[PATCH] data: klog.json: cater for CMOS handler errors

[Alex Hung]

On 03/06/2013 07:46 PM, Colin King wrote:
> From: Colin Ian King <colin.king at canonical.com>
>
> We're seeing some attempts at firmware to access CMOS memory
> even though the kernel does not (yet) support this.  This could
> be because of implementations of Peak Shift which accesses CMOS
> memory to figure out the time.  Add some better advice to explain
> this error.
>
> Signed-off-by: Colin Ian King <colin.king at canonical.com>
> ---
>   data/klog.json | 16 ++++++++++++++++
>   1 file changed, 16 insertions(+)
>
> diff --git a/data/klog.json b/data/klog.json
> index 7c99c03..a9477fc 100644
> --- a/data/klog.json
> +++ b/data/klog.json
> @@ -2654,6 +2654,14 @@
>      "compare_mode": "regex",
>      "log_level": "LOG_LEVEL_HIGH",
>      "tag": "FWTS_TAG_ACPI_NO_HANDLER",
> +   "pattern": "ACPI Error.*No handler for Region.*SystemCMOS",
> +   "advice": "An attempt to dispatch an address space or operation region access to CMOS memory has failed because the Linux kernel does not yet have a CMOS handler.  This may have little or no detrimental effect, however it is worth examining the AML to see what kind of access is occurring to see if the firmware really requires CMOS access.  For example, recent firmware may implement Peak Shift, which requires access CMOS memory and this can trigger this error.",
> +   "label": "KlogAcpiNoHandlerForRegion"
> +  },
> +  {
> +   "compare_mode": "regex",
> +   "log_level": "LOG_LEVEL_HIGH",
> +   "tag": "FWTS_TAG_ACPI_NO_HANDLER",
>      "pattern": "ACPI Error.*No handler for Region",
>      "advice": "",
>      "label": "KlogAcpiNoHandlerForRegion"
> @@ -2662,6 +2670,14 @@
>      "compare_mode": "regex",
>      "log_level": "LOG_LEVEL_HIGH",
>      "tag": "FWTS_TAG_ACPI_NO_HANDLER",
> +   "pattern": "ACPI Error.*Region SystemCMOS.* has no handler",
> +   "advice": "An attempt to read or write a single field to a CMOS Operation Region has failed because the Linux kernel does not yet have a CMOS handler.  This may have little or no detrimental effect, however it is worth examining the AML to see what kind of access is occurring to see if the firmware really requires CMOS access.  For example, recent firmware may implement Peak Shift, which requires access CMOS memory and this can trigger this error.",
> +   "label": "KlogAcpiRegionHasNoHandler"
> +  },
> +  {
> +   "compare_mode": "regex",
> +   "log_level": "LOG_LEVEL_HIGH",
> +   "tag": "FWTS_TAG_ACPI_NO_HANDLER",
>      "pattern": "ACPI Error.*Region .* has no handler",
>      "advice": "",
>      "label": "KlogAcpiRegionHasNoHandler"
>

I submitted a patch to upstream and hopefully they will accept it. :).