Sven Schnelle (svens@stackframe.org) just uploaded a new patch set to gerrit, which you can find at http://review.coreboot.org/67
-gerrit
commit 1b30ff69f083f2416abfa0f0977ea315368d28e6 Author: Sven Schnelle svens@stackframe.org Date: Wed Jun 29 15:05:28 2011 +0200
i82801gx: read RTC status register to prevent IRQ storm
My Thinkpad appeared dead. After investigation, it turned out that the RTC Alarm was triggering an RTC PM1 SMI, but the SMI handler didn't read the status register, so it was triggered again.
This is a really nasty situation, as it means you have to dissemble your Notebook just to unplug the RTC battery.
Change-Id: I5ac611e8a72deb5f38c86486dbe0693804935723 Signed-off-by: Sven Schnelle svens@stackframe.org --- src/southbridge/intel/i82801gx/smihandler.c | 6 ++++++ 1 files changed, 6 insertions(+), 0 deletions(-)
diff --git a/src/southbridge/intel/i82801gx/smihandler.c b/src/southbridge/intel/i82801gx/smihandler.c index 434bfeb..81a88d8 100644 --- a/src/southbridge/intel/i82801gx/smihandler.c +++ b/src/southbridge/intel/i82801gx/smihandler.c @@ -26,6 +26,7 @@ #include <cpu/x86/cache.h> #include <cpu/x86/smm.h> #include <device/pci_def.h> +#include <pc80/mc146818rtc.h> #include "i82801gx.h"
/* I945 */ @@ -423,6 +424,11 @@ static void southbridge_smi_pm1(unsigned int node, smm_state_save_area_t *state_ reg32 = (7 << 10) | (1 << 13); outl(reg32, pmbase + PM1_CNT); } + + if (pm1_sts & RTC_STS) { + /* read RTC status register to disable the interrupt */ + printk(BIOS_DEBUG, "RTC IRQ status: %02X\n", cmos_read(RTC_REG_C)); + } }
static void southbridge_smi_gpe0(unsigned int node, smm_state_save_area_t *state_save)