[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] Re: [Xen-users] Debugging DomU
On 14/05/2015 20:06, Chris (Christopher) Brand wrote: Hi Julien, Hi Christopher, This is normal that address 0 is not mapped to any RAM. The memory layout of the guest (see xen/include/public/arch-arm.h) as the RAM guest defined in higher memory.Ok. That seems to match what I'm seeing in the MMU - VTTBR for that guest contains two L1 tables, one at 0 and one at 0x40000000 (GUEST_RAM_BASE). The first is almost empty while the second has quite a few mappings. I assume VTTBR is where I should be looking ?Although, vmlinux is using virtual address, so 0 is likely a virtual address. Do you know if the page table has been setup to the guest?I see a call to p2m_alloc_table(), followed by a number of calls to p2m_create_table(). p2m_alloc_table() and p2m_create_table() are used in Xen for creating the stage-2 page table. I was asking if the kernel already setup his PT and enable pagination. The error you describe in your mail is usually happening in the early setup.That seems to be the case, yes. I haven't been able to see anything that's happening in the VM before this.I don't think this is the bug, as said above we don't map anything to address 0 and this is normal because it's not part of the layout.Ok. It certainly wouldn't be surprising if I'm getting my address types mixed up :-)It would be useful if you provide the version of Xen and Linux you use.Xen is a recent git clone - cb34a7c8d741, with a couple of small hacks that we needed to get to the point of Dom0 running on our platform (there's a change to xen/drivers/char/ns16550.c to workaround a uart bug, and one to xen/arch/platforms/brcm.c to set dom0_gnttab_start and dom0_gnttab_end). Linux Kernel is 3.14. I took our kernel, enabled the various CONFIG options for Xen and rebuilt it. The important CONFIG_* that should be enable are: - CONFIG_XEN_* - CONFIG_ARCH_MULTIPLATFORM (required to enable the next option) - CONFIG_ARCH_VIRT - CONFIG_AUTO_ZRELADDR - CONFIG_ARM_PATCH_PHYS_VIRTI would also advice you to backport the commit 238962ac71910d6c20162ea5230685fead1836a4 "ARM: 8191/1: decompressor: ensure I-side picks up relocated code". It may help to boot further as it fixes an issue with similar symptoms as yours. Regards, -- Julien Grall _______________________________________________ Xen-users mailing list Xen-users@xxxxxxxxxxxxx http://lists.xen.org/xen-users
|
Lists.xenproject.org is hosted with RackSpace, monitoring our |