Re: S3 suspend/resume driver questions

Re: S3 suspend/resume driver questions

[Ilya Bagrak]

I can access all these functions through the "Fn" key on my keyboard, so I
have not experienced a real need for this patch.  Is the keyboard layout
different or are the special keys are not working on other models?  I
guess with the patch one could write a little GUI applet that can expose
these things to the user, but I find Fn keys just as easy.  Anyways I'll
let you know if the patch works for me. Thanks,

Ilya


> Ilya,
>
>> Your patch worked beautifully on my Fujitsu E7010 laptop.  Prior to
>> the patch, the system would lock up hard if I tried to load
>> snd_intel8x0 after waking up from S3.  If I didn't not unload the
>> sound modules before going to sleep, the system would not wake up at
>> all (lock up would happen on resume).  In any case, the sound "just
>> works" (tm) now.
>
> Nice to know.  From what I've seen, it looks like Fujitsu is pretty good
> with keeping things consistent between their different lines of
> notebooks.  They usually reuse components (the prism2 wifi card in your
> notebook is probably the same card as the one that's in my notebook
> [which is also used in the P2xxx series too]).  I'm guessing this patch
> I've modified would work on your notebook too (read the header of the
> patch):
>
> http://devrandom.com/fujitsu/fujitsu_acpi-2.6.patch
>
> It adds LCD brightness, volume and pointer status functionality and the
> ability to change them too.  Let me know if it works for you.
>
> Adrian
>
>
> -------------------------------------------------------
> This SF.Net email is sponsored by: IBM Linux Tutorials
> Free Linux tutorial presented by Daniel Robbins, President and CEO of
> GenToo technologies. Learn everything from fundamentals to system
> administration.http://ads.osdn.com/?ad_id\x1470&alloc_id638&opÌk
> _______________________________________________
> Acpi-devel mailing list
> Acpi-devel-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org
> https://lists.sourceforge.net/lists/listinfo/acpi-devel





-------------------------------------------------------
This SF.Net email is sponsored by: IBM Linux Tutorials
Free Linux tutorial presented by Daniel Robbins, President and CEO of
GenToo technologies. Learn everything from fundamentals to system
administration.http://ads.osdn.com/?ad_id\x1470&alloc_id638&op=click

Re: S3 suspend/resume driver questions

[Tomasz Ciolek]

The keyboard is working fine.

The symptoms of bad S3 state on the P2120 is that after a resume the
laptop does NOT redo the Radeon M6 post correctly (at all) as a
consequence the screen remils blank (off) and no video/console... 

I will try this patch and see what I can get 

TMC

PS the Fn Key work fine...

On Mon, Apr 19, 2004 at 10:47:25PM -0400, Ilya Bagrak wrote:
> I can access all these functions through the "Fn" key on my keyboard, so I
> have not experienced a real need for this patch.  Is the keyboard layout
> different or are the special keys are not working on other models?  I
> guess with the patch one could write a little GUI applet that can expose
> these things to the user, but I find Fn keys just as easy.  Anyways I'll
> let you know if the patch works for me. Thanks,
> 
> Ilya
> 
> 
> > Ilya,
> >
> >> Your patch worked beautifully on my Fujitsu E7010 laptop.  Prior to
> >> the patch, the system would lock up hard if I tried to load
> >> snd_intel8x0 after waking up from S3.  If I didn't not unload the
> >> sound modules before going to sleep, the system would not wake up at
> >> all (lock up would happen on resume).  In any case, the sound "just
> >> works" (tm) now.
> >
> > Nice to know.  From what I've seen, it looks like Fujitsu is pretty good
> > with keeping things consistent between their different lines of
> > notebooks.  They usually reuse components (the prism2 wifi card in your
> > notebook is probably the same card as the one that's in my notebook
> > [which is also used in the P2xxx series too]).  I'm guessing this patch
> > I've modified would work on your notebook too (read the header of the
> > patch):
> >
> > http://devrandom.com/fujitsu/fujitsu_acpi-2.6.patch
> >
> > It adds LCD brightness, volume and pointer status functionality and the
> > ability to change them too.  Let me know if it works for you.
> >
> > Adrian
> >
> >
> > -------------------------------------------------------
> > This SF.Net email is sponsored by: IBM Linux Tutorials
> > Free Linux tutorial presented by Daniel Robbins, President and CEO of
> > GenToo technologies. Learn everything from fundamentals to system
> > administration.http://ads.osdn.com/?ad_id\x1470&alloc_id638&op?k
> > _______________________________________________
> > Acpi-devel mailing list
> > Acpi-devel-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org
> > https://lists.sourceforge.net/lists/listinfo/acpi-devel
> 
> 
> 
> 
> 
> -------------------------------------------------------
> This SF.Net email is sponsored by: IBM Linux Tutorials
> Free Linux tutorial presented by Daniel Robbins, President and CEO of
> GenToo technologies. Learn everything from fundamentals to system
> administration.http://ads.osdn.com/?ad_id\x1470&alloc_id638&op=click
> _______________________________________________
> Acpi-devel mailing list
> Acpi-devel-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org
> https://lists.sourceforge.net/lists/listinfo/acpi-devel

