[PATCH] char/ipmi: fix OOPS caused by pnp_unregister_driver on unregistered driver

[PATCH] char/ipmi: fix OOPS caused by pnp_unregister_driver on unregistered driver

[Peter Huewe]

This patch fixes an OOPS triggered when calling modprobe ipmi_si a
second time after the first modprobe returned without finding any ipmi
devices.

The OOPS is caused by calling pnp_unregister_driver on a (possibly)
unregistered pnp driver. (Although I'm not sure why it works the first
time)

IPMI System Interface driver.
ipmi_si: Adding default-specified kcs state machine
ipmi_si: Trying default-specified kcs state machine at i/o address 0xca2, slave address 0x0, irq 0
ipmi_si: Interface detection failed
ipmi_si: Adding default-specified smic state machine
ipmi_si: Trying default-specified smic state machine at i/o address 0xca9, slave address 0x0, irq 0
ipmi_si: Interface detection failed
ipmi_si: Adding default-specified bt state machine
ipmi_si: Trying default-specified bt state machine at i/o address 0xe4, slave address 0x0, irq 0
ipmi_si: Interface detection failed
ipmi_si: Unable to find any System Interface(s)

IPMI System Interface driver.
BUG: unable to handle kernel paging request at c9ccb990
IP: [<c1380c98>] put_driver+0x8/0x20
*pde = 05e2f067 *pte = 00000000
Oops: 0000 [#1]
last sysfs file: /sys/module/ipmi_msghandler/initstate
Modules linked in: ipmi_si(+) ipmi_msghandler snd_seq_dummy snd_seq_oss snd_seq_midi_event snd_seq snd_pcm_oss snd_mixer_oss ipv6 pcmcia pcmcia_core lp fuse snd_ens1371 snd_via82xx gameport snd_ac97_codec ac97_bus snd_pcm ppdev snd_timer snd_mpu401_uart snd_rawmidi snd_seq_device snd via686a snd_page_alloc hwmon soundcore parport_pc shpchp rtc_cmos psmouse rtc_core parport rtc_lib button evdev serio_raw sg [last unloaded: pcmcia_core]

Pid: 1653, comm: insmod Not tainted 2.6.37 #1 8363-686A/VT8363
EIP: 0060:[<c1380c98>] EFLAGS: 00010286 CPU: 0
EIP is at put_driver+0x8/0x20
EAX: c9ccb960 EBX: c12cc2e0 ECX: 00000000 EDX: c9ccb960
ESI: c9ceb960 EDI: 00000000 EBP: c5e8fee0 ESP: c5e8fee0
 DS: 007b ES: 007b FS: 0000 GS: 00e0 SS: 0068
Process insmod (pid: 1653, ti=c5e8e000 task=c6e0a000 task.ti=c5e8e000)
Stack:
 c5e8ff04 c1380db5 c1118ba3 c9ceb8c0 c5e8fef8 c1380e99 00000000 00000004
 00000000 c5e8ff0c c12cc49c c5e8ff58 c9ce9640 c9cea71c 000001a4 00000000
 c9ceb65c c19fc3f8 fffffffc 00000000 c5e8ff38 c107b185 c5e8ff54 c105063c
Call Trace:
 [<c1380db5>] ? driver_register+0x105/0x130
 [<c1118ba3>] ? sysfs_create_file+0x33/0x40
 [<c1380e99>] ? driver_create_file+0x19/0x20
 [<c12cc49c>] ? pnp_register_driver+0x1c/0x20
 [<c9ce9640>] ? init_ipmi_si+0x294/0x88f [ipmi_si]
 [<c107b185>] ? tracepoint_module_notify+0x25/0x30
 [<c105063c>] ? notifier_call_chain+0x4c/0x70
 [<c1001206>] ? do_one_initcall+0x116/0x160
 [<c9ce93ac>] ? init_ipmi_si+0x0/0x88f [ipmi_si]
 [<c1062f41>] ? sys_init_module+0x91/0x1d0
 [<c17b7ebc>] ? syscall_call+0x7/0xb
Code: 01 00 00 c7 04 24 c7 1c 91 c1 e8 64 14 cb ff 83 c4 0c 5b 5e c9 c3 8d b6 00 00 00 00 8d bc 27 00 00 00 00 55 89 e5 3e 8d 74 26 00 <8b> 40 30 e8 70 18 ee ff c9 c3 8d b4 26 00 00 00 00 8d bc 27 00
EIP: [<c1380c98>] put_driver+0x8/0x20 SS:ESP 0068:c5e8fee0
CR2: 00000000c9ccb990
---[ end trace 9b4a0dc647b099c4 ]---

Since the OOPS occurs in the current 2.6.37 release, I added stable on
CC.
The code exists in this form since 27d0567ab6 (2009-12-17), but could
not reproduce it in 2.6.33.

