[DB] BIOS - ACPI - data collecting

[Adarion]

Back with more laptop stuff:
I was very fond to get my hands and FWTS on the following machine yesterday:

HP Pavilion dm1-3100 series (dm1-3101eg) powered by an AMD E-350. The pal was running Ubuntu on it.

the tables

Signature Length Revision  OemId     OemTableId   OemRevision CompilerId CompilerRevision

    DSDT   36711     01    "HP    "  "INSYDE  "    F0000000    "ACPI"     00040000
    FACS      64
    FACP     244     04    "HPQOEM"  "SLIC-MPC"    00000003    "ACPI"     00040000
    HPET      56     01    "HP    "  "INSYDE  "    00000001    "ACPI"     00040000
    APIC     132     02    "HP    "  "INSYDE  "    00000001    "ACPI"     00040000
    MCFG      60     01    "HP    "  "INSYDE  "    00000001    "ACPI"     00040000
    BOOT      40     01    "HP    "  "INSYDE  "    00000001    "ACPI"     00040000
    SLIC     374     01    "HPQOEM"  "SLIC-MPC"    00000001    "ACPI"     00040000
    SSDT     990     01    "HP    "  "INSYDE  "    00000001    "AMD "     00000001
    SSDT    4858     02    "HP    "  "INSYDE  "    00000001    "MSFT"     04000000
    XSDT     100     01    "HPQOEM"  "SLIC-MPC"    00000003    "    "     01000013
    RSDT      68     01    "HPQOEM"  "SLIC-MPC"    00000003    "    "     01000013
    RSDP                   "HPQOEM"

Found 13 ACPI tables [20100107]

cut from dmidecode

# dmidecode 2.9
SMBIOS 2.7 present.
24 structures occupying 1248 bytes.
Table at 0x000E3890.

Handle 0x0000, DMI type 0, 24 bytes
BIOS Information
        Vendor: Hewlett-Packard
        Version: F.05
        Release Date: 03/04/2011
        ROM Size: 2048 kB
        Characteristics:
                PCI is supported
                BIOS is upgradeable
                BIOS shadowing is allowed
                Boot from CD is supported
                Selectable boot is supported
                BIOS ROM is socketed
                EDD is supported
                Japanese floppy for NEC 9800 1.2 MB is supported (int 13h)
                Japanese floppy for Toshiba 1.2 MB is supported (int 13h)
                5.25"/360 KB floppy services are supported (int 13h)
                5.25"/1.2 MB floppy services are supported (int 13h)
                3.5"/720 KB floppy services are supported (int 13h)
                3.5"/2.88 MB floppy services are supported (int 13h)
                8042 keyboard services are supported (int 9h)
                CGA/mono video services are supported (int 10h)
                ACPI is supported
                USB legacy is supported
                Targeted content distribution is supported
        BIOS Revision: 15.5
        Firmware Revision: 150.31

Handle 0x0001, DMI type 1, 27 bytes
System Information
        Manufacturer: Hewlett-Packard
        Product Name: HP Pavilion dm1 Notebook PC
        Wake-up Type: Power Switch
        SKU Number: LE942EA#ABD
        Family: 103C_5335KV G=N L=CON B=HP S=PAV

FWTS Results, all tests
Results generated by fwts: Version V0.18.04 (built Wed Sep 15 10:41:31 BST 2010).

Summary: 61 passed, 15 failed, 0 warnings, 1 aborted.
abort=intel microcode
1 high
9 medium
5 low
fails:
Found 1 errors in kernel log.
No fan information present
3 in section dmi_decode
1 in fadt 1 in maxreadreq 1 in mtrr 1 in osilinux 1 in syntaxcheck 1 in thermal_trip

[Adarion]

And the next machine on the same evening was a (total overkill for the purpose...)

Acer Aspire 5950g (i7-2630QM and a big radeon chip, omg. The guy ain't doing much more than typing on that machine...)

