file: oops4.ksyms.txt: the OOPS that I got.
file: oops4.bt.txt: ksymoops-derived backtrace of OOPS.
file: .config - my .config file.
To verify that this is not a compilation problem, I ran a 'make
mrproper' and compiled from scratch.
Something to note: module support is disabled and I remember that
sometime back you needed to have ALSA compiled as a module, and this
OOPS is in ALSA. Is it still necessary to compile ALSA as a module ?
I am not on the kernel mailing list, (only on linux-net and netdev). So
please email me personally if you need any other information.
thanks,
--
Ranjeet Shetye
Senior Software Engineer
Zultys Technologies
Ranjeet dot Shetye2 at Zultys dot com
http://www.zultys.com/
The views, opinions, and judgements expressed in this message are solely
those of the author. The message contents have not been reviewed or
approved by Zultys.
Ranjeet Shetye <[email protected]> wrote:
>
> Unable to handle kernel paging request at virtual address ffffffef
> printing eip:
> c027184c
> *pde = 00001067
> *pte = 00000000
> Oops: 0000 [#1]
> CPU: 0
> EIP: 0060:[<c027184c>] Not tainted
> EFLAGS: 00010282
> EIP is at atomic_dec_and_lock+0x8/0x54
> eax: ffffffef ebx: ffffffef ecx: ffffffef edx: cf372254
> esi: ffffffef edi: cf743e20 ebp: c12efea8 esp: c12efea0
> ds: 007b es: 007b ss: 0068
> Process swapper (pid: 1, threadinfo=c12ee000 task=c12ed8c0)
> Stack: ffffffef ffffffef c12efec4 c01675e4 ffffffef c05a21b0 ffffffef cf7469b4
> cf743e20 c12efee4 c018aa52 ffffffef 000041ed c018a998 c05f0f80 00000000
> c05f0f80 c12eff00 c018aab1 c05f0f80 cf743e20 c05f0f84 00000000 c05f0f80
> Call Trace:
> [<c01675e4>] dput+0x24/0x227
> [<c018aa52>] create_dir+0x9e/0xa4
> [<c018a998>] init_dir+0x0/0x1c
> [<c018aab1>] sysfs_create_dir+0x36/0x6c
> [<c026eee0>] create_dir+0x1f/0x49
> [<c026f331>] kobject_add+0x4d/0x124
Bug in fs/sysfs/dir.c:create_dir() - sysfs_create() returned -EEXIST and we
turned that into a pointer and did a dput() on it.
Something like this should fix it. The -EEXIST return may be another bug?
diff -puN fs/sysfs/dir.c~sysfs-create_dir-oops-fix fs/sysfs/dir.c
--- 25/fs/sysfs/dir.c~sysfs-create_dir-oops-fix Wed Sep 10 15:41:35 2003
+++ 25-akpm/fs/sysfs/dir.c Wed Sep 10 15:44:42 2003
@@ -24,10 +24,11 @@ static int init_dir(struct inode * inode
static struct dentry *
create_dir(struct kobject * k, struct dentry * p, const char * n)
{
- struct dentry * dentry;
+ struct dentry *dentry, *ret;
down(&p->d_inode->i_sem);
dentry = sysfs_get_dentry(p,n);
+ ret = dentry;
if (!IS_ERR(dentry)) {
int error = sysfs_create(dentry,
S_IFDIR| S_IRWXU | S_IRUGO | S_IXUGO,
@@ -36,11 +37,11 @@ create_dir(struct kobject * k, struct de
dentry->d_fsdata = k;
p->d_inode->i_nlink++;
} else
- dentry = ERR_PTR(error);
+ ret = ERR_PTR(error);
dput(dentry);
}
up(&p->d_inode->i_sem);
- return dentry;
+ return ret;
}
_
On Wed, Sep 10, 2003 at 03:16:50PM -0700, Ranjeet Shetye wrote:
>
> file: oops4.ksyms.txt: the OOPS that I got.
> file: oops4.bt.txt: ksymoops-derived backtrace of OOPS.
> file: .config - my .config file.
>
> To verify that this is not a compilation problem, I ran a 'make
> mrproper' and compiled from scratch.
>
> Something to note: module support is disabled and I remember that
> sometime back you needed to have ALSA compiled as a module, and this
> OOPS is in ALSA. Is it still necessary to compile ALSA as a module ?
>
> I am not on the kernel mailing list, (only on linux-net and netdev). So
> please email me personally if you need any other information.
What were you doing that caused this oops?
thanks,
greg k-h
On Wed, 2003-09-10 at 17:24, Greg KH wrote:
> On Wed, Sep 10, 2003 at 05:19:05PM -0700, Ranjeet Shetye wrote:
> >
> > Your changes fixed the issue. Thanks a lot for your help. I still get
> > this call trace, but no more OOPS on bootup.
