MEI Module Breaks MX-17 Suspend On C710 Chromebook

Report Bugs, Issues and non- package Requests
Post Reply
Message
Author
User avatar
calinb
Posts: 80
Joined: Tue Jun 27, 2017 1:57 am

MEI Module Breaks MX-17 Suspend On C710 Chromebook

#1 Post by calinb »

I'm starting a new thread here to bring this issue to the attention of the MX and antiX kernel gurus. It sure looks like a kernel (module) bug to me. Perhaps a kernel.org bug report is in order.

Here is the fix:
viewtopic.php?p=437144#p437144

Here is a history of a discussion (with dmesg snippets, etc.) that I had with Chromebook ROM dev, John Lewis, who tipped me off to the root cause and solution:
https://johnlewis.ie/custom-chromebook- ... ment-35772

I don't know what the consequence of blacklisting the modules containing the "support" code and binary blobs for Intel's MEI might be. From what I've read elsewhere, including the antiX forum, I know that it's impossible to completely disable MEI and I also know that MEI is evil! :eek:

User avatar
Stevo
Developer
Posts: 12838
Joined: Fri Dec 15, 2006 8:07 pm

Re: MEI Module Breaks MX-17 Suspend On C710 Chromebook

#2 Post by Stevo »

It seems to me that you could also manually unload the modules before trying to suspend in order to test it out. The configuration files for suspend also have provisions for unloading specified modules before suspend and reloading them after resume. I don't know if that's in tlp, too; let me look at the script.

User avatar
calinb
Posts: 80
Joined: Tue Jun 27, 2017 1:57 am

Re: MEI Module Breaks MX-17 Suspend On C710 Chromebook

#3 Post by calinb »

Stevo wrote:It seems to me that you could also manually unload the modules before trying to suspend in order to test it out. The configuration files for suspend also have provisions for unloading specified modules before suspend and reloading them after resume. I don't know if that's in tlp, too; let me look at the script.
Yes--unloading the modules should work too, as long as mei_me is unloaded before mei, because mei_me is using mei. I'm not familiar with the suspend configuration files or script but it's probably worth trying it out and it's a good tip to perhaps keep mei up and running except during suspend/resume cycles. John Lewis noted that mei slows the performance of suspend on his Ubuntu web server (a C710), even though suspend remains functional on his computer, so I'm also rather interested in learning about any potential issues associated with just keeping it blacklisted 24/7/365!!!

Again, that's a good tip you've suggested. In all my hours of online searching through various kernel module problems that block suspend from working on various machines (not just Chromebooks), this is the first I've heard of that option.

Post Reply

Return to “Bugs and Non-Package Requests Forum”