KernelVersion: 2.6.37
Cc: <stable@kernel.org>
Signed-off-by: Peter Huewe <peterhuewe@gmx.de>
---
 drivers/char/ipmi/ipmi_si_intf.c |    5 +++++
 1 files changed, 5 insertions(+), 0 deletions(-)

diff --git a/drivers/char/ipmi/ipmi_si_intf.c b/drivers/char/ipmi/ipmi_si_intf.c
index 035da9e..26c46af 100644
--- a/drivers/char/ipmi/ipmi_si_intf.c
+++ b/drivers/char/ipmi/ipmi_si_intf.c
@@ -3440,6 +3440,11 @@ static int __devinit init_ipmi_si(void)
 			pci_unregister_driver(&ipmi_pci_driver);
 #endif
 
+#ifdef CONFIG_ACPI
+		if (pnp_registered)
+			pnp_unregister_driver(&ipmi_pnp_driver);
+#endif
+
 #ifdef CONFIG_PPC_OF
 		if (of_registered)
 			of_unregister_platform_driver(&ipmi_of_platform_driver);
-- 
1.7.2.2

Re: [PATCH] char/ipmi: fix OOPS caused by pnp_unregister_driver on unregistered driver

[Peter Hüwe]

Am Sonntag 09 Januar 2011, 16:41:20 schrieb Peter Huewe:
> This patch fixes an OOPS triggered when calling modprobe ipmi_si a
> second time after the first modprobe returned without finding any ipmi
> devices.
> 
> The OOPS is caused by calling pnp_unregister_driver on a (possibly)
> unregistered pnp driver. (Although I'm not sure why it works the first
> time)
> 
> IPMI System Interface driver.
> ipmi_si: Adding default-specified kcs state machine
> ipmi_si: Trying default-specified kcs state machine at i/o address 0xca2,
> slave address 0x0, irq 0 ipmi_si: Interface detection failed
> ipmi_si: Adding default-specified smic state machine
> ipmi_si: Trying default-specified smic state machine at i/o address 0xca9,
> slave address 0x0, irq 0 ipmi_si: Interface detection failed
> ipmi_si: Adding default-specified bt state machine
> ipmi_si: Trying default-specified bt state machine at i/o address 0xe4,
> slave address 0x0, irq 0 ipmi_si: Interface detection failed
> ipmi_si: Unable to find any System Interface(s)
> 
> IPMI System Interface driver.
> BUG: unable to handle kernel paging request at c9ccb990
> IP: [<c1380c98>] put_driver+0x8/0x20
> *pde = 05e2f067 *pte = 00000000
> Oops: 0000 [#1]
> last sysfs file: /sys/module/ipmi_msghandler/initstate
> Modules linked in: ipmi_si(+) ipmi_msghandler snd_seq_dummy snd_seq_oss
> snd_seq_midi_event snd_seq snd_pcm_oss snd_mixer_oss ipv6 pcmcia
> pcmcia_core lp fuse snd_ens1371 snd_via82xx gameport snd_ac97_codec
> ac97_bus snd_pcm ppdev snd_timer snd_mpu401_uart snd_rawmidi
> snd_seq_device snd via686a snd_page_alloc hwmon soundcore parport_pc
> shpchp rtc_cmos psmouse rtc_core parport rtc_lib button evdev serio_raw sg
> [last unloaded: pcmcia_core]
> 
> Pid: 1653, comm: insmod Not tainted 2.6.37 #1 8363-686A/VT8363
> EIP: 0060:[<c1380c98>] EFLAGS: 00010286 CPU: 0
> EIP is at put_driver+0x8/0x20
> EAX: c9ccb960 EBX: c12cc2e0 ECX: 00000000 EDX: c9ccb960
> ESI: c9ceb960 EDI: 00000000 EBP: c5e8fee0 ESP: c5e8fee0
>  DS: 007b ES: 007b FS: 0000 GS: 00e0 SS: 0068
> Process insmod (pid: 1653, ti=c5e8e000 task=c6e0a000 task.ti=c5e8e000)
> Stack:
>  c5e8ff04 c1380db5 c1118ba3 c9ceb8c0 c5e8fef8 c1380e99 00000000 00000004
>  00000000 c5e8ff0c c12cc49c c5e8ff58 c9ce9640 c9cea71c 000001a4 00000000
>  c9ceb65c c19fc3f8 fffffffc 00000000 c5e8ff38 c107b185 c5e8ff54 c105063c
> Call Trace:
>  [<c1380db5>] ? driver_register+0x105/0x130
>  [<c1118ba3>] ? sysfs_create_file+0x33/0x40
>  [<c1380e99>] ? driver_create_file+0x19/0x20
>  [<c12cc49c>] ? pnp_register_driver+0x1c/0x20
>  [<c9ce9640>] ? init_ipmi_si+0x294/0x88f [ipmi_si]
>  [<c107b185>] ? tracepoint_module_notify+0x25/0x30
>  [<c105063c>] ? notifier_call_chain+0x4c/0x70
>  [<c1001206>] ? do_one_initcall+0x116/0x160
>  [<c9ce93ac>] ? init_ipmi_si+0x0/0x88f [ipmi_si]
>  [<c1062f41>] ? sys_init_module+0x91/0x1d0
>  [<c17b7ebc>] ? syscall_call+0x7/0xb
> Code: 01 00 00 c7 04 24 c7 1c 91 c1 e8 64 14 cb ff 83 c4 0c 5b 5e c9 c3 8d
> b6 00 00 00 00 8d bc 27 00 00 00 00 55 89 e5 3e 8d 74 26 00 <8b> 40 30 e8
> 70 18 ee ff c9 c3 8d b4 26 00 00 00 00 8d bc 27 00 EIP: [<c1380c98>]
> put_driver+0x8/0x20 SS:ESP 0068:c5e8fee0
> CR2: 00000000c9ccb990
> ---[ end trace 9b4a0dc647b099c4 ]---
> 
> Since the OOPS occurs in the current 2.6.37 release, I added stable on
> CC.
> The code exists in this form since 27d0567ab6 (2009-12-17), but could
> not reproduce it in 2.6.33.
> 
> KernelVersion: 2.6.37
> Cc: <stable@kernel.org>
> Signed-off-by: Peter Huewe <peterhuewe@gmx.de>
> ---
 Ping - any updates on this one?