> >
> > kobject_register failed for Ensoniq AudioPCI (-17)
> > Call Trace:
> > [<c026f45c>] kobject_register+0x50/0x59
> > [<c02f8003>] bus_add_driver+0x4c/0xaf
> > [<c02f8453>] driver_register+0x31/0x35
> > [<c027c3bf>] pci_populate_driver_dir+0x29/0x2b
> > [<c027c491>] pci_register_driver+0x5e/0x83
> > [<c06a145f>] alsa_card_ens137x_init+0x15/0x41
> > [<c068475a>] do_initcalls+0x2a/0x97
> > [<c012e920>] init_workqueues+0x12/0x2a
> > [<c01050a3>] init+0x39/0x196
> > [<c010506a>] init+0x0/0x196
> > [<c0108f31>] kernel_thread_helper+0x5/0xb
>
> Odds are that the pci driver is trying to register 2 drivers with the
> pci core with the same name. What does /sys/bus/pci/drivers show?
>
> thanks,
>
> greg k-h
Hi Greg,
I didn't find a /proc/sys/bus/pci/drivers, but I did find a
/proc/bus/pci/devices - which is what I am guessing you meant. If you
did in fact mean '/proc/sys/bus/pci/drivers' then I dont have any such
file. In fact I dont have a bus sub-directory under /proc/sys/
Anyways, here's the output:
cat /proc/bus/pci/devices
0000 80862560 0 f0000008 00000000
00000000 00000000 00000000 00000000
00000000 08000000 00000000 00000000
00000000 00000000 00000000 00000000
agpgart-intel
0010 80862562 10 e8000008 ff680000
00000000 00000000 00000000 00000000
00000000 08000000 00080000 00000000
00000000 00000000 00000000 00000000
00e8 808624c2 10 00000000 00000000
00000000 00000000 0000ff81 00000000
00000000 00000000 00000000 00000000
00000000 00000020 00000000 00000000 uhci-hcd
00e9 808624c4 13 00000000 00000000
00000000 00000000 0000ff61 00000000
00000000 00000000 00000000 00000000
00000000 00000020 00000000 00000000 uhci-hcd
00ea 808624c7 12 00000000 00000000
00000000 00000000 0000ff41 00000000
00000000 00000000 00000000 00000000
00000000 00000020 00000000 00000000 uhci-hcd
00ef 808624cd 17 ffa00800 00000000
00000000 00000000 00000000 00000000
00000000 00000400 00000000 00000000
00000000 00000000 00000000 00000000 ehci_hcd
00f0 8086244e 0 00000000 00000000
00000000 00000000 00000000 00000000
00000000 00000000 00000000 00000000
00000000 00000000 00000000 00000000
00f8 808624c0 0 00000000 00000000
00000000 00000000 00000000 00000000
00000000 00000000 00000000 00000000
00000000 00000000 00000000 00000000
00f9 808624cb 12 00000000 00000000
00000000 00000000 0000ffa1 10000000
00000000 00000000 00000000 00000000
00000000 00000010 00000400 00000000 PIIX IDE
00fb 808624c3 11 00000000 00000000
00000000 00000000 0000dc81 00000000
00000000 00000000 00000000 00000000
00000000 00000020 00000000 00000000
00fd 808624c5 11 0000d801 0000dc41
ffa00400 ffa00000 00000000 00000000
00000000 00000100 00000040 00000200
00000100 00000000 00000000 00000000 Intel
ICH
0138 16ae1141 10 f80fe008 00000000
00000000 00000000 00000000 00000000
00000000 00002000 00000000 00000000
00000000 00000000 00000000 00000000
0140 10ec8139 11 0000ec01 ff8ffc00
00000000 00000000 00000000 00000000
00000000 00000100 00000100 00000000
00000000 00000000 00000000 00000000 8139too
0160 8086100e 12 ff8c0000 00000000
0000e8c1 00000000 00000000 00000000
00000000 00020000 00000000 00000040
00000000 00000000 00000000 00000000 e1000
thanks,
--
Ranjeet Shetye
Senior Software Engineer
Zultys Technologies
Ranjeet dot Shetye2 at Zultys dot com
http://www.zultys.com/
The views, opinions, and judgements expressed in this message are solely
those of the author. The message contents have not been reviewed or
approved by Zultys.
