User Details
- User Since
- May 10 2014, 2:21 PM (588 w, 2 d)
Yesterday
It was a beri driver in sys/dev/beri/virtio/virtio_mmio_platform.c
Sat, Aug 16
Thu, Aug 14
Wed, Aug 13
Tue, Aug 12
Mon, Aug 11
Fri, Aug 8
Thu, Aug 7
Rebase past 5d56a04d1b99c0187c7799be0d1e93586312bd55
I'm not sure there is anything to assert on. CPU_OFF is defined to "Power down the calling core" & doesn't take any arguments.
Update based on feedback from @markj
Add a missing header file
CPU_OFF can only be called from the target CPU
Wed, Aug 6
Tue, Aug 5
Doesn't appear to be needed, the host kernel won't use it as it's not enabled.
Mon, Aug 4
Fri, Aug 1
Thu, Jul 31
Wed, Jul 30
Only handle copyin/out/etc
Will keep Harry as Author
It looks like it's in downstream dts files, e.g. https://github.com/raspberrypi/linux/blob/45079112ab81154eaf418ff05a989e6fdf63db79/arch/arm/boot/dts/broadcom/bcm2710-rpi-cm3.dts#L143-L148
Do you plan to return to EL2 on systems that booted from it but don't have VHE?
Mon, Jul 28
VM_MEMATTR_UNCACHEABLE should be ok for this as the firmware is running on the GPU
Thu, Jul 24
Is the firmware cache-coherent? If not we will need to ensure the value is written back to memory.
Mon, Jul 21
Jul 18 2025
Rebase + remove libopencsd-all.ucl
Jul 17 2025
I'm unsure why the libopencsd needs to exist, it will be used by the userspace hwt tool (D40728) to decode trace data from the CPU.