2.6.18-rc5-mm1: BUG in pci_disable_msi on attempts to rmmod snd_hda_intel (64-bit kernel)

2.6.18-rc5-mm1: BUG in pci_disable_msi on attempts to rmmod snd_hda_intel (64-bit kernel)

[Rafael J. Wysocki]

Hi,

I get the appended BUG when I try to rmmod snd_hda_intel on 2.6.18-rc5-mm1
(the box is an HPC nx6325).

The driver also breaks the suspend to disk (causes an oops to appear during
the suspend phase, but I haven't been able to catch it yet).

Greetings,
Rafael


Kernel BUG at drivers/pci/msi.c:910
invalid opcode: 0000 [1] SMP 
last sysfs file: /power/disk
CPU 0 
Modules linked in: cpufreq_ondemand cpufreq_userspace cpufreq_powersave powernow_k8 freq_table af_packet button edd snd_pcm_oss snd_mixer_oss battery snd_seq snd_seq_device ac ip6t_REJECT xt_tcpudp ipt_REJECT xt_state iptable_mangle iptable_nat ip_nat iptable_filter ip6table_mangle ip_conntrack nfnetlink ip_tables ip6table_filter ip6_tables x_tables ipv6 loop dm_mod usbhid ff_memless hci_usb bluetooth snd_hda_intel bcm43xx shpchp snd_hda_codec pcmcia pci_hotplug snd_pcm firmware_class ieee80211softmac ohci1394 sdhci snd_timer snd soundcore snd_page_alloc ieee1394 mmc_core ohci_hcd i2c_piix4 yenta_socket rsrc_nonstatic pcmcia_core ide_cd i2c_core cdrom k8temp hwmon ehci_hcd usbcore ieee80211 ieee80211_crypt tg3 ext3 jbd fan thermal processor atiixp ide_disk ide_core sg
Pid: 3684, comm: rmmod Tainted: G   M  2.6.18-rc5-mm1 #14
RIP: 0010:[<ffffffff80326c7b>]  [<ffffffff80326c7b>] pci_disable_msi+0xdf/0x11a
RSP: 0018:ffff810035559c68  EFLAGS: 00010202
RAX: 0000000000000052 RBX: ffff810037145620 RCX: ffffffff804e7048
RDX: ffff810035559a78 RSI: 0000000000000092 RDI: ffffffff804e7040
RBP: ffff810035559c88 R08: ffffffff804e7048 R09: 0000000000000002
R10: ffff810035559938 R11: 0000000000000000 R12: ffff810037ba7678
R13: 0000000000000008 R14: 0000000000000880 R15: 00007fff6dc093c0
FS:  00002ba63d1ea6d0(0000) GS:ffffffff8064e000(0000) knlGS:0000000000000000
CS:  0010 DS: 0000 ES: 0000 CR0: 000000008005003b
CR2: 00002ba63d0c6e40 CR3: 000000002f1f1000 CR4: 00000000000006e0
Process rmmod (pid: 3684, threadinfo ffff810035558000, task ffff810034552790)
Stack:  ffff810035559c78 0081ffff8025e80c ffff810035588b50 0000000000000280
 ffff810035559cb8 ffffffff882f4405 ffff810035559cc8 ffff810034a25878
 0000000000002000 0000000000002fff ffff810035559cc8 ffffffff882f449a
Call Trace:
 [<ffffffff882f4405>] :snd_hda_intel:azx_free+0xc3/0x14b
 [<ffffffff882f449a>] :snd_hda_intel:azx_dev_free+0xd/0xf
 [<ffffffff881a688b>] :snd:snd_device_free+0xea/0x157
 [<ffffffff881a696b>] :snd:snd_device_free_all+0x73/0x98
 [<ffffffff881a1735>] :snd:snd_card_do_free+0x7f/0x10f
 [<ffffffff881a21ef>] :snd:snd_card_free+0xa4/0xb4
 [<ffffffff882f4321>] :snd_hda_intel:azx_remove+0x18/0x27
 [<ffffffff80320610>] pci_device_remove+0x25/0x3c
 [<ffffffff80377671>] __device_release_driver+0x80/0x9c
 [<ffffffff80377acb>] driver_detach+0xac/0xee
 [<ffffffff80376b74>] bus_remove_driver+0x75/0x97
 [<ffffffff80377b4a>] driver_unregister+0x15/0x21
 [<ffffffff803207be>] pci_unregister_driver+0x13/0x8e
 [<ffffffff882f5988>] :snd_hda_intel:alsa_card_azx_exit+0x10/0x12
 [<ffffffff8029ada1>] sys_delete_module+0x1b5/0x1e6
 [<ffffffff8025910e>] system_call+0x7e/0x83
DWARF2 unwinder stuck at system_call+0x7e/0x83

