Skip to contentSkip to conversation

double fault in 2.6.19rc4-git5 while unplugging/replugging a USB headset

DISCUSSION Andi Kleen 2 participants 5 messages Nov 6, 2006 – Nov 7, 2006 Sound

lore
Share & tools
lore

Andi Kleen

author
#0 lore raw

FYI

usb 5-2: USB disconnect, address 2
ALSA /home/lsrc/quilt/linux/sound/usb/usbaudio.c:845: cannot submit datapipe for urb 0, error -19: no device [... repeated a lot ...]
ACPI: PCI interrupt for device 0000:00:1b.0 disabled
usb 5-2: new full speed USB device using uhci_hcd and address 3
usb 5-2: configuration #1 chosen from 1 choice
input: C-Media USB Headphone Set as /class/input/input3
input: USB HID v1.00 Device [C-Media USB Headphone Set ] on usb-0000:00:1d.3-2
usb 5-2: USB disconnect, address 3

Unable to handle kernel paging request at ffffffff880e57d4 RIP: 
 [<ffffffff880e57d4>]
PGD 203027 PUD 205027 PMD 7f41e067 PTE 0
Oops: 0010 [1] SMP 
CPU 0 
Modules linked in: snd_pcm_oss snd_mixer_oss snd_seq snd_atiixp snd_ac97_codec snd_ac97_bus speedstep_centrino pppoe pppox ppp_generic slhc ipt_MASQUERADE iptable_nat ip_nat_h323 ip_conntrack_h323 ip_nat_sip ip_conntrack_sip ip_nat_ftp ip_nat_irc i2c_isa ip_nat ip6t_LOG ip6t_REJECT ip6table_filter ip6_tables xt_tcpudp xt_state ip_conntrack_ftp ip_conntrack_irc ip_conntrack ipt_LOG ipt_REJECT iptable_filter ip_tables x_tables pl2303 snd_usb_audio snd_pcm snd_timer snd_page_alloc snd_usb_lib snd_rawmidi snd_seq_device snd_hwdep snd usbserial i2c_i801 i2c_core
Pid: 8104, comm: xmms Not tainted 2.6.19-rc4-git5-BASIL #99
RIP: 0010:[<ffffffff880e57d4>]  [<ffffffff880e57d4>]
RSP: 0018:ffff81004730fd50  EFLAGS: 00010286
RAX: ffffffff880e57d4 RBX: ffff810075368000 RCX: 0000000000000000
RDX: 000000000000001d RSI: 0000000000000002 RDI: ffff810075368230
RBP: ffff810075368230 R08: ffff81007a5711f0 R09: ffff81005e5345c0
R10: 00000000454e8c24 R11: ffff81007bd030b8 R12: 00000000000001f8
R13: ffff810057422460 R14: ffff810045b7a5c0 R15: ffff81007a4b9080
FS:  0000000000000000(0000) GS:ffffffff80662000(0000) knlGS:0000000000000000
CS:  0010 DS: 0000 ES: 0000 CR0: 000000008005003b
CR2: ffffffff880e57d4 CR3: 0000000000201000 CR4: 00000000000006e0
Process xmms (pid: 8104, threadinfo ffff81004730e000, task ffff810076fee180)
Stack:  ffffffff880a2cd6 0000000000000000 0000000000000001 ffff81007604bcf0
 ffffffff8024c42b 0000000000000000 ffffffff8058b290 0000001f45880508
 ffff810001f3fe50 0000000300000000 ffff81007a5711e8 0000000000000000
Call Trace:
 [<ffffffff880a2cd6>] :snd_mixer_oss:snd_mixer_oss_notify_handler+0x23e/0x271
 [<ffffffff8024c42b>] free_pages_bulk+0x2c/0x1f4
 [<ffffffff802229d2>] __wake_up+0x36/0x4d
 [<ffffffff880166d7>] :snd:snd_card_do_free+0x17/0xe4
 [<ffffffff88016e3d>] :snd:snd_card_file_remove+0xfd/0x135
 [<ffffffff88041fc7>] :snd_pcm:snd_pcm_release+0xb6/0xc1
 [<ffffffff8026745a>] __fput+0xac/0x18a
 [<ffffffff802653d4>] filp_close+0x5c/0x64
 [<ffffffff8022a913>] put_files_struct+0x6b/0xbd
 [<ffffffff8022ba84>] do_exit+0x233/0x814
 [<ffffffff8022c0eb>] sys_exit_group+0x0/0xe
 [<ffffffff802093fe>] system_call+0x7e/0x83


Code:  Bad RIP value.