On Wed, 2003-09-10 at 15:46, Andrew Morton wrote:
> Ranjeet Shetye <[email protected]> wrote:
> >
> > Unable to handle kernel paging request at virtual address ffffffef
> > printing eip:
> > c027184c
> > *pde = 00001067
> > *pte = 00000000
> > Oops: 0000 [#1]
> > CPU: 0
> > EIP: 0060:[<c027184c>] Not tainted
> > EFLAGS: 00010282
> > EIP is at atomic_dec_and_lock+0x8/0x54
> > eax: ffffffef ebx: ffffffef ecx: ffffffef edx: cf372254
> > esi: ffffffef edi: cf743e20 ebp: c12efea8 esp: c12efea0
> > ds: 007b es: 007b ss: 0068
> > Process swapper (pid: 1, threadinfo=c12ee000 task=c12ed8c0)
> > Stack: ffffffef ffffffef c12efec4 c01675e4 ffffffef c05a21b0 ffffffef cf7469b4
> > cf743e20 c12efee4 c018aa52 ffffffef 000041ed c018a998 c05f0f80 00000000
> > c05f0f80 c12eff00 c018aab1 c05f0f80 cf743e20 c05f0f84 00000000 c05f0f80
> > Call Trace:
> > [<c01675e4>] dput+0x24/0x227
> > [<c018aa52>] create_dir+0x9e/0xa4
> > [<c018a998>] init_dir+0x0/0x1c
> > [<c018aab1>] sysfs_create_dir+0x36/0x6c
> > [<c026eee0>] create_dir+0x1f/0x49
> > [<c026f331>] kobject_add+0x4d/0x124
>
> Bug in fs/sysfs/dir.c:create_dir() - sysfs_create() returned -EEXIST and we
> turned that into a pointer and did a dput() on it.
>
> Something like this should fix it. The -EEXIST return may be another bug?
>
>
>
> diff -puN fs/sysfs/dir.c~sysfs-create_dir-oops-fix fs/sysfs/dir.c
> --- 25/fs/sysfs/dir.c~sysfs-create_dir-oops-fix Wed Sep 10 15:41:35 2003
> +++ 25-akpm/fs/sysfs/dir.c Wed Sep 10 15:44:42 2003
> @@ -24,10 +24,11 @@ static int init_dir(struct inode * inode
> static struct dentry *
> create_dir(struct kobject * k, struct dentry * p, const char * n)
> {
> - struct dentry * dentry;
> + struct dentry *dentry, *ret;
>
> down(&p->d_inode->i_sem);
> dentry = sysfs_get_dentry(p,n);
> + ret = dentry;
> if (!IS_ERR(dentry)) {
> int error = sysfs_create(dentry,
> S_IFDIR| S_IRWXU | S_IRUGO | S_IXUGO,
> @@ -36,11 +37,11 @@ create_dir(struct kobject * k, struct de
> dentry->d_fsdata = k;
> p->d_inode->i_nlink++;
> } else
> - dentry = ERR_PTR(error);
> + ret = ERR_PTR(error);
> dput(dentry);
> }
> up(&p->d_inode->i_sem);
> - return dentry;
> + return ret;
> }
>
>
>
> _
Hi Andrew,
Your changes fixed the issue. Thanks a lot for your help. I still get
this call trace, but no more OOPS on bootup.
kobject_register failed for Ensoniq AudioPCI (-17)
Call Trace:
[<c026f45c>] kobject_register+0x50/0x59
[<c02f8003>] bus_add_driver+0x4c/0xaf
[<c02f8453>] driver_register+0x31/0x35
[<c027c3bf>] pci_populate_driver_dir+0x29/0x2b
[<c027c491>] pci_register_driver+0x5e/0x83
[<c06a145f>] alsa_card_ens137x_init+0x15/0x41
[<c068475a>] do_initcalls+0x2a/0x97
[<c012e920>] init_workqueues+0x12/0x2a
[<c01050a3>] init+0x39/0x196
[<c010506a>] init+0x0/0x196
[<c0108f31>] kernel_thread_helper+0x5/0xb
Also, I wanted to ask one more thing:
In the top level Makefile, I have replaced
EXTRAVERSION = -test5
with
TESTVERSION = -test5
BKVERSION = $(shell if [ -d "./BitKeeper" ]; then echo '-bk-`bk changes
| head -1 | cut -f 2 -d @ | cut -f 1 -d ,`' ; else echo ''; fi)
EXTRAVERSION = $(TESTVERSION)$(BKVERSION)
As a result 'uname -r' on a bk-based kernel now returns a more useful
'2.6.0-test5-bk-1.1227.1.49' instead of just '2.6.0-test5'. For non-bk
based source code trees, the version is still '2.6.0-test5'. This is
very useful to me because I have 2 seperate 2.6 trees (bk and vanilla),
and this lets me figure out where I got my kernel from and esp. at what
bk changeset it was built.
I am not very familiar with bk (still stumbling around) and hence I
couldn't generate a bk diff or do a bk send etc; if you think this
change is useful, please send it over to be included in the main tree.
thanks again,
Ranjeet.