Tables:

Signature Length Revision  OemId     OemTableId   OemRevision CompilerId CompilerRevision

    DSDT   41377     01    "ACRSYS"  "ACRPRDCT"    00000000    "1025"     00040000
    FACS      64
    FACP     244     04    "ACRSYS"  "ACRPRDCT"    00000001    "1025"     00040000
    ASF!     165     20    "ACRSYS"  "ACRPRDCT"    00000001    "1025"     00040000
    HPET      56     01    "ACRSYS"  "ACRPRDCT"    00000001    "1025"     00040000
    APIC     140     02    "ACRSYS"  "ACRPRDCT"    00000001    "1025"     00040000
    MCFG      60     01    "ACRSYS"  "ACRPRDCT"    00000001    "1025"     00040000
    SLIC     374     01    "ACRSYS"  "ACRPRDCT"    00000001    "1025"     00040000
    WDAT     548     01    "ACRSYS"  "ACRPRDCT"    00000001    "1025"     00040000
    SSDT    3010     01    "ACRSYS"  "ACRPRDCT"    00001000    "1025"     00040000
    BOOT      40     01    "ACRSYS"  "ACRPRDCT"    00000001    "1025"     00040000
    ASPT      52     07    "ACRSYS"  "ACRPRDCT"    00000001    "1025"     00040000
    SSDT    2316     01    "ACRSYS"  "ACRPRDCT"    00003000    "1025"     00040000
    SSDT    2454     01    "ACRSYS"  "ACRPRDCT"    00003000    "1025"     00040000
    SSDT    1078     01    "ACRSYS"  "ACRPRDCT"    00001000    "1025"     00040000
    XSDT     140     01    "ACRSYS"  "ACRPRDCT"    00000001    "    "     01000013
    RSDT      88     01    "ACRSYS"  "ACRPRDCT"    00000001    "    "     01000013
    RSDP                   "ACRSYS"

Found 18 ACPI tables [20100107]

Um, well. Looks strange like Acer had overwritten everything with their own id.

dmidecode:

# dmidecode 2.9
SMBIOS 2.7 present.
52 structures occupying 2472 bytes.
Table at 0x000E4640.

Handle 0x0000, DMI type 0, 24 bytes
BIOS Information
	Vendor: Acer
	Version: V1.16
	Release Date: 05/04/2011
	ROM Size: 2560 kB
	Characteristics:
		PCI is supported
		BIOS is upgradeable
		BIOS shadowing is allowed
		Boot from CD is supported
		Selectable boot is supported
		EDD is supported
		Japanese floppy for NEC 9800 1.2 MB is supported (int 13h)
		Japanese floppy for Toshiba 1.2 MB is supported (int 13h)
		5.25"/360 KB floppy services are supported (int 13h)
		5.25"/1.2 MB floppy services are supported (int 13h)
		3.5"/720 KB floppy services are supported (int 13h)
		3.5"/2.88 MB floppy services are supported (int 13h)
		8042 keyboard services are supported (int 9h)
		CGA/mono video services are supported (int 10h)
		ACPI is supported
		USB legacy is supported
		BIOS boot specification is supported
		Targeted content distribution is supported
	BIOS Revision: 16.240

Handle 0x0001, DMI type 1, 27 bytes
System Information
	Manufacturer: Acer
	Product Name: Aspire 5950G
	Version: V1.16
	Wake-up Type: Power Switch
	SKU Number:  
	Family:

FWTS
Results generated by fwts: Version V0.18.04
Summary: 65 passed, 31 failed, 0 warnings, 2 aborted.
high 4
med 25
low 2
abort: microcode thingy, damn I should leave this test away
and No thermal zones found on this machine

[edmon]

I'll get a dump using the Intel CD later

i have some problems using firmwarekit-r3.iso
so i tried ubuntu fwts:

http://pastebin.com/XuaExZL8

[edmon]

I'll get a dump using the Intel CD later

i have some problems using firmwarekit-r3.iso
so i tried ubuntu fwts:

http://pastebin.com/6Cymkm7k

[whitecat]

I will update the last posts soon.

