03-08-2011 11:50 PM
Hello people.
I'm trying to get a DAQ card installed at the University, and have insisted on doing so on OpenSUSE 11.3.
This took some time, but I believe everything is installed correctly now.
Still, nipal doesn't start correctly, reporting (in dmesg) :
More than 4GB of addressable memory detected.
I have 4 GB installed - free -b reports: 4203069440 bytes
(which is actually less than 2^32)
I did notice that 11.3's kernel is, by default, configured to 64 GB. (I'll disable that and recompile
the kernel)
I disabled swap, just in case... no luck. Do I need anything else?
John
03-09-2011 07:37 AM
Hello Martin,
Thanks for the answer!
When I configured the kernel, setting the size range from 1-4GB, the PAE options
disappeared from the menu, so I'm guessing this might have the same result?
I have installed NIKAL 2.0.
John
03-09-2011 08:39 AM
> In short: Why roll your own kernel if you simply can configure the upstream offering?
Well, you've convinced me, Martin! I've added the memmap directive and am rebooting now.
The risk factor is the main argument for me, as I'm doing the entire install process over SSH.
I'll report back with the results.
Cheers,
John
03-09-2011 09:15 AM
Martin,
I added your memmap to the kernel commandline, and rebooted the machine (still remotely). It doesn't come up
anymore, so a trip to the university is in order to check what happened.
Any idea what could have gone wrong? Does the command have to be on a specific spot on the command
line?
John
03-09-2011 10:06 AM
>> Any idea what could have gone wrong?
> I guess that there is wrong spelling. The numbers of zeros must be exactly 8. Please double check the spelling.
> How did you add the memmap option to the commandline?
I added it on the kernel command line entry in YaST2 bootloader menu.
> Maybe you accidentally added an invisible character in grub configuration etc.
Yes - quite possibly. The mouse clicks sometimes give problems with the
remote terminal window.
I'll probably visit the university after lunch - it's 40km, so I want to combine
this with some other errands...
Thanks for the patience, will report back later.
John
03-09-2011 07:21 PM
Hi Martin,
It seems the parameter wasn't wrong - apparently the optical mouse broke down (it's a perfectly new one). As it was,
the mouse device detached and re-attached itself continuously (USB), and caused problems with the login system.
But... there do still to be some problems left. The ni utilities do not report any device (though now the 4GB error has
disappeared). ni-kal seems to crash on startup. There is a crash report in the syslog (which I can't access at the
moment).
John
03-10-2011 06:59 AM
>> The ni utilities do not report any device (though now the 4GB error has disappeared).
>
>What is the output of lsmod?
I'll check this afternoon.
>> ni-kal seems to crash on startup. There is a crash report in the syslog (which I can't access at the
> moment).
>
> So it crashes the kernel?
It's the module which crashes, not the application. So the boot process continues, and several other
NI modules load correctly.
> I have many machines running opensuse 11.3 and nikal though I guess we are using different versions.
> Which version are you using?
NI-KAL 2.0 - I couldn't get any of the other installed. It took me a while to detect this version.
Cheers,
John
03-10-2011 08:57 AM
Thanks Martin,
> rpm -qa | grep -i nika
> nikali-2.1.0-f0.noarch
Is there some systematic way to get a listing of the latest Linux packages? Even the 2.0 nikali package wasn't easy
to find! 2.1 doesn't show up on the NI search engine (last is 2.0). Maybe it is part of another package?
Please help!
John
03-10-2011 07:18 PM
> I guess that you must ask your sales representative for a more recent version.
Complicated - I'm not the buyer. The NI converter is on loan from a U.S. university.. I'll ask them to contact their
NI representative.
Today, in between classes, I restarted the machine in failsafe mode, and I'm now able to access it remotely
again. There is something that makes it unstable in normal mode.
Here's a dump from dmesg which was done with the memmap directive active. I have the impression this has
something to do with the mapping:
[ 22.941242] ADDRCONF(NETDEV_CHANGE): eth0: link becomes ready
[ 23.497159] nikal: module license 'Copyright (c) 2002-2010 National
Instruments Corporation. All Rights Reserved. Any and all use of the
copyrighted materials is subject to the then current terms and conditions of
the applicable license agreement, which can be found at
<http://www.ni.com/linux/>.' taints kernel.
