Can someone please test this on Geode? Thanks.
On 13.02.2008 01:02, Carl-Daniel Hailfinger wrote:
> The patch is correct if you compare it with latest v3 code, but
> incorrect if you compare it to v2 because v3 has inverted logic. See below.
>
> Ron? This one is for you.
>
> On 12.02.2008 19:26, Carl-Daniel Hailfinger wrote:
>
>> Enable Suspend-to-RAM code based on Kconfig option.
>> Does not work yet, so the Kconfig option depends on BROKEN.
>>
>> Signed-off-by: Carl-Daniel Hailfinger <[EMAIL PROTECTED]>
>>
>> Index: LinuxBIOSv3-if0crud/southbridge/amd/cs5536/cs5536.c
>> ===================================================================
>> --- LinuxBIOSv3-if0crud/southbridge/amd/cs5536/cs5536.c (Revision 587)
>> +++ LinuxBIOSv3-if0crud/southbridge/amd/cs5536/cs5536.c (Arbeitskopie)
>> @@ -488,7 +488,7 @@
>> }
>> sb = (struct southbridge_amd_cs5536_dts_config
>> *)dev->device_configuration;
>>
>> -#if 0
>> +#ifdef CONFIG_SUSPEND_TO_RAM
>> if (!IsS3Resume())
>> {
>> struct acpi_init *aci = acpi_init_table;
>>
>>
>
> Looking at the equivalent v2 code and the history of these lines in v3,
> I found a logic inversion committed in
>
>> r385 | rminnich | 2007-06-27 22:52:49 +0200 (Mi, 27 Jun 2007) | 9 lines
>>
>> Trivial fixes for some typos, and a major fix for an unitialized
>> variable.
>>
>> Add a license to dts.
>>
>> Signed-off-by: Ronald G. Minnich <[EMAIL PROTECTED]>
>> Acked-by: Stefan Reinauer <[EMAIL PROTECTED]>
>> void chipsetinit(void)
>> {
>> struct device *dev;
>> msr_t msr;
>> u32 msrnum;
>> - struct southbridge_amd_cs5536_config *sb =
>> - (struct southbridge_amd_cs5536_config *)dev->device_configuration;
>> + struct southbridge_amd_cs5536_config *sb;
>> struct msrinit *csi;
>>
>> post_code(P80_CHIPSET_INIT);
>> -
>> + dev = dev_find_device(PCI_VENDOR_ID_AMD, PCI_DEVICE_ID_AMD_CS5536_ISA,
>> 0);
>> + if (! dev) {
>> + printk(BIOS_ERR, "%s: Could not find the south bridge!\n",
>> __FUNCTION));
>> + return;
>> + }
>> + sb = (struct southbridge_amd_cs5536_config *)dev->device_configuration;
>> - /* we hope NEVER to be in linuxbios when S3 resumes
>> +#if 0
>> if (! IsS3Resume()) */
>>
>
> In the old (before r385) version, the code below was always executed.
> In the new version, the code below is never executed.
> v2 behaves like the code before r385.
>
>
>> {
>> struct acpiinit *aci = acpi_init_table;
>> @@ -531,6 +534,7 @@
>>
>> pm_chipset_init();
>> }
>> +#endif
>>
>> /* set hd IRQ */
>> outl(GPIOL_2_SET, GPIO_IO_BASE + GPIOL_INPUT_ENABLE);
>>
>
> I believe we have to revert the semantic change.
>
Enable Suspend-to-RAM code based on config option. Revert semantics to
those we had in v2 and before r385.
This causes pm_chipset_init() to be called on Geode LX.
Signed-off-by: Carl-Daniel Hailfinger <[EMAIL PROTECTED]>
Index: LinuxBIOSv3-s3/southbridge/amd/cs5536/cs5536.c
===================================================================
--- LinuxBIOSv3-s3/southbridge/amd/cs5536/cs5536.c (Revision 590)
+++ LinuxBIOSv3-s3/southbridge/amd/cs5536/cs5536.c (Arbeitskopie)
@@ -488,8 +488,9 @@
}
sb = (struct southbridge_amd_cs5536_dts_config
*)dev->device_configuration;
-#if 0
+#ifdef CONFIG_SUSPEND_TO_RAM
if (!IsS3Resume())
+#endif
{
struct acpi_init *aci = acpi_init_table;
for (; aci->ioreg; aci++) {
@@ -498,7 +499,6 @@
}
pm_chipset_init();
}
-#endif
/* Set HD IRQ. */
outl(GPIOL_2_SET, GPIO_IO_BASE + GPIOL_INPUT_ENABLE);
Index: LinuxBIOSv3-s3/device/Kconfig
===================================================================
--- LinuxBIOSv3-s3/device/Kconfig (Revision 590)
+++ LinuxBIOSv3-s3/device/Kconfig (Arbeitskopie)
@@ -87,3 +87,13 @@
endmenu
+menu "Power management"
+
+config SUSPEND_TO_RAM
+ bool "Suspend-to-RAM (S3)"
+ depends EXPERIMENTAL && BROKEN
+ help
+ Enable support for Suspend-to-RAM (S3) functionality.
+
+endmenu
+
--
http://www.hailfinger.org/
--
coreboot mailing list
[email protected]
http://www.coreboot.org/mailman/listinfo/coreboot