Leftover inexact backtrace:



Code: 0f 0b 68 9e c3 42 80 c2 8e 03 eb 27 48 c7 c7 44 98 50 80 8b 
RIP  [<ffffffff80326c7b>] pci_disable_msi+0xdf/0x11a
 RSP <ffff810035559c68>
 


-- 
You never change things by fighting the existing reality.
		R. Buckminster Fuller

-------------------------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642

Re: 2.6.18-rc5-mm1: BUG in pci_disable_msi on attempts to rmmod snd_hda_intel (64-bit kernel)

[Andrew Morton]

On Wed, 6 Sep 2006 22:59:53 +0200
"Rafael J. Wysocki" <rjw@sisk.pl> wrote:

> Hi,
> 
> I get the appended BUG when I try to rmmod snd_hda_intel on 2.6.18-rc5-mm1
> (the box is an HPC nx6325).
> 
> The driver also breaks the suspend to disk (causes an oops to appear during
> the suspend phase, but I haven't been able to catch it yet).
> 

I'm told that the MSI patehs in rc5-mm1 are no good and have been redone
and replaced.  So could you please remember to retest next -mm?  Fingers
crossed, it'll be fixed then.

(Where did the taint come from?)

Thanks.

> 
> Kernel BUG at drivers/pci/msi.c:910
> invalid opcode: 0000 [1] SMP 
> last sysfs file: /power/disk
> CPU 0 
> Modules linked in: cpufreq_ondemand cpufreq_userspace cpufreq_powersave powernow_k8 freq_table af_packet button edd snd_pcm_oss snd_mixer_oss battery snd_seq snd_seq_device ac ip6t_REJECT xt_tcpudp ipt_REJECT xt_state iptable_mangle iptable_nat ip_nat iptable_filter ip6table_mangle ip_conntrack nfnetlink ip_tables ip6table_filter ip6_tables x_tables ipv6 loop dm_mod usbhid ff_memless hci_usb bluetooth snd_hda_intel bcm43xx shpchp snd_hda_codec pcmcia pci_hotplug snd_pcm firmware_class ieee80211softmac ohci1394 sdhci snd_timer snd soundcore snd_page_alloc ieee1394 mmc_core ohci_hcd i2c_piix4 yenta_socket rsrc_nonstatic pcmcia_core ide_cd i2c_core cdrom k8temp hwmon ehci_hcd usbcore ieee80211 ieee80211_crypt tg3 ext3 jbd fan thermal processor atiixp ide_disk ide_core sg
> Pid: 3684, comm: rmmod Tainted: G   M  2.6.18-rc5-mm1 #14
> RIP: 0010:[<ffffffff80326c7b>]  [<ffffffff80326c7b>] pci_disable_msi+0xdf/0x11a
> RSP: 0018:ffff810035559c68  EFLAGS: 00010202
> RAX: 0000000000000052 RBX: ffff810037145620 RCX: ffffffff804e7048
> RDX: ffff810035559a78 RSI: 0000000000000092 RDI: ffffffff804e7040
> RBP: ffff810035559c88 R08: ffffffff804e7048 R09: 0000000000000002
> R10: ffff810035559938 R11: 0000000000000000 R12: ffff810037ba7678
> R13: 0000000000000008 R14: 0000000000000880 R15: 00007fff6dc093c0
> FS:  00002ba63d1ea6d0(0000) GS:ffffffff8064e000(0000) knlGS:0000000000000000
