On Mon, 27 Feb 2006, Tomasz Chmielewski wrote:
> Pavel Tsekov wrote:
> >> Pressing ctrl + o shows the background, but I can't execute any shell
> >> commands (I return to mc immediately when pressing any key).
> >> Strace available at http://wpkg.org/subshell.strace.gz (strace -o
> >> subshell.strace) and http://wpkg.org/sub.strace.gz (strace -o sub.strace
> >> mc -U).
> > The strace output suggests that the subshell is being killed by a SIGPWR
> > signal. I cannot tell why this signal is being sent to the subshell
> > though. Maybe the OS you are running on your embedded device is
> > sending this signal for some specific purpose and expects the application
> > to deal with it... In any case the fact that the subshell is not working
> > on your platform is not a bug in MC.
> Perhaps the subshell consumes so much resources on this tiny router that
> the kernel reports a power failure (that's what SIGPWR is) and kills
> the subshell :)
I cannot tell for sure but this doesn't seem to be the right signal for
such scenarios ... The subshell process is just another shell (i.e. bash)
so it consumes as much resources as your login shell.
> It's a regular Linux 2.4.30, a major difference is that it's not using
> glibc, but uClibc, and not a x86 processor, but a mipsel one.
Are you perhaps targetting an OpenWRT capable device ? If this is the case
I might be able to get a shell access on an ASUS W500G device and try
> Not being a kernel hacker, this tells me that the only way to get
> subshell working for me would be to make it ignore the SIGPWR I guess?
It will help - yes. However there must be some reason for the signal being
sent. I suggest you to try and find out what it might be.
Mc mailing list