Hello Brian,
I've committed this as well. Thanks!
Camiel.
On Tue, Feb 26, 2008 at 3:36 PM, Brian Wheeler <bdw...@in...> wrote:
> fixes for the DMA device:
> * output wrapped in DEBUG_DMA ifdefs.
> * no-op implementation of the different registers
>
> System fixes:
> * ignore reads of 0x480 & 0x4c0 (TLBIV, TLBIA). It looks like
> linux (and others) do a read immediately after a write. Probably to
> keep the pchip's addresses in the TLB?
>
> In Linux's arch/alpha/kernel/core_tsunami.c:
>
> 184 csr = &pchip->tlbia.csr;
> 185 if (((start ^ end) & 0xffff0000) == 0)
> 186 csr = &pchip->tlbiv.csr;
> 187
> 188 /* For TBIA, it doesn't matter what value we write. For TBI,
> 189 it's the shifted tag bits. */
> 190 value = (start & 0xffff0000) >> 12;
> 191
> 192 *csr = value;
> 193 mb();
> 194 *csr;
>
> Debugging fixes:
> * Require a <RETURN> after a failure call to stop before termination.
> This lets the xterm stay alive so I can see what the last message was :)
>
> Brian
>
> -------------------------------------------------------------------------
> This SF.net email is sponsored by: Microsoft
> Defy all challenges. Microsoft(R) Visual Studio 2008.
> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
> _______________________________________________
> Es40-developers mailing list
> Es4...@li...
> https://lists.sourceforge.net/lists/listinfo/es40-developers
>
>
|
