 (order: 2, 16384 bytes)
[ 0.010000] Inode-cache hash table entries: 2048 (order: 1, 8192 bytes)
[ 0.010000] Memory: 32MB = 32MB total
[ 0.010000] Memory: 25412KB available (2556K code, 270K data, 104K init)
[ 0.240000] Security Framework initialized
[ 0.240000] SELinux: Disabled at boot.
[ 0.240000] Capability LSM initialized
[ 0.240000] Mount-cache hash table entries: 512
[ 0.240000] Initializing cgroup subsys ns
[ 0.240000] Initializing cgroup subsys cpuacct
[ 0.240000] CPU: Testing write buffer coherency: ok
[ 0.240000] net_namespace: 540 bytes
[ 0.240000] NET: Registered protocol family 16
[ 0.250000] IXP4xx: Using 16MiB expansion bus window size
[ 0.250000] NSLU2: Using MAC address 00:14:bf:65:db:3d for port 0
[ 0.250000] PCI: IXP4xx is host
[ 0.250000] PCI: IXP4xx Using direct access for memory space
[ 0.250000] PCI: bus0: Fast back to back transfers disabled
[ 0.250000] dmabounce: registered device 0000:00:01.0 on pci bus
[ 0.250000] dmabounce: registered device 0000:00:01.1 on pci bus
[ 0.250000] dmabounce: registered device 0000:00:01.2 on pci bus
[ 0.290000] NET: Registered protocol family 2
[ 0.380000] IP route cache hash table entries: 1024 (order: 0, 4096 bytes)
[ 0.380000] TCP established hash table entries: 1024 (order: 1, 8192 bytes)
[ 0.380000] TCP bind hash table entries: 1024 (order: 0, 4096 bytes)
[ 0.380000] TCP: Hash tables configured (established 1024 bind 1024)
[ 0.380000] TCP reno registered
[ 0.410000] checking if image is initramfs...it isn't (invalid compressed format (err=1)); looks like an initrd
[ 2.730000] Freeing initrd memory: 4096K
[ 2.730000] NetWinder Floating Point Emulator V0.97 (double precision)
[ 2.730000] audit: initializing netlink socket (disabled)
[ 2.730000] type=2000 audit(2.754:1): initialized
[ 2.740000] VFS: Disk quotas dquot_6.5.1
[ 2.740000] Dquot-cache hash table entries: 1024 (order 0, 4096 bytes)
[ 2.740000] JFFS2 version 2.2. (NAND) © 2001-2006 Red Hat, Inc.
[ 2.740000] io scheduler noop registered
[ 2.740000] io scheduler anticipatory registered
[ 2.740000] io scheduler deadline registered
[ 2.740000] io scheduler cfq registered (default)
[ 2.780000] Serial: 8250/16550 driver $Revision: 1.90 $ 2 ports, IRQ sharing disabled
[ 2.780000] serial8250.0: ttyS0 at MMIO 0xc8000000 (irq = 15) is a XScale
[ 2.780000] console [ttyS0] enabled
[ 2.790000] serial8250.0: ttyS1 at MMIO 0xc8001000 (irq = 13) is a XScale
[ 2.810000] brd: module loaded
[ 2.820000] IXP4XX-Flash.0: Found 1 x16 devices at 0x0 in 16-bit bank
[ 2.830000] Intel/Sharp Extended Query Table at 0x0031
[ 2.830000] Using buffer write method
[ 2.840000] cfi_cmdset_0001: Erase suspend on write enabled
[ 2.840000] Searching for RedBoot partition table in IXP4XX-Flash.0 at offset 0x7e0000
[ 2.940000] 6 RedBoot partitions found on MTD device IXP4XX-Flash.0
[ 2.950000] Creating 6 MTD partitions on "IXP4XX-Flash.0":
[ 2.950000] 0x00000000-0x00040000 : "RedBoot"
[ 2.960000] 0x00040000-0x00060000 : "SysConf"
[ 2.970000] 0x00060000-0x00080000 : "Loader"
[ 2.970000] 0x00080000-0x001e0000 : "Kernel"
[ 2.980000] 0x001e0000-0x007e0000 : "Ramdisk"
[ 2.990000] 0x007e0000-0x00800000 : "FIS directory"
[ 2.990000] mice: PS/2 mouse device common for all mice
[ 3.000000] i2c /dev entries driver
[ 3.020000] rtc-x1205 0-006f: chip found, driver version 1.0.8
[ 3.020000] rtc-x1205 0-006f: rtc core: registered rtc-x1205 as rtc0
[ 3.030000] i2c-gpio i2c-gpio.0: using pins 7 (SDA) and 6 (SCL)
[ 3.040000] IXP4xx Watchdog Timer: heartbeat 60 sec
[ 3.050000] Registered led device: nslu2:green:ready
[ 3.060000] Registered led device: nslu2:red:status
[ 3.060000] Registered led device: nslu2:green:disk-1
[ 3.070000] Registered led device: nslu2:green:disk-2
[ 3.070000] NET: Registered protocol family 26
[ 3.080000] TCP bic registered
[ 3.090000] NET: Registered protocol family 1
[ 3.090000] NET: Registered protocol family 17
[ 3.100000] XScale DSP coprocessor detected.
