Failures with exelib.exp testcase (was Re: minutes 2010-08-19)

[Mark Wielaard]

On Tue, 2011-01-18 at 20:04 +0530, K.Prasad wrote:
> It appears that the probe is placed on top of the stack...

What do you mean by that?

> pasting the relevant log here (while complete screen logs are attached).
> 
> selected function print_hw
> prologue searching function 'print_hw' 0x10000530-0x10000570@/home/prasadkr/helloworld.c:4
> checking line record 0x10000530@/home/prasadkr/helloworld.c:5
> prologue found function 'print_hw' (naked) = 0x10000530
> probe print_hw@/home/prasadkr/helloworld.c:4 process=/home/prasadkr/helloworld reloc=.absolute pc=0x10000530

So, we actually are putting the probe at address 0x10000530, and uaddr()
does return that address from within the probe context. Good. But our
translation table doesn't contain "print_hw" at that address.

Could you check some things?
- Can you run it under gdb and put a breakpoint at print_hw,
  which address does that give?
- What is the output of eu-readelf -s helloworld?

Thanks,

Mark