Nevertheless, given what have been told here, I will create a new datasheet with this new firmware kit, since its results are probably more accurate.

Also, I will no more include the ACPI compiler ID. Given all the data we already have, I don't think this information is finally worth noting.

I will update the first post of the topic in these ways.

[Adarion]

Nevertheless, given what have been told here, I will create a new datasheet with this new firmware kit, since its results are probably more accurate.

Well, please keep in mind that is has some experimental tests and some fail probably on every machine. So "failure" is in that case not a problem of the very machine it is tested on. Also things like "fan could not be read" is useless on a fanless, silent system without any fan at all. In my case I just ran all tests out of curiosity. I should still have the full logs and dumps around here.

After all the question is: Would it be fair to compare machines when we used "different" "benchmarks"? (different in "" cause fwts is based on LFK anyway)

Also, I will no more include the ACPI compiler ID. Given all the data we already have, I don't think this information is finally worth noting.

Why?
Yes, some things were strange like Acer writing their own id over all fields, but still a DSDT compiled with "MSFT" might give you a hint how things might work out (or not). If you need support entering data I might offer it.


FWTS it is a package in Ubuntu so people might check even without any boot media and hopefully there will be some development.
Interesting to read from the blog entry of the main developer:

http://kernel.ubuntu.com/%7Ecking/presentations/firmware/fwts-overview.odp

gives an overview over FWTS.

Though on

https://launchpad.net/fwts

I just found a recent version
Version 0.23.25 uploaded on 2011-10-11
so Deb/Ubuntu users can probably just apt-get it.
webgit here

http://kernel.ubuntu.com/git?p=cking/fwts/.git

As soon as I find the time I'll try to check out a more recent version. I was using the live distribution on a CFDisk which included it but in an older version. That was good since I didn't have an installed Linux available on all the tested machines.


Small addition to my M4A78-E:
there was a "wrong checksum" error during acpidump and it is "just" Socket AM2+ (which can also fit an AM3 nicely)
Bios is v 2511


I'm still fighting with a few ooolder machines to get something booted there. (Without freezing the machine.) Either they don't like USB-Boots or the CD with LFK 3 froze the system during kernel bootup.


Oh, and does anybody know how to fix a DSDT?

There was once a collection of them on a SF project but that was about dead last time I checked. I mean, wouldn't it be awesome to have some database with fixed ACPI tables so users could download them and hand them to their Kernel at boot time? (Or even, if that is legally possible to have known to be good tables right in the kernel?)

[whitecat]

Why?
Yes, some things were strange like Acer writing their own id over all fields, but still a DSDT compiled with "MSFT" might give you a hint how things might work out (or not).

In fact I do not see any correlation between the ACPI compiler used and the ftk3 results.

[Adarion]

Well, I don't know exactly what the lfk tests, but maybe there are also tests for disassembling recomp. the DSDT or things that are influenced by the implementation of the DSDT.
And since Microsoft's compiler produces faulty code (I call it intention) non-Windows systems will have problems with that tables since they deviate from specifications. So they need a lot of workarounds. Of course Windows knows about these "shortcomings" of their compiler and will work perfectly with these things. I find it rather interesting to see which compiler was used in that context.

[zchronos]

Laptop TOSHIBA Satellite A505-SP6022L with OpenSUSE 12.1 RC2 (x86_64)

# acpixtract -l acpi.dump

Signature Length Revision  OemId     OemTableId   OemRevision CompilerId CompilerRevision

