Re: 11.0-RC2 suspend/resume on thinkpad x201 kills poweroff

[John Baldwin]

On Monday, September 12, 2016 10:35:35 AM Philip Homburg wrote:
> In your letter dated Fri, 9 Sep 2016 07:51:35 -0700 you wrote:
> >is it too new to "kldload i915kms" ?
> 
> Yes, i915kms does the trick.
> 
> In X, suspend/resume works. On a console it requires a console switch
> go get screen output. It would be nice if the driver could do this
> automaticaly.
> 
> I guess now the question is where to document this. The handbook seems
> to be silent about this issue.

The console resume thing is a bug on all systems in vt(4).  With sc(4)
resume on console brings the screen back.  It seems that vt(4) needs
some sort of "repaint" call on resume.

-- 
John Baldwin
___
freebsd-current@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-current
To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"

Re: 11.0-RC2 suspend/resume on thinkpad x201 kills poweroff

[Philip Homburg]

In your letter dated Fri, 9 Sep 2016 07:51:35 -0700 you wrote:
>is it too new to "kldload i915kms" ?

Yes, i915kms does the trick.

In X, suspend/resume works. On a console it requires a console switch
go get screen output. It would be nice if the driver could do this
automaticaly.

I guess now the question is where to document this. The handbook seems
to be silent about this issue.

___
freebsd-current@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-current
To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"

Re: 11.0-RC2 suspend/resume on thinkpad x201 kills poweroff

[Philibert-Caillat Fabrice]

Thank you for the answer, I'll try this then.

Fabrice Philibert-Caillat
*http://auroraelibri.com *
+33.972295359
+33.622507341

2016-09-10 8:53 GMT+02:00 Philip Homburg :

> In your letter dated Fri, 9 Sep 2016 17:39:42 +0200 you wrote:
> >Hello, I am sorry if I did not introduce myself properly: I am a
> subscriber
> >on this list, I have read this thread and I think that my own problem is
> >related: I also have a Thinkpad X201 and I have been unable to
> >suspend/resume with the i915kms driver loaded (with 11.0-RC2 currently). I
> >now suspect (since I did a lot of tests previously) that I might have also
> >encountered the same corruption, as my Thinkpad is now unable to shut down
> >properly. May I ask if it is possible to reset properly without opening
> the
> >computer? In any case, if I can perform any further test to help here, I
> >would be happy.
> >
> >Regarding suspend/resume (which, I understand, is another problem), the
> >most I could achieve so far is having the suspend led blink a few times,
> >then stop blinking. The computer seems to enter suspend at first, but it
> is
> >unable to resume. If I understood correctly what I have read so far, the
> >fixed suspend led would instead mean that suspend did not work properly in
> >fact. I have tried with these options:
>
> When the CMOS is corrupt I see the same problem with suspending.
>
> In the past it was possible to write to the CMOS using the CMOS
> chip. But it is a very long time since I last played with that.
>
> Opening the x201 is annoying, but not very complex.
>
___
freebsd-current@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-current
To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"

Re: 11.0-RC2 suspend/resume on thinkpad x201 kills poweroff

[Philip Homburg]

In your letter dated Fri, 9 Sep 2016 17:39:42 +0200 you wrote:
>Hello, I am sorry if I did not introduce myself properly: I am a subscriber
>on this list, I have read this thread and I think that my own problem is
>related: I also have a Thinkpad X201 and I have been unable to
>suspend/resume with the i915kms driver loaded (with 11.0-RC2 currently). I
>now suspect (since I did a lot of tests previously) that I might have also
>encountered the same corruption, as my Thinkpad is now unable to shut down
>properly. May I ask if it is possible to reset properly without opening the
>computer? In any case, if I can perform any further test to help here, I
>would be happy.
>
>Regarding suspend/resume (which, I understand, is another problem), the
>most I could achieve so far is having the suspend led blink a few times,
>then stop blinking. The computer seems to enter suspend at first, but it is
>unable to resume. If I understood correctly what I have read so far, the
>fixed suspend led would instead mean that suspend did not work properly in
>fact. I have tried with these options:

When the CMOS is corrupt I see the same problem with suspending.

In the past it was possible to write to the CMOS using the CMOS
chip. But it is a very long time since I last played with that.

