From Bugzilla Helper: User-Agent: Mozilla/4.78 [en] (X11; U; Linux 2.4.9-31 i686; Nav) Description of problem: Crash when attempting LVM snapshot using IBM ServeRAID with ips driver. Version-Release number of selected component (if applicable): How reproducible: Always Steps to Reproduce: 1. Boot. 2. Create LV and filesystem. 3. Write data. 4. Try to take snapshot (lvcreate -s) Actual Results: invalid operand: 0000 Unable to handle kernel paging request at virtual address 37aedeee printing eip: c0131036 *pde = 00000000 Expected Results: Not crash. Additional info: IBM x350 with 4 Xeon 700MHz, 512M RAM. LVM is on a ServeRAID volume, which consists of a single 18G drive (no actual RAID running).
What kernel version? And could you please provide the full kernel oops trace?
Update from original sender: From: Thornton Prime <thornton> Subject: Re: [Bug 63527] Changed - Crash During LVM Snapshot > Kernel was 2.4.18-0.13 SMP/i686. > > Unfortunately, that was as much of a trace as I could get off the screen > (it always seems to stop the output right there). kernel 2.4.18-0.22 debug/i386 gave me a good trace: kernel BUG at vmalloc.c:246! invalid operand: 0000 eepro100 loop usb-ohci usbcore ext3 jbd lvm-mod ips sd_mod scsi_mod CPU: 0 EIP: 0010:[__vmalloc+54/496] Not tainted EIP: 0010:[<c0130886>] Not tainted EFLAGS: 00010282 EIP is at __vmalloc [kernel] 0x36 (2.4.18-0.22debug) eax: 0000001d ebx: 00000000 ecx: 00000001 edx: 000022c1 esi: 00000000 edi: f6a0f200 ebp: fffffff4 esp: f69d5cd0 ds: 0018 es: 0018 ss: 0018 Process lvcreate (pid: 760, stackpage=f69d5000) Stack: c0238edc 000000f6 c02dd07c c02dd07c c02dd238 00000000 00000500 c0135665 c01516f1 00000001 00000000 00000000 f6a0f200 fffffff4 f88426a5 00000000 000001f2 00000163 00000000 f6a0f200 f6a0f36c f6a0f200 f884275c f6a0f200 Call Trace: [__alloc_pages+117/768] __alloc_pages [kernel] 0x75 Call Trace: [<c0135665>] __alloc_pages [kernel] 0x75 [alloc_kiobuf_bhs+33/128] alloc_kiobuf_bhs [kernel] 0x21 [<c01516f1>] alloc_kiobuf_bhs [kernel] 0x21 [eepro100:__insmod_eepro100_O/lib/modules/2.4.18-0.22debug/kernel/dri+-1575259/96] lvm_snapshot_alloc_hash_table [lvm-mod] 0x45 [<f88426a5>] lvm_snapshot_alloc_hash_table [lvm-mod] 0x45 [eepro100:__insmod_eepro100_O/lib/modules/2.4.18-0.22debug/kernel/dri+-1575076/96] lvm_snapshot_alloc [lvm-mod] 0x6c [<f884275c>] lvm_snapshot_alloc [lvm-mod] 0x6c [eepro100:__insmod_eepro100_O/lib/modules/2.4.18-0.22debug/kernel/dri+-1584425/96] lvm_do_lv_create [lvm-mod] 0x517 [<f88402d7>] lvm_do_lv_create [lvm-mod] 0x517 [eepro100:__insmod_eepro100_O/lib/modules/2.4.18-0.22debug/kernel/dri+-1594763/96] lvm_chr_ioctl [lvm-mod] 0x715 [<f883da75>] lvm_chr_ioctl [lvm-mod] 0x715 [eepro100:__insmod_eepro100_O/lib/modules/2.4.18-0.22debug/kernel/dri+-1555168/96] lv_req [lvm-mod] 0x0 [<f8847520>] lv_req [lvm-mod] 0x0 [locate_hd_struct+39/112] locate_hd_struct [kernel] 0x27 [<c018ba97>] locate_hd_struct [kernel] 0x27 [req_new_io+53/96] req_new_io [kernel] 0x35 [<c018bbe5>] req_new_io [kernel] 0x35 [kmem_cache_free+536/688] kmem_cache_free [kernel] 0x218 [<c0131fe8>] kmem_cache_free [kernel] 0x218 [sys_ioctl+535/560] sys_ioctl [kernel] 0x217 [<c014a3e7>] sys_ioctl [kernel] 0x217 [system_call+51/56] system_call [kernel] 0x33 [<c0108923>] system_call [kernel] 0x33 Code: 0f 0b 5d 58 31 c0 e9 a3 01 00 00 6a 02 53 e8 57 fe ff ff 5e
Ugh, there's clearly an overflow possibility there. We've checked in a fix for the most obvious problem causing the oops. What size of snapshot were you trying to create?
On Tue, Apr 16, 2002 at 08:56:33AM -0700, Thornton Prime wrote: > 100M on a 18G volume (with 3G free), I believe. I actually tried a few > different sizes in the same range and all oopsed.
We've got a fix for the oops in our current tree. I can reproduce the problem with old kernels, and the current kernel seems to work OK.
