Bug#503821: linux-image-2.6.26-1-xen-amd64: Kernel crash in Dom0 (Eeek! page_mapcount(page) went negative! (-1))
Package: linux-image-2.6.26-1-xen-amd64
Version: 2.6.26-9
Severity: grave
Justification: renders package unusable
I'm trying to run Xen in Lenny with the new 2.6.26-kernel as Dom0.
It seems to be some problems on the amd64 architecture.
After approx 10-20 hours of uptime, the Dom0 crash (even if there is no
DomU running) with the following error:
[50079.669383] Eeek! page_mapcount(page) went negative! (-1)
[50079.669383] page pfn = 5
[50079.669383] page->flags = 0
[50079.669383] page->count = 0
[50079.669383] page->mapping = 0000000000000000
[50079.669383] vma->vm_ops = 0x0
[50079.669383] ------------[ cut here ]------------
[50079.669383] kernel BUG at mm/rmap.c:673!
[50079.669383] invalid opcode: 0000 [1] SMP
[50079.669383] CPU 0
[50079.669383] Modules linked in: xt_tcpudp xt_physdev iptable_filter ip_tables x_tables bridge netloop video output ac battery microcode firmware_class nfsd auth_rpcgss exportfs nfs lockd nfs_acl sunrpc ipv6 xfs reiserfs ext2 sha256_generic aes_x86_64 aes_generic cbc dm_crypt crypto_blkcipher raid456 async_xor async_memcpy async_tx xor loop iTCO_wdt serio_raw i2c_i801 psmouse pcspkr i2c_core rng_core container button i3000_edac edac_core shpchp pci_hotplug evdev joydev ext3 jbd mbcache dm_mirror dm_log dm_snapshot dm_mod raid1 md_mod ide_cd_mod cdrom ide_pci_generic usbhid hid ff_memless piix ide_core ata_piix sd_mod floppy ata_generic ehci_hcd uhci_hcd sata_sil24 e1000e libata dock 3w_9xxx scsi_mod thermal processor fan thermal_sys
[50079.669383] Pid: 9197, comm: mutt Not tainted 2.6.26-1-xen-amd64 #1
[50079.669383] RIP: e030:[<ffffffff8027c550>] [<ffffffff8027c550>] page_remove_rmap+0xfb/0x117
[50079.669383] RSP: e02b:ffff880074601dc8 EFLAGS: 00010246
[50079.669383] RAX: 0000000000000000 RBX: ffff880002359118 RCX: 000051510001509d
[50079.669383] RDX: ffffffffff5f7000 RSI: 0000000000000001 RDI: ffffffff805aaab0
[50079.669383] RBP: ffff8800746d1918 R08: 0000000000000023 R09: ffff880074601800
[50079.669383] R10: 0000000000000000 R11: 0000014221337ed7 R12: ffff880002359118
[50079.669383] R13: ffff880014e61320 R14: ffff88007ff34b80 R15: ffff8800027eb548
[50079.669383] FS: 00007f2b26ff4700(0000) GS:ffffffff80539000(0000) knlGS:0000000000000000
[50079.669383] CS: e033 DS: 0000 ES: 0000
[50079.669383] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
[50079.669383] DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400
[50079.669383] Process mutt (pid: 9197, threadinfo ffff880074600000, task ffff8800745929f0)
[50079.669383] Stack: ffff880014e610e8 0000000000005100 0000000005e64000 ffffffff80273239
[50079.669383] ffff88007475c000 0000000000000000 ffff880074601ec8 ffffffffffffffff
[50079.669383] 0000000000000000 ffff8800746d1918 ffff880074601ed0 00000000003b9000
[50079.669383] Call Trace:
[50079.669383] [<ffffffff80273239>] ? unmap_vmas+0x744/0xa49
[50079.669383] [<ffffffff80278567>] ? exit_mmap+0x7b/0xf7
[50079.669383] [<ffffffff8022a73d>] ? mmput+0x2c/0xc0
[50079.669383] [<ffffffff8022fef8>] ? do_exit+0x25a/0x6ce
[50079.669383] [<ffffffff80230412>] ? do_group_exit+0xa6/0xdc
[50079.669383] [<ffffffff8020b528>] ? system_call+0x68/0x6d
[50079.669383] [<ffffffff8020b4c0>] ? system_call+0x0/0x6d
[50079.669383]
[50079.669383]
[50079.669383] Code: 80 e8 18 0c fd ff 48 8b 85 90 00 00 00 48 85 c0 74 19 48 8b 40 20 48 85 c0 74 10 48 8b 70 58 48 c7 c7 e1 52 4b 80 e8 f3 0b fd ff <0f> 0b eb fe 8b 77 18 41 58 5b 5d 83 e6 01 f7 de 83 c6 04 e9 64
[50079.669383] RIP [<ffffffff8027c550>] page_remove_rmap+0xfb/0x117
[50079.669383] RSP <ffff880074601dc8>
[50079.673388] ---[ end trace c445527cbda75056 ]---
[50079.673479] Fixing recursive fault but reboot is needed!
The same system is stable when running linux-image-2.6.26-1-amd64.
I've also successfully ran linux-image-2.6.26-1-xen-686 using
i386-architecture on the same hardware, so it seems to happen only in Xen-variants
on amd64 architecture.
I have no reliable way of forcing the error to occur. The best way I've found
so far to get this crash is to do a couple of kernel-recompilations, but sometimes
I can do a couple of compile-runs without crash.
The hardware on this box is a Supermicro PDSME+-motherboard, with a E6600 Core2Duo
and 8 Gigs of RAM (ECC)
I've run memtest86+ for over 24 hours with no problems reported.
I'm not sure if I should tag this as grave or critical, but I feel that it's impossible
to run a Xen-system on AMD64 with Debian Lenny currently.
-- System Information:
Debian Release: lenny/sid
APT prefers testing
APT policy: (500, 'testing')
Architecture: amd64 (x86_64)
Kernel: Linux 2.6.26-1-xen-amd64 (SMP w/2 CPU cores)
Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8)
Shell: /bin/sh linked to /bin/bash
Versions of packages linux-image-2.6.26-1-xen-amd64 depends on:
ii initramfs-tools 0.92j tools for generating an initramfs
ii linux-modules-2.6.26-1-xen-am 2.6.26-9 Linux 2.6.26 modules on AMD64
linux-image-2.6.26-1-xen-amd64 recommends no packages.
Versions of packages linux-image-2.6.26-1-xen-amd64 suggests:
ii grub 0.97-47 GRand Unified Bootloader (Legacy v
ii linux-doc-2.6.26 2.6.26-8 Linux kernel specific documentatio
-- no debconf information
Reply to: