Re: 2.6.27-rc1 crashing under Xen 3.1.2, 3.1.3

[Jeremy Fitzhardinge <jeremy@goop.org>]

Carl Jones wrote:
> Hi,
>
> I'm testing 2.6.27-rc1 and it appears to crash when booting under
> 3.1.2 and 3.1.3.
>
> Host is 2x Xeon 5410/32GB, 3.1.2 on x86_64 (from source).
>
> I see the same issue using 2.6.26. The same kernel will boot fine
> under 3.2.x, and 3.1.4 (just not 3.1.2 or 3.1.3). Older kernels e.g.
> 2.6.24 boot fine.
>
> domU has 256MB memory, 1 VCPU.
>
> dmesg output is below.
> Kernel config: http://radium.outervoid.net/~carl/2.6.27-rc1
>
> Regards,
> Carl
>
> Reserving virtual address space above 0xf57fe000
> Linux version 2.6.27-rc1-xenU (root@x) (gcc version 4.1.2 20071124
> (Red Hat 4.1.2-42)) #4 SMP Mon Aug 4 20:52:48 EDT 2008
> ------------[ cut here ]------------
> WARNING: at arch/x86/mm/ioremap.c:469 early_ioremap_init+0x126/0x210()
> Pid: 0, comm: swapper Not tainted 2.6.27-rc1-xenU #4
>  [<c012365a>] warn_on_slowpath+0x5a/0xa0
>  [<c0103a94>] xen_mc_flush+0xc4/0x190
>  [<c0104b02>] xen_set_pmd_hyper+0x72/0x90
>  [<c03095c6>] early_ioremap_init+0x126/0x210
>  [<c02ffc99>] setup_arch+0x29/0x5d0
>  [<c02fbb41>] start_kernel+0x61/0x2a0
>  [<c02ff16a>] xen_start_kernel+0x58a/0x650
>  =======================
> ---[ end trace 4eaa2a86a8e2da22 ]---
> pmd c033cd50 != c033cd58
> fix_to_virt(FIX_BTMAP_BEGIN): f55fe000
> fix_to_virt(FIX_BTMAP_END):   f56fd000
> FIX_BTMAP_END:       256
> FIX_BTMAP_BEGIN:     511
> BIOS-provided physical RAM map:
>  Xen: 0000000000000000 - 00000000000a0000 (usable)
>  Xen: 00000000000a0000 - 0000000000100000 (reserved)
>  Xen: 0000000000100000 - 0000000000366000 (usable)
>  Xen: 0000000000366000 - 00000000003a9000 (reserved)
>  Xen: 00000000003a9000 - 0000000010000000 (usable)
> last_pfn = 0x10000 max_arch_pfn = 0x1000000
> NX (Execute Disable) protection: active
> DMI not present or invalid.
> 0MB HIGHMEM available.
> 256MB LOWMEM available.
>   mapped low ram: 0 - 10000000
>   low ram: 00000000 - 10000000
>   bootmap 00002000 - 00004000
> (7 early reservations) ==> bootmem [0000000000 - 0010000000]
>   #0 [0000000000 - 0000001000]   BIOS data page ==> [0000000000 - 0000001000]
>   #1 [0000001000 - 0000002000]    EX TRAMPOLINE ==> [0000001000 - 0000002000]
>   #2 [0000006000 - 0000007000]       TRAMPOLINE ==> [0000006000 - 0000007000]
>   #3 [0000100000 - 00003656f0]    TEXT DATA BSS ==> [0000100000 - 00003656f0]
>   #4 [00003a9000 - 00003af000]    INIT_PG_TABLE ==> [00003a9000 - 00003af000]
>   #5 [0000007000 - 0000081000]          PGTABLE ==> [0000007000 - 0000081000]
>   #6 [0000002000 - 0000004000]          BOOTMAP ==> [0000002000 - 0000004000]
> Zone PFN ranges:
>   DMA      0x00000000 -> 0x00001000
>   Normal   0x00001000 -> 0x00010000
>   HighMem  0x00010000 -> 0x00010000
> Movable zone start PFN for each node
> early_node_map[3] active PFN ranges
>     0: 0x00000000 -> 0x000000a0
>     0: 0x00000100 -> 0x00000366
>     0: 0x000003a9 -> 0x00010000
> Local APIC disabled by BIOS -- you can enable it with "lapic"
> Allocating PCI resources starting at 20000000 (gap: 10000000:f0000000)
> PERCPU: Allocating 32180 bytes of per cpu data
> Built 1 zonelists in Zone order, mobility grouping on.  Total pages: 64861
> Kernel command line: root=/dev/xvda1 ro console=hvc
> Enabling fast FPU save and restore... done.
> Enabling unmasked SIMD FPU exception support... done.
> Initializing CPU#0
> PID hash table entries: 1024 (order: 10, 4096 bytes)
> Detected 1600.003 MHz processor.
> Console: colour dummy device 80x25
> console [tty0] enabled
> console [hvc0] enabled
> Dentry cache hash table entries: 32768 (order: 5, 131072 bytes)
> Inode-cache hash table entries: 16384 (order: 4, 65536 bytes)
> Memory: 256076k/262144k available (1505k kernel code, 5272k reserved,
> 495k data, 232k init, 0k highmem)
> virtual kernel memory layout:
>     fixmap  : 0xf574f000 - 0xf57fd000   ( 696 kB)
>     pkmap   : 0xf5200000 - 0xf5400000   (2048 kB)
>     vmalloc : 0xd0800000 - 0xf51fe000   ( 585 MB)
>     lowmem  : 0xc0000000 - 0xd0000000   ( 256 MB)
>       .init : 0xc02fb000 - 0xc0335000   ( 232 kB)
>       .data : 0xc027840f - 0xc02f4218   ( 495 kB)
>       .text : 0xc0100000 - 0xc027840f   (1505 kB)
> Checking if this processor honours the WP bit even in supervisor mode...Ok.
> 1 multicall(s) failed: cpu 0
> Pid: 0, comm: swapper Tainted: G        W 2.6.27-rc1-xenU #4
>  [<c0103afa>] xen_mc_flush+0x12a/0x190
>  [<c0104a3d>] xen_set_pud_hyper+0x8d/0x90
>  [<c0115965>] zap_low_mappings+0x55/0x80
>  [<c02fbcb4>] start_kernel+0x1d4/0x2a0
>  [<c02fb670>] unknown_bootoption+0x0/0x1f0
>  [<c02ff16a>] xen_start_kernel+0x58a/0x650
>  =======================
>   call  1/1: op=1 arg=[c1203860] result=-22
> ------------[ cut here ]------------
> kernel BUG at arch/x86/xen/multicalls.c:104!
> invalid opcode: 0000 [#1] SMP
>   

Keir, this looks like another instance of the "Xen won't allow guest to
update level 3" bug, though in this case it's 32-on-32.

Carl, are there any messages on the Xen console when this happens ("xm
dmesg")?

This looks like the same bug:
https://bugzilla.redhat.com/show_bug.cgi?id=459067

    J