Radeon 9000 Pro corruption problem on PPC when DRI is enabled

[Ari Entlich]

On Thu, 2006-08-24 at 13:48 +0200, Roland Scheidegger wrote:
> Ian Romanick wrote:
> >> That said, I don't understand why it would get identified as a x86 bios
> >> if it isn't, as xf86ReadPciBIOS checks that itself, so in fact there
> >> shouldn't be any need to put in the same verification again in the
> >> radeon driver.
> > 
> > I don't see any test for that in linuxPciHandleBIOS.  If
> > linuxPciHandleBIOS returns success, none of the rest of the code in
> > HandlePciBIOS gets executed.
> Argh missed that one. Is that a bug or a feature? I find it strange that 
> the sanity checks are vastly different for both methods.
> And it still doesn't explain why mapping the bios failed with 6.8.2 but 
> "succeeds" with 7.1.
> 
> Anyway, here's a new patch which should hopefully be more correct. If 
> that still doesn't rule out the bios as not being x86 video bios I don't 
> know what could be done (it should be the same test as radeonfb uses, 
> which does figure out it's OF after all).
> 
> Roland
> 
> plain text document attachment (radeon_bios.c.diff)
> --- radeon_bios.c~	2006-05-03 19:21:45.000000000 +0200
> +++ radeon_bios.c	2006-08-24 13:38:26.000000000 +0200
> @@ -46,6 +46,7 @@
>  {
>      RADEONInfoPtr info     = RADEONPTR(pScrn);
>      int tmp;
> +    unsigned short dptr;
>  
>      if (!(info->VBIOS = xalloc(RADEON_VBIOS_SIZE))) {
>  	xf86DrvMsg(pScrn->scrnIndex, X_ERROR,
> @@ -77,7 +78,22 @@
>  	xfree (info->VBIOS);
>  	info->VBIOS = NULL;
>  	return FALSE;
> -    } 
> +    }
> +
> +    /* verify it's x86 bios not OF firmware, copied from radeonfb */
> +    dptr = RADEON_BIOS16(0x18);
> +    /* if pci data signature is wrong assume x86 video bios anyway */
> +    if (RADEON_BIOS32(dptr) != (('R' << 24) | ('I' << 16) | ('C' << 8) | 'P')) {
> +       xf86DrvMsg(pScrn->scrnIndex, X_WARNING,
> +		   "ROM pci data signature incorrect, ignoring\n");
> +    }
> +    else if (info->VBIOS[dptr + 0x14] != 0x0) {
> +	xf86DrvMsg(pScrn->scrnIndex, X_WARNING,
> +		   "Not a x86 BIOS ROM Image, BIOS data will not be used\n");
> +	xfree (info->VBIOS);
> +	info->VBIOS = NULL;
> +	return FALSE;
> +    }
>  
>      if (info->VBIOS) info->ROMHeaderStart = RADEON_BIOS16(0x48);
>  

Yup, this patch fixed it! Thanks a bunch :).

Thanks,
 Ari