Thanks,
Peter

Re: [PATCH] char/ipmi: fix OOPS caused by pnp_unregister_driver on unregistered driver

[Peter Hüwe]

Am Sonntag 09 Januar 2011, 16:41:20 schrieb Peter Huewe:
> This patch fixes an OOPS triggered when calling modprobe ipmi_si a
> second time after the first modprobe returned without finding any ipmi
> devices.
> 
> The OOPS is caused by calling pnp_unregister_driver on a (possibly)
> unregistered pnp driver. (Although I'm not sure why it works the first
> time)
> 
> IPMI System Interface driver.
> ipmi_si: Adding default-specified kcs state machine
> ipmi_si: Trying default-specified kcs state machine at i/o address 0xca2,
> slave address 0x0, irq 0 ipmi_si: Interface detection failed
> ipmi_si: Adding default-specified smic state machine
> ipmi_si: Trying default-specified smic state machine at i/o address 0xca9,
> slave address 0x0, irq 0 ipmi_si: Interface detection failed
> ipmi_si: Adding default-specified bt state machine
> ipmi_si: Trying default-specified bt state machine at i/o address 0xe4,
> slave address 0x0, irq 0 ipmi_si: Interface detection failed
> ipmi_si: Unable to find any System Interface(s)
> 
> IPMI System Interface driver.
> BUG: unable to handle kernel paging request at c9ccb990
> IP: [<c1380c98>] put_driver+0x8/0x20
> *pde = 05e2f067 *pte = 00000000
> Oops: 0000 [#1]
> last sysfs file: /sys/module/ipmi_msghandler/initstate
> Modules linked in: ipmi_si(+) ipmi_msghandler snd_seq_dummy snd_seq_oss
> snd_seq_midi_event snd_seq snd_pcm_oss snd_mixer_oss ipv6 pcmcia
> pcmcia_core lp fuse snd_ens1371 snd_via82xx gameport snd_ac97_codec
> ac97_bus snd_pcm ppdev snd_timer snd_mpu401_uart snd_rawmidi
> snd_seq_device snd via686a snd_page_alloc hwmon soundcore parport_pc
> shpchp rtc_cmos psmouse rtc_core parport rtc_lib button evdev serio_raw sg
> [last unloaded: pcmcia_core]
> 
> Pid: 1653, comm: insmod Not tainted 2.6.37 #1 8363-686A/VT8363
> EIP: 0060:[<c1380c98>] EFLAGS: 00010286 CPU: 0
> EIP is at put_driver+0x8/0x20
> EAX: c9ccb960 EBX: c12cc2e0 ECX: 00000000 EDX: c9ccb960
> ESI: c9ceb960 EDI: 00000000 EBP: c5e8fee0 ESP: c5e8fee0
>  DS: 007b ES: 007b FS: 0000 GS: 00e0 SS: 0068
> Process insmod (pid: 1653, ti=c5e8e000 task=c6e0a000 task.ti=c5e8e000)
> Stack:
>  c5e8ff04 c1380db5 c1118ba3 c9ceb8c0 c5e8fef8 c1380e99 00000000 00000004
>  00000000 c5e8ff0c c12cc49c c5e8ff58 c9ce9640 c9cea71c 000001a4 00000000
>  c9ceb65c c19fc3f8 fffffffc 00000000 c5e8ff38 c107b185 c5e8ff54 c105063c
