Re: PowerBook 5xx instability fix.

[Nat Sloss <nathanialsloss%yahoo.com.au@localhost>]

On Sun, 30 Mar 2025 12:42:57 Taylor R Campbell wrote:
> > Date: Sun, 30 Mar 2025 11:29:38 +1100
> > From: Nat Sloss <nathanialsloss%yahoo.com.au@localhost>
> > 
> > The attached patch fixes it and I'd like to commit it if there are no
> > objections.
> > 
> >  	maxaddr = high[numranges - 1] - m68k_ptob(1);
> > 
> > -	high[numranges - 1] -= (m68k_round_page(MSGBUFSIZE) + m68k_ptob(1));
> > +	high[numranges - 1] -= (m68k_round_page(MSGBUFSIZE) + m68k_ptob(8));
> 
> Please put a comment here explaining what's going on.
> 
> Is maxaddr still correct?  Should you apply the same substitution on
> that line too?  If not, the comment must explain why maxaddr has
> m68k_ptob(1) subtracted while high[numranges - 1] has m68k_ptob(8)
> subtracted.

Is something like this OK?

Best regards,

Nat

diff -r d60e61cd41cf sys/arch/mac68k/mac68k/pmap_bootstrap.c
--- a/sys/arch/mac68k/mac68k/pmap_bootstrap.c	Sat Mar 29 01:43:38 2025 +0000
+++ b/sys/arch/mac68k/mac68k/pmap_bootstrap.c	Sun Mar 30 13:36:45 2025 +1100
@@ -503,7 +503,14 @@ pmap_bootstrap(paddr_t nextpa, paddr_t f
 	physmem = m68k_btop(avail_remaining + nextpa - firstpa);
 
 	maxaddr = high[numranges - 1] - m68k_ptob(1);
-	high[numranges - 1] -= (m68k_round_page(MSGBUFSIZE) + m68k_ptob(1));
+
+	/*
+	 * Reduce high by an extra 7 pages which seem to be used by the EASC
+	 * ascaudio(4) on some machines.  maxaddr is unchanged as the last page
+	 * can still be safetly used to reboot the system.
+	 */
+	high[numranges - 1] -= (m68k_round_page(MSGBUFSIZE) + m68k_ptob(8));
+
 	avail_end = high[numranges - 1];
 	mem_size = m68k_ptob(physmem);
 	virtual_end = VM_MAX_KERNEL_ADDRESS;