[ 3.100000] registered taskstats version 1
[ 3.110000] rtc-x1205 0-006f: setting system clock to 2008-05-12 13:20:48 UTC (1210598448)
[ 3.120000] RAMDISK: Compressed image found at block 0
[ 5.410000] RAMDISK: incomplete write (-28 != 32768) 8388608
[ 5.770000] RAMDISK: ran out of compressed data
[ 5.780000] invalid compressed format (err=1)
[ 5.850000] VFS: Cannot open root device "<NULL>" or unknown-block(0,0)
[ 5.860000] Please append a correct "root=" boot option; here are the available partitions:
[ 5.870000] 0100 8192 ram0 (driver?)
[ 5.870000] 0101 8192 ram1 (driver?)
[ 5.880000] 0102 8192 ram2 (driver?)
[ 5.880000] 0103 8192 ram3 (driver?)
[ 5.890000] 0104 8192 ram4 (driver?)
[ 5.890000] 0105 8192 ram5 (driver?)
[ 5.890000] 0106 8192 ram6 (driver?)
[ 5.900000] 0107 8192 ram7 (driver?)
[ 5.900000] 0108 8192 ram8 (driver?)
[ 5.910000] 0109 8192 ram9 (driver?)
[ 5.910000] 010a 8192 ram10 (driver?)
[ 5.920000] 010b 8192 ram11 (driver?)
[ 5.920000] 010c 8192 ram12 (driver?)
[ 5.920000] 010d 8192 ram13 (driver?)
[ 5.930000] 010e 8192 ram14 (driver?)
[ 5.930000] 010f 8192 ram15 (driver?)
[ 5.940000] 1f00 256 mtdblock0 (driver?)
[ 5.940000] 1f01 128 mtdblock1 (driver?)
[ 5.950000] 1f02 128 mtdblock2 (driver?)
[ 5.950000] 1f03 1408 mtdblock3 (driver?)
[ 5.950000] 1f04 6144 mtdblock4 (driver?)
[ 5.960000] 1f05 128 mtdblock5 (driver?)
[ 5.960000] Kernel panic - not syncing: VFS: Unable to mount root fs on unknown-block(0,0)
--------------------------------
--
http://www.kevin-price.de/
Martin Michlmayr
> > Cool, that'd be great!
> Here we go:
Thanks.
Can you show me the output of
ls -l /boot
Kevin Price
> Can you show me the output of
> ls -l /boot
Are you suspecting the initrd size?
My /etc/initramfs-tools/initramfs.conf says "MODULES=most", btw.
ls -l /boot
-rw-r--r-- 1 root root 748537 Apr 21 22:05 System.map-2.6.24-1-ixp4xx
-rw-r--r-- 1 root root 762523 Apr 30 10:51 System.map-2.6.25-1-ixp4xx
-rw-r--r-- 1 root root 43084 May 27 2007 apex.flash
-rw-r--r-- 1 root root 56582 Apr 20 13:41 config-2.6.24-1-ixp4xx
-rw-r--r-- 1 root root 59504 Apr 28 17:06 config-2.6.25-1-ixp4xx
lrwxrwxrwx 1 root root 26 May 12 13:09 initrd.img -> initrd.img-2.6.25-1-ixp4xx
-rw-r--r-- 1 root root 4149652 Apr 30 01:02 initrd.img-2.6.24-1-ixp4xx
-rw-r--r-- 1 root root 4355143 May 12 13:09 initrd.img-2.6.25-1-ixp4xx
lrwxrwxrwx 1 root root 26 Apr 29 23:46 initrd.img.old -> initrd.img-2.6.24-1-ixp4xx
lrwxrwxrwx 1 root root 23 May 12 13:09 vmlinuz -> vmlinuz-2.6.25-1-ixp4xx
-rw-r--r-- 1 root root 1339880 Apr 21 22:05 vmlinuz-2.6.24-1-ixp4xx
-rw-r--r-- 1 root root 1364620 Apr 30 10:51 vmlinuz-2.6.25-1-ixp4xx
lrwxrwxrwx 1 root root 23 Feb 3 23:01 vmlinuz.old -> vmlinuz-2.6.24-1-ixp4xx
Kevin Price
Kevin Price schrieb:
> Are you suspecting the initrd size?
> My /etc/initramfs-tools/initramfs.conf says "MODULES=most", btw.
That was the cause. Chaning modules to "dep" created an initrd.img with
a size of 2176130, which is less than half the size of
> -rw-r--r-- 1 root root 4355143 May 12 13:09 initrd.img-2.6.25-1-ixp4xx
With that, my slug boots fine. I thought, flash-kernel is supposed to
complain about such problems. I used the current lenny version.
ii flash-kernel 1.7 utility to write kernel and initramfs
I suggest moving this bug to package flash-kernel. Any objections?
cheers
Kevin
--
http://www.kevin-price.de/
Martin Michlmayr
> Are you suspecting the initrd size?