--
Ranjeet Shetye
Senior Software Engineer
Zultys Technologies
Ranjeet dot Shetye2 at Zultys dot com
http://www.zultys.com/
The views, opinions, and judgements expressed in this message are solely
those of the author. The message contents have not been reviewed or
approved by Zultys.
On Wed, Sep 10, 2003 at 05:19:05PM -0700, Ranjeet Shetye wrote:
>
> Your changes fixed the issue. Thanks a lot for your help. I still get
> this call trace, but no more OOPS on bootup.
>
> kobject_register failed for Ensoniq AudioPCI (-17)
> Call Trace:
> [<c026f45c>] kobject_register+0x50/0x59
> [<c02f8003>] bus_add_driver+0x4c/0xaf
> [<c02f8453>] driver_register+0x31/0x35
> [<c027c3bf>] pci_populate_driver_dir+0x29/0x2b
> [<c027c491>] pci_register_driver+0x5e/0x83
> [<c06a145f>] alsa_card_ens137x_init+0x15/0x41
> [<c068475a>] do_initcalls+0x2a/0x97
> [<c012e920>] init_workqueues+0x12/0x2a
> [<c01050a3>] init+0x39/0x196
> [<c010506a>] init+0x0/0x196
> [<c0108f31>] kernel_thread_helper+0x5/0xb
Odds are that the pci driver is trying to register 2 drivers with the
pci core with the same name. What does /sys/bus/pci/drivers show?
thanks,
greg k-h
On Wed, 2003-09-10 at 16:25, Greg KH wrote:
> On Wed, Sep 10, 2003 at 03:16:50PM -0700, Ranjeet Shetye wrote:
> >
> > file: oops4.ksyms.txt: the OOPS that I got.
> > file: oops4.bt.txt: ksymoops-derived backtrace of OOPS.
> > file: .config - my .config file.
> >
> > To verify that this is not a compilation problem, I ran a 'make
> > mrproper' and compiled from scratch.
> >
> > Something to note: module support is disabled and I remember that
> > sometime back you needed to have ALSA compiled as a module, and this
> > OOPS is in ALSA. Is it still necessary to compile ALSA as a module ?
> >
> > I am not on the kernel mailing list, (only on linux-net and netdev). So
> > please email me personally if you need any other information.
>
> What were you doing that caused this oops?
>
> thanks,
>
> greg k-h
Hi Greg,
The OOPS was happening at boot time. Andrew's fix has resolved the
issue.
thanks,
Ranjeet.
--
Ranjeet Shetye
Senior Software Engineer
Zultys Technologies
Ranjeet dot Shetye2 at Zultys dot com
http://www.zultys.com/
The views, opinions, and judgements expressed in this message are solely
those of the author. The message contents have not been reviewed or
approved by Zultys.
On Wed, Sep 10, 2003 at 05:36:51PM -0700, Ranjeet Shetye wrote:
> On Wed, 2003-09-10 at 17:24, Greg KH wrote:
> > On Wed, Sep 10, 2003 at 05:19:05PM -0700, Ranjeet Shetye wrote:
> > >
> > > Your changes fixed the issue. Thanks a lot for your help. I still get
> > > this call trace, but no more OOPS on bootup.
> > >
> > > kobject_register failed for Ensoniq AudioPCI (-17)
> > > Call Trace:
> > > [<c026f45c>] kobject_register+0x50/0x59
> > > [<c02f8003>] bus_add_driver+0x4c/0xaf
> > > [<c02f8453>] driver_register+0x31/0x35
> > > [<c027c3bf>] pci_populate_driver_dir+0x29/0x2b
> > > [<c027c491>] pci_register_driver+0x5e/0x83
> > > [<c06a145f>] alsa_card_ens137x_init+0x15/0x41
> > > [<c068475a>] do_initcalls+0x2a/0x97
> > > [<c012e920>] init_workqueues+0x12/0x2a
> > > [<c01050a3>] init+0x39/0x196
> > > [<c010506a>] init+0x0/0x196
> > > [<c0108f31>] kernel_thread_helper+0x5/0xb
> >
> > Odds are that the pci driver is trying to register 2 drivers with the
> > pci core with the same name. What does /sys/bus/pci/drivers show?
