* 2.6.0-test5-mm4 Oops
@ 2003-09-26 18:49 Eric Balsa
2003-09-27 9:27 ` [PATCH][2.6] fix atp870u boot oops Zwane Mwaikambo
0 siblings, 1 reply; 4+ messages in thread
From: Eric Balsa @ 2003-09-26 18:49 UTC (permalink / raw)
To: linux-kernel
Hello,
I am seeing a kernel oops when booting 2.6.0-test5-mm4. Below is the oops.
Looks like it is from the ACARD driver, since when I disable it in the
kernel, the oops no longer happens. This is a dual Xeon 866 (Dell Poweredge
4400) with an SMP kernel. I hope the ksymoops is useful.
Regards,
--Eric
Preempt SMP
CPU: 0
EIP: 0060:[<02293d33>] Not tainted VLI
EFLAGS: 00010202
EIP is at atp870u_detect+0x18b/0x9c8
eax: 00000000 ebx: 023e8cd4 ecx: 41fa5f7c edx: 000006f2
esi: 00000000 edi: 00000000 ebp: 0000000e esp: 41fa5c08
ds: 007b es: 007b ss:0068
Process Swapper (pid: 1,threadinfo=41fa4000 task=03a57900)
Stack: 023e8cd4 023e8c60 00000000 00000000 00000001 00000100 00000005
41fa5f7c
55789004 00000002 00000001 00000100 00000006 00060020 21789004 00000003
00804001 00000101 00000007 00060020 08389004 00000000 00804001 41fa5cb2
Call Trace:
[<0243a1bb>] init_this_scsi_driver+0x4b/0xe0
[<0242280d>] do_initcalls+0x3d/0x8c
[<02422875>] do_basic_setup+0x19/0x24
[<021070ef>] init+0x5b/0x15c
[<02107094>] init+0x0/0x15c
[<0210a991>] kernel_thread_helper+0x5/0xc
Code: 60 8d 8c 24 aa 00 00 00 89 4c 24 5c 8b 4c 24 1c 0f b6 44 24 72 8d 3c
85 0000 00 00 8b 14 0f 89 44 24 54 85 d2 0f 84 4d 08 00 00 <8b> aa fc 00 00
00 8b 74 24 54 8a 9a f4 00 00 00 01 f6 66 81 bc
<0> Kernel Panic: Attempted to kill init!
Output from ksymoops:
ksymoops 2.4.5 on i686 2.6.0-test5-mm4. Options used
-v /root/linux-2.6.0-mm4 (specified)
-k /proc/ksyms (default)
-l /proc/modules (default)
-o /lib/modules/2.6.0-test5-mm4/ (default)
-m ./System.map (specified)
Error (pclose_local): read_nm_symbols pclose failed 0x100
Warning (read_vmlinux): no kernel symbols in vmlinux, is
/root/linux-2.6.0-mm4 a valid vmlinux file?
Error (regular_file): read_ksyms stat /proc/ksyms failed
No modules in ksyms, skipping objects
No ksyms, skipping lsmod
CPU: 0
EIP: 0060:[<02293d33>] Not tainted VLI
Using defaults from ksymoops -t elf32-i386 -a i386
EFLAGS: 00010202
eax: 00000000 ebx: 023e8cd4 ecx: 41fa5f7c edx: 000006f2
esi: 00000000 edi: 00000000 ebp: 0000000e esp: 41fa5c08
ds: 007b es: 007b ss:0068
Stack: 023e8cd4 023e8c60 00000000 00000000 00000001 00000100 00000005
41fa5f7c
55789004 00000002 00000001 00000100 00000006 00060020 21789004 00000003
00804001 00000101 00000007 00060020 08389004 00000000 00804001 41fa5cb2
[<0243a1bb>] init_this_scsi_driver+0x4b/0xe0
