step_resume breakpoints native+remote

[Markus Deuling <deuling at de dot ibm dot com>]

Hi,

I still debug that pthread + dlopen application. The routine called
by pthread_create is test(). I set a breakpoint to test() and step over
pthread_create(). When GDB stops at test(), I see the 'step resume' breakpoint
with number 0.  

Num Type           Disp Enb Address    What
1   breakpoint     keep y   0x08048525 in main at main.c:23
       breakpoint already hit 1 time
-3  shlib events   keep y   0x45d3c3c9 <_dl_debug_state>
       breakpoint already hit 2 times
-8  thread events  keep y   0x4961ba84 <__nptl_create_event>
       breakpoint already hit 1 time
-9  thread events  keep y   0x4961ba89 <__nptl_death_event>
-13 longjmp        keep n   0x496218f0 <siglongjmp>
-14 longjmp        keep n   0x45d72d00 <siglongjmp>
-15 longjmp        keep n   0x496218f0 <siglongjmp>
-16 longjmp resume keep n   0x00000000
2   breakpoint     keep y   0x080485d6 in test at main.c:42
       breakpoint already hit 1 time
0   step resume    keep y   0x0804855b <main+71> thread 1
       stop only in stack frame at 0xbfa74f40
       stop only in thread 1

I can see that only when debugging native. If I use gdbserver, there appears no 'step resume'
breakpoint:

Num Type           Disp Enb Address    What
-3  shlib events   keep y   0x45d3c3c9 <_dl_debug_state>
       breakpoint already hit 2 times
1   breakpoint     keep y   0x08048525 in main at main.c:23
       breakpoint already hit 1 time
2   breakpoint     keep y   0x080485d6 in test at main.c:42
       breakpoint already hit 2 times
-11 longjmp        keep n   0x496218f0 <siglongjmp>
-12 longjmp        keep n   0x45d72d00 <siglongjmp>
-13 longjmp        keep n   0x496218f0 <siglongjmp>
-14 longjmp resume keep n   0x00000000

Why isn't there a step_resume breakpoint then? Does gdbserver handle that in a different way ?


Regards,
Markus

--
Markus Deuling
GNU Toolchain for Linux on Cell BE
deuling@de.ibm.com