Sorry, I set HIVE_HIGH_TABLES=0 not 1. Please delete my last mail. Based on my test, if I set HIVE_HIGH_TABLES=0 in northbridge/amd/amdk8/Config.lb, Linux will not crash and ACPI works well.
Note. I use the old version filo without libpayload. Zheng -----Original Message----- From: Bao, Zheng Sent: Thursday, June 04, 2009 10:46 AM To: 'Carl-Daniel Hailfinger' Cc: '[email protected]' Subject: RE: [coreboot] [dbm690t] The new acpi table doesn't seem to be correct. Based on my test, if I set HIVE_HIGH_TABLES=1 in northbridge/amd/amdk8/Config.lb, Linux will not crash and ACPI works well. Note. I use the old version filo without libpayload. Zheng -----Original Message----- From: Carl-Daniel Hailfinger [mailto:[email protected]] Sent: Thursday, June 04, 2009 10:42 AM To: Bao, Zheng Cc: [email protected] Subject: Re: [coreboot] [dbm690t] The new acpi table doesn't seem to be correct. I found the HIGH_TABLES bug on the DBM690T. The K8 HIGH_TABLES code places these tables in UMA video memory. Of course that memory is cleared by some payloads and operating systems. However, if I fix that, Linux still crashes during boot because it only sees 32 MB instead of 2 MB RAM. If I disable high tables completely, Linux complains about corrupt ACPI tables and other stuff. Has the tables code been tested recently if it works even in the case where we don't want high tables? Regards, Carl-Daniel -- http://www.hailfinger.org/ -- coreboot mailing list: [email protected] http://www.coreboot.org/mailman/listinfo/coreboot