RIP [<ffffffff880e57d4>]
RSP <ffff81004730fd50>
CR2: ffffffff880e57d4
<1>Fixing recursive fault but reboot is needed!

-Andi

#1 lore raw

At Mon, 6 Nov 2006 02:20:06 +0100,

Quoted 39 lines · Andi Kleen

FYI

usb 5-2: USB disconnect, address 2
ALSA /home/lsrc/quilt/linux/sound/usb/usbaudio.c:845: cannot submit datapipe for urb 0, error -19: no device [... repeated a lot ...]
ACPI: PCI interrupt for device 0000:00:1b.0 disabled
usb 5-2: new full speed USB device using uhci_hcd and address 3
usb 5-2: configuration #1 chosen from 1 choice
input: C-Media USB Headphone Set as /class/input/input3
input: USB HID v1.00 Device [C-Media USB Headphone Set ] on usb-0000:00:1d.3-2
usb 5-2: USB disconnect, address 3

Unable to handle kernel paging request at ffffffff880e57d4 RIP: 
 [<ffffffff880e57d4>]
PGD 203027 PUD 205027 PMD 7f41e067 PTE 0
Oops: 0010 [1] SMP 
CPU 0 
Modules linked in: snd_pcm_oss snd_mixer_oss snd_seq snd_atiixp snd_ac97_codec snd_ac97_bus speedstep_centrino pppoe pppox ppp_generic slhc ipt_MASQUERADE iptable_nat ip_nat_h323 ip_conntrack_h323 ip_nat_sip ip_conntrack_sip ip_nat_ftp ip_nat_irc i2c_isa ip_nat ip6t_LOG ip6t_REJECT ip6table_filter ip6_tables xt_tcpudp xt_state ip_conntrack_ftp ip_conntrack_irc ip_conntrack ipt_LOG ipt_REJECT iptable_filter ip_tables x_tables pl2303 snd_usb_audio snd_pcm snd_timer snd_page_alloc snd_usb_lib snd_rawmidi snd_seq_device snd_hwdep snd usbserial i2c_i801 i2c_core
Pid: 8104, comm: xmms Not tainted 2.6.19-rc4-git5-BASIL #99
RIP: 0010:[<ffffffff880e57d4>]  [<ffffffff880e57d4>]
RSP: 0018:ffff81004730fd50  EFLAGS: 00010286
RAX: ffffffff880e57d4 RBX: ffff810075368000 RCX: 0000000000000000
RDX: 000000000000001d RSI: 0000000000000002 RDI: ffff810075368230
RBP: ffff810075368230 R08: ffff81007a5711f0 R09: ffff81005e5345c0
R10: 00000000454e8c24 R11: ffff81007bd030b8 R12: 00000000000001f8
R13: ffff810057422460 R14: ffff810045b7a5c0 R15: ffff81007a4b9080
FS:  0000000000000000(0000) GS:ffffffff80662000(0000) knlGS:0000000000000000
CS:  0010 DS: 0000 ES: 0000 CR0: 000000008005003b
CR2: ffffffff880e57d4 CR3: 0000000000201000 CR4: 00000000000006e0
Process xmms (pid: 8104, threadinfo ffff81004730e000, task ffff810076fee180)
Stack:  ffffffff880a2cd6 0000000000000000 0000000000000001 ffff81007604bcf0
 ffffffff8024c42b 0000000000000000 ffffffff8058b290 0000001f45880508
 ffff810001f3fe50 0000000300000000 ffff81007a5711e8 0000000000000000
Call Trace:
 [<ffffffff880a2cd6>] :snd_mixer_oss:snd_mixer_oss_notify_handler+0x23e/0x271
 [<ffffffff8024c42b>] free_pages_bulk+0x2c/0x1f4
 [<ffffffff802229d2>] __wake_up+0x36/0x4d
 [<ffffffff880166d7>] :snd:snd_card_do_free+0x17/0xe4
Quoted 8 lines
 [<ffffffff88016e3d>] :snd:snd_card_file_remove+0xfd/0x135
 [<ffffffff88041fc7>] :snd_pcm:snd_pcm_release+0xb6/0xc1
 [<ffffffff8026745a>] __fput+0xac/0x18a
 [<ffffffff802653d4>] filp_close+0x5c/0x64
 [<ffffffff8022a913>] put_files_struct+0x6b/0xbd
 [<ffffffff8022ba84>] do_exit+0x233/0x814
 [<ffffffff8022c0eb>] sys_exit_group+0x0/0xe
 [<ffffffff802093fe>] system_call+0x7e/0x83

