patch for Regressions from "[rfc] -gdwarf-3 and DW_OP_call_frame_cfa"

[H.J. Lu]

On Fri, Jun 12, 2009 at 6:10 AM, Hans-Peter
Nilsson<hans-peter.nilsson@axis.com> wrote:
>> Date: Fri, 12 Jun 2009 14:26:33 +0200
>> From: Hans-Peter Nilsson <hp@axis.com>
>
>> The testsuite bogusly produces e.g. -gdwarf-21 and -gdwarf-23
>> making this patch cause numerous (as in 83) regressions.
>> Obviously now made a bug, but I'm not sure of the intent of that
>> part of the proc gcc-dg-debug-runtest (the 'foreach level {1 ""
>> 3} {' loop); it looks as if it really was intended to produce
>> those options.  Maybe it's even used in real code?  Mark is
>> fingered by svn blame but it appears he just consolidated this
>> code.
>
> Looks like the following is what is now intended.  Still, I'm
> not sure this is right, as there are testsuite regressions even
> though the gcc.log says the options are handled right.  So, I
> have to do a full regtest to see whether it's your patch or a
> bug in the following patch that causes e.g.:

See:

http://gcc.gnu.org/ml/gcc-regression/2009-06/msg00080.html

Revision 148408 is the cause.

-- 
H.J.