    DSDT   47251     02    "TOSINV"  "TOSINV00"    00000001    "MSFT"     01000013
    FACS      64
    FACP     244     04    "TOSINV"  "TOSINV00"    00000001    "MSFT"     01000013
    ASF!     165     20    "TOSINV"  "TOSINV00"    00000001    "MSFT"     01000013
    HPET      56     01    "TOSINV"  "TOSINV00"    00000001    "MSFT"     01000013
    APIC     140     02    "TOSINV"  "TOSINV00"    00000001    "MSFT"     01000013
    MCFG      60     01    "TOSINV"  "TOSINV00"    00000001    "MSFT"     01000013
    SLIC     374     01    "TOSINV"  "TOSINV00"    00000001    "MSFT"     01000013
    BOOT      40     01    "TOSINV"  "TOSINV00"    00000001    "MSFT"     01000013
    SSDT     750     01    "INTEL "  "SataAhci"    00001000    "INTL"     20051117
    ASPT      52     04    "INTEL "  "Calpella"    00000001    "MSFT"     01000013
    WDAT     548     01    "INTEL "  "Calpella"    00000001    "MSFT"     01000013
    SSDT    2545     01    "PmRef "  "CpuPm   "    00003000    "INTL"     20051117
    XSDT     124     01    "TOSINV"  "TOSINV00"    00000001    "    "     01000013
    RSDP                   "TOSINV"
    SSDT     934     01    "PmRef "  "Cpu0Ist "    00003000    "INTL"     20051117
    SSDT    1587     01    "PmRef "  "Cpu0Cst "    00003001    "INTL"     20051117
    SSDT     771     01    "PmRef "  "ApIst   "    00003000    "INTL"     20051117
    SSDT     281     01    "PmRef "  "ApCst   "    00003000    "INTL"     20051117

Found 19 ACPI tables [20100107]

The out from dmidecode

# dmidecode 2.11
SMBIOS 2.6 present.
42 structures occupying 1714 bytes.
Table at 0x000EA8B0.

Handle 0x0000, DMI type 0, 24 bytes
BIOS Information
        Vendor: INSYDE
        Version: 1.40
        Release Date: 08/31/2010
        ROM Size: 1536 kB
        Characteristics:
                PCI is supported
                PNP is supported
                BIOS is upgradeable
                BIOS shadowing is allowed
                Boot from CD is supported
                Selectable boot is supported
                EDD is supported
                Japanese floppy for NEC 9800 1.2 MB is supported (int 13h)
                Japanese floppy for Toshiba 1.2 MB is supported (int 13h)
                3.5"/720 kB floppy services are supported (int 13h)
                3.5"/2.88 MB floppy services are supported (int 13h)
                8042 keyboard services are supported (int 9h)
                CGA/mono video services are supported (int 10h)
                ACPI is supported
                USB legacy is supported
                AGP is supported
                Smart battery is supported
                BIOS boot specification is supported
                Function key-initiated network boot is supported
                Targeted content distribution is supported
        BIOS Revision: 1.40                                                                                                                                                                     
        Firmware Revision: 5.40                                                                                                                                                                 
                                                                                                                                                                                                
Handle 0x0001, DMI type 1, 27 bytes                                                                                                                                                             
System Information                                                                                                                                                                              
        Manufacturer: TOSHIBA                                                                                                                                                                   
        Product Name: Satellite A505                                                                                                                                                            
        Version: PSAT9U-009LM1
        Serial Number: 3A290854Q
        UUID: 205E3557-B536-DF11-BBA9-00266C4DE9F1
        Wake-up Type: Power Switch
        SKU Number: Calpella
        Family: Intel_Mobile

The out from the iso image:

-------------------------------------------------
  Date: Wed Nov 16 11:20:44 2011                                       
*                                               *
* Firmwarekit (release 3)                       *
* http://www.linuxfirmwarekit.org               *
*                                               *
*                                               *
* For more information on test descriptions     *
* and details on what the PASS/INFO/WARN/FAIL   *
* results mean, go to: Documentation/TestsInfo. *
*                                               *
-------------------------------------------------

KERNEL VERSION: Linux testmachine 2.6.22.9 #1 SMP Mon Oct 1 15:17:15 PDT 2007 i686 i686 i386 GNU/Linux


SUMMARY:  4 Fails, 3 Warns, 13 Pass, 20 Total

=================================================
* Plugin name: pciresource
* Result: FAIL

* Title: Validate assigned PCI resources
* Description: This test is currently a placeholder and just checks the kernel log for complaints about PCI resource errors. In the future the idea is to actually perform a validation step on all PCI resources against a certain rule-set.
================================================

[FAIL]-HPET resources incorrect

hpet_resources: 0xfed00000 is busy

=================================================
* Plugin name: thermal_trip
* Result: FAIL

* Title: ACPI passive thermal trip points
* Description: This test determines if the passive trip point works as expected.
================================================

[FAIL]-Changing passive trip point seems uneffective in Zone THRM.

