No, and actually, the power circuit doesn't need any CPU to charge the device.
Maybe, it depends upon how things are wired up inside.
Uses for the Cortex-M4 cores on the OMAP5?
- Thread starter fahrstuhl
- Start date
pmprog
DNF (Did Not Finish)
- Joined
- Apr 25, 2011
- Messages
- 4,150
Must have missed that then, so it won't boot up when you start charging like the Pandora does?
ible
professional vim user
well, maybe not full fledged linux -- since 200 MHz is pretty slow (maybe an RTOS?), though if it can offload stuff to the GPU then it makes for interesting stuff.
at any rate, it's not possible to use the M4s without the A15s on?
Exophase
Nothing good will ever come of Exophase.
If by GPU you mean the SGX I think there's going to be a real difficulty using the user space binary blob drivers on the M4s.
It'd be really strange if this were not possible.
levi
Still fresh, damnit!
I'm idly wondering if one of the M4s could be used to keep the DSP fed with data. Is there a need to make that easier to schedule?
An M4 processor could be used as a flashlight controller if there is an external white LED that the M4 can control. A flashlight called the Eternalight has a mode in which it turns an LED on and off at a rate that is faster than the human optic nerve response time so that the light looks steady to a human eye but consumes less energy than having the LED constantly on all the time. The Eternalight can operate for 700 hours in this mode. Perhaps the M4 can do something similar.
WizardStan
Mega GP Mania
- Joined
- May 24, 2008
- Messages
- 16,731
That's called PWM: pulse width modulation. That's literally how dimmable backlighting works: it consumes less power by being on less and therefore being not as bright. I believe all of the LEDs on the Pyra will be dimmable in this manner, just replace the RGBs behind the logo with bright whites and set them to 50% and get ages and ages of flashlight.
There's all kinds of instructions for building a PWM USB device, it's not very difficult. It would probably consume more energy using the M4s to regulate it than using the dedicated hardware.
There's all kinds of instructions for building a PWM USB device, it's not very difficult. It would probably consume more energy using the M4s to regulate it than using the dedicated hardware.
kuru
Who the heck is Stephan Kramer?
What would an M4 be used for in automotive context? To drive the always-on digital clock?
txus
Member
- Joined
- Mar 23, 2014
- Messages
- 115
AFAIK the M4s are used to access the hardware video accelerator. There is very little info around, but this seems to be the latest code (if interested, look into libdce and gst-ducati): https://gitlab.com/groups/gstreamer-omap
Exophase
Nothing good will ever come of Exophase.
There is PWM capability in the OMAP5's timers too, although I have no idea if any can go to exposed GPIOs that you can attach anything to (but it'd be the same with the M4s)
Just a probably stupid n00b question:
Would it be possible to utilize the small core to keep exagear (and wine) running without a performance loss?
This would solve the whole problem with people disliking ARM (and Linux) for some reason.
Would it be possible to utilize the small core to keep exagear (and wine) running without a performance loss?
This would solve the whole problem with people disliking ARM (and Linux) for some reason.
I think you have a wrong idea of how "small" those two cores are. Running wine on a 200MHz (I think?) Cortex-M4? I don't think you'd have a lot of fun with that ^^. I can imagine that running things like sound effects and music on them (maybe even emulating old sound chips on them) being feasible (and useful). Otherwise, I don't really see much use in them in this kind of device, tbh.
slaeshjag
¯\_(ツ)_/¯
Keep in mind that the M4 cores can't run code compiled for the Cortex-A cores. They are also special in terms of memory layout. Porting normal userspace applications is most probably way too cumbersome for no benefit. The M4 cores have about as many usecases as a DSP, except it's a slow general purpose processor whose main advantage is predictable performance. Performance will suck on a sub-GP32 level, but it will be predictable and relatively uninterrupted. It will most probably require software written especially for it for it to be even vaguely useful. Just shotgunning usage scenarios without knowing the technical details is unlikely to find anything that sticks.
stevenc99
Member
- Joined
- May 10, 2016
- Messages
- 99
These are most useful when you need to sample inputs very fast, like an oscilloscope or logic analyser does, or odometer attached to a bicycle wheel. They can usually read a portion of system RAM and use it to generate complex output, like a video signal. Or respond quickly inputs by changing some output, like controlling the rotors on a quadrocopter. I really hope to see people making these things, especially a flying Pyra.
Orbstheorem
Member
- Joined
- May 24, 2016
- Messages
- 47
I'd love to be able to *suspend* my pyra and have it vibrate when a new sms arrives so I can wake it up from suspension
[doublepost=1464130511,1464129562][/doublepost]How does one actually access to M4 cores from linux?
Do you have to write into some "magic offset" to talk to the M4?
Any real world code examples available?
As stated up ^^ I'd like to implement something so I can suspend my pyra when not using it but still vibrate when I get an SMS or a call...
SNESFAN
Retro game fanatic
How about a hypervisor boot environment on the m4's that could then boot different VMs running different kernels for the a15 cores?
ible
professional vim user
i think whenever you say "hypervisor", you can't use the M4s for it...
