C2 or C3 on Presario 2110US?

C2 or C3 on Presario 2110US?

[Carl Thompson]

Hello,

     I just upgraded my laptop to a Compaq Presario 2110US and am
pleased to report for the most part Linux ACPI works excellently on it
(without the problems I have on my Sony Vaios).  The one exception is
that Linux ACPI does not have any processor support for it (everything
in /proc/acpi/processor/CPU0 reads "<not supported>" except power which
tells me that only the C1 state is available.  I believe the problem is
that the chipset used in the laptop is reasonably new (ATI/ALI combo),
in particular the ATI bridges.  Is there any effort to get ACPI working
on this chipset?  In particular I would like to see C2 or C3 because the
2110 has the loudest fan I've heard on a laptop.  I am willing to help
in any way I can and here is the output of lspci:

00:00.0 Host bridge: ATI Technologies Inc: Unknown device cab0 (rev 13)
00:01.0 PCI bridge: ATI Technologies Inc U1/A3 AGP Bridge [IGP 320M]
(rev 01)
00:02.0 USB Controller: ALi Corporation USB 1.1 Controller (rev 03)
00:06.0 Multimedia audio controller: ALi Corporation M5451 PCI AC-Link
Controller Audio Device (rev 02)
00:07.0 ISA bridge: ALi Corporation M1533 PCI to ISA Bridge [Aladdin IV]
00:08.0 Modem: ALi Corporation Intel 537 [M5457 AC-Link Modem]
00:0a.0 CardBus bridge: O2 Micro, Inc. OZ6912 Cardbus Controller
00:10.0 IDE interface: ALi Corporation M5229 IDE (rev c4)
00:11.0 Bridge: ALi Corporation M7101 PMU
00:12.0 Ethernet controller: National Semiconductor Corporation DP83815
(MacPhyter) Ethernet Controller
01:05.0 VGA compatible controller: ATI Technologies Inc Radeon Mobility U1

The processor itself is a Mobile AMD Athlon 2000+.

Thanks,
Carl Thompson




-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf

C2 or C3 on Presario 2110US?

[Carl Thompson]

     I am resending this because I got no responses to my original
message from the ACPI team on this issue.  I would really apppreciate it
if someone could let me know if this is something that is being looked
at, will be looked at or will never be looked at!  Actually, any sort of
acknowledgement would be nice.  What ever happened to the talk of
implementing a bug tracking system?

     If there's any other information I can provide or anything I can do
to help please let me know.  I am also willing to try to help more
actively if someone can tell me what information I need and where in the
ACPI code it is handled.

Thank you for responding,
Carl Thompson

------ Original Message ------

Hello,

     I just upgraded my laptop to a Compaq Presario 2110US and am
pleased to report for the most part Linux ACPI works excellently on it
(without the problems I have on my Sony Vaios).  The one exception is
that Linux ACPI does not have any processor support for it (everything
in /proc/acpi/processor/CPU0 reads "<not supported>" except power which
tells me that only the C1 state is available.  I believe the problem is
that the chipset used in the laptop is reasonably new (ATI/ALI combo),
in particular the ATI bridges.  Is there any effort to get ACPI working
on this chipset?  In particular I would like to see C2 or C3 because the
2110 has the loudest fan I've heard on a laptop.  I am willing to help
in any way I can and here is the output of lspci:

00:00.0 Host bridge: ATI Technologies Inc: Unknown device cab0 (rev 13)
00:01.0 PCI bridge: ATI Technologies Inc U1/A3 AGP Bridge [IGP 320M]
(rev 01)
00:02.0 USB Controller: ALi Corporation USB 1.1 Controller (rev 03)
00:06.0 Multimedia audio controller: ALi Corporation M5451 PCI AC-Link
Controller Audio Device (rev 02)
00:07.0 ISA bridge: ALi Corporation M1533 PCI to ISA Bridge [Aladdin IV]
00:08.0 Modem: ALi Corporation Intel 537 [M5457 AC-Link Modem]
00:0a.0 CardBus bridge: O2 Micro, Inc. OZ6912 Cardbus Controller
00:10.0 IDE interface: ALi Corporation M5229 IDE (rev c4)
00:11.0 Bridge: ALi Corporation M7101 PMU
00:12.0 Ethernet controller: National Semiconductor Corporation DP83815
(MacPhyter) Ethernet Controller
01:05.0 VGA compatible controller: ATI Technologies Inc Radeon Mobility U1

The processor itself is a Mobile AMD Athlon 2000+.

