paport related OOPS since 2.6.11-bk7
The OOPS below gets generated consistently when FC3 kudzu is run during boot (tested between 2.6.11-bk7 and 2.6.11.6-bk1). It seems to be caused by the hotplug-parport changeset:
http://linux.bkbits.net:8080/linux-2.5/cset@4230791b6YtcIhZDSvvWbzSdUpg2zg?nav=index.html|ChangeSet@-4w
(reverting this changeset eliminates the oops).
parport_pc: VIA 686A/8231 detectedparport_pc: probing current configurationparport_pc: Current parallel port base: 0x378
parport0: PC-style at 0x378, irq 7 [PCSPP,EPP]parport_pc: VIA parallel port: io=0x378, irq=7
Unable to handle kernel NULL pointer dereference at virtual address 00000000 printing eip: d3f37368 *pde = 00000000 Oops: 0000 [#1] PREEMPT Modules linked in: parport_pc parport binfmt_misc nls_cp437 nls_iso8859_1 ntfs video thermal processor fan container button battery ac uhci_hcd usbcore 3c59x mii CPU: 0 EIP: 0060:[<d3f37368>] Not tainted VLI EFLAGS: 00010286 (2.6.12-rc1)
EIP is at parport_pc_pci_remove+0x18/0x40 [parport_pc]
eax: c12ef844 ebx: c12ef800 ecx: c12ef844 edx: d3f37350
esi: 00000000 edi: d3f3cc68 ebp: ce53aecc esp: ce53aec0
ds: 007b es: 007b ss: 0068
Process modprobe (pid: 3187, threadinfo=ce53a000 task=cfbe2570)
Stack: c018f481 c12ef800 c12ef844 ce53aedc c0239636 c12ef800 c12ef868 ce53aef8
c02fe67c c12ef844 c040412e d3f3ccb0 d3f3ccb0 d3f375a0 ce53af0c
c02fe6a0
c12ef844 d3f3cc68 00000000 ce53af24 c02feb9a d3f3cc68 d3f3cc68
d3f3cc68
Call Trace:
[<c0103ebf>] show_stack+0x7f/0xa0
[<c0104056>] show_registers+0x156/0x1d0
[<c010427a>] die+0xea/0x180
[<c01149a2>] do_page_fault+0x482/0x6ba
[<c0103b23>] error_code+0x2b/0x30
[<c0239636>] pci_device_remove+0x36/0x40
[<c02fe67c>] device_release_driver+0x7c/0x80
[<c02fe6a0>] driver_detach+0x20/0x30
[<c02feb9a>] bus_remove_driver+0x4a/0x90
[<c02ff162>] driver_unregister+0x12/0x20
[<c0239875>] pci_unregister_driver+0x15/0x20
[<d3f3763e>] parport_pc_exit+0x9e/0xae [parport_pc]
[<c0133fce>] sys_delete_module+0x17e/0x1b0
[<c010303b>] sysenter_past_esp+0x54/0x75
Code: 44 24 04 89 0c 24 ff d2 eb 94 89 f6 8d bc 27 00 00 00 00 55 89 e5 56 53 83 ec 04 8b 45 08 83 c0 44 8b 70 74 c7 40 74 00 00 00 00 <8b> 1e eb 10 8d 74 26 00 8b 44 9e 04 89 04 24 e8 d4 f4 ff ff 4bSignature · 3 lines
Eran Mann MRV International www.mrv.com
Quoted 5 lines · Eran Mann
The OOPS below gets generated consistently when FC3 kudzu is run during boot (tested between 2.6.11-bk7 and 2.6.11.6-bk1). It seems to be caused by the hotplug-parport changeset:
http://linux.bkbits.net:8080/linux-2.5/cset@4230791b6YtcIhZDSvvWbzSdUpg2zg?nav=index.html|ChangeSet@-4w
(reverting this changeset eliminates the oops).
Please try this instead.
