Date: Wed, 17 Dec 2014 13:36:29 +0000 From: Ruslan Bukin <br@FreeBSD.org> To: Ian Lepore <ian@freebsd.org> Cc: svn-src-head@freebsd.org, svn-src-all@freebsd.org, src-committers@freebsd.org Subject: Re: svn commit: r275858 - head/sys/mips/beri Message-ID: <20141217133629.GA29458@bsdpad.com> In-Reply-To: <1418820674.1015.14.camel@freebsd.org> References: <201412170934.sBH9Ytqp071808@svn.freebsd.org> <1418820674.1015.14.camel@freebsd.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On Wed, Dec 17, 2014 at 05:51:14AM -0700, Ian Lepore wrote:
> > Modified: head/sys/mips/beri/beri_machdep.c
> > ==============================================================================
> > --- head/sys/mips/beri/beri_machdep.c Wed Dec 17 07:47:25 2014 (r275857)
> > +++ head/sys/mips/beri/beri_machdep.c Wed Dec 17 09:34:54 2014 (r275858)
> > @@ -88,6 +88,11 @@ static void
> > mips_init(void)
> > {
> > int i;
> > +#ifdef FDT
> > + struct mem_region mr[FDT_MEM_REGIONS];
> > + int mr_cnt, val;
> > + int j;
> > +#endif
> >
> > for (i = 0; i < 10; i++) {
> > phys_avail[i] = 0;
> > @@ -102,6 +107,29 @@ mips_init(void)
> >
> > physmem = realmem;
> >
> > +#ifdef FDT
> > + if (fdt_get_mem_regions(mr, &mr_cnt, &val) == 0) {
> > +
> > + physmem = btoc(val);
> > +
> > + KASSERT((phys_avail[0] >= mr[0].mr_start) && \
> > + (phys_avail[0] < (mr[0].mr_start + mr[0].mr_size)),
> > + ("First region is not within FDT memory range"));
> > +
> > + /* Limit size of the first region */
> > + phys_avail[1] = MIN(mr[0].mr_size, ctob(realmem));
> > + dump_avail[1] = phys_avail[1];
> > +
> > + /* Add the rest of regions */
> > + for (i = 1, j = 2; i < mr_cnt; i++, j+=2) {
> > + phys_avail[j] = mr[i].mr_start;
> > + phys_avail[j+1] = mr[i].mr_size;
> > + dump_avail[j] = phys_avail[j];
> > + dump_avail[j+1] = phys_avail[j+1];
> > + }
> > + }
> > +#endif
> > +
> > init_param1();
> > init_param2(physmem);
> > mips_cpu_init();
> >
>
> This doesn't look right, the values in the odd-numbered array slots for
> phys_avail and dump_avail need to be addresses, not sizes. That is,
> avail[j] = start, avail[j+1] = start+size (note: not start+size-1).
>
>
Right, thanks. Sounds like it should be like that
Index: sys/mips/beri/beri_machdep.c
===================================================================
--- sys/mips/beri/beri_machdep.c (revision 275861)
+++ sys/mips/beri/beri_machdep.c (working copy)
@@ -117,13 +117,13 @@
("First region is not within FDT memory range"));
/* Limit size of the first region */
- phys_avail[1] = MIN(mr[0].mr_size, ctob(realmem));
+ phys_avail[1] = (mr[0].mr_start + MIN(mr[0].mr_size, ctob(realmem)));
dump_avail[1] = phys_avail[1];
/* Add the rest of regions */
for (i = 1, j = 2; i < mr_cnt; i++, j+=2) {
phys_avail[j] = mr[i].mr_start;
- phys_avail[j+1] = mr[i].mr_size;
+ phys_avail[j+1] = (mr[i].mr_start + mr[i].mr_size);
dump_avail[j] = phys_avail[j];
dump_avail[j+1] = phys_avail[j+1];
}
Ruslan
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20141217133629.GA29458>