Re: ARM, registers, "frame 0", where's the PC?

[Daniel Jacobowitz <drow at mvista dot com>]

On Thu, Jan 09, 2003 at 11:46:45PM +0100, Torsten Mohr wrote:
> Hi,
> 
> > Have you tried this without using Insight?  It looks to me like
> > something's not flushing the frame cache that ought to be.
> 
> i just did some more testing, it seems it disturbed GDB and INSIGHT
> that it is not ok to answer a "continue" with a "S0" after the
> target stopped again.  I now use a "S00", which improves things
> a lot.  Yes, i just doublechecked that, when the target stops
> after a "continue" and i send "S0", it tells me it stopped at
> 0x12345678, when i answer "S00", insight shows me where it stopped
> in the window.  Looking at "info gdb" this is specified, so
> it is my mistake.  I took that information from a file i found
> in the web where it didn't state that S<signal> has to be
> S<%02-Signal>.  Maybe gdb/insight throws the answer "S0" away
> and just ignores it and uses its last informations?
> 
> 
> But anyway, when i answer "S00" there's a popup window every
> time i stop.  Looking at "info gdb", i'd say an answer with
> "S" is preferred before "W", "X" or "T".
> Looking at "man -S 7 signal" i'm confused and don't know
> what signal number to use.
> 
> "W00": program exited normally
> "X00": program exited with signal 0, Signal 0
> "T00": program received signal 0, Signal 0
> 
> "S": doesn't work, same problem as before.
> 
> So i guess there has to be a certain number that doesn't bring
> up that window/message.
> 
> "S01": SIGHUP
> "S11": SIGSTOP
> "OK": doesn't work, insight hangs.
> 
> 
> Hmm, don't know, what could the preferred answer be?

05, SIGTRAP.  You should really look at existing stubs.  You probably
want to be using T anyway.

-- 
Daniel Jacobowitz
MontaVista Software                         Debian GNU/Linux Developer