> CS:  0010 DS: 0000 ES: 0000 CR0: 000000008005003b
> CR2: 00002ba63d0c6e40 CR3: 000000002f1f1000 CR4: 00000000000006e0
> Process rmmod (pid: 3684, threadinfo ffff810035558000, task ffff810034552790)
> Stack:  ffff810035559c78 0081ffff8025e80c ffff810035588b50 0000000000000280
>  ffff810035559cb8 ffffffff882f4405 ffff810035559cc8 ffff810034a25878
>  0000000000002000 0000000000002fff ffff810035559cc8 ffffffff882f449a
> Call Trace:
>  [<ffffffff882f4405>] :snd_hda_intel:azx_free+0xc3/0x14b
>  [<ffffffff882f449a>] :snd_hda_intel:azx_dev_free+0xd/0xf
>  [<ffffffff881a688b>] :snd:snd_device_free+0xea/0x157
>  [<ffffffff881a696b>] :snd:snd_device_free_all+0x73/0x98
>  [<ffffffff881a1735>] :snd:snd_card_do_free+0x7f/0x10f
>  [<ffffffff881a21ef>] :snd:snd_card_free+0xa4/0xb4
>  [<ffffffff882f4321>] :snd_hda_intel:azx_remove+0x18/0x27
>  [<ffffffff80320610>] pci_device_remove+0x25/0x3c
>  [<ffffffff80377671>] __device_release_driver+0x80/0x9c
>  [<ffffffff80377acb>] driver_detach+0xac/0xee
>  [<ffffffff80376b74>] bus_remove_driver+0x75/0x97
>  [<ffffffff80377b4a>] driver_unregister+0x15/0x21
>  [<ffffffff803207be>] pci_unregister_driver+0x13/0x8e
>  [<ffffffff882f5988>] :snd_hda_intel:alsa_card_azx_exit+0x10/0x12
>  [<ffffffff8029ada1>] sys_delete_module+0x1b5/0x1e6
>  [<ffffffff8025910e>] system_call+0x7e/0x83
> DWARF2 unwinder stuck at system_call+0x7e/0x83
> 
> Leftover inexact backtrace:
> 
> 
> 
> Code: 0f 0b 68 9e c3 42 80 c2 8e 03 eb 27 48 c7 c7 44 98 50 80 8b 
> RIP  [<ffffffff80326c7b>] pci_disable_msi+0xdf/0x11a
>  RSP <ffff810035559c68>
>  
> 
> 
> -- 
> You never change things by fighting the existing reality.
> 		R. Buckminster Fuller

-------------------------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642

Re: 2.6.18-rc5-mm1: BUG in pci_disable_msi on attempts to rmmod snd_hda_intel (64-bit kernel)

[Greg KH]

On Wed, Sep 06, 2006 at 08:24:29PM -0700, Andrew Morton wrote:
> On Wed, 6 Sep 2006 22:59:53 +0200
> "Rafael J. Wysocki" <rjw@sisk.pl> wrote:
> 
> > Hi,
> > 
> > I get the appended BUG when I try to rmmod snd_hda_intel on 2.6.18-rc5-mm1
> > (the box is an HPC nx6325).
> > 
> > The driver also breaks the suspend to disk (causes an oops to appear during
> > the suspend phase, but I haven't been able to catch it yet).
> > 
> 
> I'm told that the MSI patehs in rc5-mm1 are no good and have been redone
> and replaced.  So could you please remember to retest next -mm?  Fingers
> crossed, it'll be fixed then.

