[Top] [All Lists]

Grrr: Hello

Subject: Grrr: Hello
From: "David S. Miller" <>
Date: Mon, 28 Aug 1995 14:53:24 -0400
In-reply-to: <> (message from Andreas Busse on Mon, 28 Aug 1995 20:46:12 +0200)
   Date: Mon, 28 Aug 1995 20:46:12 +0200
   From: Andreas Busse <>

   Well, now you know where we're living :-) 
   Thanks for your interest; I'm quite sure that we have similar
   problems with our RISC(y) machines and non-standard I/O devices
   etc. How is it going with Linux/SPARC ?

I can mount MINIX and EXT2 filesystems on sun4c floppies, a.out SunOS
binary loading works and we have the following debugging facilities
for the kernel:

1) kadb - this is sun's assembler level debugger, not too bad for
          traps and entry.S type code

2) kgdb - incredible, debugging the kernel on the source level is
          such a big win at this point

3) printk - Our old friend ;)

I almost got "hello world" to execute, and should be able to get a
statically linked bourne shell up soon after that, we'll see.  I do
have a full serial driver (necessary for kgdb too) which runs two
serial lines up to 9600 baud, and the mouse/keyboard at 1200.  My
keyboard is fully functional (so is the mouse) and does all the nifty
shift-scroll_lock etc. functions that dump kernel information at the
console etc.  Currently I am working also on frame buffer routines to
do virtual console switching on SparcSTations.  Most of my effort is
towards sun4c support, however I was doing the mainline code for the
sun4m and was able to get a quad-processor HyperSparc SS10 to boot ;)
Sun4m MMU matches the i386/Alpha page table mechanism more closely and
as such will be easier, sun4c mmu has the extended tlb model and has
been no end of pains especially with it's virtual cache ;(

   To Stoned and/or Luc: Could you please add David to the list? Thanks!

Thanks Andy ;)

David S. Miller

<Prev in Thread] Current Thread [Next in Thread>
  • Hello, David S. Miller
    • Re: Hello, Andreas Busse
      • Grrr: Hello, David S. Miller <=