P9.0.0_2.2.0-ga on imx6 getting hot and shutting down

cancel
Showing results for 
Search instead for 
Did you mean: 

P9.0.0_2.2.0-ga on imx6 getting hot and shutting down

460 Views
sasidharan_sree
Contributor I

Hi,

Using i.MX Android™ P9.0.0_2.2.0-ga release on imx6. On running a graphics test app (GPUBenchmark3D), system is getting hot and shutting down. Is this a known issue? What could be the reason? 

dmesg log:

audit: rate limit exceeded
audit: audit_lost=17193 audit_rate_limit=5 audit_backlog_limit=64
type=1400 audit(1579179410.160:17489): avc: denied { dac_read_search } for pid=216 comm="lmkd" capability=2 scontext=u:r:lmkd:s0 tcontext=u:r:lmkd:s0 tclass=capability permissive=0 duplicate messages suppressed
audit: rate limit exceeded
type=1400 audit(1579179410.190:17490): avc: denied { dac_read_search } for pid=1671 comm="main" capability=2 scontext=u:r:zygote:s0 tcontext=u:r:zygote:s0 tclass=capability permissive=0
System is a little hot. GPU3D clock will work at 32/64
thermal thermal_zone0: critical temperature reached (100 C), shutting down
reboot: Failed to start orderly shutdown: forcing the issue
Emergency Sync complete
type=1400 audit(1579179410.190:17492): avc: denied { dac_read_search } for pid=1671 comm="main" capability=2 scontext=u:r:zygote:s0 tcontext=u:r:zygote:s0 tclass=capability permissive=0 duplicate messages suppressed
type=1400 audit(1579179497.140:17499): avc: denied { dac_read_search } for pid=1710 comm="main" capability=2 scontext=u:r:zygote:s0 tcontext=u:r:zygote:s0 tclass=capability permissive=0
audit: audit_lost=17199 audit_rate_limit=5 audit_backlog_limit=64
type=1400 audit(1579179497.140:17502): avc: denied { dac_read_search } for pid=1710 comm="main" capability=2 scontext=u:r:zygote:s0 tcontext=u:r:zygote:s0 tclass=capability permissive=0 duplicate messages suppressed
audit: rate limit exceeded
imx2-wdt 20bc000.wdog: Device shutdown: Expect reboot!
type=1400 audit(1579179497.240:17503): avc: denied { dac_read_search } for pid=216 comm="lmkd" capability=2 scontext=u:r:lmkd:s0 tcontext=u:r:lmkd:s0 tclass=capability permissive=0
android_work: sent uevent USB_STATE=DISCONNECTED
reboot: Power down

Regards,

Sreejith

Labels (2)
Tags (1)
0 Kudos
3 Replies

343 Views
gusarambula
NXP TechSupport
NXP TechSupport

Hello Sreejith,

I have not heard of a known issue. However, I would recommend to review the conditions in which the benchmark is being run and the thermal design of the board.

The thermal monitor driver signals when there is an overheat event. Depending on the configuration there may be different results. The thermal driver can signal de GPU driver to reduce frequency (and thus, performance) or can even shutdown to avoid damage to the processor.

If you are using an NXP board, which is likely the case, it should not overheat under normal conditions, but a heatsink may be necessary if you will demand a lot from the cores or GPU.

I hope this helps!
Regards,

0 Kudos

343 Views
sasidharan_sree
Contributor I

Hi,

Heatsink is not present. But the same benchmark is running on Android Oreo 8.0.0 release for the imx6 without any overheat issue.

Regards,

Sreejith

0 Kudos

343 Views
gusarambula
NXP TechSupport
NXP TechSupport

Hello Sreejith,

It would be hard to point out what could be the cause here as there are several changes between Oreo and Pie, even the GPU driver. However, this is not a known issue. It would be good to review the overhead Android is giving on each version, which may cause additional CPU workload while the GPU is running.

Regards,

0 Kudos