[PATCH] arm64: fix single step issue in kgdb

[Daniel Thompson]

On Fri, Apr 20, 2018 at 08:57:08PM +0800, Wei Li wrote:
> Because the former mail is in html by my mistake, it is rejected by
> the mailing list server.

I'm afraid that switching to plain text isn't quite enough; the patch
appears to be corrupt.

I'd personally suggest switching over to `git send-email` but there are
many other ways to send out patches. If you can't use git perhaps take a
look at the following for configuration hints:
https://www.kernel.org/doc/html/latest/process/email-clients.html


> I just update the diffstat info and resend it
> in plain text. There is no difference in patch. Sorry to make you
> confused, please ignore the former one.

Didn't you also change the patch description? That would usually be a
reason to increment the patch version.


> >> diff -uprN linux-4.16.2_org/arch/arm64/include/asm/debug-monitors.h linux-4.16.2_mod/arch/arm64/include/asm/debug-monitors.h
> >> --- linux-4.16.2_org/arch/arm64/include/asm/debug-monitors.h    2018-04-12 18:30:01.000000000 +0800
> >> +++ linux-4.16.2_mod/arch/arm64/include/asm/debug-monitors.h    2018-04-19 00:19:10.134999268 +0800
> >> @@ -135,6 +135,9 @@ static inline int reinstall_suspended_bp
> >>  
> >>  int aarch32_break_handler(struct pt_regs *regs);
> >>  
> >> +void set_regs_spsr_ss(struct pt_regs *regs);
> >> +void clear_regs_spsr_ss(struct pt_regs *regs);

Why do we need to bring out clear_regs_spsr_ss()?  It does not seem to 
be being used.


> >> @@ -252,7 +253,7 @@ static int kgdb_step_brk_fn(struct pt_re
> >>      if (!kgdb_single_step)
> >>          return DBG_HOOK_ERROR;
> >>  
> >> -    kgdb_handle_exception(1, SIGTRAP, 0, regs);
> >> +    kgdb_handle_exception(0, SIGTRAP, 0, regs);

Based on a quick glance this change looks sensible to me, but I don't
think it was mentioned clearly in the patch header.


Daniel.