Thanks,
Carl Thompson





-------------------------------------------------------
This SF.net email is sponsored by:Crypto Challenge is now open! 
Get cracking and register here for some mind boggling fun and 
the chance of winning an Apple iPod:
http://ads.sourceforge.net/cgi-bin/redirect.pl?thaw0031en

RE: C2 or C3 on Presario 2110US?

[Breathnach, Proinnsias (Dublin)]

Carl,
	I'm no ACPI expert here, but I'd imagine the problem is probably
closer related to the DSDT (the tables the BIOS manufacturer included to
tell the OS about what it can / can't do ...

In your case it seems that the manufacturer (of either the BIOS itself or
probably the laptop) didn't include any C-State information in the ACPI
tables, or if they did that this information is unreadable by the Linux ACPI
parser.

Can you output your DSDT and copy the list on it ... someone with more
experience may have an answer as to why it's failing like that !

What versions of the kernel / ACPI are you using are also important factors.

P

> -----Original Message-----
> Actually, any sort of
> acknowledgement would be nice.  What ever happened to the talk of
> implementing a bug tracking system?
> 
> 



-------------------------------------------------------
This SF.net email is sponsored by:Crypto Challenge is now open! 
Get cracking and register here for some mind boggling fun and 
the chance of winning an Apple iPod:
http://ads.sourceforge.net/cgi-bin/redirect.pl?thaw0031en

RE: C2 or C3 on Presario 2110US?

[Carl Thompson]

Hmm, I didn't realize the C states worked through the DSDT too!  I am
currently using kernel 2.4.21pre4 with ACPI 20030218.  Today I will try
ACPI 20030228 and see whether that makes a difference.  OK, so I
disassembled the DSDT and recompiled it, and apparently Compaq is pretty
good with its DSDTs these days because I just got three minor warnings
for methods not returning a value.  When I recompile the kernel I will
use the fixed DSDT.  But I don't think that these warnings would be
enough to cause no C2-C3 support.

I also forgot to mention that I have no ACPI fan support, but the fan
does in fact vary speeds depending on the temperature so I think it is
just designed to be controlled by the BIOS (or whatever).  That's OK
because it is varying speeds and seems to be working as it should (even
though it would be nice to control it).

Is it possible this machine just doesn't do C2-C3, processor limit
control or processor performance control?  If that is the case, then why
does Windows seem to be running cooler?  The fan is often off in Windows
but never off in Linux.  Also battery life seems to be better in Windows
whereas on my other laptops it's better in Linux.  And yes, I am using
CPUFreq's PowerNow support to throttle down the CPU speed and voltage to
the minimum.

Does anyone know of any tools to display the exact CPU speed, voltage
and temperature under Windows so I can see exactly what it's doing?  The
PowerNow dashboard from AMD just gives a (simulated) analog guage which
does not give accurate measurements.

Thanks for your help,
Carl Thompson


Quoting "Breathnach, Proinnsias (Dublin)" <breatpro-Lp/cVzEoVybn/p8AFm7ortBPR1lH4CV8@public.gmane.org>:

> Carl,
> 	I'm no ACPI expert here, but I'd imagine the problem is probably
> closer related to the DSDT (the tables the BIOS manufacturer included
> to
> tell the OS about what it can / can't do ...
> 
> In your case it seems that the manufacturer (of either the BIOS
> itself or
> probably the laptop) didn't include any C-State information in the
> ACPI
> tables, or if they did that this information is unreadable by the
> Linux ACPI
> parser.
> 
> Can you output your DSDT and copy the list on it ... someone with
> more
> experience may have an answer as to why it's failing like that !
> 
> What versions of the kernel / ACPI are you using are also important
> factors.
> 
> P
> 
> > -----Original Message-----
> > Actually, any sort of
> > acknowledgement would be nice.  What ever happened to the talk of
> > implementing a bug tracking system?
> >
> >
> 
> 
> 






