> Date: Tue, 10 Aug 2010 23:15:32 +0200
> From: Mike Belopuhov <[email protected]>
>
> Hi,
>
> I have a Thinkpad X41 which doesn't do ACPI suspend properly
> (it immediately wakes up after 'zzz'). So I wonder if anyone
> can give me hints what to look at, or if someone is willing
> to take a closer look at acpidump? I won't mind doing some
> debugging myself, but at the moment I'm clueless what to start
> looking at. As I said it tries to suspend, blinks the "moon",
> then it comes back as if something woke it up. I wasn't
> running X while doing that test.
Might be worth trying the attached diff. Fixes a similar issue on my
old Dell laptop.
Index: acpi_machdep.c
===================================================================
RCS file: /cvs/src/sys/arch/i386/i386/acpi_machdep.c,v
retrieving revision 1.38
diff -u -p -r1.38 acpi_machdep.c
--- acpi_machdep.c 8 Aug 2010 21:23:41 -0000 1.38
+++ acpi_machdep.c 10 Aug 2010 22:07:39 -0000
@@ -216,6 +216,9 @@ acpi_cpu_flush(struct acpi_softc *sc, in
wbinvd();
}
+#include <dev/ic/mc146818reg.h>
+void mc146818_write(void *, u_int, u_int);
+
int
acpi_sleep_machdep(struct acpi_softc *sc, int state)
{
@@ -223,6 +226,8 @@ acpi_sleep_machdep(struct acpi_softc *sc
printf("%s: acpi_sleep_machdep: no FACS\n", DEVNAME(sc));
return (ENXIO);
}
+
+ mc146818_write(NULL, MC_REGB, MC_REGB_24HR);
/* i386 does lazy pmap_activate */
pmap_activate(curproc);