> >
> > thanks,
> >
> > greg k-h
>
> Hi Greg,
>
> I didn't find a /proc/sys/bus/pci/drivers, but I did find a
> /proc/bus/pci/devices - which is what I am guessing you meant. If you
> did in fact mean '/proc/sys/bus/pci/drivers' then I dont have any such
> file. In fact I dont have a bus sub-directory under /proc/sys/
No, look at what I asked for above, "/sys/bus/pci..." I don't care about
/proc at all :)
sysfs is mounted at /sys.
thanks,
greg k-h
On Wed, 2003-09-10 at 19:40, Greg KH wrote:
> On Wed, Sep 10, 2003 at 05:36:51PM -0700, Ranjeet Shetye wrote:
> > On Wed, 2003-09-10 at 17:24, Greg KH wrote:
> > > On Wed, Sep 10, 2003 at 05:19:05PM -0700, Ranjeet Shetye wrote:
> > > >
> > > > Your changes fixed the issue. Thanks a lot for your help. I still get
> > > > this call trace, but no more OOPS on bootup.
> > > >
> > > > kobject_register failed for Ensoniq AudioPCI (-17)
> > > > Call Trace:
> > > > [<c026f45c>] kobject_register+0x50/0x59
> > > > [<c02f8003>] bus_add_driver+0x4c/0xaf
> > > > [<c02f8453>] driver_register+0x31/0x35
> > > > [<c027c3bf>] pci_populate_driver_dir+0x29/0x2b
> > > > [<c027c491>] pci_register_driver+0x5e/0x83
> > > > [<c06a145f>] alsa_card_ens137x_init+0x15/0x41
> > > > [<c068475a>] do_initcalls+0x2a/0x97
> > > > [<c012e920>] init_workqueues+0x12/0x2a
> > > > [<c01050a3>] init+0x39/0x196
> > > > [<c010506a>] init+0x0/0x196
> > > > [<c0108f31>] kernel_thread_helper+0x5/0xb
> > >
> > > Odds are that the pci driver is trying to register 2 drivers with the
> > > pci core with the same name. What does /sys/bus/pci/drivers show?
> > >
> > > thanks,
> > >
> > > greg k-h
> >
> > Hi Greg,
> >
> > I didn't find a /proc/sys/bus/pci/drivers, but I did find a
> > /proc/bus/pci/devices - which is what I am guessing you meant. If you
> > did in fact mean '/proc/sys/bus/pci/drivers' then I dont have any such
> > file. In fact I dont have a bus sub-directory under /proc/sys/
>
> No, look at what I asked for above, "/sys/bus/pci..." I don't care about
> /proc at all :)
>
> sysfs is mounted at /sys.
>
> thanks,
>
> greg k-h
hi,
Actually I don't know what part I am missing, but I dont have a /sys
directory at all. That's why I assumed you might be referring to
/proc/sys. I've attached my .config. Is there some option I need to turn
on ? I looked and couldn't find it. (The fs/Makefile has a obj-y for
sysfs and there is no CONFIG_ parameter in any file in fs/sysfs). I dont
have a top level /sys directory. Is that the problem ? How do i get my
sysfs running ?
I am running Gentoo Linux 1.4 on x86 with kernel
2.6.0-test5-bk-1.1227.1.49 and gcc 3.2.3.
thanks,
--
Ranjeet Shetye
Senior Software Engineer
Zultys Technologies
Ranjeet dot Shetye2 at Zultys dot com
http://www.zultys.com/
The views, opinions, and judgements expressed in this message are solely
those of the author. The message contents have not been reviewed or
approved by Zultys.
On Thu, Sep 11, 2003 at 01:26:26PM -0700, Ranjeet Shetye wrote:
> On Wed, 2003-09-10 at 19:40, Greg KH wrote:
> > On Wed, Sep 10, 2003 at 05:36:51PM -0700, Ranjeet Shetye wrote:
> > > On Wed, 2003-09-10 at 17:24, Greg KH wrote:
> > > > On Wed, Sep 10, 2003 at 05:19:05PM -0700, Ranjeet Shetye wrote:
> > > > >
> > > > > Your changes fixed the issue. Thanks a lot for your help. I still get
> > > > > this call trace, but no more OOPS on bootup.
> > > > >
> > > > > kobject_register failed for Ensoniq AudioPCI (-17)
> > > > > Call Trace:
> > > > > [<c026f45c>] kobject_register+0x50/0x59
> > > > > [<c02f8003>] bus_add_driver+0x4c/0xaf
> > > > > [<c02f8453>] driver_register+0x31/0x35
> > > > > [<c027c3bf>] pci_populate_driver_dir+0x29/0x2b
> > > > > [<c027c491>] pci_register_driver+0x5e/0x83
> > > > > [<c06a145f>] alsa_card_ens137x_init+0x15/0x41
> > > > > [<c068475a>] do_initcalls+0x2a/0x97
> > > > > [<c012e920>] init_workqueues+0x12/0x2a
> > > > > [<c01050a3>] init+0x39/0x196
> > > > > [<c010506a>] init+0x0/0x196
> > > > > [<c0108f31>] kernel_thread_helper+0x5/0xb
> > > >
> > > > Odds are that the pci driver is trying to register 2 drivers with the
> > > > pci core with the same name. What does /sys/bus/pci/drivers show?