-- 
Tomasz M. Ciolek	
*******************************************************************************
 tmc at dreamcraft dot com dot au 
*******************************************************************************
   GPG Key ID:		0x41C4C2F0
   GPG Key Fingerprint: 3883 B308 8256 2246 D3ED  A1FF 3A1D 0EAD 41C4 C2F0
   Key available on www.pgp.net	
*******************************************************************************


-------------------------------------------------------
This SF.Net email is sponsored by: IBM Linux Tutorials
Free Linux tutorial presented by Daniel Robbins, President and CEO of
GenToo technologies. Learn everything from fundamentals to system
administration.http://ads.osdn.com/?ad_id\x1470&alloc_id638&op=click

Re: S3 suspend/resume driver questions

[Adrian Yee]

Hi,

> The symptoms of bad S3 state on the P2120 is that after a resume the
> laptop does NOT redo the Radeon M6 post correctly (at all) as a
> consequence the screen remils blank (off) and no video/console... 

Have you tried the things mentioned in Documentation/power/video.txt?  I had to do acpi_sleep=s3_bios to get video working on my system.

Adrian


-------------------------------------------------------
This SF.Net email is sponsored by: IBM Linux Tutorials
Free Linux tutorial presented by Daniel Robbins, President and CEO of
GenToo technologies. Learn everything from fundamentals to system
administration.http://ads.osdn.com/?ad_id\x1470&alloc_id638&op=click

Re: S3 suspend/resume driver questions

[Tomasz Ciolek]

On Mon, Apr 19, 2004 at 08:15:34PM -0700, Adrian Yee wrote:
> Hi,
> 
> > The symptoms of bad S3 state on the P2120 is that after a resume the
> > laptop does NOT redo the Radeon M6 post correctly (at all) as a
> > consequence the screen remils blank (off) and no video/console... 
> 
> Have you tried the things mentioned in Documentation/power/video.txt?  I had to do acpi_sleep=s3_bios to get video working on my system.
> 
> Adrian

yes. I have tried that. Im assuming thats a kernel boot option?

TMC

-- 
Tomasz M. Ciolek	
*******************************************************************************
 tmc at dreamcraft dot com dot au 
*******************************************************************************
   GPG Key ID:		0x41C4C2F0
   GPG Key Fingerprint: 3883 B308 8256 2246 D3ED  A1FF 3A1D 0EAD 41C4 C2F0
   Key available on www.pgp.net	
*******************************************************************************


-------------------------------------------------------
This SF.Net email is sponsored by: IBM Linux Tutorials
Free Linux tutorial presented by Daniel Robbins, President and CEO of
GenToo technologies. Learn everything from fundamentals to system
administration.http://ads.osdn.com/?ad_id=1470&alloc_id=3638&op=click

Re: S3 suspend/resume driver questions

[Adrian Yee]

> I can access all these functions through the "Fn" key on my keyboard, so I
> have not experienced a real need for this patch.  Is the keyboard layout
> different or are the special keys are not working on other models?  I
> guess with the patch one could write a little GUI applet that can expose
> these things to the user, but I find Fn keys just as easy.  Anyways I'll
> let you know if the patch works for me. Thanks,

They work fine, just that this gives you the ability to do things like displaying an OSD on change and the ability to change settings via software lets you do things like dimming the LCD when idle.