Yes, the msi code was redone again, so please test the next -mm release
for this.

thanks,

greg k-h

-------------------------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642

Re: 2.6.18-rc5-mm1: BUG in pci_disable_msi on attempts to rmmod snd_hda_intel (64-bit kernel)

[Rafael J. Wysocki]

On Thursday, 7 September 2006 05:24, Andrew Morton wrote:
> On Wed, 6 Sep 2006 22:59:53 +0200
> "Rafael J. Wysocki" <rjw@sisk.pl> wrote:
> 
> > Hi,
> > 
> > I get the appended BUG when I try to rmmod snd_hda_intel on 2.6.18-rc5-mm1
> > (the box is an HPC nx6325).
> > 
> > The driver also breaks the suspend to disk (causes an oops to appear during
> > the suspend phase, but I haven't been able to catch it yet).
> > 
> 
> I'm told that the MSI patehs in rc5-mm1 are no good and have been redone
> and replaced.  So could you please remember to retest next -mm?  Fingers
> crossed, it'll be fixed then.

I will.

> (Where did the taint come from?)

No idea.  The system is an out-of-the-box SUSE 10.1 + current updates + the
kernel.

Greetings,
Rafael


-- 
You never change things by fighting the existing reality.
		R. Buckminster Fuller

-------------------------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642

Re: 2.6.18-rc5-mm1: BUG in pci_disable_msi on attempts to rmmod snd_hda_intel (64-bit kernel)

[Takashi Iwai]

At Wed, 6 Sep 2006 20:24:29 -0700,
Andrew Morton wrote:
> 
> On Wed, 6 Sep 2006 22:59:53 +0200
> "Rafael J. Wysocki" <rjw@sisk.pl> wrote:
> 
> > Hi,
> > 
> > I get the appended BUG when I try to rmmod snd_hda_intel on 2.6.18-rc5-mm1
> > (the box is an HPC nx6325).
> > 
> > The driver also breaks the suspend to disk (causes an oops to appear during
> > the suspend phase, but I haven't been able to catch it yet).
> > 
> 
> I'm told that the MSI patehs in rc5-mm1 are no good and have been redone
> and replaced.  So could you please remember to retest next -mm?  Fingers
> crossed, it'll be fixed then.

Hm, then is it safer to disable MSI support in the driver as default,
or is it supposed to be stable enough?
Right now, pci_enable_msi() is always called unless disable_msi=1
option is given explicilty to the driver.


Takashi

-------------------------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642

Re: 2.6.18-rc5-mm1: BUG in pci_disable_msi on attempts to rmmod snd_hda_intel (64-bit kernel)

[Dave Jones]

On Wed, Sep 06, 2006 at 08:24:29PM -0700, Andrew Morton wrote:
 
 > I'm told that the MSI patehs in rc5-mm1 are no good and have been redone
 > and replaced.  So could you please remember to retest next -mm?  Fingers
 > crossed, it'll be fixed then.
 > 
 > > Pid: 3684, comm: rmmod Tainted: G   M  2.6.18-rc5-mm1 #14
 > (Where did the taint come from?)

M = machine check exception happened.

(G=nonsense)

	Dave

-------------------------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642

Re: 2.6.18-rc5-mm1: BUG in pci_disable_msi on attempts to rmmod snd_hda_intel (64-bit kernel)

[Takashi Iwai]

At Wed, 6 Sep 2006 22:59:53 +0200,
Rafael J. Wysocki wrote:
> 
> Hi,
> 
> I get the appended BUG when I try to rmmod snd_hda_intel on 2.6.18-rc5-mm1
> (the box is an HPC nx6325).
> 
> The driver also breaks the suspend to disk (causes an oops to appear during
> the suspend phase, but I haven't been able to catch it yet).

Could you try the patch below?  It's pending on my local tree for
testing...  Also, you can use disable_msi=1 option to avoid MSI.


