Contribute
Register

Laptop won't sleep/shutdown after a few hours of uptime

Status
Not open for further replies.
JVital: I wrote a new patch below: instead of preventing the ME driver from starting, this one prevents the driver from sending any messages to the intel ME. It solves the sleep/shutdown issues without any of the side effects from the previous patch (i.e. some applications taking a long time to start after sleep). Let me know if it works for you too.
Code:
<dict>
    <key>Name</key>
    <string>AppleIntelFramebufferAzul</string>
    <key>Comment</key>
    <string>Prevent ME driver from writing to ME</string>
    <key>Find</key>
    <string>84C90F84BC0100004489F1</string>
    <key>Replace</key>
    <string>84C9E9BD010000904489F1</string>
</dict>

Works prefect with me till now Dell 3537, i didn't face any of the side effects html5 or vlc

Thanks a lot
 
Works prefect with me till now Dell 3537, i didn't face any of the side effects html5 or vlc

Thanks a lot

Good to know, thanks for the feedback. So far, that's 2 Dell and one Acer model with the same issue, all haswell. Out of curiosity, can you guys confirm if your bioses are Phoenix SecureCore Tiano?
 
I don't think anyone expected it would.

You mentioned earlier that it might be related, but you weren't sure. Just confirming that it wasn't :p

JVital: I wrote a new patch below: instead of preventing the ME driver from starting, this one prevents the driver from sending any messages to the intel ME. It solves the sleep/shutdown issues without any of the side effects from the previous patch (i.e. some applications taking a long time to start after sleep). Let me know if it works for you too.

This patch works wonders! I have no idea how you figure this stuff out; I'd love to learn how to pull code apart like you do if I ever get some free time. I really appreciate all your work to fix this issue. I also want to apologize for replying slowly; college is really killing me.

Good to know, thanks for the feedback. So far, that's 2 Dell and one Acer model with the same issue, all haswell. Out of curiosity, can you guys confirm if your bioses are Phoenix SecureCore Tiano?

Unfortunately not. My laptop has a version of the InsydeH2O UFEI BIOS.
 
This patch works wonders! I have no idea how you figure this stuff out; I'd love to learn how to pull code apart like you do if I ever get some free time. I really appreciate all your work to fix this issue. I also want to apologize for replying slowly; college is really killing me.
Good to hear it works fine on different laptops. It's been perfectly stable on my side for a week without reboot.

Unfortunately not. My laptop has a version of the InsydeH2O UFEI BIOS.
Well there goes that theory :) That's actually probably good news, meaning it may not be a bios bug. Someone smarter than me will have to come along to use that information in a meaningful way though as I have no idea how to make the ME work properly. At least we have a functional workaround in the meantime.
 
Yet again, after a few days of testing and some tinkering, I found an issue with this patch. Every time it wakes from sleep, I have low screen brightness and html5 video surfaces do not work for about 2 minutes (I didn't test Quicktime or VLC in this time). If I let it go, the brightness will eventually come back up to what I had it set to from before and all video surfaces work normally. If I try changing the brightness while it's low, all input freezes until the brightness comes back up. Interestingly enough, the screen itself does not freeze - if a web page is loading, for example, it will finish loading.

I didn't say anything about this before because it was a sporadic issue I used to have every ~20 sleep cycles or so, but now it happens all the time and lasts for longer. I tried changing several things in my DSDT/SSDT to see if I patched it wrong, but I think everything's OK - all of my new changes did nothing to solve the issue. I also used the debug version of ACPIBacklight.kext and it showed that it was being called to set the brightness immediately after wake, but it does't have an apparent effect for the 2 minutes. It's really not that bad so I'll keep using the patch, but insight would be nice :). If I had to guess, I'd say that the MEI driver is trying to write something after it comes out of sleep and, after some kind of timeout, it gives up. In reality though, I have no idea what I'm talking about ;)
 
Yet again, after a few days of testing and some tinkering, I found an issue with this patch. Every time it wakes from sleep, I have low screen brightness and html5 video surfaces do not work for about 2 minutes (I didn't test Quicktime or VLC in this time). If I let it go, the brightness will eventually come back up to what I had it set to from before and all video surfaces work normally. If I try changing the brightness while it's low, all input freezes until the brightness comes back up. Interestingly enough, the screen itself does not freeze - if a web page is loading, for example, it will finish loading.

I didn't say anything about this before because it was a sporadic issue I used to have every ~20 sleep cycles or so, but now it happens all the time and lasts for longer. I tried changing several things in my DSDT/SSDT to see if I patched it wrong, but I think everything's OK - all of my new changes did nothing to solve the issue. I also used the debug version of ACPIBacklight.kext and it showed that it was being called to set the brightness immediately after wake, but it does't have an apparent effect for the 2 minutes. It's really not that bad so I'll keep using the patch, but insight would be nice :). If I had to guess, I'd say that the MEI driver is trying to write something after it comes out of sleep and, after some kind of timeout, it gives up. In reality though, I have no idea what I'm talking about ;)

That's odd, I do not see anything like this on my end - backlight always come back to the same level as it was before sleep. Does it go back to being a sporadic issue if you disable/remove the ME patch?
 
That's odd, I do not see anything like this on my end - backlight always come back to the same level as it was before sleep. Does it go back to being a sporadic issue if you disable/remove the ME patch?

I just double checked and yep, it does. Or, at least it doesn't happen all the time anymore. It happened extremely rarely so it's hard to test. I apply the patch again and bam, brightness issues.
 
I just double checked and yep, it does. Or, at least it doesn't happen all the time anymore. It happened extremely rarely so it's hard to test. I apply the patch again and bam, brightness issues.

Sorry to say I have no idea how to help on this one.. I'll let you know if I see something like this happening on my side though.
 
Works prefect with me till now Dell 3537, i didn't face any of the side effects html5 or vlc

Thanks a lot
<dict>
<key>Name</key>
<string>AppleIntelFramebufferAzul</string>
<key>Comment</key>
<string>Prevent ME driver from writing to ME</string>
<key>Find</key>
<string>84C90F84BC0100004489F1</string>
<key>Replace</key>
<string>84C9E9BD010000904489F1</string>
</dict>
how did you fix shutdown/sleep problem?How did you patch this above? Please tell I don't know where to start. Thank you very much
 
Status
Not open for further replies.
Back
Top