-------------------------------------------------------
This SF.net email is sponsored by:Crypto Challenge is now open! 
Get cracking and register here for some mind boggling fun and 
the chance of winning an Apple iPod:
http://ads.sourceforge.net/cgi-bin/redirect.pl?thaw0031en

RE: C2 or C3 on Presario 2110US?

[Grover, Andrew]

> From: Carl Thompson [mailto:cet-GxmFRYwVNOxKOnpN5g9PlQ@public.gmane.org] 
> Is it possible this machine just doesn't do C2-C3, processor limit
> control or processor performance control?  If that is the 
> case, then why
> does Windows seem to be running cooler?  The fan is often off 
> in Windows
> but never off in Linux.  Also battery life seems to be better 
> in Windows
> whereas on my other laptops it's better in Linux.  And yes, I am using
> CPUFreq's PowerNow support to throttle down the CPU speed and 
> voltage to
> the minimum.

Does the system feel slow, or are there an abnormal number of interrupts
shown in /proc/interrupts?

Also, the latest code required that the Processor P_BLK have a length of
0 or 6, without exception.

Finally, C2 and C3 statistics should be exposed through perfmon under
Windows. You might try finding that information, and seeing if indeed it
is using C2 or C3.

> Does anyone know of any tools to display the exact CPU speed, voltage
> and temperature under Windows so I can see exactly what it's 
> doing?  The
> PowerNow dashboard from AMD just gives a (simulated) analog 
> guage which
> does not give accurate measurements.

Wcpuid by H. Oda? I know it shows speed, but I don't have a PowerNow CPU
to test the other part. :)

Regards -- Andy



-------------------------------------------------------
This SF.net email is sponsored by:Crypto Challenge is now open! 
Get cracking and register here for some mind boggling fun and 
the chance of winning an Apple iPod:
http://ads.sourceforge.net/cgi-bin/redirect.pl?thaw0031en

RE: C2 or C3 on Presario 2110US?

[Carl Thompson]

Quoting "Grover, Andrew" <andrew.grover-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>:

> > From: Carl Thompson [mailto:cet-GxmFRYwVNOxKOnpN5g9PlQ@public.gmane.org]
> > Is it possible this machine just doesn't do C2-C3, processor limit
> > control or processor performance control?  If that is the case, then 
> > why does Windows seem to be running cooler?  The fan is often off
> > in Windows but never off in Linux.  Also battery life seems to be 
> > better in Windows whereas on my other laptops it's better in Linux. 
> > And yes, I am using CPUFreq's PowerNow support to throttle down the 
> > CPU speed and voltage to the minimum.
> 
> Does the system feel slow, or are there an abnormal number of
> interrupts shown in /proc/interrupts?

No, the system operates correctly and the interrupts seem reasonable.

> Also, the latest code required that the Processor P_BLK have a length
> of 0 or 6, without exception.

I'm not sure what this is, how I would check it or what I can do about it.

> Finally, C2 and C3 statistics should be exposed through perfmon under
> Windows. You might try finding that information, and seeing if indeed
> it is using C2 or C3.

Thanks for the info.  According to perfmon, Windows goes into C2 when
idle, but never into C3.  So C2 is in fact supported on this hardware.

> ...