Adrian


-------------------------------------------------------
This SF.Net email is sponsored by: IBM Linux Tutorials
Free Linux tutorial presented by Daniel Robbins, President and CEO of
GenToo technologies. Learn everything from fundamentals to system
administration.http://ads.osdn.com/?ad_id\x1470&alloc_id638&op=click

Re: S3 suspend/resume driver questions

[Tomasz Ciolek]

I have done some tests - will full module suite and without. No go.
Screen remains blan and laptop does NOT repsond to keyboard. (BTW
suspend to disk works just fine). I think it is the graphics driver that
just hoses the whole shebang... 

Anyways I will try a fb and non fb kernel next and see.

TMC

On Mon, Apr 19, 2004 at 08:09:36PM -0700, Adrian Yee wrote:
> > I can access all these functions through the "Fn" key on my keyboard, so I
> > have not experienced a real need for this patch.  Is the keyboard layout
> > different or are the special keys are not working on other models?  I
> > guess with the patch one could write a little GUI applet that can expose
> > these things to the user, but I find Fn keys just as easy.  Anyways I'll
> > let you know if the patch works for me. Thanks,
> 
> They work fine, just that this gives you the ability to do things like displaying an OSD on change and the ability to change settings via software lets you do things like dimming the LCD when idle.
> 
> Adrian
> 
> 
> -------------------------------------------------------
> This SF.Net email is sponsored by: IBM Linux Tutorials
> Free Linux tutorial presented by Daniel Robbins, President and CEO of
> GenToo technologies. Learn everything from fundamentals to system
> administration.http://ads.osdn.com/?ad_id\x1470&alloc_id638&op=click
> _______________________________________________
> Acpi-devel mailing list
> Acpi-devel-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org
> https://lists.sourceforge.net/lists/listinfo/acpi-devel

-- 
Tomasz M. Ciolek	
*******************************************************************************
 tmc at dreamcraft dot com dot au 
*******************************************************************************
   GPG Key ID:		0x41C4C2F0
   GPG Key Fingerprint: 3883 B308 8256 2246 D3ED  A1FF 3A1D 0EAD 41C4 C2F0
   Key available on www.pgp.net	
*******************************************************************************


-------------------------------------------------------
This SF.Net email is sponsored by: IBM Linux Tutorials
Free Linux tutorial presented by Daniel Robbins, President and CEO of
GenToo technologies. Learn everything from fundamentals to system
administration.http://ads.osdn.com/?ad_id\x1470&alloc_id638&op=click

Re: S3 suspend/resume driver questions

[Ilya Bagrak]

Adrian,

Your patch worked beautifully on my Fujitsu E7010 laptop.  Prior to the
patch, the system would lock up hard if I tried to load snd_intel8x0 after
waking up from S3.  If I didn't not unload the sound modules before going
to sleep, the system would not wake up at all (lock up would happen on
resume).  In any case, the sound "just works" (tm) now.

There is more good news.  I adapted your patch to orinoco_pci driver which
manages my built in Prism2 wireless adapter.  The unpatched driver failed
to reinit the card upon resume.  Even though that didn't not lock up my
computer like sound did, the card was not functional.  I was absolutely
stunned when this 3-liner-of-a-patch returned full functionality to the
card upon resume.  You don't even have to unload the driver before going
to sleep.  I have a little signal strength applet running in the toolbar,
and you can watch the signal strength gradually increase after wakeup and
the card is back to normal in a couple of seconds.

The wireless NIC was the last stumbling block in getting S3 to work.  I
now officially declare Fujitsu E7010 Linux ACPI-friendly.  All major
devices (have not tested Firewire or IrDA) recover and work properly after
S3. Thank you all.