Yes, while the NSLU2 has 6 MB for the ramdisk, APEX passes a parameter
to the kernel saying the ramdisk is 4 MB, so everything after 4 MB is
ignored. Your ramdisk is larger than 4 MB, which explains why the
kernel is complaing that the ramdisk is not complete.
The reason this shows up on armel and not on arm is that some SCSI
drivers are disabled on arm but not on armel (for historical reasons).
The short term solution is to disable some more drivers. I've done
that for 2.6.26-rc1 and will do it for the next 2.6.25 upload too.
The right solution is to fix APEX. A bug about this has been open for
quite a while: #451882
Kevin, since you have a serial console, are you interested in testing
the patches in #451882 and #421359?
debian-arm people in general: anyone with a armel compiler who's
interested in compiling and testing 2.6.26-rc1 on armel?
Kevin Price
> Kevin, since you have a serial console, are you interested in testing
> the patches in #451882 and #421359?
Yes. My slug is now compiling apex-1.4.15 with the patch from #451882
(CONFIG_RAMDISK_SIZE=0x005FFFF0) I am curious if this solves this bug
and I will let you know.
BTW: What is keeping the patches from moving into sid? I'd love to see a
more current apex version in lenny.
> debian-arm people in general: anyone with a armel compiler who's
> interested in compiling and testing 2.6.26-rc1 on armel?
Would that need to be a native or cross-compiler or is the qemu arm
emulation reliable enough for such compilation jobs? Somehow I recall
that the emulation was said to be unsuitable to be employed as buildds.
Kevin
--
http://www.kevin-price.de/
Rod Whitby
> * Kevin Price <k...@kevin-price.de> [2008-05-12 17:43]:
>>> Kevin, since you have a serial console, are you interested in testing
>>> the patches in #451882 and #421359?
>> Yes. My slug is now compiling apex-1.4.15 with the patch from #451882
>> (CONFIG_RAMDISK_SIZE=0x005FFFF0) I am curious if this solves this bug
>> and I will let you know.
>
> written to flash.
>
>> BTW: What is keeping the patches from moving into sid? I'd love to
>> see a more current apex version in lenny.
>
> bugs.
BTW, the nslu2-linux project keeps some apex patches at:
http://svn.nslu2-linux.org/svnroot/kernel/trunk/patches/apex/
If Marc is not responsive, I'm happy to give relevant people write
access to there to be used as a central place for apex patches ...
Once the initramfs has some recovery support, the variations feature of
Apex will become useful.
-- Rod
--
To UNSUBSCRIBE, email to debian-ar...@lists.debian.org
Kevin Price
> This has been fixed in version 2.6.25-3 of linux-image-2.6.25-2-ixp4xx
> which is now in unstable. Would be great if you (Paul, Kevin) could
> confirm it works.
Yes, confirmed. It works nicely for me. I re-flashed the old apex from
debian, and set modules=most in initramfs.conf. before installing
2.6.25-3. The initrd became 3840262 bytes small, which is fine. All the
necessary modules are there too.
Thanks a lot for that good piece of work!
--
Kevin Price
http://www.kevin-price.de/
Paul Collins
> Martin Michlmayr schrieb:
>> This has been fixed in version 2.6.25-3 of linux-image-2.6.25-2-ixp4xx
>> which is now in unstable. Would be great if you (Paul, Kevin) could
>> confirm it works.
>
> Yes, confirmed. It works nicely for me.
Works great here too. Thanks, folks!
--
Paul Collins
Wellington, New Zealand
Dag vijandelijk luchtschip de huismeester is dood
Marc Singer
> * Kevin Price <k...@kevin-price.de> [2008-05-12 17:43]:
> > > Kevin, since you have a serial console, are you interested in testing
> > > the patches in #451882 and #421359?
> >
> > Yes. My slug is now compiling apex-1.4.15 with the patch from #451882
> > (CONFIG_RAMDISK_SIZE=0x005FFFF0) I am curious if this solves this bug
> > and I will let you know.
>
> written to flash.
>
> > see a more current apex version in lenny.
>
> feedback on the patch in #421359.
...perhaps it has taken me too long to do so, but I am glad to make
changes.
Is the 1.4.15.2 update insufficient?
>
> > > debian-arm people in general: anyone with a armel compiler who's
> > > interested in compiling and testing 2.6.26-rc1 on armel?
> > Would that need to be a native or cross-compiler or is the qemu arm
>
> cross-compile would be fine. I just don't have a cross-compiler for
> armel. Native or qemu would work too, but are slower.
> --
> Martin Michlmayr
> http://www.cyrius.com/
>
>
> --
> To UNSUBSCRIBE, email to debian-ar...@lists.debian.org
Kevin Price
> ...perhaps it has taken me too long to do so, but I am glad to make
> changes.
Thanks fo doing so!
> Is the 1.4.15.2 update insufficient?
As far as I am concerned, this works nicely. Since the urgency is low,
it'll take a few more days for apex to advance into testing.
see http://bjorn.haxx.se/debian/testing.pl?package=apex
It seems to me that packages.debian.org is not the right place to look,
because it seems to have been out of date during the last few days.
If lenny's freeze keeps apex from moving on: Will we need to ask the
release manager?
cheers