From Bugzilla Helper: User-Agent: Mozilla/5.0 (X11; U; Linux x86_64; rv:1.7.3) Gecko/20041020 Firefox/0.10.1 Description of problem: acpi_processor_get_performance_states fails on some machines that have empty _PSS entries. The result is that the powernow-k8 module fails to load on some SMP machines (this is hardware specific, some machines may work, many dont). example (see the "Zero" entries after the valid power state entries): Scope(\_PR_) { Processor(CPU1, 1, 0x5010, 0x6) { Name(_PCT, Package(0x2) { Buffer(0x11) {0x82, 0xc, 0x0, 0x7f, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x79, 0x0 }, Buffer(0x11) {0x82, 0xc, 0x0, 0x7f, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x79, 0x0 }, }) Name(_PSS, Package(0x4) { Package(0x6) { 0x00000898, 0x00017318, 0x00000064, 0x00000007, 0xc020120e, 0x0000020e, }, Package(0x6) { 0x000007d0, 0x000160bc, 0x00000064, 0x00000007, 0xc020128c, 0x0000028c, }, Package(0x6) { 0x00000708, 0x0001287c, 0x00000064, 0x00000007, 0xc020130a, 0x0000030a, }, Package(0x6) { 0x000003e8, 0x00008d04, 0x00000064, 0x00000007, 0xc0201482, 0x00000482, }, Zero, Zero, Zero, Zero, Zero, Zero, Zero, Zero, Version-Release number of selected component (if applicable): linux-2.6.9 How reproducible: Sometimes Steps to Reproduce: 1. find a machine with ACPI _PSS enablement and "Zero" entries in the _PSS table (use acpidump to get this info if needed) 2. boot SMP enabled kernel 3. modprobe powernow-k8 Actual Results: FATAL: Error inserting powernow_k8 (/lib/modules/2.6.9-prep/kernel/arch/x86_64/kernel/cpufreq/powernow-k8.ko): No such device exact error: bad ACPI data: _PSS Expected Results: module loads correctly and finds all power states in _PSS table Additional info: this appears to be fixed in both the 2.4.31 and 2.6.12 kernels this should be a MUSTFIX as we will otherwise have no SMP Powernow enablement
Created attachment 117479 [details] patch to limit max performance supported states (_PSS) This patch scans PSS entries and counts the valid ones, and adds a sanity check to weed out invalid (or zero) PSS entries. This patch was tested on a variety of machine and resolves the problem (the powernow-k8 module loads properly). It seems like more machines than not have PSS tables that have been "padded out" with zeros.
this should be a MUSTFIX instead of a CANFIX. The Powernow driver will otherwise fail to load on many new machines (including most of the machines we test with)
An advisory has been issued which should help the problem described in this bug report. This report is therefore being closed with a resolution of ERRATA. For more information on the solution and/or where to find the updated files, please follow the link below. You may reopen this bug report if the solution does not work for you. http://rhn.redhat.com/errata/RHSA-2005-514.html