> Call Trace:
>  [<c1380db5>] ? driver_register+0x105/0x130
>  [<c1118ba3>] ? sysfs_create_file+0x33/0x40
>  [<c1380e99>] ? driver_create_file+0x19/0x20
>  [<c12cc49c>] ? pnp_register_driver+0x1c/0x20
>  [<c9ce9640>] ? init_ipmi_si+0x294/0x88f [ipmi_si]
>  [<c107b185>] ? tracepoint_module_notify+0x25/0x30
>  [<c105063c>] ? notifier_call_chain+0x4c/0x70
>  [<c1001206>] ? do_one_initcall+0x116/0x160
>  [<c9ce93ac>] ? init_ipmi_si+0x0/0x88f [ipmi_si]
>  [<c1062f41>] ? sys_init_module+0x91/0x1d0
>  [<c17b7ebc>] ? syscall_call+0x7/0xb
> Code: 01 00 00 c7 04 24 c7 1c 91 c1 e8 64 14 cb ff 83 c4 0c 5b 5e c9 c3 8d
> b6 00 00 00 00 8d bc 27 00 00 00 00 55 89 e5 3e 8d 74 26 00 <8b> 40 30 e8
> 70 18 ee ff c9 c3 8d b4 26 00 00 00 00 8d bc 27 00 EIP: [<c1380c98>]
> put_driver+0x8/0x20 SS:ESP 0068:c5e8fee0
> CR2: 00000000c9ccb990
> ---[ end trace 9b4a0dc647b099c4 ]---
> 
> Since the OOPS occurs in the current 2.6.37 release, I added stable on
> CC.
> The code exists in this form since 27d0567ab6 (2009-12-17), but could
> not reproduce it in 2.6.33.
> 
> KernelVersion: 2.6.37
> Cc: <stable@kernel.org>
> Signed-off-by: Peter Huewe <peterhuewe@gmx.de>
> ---
>  drivers/char/ipmi/ipmi_si_intf.c |    5 +++++
>  1 files changed, 5 insertions(+), 0 deletions(-)
> 
> diff --git a/drivers/char/ipmi/ipmi_si_intf.c
> b/drivers/char/ipmi/ipmi_si_intf.c index 035da9e..26c46af 100644
> --- a/drivers/char/ipmi/ipmi_si_intf.c
> +++ b/drivers/char/ipmi/ipmi_si_intf.c
> @@ -3440,6 +3440,11 @@ static int __devinit init_ipmi_si(void)
>  			pci_unregister_driver(&ipmi_pci_driver);
>  #endif
> 
> +#ifdef CONFIG_ACPI
> +		if (pnp_registered)
> +			pnp_unregister_driver(&ipmi_pnp_driver);
> +#endif
> +
>  #ifdef CONFIG_PPC_OF
>  		if (of_registered)
>  			of_unregister_platform_driver(&ipmi_of_platform_driver);


Ping? 
Any updates on this one?

Thanks,
Peter

[PATCH] char/ipmi: fix OOPS caused by pnp_unregister_driver on unregistered driver

[Corey Minyard]

This patch fixes an OOPS triggered when calling modprobe ipmi_si a
second time after the first modprobe returned without finding any ipmi
devices.  This can happen if you reload the module after having the
first module load fail.  The driver was not deregistering from
PNP in that case.

Peter Huewe originally reported this patch and supplied a fix, I have
a different patch based on Linus' suggestion that cleans things up a
bit more.