Ilya

> Hi,
>
>> mine isn't a technical reply, but you can post in the 'alsa-devel'
>> mailing-list, as it seems that Takashi Iwai is working on PM-lacking
>> drivers. Here some links:
>>    http://sourceforge.net/mailarchive/forum.php?thread_id=4201761&forum_id=1752
>
> Doing a grep through the alsa code, it doesn't look like any driver
> save/restores pci state.
>
>>    http://sourceforge.net/mailarchive/forum.php?thread_id=4107033&forum_id=1752
>>
>> BTW-1, could you post your 'intel8x0' patch? I could give it a try
>> ASAP...
>
> That's essentially what I did.  It is a little different though:
>
> --- /usr/src/linux-2.6.5-mm6/sound/pci/intel8x0.c       2004-04-05
> 18:46:17.000000000 -0700 +++ intel8x0.c  2004-04-18 04:03:34.000000000
> -0700
> @@ -445,6 +445,7 @@ struct _snd_intel8x0 {
>
>  #ifdef CONFIG_PM
>         int in_suspend;
> +       u32 pci_state[16];
>  #endif
>  };
>
> @@ -2223,11 +2224,13 @@ static int snd_intel8x0_suspend(struct p
>  {
>         intel8x0_t *chip = snd_magic_cast(intel8x0_t,
> pci_get_drvdata(dev), return -ENXIO); intel8x0_suspend(chip);
> +       pci_save_state(dev, chip->pci_state);
>         return 0;
>  }
>  static int snd_intel8x0_resume(struct pci_dev *dev)
>  {
>         intel8x0_t *chip = snd_magic_cast(intel8x0_t,
> pci_get_drvdata(dev), return -ENXIO);
> +       pci_restore_state(dev, chip->pci_state);
>         intel8x0_resume(chip);
>         return 0;
>  }
>
>> BTW-2, in the few tests I made on my ASUS M6842NWH ('intel8x0' based),
>> I have to remove the sound modules even to have a successfull S1.
>
> Actually, I haven't tried out S1 yet, but I don't have to remove the
> sound modules for S3 to work.  And even if I do, my system will lock up
> when trying to load the alsa drivers due to the device's configuration
> state.
>
> Adrian
>
>
> -------------------------------------------------------
> This SF.Net email is sponsored by: IBM Linux Tutorials
> Free Linux tutorial presented by Daniel Robbins, President and CEO of
> GenToo technologies. Learn everything from fundamentals to system
> administration.http://ads.osdn.com/?ad_id\x1470&alloc_id638&opÌk
> _______________________________________________
> Acpi-devel mailing list
> Acpi-devel-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org
> https://lists.sourceforge.net/lists/listinfo/acpi-devel





-------------------------------------------------------
This SF.Net email is sponsored by: IBM Linux Tutorials
Free Linux tutorial presented by Daniel Robbins, President and CEO of
GenToo technologies. Learn everything from fundamentals to system
administration.http://ads.osdn.com/?ad_id\x1470&alloc_id638&op=click

Re: S3 suspend/resume driver questions

[Tomasz Ciolek]

[-- Attachment #1: Type: text/plain, Size: 5062 bytes --]

 Hooray.
 
 If only someone at ATI or Fujitsu finnaly mamanged to cough up
 the magic incantations for the radeon cards so that the POST
 correctly... 