How did it happen? You plugged out a usb adaptor during xmms is running with OSS output mode?

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

author
#2 lore raw
Quoted 4 lines

How did it happen? You plugged out a usb adaptor during xmms is running with OSS output mode?

xmms was running yes. But I think the oops happened on replug.

Long story: i had the headset connected during boot and my XMMS insisted on outputing its music on it, which wasn't intended -- i wanted it on the speakers connected to the motherboard sound chip.

So i unplug the headset and restarted alsa initialization, but it was very unhappy about cleaning up and couldn't initialize the other sound card. I replugged the headset and then I got the oops.

-Andi

-------------------------------------------------------------------------
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

#3 lore raw

At Mon, 6 Nov 2006 18:03:42 +0100,

Quoted 8 lines · Andi Kleen
Quoted 5 lines

How did it happen? You plugged out a usb adaptor during xmms is running with OSS output mode?

xmms was running yes. But I think the oops happened on replug.

Quoted 8 lines

Long story: i had the headset connected during boot and my XMMS insisted on outputing its music on it, which wasn't intended -- i wanted it on the speakers connected to the motherboard sound chip.

So i unplug the headset and restarted alsa initialization, but it was very unhappy about cleaning up and couldn't initialize the other sound card. I replugged the headset and then I got the oops.

OK I got a lockup with my laptop and usb-audio replugging and xmms, too.

The patch below fixed the lockup problem on my machine. Could you check whether it fixes your case, too?

thanks,

Takashi

sound/core/oss/pcm_oss.c +2 -1
index 505b23e..e0821eb 100644
@@ -2359,7 +2359,8 @@ static int snd_pcm_oss_release(struct in
 		substream = pcm_oss_file->streams[SNDRV_PCM_STREAM_CAPTURE];
 	snd_assert(substream != NULL, return -ENXIO);
 	pcm = substream->pcm;
-	snd_pcm_oss_sync(pcm_oss_file);
+	if (!pcm->card->shutdown)
+		snd_pcm_oss_sync(pcm_oss_file);
 	mutex_lock(&pcm->open_mutex);
 	snd_pcm_oss_release_file(pcm_oss_file);
 	mutex_unlock(&pcm->open_mutex);
sound/core/pcm_native.c +4 -2
index 37b4b10..66e24b5 100644
@@ -1310,7 +1310,8 @@ static int snd_pcm_pre_prepare(struct sn
 			       int f_flags)
 {
 	struct snd_pcm_runtime *runtime = substream->runtime;
-	if (runtime->status->state == SNDRV_PCM_STATE_OPEN)
+	if (runtime->status->state == SNDRV_PCM_STATE_OPEN ||
+	    runtime->status->state == SNDRV_PCM_STATE_DISCONNECTED)
 		return -EBADFD;
 	if (snd_pcm_running(substream))
 		return -EBUSY;
@@ -1568,7 +1569,8 @@ static int snd_pcm_drop(struct snd_pcm_s
 	runtime = substream->runtime;
 	card = substream->pcm->card;
 
