Andreas, Okay, that sounds like a good plan, lets try to meet Sunday at some time to recap progress over the weekend. We are on the same page, FYI for ARM it is coprocessor 15 register 2 :) Regards, Brian Andreas Nilsson wrote: > Brian, > > I'll let Christoffer answer the svn question. > > Just merge the changes into the main branch and we'll look at them there. > > As for our status we're essentially working with booting the kernel. I wrote > an entry in the KVM userspace wiki page about the bootloader and kernel > decompression process. The first problem we naturally encountered (after > running 6 instructions :-) is the memory mapping. Hence we're currently > looking at the tlb/mmu emulation and it would make sense if you started with > the synchronous exception implementation and extended emulation work during > the weekend. > > Just so we're on the same page, the "page table base reg" in the coprocessor > corresponds to the x86 cr3 right? > > Dave continues with QEMU graphics this week and then he'll probably start > testing our KVM prototype on the physical device next week. > > /Andreas > > -----Original Message----- > From: android-virt-bounces at lists.cs.columbia.edu > [mailto:android-virt-bounces at lists.cs.columbia.edu] On Behalf Of Brian Smith > Sent: den 15 april 2009 22:03 > To: android-virt at lists.cs.columbia.edu > Subject: SVN repo > > Hi all, > I have finished up my testing of the emulation code and have checked > it into the emulation branch. If you are interested take > a look at it before I merge it into the main branch (probably Friday or > early Saturday). I have a new test program to update > in the SVN repository, but I noticed after pulling the most recent > updates the Makefile still contains my path information, > which I though Christoffer said he had changed to rely on $PATH. Can > someone let me know if I have the most recent > Makefile in svn, or if something got messed up (I am at revision 438), I > don't want to re-mess up anything that has already > been fixed. > Now that the "core emulation" is more solid I was hoping to get a > status update on what everyone is working on so I can > get a better feel for what to concentrate on this weekend. Unless I > hear differently, here is what I'm going to shoot for: > > * Take control of all synchronous exceptions for guest (PREFETCH ABORT, > DATA ABORT, UNDEFINED, SWI) > * Begin "exception emulation" work > * Begin coprocessor emulation for more critical functions (page table > base reg, fault status regs, etc) > * mmu/tlb type work > > Brian > _______________________________________________ > Android-virt mailing list > Android-virt at lists.cs.columbia.edu > https://lists.cs.columbia.edu/cucslists/listinfo/android-virt > > > _______________________________________________ > Android-virt mailing list > Android-virt at lists.cs.columbia.edu > https://lists.cs.columbia.edu/cucslists/listinfo/android-virt > >
