a strange problem with rtems_io_register_name

Cedric Aubert cedric_aubert at yahoo.fr
Fri Aug 29 11:36:05 UTC 2003


I have the same problem than Stan with
a modified BSP from efi68k with a 68000.

rtems_io_register_name return RTEMS_TOO_MANY
because mknod return -1

I use the hello world with no modification, witch
is provided with rtems package not the example
package.

Cedric

--- Joel Sherrill <joel.sherrill at OARcorp.com> wrote:
> Stan wrote:
> 
> >>Can't you make it bigger than that?  I would
> rather see something
> >>like 64K until you know that low memory isn't an
> issue.
> >>
> >>    
> >>
> >
> >I tryed to change _HeapSize in linkcmds with
> 0x10000 and 0x0.
> >RTEMS_Malloc_Initialize don't fail.
> >With _HeapSize=0, rtems allocate 448Ko for heap.
> >
> >
> >  
> >
> >>>>This should be one of the first attempts to
> malloc
> >>>>memory in the boot sequence.
> >>>>        
> >>>>
> >>>In that case, why mknod don't failed in main task
> ?
> >>>      
> >>>
> >>This runs before the first task.
> >>    
> >>
> >
> >If mknod fail in the boot sequence, why don't fail
> after ?
> >Reverse it stands to reason.
> >  
> >
> Let's go back to basics then.
> 
> What test are you running and have you modified it? 
> 
> 
> > 
> >  
> >
> >>>How to check simply the filesystem for remove
> this ambiguity ?
> >>>      
> >>>
> >>I suppose it could do something more specific with
> the error
> >>returned by mknod().  Suggestions welcomed.
> >>
> >>    
> >>
> >
> >Has anybody used the mcf5206elite bsp (on latest
> ss-20030703) ?
> >I have other custom bsp based upon efi68k on this
> latest;
> >It work properly ...
> >
> >I don't how to solve this problem :(
> >
> >
> >Stan, sad.
> >
> >  
> >
> 
> 
> 


__________________________________
Do you Yahoo!?
Yahoo! SiteBuilder - Free, easy-to-use web site design software
http://sitebuilder.yahoo.com



More information about the users mailing list