It appears that the parport driver claims on-board superio devices without actually doing anything. When the driver is removed, we try to dereference non-existent driver data to unregister the ports. Since we didn't register anything, it's safe to ignore these devices in the remove function.
drivers/parport/parport_pc.c +5 -3
@@ -2976,10 +2976,12 @@ static void __devexit parport_pc_pci_rempci_set_drvdata (dev,NULL ); -for (i = data->num -1 ; i >=0 ; i--) -parport_pc_unregister_port (data->ports[i]); +if (data) { +for (i = data->num -1 ; i >=0 ; i--) +parport_pc_unregister_port (data->ports[i]); -kfree (data); +kfree (data); + } }static struct pci_driver parport_pc_pci_driver = {
Signature · 3 lines
Russell King Linux kernel 2.6 ARM Linux - http://www.arm.linux.org.uk/ maintainer of: 2.6 Serial core
Quoted 27 lines · Russell King
Quoted 3 lines · Eran Mann
The OOPS below gets generated consistently when FC3 kudzu is run during boot (tested between 2.6.11-bk7 and 2.6.11.6-bk1). It seems to be caused by the hotplug-parport changeset:
http://linux.bkbits.net:8080/linux-2.5/cset@4230791b6YtcIhZDSvvWbzSdUpg2zg?nav=index.html|ChangeSet@-4w
(reverting this changeset eliminates the oops).Please try this instead.
It appears that the parport driver claims on-board superio devices without actually doing anything. When the driver is removed, we try to dereference non-existent driver data to unregister the ports. Since we didn't register anything, it's safe to ignore these devices in the remove function.
Signed-off-by Russell King <[email protected]>diff -up -x BitKeeper -x ChangeSet -x SCCS -x
_xlk-x *.orig -x *.rejorig/drivers/parport/parport_pc.clinux/drivers/parport/parport_pc.cdrivers/parport/parport_pc.c +0
-2@@ -2976,10 +2976,12 @@ static void __devexit parport_pc_pci_rempci_set_drvdata (dev,NULL ); -for (i = data->num -1 ; i >=0 ; i--) -parport_pc_unregister_port (data->ports[i]);
drivers/parport/parport_pc.c static void __devexit parport_pc_pci_rem · line 2987
Quoted 10 lines
drivers/parport/parport_pc.c +5
-1+if (data) { +for (i = data->num -1 ; i >=0 ; i--) +parport_pc_unregister_port (data->ports[i]); -kfree (data); +kfree (data); + } }static struct pci_driver parport_pc_pci_driver = {
Doesn't seem to help. Do you want me to check anything else?
Mar 28 17:13:44 eran kernel: parport0: PC-style at 0x378, irq 7 [PCSPP,EPP] Mar 28 17:13:44 eran kernel: parport_pc: VIA parallel port: io=0x378, irq=7 Mar 28 17:13:44 eran kernel: Unable to handle kernel NULL pointer dereference at virtual address 00000000 Mar 28 17:13:44 eran kernel: printing eip: Mar 28 17:13:44 eran kernel: d3f37368 Mar 28 17:13:44 eran kernel: *pde = 00000000 Mar 28 17:13:44 eran kernel: Oops: 0000 [#1] Mar 28 17:13:44 eran kernel: PREEMPT Mar 28 17:13:44 eran kernel: Modules linked in: parport_pc parport binfmt_misc nls_cp437 nls_iso8859_1 ntfs video thermal processor fan container button battery ac uhci_hcd usbcore 3c59x mii