This is what I get in /proc/acpi/processor/CPU0:

[carl@Tweak carl]$ cd /proc/acpi/processor/CPU0

[carl@Tweak CPU0]$ ls
info limit performance power throttling

[carl@Tweak CPU0]$ cat info
processor id:            0
acpi id:                 0
bus mastering control:   yes
power management:        no
throttling control:      no
performance management:  no
limit interface:         no

[carl@Tweak CPU0]$ cat limit
<not supported>

[carl@Tweak CPU0]$ cat performance
<not supported>

[carl@Tweak CPU0]$ cat power
active state:            C1
default state:           C1
bus master activity:     00000000
states:
   *C1:                  promotion[--] demotion[--] latency[000]
usage[00000000]
    C2:                  <not supported>
    C3:                  <not supported>

[carl@Tweak CPU0]$ cat throttling
<not supported>


> Regards -- Andy

Thank you,
Carl Thompson




-------------------------------------------------------
This SF.net email is sponsored by:Crypto Challenge is now open! 
Get cracking and register here for some mind boggling fun and 
the chance of winning an Apple iPod:
http://ads.sourceforge.net/cgi-bin/redirect.pl?thaw0031en

RE: C2 or C3 on Presario 2110US?

[Carl Thompson]

Just wondering if anyone can help me figure out how to get C2 working on
my laptop...  Someone suggested that my processor P_BLK might be the
wrong size but I have no idea what that is and searching for that string
in the source yields nothing.

Can anyone help me?

Thank you,
Carl Thompson

Quoting Carl Thompson <cet-GxmFRYwVNOxKOnpN5g9PlQ@public.gmane.org>:

> Quoting "Grover, Andrew" <andrew.grover-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>:
> 
> > > From: Carl Thompson [mailto:cet-GxmFRYwVNOxKOnpN5g9PlQ@public.gmane.org]
> > > Is it possible this machine just doesn't do C2-C3, processor
> limit
> > > control or processor performance control?  If that is the case,
> then
> > > why does Windows seem to be running cooler?  The fan is often off
> > > in Windows but never off in Linux.  Also battery life seems to be
> > > better in Windows whereas on my other laptops it's better in
> Linux.
> > > And yes, I am using CPUFreq's PowerNow support to throttle down
> the
> > > CPU speed and voltage to the minimum.
> >
> > Does the system feel slow, or are there an abnormal number of
> > interrupts shown in /proc/interrupts?
> 
> No, the system operates correctly and the interrupts seem reasonable.
> 
> > Also, the latest code required that the Processor P_BLK have a
> length
> > of 0 or 6, without exception.
> 
> I'm not sure what this is, how I would check it or what I can do
> about it.
> 
> > Finally, C2 and C3 statistics should be exposed through perfmon
> under
> > Windows. You might try finding that information, and seeing if
> indeed
> > it is using C2 or C3.
> 
> Thanks for the info.  According to perfmon, Windows goes into C2 when
> idle, but never into C3.  So C2 is in fact supported on this
> hardware.
> 
> > ...
> 
> This is what I get in /proc/acpi/processor/CPU0:
> 
> [carl@Tweak carl]$ cd /proc/acpi/processor/CPU0
> 
> [carl@Tweak CPU0]$ ls
> info limit performance power throttling
> 
> [carl@Tweak CPU0]$ cat info
> processor id:            0
> acpi id:                 0
> bus mastering control:   yes
> power management:        no
> throttling control:      no
> performance management:  no
> limit interface:         no
> 
> [carl@Tweak CPU0]$ cat limit
> <not supported>
> 
> [carl@Tweak CPU0]$ cat performance
> <not supported>
> 
> [carl@Tweak CPU0]$ cat power
> active state:            C1
> default state:           C1
> bus master activity:     00000000
> states:
>    *C1:                  promotion[--] demotion[--] latency[000]
> usage[00000000]
>     C2:                  <not supported>
>     C3:                  <not supported>
> 
> [carl@Tweak CPU0]$ cat throttling
> <not supported>
> 
> 
> > Regards -- Andy
> 
> Thank you,
> Carl Thompson
> 
> 






-------------------------------------------------------
This SF.net email is sponsored by: Does your code think in ink? 
You could win a Tablet PC. Get a free Tablet PC hat just for playing. 
What are you waiting for?
http://ads.sourceforge.net/cgi-bin/redirect.pl?micr5043en

Re: C2 or C3 on Presario 2110US?