=================================================
* Plugin name: cpufreq
* Result: FAIL

* Title: CPU frequency scaling tests (1-2 mins)
* Description: For each processor in the system, this test steps through the various frequency states (P-states) that the BIOS advertises for the processor. For each processor/frequency combination, a quick performance value is measured. The test then validates that: 
  1) Each processor has the same number of frequency states
  2) Higher advertised frequencies have a higher performance
  3) No duplicate frequency values are reported by the BIOS
  4) Is BIOS wrongly doing Sw_All P-state coordination across cores
  5) Is BIOS wrongly doing Sw_Any P-state coordination across cores

================================================

[INFO]-10 CPU frequency steps supported

 Frequency | Speed 
-----------+---------
  2.32 Ghz | 100.0 %
  2.32 Ghz |  89.5 %
  2.18 Ghz |  84.3 %
  2.05 Ghz |  79.0 %
  1.92 Ghz |  73.8 %
  1.78 Ghz |  68.5 %
  1.65 Ghz |  63.2 %
  1466 Mhz |  58.0 %
  1333 Mhz |  52.7 %
  1199 Mhz |  47.5 %


[PASS]-P-state coordination done by Harware

[FAIL]-Firmware not implementing hardware coordination cleanly. Firmware using SW_ALL instead?


[FAIL]-Firmware not implementing hardware coordination cleanly. Firmware using SW_ANY instead?


=================================================
* Plugin name: dmi
* Result: FAIL

* Title: DMI information check
* Description: This test checks the DMI/SMBIOS tables for common errors.
================================================

[FAIL]-No SMBIOS nor DMI entry point found.

[FAIL]-No SMBIOS nor DMI entry point found.

[FAIL]-No SMBIOS nor DMI entry point found.

[FAIL]-No SMBIOS nor DMI entry point found.

[FAIL]-No SMBIOS nor DMI entry point found.

[FAIL]-No SMBIOS nor DMI entry point found.

[FAIL]-No SMBIOS nor DMI entry point found.

[FAIL]-No SMBIOS nor DMI entry point found.

[FAIL]-No SMBIOS nor DMI entry point found.

[FAIL]-No SMBIOS nor DMI entry point found.

[FAIL]-No SMBIOS nor DMI entry point found.

[FAIL]-No SMBIOS nor DMI entry point found.

=================================================
* Plugin name: maxreadreq
* Result: WARN

* Title: PCI Express MaxReadReq tuning
* Description: This test checks if the firmware has set MaxReadReq to a higher value on non-montherboard devices
================================================

[WARN]-MaxReadReq for device pci://00:00:1b.0 is low (128)

=================================================
* Plugin name: edd
* Result: WARN

* Title: EDD Boot disk hinting
* Description: This test verifies if the BIOS directs the operating system on which storage device to use for booting (EDD information). This is important for systems that (can) have multiple disks. Linux distributions increasingly depend on this info to find out on which device to install the bootloader.
================================================

[INFO]-device 80: The system boots from device 0000:00:1f.2 channel: 255 device: 1 


[WARN]-device 80: No matching MBR signature (0x5b2d21cf) found for the boot disk

=================================================
* Plugin name: chk_hpet
* Result: WARN

* Title: HPET configuration test
* Description: This test checks the HPET PCI BAR for each timer block in the timer.The base address is passed by the firmware via an ACPI table.IRQ routing and initialization is also verified by the test.
================================================

[WARN]-HPET driver in the kernel is enabled, inaccurate results follow.