Mar 28 17:13:44 eran kernel: CPU: 0 Mar 28 17:13:44 eran kernel: EIP: 0060:[<d3f37368>] Not tainted VLI Mar 28 17:13:44 eran kernel: EFLAGS: 00010286 (2.6.11-bk6-patched) Mar 28 17:13:44 eran kernel: EIP is at parport_pc_pci_remove+0x18/0x40 [parport_pc] Mar 28 17:13:44 eran kernel: eax: c12ef844 ebx: c12ef800 ecx: c12ef844 edx: d3f37350 Mar 28 17:13:44 eran kernel: esi: 00000000 edi: d3f3cd48 ebp: ce585ecc esp: ce585ec0 Mar 28 17:13:44 eran kernel: ds: 007b es: 007b ss: 0068 Mar 28 17:13:44 eran kernel: Process modprobe (pid: 3503, threadinfo=ce585000 task=cffdf040) Mar 28 17:13:44 eran kernel: Stack: c0190441 c12ef800 c12ef844 ce585edc c0237c06 c12ef800 c12ef868 ce585ef8 Mar 28 17:13:44 eran kernel: c02f9c5c c12ef844 c03ff1cd d3f3cd90 d3f3cd90 d3f375a0 ce585f0c c02f9c80 Mar 28 17:13:44 eran kernel: c12ef844 d3f3cd48 00000000 ce585f24 c02fa17a d3f3cd48 d3f3cd48 d3f3cd48 Mar 28 17:13:44 eran kernel: Call Trace: Mar 28 17:13:44 eran kernel: [<c0103ecf>] show_stack+0x7f/0xa0 Mar 28 17:13:44 eran kernel: [<c0104066>] show_registers+0x156/0x1d0 Mar 28 17:13:44 eran kernel: [<c010428a>] die+0xea/0x180 Mar 28 17:13:44 eran kernel: [<c0114e22>] do_page_fault+0x482/0x6ba Mar 28 17:13:44 eran kernel: [<c0103b33>] error_code+0x2b/0x30 Mar 28 17:13:44 eran kernel: [<c0237c06>] pci_device_remove+0x36/0x40 Mar 28 17:13:44 eran kernel: [<c02f9c5c>] device_release_driver+0x7c/0x80 Mar 28 17:13:44 eran kernel: [<c02f9c80>] driver_detach+0x20/0x30 Mar 28 17:13:44 eran kernel: [<c02fa17a>] bus_remove_driver+0x4a/0x90 Mar 28 17:13:44 eran kernel: [<c02fa742>] driver_unregister+0x12/0x20 Mar 28 17:13:44 eran kernel: [<c0237e45>] pci_unregister_driver+0x15/0x20 Mar 28 17:13:44 eran kernel: [<d3f3763e>] parport_pc_exit+0x9e/0xae [parport_pc] Mar 28 17:13:44 eran kernel: [<c01343fe>] sys_delete_module+0x17e/0x1b0 Mar 28 17:13:44 eran kernel: [<c010304b>] sysenter_past_esp+0x54/0x75 Mar 28 17:13:44 eran kernel: Code: 44 24 04 89 0c 24 ff d2 eb 94 89 f6 8d bc 27 00 00 00 00 55 89 e5 56 53 83 ec 04 8b 45 08 83 c0 44 8b 70 74 c7 40 74 00 00 00 00 <8b> 1e eb 10 8d 74 26 00 8b 44 9e 04 89 04 24 e8 d4 f4 ff ff 4b
Signature · 6 lines
Eran Mann Senior Software Engineer MRV International Tel: 972-4-9936297 Fax: 972-4-9890430 www.mrv.com
Quoted 1 line · Eran Mann
Mar 28 17:13:44 eran kernel: EIP is at parport_pc_pci_remove+0x18/0x40 [parport_pc]
Are you sure you reinstalled the kernel modules after rebuilding? parport_pc_pci_remove() was 0x40 bytes long before, and still seems to be. Also, the Code: lines are identical. This means you're running the same (unpatched) code as you were before, so you aren't actually running my fix.
Can you double-check please?
Signature · 3 lines
Russell King Linux kernel 2.6 ARM Linux - http://www.arm.linux.org.uk/ maintainer of: 2.6 Serial core
Quoted 3 lines · Russell King
Can you double-check please?
I'm sorry, I booted the wrong image.
The fix does work (tested against 2.6.11.6-bk1).