Somehow as root, I changed my shell to a non-existent shell '/bin/tcsh' on OpenBSD 4.2 When I try to su, I get this error: su: /bin/tcsh: No such file or directory I can't login as root, ssh in as root or su to root. I'm not in the sudoers file (but I am in the wheel group) so I can't sudo chsh for root. Any suggestions. I just set the box up today. It's been years since I used OpenBSD. I'm a bit rusty and duller than I once was :) Thanks, Brad -- View this message in context: http://www.nabble.com/I%27ve-done-something-stupid-tf4775501.html#a13660801 Sent from the openbsd user - misc mailing list archive at Nabble.com.
-- View this message in context: http://www.nabble.com/I%27ve-done-something-stupid-tf4775501.html#a13667702 Sent from the openbsd user - misc mailing list archive at Nabble.com.
-- View this message in context: http://www.nabble.com/I%27ve-done-something-stupid-tf4775501.html#a13671503 Sent from the openbsd user - misc mailing list archive at Nabble.com.
For the future, I find it useful to have headless boxes set up for serial console attached (or attachable) to a modem. Saves the drive. Doug.
No problem. Put in the boot media you used to install OpenBSD and boot with it. At the (U)pgrade, (I)nstall, (S)hell prompt type 's' for shell. Next, you can "really" fix the problem. Maybe someone else will tell you how. I'm going to tell you the first short way that comes to mind... # mount /dev/wd0a /mnt # cd /mnt/bin # cp ksh tcsh # echo /bin/tcsh >> /mnt/etc/shells <eject boot media> # reboot <login and fix your shell to be one of the installed ones> # rm /bin/tcsh -- Darrin Chandler | Phoenix BSD User Group | MetaBUG dwchandler@stilyagin.com | http://phxbug.org/ | http://metabug.org/ http://www.stilyagin.com/ | Daemons in the Desert | Global BUG Federation
Get your boot CD out. Go into the shell, and then you can mount the root partition on /mnt, and start fixing things. --STeve Andre'
Hi Folks, I've just been upgrading some of our old war horses (Nokia IP440) to 4.2. They run Intel made BX PIII chipset motherboards, dmesg below. Whilst not extensive the boards do have some sensor data that we grab to check on the health of the old girls. After a fresh install of 4.2 I noticed we had lost the FAN readout from the list of sensors, see output below (taken from different boxes but I've confirmed the loss using the same box switching between 4.1 and 4.2). Whilst this isn't critical for us on these units whatever is causing the omission may have bigger problems for other people so I thought I'd bring it to the lists attention. Many thanks to all the developers for yet another excellence release in 4.2, the bulk CD order is going through soon! Sensor Output from 4.1 i386 (sysctl -a hw) hw.machine=i386 hw.model=Intel Pentium III ("GenuineIntel" 686-class) hw.ncpu=1 hw.byteorder=1234 hw.physmem=267993088 hw.usermem=267988992 hw.pagesize=4096 hw.disknames=wd0,cd0,fd0 hw.diskcount=3 hw.sensors.lmenv0.temp1=23.00 degC (Internal) *hw.sensors.lmenv0.fan0=2647 RPM ******* MISSING ******* hw.sensors.lmenv0.fan1=3970 RPM ********* MISSING ******** hw.sensors.lmenv0.volt0=1.52 VDC (+2.5Vin) hw.sensors.lmenv0.volt1=1.66 VDC (Vccp) hw.sensors.lmenv0.volt2=3.30 VDC (+Vcc) hw.sensors.lmenv0.volt3=5.08 VDC (+5Vin/Vcc) hw.sensors.lmenv0.volt4=12.38 VDC (+12Vin) hw.sensors.lmenv0.volt5=2.43 VDC (Vccp) hw.cpuspeed=599 hw.vendor=Intel Corporation hw.product=SE440BX-2 hw.uuid=ebf758f0-b47b-11d4-af0d-0030d3006ea4 Sensor Output from 4.2 i386 (sysctl -a hw) hw.machine=i386 hw.model=Intel Pentium III ("GenuineIntel" 686-class) hw.ncpu=1 hw.byteorder=1234 hw.physmem=267993088 hw.usermem=267984896 hw.pagesize=4096 hw.disknames=wd0,cd0,fd0 hw.diskcount=3 hw.sensors.lmenv0.temp1=28.00 degC (Internal) hw.sensors.lmenv0.volt0=1.50 VDC (+2.5Vin) hw.sensors.lmenv0.volt1=1.69 VDC (Vccp) hw.sensors.lmenv0.volt2=3.27 VDC (+Vcc) hw.sensors.lmenv0.volt3=5.05 ...
fixed, lm87.c#rev1.20. :) The bug was caused by an ininitialised value, such that fan sensors in certain chips (lm81, adm9240 and ds1780) might have pseudo-randomly never appeared. Just to make it clear -- this was not a regression in 4.2, the fact that it was missing from 4.2 is simply a pseudo-random occurrence. :) In any case, the bug should be gone for good, thanks to LLVM/Clang Static Analyser. br, cnst.su.