Takashi

diff -r f38b7e4c44c3 Documentation/sound/alsa/ALSA-Configuration.txt
--- a/sound/alsa/Documentation/ALSA-Configuration.txt	Wed Sep 06 16:03:11 2006 +0200
+++ b/sound/alsa/Documentation/ALSA-Configuration.txt	Thu Sep 07 11:15:07 2006 +0200
@@ -758,6 +758,7 @@ Prior to version 0.9.0rc4 options had a 
     position_fix - Fix DMA pointer (0 = auto, 1 = none, 2 = POSBUF, 3 = FIFO size)
     single_cmd  - Use single immediate commands to communicate with
 		codecs (for debugging only)
+    disable_msi - Disable Message Signaled Interrupt (MSI)
 
     This module supports one card and autoprobe.
 
diff -r f38b7e4c44c3 sound/pci/hda/hda_intel.c
--- a/sound/pci/hda/hda_intel.c	Wed Sep 06 16:03:11 2006 +0200
+++ b/sound/pci/hda/hda_intel.c	Thu Sep 07 11:16:34 2006 +0200
@@ -1422,8 +1422,9 @@ static int azx_free(struct azx *chip)
 	}
 
 	if (chip->irq >= 0) {
-		pci_disable_msi(chip->pci);
 		free_irq(chip->irq, (void*)chip);
+		if (!disable_msi)
+			pci_disable_msi(chip->pci);
 	}
 	if (chip->remap_addr)
 		iounmap(chip->remap_addr);

-------------------------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642

Re: 2.6.18-rc5-mm1: BUG in pci_disable_msi on attempts to rmmod snd_hda_intel (64-bit kernel)

[Rafael J. Wysocki]

On Thursday, 7 September 2006 11:20, Takashi Iwai wrote:
> At Wed, 6 Sep 2006 22:59:53 +0200,
> Rafael J. Wysocki wrote:
> > 
> > Hi,
> > 
> > I get the appended BUG when I try to rmmod snd_hda_intel on 2.6.18-rc5-mm1
> > (the box is an HPC nx6325).
> > 
> > The driver also breaks the suspend to disk (causes an oops to appear during
> > the suspend phase, but I haven't been able to catch it yet).
> 
> Could you try the patch below?  It's pending on my local tree for
> testing...  Also, you can use disable_msi=1 option to avoid MSI.

The patch fixes the rmmod issue.

Also setting disable_msi=1 fixes the suspend issue.

If the MSI is not disabled, the suspend fails with an oops (on an attempt to
access disk after the resume-during-suspend), but this will hopefully get
fixed with the new MSI code.

Thanks,
Rafael


-- 
You never change things by fighting the existing reality.
		R. Buckminster Fuller

-------------------------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642

Re: 2.6.18-rc5-mm1: BUG in pci_disable_msi on attempts to rmmod snd_hda_intel (64-bit kernel)

[Takashi Iwai]

At Thu, 7 Sep 2006 12:29:54 +0200,
Rafael J. Wysocki wrote:
> 
> On Thursday, 7 September 2006 11:20, Takashi Iwai wrote:
> > At Wed, 6 Sep 2006 22:59:53 +0200,
> > Rafael J. Wysocki wrote:
> > > 
> > > Hi,
> > > 
> > > I get the appended BUG when I try to rmmod snd_hda_intel on 2.6.18-rc5-mm1
> > > (the box is an HPC nx6325).
> > > 
> > > The driver also breaks the suspend to disk (causes an oops to appear during
> > > the suspend phase, but I haven't been able to catch it yet).
> > 
> > Could you try the patch below?  It's pending on my local tree for
> > testing...  Also, you can use disable_msi=1 option to avoid MSI.
> 
> The patch fixes the rmmod issue.
> 
> Also setting disable_msi=1 fixes the suspend issue.
> 
> If the MSI is not disabled, the suspend fails with an oops (on an attempt to
> access disk after the resume-during-suspend), but this will hopefully get
> fixed with the new MSI code.

The PM problem might be fixed by the patch below.  Give it a try.


