> >> > So far I have made sure that 4.2.0 works fine on V240 overnight with the
> >> > following command to find any other errors too.
> >> > while true; do make clean; make -j2 || break; done
For now it seems 4.3.0-00063-gea2d67b and 4.3.0-08824-g7c623ca also
survive for many hours. That is strange - I think I saw it earlier on
v240 where I am currently testing (I have seen it multiple times on
v240, only once or twice on v440).
Now I have also ran 4.4-rc8 on v440 for some hours and it seem to
survive.
So I do not know how to reproduce the reboot at will :(
> >> let me restart a loop with your .config on my v440 and let this run
> >> overnight, to see if I can reproduce your case.
> >
> > Hmm. I've rebooted my v440 with a kernel built with your .config,
> > I've been able to do the above make loop for several hours now.
> >
> > perhaps it is some service/module that you have enabled/loaded
> > that I dont? Would you like to compare output of lsmod and/or
> > chkconfig and/or service--status-all? (let me know which one).
lsmod on v440:
ipv6 371138 22
loop 18495 0
sr_mod 15039 0
cdrom 30798 1 sr_mod
ohci_pci 3362 0
ohci_hcd 32471 1 ohci_pci
usbcore 191889 2 ohci_hcd,ohci_pci
usb_common 3832 1 usbcore
pata_ali 9305 0
libata 195005 1 pata_ali
sg 19827 0
cassini 46893 0
flash 3435 0
lsmod on v240:
ipv6 496922 26
loop 19367 0
tg3 167037 0
skge 40515 0
hwmon 3838 1 tg3
sg 20793 0
i2c_ali15x3 6252 0
i2c_ali1535 6046 0
i2c_core 28484 2 i2c_ali1535,i2c_ali15x3
ptp 12769 1 tg3
pps_core 8737 1 ptp
flash 3875 0
skge is for Ethernet controller: SysKonnect SK-9872 Gigabit
Ethernet Server Adapter (SK-NET GE-ZX dual link) (rev 12) that is in the
v240 as an additional card with OF ROM.
> > Other thing is, there is nothing in the console/logs to
> > give a hint why it rebooted?
Nothing in ALOM console log history or in syslog.
> It could also be influenced by either the amount of memory he has
> installed, or what userland he is using.
v240 has 6G RAM, v440 has 8G.
Userland is Debian unstable, as it was as of July 25, 2015. gcc 4.9.3-2,
binutils 2.25-10.
My usage pattern on all the machines was powering up the machine, doing
git pull && make -j4 && echo OK
where -j number corresponds to number of CPUs in the system. Maybe git
pull created some memory pressure before?
ps shows I have running
init udevd dhclient uuidd rsyslogd atd cron irqbalance sshd exim4 ntpd
getty + sshd,bash for my login session.
v440 also has /usr/bin/daemon /etc/init.d/mpt-statusd check_mpt
init is sysvinit on both machines.
irqbalance and ntpd are the only slightly suspicious ones for my eye, or
i2c modules but these are only on v240.
mroos@v440:~$ cat /proc/interrupts
CPU0 CPU1 CPU2 CPU3
0: 7561 3617 5495 4128 none timer
1: 0 0 0 0 sun4u-IVEC TOMATILLO_PCIERR
2: 0 0 0 0 sun4u-IVEC TOMATILLO_UE
3: 0 0 0 0 sun4u-IVEC TOMATILLO_CE
4: 0 0 0 0 sun4u-IVEC TOMATILLO_SERR
6: 0 0 453 0 sun4u-IVEC eth0
8: 0 0 0 0 sun4u-IVEC TOMATILLO_PCIERR
13: 0 0 0 0 sun4u-IVEC TOMATILLO_PCIERR
14: 0 0 0 0 sun4u-IVEC TOMATILLO_UE
15: 0 0 0 0 sun4u-IVEC TOMATILLO_CE
16: 0 0 0 0 sun4u-IVEC TOMATILLO_SERR
19: 0 0 0 0 sun4u-IVEC power
20: 230 0 0 0 sun4u-IVEC su(serial)
21: 0 0 0 0 sun4u-IVEC ohci_hcd:usb1
22: 0 0 0 0 sun4u-IVEC ohci_hcd:usb2
23: 0 172 0 199 sun4u-IVEC pata_ali
24: 0 0 0 0 sun4u-IVEC TOMATILLO_PCIERR
31: 0 0 0 3311 sun4u-IVEC ioc0
32: 49 0 0 0 sun4u-IVEC ioc1
NMI: 0 0 0 0 Non-maskable interrupts
mroos@v240:~$ cat /proc/interrupts
CPU0 CPU1
0: 10073372 10074564 none timer
1: 0 0 sun4u-IVEC TOMATILLO_PCIERR
2: 0 0 sun4u-IVEC TOMATILLO_UE
3: 0 0 sun4u-IVEC TOMATILLO_CE
4: 0 0 sun4u-IVEC TOMATILLO_SERR
6: 44675 52105 sun4u-IVEC eth0
8: 0 0 sun4u-IVEC TOMATILLO_PCIERR
14: 0 0 sun4u-IVEC power
15: 0 266 sun4u-IVEC su(serial)
20: 0 0 sun4u-IVEC TOMATILLO_PCIERR
21: 0 0 sun4u-IVEC TOMATILLO_UE
22: 0 0 sun4u-IVEC TOMATILLO_CE
23: 0 0 sun4u-IVEC TOMATILLO_SERR
25: 46126 73792 sun4u-IVEC sym53c8xx
26: 30 0 sun4u-IVEC sym53c8xx
27: 0 0 sun4u-IVEC TOMATILLO_PCIERR
NMI: 42625 42625 Non-maskable interrupts
--
Meelis Roos (mroos@xxxxxxxx)
--
To unsubscribe from this list: send the line "unsubscribe sparclinux" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
