Grey Screen On Waking MBP

[Dave Braine]

Every now and then, as I open the lid of my MBP to wake it, the Superdrive makes the normal power on noise, and the screen comes up as shades of grey, and a "progress bar" in the form of white dots appears.

It can take up to 3 or 4 minutes for the screen to get back to normal, and another 3 or 4 minutes to get control back; ie, keyboard and trackpad.

Has anyone else come across this? it used to happen with my old White MacBook as well.

 

[Dave Braine]

Has anyone else come across this?

I guess not then.

I'll rephrase the question: anyone any idea why this happens?

 

[chown33]

I think the machine is coming out of hibernatemode or autopoweroff. Those are the terms used with 'pmset':
https://developer.apple.com/legacy/library/documentation/Darwin/Reference/ManPages/man1/pmset.1.html

Simplest is to try setting autopoweroff to 0, and the same for hibernatemode. If that avoids the problem, then play around with the time interval and the mode.

The reason it takes so long is the machine is actually rebooting, then restoring a previously saved full memory image, which picks up where it left off. You're still paying the reboot cost, though.

Shortly after I got my MacBook Pro some years ago, I got fed up with waiting so long after it had been asleep for a day or so. So I started playing around extending the time interval (autopoweroffdelay). IIRC the interval is measured in seconds, and I'm at 5 days now (432000). Since my MBP is usually on the charger, and only infrequently off, this works for me. YMMV.

 

[Dave Braine]

Thanks for that. chown33. Your mention of yours usually being on the charger is interesting. The problem only seems to occur when mine is plugged in.

 

[chown33]

Thanks for that. chown33. Your mention of yours usually being on the charger is interesting. The problem only seems to occur when mine is plugged in.

I don't recall if that particular parameter has separate values for battery vs. charger. I was definitely seeing it in both cases, which I thought was stupid given the circumstance (on charger).

I would still start the exploration by turning off both hibernatemode and autopoweroff. Maybe a reboot, too, and then check with 'pmset -g' to list all the values. Then leave it asleep for 24 hrs and see what happens at next wakeup.

When I first tracked down the setting, I wasn't sure what to look for, so I started with "deep sleep" or "hibernate". That still didn't produce the results I wanted, so it took a while longer to find the autopoweroff setting and how it relates to hibernatemode.

FWIW, here's my current settings on charger:

pmset -g
Active Profiles:
Battery Power        -1
AC Power        -1*
Currently in use:
standbydelay         4200
standby              0
womp                 0
halfdim              1
hibernatefile        /var/vm/sleepimage
sms                  1
networkoversleep     0
disksleep            10
sleep                0
autopoweroffdelay    432000
hibernatemode        3
autopoweroff         1
ttyskeepawake        1
displaysleep         15
acwake               0
lidwake              1

And on battery:

Active Profiles:
Battery Power        -1*
AC Power        -1
Currently in use:
standbydelay         4200
standby              0
halfdim              1
sms                  1
hibernatefile        /var/vm/sleepimage
disksleep            10
sleep                20
autopoweroffdelay    432000
hibernatemode        3
autopoweroff         1
ttyskeepawake        1
displaysleep         4
acwake               0
lidwake              1

You can see some values differ, but most are the same.

 

[Dave Braine]

Thanks. I'm reading up on "safe sleep", hibernation(always thought it was a PC thing and not Macs) at the moment.

 

[Dave Braine]

So, after much reading and experimentation, I've set Hibernation Mode to 0, set autpoweroff to OFF and standby delay to 420000...and it still happens.

 

[chown33]

So, after much reading and experimentation, I've set Hibernation Mode to 0, set autpoweroff to OFF and standby delay to 420000...and it still happens.

autopoweroff should be 0 (a number).

Try setting autopoweroffdelay to 420000, not standbydelay.

Post the output of pmset -g .

Which OS version? Which Mac model (cmd: sysctl hw.model)?

 

[Dave Braine]

autopoweroff should be 0

It is. I just used OFF for the post.

Which OS version? Which Mac model

As per my signature.

Post the output of pmset -g

Active Profiles:
Battery Power -1*
AC Power -1
Currently in use:
standbydelay 420000
standby 0
halfdim 1
sms 1
hibernatefile /var/vm/sleepimage
disksleep 10
sleep 15
autopoweroffdelay 420000
hibernatemode 0
autopoweroff 1
ttyskeepawake 1
displaysleep 5
acwake 0
lidwake 1

 

[chown33]

It is. I just used OFF for the post.


As per my signature.

My MR profile preference has signatures off. Turning them on temporarily, shows me:

MacBook Pro(mid 2013), 10.10.5;​

That model seems like it would support at least the same options my MacBookPro9,2 does.

Have you rebooted since changing the settings?


Comparing your pmset -g output to mine, the main difference I see:

Yours: hibernatemode 0
Mine: hibernatemode 3​

You could try setting autopoweroff to 0.

Might also want to review the pmset man page and read the whole thing, paying close attention to the coupling of standby, hibernatemode, autopoweroff, etc.
https://developer.apple.com/legacy/library/documentation/Darwin/Reference/ManPages/man1/pmset.1.html

 

[Dave Braine]

Thanks for your help with this, chown33.

My MR profile preference has signatures off.

OK, didn't know you could do that.

Comparing your pmset -g output to mine, the main difference I see:
Yours: hibernatemode 0
Mine: hibernatemode 3

Yes, to go back to plain old fashioned sleep.

You could try setting autopoweroff to 0

I'll try that next. Thanks.

 

[chown33]

Yes, to go back to plain old fashioned sleep.

Try a mode of 3; it's worth a shot. I mention this because the man page isn't always completely accurate. For example, the docs say autopoweroffdelay is measured in minutes, but that's inconsistent with the default value, and inconsistent with my experience of non-default values. Those imply the value is measured in seconds.

 

[Dave Braine]

It was on mode 3 originally.

 

[chown33]

It was on mode 3 originally.

Granted.

When trying to diagnose a problem, I generally try to replicate the problem first. In other words, replicate the failure. If I can't do that, then I try to setup the system under test to match a known working system, i.e. replicate a success.

If the settings are exactly the same, but the failing system still malfunctions, then it's possible that the problem doesn't lie with that particular collection of settings. The malfunction may lie with the system component that's interpreting the settigns. On the other hand, if the known-working settings cause the failure to disappear, then whatever the settings are, even if they're illogical, they indicate a success.

At this point, AFAICT, you haven't completely replicated the settings I provided earlier. If you have, and the system still failed in the same way, under the same conditions, then I'm inclined to think that altering the settings may not, by itself, lead to success. Of course, that's just a gut feeling, considering there are other things that differ between my system and yours, not least of which is OS version.

Given the differences in OS version, and my own experiences in the past with Apple occasionally borking the sleep mgmt, I wouldn't be too surprised to learn there's a bug inherent tp the OS version. I also wouldn't be surprised to learn there's some magical collection of settings that causes things to work, and that collection isn't obvious, or is even counter-intuitive, or outright contradicts the docs.


As a last resort, or just as a "nuke it" strategy, you could try moving the plist file the settings are stored in to a location outside of /Library/Preferences, then reboot and see if the OS makes a default one. If it does, start over by seeing what the defaults actually do, and if the failure continues, then replicate the settings I posted.

 

[Dave Braine]

Thanks, chown33. Just waiting to see if the problem is still there.