Hi, On 07/05/2018 09:51 AM, Peter Maydell wrote: > On 4 July 2018 at 16:51, Will Deacon <will.deacon@xxxxxxx> wrote: >> On Wed, Jul 04, 2018 at 03:41:18PM +0100, Marc Zyngier wrote: >>> Not really. Let's say I want my IPA space split in two: memory covers >>> the low 47 bit, and I want MMIO spanning the top 47 bit. With your >>> scheme, you'd end-up with a 47bit IPA space, while you really want 48 >>> bits (MMIO space implemented by userspace isn't registered to the >>> kernel). >> >> That still sounds quite niche for a VM. Does QEMU do that? > > Not at 47 bits, but we have RAM up to the 256GB mark, and > MMIO above that (including a large PCI window), so the general > arrangement of having the top end of the IPA space not > necessarily be things we've told the kernel about definitely > exists. Is this document (2012) still a reference document? http://infocenter.arm.com/help/topic/com.arm.doc.den0001c/DEN0001C_principles_of_arm_memory_maps.pdf (especially Fig 5?) Peter, comments in QEMU hw/arm/virt.c suggested next RAM chunk should be added at 2TB. This doc suggests to put it at 8TB. I understand the PA memory map only is suggested but shouldn't we align? Thanks Eric > > thanks > -- PMM > _______________________________________________ > kvmarm mailing list > kvmarm@xxxxxxxxxxxxxxxxxxxxx > https://lists.cs.columbia.edu/mailman/listinfo/kvmarm > _______________________________________________ kvmarm mailing list kvmarm@xxxxxxxxxxxxxxxxxxxxx https://lists.cs.columbia.edu/mailman/listinfo/kvmarm
