Re: [PATCH] [gdbserver] Disable conditional breakpoints on no-hardware-single-step targets

[Luis Machado <lgustavo@codesourcery.com>]

On 04/30/2015 02:10 PM, Antoine Tremblay wrote:
>
>
> On 04/30/2015 12:23 PM, Yao Qi wrote:
>> From: Yao Qi <yao.qi@linaro.org>
>>    /* Define an ARM-mode breakpoint; we only set breakpoints in the C
>>       library, which is most likely to be ARM.  If the kernel supports
>>       clone events, we will never insert a breakpoint, so even a Thumb
>>       C library will work; so will mixing EABI/non-EABI gdbserver and
>>       application.  */
>> #ifndef __ARM_EABI__
>>    (const unsigned char *) &arm_breakpoint,
>> #else
>>    (const unsigned char *) &arm_eabi_breakpoint,
>> #endif
>>
>> note that the comments are no longer valid as C library can be compiled
>> in thumb mode.
>
> Could we update the comments at the same time ?...
>>
>> When GDBserver steps over a breakpoint in arm mode function, which
>> returns to thumb mode, GDBserver will insert arm mode breakpoint by
>> mistake and the program will crash.  GDBserver alone is unable to
>> determine the arm/thumb mode given a PC address.  See how GDB does
>> it in arm-tdep.c:arm_pc_is_thumb.
>>
>> After thinking about how to teach GDBserver inserting right breakpoint
>> (arm or thumb) for a while, I reconsider it from a different direction
>> that it may be unreasonable to run target-side conditional breakpoint for
>> targets without hardware single step.  Pedro also pointed this out here
>> https://sourceware.org/ml/gdb-patches/2015-04/msg00337.html
>>
>
> I'm looking into teaching gdbserver about inserting the right
> breakpoint and possibly software single-step like this post :
> https://sourceware.org/ml/gdb/2012-10/msg00077.html wanted to do...
>
> It would also fix the problem where we would step-over other breakpoints...
>
> But it does seems very complex to say the least, I'm still early in my
> investigation but if you could share your thoughts on how you came to
> think of it as unreasonable to fix arm-mode and possibly the
> single-stepping it would be appreciated ?

I like the idea of having gdbserver learn how to properly 
software-single-step, allowing us to share the knowledge GDB already has.

Disabling a feature, on the other hand, sound like a backward movement.

People with knowledge on each architecture can probably help fine tune 
those to their needs.