[Ducrot Bruno]

On Thu, Mar 13, 2003 at 05:34:01PM -0800, Grover, Andrew wrote:
> > From: Carl Thompson [mailto:cet-GxmFRYwVNOxKOnpN5g9PlQ@public.gmane.org] 
> > Is it possible this machine just doesn't do C2-C3, processor limit
> > control or processor performance control?  If that is the 
> > case, then why
> > does Windows seem to be running cooler?  The fan is often off 
> > in Windows
> > but never off in Linux.  Also battery life seems to be better 
> > in Windows
> > whereas on my other laptops it's better in Linux.  And yes, I am using
> > CPUFreq's PowerNow support to throttle down the CPU speed and 
> > voltage to
> > the minimum.
> 
> Does the system feel slow, or are there an abnormal number of interrupts
> shown in /proc/interrupts?
> 
> Also, the latest code required that the Processor P_BLK have a length of
> 0 or 6, without exception.
> 
> Finally, C2 and C3 statistics should be exposed through perfmon under
> Windows. You might try finding that information, and seeing if indeed it
> is using C2 or C3.
> 

Sound like (after looking Carl's FACS) that this patch will be OK:

The latency for C2 is at 100 (which I think is allowed).


--- linux-2.5.65/drivers/acpi/processor.c.old	2003-03-19 15:12:16.000000000 +0100
+++ linux-2.5.65/drivers/acpi/processor.c	2003-03-19 15:24:09.000000000 +0100
@@ -639,7 +639,7 @@
 		/*
 		 * C2 latency must be less than or equal to 100 microseconds.
 		 */
-		if (acpi_fadt.plvl2_lat >= ACPI_PROCESSOR_MAX_C2_LATENCY)
+		if (acpi_fadt.plvl2_lat > ACPI_PROCESSOR_MAX_C2_LATENCY)
 			ACPI_DEBUG_PRINT((ACPI_DB_INFO,
 				"C2 latency too large [%d]\n",
 				acpi_fadt.plvl2_lat));
@@ -673,7 +673,7 @@
 		/*
 		 * C3 latency must be less than or equal to 1000 microseconds.
 		 */
-		if (acpi_fadt.plvl3_lat >= ACPI_PROCESSOR_MAX_C3_LATENCY)
+		if (acpi_fadt.plvl3_lat > ACPI_PROCESSOR_MAX_C3_LATENCY)
 			ACPI_DEBUG_PRINT((ACPI_DB_INFO,
 				"C3 latency too large [%d]\n", 
 				acpi_fadt.plvl3_lat));

-- 
Ducrot Bruno

--  Which is worse:  ignorance or apathy?
--  Don't know.  Don't care.


-------------------------------------------------------
This SF.net email is sponsored by: Does your code think in ink? 
You could win a Tablet PC. Get a free Tablet PC hat just for playing. 
What are you waiting for?
http://ads.sourceforge.net/cgi-bin/redirect.pl?micr5043en

Re: C2 or C3 on Presario 2110US?

[Carl Thompson]

Wow!  Thank you very much!  Yes, this works!  I also notice that this
problem also exists in the code for the C3 state.  I still don't have
performance or limit support but I do have PowerNow which I think is a
better way of doing the same thing, right?

Now I will wait and see if my fan goes off...  ;-)

Thank you from a very happy camper,
Carl Thompson

Quoting Ducrot Bruno <ducrot-kk6yZipjEM5g9hUCZPvPmw@public.gmane.org>:

> Hi!
> 
> On Wed, Mar 19, 2003 at 09:31:44AM -0800, Carl Thompson wrote:
> > OK, I have attached my ACPI tables and lspci -xxx output for the
> PMU device.
> >
> > Thank you,
> > Carl Thompson
> >
> > Quoting Ducrot Bruno <ducrot-kk6yZipjEM5g9hUCZPvPmw@public.gmane.org>:
> >
> > > Hi Carl,
> > >
> > > On Tue, Mar 18, 2003 at 09:11:49AM -0800, Carl Thompson wrote:
> > > > Just wondering if anyone can help me figure out how to get C2
> > > > working on my laptop...  Someone suggested that my processor
> P_BLK
> > > > might be the wrong size but I have no idea what that is and
> > > > searching for that string in the source yields nothing.
> > > >
> > > > Can anyone help me?
> > > >
> 
> In fact, it not your processor P_BLK size that is wrong:
> 
>     Scope (_PR)
>     {
>         Processor (CPU0, 0x00, 0x00008010, 0x06) {}
>                                            ^^^^
>                                          6 is correct.
>     }
> 
> but the FACS one:
> 
> ducrot@neptune:~/acpi_hack/carl$ cat FACP |
> ../../acpi/pmtools/acpidmp/acpitbl
> Signature:        FACP
> Length:           116
> Revision:         0x01
> Checksum:         0x97
> OEMID:            ATI
> OEM Table ID:     Raptor
> OEM Revision:     0x06040000
> Creator ID:       ATI
> Creator Revision: 0x000f4240
> FIRMWARE_CTRL:    0x1befffc0
> DSDT:             0x1bef8cf4
> INT_MODEL:        0x00
> SCI_INT:          9
> SMI_CMD:          0x000000b1
> ACPI_ENABLE:      0xf0
> ACPI_DISABLE:     0xf1
> S4BIOS_REQ:       0xf2
> PM1a_EVT_BLK:     0x00008000
> PM1b_EVT_BLK:     0x00000000
> PM1a_CNT_BLK:     0x0000fe10
> PM1b_CNT_BLK:     0x00000000
> PM2_CNT_BLK:      0x00008070
> PM_TMR_BLK:       0x00008008
> GPE0_BLK:         0x00008018
> GPE1_BLK:         0x00000000
> PM1_EVT_LEN:      4
> PM1_CNT_LEN:      2
> PM2_CNT_LEN:      1
> PM_TM_LEN:        4
> GPE0_BLK_LEN:     16
> GPE1_BLK_LEN:     0
> GPE1_BASE:        0
> P_LVL2_LAT:       100
> P_LVL3_LAT:       1001
> FLUSH_SIZE:       0
> FLUSH_STRIDE:     0
> DUTY_OFFSET:      1
> DUTY_WIDTH:       0
> DAY_ALRM:         0x0d
> MON_ALRM:         0x00
> CENTURY:          0x32
> Flags:            0x00000000
> 
> 
> The P_LVL2_LAT is too high: 100.
> It should be < 100.
> 
> I guess that actually windows will do:
> P_LVL2_LAT <= 100 then C2 OK
> 
> whereas ACPI-CA will do:
> P_LVL2_LAT < 100  then C2 OK
> 
> which is wrong (the keyword in spec is
> 'must be less than or equal to').
> 
> 
> Ah...  processor.c is wrong...
> 
> Try this:
> 
> --- linux-2.5.65/drivers/acpi/processor.c.old	2003-03-19
> 15:12:16.000000000 +0100
> +++ linux-2.5.65/drivers/acpi/processor.c	2003-03-19
> 15:13:22.000000000 +0100
> @@ -639,7 +639,7 @@
>  		/*
>  		 * C2 latency must be less than or equal to 100 microseconds.
>  		 */
> -		if (acpi_fadt.plvl2_lat >= ACPI_PROCESSOR_MAX_C2_LATENCY)
> +		if (acpi_fadt.plvl2_lat > ACPI_PROCESSOR_MAX_C2_LATENCY)
>  			ACPI_DEBUG_PRINT((ACPI_DB_INFO,
>  				"C2 latency too large [%d]\n",
>  				acpi_fadt.plvl2_lat));
> 
> 
> 
> Cheers,
> 
> --
> Ducrot Bruno
> 
> --  Which is worse:  ignorance or apathy?
> --  Don't know.  Don't care.
> 
> 






-------------------------------------------------------
This SF.net email is sponsored by: Does your code think in ink? 
You could win a Tablet PC. Get a free Tablet PC hat just for playing. 
What are you waiting for?
http://ads.sourceforge.net/cgi-bin/redirect.pl?micr5043en