[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[pc386] "Non-existent port" interferences with cache on old P90 (solved)



>>>>> "David" == David Decotigny <David.Decotigny at irisa.fr> writes:


David> -----------------------------8<--------------------------------
David> diff -ru rtems-4.0.0-orig/c/src/lib/libbsp/i386/pc386/startup/ldsegs.s rtems-4.0.0/c/src/lib/libbsp/i386/pc386/startup/ldsegs.s
David> --- rtems-4.0.0-orig/c/src/lib/libbsp/i386/pc386/startup/ldsegs.s	Fri Jul 24 00:02:28 1998
David> +++ rtems-4.0.0/c/src/lib/libbsp/i386/pc386/startup/ldsegs.s	Wed Jul 14 14:28:54 1999
David> @@ -56,12 +56,10 @@
David>  	.p2align 4
David>  /*----------------------------------------------------------------------------+
David>  | delay
David> -+------------------------------------------------------------------------------
David> -| Delay is needed after doing I/O. We do it by writing to a non-existent port.
David>  +----------------------------------------------------------------------------*/
David>  SYM(delay):
David> -	outb	al, $0xED	# about 1uS delay
David> -	ret
David> +	jmp	1f
David> +1:	ret
 
David>  /*-------------------------------------------------------------------------+
David>  |         Function: _load_segments
-----------------------------> 8--------------------------------

If have checked linux sources, asm-i386/io.h 

-----------------------------------------------------
#ifdef SLOW_IO_BY_JUMPING
#define __SLOW_DOWN_IO "\njmp 1f\n1:\tjmp 1f\n1:"
#else
#define __SLOW_DOWN_IO "\noutb %%al,$0x80"
#endif
-----------------------------------------------------
I would say, that the oub methods is far less processor speed dependent
and thus better (default for linux...)

Could you try the 0x80 address as linux has been testet on so much machines...



David> Thank you for RTEMS being a freely available product !

David> Nice day everybody,

Thanks for your debug...

-- 
   __                 
  /  `                   	Eric Valette
 /--   __  o _.          	Canon CRF - Communication Dept
(___, / (_(_(__         	Rue de la touche lambert
				35517 Cesson-Sevigne  Cedex
				FRANCE
Tel: +33 (0)2 99 87 68 91	Fax: +33 (0)2 99 84 11 30
E-mail: valette at crf.canon.fr	http://www.crf.canon.fr