Re: [patch ld]: Testsuite fix of vers-script-3 and vers-script-4 tests

[Dave Korn <dave.korn.cygwin@gmail.com>]

On 15/02/2011 16:03, Dave Korn wrote:
> On 15/02/2011 15:55, Dave Korn wrote:
> 
>>   Nope.  This causes two regressions on i686-pc-cygwin:
>>
>>> FAIL: vers-script-3
>>> FAIL: vers-script-4
>>   They pass currently.  They fail with your patch.  
> 
> 
> Ah, hang on: I just updated my sandbox, and I have found a suspect change:
> 
>> 2011-02-14  Mike Frysinger  <...
>>
>> 	* ldlang.c (lang_vers_match): Declare a new c_sym, assign it to
>> 	the bfd_demangle of sym, change users of sym to c_sym when not
>> 	already demangling, and free when done.  Change callers of
>> 	cplus_demangle to bfd_demangle.
> 
> 
>   This is likely to have changed the behaviour I guess.

  Now reading the recent "version scripts and default/C language mangling" and
"[PATCH] bfd/ld: handle ABI prefixes in version scripts" threads.. looks like
the core handling was made language independent by adding demangling, so
removing the pseudo-demangling code in binutils is the correct solution.
(I've been a couple of days behind HEAD trying to debug a patchset for the ld
plugin api.)

  Patch is OK after all, thanks!

    cheers,
      DaveK