Re: c++filt lossage to gcj 4.0.x

[Nick Clifton <nickc at redhat dot com>]

Hi Andrew,

> This patch:

> removed a critical fflush() from c++filt.  As a result gcj 4.0.x no
> longer works on any system with current binutils.

Fascinating.

> This patch restores the previous binutils behaviour.

> 2005-11-17  Andrew Haley  <aph@redhat.com>
>
> 	* cxxfilt.c (main): Flush ouput at newline.


> Index: cxxfilt.c
> ===================================================================
> RCS file: /cvs/src/src/binutils/cxxfilt.c,v
> retrieving revision 1.10
> diff -u -r1.10 cxxfilt.c
> --- cxxfilt.c	11 Oct 2005 16:33:16 -0000	1.10
> +++ cxxfilt.c	17 Nov 2005 12:29:26 -0000
> @@ -278,8 +278,9 @@
>        /* Echo the whitespace characters so that the output looks
>  	 like the input, only with the mangled names demangled.  */
>        putchar (c);
> +      if (c == '\n')
> +	fflush (stdout);
>      }
>  
> -  fflush (stdout);
>    return 0;
>  }

Restoring the removed fflush is OK, but I think that you should leave in 
the final fflush().  You cannot guarantee that input will terminate with 
a line feed character, so a final flush is a good idea.

Approved with this modification.

Cheers
  Nick