Takashi

diff -r 4f27a5064cda sound/pci/hda/hda_intel.c
--- a/sound/pci/hda/hda_intel.c	Thu Sep 07 12:40:00 2006 +0200
+++ b/sound/pci/hda/hda_intel.c	Thu Sep 07 15:20:56 2006 +0200
@@ -1375,6 +1375,10 @@ static int azx_suspend(struct pci_dev *p
 		snd_pcm_suspend_all(chip->pcm[i]);
 	snd_hda_suspend(chip->bus, state);
 	azx_free_cmd_io(chip);
+	if (chip->irq >= 0)
+		free_irq(chip->irq, chip);
+	if (!disable_msi)
+		pci_disable_msi(chip->pci);
 	pci_disable_device(pci);
 	pci_save_state(pci);
 	return 0;
@@ -1387,6 +1391,12 @@ static int azx_resume(struct pci_dev *pc
 
 	pci_restore_state(pci);
 	pci_enable_device(pci);
+	if (!disable_msi)
+		pci_enable_msi(pci);
+	/* FIXME: need proper error handling */
+	request_irq(pci->irq, azx_interrupt, IRQF_DISABLED|IRQF_SHARED,
+		    "HDA Intel", chip);
+	chip->irq = pci->irq;
 	pci_set_master(pci);
 	azx_init_chip(chip);
 	snd_hda_resume(chip->bus);

-------------------------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642

Re: 2.6.18-rc5-mm1: BUG in pci_disable_msi on attempts to rmmod snd_hda_intel (64-bit kernel)

[Rafael J. Wysocki]

On Thursday, 7 September 2006 15:17, Takashi Iwai wrote:
> At Thu, 7 Sep 2006 12:29:54 +0200,
> Rafael J. Wysocki wrote:
> > 
> > On Thursday, 7 September 2006 11:20, Takashi Iwai wrote:
> > > At Wed, 6 Sep 2006 22:59:53 +0200,
> > > Rafael J. Wysocki wrote:
> > > > 
> > > > Hi,
> > > > 
> > > > I get the appended BUG when I try to rmmod snd_hda_intel on 2.6.18-rc5-mm1
> > > > (the box is an HPC nx6325).
> > > > 
> > > > The driver also breaks the suspend to disk (causes an oops to appear during
> > > > the suspend phase, but I haven't been able to catch it yet).
> > > 
> > > Could you try the patch below?  It's pending on my local tree for
> > > testing...  Also, you can use disable_msi=1 option to avoid MSI.
> > 
> > The patch fixes the rmmod issue.
> > 
> > Also setting disable_msi=1 fixes the suspend issue.
> > 
> > If the MSI is not disabled, the suspend fails with an oops (on an attempt to
> > access disk after the resume-during-suspend), but this will hopefully get
> > fixed with the new MSI code.
> 
> The PM problem might be fixed by the patch below.  Give it a try.

Yes, the patch fixes the problem.

Thanks,
Rafael


-- 
You never change things by fighting the existing reality.
		R. Buckminster Fuller

-------------------------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642

Re: 2.6.18-rc5-mm1: BUG in pci_disable_msi on attempts to rmmod snd_hda_intel (64-bit kernel)

[Takashi Iwai]

At Thu, 7 Sep 2006 23:14:41 +0200,
Rafael J. Wysocki wrote:
> 
> On Thursday, 7 September 2006 15:17, Takashi Iwai wrote:
> > At Thu, 7 Sep 2006 12:29:54 +0200,
> > Rafael J. Wysocki wrote:
> > > 
> > > On Thursday, 7 September 2006 11:20, Takashi Iwai wrote:
> > > > At Wed, 6 Sep 2006 22:59:53 +0200,
> > > > Rafael J. Wysocki wrote:
> > > > > 
> > > > > Hi,
> > > > > 
> > > > > I get the appended BUG when I try to rmmod snd_hda_intel on 2.6.18-rc5-mm1
> > > > > (the box is an HPC nx6325).
> > > > > 
> > > > > The driver also breaks the suspend to disk (causes an oops to appear during
> > > > > the suspend phase, but I haven't been able to catch it yet).
> > > > 
> > > > Could you try the patch below?  It's pending on my local tree for
> > > > testing...  Also, you can use disable_msi=1 option to avoid MSI.
> > > 
> > > The patch fixes the rmmod issue.
> > > 
> > > Also setting disable_msi=1 fixes the suspend issue.
> > > 
> > > If the MSI is not disabled, the suspend fails with an oops (on an attempt to
> > > access disk after the resume-during-suspend), but this will hopefully get
> > > fixed with the new MSI code.
> > 
> > The PM problem might be fixed by the patch below.  Give it a try.
> 
> Yes, the patch fixes the problem.