> > >
> > > I didn't find a /proc/sys/bus/pci/drivers, but I did find a
> > > /proc/bus/pci/devices - which is what I am guessing you meant. If you
> > > did in fact mean '/proc/sys/bus/pci/drivers' then I dont have any such
> > > file. In fact I dont have a bus sub-directory under /proc/sys/
> >
> > No, look at what I asked for above, "/sys/bus/pci..." I don't care about
> > /proc at all :)
> >
> > sysfs is mounted at /sys.
>
> Actually I don't know what part I am missing, but I dont have a /sys
> directory at all. That's why I assumed you might be referring to
> /proc/sys. I've attached my .config. Is there some option I need to turn
> on ? I looked and couldn't find it. (The fs/Makefile has a obj-y for
> sysfs and there is no CONFIG_ parameter in any file in fs/sysfs). I dont
> have a top level /sys directory. Is that the problem ? How do i get my
> sysfs running ?
mkdir /sys
mount -t sysfs none /sys
thanks,
greg k-h
On Thu, 2003-09-11 at 13:28, Greg KH wrote:
> On Thu, Sep 11, 2003 at 01:26:26PM -0700, Ranjeet Shetye wrote:
> > On Wed, 2003-09-10 at 19:40, Greg KH wrote:
> > > On Wed, Sep 10, 2003 at 05:36:51PM -0700, Ranjeet Shetye wrote:
> > > > On Wed, 2003-09-10 at 17:24, Greg KH wrote:
> > > > > On Wed, Sep 10, 2003 at 05:19:05PM -0700, Ranjeet Shetye wrote:
> > > > > >
> > > > > > Your changes fixed the issue. Thanks a lot for your help. I still get
> > > > > > this call trace, but no more OOPS on bootup.
> > > > > >
> > > > > > kobject_register failed for Ensoniq AudioPCI (-17)
> > > > > > Call Trace:
> > > > > > [<c026f45c>] kobject_register+0x50/0x59
> > > > > > [<c02f8003>] bus_add_driver+0x4c/0xaf
> > > > > > [<c02f8453>] driver_register+0x31/0x35
> > > > > > [<c027c3bf>] pci_populate_driver_dir+0x29/0x2b
> > > > > > [<c027c491>] pci_register_driver+0x5e/0x83
> > > > > > [<c06a145f>] alsa_card_ens137x_init+0x15/0x41
> > > > > > [<c068475a>] do_initcalls+0x2a/0x97
> > > > > > [<c012e920>] init_workqueues+0x12/0x2a
> > > > > > [<c01050a3>] init+0x39/0x196
> > > > > > [<c010506a>] init+0x0/0x196
> > > > > > [<c0108f31>] kernel_thread_helper+0x5/0xb
> > > > >
> > > > > Odds are that the pci driver is trying to register 2 drivers with the
> > > > > pci core with the same name. What does /sys/bus/pci/drivers show?
> > > >
> > > > I didn't find a /proc/sys/bus/pci/drivers, but I did find a
> > > > /proc/bus/pci/devices - which is what I am guessing you meant. If you
> > > > did in fact mean '/proc/sys/bus/pci/drivers' then I dont have any such
> > > > file. In fact I dont have a bus sub-directory under /proc/sys/
> > >
> > > No, look at what I asked for above, "/sys/bus/pci..." I don't care about
> > > /proc at all :)
> > >
> > > sysfs is mounted at /sys.
> >
> > Actually I don't know what part I am missing, but I dont have a /sys
> > directory at all. That's why I assumed you might be referring to
> > /proc/sys. I've attached my .config. Is there some option I need to turn
> > on ? I looked and couldn't find it. (The fs/Makefile has a obj-y for
> > sysfs and there is no CONFIG_ parameter in any file in fs/sysfs). I dont
> > have a top level /sys directory. Is that the problem ? How do i get my
> > sysfs running ?
>
> mkdir /sys
> mount -t sysfs none /sys
>
> thanks,
>
> greg k-h
he, he. thanks for that. Didn't know what I was missing. Will drop a
line to Gentoo folks to update their x86 installation guide with this
gem.