I'm not sure id 2.4.18-0.26smp is supposed to have the fix, but I got another oops with the latest rawhide trying a 'lvcreate -s -n snap -L 100M /dev/vg01/lvol1'. pr 16 10:19:36 abe kernel: kernel BUG at vmalloc.c:246! Apr 16 10:19:36 abe kernel: invalid operand: 0000 Apr 16 10:19:36 abe kernel: sg e100 loop usb-ohci usbcore ext3 jbd lvm-mod ips sd_mod scsi_mod Apr 16 10:19:36 abe kernel: CPU: 0 Apr 16 10:19:36 abe kernel: EIP: 0010:[__vmalloc+54/528] Not tainted Apr 16 10:19:36 abe kernel: EIP: 0010:[<c0136626>] Not tainted Apr 16 10:19:36 abe kernel: EFLAGS: 00010282 Apr 16 10:19:36 abe kernel: Apr 16 10:19:36 abe kernel: EIP is at __vmalloc [kernel] 0x36 (2.4.18-0.26smp) Apr 16 10:19:36 abe kernel: eax: 0000001d ebx: 00000000 ecx: c02fde80 edx: 00003ca4 Apr 16 10:19:36 abe kernel: esi: 00000000 edi: f6a99e00 ebp: fffffff4 esp: f6a5fcd0 Apr 16 10:19:36 abe kernel: ds: 0018 es: 0018 ss: 0018 Apr 16 10:19:36 abe kernel: Process lvcreate (pid: 805, stackpage=f6a5f000) Apr 16 10:19:36 abe kernel: Stack: c025399c 000000f6 00000000 c02ff37c c02ff37c c02ff538 00000000 00000500 Apr 16 10:19:36 abe kernel: c013b702 00000001 00000000 00000000 f6a99e00 fffffff4 f8843735 00000000 Apr 16 10:19:36 abe kernel: 000001f2 00000163 00000000 f6a99e00 f6a99f6c f6a99e00 f88437ec f6a99e00 Apr 16 10:19:36 abe kernel: Call Trace: [__alloc_pages+114/736] __alloc_pages [kernel] 0x72 Apr 16 10:19:36 abe kernel: Call Trace: [<c013b702>] __alloc_pages [kernel] 0x72 Apr 16 10:19:36 abe kernel: [e100:__insmod_e100_O/lib/modules/2.4.18-0.26smp/kernel/drivers/a+-1591499/96] lvm_snapshot_alloc_hash_table [lvm-mod] 0x45 Apr 16 10:19:36 abe kernel: [<f8843735>] lvm_snapshot_alloc_hash_table [lvm-mod] 0x45 Apr 16 10:19:36 abe kernel: [e100:__insmod_e100_O/lib/modules/2.4.18-0.26smp/kernel/drivers/a+-1591316/96] lvm_snapshot_alloc [lvm-mod] 0x6c Apr 16 10:19:36 abe kernel: [<f88437ec>] lvm_snapshot_alloc [lvm-mod] 0x6c Apr 16 10:19:36 abe kernel: [e100:__insmod_e100_O/lib/modules/2.4.18-0.26smp/kernel/drivers/a+-1600713/96] lvm_do_lv_create [lvm-mod] 0x527 Apr 16 10:19:36 abe kernel: [<f8841337>] lvm_do_lv_create [lvm-mod] 0x527 Apr 16 10:19:36 abe kernel: [e100:__insmod_e100_O/lib/modules/2.4.18-0.26smp/kernel/drivers/a+-1611115/96] lvm_chr_ioctl [lvm-mod] 0x715 Apr 16 10:19:36 abe kernel: [<f883ea95>] lvm_chr_ioctl [lvm-mod] 0x715 Apr 16 10:19:36 abe kernel: [e100:__insmod_e100_O/lib/modules/2.4.18-0.26smp/kernel/drivers/a+-1571008/96] lv_req [lvm-mod] 0x0 Apr 16 10:19:36 abe kernel: [<f8848740>] lv_req [lvm-mod] 0x0 Apr 16 10:19:37 abe kernel: [e100:__insmod_e100_O/lib/modules/2.4.18-0.26smp/kernel/drivers/a+-1660385/96] ips_next [ips] 0x6ef Apr 16 10:19:37 abe kernel: [<f8832a1f>] ips_next [ips] 0x6ef Apr 16 10:19:37 abe kernel: [account_io_start+68/80] account_io_start [kernel] 0x44 Apr 16 10:19:37 abe kernel: [<c0199b24>] account_io_start [kernel] 0x44 Apr 16 10:19:37 abe kernel: [locate_hd_struct+39/112] locate_hd_struct [kernel] 0x27 Apr 16 10:19:37 abe kernel: [<c0199a67>] locate_hd_struct [kernel] 0x27 Apr 16 10:19:37 abe kernel: [sys_ioctl+599/670] sys_ioctl [kernel] 0x257 Apr 16 10:19:37 abe kernel: [<c0153be7>] sys_ioctl [kernel] 0x257 Apr 16 10:19:37 abe kernel: [system_call+51/56] system_call [kernel] 0x33 Apr 16 10:19:37 abe kernel: [<c0108c6b>] system_call [kernel] 0x33 Apr 16 10:19:37 abe kernel: Apr 16 10:19:37 abe kernel: Apr 16 10:19:37 abe kernel: Code: 0f 0b 5f 5d 31 c0 e9 c7 01 00 00 6a 02 53 e8 e7 fd ff ff 59
No, it was fixed after that build.