Re: did someone break SMP on amd64-current?

[Jukka Ruohonen <jruohonen%iki.fi@localhost>]

On Wed, Feb 16, 2011 at 10:55:55PM -0800, Hisashi T Fujinaka wrote:
> acpicpu0: failed to evaluate _TSS: AE_AML_BAD_RESOURCE_VALUE
> acpicpu1 at acpi0 (CPU1): ACPI CPU
>
> Is that expected?

Of course not. Can you try what the following printf-diff yields?

- Jukka.

Index: acpi_cpu_tstate.c
===================================================================
RCS file: /cvsroot/src/sys/dev/acpi/acpi_cpu_tstate.c,v
retrieving revision 1.19
diff -u -p -r1.19 acpi_cpu_tstate.c
--- acpi_cpu_tstate.c   30 Jan 2011 08:55:52 -0000      1.19
+++ acpi_cpu_tstate.c   17 Feb 2011 07:09:42 -0000
@@ -474,13 +474,13 @@ acpicpu_tstate_ptc(struct acpicpu_softc 
                         * the fourth bit enables the modulation.
                         */
                        if (reg[i]->reg_bitwidth != 4) {
-                               rv = AE_AML_BAD_RESOURCE_VALUE;
-                               goto out;
+                               aprint_error_dev(sc->sc_dev,
+                                   "bitwidth = %u\n", reg[i]->reg_bitwidth);
                        }
 
                        if (reg[i]->reg_bitoffset != 1) {
-                               rv = AE_AML_BAD_RESOURCE_VALUE;
-                               goto out;
+                               aprint_error_dev(sc->sc_dev,
+                                   "bitoffset = %u\n", reg[i]->reg_bitoffset);
                        }
 
                        break;