This needed only very minor tweaks, and interestingly only involved more unfolding of Kernel_Config.config_ARM_PA_SIZE_BITS_40_def.
Most of the unfoldings involved goals where concrete vmlevel numbers now overflow earlier (3 = 0 for config_ARM_PA_SIZE_BITS_40, instead of 4 = 0).
Together with the maxIRQ pull request #773, this enables the proof to work for the zynqmp (zcu102) and bcm2711 (rpi4) platforms. See also seL4/seL4#1290.
This needed only very minor tweaks, and interestingly only involved more unfolding of
Kernel_Config.config_ARM_PA_SIZE_BITS_40_def
.Most of the unfoldings involved goals where concrete
vmlevel
numbers now overflow earlier (3 = 0 for config_ARM_PA_SIZE_BITS_40, instead of 4 = 0).Together with the
maxIRQ
pull request #773, this enables the proof to work for thezynqmp
(zcu102
) andbcm2711
(rpi4
) platforms. See also seL4/seL4#1290.