Glad to hear!

Jaroslav, could you update alsa-current.git again?
I already committed the patches to HG tree.


thanks,

Takashi

-------------------------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642

Re: 2.6.18-rc5-mm1: BUG in pci_disable_msi on attempts to rmmod snd_hda_intel (64-bit kernel)

[Jaroslav Kysela]

On Fri, 8 Sep 2006, Takashi Iwai wrote:

> At Thu, 7 Sep 2006 23:14:41 +0200,
> Rafael J. Wysocki wrote:
> > 
> > On Thursday, 7 September 2006 15:17, Takashi Iwai wrote:
> > > At Thu, 7 Sep 2006 12:29:54 +0200,
> > > Rafael J. Wysocki wrote:
> > > > 
> > > > On Thursday, 7 September 2006 11:20, Takashi Iwai wrote:
> > > > > At Wed, 6 Sep 2006 22:59:53 +0200,
> > > > > Rafael J. Wysocki wrote:
> > > > > > 
> > > > > > Hi,
> > > > > > 
> > > > > > I get the appended BUG when I try to rmmod snd_hda_intel on 2.6.18-rc5-mm1
> > > > > > (the box is an HPC nx6325).
> > > > > > 
> > > > > > The driver also breaks the suspend to disk (causes an oops to appear during
> > > > > > the suspend phase, but I haven't been able to catch it yet).
> > > > > 
> > > > > Could you try the patch below?  It's pending on my local tree for
> > > > > testing...  Also, you can use disable_msi=1 option to avoid MSI.
> > > > 
> > > > The patch fixes the rmmod issue.
> > > > 
> > > > Also setting disable_msi=1 fixes the suspend issue.
> > > > 
> > > > If the MSI is not disabled, the suspend fails with an oops (on an attempt to
> > > > access disk after the resume-during-suspend), but this will hopefully get
> > > > fixed with the new MSI code.
> > > 
> > > The PM problem might be fixed by the patch below.  Give it a try.
> > 
> > Yes, the patch fixes the problem.
> 
> Glad to hear!
> 
> Jaroslav, could you update alsa-current.git again?
> I already committed the patches to HG tree.

Done.

						Jaroslav

-----
Jaroslav Kysela <perex@suse.cz>
Linux Kernel Sound Maintainer
ALSA Project, SUSE Labs

-------------------------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642

2.6.18-rc5-mm1: BUG in pci_disable_msi on attempts to rmmod snd_hda_intel (64-bit kernel)

[Rafael J. Wysocki]

Hi,

I get the appended BUG when I try to rmmod snd_hda_intel on 2.6.18-rc5-mm1
(the box is an HPC nx6325).

The driver also breaks the suspend to disk (causes an oops to appear during
the suspend phase, but I haven't been able to catch it yet).

