[Ltrace-devel] ltrace v0.7.0-git: Session on a Freetz MIPSEL router box

[Petr Machata]

Sedat Dilek <sedat.dilek at gmail.com> writes:

> root at fritz:/var/tmp# ./ltrace --config=ltrace.conf --debug=71 ./a.out
> 2>&1 | tee ltrace-debug-71.txt
>
> - Sedat -
>
> P.S.: ltrace-debug-71.txt (XZ-compressed) file attached!

This looks like setup of _start breakpoint:

DEBUG: plt.c:41: plt_addr 410830 ndx 0x3
DEBUG: ltrace-elf.c:511: do_close_elf()
DEBUG: proc.c:748: proc_add_breakpoint(pid=1658, (null)@0x4005a0)
DEBUG: dict.c:158: dict_find_entry()
DEBUG: dict.c:98: dict_enter()
DEBUG: dict.c:124: new dict entry at 0x47ac68[368]: (0x4005a0,0x4570c8)
DEBUG: breakpoint.c:86: enable_breakpoint: pid=1658, addr=0x4005a0, symbol=(null)
DEBUG: breakpoint.c:49: arch_enable_breakpoint: pid=1658, addr=0x4005a0, symbol=(null)
DEBUG: proc.c:686: added library a.out at 0x400000 (./a.out) to 1658

Later it hits:

DEBUG: events.c:143: next_event()
DEBUG: events.c:200: event from pid 1658
DEBUG: events.c:334: event: BREAKPOINT: pid=1658, addr=0x4005a0
DEBUG: handle_event.c:87: handle_event(pid=1658, type=11)
DEBUG: handle_event.c:175: [1658] event: breakpoint 0x4005a0
DEBUG: handle_event.c:591: handle_breakpoint(pid=1658, addr=0x4005a0)
DEBUG: breakpoints.c:73: address2bpstruct(pid=1658, addr=0x4005a0)
DEBUG: dict.c:158: dict_find_entry()
DEBUG: breakpoints.c:250: delete_breakpoint(pid=1658, addr=0x4005a0)
DEBUG: dict.c:158: dict_find_entry()
DEBUG: breakpoint.c:134: disable_breakpoint: pid=1658, addr=0x4005a0, symbol=(null)
DEBUG: breakpoint.c:100: arch_disable_breakpoint: pid=1658, addr=0x4005a0, symbol=(null)
DEBUG: proc.c:770: proc_remove_breakpoint(pid=1658, (null)@0x4005a0)
DEBUG: dict.c:132: dict_remove(0x4005a0)
DEBUG: proc.c:566: linkmap_init()
DEBUG: proc.c:324: find_dynamic_entry()
DEBUG: proc.c:522: load_debug_struct
DEBUG: breakpoints.c:73: address2bpstruct(pid=1658, addr=0x4005a0)
DEBUG: dict.c:158: dict_find_entry()
DEBUG: trace.c:190: continue_process: pid=1658

But linkmap_init apparently doesn't populate the library list, otherwise
we would see a flutter of new symbols being enabled all over the place.
It seems like ltrace is misenterpreting the debug struct.  That will
need a hands-on work I'm afraid, it's not anything that we can solve by
ping-ponging e-mail messages.

Thank you,
PM