Opening the x201 is annoying, but not very complex.
___
freebsd-current@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-current
To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"

Re: 11.0-RC2 suspend/resume on thinkpad x201 kills poweroff

[Philibert-Caillat Fabrice]

Hello, I am sorry if I did not introduce myself properly: I am a subscriber
on this list, I have read this thread and I think that my own problem is
related: I also have a Thinkpad X201 and I have been unable to
suspend/resume with the i915kms driver loaded (with 11.0-RC2 currently). I
now suspect (since I did a lot of tests previously) that I might have also
encountered the same corruption, as my Thinkpad is now unable to shut down
properly. May I ask if it is possible to reset properly without opening the
computer? In any case, if I can perform any further test to help here, I
would be happy.

Regarding suspend/resume (which, I understand, is another problem), the
most I could achieve so far is having the suspend led blink a few times,
then stop blinking. The computer seems to enter suspend at first, but it is
unable to resume. If I understood correctly what I have read so far, the
fixed suspend led would instead mean that suspend did not work properly in
fact. I have tried with these options:

hw.vga.textmode=0 / 1
hw.acpi.reset_video=0 / 1
dev.acpi_ibm.0.events=0 / 1
hw.pci.do_power_nodriver=3 (or with no such sysctl)

Furthermore, currently I have these sysctls :
hw.acpi.sleep_button_state=S3
hw.acpi.power_button_state=S5
hw.acpi.sleep_delay=3
hw.acpi.verbose=1
hw.usb.no_suspend_wait=1
hw.syscons.sc_no_suspend_vtswitch=0

In /boot/loader.conf, I tried with/without:

i915kms_load="YES"
drm.i915.enable_rc6=7

Since my computer doesn't shut down, I have tried booting a Ubuntu live usb
key and suspend/resume brought me the same results as with FreeBSD (I do
not know really if it should be working properly, but I suspect that the
answer is yes).

Nothing I tried since then seems to help, but I did not reset CMOS.

Regards,

Fabrice Philibert-Caillat

2016-09-09 16:51 GMT+02:00 Adrian Chadd :

> is it too new to "kldload i915kms" ?
>
>
> -adrian
>
>
> On 9 September 2016 at 00:45, Philip Homburg 
> wrote:
> >>What graphics driver are you using?
> >
> > Just the default:
> > 'VT(vga): resolution 640x480'
> >
> > Booting with -v and switching I now noticed that each time I switch to a
> > different console I get a line
> > 'Timeout initializing vt_vga'
> >
> >
> > ___
> > freebsd-current@freebsd.org mailing list
> > https://lists.freebsd.org/mailman/listinfo/freebsd-current
> > To unsubscribe, send any mail to "freebsd-current-unsubscribe@
> freebsd.org"
> ___
> freebsd-current@freebsd.org mailing list
> https://lists.freebsd.org/mailman/listinfo/freebsd-current
> To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"
>
___
freebsd-current@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-current
To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"

Re: 11.0-RC2 suspend/resume on thinkpad x201 kills poweroff

[Adrian Chadd]

is it too new to "kldload i915kms" ?


-adrian


On 9 September 2016 at 00:45, Philip Homburg  wrote:
>>What graphics driver are you using?
>
> Just the default:
> 'VT(vga): resolution 640x480'
>
> Booting with -v and switching I now noticed that each time I switch to a
> different console I get a line
> 'Timeout initializing vt_vga'
>
>
> ___
> freebsd-current@freebsd.org mailing list
> https://lists.freebsd.org/mailman/listinfo/freebsd-current
> To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"
___
freebsd-current@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-current
To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"

Re: 11.0-RC2 suspend/resume on thinkpad x201 kills poweroff

[Philip Homburg]

>What graphics driver are you using?

Just the default:
'VT(vga): resolution 640x480'

Booting with -v and switching I now noticed that each time I switch to a
different console I get a line 
'Timeout initializing vt_vga'


___
freebsd-current@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-current
To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"

Re: 11.0-RC2 suspend/resume on thinkpad x201 kills poweroff

[Ccs189]

Not sure if it helps. I have X1 Carbon 1st generation. What I did was recompile 
the kernel without vesa driver since the 11-current. And I am able to resume 
successful without the reset.video=1 even now with 11-RC2 I still ok. 

Best regards,
Chan


> On 9 Sep 2016, at 8:10 AM, Adrian Chadd  wrote:
> 
> Try forcing a console switch - hit alt-f1, alt-f2, etc.
> 
> 
> -a
> 
> 
>> On 8 September 2016 at 15:10, Glen Barber  wrote:
>>> On Thu, Sep 08, 2016 at 11:32:24PM +0200, Philip Homburg wrote:
>>> The main culprit seems to be putting 'hw.acpi.reset_video=1' in
>>> /boot/loader.conf
>> 
>> Yes, I have seen this cause problems on some (not all) hardware.
>> 
>>> this causes resume to hang half way through. I managed to trigger
>>> CMOS corruption once, but I can't systematically trigger it.
>>> 
>>> Without hw.acpi.reset_video=1 in /boot/loader.conf, resume works except that
>>> the screen stays black.
>> 
>> What graphics driver are you using?
>> 
>> Glen
> ___
> freebsd-current@freebsd.org mailing list
> https://lists.freebsd.org/mailman/listinfo/freebsd-current
> To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"
___
freebsd-current@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-current
To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"

Re: 11.0-RC2 suspend/resume on thinkpad x201 kills poweroff

[Adrian Chadd]

Try forcing a console switch - hit alt-f1, alt-f2, etc.