 Regards
 TMC

On Mon, Apr 19, 2004 at 09:42:12PM -0400, Ilya Bagrak wrote:
> Adrian,
> 
> Your patch worked beautifully on my Fujitsu E7010 laptop.  Prior to the
> patch, the system would lock up hard if I tried to load snd_intel8x0 after
> waking up from S3.  If I didn't not unload the sound modules before going
> to sleep, the system would not wake up at all (lock up would happen on
> resume).  In any case, the sound "just works" (tm) now.
> 
> There is more good news.  I adapted your patch to orinoco_pci driver which
> manages my built in Prism2 wireless adapter.  The unpatched driver failed
> to reinit the card upon resume.  Even though that didn't not lock up my
> computer like sound did, the card was not functional.  I was absolutely
> stunned when this 3-liner-of-a-patch returned full functionality to the
> card upon resume.  You don't even have to unload the driver before going
> to sleep.  I have a little signal strength applet running in the toolbar,
> and you can watch the signal strength gradually increase after wakeup and
> the card is back to normal in a couple of seconds.
> 
> The wireless NIC was the last stumbling block in getting S3 to work.  I
> now officially declare Fujitsu E7010 Linux ACPI-friendly.  All major
> devices (have not tested Firewire or IrDA) recover and work properly after
> S3. Thank you all.
> 
> Ilya
> 
> > Hi,
> >
> >> mine isn't a technical reply, but you can post in the 'alsa-devel'
> >> mailing-list, as it seems that Takashi Iwai is working on PM-lacking
> >> drivers. Here some links:
> >>    http://sourceforge.net/mailarchive/forum.php?thread_id=4201761&forum_id=1752
> >
> > Doing a grep through the alsa code, it doesn't look like any driver
> > save/restores pci state.
> >
> >>    http://sourceforge.net/mailarchive/forum.php?thread_id=4107033&forum_id=1752
> >>
> >> BTW-1, could you post your 'intel8x0' patch? I could give it a try
> >> ASAP...
> >
> > That's essentially what I did.  It is a little different though:
> >
> > --- /usr/src/linux-2.6.5-mm6/sound/pci/intel8x0.c       2004-04-05
> > 18:46:17.000000000 -0700 +++ intel8x0.c  2004-04-18 04:03:34.000000000
> > -0700
> > @@ -445,6 +445,7 @@ struct _snd_intel8x0 {
> >
> >  #ifdef CONFIG_PM
> >         int in_suspend;
> > +       u32 pci_state[16];
> >  #endif
> >  };
> >
> > @@ -2223,11 +2224,13 @@ static int snd_intel8x0_suspend(struct p
> >  {
> >         intel8x0_t *chip = snd_magic_cast(intel8x0_t,
> > pci_get_drvdata(dev), return -ENXIO); intel8x0_suspend(chip);
> > +       pci_save_state(dev, chip->pci_state);
> >         return 0;
> >  }
> >  static int snd_intel8x0_resume(struct pci_dev *dev)
> >  {
> >         intel8x0_t *chip = snd_magic_cast(intel8x0_t,
> > pci_get_drvdata(dev), return -ENXIO);
> > +       pci_restore_state(dev, chip->pci_state);
> >         intel8x0_resume(chip);
> >         return 0;
> >  }
> >
> >> BTW-2, in the few tests I made on my ASUS M6842NWH ('intel8x0' based),
> >> I have to remove the sound modules even to have a successfull S1.
> >
> > Actually, I haven't tried out S1 yet, but I don't have to remove the
> > sound modules for S3 to work.  And even if I do, my system will lock up
> > when trying to load the alsa drivers due to the device's configuration
> > state.
> >
> > Adrian
> >
> >
> > -------------------------------------------------------
> > This SF.Net email is sponsored by: IBM Linux Tutorials
> > Free Linux tutorial presented by Daniel Robbins, President and CEO of
> > GenToo technologies. Learn everything from fundamentals to system
> > administration.http://ads.osdn.com/?ad_id\x1470&alloc_id638&op?k
> > _______________________________________________
> > Acpi-devel mailing list
> > Acpi-devel-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org
> > https://lists.sourceforge.net/lists/listinfo/acpi-devel
> 
> 
> 
> 
> 
> -------------------------------------------------------
> This SF.Net email is sponsored by: IBM Linux Tutorials
> Free Linux tutorial presented by Daniel Robbins, President and CEO of
> GenToo technologies. Learn everything from fundamentals to system
> administration.http://ads.osdn.com/?ad_id\x1470&alloc_id638&op=click
> _______________________________________________
> Acpi-devel mailing list
> Acpi-devel-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org
> https://lists.sourceforge.net/lists/listinfo/acpi-devel