[<0242280d>] do_initcalls+0x3d/0x8c
[<02422875>] do_basic_setup+0x19/0x24
[<021070ef>] init+0x5b/0x15c
[<02107094>] init+0x0/0x15c
[<0210a991>] kernel_thread_helper+0x5/0xc
Code: 60 8d 8c 24 aa 00 00 00 89 4c 24 5c 8b 4c 24 1c 0f b6 44 24 72 8d 3c
85 0000 00 00 8b 14 0f 89 44 24 54 85 d2 0f 84 4d 08 00 00 <8b> aa fc 00 00
00 8b 74 24 54 8a 9a f4 00 00 00 01 f6 66 81 bc
>>EIP; 02293d33 <atp870u_detect+18b/9e8> <=====
>>ebx; 023e8cd4 <driver_template+74/80>
>>ecx; 41fa5f7c <_end+3fb1f514/fdb64598>
>>edx; 000006f2 Before first symbol
>>esp; 41fa5c08 <_end+3fb1f1a0/fdb64598>
Code; 02293d08 <atp870u_detect+160/9e8>
00000000 <_EIP>:
Code; 02293d08 <atp870u_detect+160/9e8>
0: 60 pusha
Code; 02293d09 <atp870u_detect+161/9e8>
1: 8d 8c 24 aa 00 00 00 lea 0xaa(%esp,1),%ecx
Code; 02293d10 <atp870u_detect+168/9e8>
8: 89 4c 24 5c mov %ecx,0x5c(%esp,1)
Code; 02293d14 <atp870u_detect+16c/9e8>
c: 8b 4c 24 1c mov 0x1c(%esp,1),%ecx
Code; 02293d18 <atp870u_detect+170/9e8>
10: 0f b6 44 24 72 movzbl 0x72(%esp,1),%eax
Code; 02293d1d <atp870u_detect+175/9e8>
15: 8d 3c 85 00 00 00 00 lea 0x0(,%eax,4),%edi
Code; 02293d24 <atp870u_detect+17c/9e8>
1c: 8b 14 0f mov (%edi,%ecx,1),%edx
Code; 02293d27 <atp870u_detect+17f/9e8>
1f: 89 44 24 54 mov %eax,0x54(%esp,1)
Code; 02293d2b <atp870u_detect+183/9e8>
23: 85 d2 test %edx,%edx
Code; 02293d2d <atp870u_detect+185/9e8>
25: 0f 84 4d 08 00 00 je 878 <_EIP+0x878> 02294580
<atp870u_detect+9d8/9e8>
Code; 02293d33 <atp870u_detect+18b/9e8> <=====
2b: 8b aa fc 00 00 00 mov 0xfc(%edx),%ebp <=====
Code; 02293d39 <atp870u_detect+191/9e8>
31: 8b 74 24 54 mov 0x54(%esp,1),%esi
Code; 02293d3d <atp870u_detect+195/9e8>
35: 8a 9a f4 00 00 00 mov 0xf4(%edx),%bl
Code; 02293d43 <atp870u_detect+19b/9e8>
3b: 01 f6 add %esi,%esi
Code; 02293d45 <atp870u_detect+19d/9e8>
3d: 66 data16
Code; 02293d46 <atp870u_detect+19e/9e8>
3e: 81 .byte 0x81
Code; 02293d47 <atp870u_detect+19f/9e8>
3f: bc .byte 0xbc
<0> Kernel Panic: Attempted to kill init!
1 warning and 2 errors issued. Results may not be reliable.
^ permalink raw reply [flat|nested] 4+ messages in thread
* [PATCH][2.6] fix atp870u boot oops
2003-09-26 18:49 2.6.0-test5-mm4 Oops Eric Balsa
@ 2003-09-27 9:27 ` Zwane Mwaikambo
2003-09-27 13:25 ` James Bottomley
0 siblings, 1 reply; 4+ messages in thread
From: Zwane Mwaikambo @ 2003-09-27 9:27 UTC (permalink / raw)
To: Linux Kernel; +Cc: James Bottomley, Eric Balsa
The driver's probe function accesses uninitialised data. I also made it
use pci_get_device instead of pci_find_device to bump the refcount on the
pci devices it finds.