[ 23.497163] Disabling lock debugging due to kernel taint
[ 23.922062] BUG: unable to handle kernel paging request at f71fe000
[ 23.922068] IP: [<f8d69bb4>] nipalk-unversioned0002875+0x294/0x4a8
[nipalk]
[ 23.922116] *pdpt = 00000000009f9001 *pde = 0000000000015067 *pte =
0000000000000000
[ 23.922119] Oops: 0000 [#1] PREEMPT SMP
[ 23.922122] last sysfs file: /sys/bus/pci/drivers/nidaqmx_NIDSA9x/uevent
[ 23.922125] Modules linked in: nipalk(P) nikal(P) edd
cpufreq_conservative cpufreq_userspace cpufreq_powersave ip6t_REJECT
acpi_cpufreq nf_conntrack_ipv6 ip6table_raw speedstep_lib xt_NOTRACK
ipt_REJECT xt_state iptable_raw iptable_filter mperf ip6table_mangle
nf_conntrack_netbios_ns nf_conntrack_ipv4 nf_conntrack nf_defrag_ipv4
ip_tables ip6table_filter ip6_tables x_tables loop dm_mod ppdev parport_pc
sg parport sr_mod cdrom mptctl joydev tg3 container iTCO_wdt
iTCO_vendor_support i2c_i801 pcspkr serio_raw i3200_edac button edac_core
preloadtrace ext4 jbd2 crc16 radeon ttm drm_kms_helper drm i2c_algo_bit fan
processor ata_generic mptsas mptscsih mptbase scsi_transport_sas thermal
thermal_sys
[ 23.922161]
[ 23.922163] Pid: 2617, comm: cat Tainted: P
2.6.34.7-0.7-desktop #1 4367/IBM System x3200 M2 -[436734U]-
[ 23.922166] EIP: 0060:[<f8d69bb4>] EFLAGS: 00010212 CPU: 0
[ 23.922204] EIP is at nipalk-unversioned0002875+0x294/0x4a8 [nipalk]
[ 23.922206] EAX: f2ea93e8 EBX: f8db2da0 ECX: 010d3740 EDX: f2eb0300
[ 23.922207] ESI: f8db2da0 EDI: f2dba738 EBP: f2ccfefc ESP: f2ccfe7c
[ 23.922209] DS: 007b ES: 007b FS: 00d8 GS: 0033 SS: 0068
[ 23.922211] Process cat (pid: 2617, ti=f2cce000 task=f44e50c0
task.ti=f2cce000)
[ 23.922212] Stack:
[ 23.922213] f2eb0300 00000004 f8db2da0 00000000 00000000 f2ea93e8
00000000 00000000
[ 23.922217] <0> 00000000 f2dba738 f8d8dcf4 f2ccfed4 00000000 f8db2da0
010d3740 00000000
[ 23.922221] <0> f2ccff00 f8db2da0 00000002 f8d61549 f2ccfed4 f8db2da0
00000002 f2ccfef8
[ 23.922226] Call Trace:
[ 23.922236] Inexact backtrace:
[ 23.922237]
[ 23.922266] [<f8d8dcf4>] ? _ZNV14tSyncAtomicU32mmEi+0x14/0x30 [nipalk]
[ 23.922307] [<f8d61549>] ? nipalk-unversioned0002590+0x55/0x7c [nipalk]
[ 23.922347] [<f8d4a969>] ? nipalk-unversioned0002405+0x29/0x48 [nipalk]
[ 23.922386] [<f8d49e34>] ? nipalk-unversioned0002396+0x1b8/0x1ec
[nipalk]
[ 23.922391] [<c0654380>] ? notify_die+0x30/0x40
[ 23.922394] [<c065227c>] ? do_debug+0xac/0x180
[ 23.922399] [<f7aa094c>] ? nNIKAL100_release+0x1c/0x40 [nikal]
[ 23.922402] [<c02feca7>] ? __fput+0xc7/0x1d0
[ 23.922404] [<c02fb459>] ? filp_close+0x49/0x70
[ 23.922406] [<c02fb4ed>] ? sys_close+0x6d/0xc0
[ 23.922409] [<c020324c>] ? sysenter_do_call+0x12/0x22
[ 23.922411] Code: a4 00 00 00 00 8b 5d 88 89 5d b4 c7 45 b8 00 00 00 00
8b 43 08 85 c0 89 5d c4 89 45 c8 74 27 8b 75 b4 8b 4d b8 8b 56 0c 8b 45 94
<39> 04 8a 74 2e 8d 51 01 8b 5d 88 89 55 b8 8b 43 08 39 c2 89 5d
[ 23.922433] EIP: [<f8d69bb4>] nipalk-unversioned0002875+0x294/0x4a8
[nipalk] SS:ESP 0068:f2ccfe7c
03-11-2011 08:56 AM
Hey John-
This crash looks similar to what was solved with the release posted in KB 5D2ERAUS. Could you give it a try, and see if it fixes the problem you are seeing?