Here's the listing from "/sys/bus/pci/drivers"
OUTPUT of "/bin/ls -c1 /sys/bus/pci/drivers"
ALI 5451
AZF3328
Digigram VX222
EMU10K1_Audigy
ES1968 (ESS Maestro
ESS ES1938 (Solo-1)
FM801
ICE1712
ICE1724
Intel ICH
Intel ICH Joystick
Maestro3
NeoMagic 256
RME Digi32
RME Digi96
RME Digi9652 (Hammerfall
RME Hammerfall DSP
S3 SonicVibes
Sound Fusion CS46xx
Trident4DWaveAudio
VIA 82xx Audio
Yamaha DS-XG PCI
korg1212
ALS4000
C-Media PCI
CS4281
Ensoniq AudioPCI
ehci_hcd
uhci-hcd
PCI IDE
PIIX IDE
RZ1000 IDE
8139too
e100
e1000
agpgart-ali
agpgart-amdk7
agpgart-intel
agpgart-serverworks
agpgart-sis
agpgart-via
parport_pc
serial
OUTPUT of "find /sys/bus/pci/drivers"
/sys/bus/pci/drivers
/sys/bus/pci/drivers/Digigram VX222
/sys/bus/pci/drivers/Digigram VX222/new_id
/sys/bus/pci/drivers/ICE1724
/sys/bus/pci/drivers/ICE1724/new_id
/sys/bus/pci/drivers/ICE1712
/sys/bus/pci/drivers/ICE1712/new_id
/sys/bus/pci/drivers/Yamaha DS-XG PCI
/sys/bus/pci/drivers/Yamaha DS-XG PCI/new_id
/sys/bus/pci/drivers/Trident4DWaveAudio
/sys/bus/pci/drivers/Trident4DWaveAudio/new_id
/sys/bus/pci/drivers/RME Hammerfall DSP
/sys/bus/pci/drivers/RME Hammerfall DSP/new_id
/sys/bus/pci/drivers/RME Digi9652 (Hammerfall
/sys/bus/pci/drivers/RME Digi9652 (Hammerfall/new_id
/sys/bus/pci/drivers/NeoMagic 256
/sys/bus/pci/drivers/NeoMagic 256/new_id
/sys/bus/pci/drivers/korg1212
/sys/bus/pci/drivers/korg1212/new_id
/sys/bus/pci/drivers/EMU10K1_Audigy
/sys/bus/pci/drivers/EMU10K1_Audigy/new_id
/sys/bus/pci/drivers/Sound Fusion CS46xx
/sys/bus/pci/drivers/Sound Fusion CS46xx/new_id
/sys/bus/pci/drivers/ALI 5451
/sys/bus/pci/drivers/ALI 5451/new_id
/sys/bus/pci/drivers/AZF3328
/sys/bus/pci/drivers/AZF3328/new_id
/sys/bus/pci/drivers/VIA 82xx Audio
/sys/bus/pci/drivers/VIA 82xx Audio/new_id
/sys/bus/pci/drivers/S3 SonicVibes
/sys/bus/pci/drivers/S3 SonicVibes/new_id
/sys/bus/pci/drivers/RME Digi96
/sys/bus/pci/drivers/RME Digi96/new_id
/sys/bus/pci/drivers/RME Digi32
/sys/bus/pci/drivers/RME Digi32/new_id
/sys/bus/pci/drivers/Maestro3
/sys/bus/pci/drivers/Maestro3/new_id
/sys/bus/pci/drivers/Intel ICH Joystick
/sys/bus/pci/drivers/Intel ICH Joystick/new_id
/sys/bus/pci/drivers/Intel ICH
/sys/bus/pci/drivers/Intel ICH/new_id
/sys/bus/pci/drivers/Intel ICH/0000:00:1f.5
/sys/bus/pci/drivers/FM801
/sys/bus/pci/drivers/FM801/new_id
/sys/bus/pci/drivers/ES1968 (ESS Maestro
/sys/bus/pci/drivers/ES1968 (ESS Maestro/new_id
/sys/bus/pci/drivers/ESS ES1938 (Solo-1)
/sys/bus/pci/drivers/ESS ES1938 (Solo-1)/new_id
/sys/bus/pci/drivers/Ensoniq AudioPCI
/sys/bus/pci/drivers/Ensoniq AudioPCI/new_id
/sys/bus/pci/drivers/CS4281
/sys/bus/pci/drivers/CS4281/new_id
/sys/bus/pci/drivers/C-Media PCI
/sys/bus/pci/drivers/C-Media PCI/new_id
/sys/bus/pci/drivers/ALS4000
/sys/bus/pci/drivers/ALS4000/new_id
/sys/bus/pci/drivers/uhci-hcd
/sys/bus/pci/drivers/uhci-hcd/new_id
/sys/bus/pci/drivers/uhci-hcd/0000:00:1d.2
/sys/bus/pci/drivers/uhci-hcd/0000:00:1d.1
/sys/bus/pci/drivers/uhci-hcd/0000:00:1d.0
/sys/bus/pci/drivers/ehci_hcd
/sys/bus/pci/drivers/ehci_hcd/new_id
/sys/bus/pci/drivers/ehci_hcd/0000:00:1d.7
/sys/bus/pci/drivers/PCI IDE
/sys/bus/pci/drivers/PCI IDE/new_id
/sys/bus/pci/drivers/RZ1000 IDE