-a


On 8 September 2016 at 15:10, Glen Barber  wrote:
> On Thu, Sep 08, 2016 at 11:32:24PM +0200, Philip Homburg wrote:
>> The main culprit seems to be putting 'hw.acpi.reset_video=1' in
>> /boot/loader.conf
>>
>
> Yes, I have seen this cause problems on some (not all) hardware.
>
>> this causes resume to hang half way through. I managed to trigger
>> CMOS corruption once, but I can't systematically trigger it.
>>
>> Without hw.acpi.reset_video=1 in /boot/loader.conf, resume works except that
>> the screen stays black.
>>
>
> What graphics driver are you using?
>
> Glen
>
___
freebsd-current@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-current
To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"

Re: 11.0-RC2 suspend/resume on thinkpad x201 kills poweroff

[Glen Barber]

On Thu, Sep 08, 2016 at 11:32:24PM +0200, Philip Homburg wrote:
> The main culprit seems to be putting 'hw.acpi.reset_video=1' in
> /boot/loader.conf
> 

Yes, I have seen this cause problems on some (not all) hardware.

> this causes resume to hang half way through. I managed to trigger
> CMOS corruption once, but I can't systematically trigger it.
> 
> Without hw.acpi.reset_video=1 in /boot/loader.conf, resume works except that
> the screen stays black.
> 

What graphics driver are you using?

Glen



signature.asc
Description: PGP signature

Re: 11.0-RC2 suspend/resume on thinkpad x201 kills poweroff

[Philip Homburg]

The main culprit seems to be putting 'hw.acpi.reset_video=1' in
/boot/loader.conf

this causes resume to hang half way through. I managed to trigger
CMOS corruption once, but I can't systematically trigger it.

Without hw.acpi.reset_video=1 in /boot/loader.conf, resume works except that
the screen stays black.


___
freebsd-current@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-current
To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"

Re: 11.0-RC2 suspend/resume on thinkpad x201 kills poweroff

[Philip Homburg]

In your letter dated Wed, 7 Sep 2016 22:58:29 + you wrote:
>Any chance you can try this with 10.3-RELEASE?

On 10.3, I get a kernel panic during suspend (in something vesa related), so I 
can't trigger the problem there.

On 11.0-RC2 suspend works, resume mostly works except that the screen stays
blank, and then at some point the CMOS corruption happens.


___
freebsd-current@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-current
To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"

Re: 11.0-RC2 suspend/resume on thinkpad x201 kills poweroff

[Glen Barber]

On Thu, Sep 08, 2016 at 12:48:28AM +0200, Philip Homburg wrote:
> Hi,
> 
> I found out that playing with suspend/resume (using acpiconf -s 3) on a
> thinkpad x201 running 11.0-RC2 changes something in CMOS that prevents
> poweroff from working.
> 
> I managed to trigger the problem om two x201 laptops (though from the
> a single 11.0-RC2 installation).
> 
> Opening up the laptops to clear the CMOS solved the problem.
> 
> If I can find some time, I'll try to see if there is an interaction with
> for example the wifi driver.
> 
> When in this state, it also prevents suspend or poweroff from Linux (I tried
> a Ubuntu live CD). 
> 

Any chance you can try this with 10.3-RELEASE?

Glen



signature.asc
Description: PGP signature

11.0-RC2 suspend/resume on thinkpad x201 kills poweroff

[Philip Homburg]

Hi,

I found out that playing with suspend/resume (using acpiconf -s 3) on a
thinkpad x201 running 11.0-RC2 changes something in CMOS that prevents
poweroff from working.

I managed to trigger the problem om two x201 laptops (though from the
a single 11.0-RC2 installation).

Opening up the laptops to clear the CMOS solved the problem.

If I can find some time, I'll try to see if there is an interaction with
for example the wifi driver.

When in this state, it also prevents suspend or poweroff from Linux (I tried
a Ubuntu live CD). 


___
freebsd-current@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-current
To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"