RE: Suggested (easier) fix to ... RE: Regressions problem (200 failures)

[Donn Terry <donnte at microsoft dot com>]

I can't answer why any better than the author of the change 
that started all this ;-) .

> 2000-02-17  Mark Mitchell  <mark@codesourcery.com>
>
>       * function.c (thread_prologue_and_epilogue_insns): Put a line note
>       after the prologue.

My suspicion is that there were situations where the "first statement"
breakpoint wasn't happening correctly, but I don't know details beyond
this.

Agreed, it only affects a subset of all gcc/gdb implementations, which
clearly
explains why everyone wasn't seeing it.

Donn

> -----Original Message-----
> From: Mark Mitchell [mailto:mark@codesourcery.com]
> Sent: Tuesday, May 23, 2000 1:07 AM
> To: Donn Terry
> Cc: kettenis@wins.uva.nl; 
> Peter.Schauer@Regent.E-Technik.TU-Muenchen.DE;
> kingdon@redhat.com; gdb@sourceware.cygnus.com
> Subject: RE: Suggested (easier) fix to ... RE: Regressions 
> problem (200
> fa ilures)
> 
> 
> >>>>> "Donn" == Donn Terry <donnte@microsoft.com> writes:
> 
>     Donn> Remember this one?  I finally found time (or rather
>     Donn> desperation :-) ) to look at it more closely.
> 
> The analysis sounds good.  So, this bug should only affect platforms
> that don't define HAS_INIT_SECTION; platforms that, for example, use
> ELF .init/.fini sections won't run into this issue.
> 
> Why are we emitting a line note at the end of the prologue?  (And
> thus before the call to __main?)  Couldn't we wait until the first
> real statement?
> 
> --
> Mark Mitchell                   mark@codesourcery.com
> CodeSourcery, LLC               http://www.codesourcery.com
>