Skip to content
This repository has been archived by the owner on Aug 5, 2022. It is now read-only.

Untracked MMIO r/w #14

Open
codyit opened this issue May 27, 2015 · 4 comments
Open

Untracked MMIO r/w #14

codyit opened this issue May 27, 2015 · 4 comments

Comments

@codyit
Copy link

codyit commented May 27, 2015

Having a lot of these for Iris5100 on mbpr13" late'13 ( 0x0a2e), a haswell chip not yet included in your device table:
https://gist.github.com/codyit/4c80895d9a65625232db

Bug, or unimplemented

kernel: [ 13.727266] vGT warning:(vgt_emulate_write:423) vGT: untracked MMIO write: vm_id(0), offset=0x43408,len=4, val=0x80000000!!!
kernel: [ 13.727389] vGT warning:(vgt_emulate_read:326) vGT: untracked MMIO read: vm_id(0), offset=0x43408,len=4, val=0x80000000!!!

@tcwu2005
Copy link

I saw similar warning of Iris5200.
Dodyit, is your guest working fine with GPU passthrough?

@codyit
Copy link
Author

codyit commented Nov 3, 2015

It was quite a while ago and I'm pretty drained these days.
IIRC although performance wasn't great, it was working.
(Also IIRC) It would crash from time to time, possibly due to these unimplemented MMIO?
I've no idea.
I guess you'd have better luck if your device is implemented as per their device list.

@codyit
Copy link
Author

codyit commented Nov 3, 2015

It's just normal considering I had to manually add my device into the list, just created this in case we have more mbpr users interested in this, oh well

@zhiyuanlv
Copy link

Sorry left your question for long time! Are you running Linux guest? The MMIOs are display related, and in the new GVT-g release, we disabled IPS feature inside guest, and there should not be such warnings. Even having such warning, it should not cause crash I think.

We have mailing list now "[email protected]", and there are more QA people there, who may give more timely response. Thanks!

Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Labels
None yet
Projects
None yet
Development

No branches or pull requests

3 participants