[patch/rfc] Use frame_type for sigtramp test in infrun.c

Andrew Cagney cagney@gnu.org
Tue Apr 6 23:33:00 GMT 2004


>>> based on other discussion, I suspect the test will need to be:
>>>
>>>>
>>>> if (frame_id_inner (this_frame_id, step_frame_id)
>>>>     || frame_id_eq (....))
>>>>
>>>> the former is to handle a signal delivery that resumed the inferior at the signal handler and not the signal trampoline.
>>
>>
>>
>> Won't anywhere you do that frame_id_inner test add another breakage for
>> sigaltstack?
> 
> 
> Hmm, true.  Sigaltstack will break any inner test.

Ok (talked this through with jeff).  The id_inner test isn't needed. 
The case of GDB doing a nexti and ending up in a signal handler should 
have been handled earlier (in time).  At the point where the signal 
first arrives - GDB should have set a breakpoint at the signal return 
address and then continued the inferior with the signal.

Time to turn infrun.c into a state machine :-/

Andrew




More information about the Gdb-patches mailing list