[INFO]-HPET found, VendorID is: 8086

=================================================
* Plugin name: battery
* Result: INFO

* Title: Battery tests
* Description: This test reports which (if any) batteries there are in the system. In addition, for charging or discharging batteries, the test validates that the reported 'current capacity' properly increments/decrements in line with the charge/discharge state. 

This test also stresses the entire battery state reporting codepath in the ACPI BIOS, and any warnings given by the ACPI interpreter will be reported.
================================================

[INFO]-Battery BAT0 is model PA353 and is currently charged

=================================================
* Plugin name: acpiinfo
* Result: INFO

* Title: General ACPI information
* Description: This test checks the output of the in-kernel ACPI CA against common error messages that indicate a bad interaction with the bios, including those that point at AML syntax errors.
================================================

[INFO]-DSDT was compiled by the Microsoft AML compiler

ACPI: DSDT BF7ED000, B893 (r2 TOSINV TOSINV00        1 MSFT  1000013)

=================================================
* Plugin name: DMAR
* Result: INFO

* Title: (experimental) DMA Remapping (VT-d) test
* Description: Verify if DMA remapping is sane.
================================================

[INFO]-No DMAR ACPI table found.

=================================================
* Plugin name: mtrr
* Result: PASS

* Title: MTRR validation
* Description: This test validates the MTRR setup against the memory map to detect any inconsistencies in cachability.
================================================

=================================================
* Plugin name: acpicompile
* Result: PASS

* Title: DSDT AML verification
* Description: This test first disassembles the DSDT of the BIOS, and then uses the IASL compiler from Intel to recompile the code. The IASL copiler is much stricter in detecting deviations from the ACPI specification and can find numerous defects that other AML compilers cannot find. Fixing these defects increases the probability that the BIOS will operate well with a variety of operating systems.
================================================

=================================================
* Plugin name: SUN
* Result: PASS

* Title: SUN duplicate test
* Description: This makes sure that each SUN (Slot Unique Number) that is called in the DSDT through the Name() method is unique, no duplicates should be found.
================================================

[PASS]-Tested _SUN ids, successfully found no duplicates

=================================================
* Plugin name: fan
* Result: PASS

* Title: Fan tests
* Description: This test reports how many fans there are in the system. It also checks for the current status of the fan(s).
================================================

[PASS]-Fan FAN status is: on

=================================================
* Plugin name: virt
* Result: PASS

* Title: VT/VMX Virtualization extensions
* Description: This test checks if VT/VMX is set up correctly
================================================

=================================================
* Plugin name: ebda
* Result: PASS

* Title: EBDA region
* Description: This test validates if the EBDA region is mapped and reserved in the E820 table.
================================================

[PASS]-EBDA region is correctly reserved in the E820 table.

=================================================
* Plugin name: os2gap
* Result: PASS

* Title: OS/2 memory hole test
* Description: This test checks if the OS/2 15Mb memory hole is absent
================================================

[PASS]-Successfully found no 15mb memory hole

=================================================
* Plugin name: ethernet
* Result: PASS

* Title: Ethernet functionality
* Description: This test is currently a placeholder for a more advanced ethernet test. Currently the only check performed is that a link is acquired within 45 seconds of enabling the interface. 45 seconds is close to the value most Linux distributions use as timeout value.

In the future the plan is to also perform actual data transfer tests as part of the ethernet test, to validate interrupt routing and other per-NIC behaviors.
================================================

=================================================
* Plugin name: apicedge
* Result: PASS

* Title: (experimental) APIC Edge/Level check
* Description: This test checks if legacy interrupts are edge and PCI interrupts are level
================================================

=================================================
* Plugin name: microcode
* Result: PASS

* Title: Processor microcode update
* Description: This test verifies if the firmware has put a recent version of the microcode into the processor at boot time. Recent microcode is important to have all the required features and errata updates for the processor.
================================================

[whitecat]

Already in the list