-- 
Tomasz M. Ciolek	
*******************************************************************************
 tmc at dreamcraft dot com dot au 
*******************************************************************************
   GPG Key ID:		0x41C4C2F0
   GPG Key Fingerprint: 3883 B308 8256 2246 D3ED  A1FF 3A1D 0EAD 41C4 C2F0
   Key available on www.pgp.net	
*******************************************************************************

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

Re: S3 suspend/resume driver questions

[ole.rohne-vJEk5272eHo]

>>>>> "Tomasz" == Tomasz Ciolek <tmc-+XVqjN4+fIJfbCVcMT5z6Ydd74u8MsAO@public.gmane.org>

Tomasz>  If only someone at ATI or Fujitsu finnaly mamanged to cough up
Tomasz>  the magic incantations for the radeon cards so that the POST
Tomasz>  correctly... 

AFIK, POST on S3 resume is not mandated by the ACPI specs, so I'll
assume we'll have to do it ourselves:

1. The "lcall $0xc000,$3"-kludge in wakeup.S doesn't even know the PCI
   slot/function number, and it is probably way too early as bridges
   etc are not awake yet.

2. POST-on-VT change in X11 can be made to work but is rather dirty
   and it doesn't fix the text consoles.

3. POST using vm86 from radeon_pci_resume is probably The Right Thing
   (tm): It knows the PCI slot/function and it happens at the right
   time wrt bridges and stuff. This requires someone with the
   knowledge to write the lines of code and the energy to make it
   work. (Having a black screen as the only diagnostics isn't very
   inspiring in the long run.)

Ole


-------------------------------------------------------
This SF.Net email is sponsored by: IBM Linux Tutorials
Free Linux tutorial presented by Daniel Robbins, President and CEO of
GenToo technologies. Learn everything from fundamentals to system
administration.http://ads.osdn.com/?ad_id=1470&alloc_id=3638&op=click

Re: S3 suspend/resume driver questions

[Adrian Yee]

Ilya,

> Your patch worked beautifully on my Fujitsu E7010 laptop.  Prior to the
> patch, the system would lock up hard if I tried to load snd_intel8x0 after
> waking up from S3.  If I didn't not unload the sound modules before going
> to sleep, the system would not wake up at all (lock up would happen on
> resume).  In any case, the sound "just works" (tm) now.

Nice to know.  From what I've seen, it looks like Fujitsu is pretty good with keeping things consistent between their different lines of notebooks.  They usually reuse components (the prism2 wifi card in your notebook is probably the same card as the one that's in my notebook [which is also used in the P2xxx series too]).  I'm guessing this patch I've modified would work on your notebook too (read the header of the patch):

http://devrandom.com/fujitsu/fujitsu_acpi-2.6.patch

It adds LCD brightness, volume and pointer status functionality and the ability to change them too.  Let me know if it works for you.

Adrian


-------------------------------------------------------
This SF.Net email is sponsored by: IBM Linux Tutorials
Free Linux tutorial presented by Daniel Robbins, President and CEO of
GenToo technologies. Learn everything from fundamentals to system
administration.http://ads.osdn.com/?ad_id\x1470&alloc_id638&op=click

Re: S3 suspend/resume driver questions

[Pavel Machek]

Hi!

> There is more good news.  I adapted your patch to orinoco_pci driver which
> manages my built in Prism2 wireless adapter.  The unpatched driver failed
> to reinit the card upon resume.  Even though that didn't not lock up my
> computer like sound did, the card was not functional.  I was absolutely
> stunned when this 3-liner-of-a-patch returned full functionality to the
> card upon resume.  You don't even have to unload the driver before going

Be sure to submit the patch to prism2 maintainer... More
people probably need that one.
-- 
64 bytes from 195.113.31.123: icmp_seq=28 ttl=51 time=448769.1 ms         



-------------------------------------------------------
This SF.net email is sponsored by: The Robotic Monkeys at ThinkGeek
For a limited time only, get FREE Ground shipping on all orders of $35
or more. Hurry up and shop folks, this offer expires April 30th!
http://www.thinkgeek.com/freeshipping/?cpg=12297

