2003-07-03 16:13:53

by Disconnect

Subject: [2.4][2.5][Trivial Patch] Bug in i386/kernel/process.c?

I was poking into how to force a warm boot (found it easily enough) and
started reading process.c, where the details live.

We have:
static int reboot_mode;
int reboot_thru_bios;

static int __init reboot_setup(char *str)
{
...set reboot_mode/reboot_thru_bios according to reboot=...
}
__setup("reboot=", reboot_setup);

Farther down:
/* Write 0x1234 to absolute memory location 0x472. The BIOS
reads
this on booting to tell it to "Bypass memory test (also warm
boot)". This seems like a fairly standard thing that gets
set by
REBOOT.COM programs, and the previous reset routine did this
too. */

*((unsigned short *)0x472) = reboot_mode;
(similar code farther down in !reboot_thru_bios)

....but reboot_mode doesn't seem to be initialized if you don't set it
via reboot=...? (Same for reboot_thru_bios)

A simple patch to use the defaults (according to the code comments) is
below; if this is right please push accordingly:

Its against 2.4 but should apply to 2.5 as well.
--- build-dis5-final/arch/i386/kernel/process.c.orig 2003-07-03 12:15:36.000000000 -0400
+++ build-dis5-final/arch/i386/kernel/process.c 2003-07-03 12:16:46.000000000 -0400
@@ -152,8 +152,8 @@
__setup("idle=", idle_setup);

static long no_idt[2];
-static int reboot_mode;
-int reboot_thru_bios;
+static int reboot_mode = 0x1234;
+int reboot_thru_bios = 1;

#ifdef CONFIG_SMP
int reboot_smp = 0;

--
Disconnect <[email protected]>

2003-07-04 16:25:39

by Disconnect

Subject: Re: [2.4][2.5][Trivial Patch] Bug in i386/kernel/process.c?

Er, that patch won't apply to 2.5 - the code is the same, but its moved
to reboot.c.

Same change, diff't file/offset.

On Thu, 2003-07-03 at 12:26, Disconnect wrote:
> I was poking into how to force a warm boot (found it easily enough) and
> started reading process.c, where the details live.
>
> We have:
> static int reboot_mode;
> int reboot_thru_bios;
>
> static int __init reboot_setup(char *str)
> {
> ...set reboot_mode/reboot_thru_bios according to reboot=...
> }
> __setup("reboot=", reboot_setup);
>
> Farther down:
> /* Write 0x1234 to absolute memory location 0x472. The BIOS
> reads
> this on booting to tell it to "Bypass memory test (also warm
> boot)". This seems like a fairly standard thing that gets
> set by
> REBOOT.COM programs, and the previous reset routine did this
> too. */
>
> *((unsigned short *)0x472) = reboot_mode;
> (similar code farther down in !reboot_thru_bios)
>
> ....but reboot_mode doesn't seem to be initialized if you don't set it
> via reboot=...? (Same for reboot_thru_bios)
>
> A simple patch to use the defaults (according to the code comments) is
> below; if this is right please push accordingly:
>
> Its against 2.4 but should apply to 2.5 as well.
> --- build-dis5-final/arch/i386/kernel/process.c.orig 2003-07-03 12:15:36.000000000 -0400
> +++ build-dis5-final/arch/i386/kernel/process.c 2003-07-03 12:16:46.000000000 -0400
> @@ -152,8 +152,8 @@
> __setup("idle=", idle_setup);
>
> static long no_idt[2];
> -static int reboot_mode;
> -int reboot_thru_bios;
> +static int reboot_mode = 0x1234;
> +int reboot_thru_bios = 1;
>
> #ifdef CONFIG_SMP
> int reboot_smp = 0;
--
Disconnect <[email protected]>