You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
This is an issue for us, as we use the that information to determine which RPi is actually running, and, how we optimize for those devices (minus cpu core detection $(nproc --all) for software installs such as php-fpm thread count). https://github.com/Fourdee/DietPi/issues/936#issuecomment-301055299
Any chance we could see a possible fix for correct Hardware information, reported in /proc/cpuinfo?
This is an unintended but unavoidable consequence of using less downstream, Pi-specific code in newer kernels. I suggest you use /proc/device-tree/model (human readable) or extract the information you need from the Revision field of /proc/cpuinfo. There is a bit in the revision number indicating that it's in the new format: if it's set then the CPU type is in another group of bits, otherwise it must be a 2708.
Activity
clivem commentedon May 13, 2017
#1725
pelwell commentedon May 13, 2017
This is an unintended but unavoidable consequence of using less downstream, Pi-specific code in newer kernels. I suggest you use /proc/device-tree/model (human readable) or extract the information you need from the Revision field of /proc/cpuinfo. There is a bit in the revision number indicating that it's in the new format: if it's set then the CPU type is in another group of bits, otherwise it must be a 2708.
Fourdee commentedon May 14, 2017
@pelwell
Thanks for the info 👍
I've made the switch to using revision codes for board detection:
https://github.com/Fourdee/DietPi/blob/testing/dietpi/dietpi-obtain_hw_model#L125-L364
I'll mark as closed, [dupe] of raspberrypi/firmware#705
karu2003 commentedon Jul 22, 2017
popcornmix
Thank you!!!!