Greetings,
Rafael


Kernel BUG at drivers/pci/msi.c:910
invalid opcode: 0000 [1] SMP 
last sysfs file: /power/disk
CPU 0 
Modules linked in: cpufreq_ondemand cpufreq_userspace cpufreq_powersave powernow_k8 freq_table af_packet button edd snd_pcm_oss snd_mixer_oss battery snd_seq snd_seq_device ac ip6t_REJECT xt_tcpudp ipt_REJECT xt_state iptable_mangle iptable_nat ip_nat iptable_filter ip6table_mangle ip_conntrack nfnetlink ip_tables ip6table_filter ip6_tables x_tables ipv6 loop dm_mod usbhid ff_memless hci_usb bluetooth snd_hda_intel bcm43xx shpchp snd_hda_codec pcmcia pci_hotplug snd_pcm firmware_class ieee80211softmac ohci1394 sdhci snd_timer snd soundcore snd_page_alloc ieee1394 mmc_core ohci_hcd i2c_piix4 yenta_socket rsrc_nonstatic pcmcia_core ide_cd i2c_core cdrom k8temp hwmon ehci_hcd usbcore ieee80211 ieee80211_crypt tg3 ext3 jbd fan thermal processor atiixp ide_disk ide_core sg
Pid: 3684, comm: rmmod Tainted: G   M  2.6.18-rc5-mm1 #14
RIP: 0010:[<ffffffff80326c7b>]  [<ffffffff80326c7b>] pci_disable_msi+0xdf/0x11a
RSP: 0018:ffff810035559c68  EFLAGS: 00010202
RAX: 0000000000000052 RBX: ffff810037145620 RCX: ffffffff804e7048
RDX: ffff810035559a78 RSI: 0000000000000092 RDI: ffffffff804e7040
RBP: ffff810035559c88 R08: ffffffff804e7048 R09: 0000000000000002
R10: ffff810035559938 R11: 0000000000000000 R12: ffff810037ba7678
R13: 0000000000000008 R14: 0000000000000880 R15: 00007fff6dc093c0
FS:  00002ba63d1ea6d0(0000) GS:ffffffff8064e000(0000) knlGS:0000000000000000
CS:  0010 DS: 0000 ES: 0000 CR0: 000000008005003b
CR2: 00002ba63d0c6e40 CR3: 000000002f1f1000 CR4: 00000000000006e0
Process rmmod (pid: 3684, threadinfo ffff810035558000, task ffff810034552790)
Stack:  ffff810035559c78 0081ffff8025e80c ffff810035588b50 0000000000000280
 ffff810035559cb8 ffffffff882f4405 ffff810035559cc8 ffff810034a25878
 0000000000002000 0000000000002fff ffff810035559cc8 ffffffff882f449a
Call Trace:
 [<ffffffff882f4405>] :snd_hda_intel:azx_free+0xc3/0x14b
 [<ffffffff882f449a>] :snd_hda_intel:azx_dev_free+0xd/0xf
 [<ffffffff881a688b>] :snd:snd_device_free+0xea/0x157
 [<ffffffff881a696b>] :snd:snd_device_free_all+0x73/0x98
 [<ffffffff881a1735>] :snd:snd_card_do_free+0x7f/0x10f
 [<ffffffff881a21ef>] :snd:snd_card_free+0xa4/0xb4
 [<ffffffff882f4321>] :snd_hda_intel:azx_remove+0x18/0x27
 [<ffffffff80320610>] pci_device_remove+0x25/0x3c
 [<ffffffff80377671>] __device_release_driver+0x80/0x9c
 [<ffffffff80377acb>] driver_detach+0xac/0xee
 [<ffffffff80376b74>] bus_remove_driver+0x75/0x97
 [<ffffffff80377b4a>] driver_unregister+0x15/0x21
 [<ffffffff803207be>] pci_unregister_driver+0x13/0x8e
 [<ffffffff882f5988>] :snd_hda_intel:alsa_card_azx_exit+0x10/0x12
 [<ffffffff8029ada1>] sys_delete_module+0x1b5/0x1e6
 [<ffffffff8025910e>] system_call+0x7e/0x83
DWARF2 unwinder stuck at system_call+0x7e/0x83

Leftover inexact backtrace:



Code: 0f 0b 68 9e c3 42 80 c2 8e 03 eb 27 48 c7 c7 44 98 50 80 8b 
RIP  [<ffffffff80326c7b>] pci_disable_msi+0xdf/0x11a
 RSP <ffff810035559c68>
 


-- 
You never change things by fighting the existing reality.
		R. Buckminster Fuller