Re: Is there single step debugging support being added to aarch64-linux?

[pinskia at gmail dot com]

> On Jun 15, 2015, at 4:48 AM, Prafull Suryawanshi <prafull.net@gmail.com> wrote:
> 
> I am not sure why hw stepper not works here (might be limitation of hw
> I am using).
> That is may be the infrun loops as it never gets stop signal. (the
> output I earlier pasted).
> Is it ok to provide patch for software single step like arm here?
> (I am preparing one. It will have simulate aarch64 instruction set and
> implement aarch64 version of functions of
> arm_linux_software_single_step, arm_insert_single_step_breakpoint and
> arm_get_next_pc.)


We do need one but for stepping over ldxr/stxr pairs. 

Thanks,
Andrew

> 
> Thanks,
> Prafull
> 
>> On Mon, Jun 15, 2015 at 4:49 PM,  <pinskia@gmail.com> wrote:
>> 
>> 
>> 
>> 
>>> On Jun 15, 2015, at 4:07 AM, Prafull Suryawanshi <prafull.net@gmail.com> wrote:
>>> 
>>> I am using gdb to talk to kgdb, so is not issue with gdb?
>>> gdb is not able to set software single step like it do for arm.
>>> There is implementation for gdbarch_software_single_step_p in arm gdb
>>> but not in aarch64 gdb.
>>> 
>>> In fact, I had small implementation to get next_pc at fixed offset
>>> (+4) and it starts working. (implementing full version)
>> 
>> 
>> Why can't you use the hardware single stepper for kgdb?
>> 
>> Thanks,
>> Andrew
>> 
>>> 
>>> Let me know.
>>> 
>>> Thanks,
>>> Prafull
>>> 
>>> 
>>> On Mon, Jun 15, 2015 at 4:27 PM, Prafull Suryawanshi
>>> <prafull.net@gmail.com> wrote:
>>>> I am using kgdb and trying to do single step (software single step).
>>>> I see that infrun loops. (I enabled remote_debug option and compiled
>>>> gdb 7.7 as 7.9 has issue with connection)
>>>> 
>>>> infrun: clear_proceed_status_thread (Thread 3462)
>>>> infrun: clear_proceed_status_thread (Thread 3444)
>>>> infrun: proceed (addr=0xffffffffffffffff, signal=GDB_SIGNAL_DEFAULT, step=1)
>>>> infrun: resume (step=1, signal=GDB_SIGNAL_0), trap_expected=1, current
>>>> thread [Thread 3462] at 0xffffffc00021014c
>>>> Sending packet: $Hcd86#7d...Ack
>>>> Packet received: OK
>>>> Sending packet: $s#73...Ack
>>>> infrun: wait_for_inferior ()
>>>> Packet received: T05thread:0d86;
>>>> infrun: target_wait (-1, status) =
>>>> infrun:   42000 [Thread 3462],
>>>> infrun:   status->kind = stopped, signal = GDB_SIGNAL_TRAP
>>>> infrun: infwait_normal_state
>>>> infrun: TARGET_WAITKIND_STOPPED
>>>> Sending packet: $g#67...Ack
>>>> Packet received:
>>>> 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
>>>> infrun: stop_pc = 0xffffffc000210150
>>>> Sending packet: $mffffffc000210150,4#4d...Ack
>>>> Packet received: fd030091
>>>> Sending packet: $mffffffc00021014c,4#7f...Ack
>>>> Packet received: fd7ba3a9
>>>> Sending packet: $mffffffc000210150,4#4d...Ack
>>>> Packet received: fd030091
>>>> infrun: stepping inside range [0xffffffc00021014c-0xffffffc000210170]
>>>> Sending packet: $Z0,ffffffc00021014c,4#c8...Ack
>>>> Packet received: OK
>>>> infrun: resume (step=1, signal=GDB_SIGNAL_0), trap_expected=0, current
>>>> thread [Thread 3462] at 0xffffffc000210150
>>>> Sending packet: $Hc0#db...Ack
>>>> Packet received: OK
>>>> Sending packet: $s#73...Ack
>>>> infrun: prepare_to_wait
>>>> 
>>>> Then I build the gdb using arch_dump enable option
>>>> gdbarch_dump: gdbarch_software_single_step_p() = 0
>>>> 
>>>> I see that in gdb code, function aarch64_linux_software_single_step not present.
>>>> 
>>>> I guess above loop expected if aarch64_linux_software_single_step absent, right?
>>>> 
>>>> Thanks,
>>>> Prafull
>>>> 
>>>>> On Mon, Jun 15, 2015 at 4:08 PM, Andreas Schwab <schwab@suse.de> wrote:
>>>>> Prafull Suryawanshi <prafull.net@gmail.com> writes:
>>>>> 
>>>>>> I am trying it on 64 bit armv8 system but not able to doit.
>>>>>> Found issue in gdb source code and absence of this function in
>>>>>> aarch64-linux-tdep.c file. Anyone knows if support getting added or
>>>>>> not?
>>>>> 
>>>>> The stepi command is working fine on aarch64.  What do you get if you
>>>>> try it?
>>>>> 
>>>>> Andreas.
>>>>> 
>>>>> --
>>>>> Andreas Schwab, SUSE Labs, schwab@suse.de
>>>>> GPG Key fingerprint = 0196 BAD8 1CE9 1970 F4BE  1748 E4D4 88E3 0EEA B9D7
>>>>> "And now for something completely different."