S3 suspend/resume driver questions

[Adrian Yee]

Hi,

Well after hacking around with the 2.6 kernels for some time now trying to get ACPI S3 suspend/resume working, I've finally gotten it to work on my notebook (Fujitsu S6010 - PIII 1GHz, i830 chipset/graphics).  I had to use the acpi_sleep=s3_bios kernel argument to get video restored on resume, but I had other problems - other devices were causing lockups on resume.  This turned out to be caused by the pci states of devices not being restored on resume (is this normal?).  To fix this, I had to make some changes to the various drivers: 8139too and the alsa intel8x0 driver.

For the 8139too driver, I had to move the pci save/restore power/state code.  It currently is located after:

if (!netif_running (dev))

And because I didn't have the device running (the module is loaded, but the device isn't configured), it wasn't restoring the device properly.  Is there a reason why it doesn't pci_save_state() if the device isn't running?  The problem with this is that I often switch between wired and wireless network, so the device is often not up.  I looked at the 3c59x driver and it also doesn't perform any suspend/resume code on the device if it isn't running.  Is this proper behaviour?

For the alsa intel8x0 driver, I just added code to save/restore the pci state on suspend and resume.  Is this the correct way to fix the problem?

One thing I couldn't fix was the i830 video state:

Before suspend:
00:02.1 Display controller: Intel Corp. 82830 CGC [Chipset Graphics Controller]
	Subsystem: Fujitsu Limited.: Unknown device 113c
	Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B-
	Status: Cap+ 66Mhz- UDF- FastB2B+ ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR-
	Latency: 0 (2750ns max)
	Region 0: Memory at f0000000 (32-bit, prefetchable) [size=128M]
	Region 1: Memory at e0100000 (32-bit, non-prefetchable) [size=512K]
	Capabilities: [d0] Power Management version 1
		Flags: PMEClk- DSI+ D1+ D2- AuxCurrent=0mA PME(D0-,D1-,D2-,D3hot-,D3cold-)
		Status: D0 PME-Enable- DSel=0 DScale=0 PME-
00: 86 80 77 35 07 00 90 00 00 00 80 03 00 00 80 00
10: 08 00 00 f0 00 00 10 e0 00 00 00 00 00 00 00 00
20: 00 00 00 00 00 00 00 00 00 00 00 00 cf 10 3c 11
30: 00 00 00 00 d0 00 00 00 00 00 00 00 00 00 00 0b

After resume:
00:02.1 Display controller: Intel Corp. 82830 CGC [Chipset Graphics Controller]
	Subsystem: Fujitsu Limited.: Unknown device 113c
	Control: I/O- Mem- BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B-
	Status: Cap+ 66Mhz- UDF- FastB2B+ ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR-
	Latency: 0 (2750ns max)
	Region 0: Memory at f0000000 (32-bit, prefetchable) [disabled] [size=128M]
	Region 1: Memory at e0100000 (32-bit, non-prefetchable) [disabled] [size=512K]
	Capabilities: [d0] Power Management version 1
		Flags: PMEClk- DSI+ D1+ D2- AuxCurrent=0mA PME(D0-,D1-,D2-,D3hot-,D3cold-)
		Status: D0 PME-Enable- DSel=0 DScale=0 PME-
00: 86 80 77 35 04 00 90 00 00 00 80 03 00 00 80 00
10: 08 00 00 f0 00 00 10 e0 00 00 00 00 00 00 00 00
20: 00 00 00 00 00 00 00 00 00 00 00 00 cf 10 3c 11
30: 00 00 00 00 d0 00 00 00 00 00 00 00 00 00 00 0b

I tried modifing the intel-agp module to save/restore the state, but it didn't seem to have any effect on it.  Any clues?

Thanks in advance for any help.

