This is the mail archive of the gcc-bugs@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]

Re: Bootstrap failure with enable-checking in current CVS (varasm.c: 3423)


>>>>> Franz Sirl writes:

Franz> At 08:16 04.11.99 , Andreas Jaeger wrote:
>> >>>>> Franz Sirl writes:
>> 
>> > At 16:24 01.11.99 , Andreas Jaeger wrote:
>> >> With the current CVS version (updated at 19991101 13:00 UTC) I get a
>> >> bootstrap failure with --enable-checking:
>> 
>> > Welcome to the club :-). I think the appended patch is the correct
>> > solution, as, according to the documentation and the comment, a 
>> SYMBOL_REF
>> > contains a string, whereas a LABEL_REF contains an expression.
>> 
>> Thanks.  The patch fixed the problem - and bootstraping ends in the
>> java frontend (see my message "ICE in Java frontend with
>> --enable-checking (lastfile in gcc/dbxout.c invalid)" - I should have
>> written while compiling Java frontend).

Franz> Good to know this happens on other platforms too. I see this 'lastfile' 
Franz> corruption already during stage2 of a --enable-checking build on 
Franz> powerpc-linux-gnu:

I get it in stage2 while compiling the java frontend.  I conjectured
that lastfile has been garbage collected and needs to be properly
allocated.

It seems to be a general problem - and I'd really like to see a full
bootstrap with --enable-checking (even if it takes ages).

Andreas
-- 
 Andreas Jaeger   
  SuSE Labs aj@suse.de	
   private aj@arthur.rhein-neckar.de


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]