/sys/bus/pci/drivers/RZ1000 IDE/new_id
/sys/bus/pci/drivers/PIIX IDE
/sys/bus/pci/drivers/PIIX IDE/new_id
/sys/bus/pci/drivers/PIIX IDE/0000:00:1f.1
/sys/bus/pci/drivers/8139too
/sys/bus/pci/drivers/8139too/new_id
/sys/bus/pci/drivers/8139too/0000:01:08.0
/sys/bus/pci/drivers/e1000
/sys/bus/pci/drivers/e1000/new_id
/sys/bus/pci/drivers/e1000/0000:01:0c.0
/sys/bus/pci/drivers/e100
/sys/bus/pci/drivers/e100/new_id
/sys/bus/pci/drivers/parport_pc
/sys/bus/pci/drivers/parport_pc/new_id
/sys/bus/pci/drivers/serial
/sys/bus/pci/drivers/serial/new_id
/sys/bus/pci/drivers/agpgart-via
/sys/bus/pci/drivers/agpgart-via/new_id
/sys/bus/pci/drivers/agpgart-serverworks
/sys/bus/pci/drivers/agpgart-serverworks/new_id
/sys/bus/pci/drivers/agpgart-sis
/sys/bus/pci/drivers/agpgart-sis/new_id
/sys/bus/pci/drivers/agpgart-intel
/sys/bus/pci/drivers/agpgart-intel/new_id
/sys/bus/pci/drivers/agpgart-intel/0000:00:00.0
/sys/bus/pci/drivers/agpgart-amdk7
/sys/bus/pci/drivers/agpgart-amdk7/new_id
/sys/bus/pci/drivers/agpgart-ali
/sys/bus/pci/drivers/agpgart-ali/new_id
thanks,
--
Ranjeet Shetye
Senior Software Engineer
Zultys Technologies
Ranjeet dot Shetye2 at Zultys dot com
http://www.zultys.com/
The views, opinions, and judgements expressed in this message are solely
those of the author. The message contents have not been reviewed or
approved by Zultys.
On Wed, Sep 10, 2003 at 05:36:51PM -0700, Ranjeet Shetye wrote:
> On Wed, 2003-09-10 at 17:24, Greg KH wrote:
> > On Wed, Sep 10, 2003 at 05:19:05PM -0700, Ranjeet Shetye wrote:
> > >
> > > Your changes fixed the issue. Thanks a lot for your help. I still get
> > > this call trace, but no more OOPS on bootup.
> > >
> > > kobject_register failed for Ensoniq AudioPCI (-17)
> > > Call Trace:
> > > [<c026f45c>] kobject_register+0x50/0x59
> > > [<c02f8003>] bus_add_driver+0x4c/0xaf
> > > [<c02f8453>] driver_register+0x31/0x35
> > > [<c027c3bf>] pci_populate_driver_dir+0x29/0x2b
> > > [<c027c491>] pci_register_driver+0x5e/0x83
> > > [<c06a145f>] alsa_card_ens137x_init+0x15/0x41
> > > [<c068475a>] do_initcalls+0x2a/0x97
> > > [<c012e920>] init_workqueues+0x12/0x2a
> > > [<c01050a3>] init+0x39/0x196
> > > [<c010506a>] init+0x0/0x196
> > > [<c0108f31>] kernel_thread_helper+0x5/0xb
> >
> > Odds are that the pci driver is trying to register 2 drivers with the
> > pci core with the same name. What does /sys/bus/pci/drivers show?
> >
> > thanks,
> >
> > greg k-h
Hi Ranjeet,
I noticed this in your .config included in your previous message.
>CONFIG_SND_ENS1370=y
>CONFIG_SND_ENS1371=y
I think the conflict might be occuring between these two drivers.
Look at sound/pci/ens1371.c and sound/pci/ens1370.c.
Out of curiosity, could you try this patch without making any changes to
your config. If it works properly, could you then verify it in sysfs.
--- a/sound/pci/ens1370.c 2003-09-13 19:28:45.000000000 +0000
+++ b/sound/pci/ens1370.c 2003-09-13 19:30:02.000000000 +0000
@@ -2354,7 +2354,11 @@
}
static struct pci_driver driver = {
- .name = "Ensoniq AudioPCI",
+#ifdef CHIP1371
+ .name = "Ensoniq 1371",
+#else
+ .name = "Ensoniq 1370",
+#endif
.id_table = snd_audiopci_ids,
.probe = snd_audiopci_probe,
.remove = __devexit_p(snd_audiopci_remove),
Thanks,
Adam