> Preempt SMP
> CPU: 0
> EIP: 0060:[<02293d33>] Not tainted VLI
> EFLAGS: 00010202
> EIP is at atp870u_detect+0x18b/0x9c8
> eax: 00000000 ebx: 023e8cd4 ecx: 41fa5f7c edx: 000006f2
> esi: 00000000 edi: 00000000 ebp: 0000000e esp: 41fa5c08
> ds: 007b es: 007b ss:0068
> Process Swapper (pid: 1,threadinfo=41fa4000 task=03a57900)
> Stack: 023e8cd4 023e8c60 00000000 00000000 00000001 00000100 00000005
> 41fa5f7c
> 55789004 00000002 00000001 00000100 00000006 00060020 21789004 00000003
> 00804001 00000101 00000007 00060020 08389004 00000000 00804001 41fa5cb2
>
> Call Trace:
> [<0243a1bb>] init_this_scsi_driver+0x4b/0xe0
> [<0242280d>] do_initcalls+0x3d/0x8c
> [<02422875>] do_basic_setup+0x19/0x24
> [<021070ef>] init+0x5b/0x15c
> [<02107094>] init+0x0/0x15c
> [<0210a991>] kernel_thread_helper+0x5/0xc
Index: linux-2.6.0-test5-mm4/drivers/scsi/atp870u.c
===================================================================
RCS file: /build/cvsroot/linux-2.6.0-test5-mm4/drivers/scsi/atp870u.c,v
retrieving revision 1.1.1.1
diff -u -p -B -r1.1.1.1 atp870u.c
--- linux-2.6.0-test5-mm4/drivers/scsi/atp870u.c 25 Sep 2003 02:36:18 -0000 1.1.1.1
+++ linux-2.6.0-test5-mm4/drivers/scsi/atp870u.c 27 Sep 2003 08:42:09 -0000
@@ -2290,7 +2290,7 @@ static int atp870u_detect(Scsi_Host_Temp
unsigned char irq, h, k, m;
unsigned long flags;
unsigned int base_io, error, tmport;
- struct pci_dev *pdev[MAX_ATP];
+ struct pci_dev *pdev[MAX_ATP] = { [0 ... MAX_ATP-1] = NULL };
unsigned char chip_ver[MAX_ATP], host_id;
unsigned short dev_id[MAX_ATP], n;
struct Scsi_Host *shpnt = NULL;
@@ -2306,8 +2306,8 @@ static int atp870u_detect(Scsi_Host_Temp
for (h = 0; devid[h]; h++) {
struct pci_dev *dev = NULL;
-
- while((dev = pci_find_device(0x1191, devid[h], dev))!=NULL)
+
+ while((dev = pci_get_device(0x1191, devid[h], dev)))
{
if(pci_enable_device(dev))
continue;
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: [PATCH][2.6] fix atp870u boot oops
2003-09-27 9:27 ` [PATCH][2.6] fix atp870u boot oops Zwane Mwaikambo
@ 2003-09-27 13:25 ` James Bottomley
2003-09-27 15:53 ` Zwane Mwaikambo
0 siblings, 1 reply; 4+ messages in thread
From: James Bottomley @ 2003-09-27 13:25 UTC (permalink / raw)
To: Zwane Mwaikambo; +Cc: Linux Kernel, Eric Balsa
On Sat, 2003-09-27 at 04:27, Zwane Mwaikambo wrote:
> The driver's probe function accesses uninitialised data. I also made it
> use pci_get_device instead of pci_find_device to bump the refcount on the
> pci devices it finds.
I don't suppose you'd like to fix this properly? i.e. convert the
driver to the sysfs based pci probing infrastructure and remove its
dependence on MAX_ATP arrays?
James
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: [PATCH][2.6] fix atp870u boot oops
2003-09-27 13:25 ` James Bottomley
@ 2003-09-27 15:53 ` Zwane Mwaikambo
0 siblings, 0 replies; 4+ messages in thread
From: Zwane Mwaikambo @ 2003-09-27 15:53 UTC (permalink / raw)
To: James Bottomley; +Cc: Linux Kernel, Eric Balsa
On Sat, 27 Sep 2003, James Bottomley wrote:
> On Sat, 2003-09-27 at 04:27, Zwane Mwaikambo wrote:
> > The driver's probe function accesses uninitialised data. I also made it
> > use pci_get_device instead of pci_find_device to bump the refcount on the
> > pci devices it finds.
>
> I don't suppose you'd like to fix this properly? i.e. convert the
> driver to the sysfs based pci probing infrastructure and remove its
> dependence on MAX_ATP arrays?
Sure, i can do that.
^ permalink raw reply [flat|nested] 4+ messages in thread
end of thread, other threads:[~2003-09-27 15:54 UTC | newest]
Thread overview: 4+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2003-09-26 18:49 2.6.0-test5-mm4 Oops Eric Balsa
2003-09-27 9:27 ` [PATCH][2.6] fix atp870u boot oops Zwane Mwaikambo
2003-09-27 13:25 ` James Bottomley
2003-09-27 15:53 ` Zwane Mwaikambo
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).