-	if (runtime->status->state == SNDRV_PCM_STATE_OPEN)
+	if (runtime->status->state == SNDRV_PCM_STATE_OPEN ||
+	    runtime->status->state == SNDRV_PCM_STATE_DISCONNECTED)
 		return -EBADFD;
 
 	snd_power_lock(card);
sound/usb/usbaudio.c +2 -1
index c82b01c..67202b9 100644
@@ -1469,7 +1469,8 @@ static int snd_usb_hw_free(struct snd_pc
 	subs->cur_audiofmt = NULL;
 	subs->cur_rate = 0;
 	subs->period_bytes = 0;
-	release_substream_urbs(subs, 0);
+	if (!subs->stream->chip->shutdown)
+		release_substream_urbs(subs, 0);
 	return snd_pcm_free_vmalloc_buffer(substream);
 }
 
#4 lore raw
In reply to Andi Kleen

At Mon, 6 Nov 2006 18:03:42 +0100,

Quoted 8 lines · Andi Kleen
Quoted 5 lines

How did it happen? You plugged out a usb adaptor during xmms is running with OSS output mode?

xmms was running yes. But I think the oops happened on replug.

Quoted 8 lines

Long story: i had the headset connected during boot and my XMMS insisted on outputing its music on it, which wasn't intended -- i wanted it on the speakers connected to the motherboard sound chip.

So i unplug the headset and restarted alsa initialization, but it was very unhappy about cleaning up and couldn't initialize the other sound card. I replugged the headset and then I got the oops.

OK I got a lockup with my laptop and usb-audio replugging and xmms, too.

The patch below fixed the lockup problem on my machine. Could you check whether it fixes your case, too?

thanks,

Takashi

sound/core/oss/pcm_oss.c +2 -1
index 505b23e..e0821eb 100644
@@ -2359,7 +2359,8 @@ static int snd_pcm_oss_release(struct in
 		substream = pcm_oss_file->streams[SNDRV_PCM_STREAM_CAPTURE];
 	snd_assert(substream != NULL, return -ENXIO);
 	pcm = substream->pcm;
-	snd_pcm_oss_sync(pcm_oss_file);
+	if (!pcm->card->shutdown)
+		snd_pcm_oss_sync(pcm_oss_file);
 	mutex_lock(&pcm->open_mutex);
 	snd_pcm_oss_release_file(pcm_oss_file);
 	mutex_unlock(&pcm->open_mutex);
sound/core/pcm_native.c +4 -2
index 37b4b10..66e24b5 100644
@@ -1310,7 +1310,8 @@ static int snd_pcm_pre_prepare(struct sn
 			       int f_flags)
 {
 	struct snd_pcm_runtime *runtime = substream->runtime;
-	if (runtime->status->state == SNDRV_PCM_STATE_OPEN)
+	if (runtime->status->state == SNDRV_PCM_STATE_OPEN ||
+	    runtime->status->state == SNDRV_PCM_STATE_DISCONNECTED)
 		return -EBADFD;
 	if (snd_pcm_running(substream))
 		return -EBUSY;
@@ -1568,7 +1569,8 @@ static int snd_pcm_drop(struct snd_pcm_s
 	runtime = substream->runtime;
 	card = substream->pcm->card;
 
-	if (runtime->status->state == SNDRV_PCM_STATE_OPEN)
+	if (runtime->status->state == SNDRV_PCM_STATE_OPEN ||
+	    runtime->status->state == SNDRV_PCM_STATE_DISCONNECTED)
 		return -EBADFD;
 
 	snd_power_lock(card);
sound/usb/usbaudio.c +2 -1
index c82b01c..67202b9 100644
@@ -1469,7 +1469,8 @@ static int snd_usb_hw_free(struct snd_pc
 	subs->cur_audiofmt = NULL;
 	subs->cur_rate = 0;
 	subs->period_bytes = 0;
-	release_substream_urbs(subs, 0);
+	if (!subs->stream->chip->shutdown)
+		release_substream_urbs(subs, 0);
 	return snd_pcm_free_vmalloc_buffer(substream);
 }
 

-------------------------------------------------------------------------
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