Adrian


-------------------------------------------------------
This SF.Net email is sponsored by: IBM Linux Tutorials
Free Linux tutorial presented by Daniel Robbins, President and CEO of
GenToo technologies. Learn everything from fundamentals to system
administration.http://ads.osdn.com/?ad_id\x1470&alloc_id638&op=click

Re: S3 suspend/resume driver questions

[Luca Capello]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Hello,

on 04/18/04 12:06, Adrian Yee wrote:
> For the alsa intel8x0 driver, I just added code to save/restore the pci
> state on suspend and resume.  Is this the correct way to fix the problem?
mine isn't a technical reply, but you can post in the 'alsa-devel' mailing-list,
as it seems that Takashi Iwai is working on PM-lacking drivers. Here some links:
   http://sourceforge.net/mailarchive/forum.php?thread_id=4201761&forum_id=1752
   http://sourceforge.net/mailarchive/forum.php?thread_id=4107033&forum_id=1752

BTW-1, could you post your 'intel8x0' patch? I could give it a try ASAP...

BTW-2, in the few tests I made on my ASUS M6842NWH ('intel8x0' based), I have to
remove the sound modules even to have a successfull S1.

Thx, bye,
Gismo / Luca
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.4 (GNU/Linux)
Comment: Using GnuPG with Debian - http://enigmail.mozdev.org

iD8DBQFAgl2FVAp7Xm10JmkRAtYxAJ4+uklo8A7tAxG9b1pgTiTpN/iNzwCfSEa8
cAGeBAq40j20Uyvdv8YMTXc=
=zQPa
-----END PGP SIGNATURE-----


-------------------------------------------------------
This SF.Net email is sponsored by: IBM Linux Tutorials
Free Linux tutorial presented by Daniel Robbins, President and CEO of
GenToo technologies. Learn everything from fundamentals to system
administration.http://ads.osdn.com/?ad_id=1470&alloc_id=3638&op=click

Re: S3 suspend/resume driver questions

[Adrian Yee]

Hi,

> mine isn't a technical reply, but you can post in the 'alsa-devel' mailing-list,
> as it seems that Takashi Iwai is working on PM-lacking drivers. Here some links:
>    http://sourceforge.net/mailarchive/forum.php?thread_id=4201761&forum_id=1752

Doing a grep through the alsa code, it doesn't look like any driver save/restores pci state.

>    http://sourceforge.net/mailarchive/forum.php?thread_id=4107033&forum_id=1752
> 
> BTW-1, could you post your 'intel8x0' patch? I could give it a try ASAP...

That's essentially what I did.  It is a little different though:

--- /usr/src/linux-2.6.5-mm6/sound/pci/intel8x0.c       2004-04-05 18:46:17.000000000 -0700
+++ intel8x0.c  2004-04-18 04:03:34.000000000 -0700
@@ -445,6 +445,7 @@ struct _snd_intel8x0 {
        
 #ifdef CONFIG_PM
        int in_suspend;
+       u32 pci_state[16];
 #endif
 };
 
@@ -2223,11 +2224,13 @@ static int snd_intel8x0_suspend(struct p
 {
        intel8x0_t *chip = snd_magic_cast(intel8x0_t, pci_get_drvdata(dev), return -ENXIO);
        intel8x0_suspend(chip);
+       pci_save_state(dev, chip->pci_state);
        return 0;
 }
 static int snd_intel8x0_resume(struct pci_dev *dev)
 {
        intel8x0_t *chip = snd_magic_cast(intel8x0_t, pci_get_drvdata(dev), return -ENXIO);
+       pci_restore_state(dev, chip->pci_state);
        intel8x0_resume(chip);
        return 0;
 }

> BTW-2, in the few tests I made on my ASUS M6842NWH ('intel8x0' based), I have to
> remove the sound modules even to have a successfull S1.

Actually, I haven't tried out S1 yet, but I don't have to remove the sound modules for S3 to work.  And even if I do, my system will lock up when trying to load the alsa drivers due to the device's configuration state.

Adrian


-------------------------------------------------------
This SF.Net email is sponsored by: IBM Linux Tutorials
Free Linux tutorial presented by Daniel Robbins, President and CEO of
GenToo technologies. Learn everything from fundamentals to system
administration.http://ads.osdn.com/?ad_id\x1470&alloc_id638&op=click