KernelVersion: 2.6.37
Cc: <stable@kernel.org>
Cc: <openipmi-developer@lists.sourceforge.net>
Cc: Peter Huewe <peterhuewe@gmx.de>
Cc: Randy Dunlap <randy.dunlap@oracle.com>
Signed-off-by: Corey Minyard <cminyard@mvista.com>
---
 drivers/char/ipmi/ipmi_si_intf.c |   12 ++----------
 1 files changed, 2 insertions(+), 10 deletions(-)

diff --git a/drivers/char/ipmi/ipmi_si_intf.c b/drivers/char/ipmi/ipmi_si_intf.c
index b6ae6e9..7855f9f 100644
--- a/drivers/char/ipmi/ipmi_si_intf.c
+++ b/drivers/char/ipmi/ipmi_si_intf.c
@@ -320,6 +320,7 @@ static int unload_when_empty = 1;
 static int add_smi(struct smi_info *smi);
 static int try_smi_init(struct smi_info *smi);
 static void cleanup_one_si(struct smi_info *to_clean);
+static void cleanup_ipmi_si(void);
 
 static ATOMIC_NOTIFIER_HEAD(xaction_notifier_list);
 static int register_xaction_notifier(struct notifier_block *nb)
@@ -3450,16 +3451,7 @@ static int __devinit init_ipmi_si(void)
 	mutex_lock(&smi_infos_lock);
 	if (unload_when_empty && list_empty(&smi_infos)) {
 		mutex_unlock(&smi_infos_lock);
-#ifdef CONFIG_PCI
-		if (pci_registered)
-			pci_unregister_driver(&ipmi_pci_driver);
-#endif
-
-#ifdef CONFIG_PPC_OF
-		if (of_registered)
-			of_unregister_platform_driver(&ipmi_of_platform_driver);
-#endif
-		driver_unregister(&ipmi_driver.driver);
+		cleanup_ipmi_si();
 		printk(KERN_WARNING PFX
 		       "Unable to find any System Interface(s)\n");
 		return -ENODEV;
-- 
1.6.3.3.333.gdf68

Re: [PATCH] char/ipmi: fix OOPS caused by pnp_unregister_driver on unregistered driver

[Peter Hüwe]

Am Donnerstag 10 Februar 2011, 23:08:38 schrieb Corey Minyard:
> This patch fixes an OOPS triggered when calling modprobe ipmi_si a
> second time after the first modprobe returned without finding any ipmi
> devices.  This can happen if you reload the module after having the
> first module load fail.  The driver was not deregistering from
> PNP in that case.
> 
> Peter Huewe originally reported this patch and supplied a fix, I have
> a different patch based on Linus' suggestion that cleans things up a
> bit more.
> 
> KernelVersion: 2.6.37
> Cc: <stable@kernel.org>
> Cc: <openipmi-developer@lists.sourceforge.net>
> Cc: Peter Huewe <peterhuewe@gmx.de>
> Cc: Randy Dunlap <randy.dunlap@oracle.com>
> Signed-off-by: Corey Minyard <cminyard@mvista.com>
> ---
>  drivers/char/ipmi/ipmi_si_intf.c |   12 ++----------
>  1 files changed, 2 insertions(+), 10 deletions(-)
> 
> diff --git a/drivers/char/ipmi/ipmi_si_intf.c
> b/drivers/char/ipmi/ipmi_si_intf.c index b6ae6e9..7855f9f 100644
> --- a/drivers/char/ipmi/ipmi_si_intf.c
> +++ b/drivers/char/ipmi/ipmi_si_intf.c
> @@ -320,6 +320,7 @@ static int unload_when_empty = 1;
>  static int add_smi(struct smi_info *smi);
>  static int try_smi_init(struct smi_info *smi);
>  static void cleanup_one_si(struct smi_info *to_clean);
> +static void cleanup_ipmi_si(void);
> 
>  static ATOMIC_NOTIFIER_HEAD(xaction_notifier_list);
>  static int register_xaction_notifier(struct notifier_block *nb)
> @@ -3450,16 +3451,7 @@ static int __devinit init_ipmi_si(void)
>  	mutex_lock(&smi_infos_lock);
>  	if (unload_when_empty && list_empty(&smi_infos)) {
>  		mutex_unlock(&smi_infos_lock);
> -#ifdef CONFIG_PCI
> -		if (pci_registered)
> -			pci_unregister_driver(&ipmi_pci_driver);
> -#endif
> -
> -#ifdef CONFIG_PPC_OF
> -		if (of_registered)
> -			of_unregister_platform_driver(&ipmi_of_platform_driver);
> -#endif
> -		driver_unregister(&ipmi_driver.driver);
> +		cleanup_ipmi_si();
>  		printk(KERN_WARNING PFX
>  		       "Unable to find any System Interface(s)\n");
>  		return -ENODEV;

Looks good to me - thanks.

Reviewed-by: Peter Huewe <peterhuewe@gmx.de>