Re: 2.6.39-rc3-git3 crashs after booting

Re: 2.6.39-rc3-git3 crashs after booting

[Andrew Morton]

[-- Attachment #1: Type: text/plain, Size: 11054 bytes --]


(cc's added.  Please retain them)

The crash appears to be due to either the scsi or block code.

On Thu, 14 Apr 2011 04:17:51 -0400 "werner" <w.landgraf@ru.ru> wrote:

> The problem with the crashs after booting which started 
> with 2.6.39-rc1 continues with 2.6.39-rc3-git3
> 
> Pls see enclosed a screen foto
> 
> Pls see below also a comparison of the config between 
> 2.6.38.2 , what don't crashs but runs stable.    Thus, the 
> problem are either the general changes from 2.6.38.2 to 
> 2.6.39, or one of the few additional features visible from 
> the diff below.
> 
> 
> W.Landgraf
> 
> SYS-Linux
> 
> 
> 
> ===========
> 3,4c3,4
> < # Linux/i386 2.6.38.2 Kernel Configuration
> < # Mon Mar 28 02:17:05 2011
> ---
> > # Linux/i386 2.6.39-rc3-git3 Kernel Configuration
> > # Thu Apr 14 01:56:21 2011
> 51d50
> < CONFIG_X86_TRAMPOLINE=y
> 64d62
> < CONFIG_LOCK_KERNEL=y
> 85a84
> > # CONFIG_FHANDLE is not set
> 97d95
> < # CONFIG_GENERIC_HARDIRQS_NO_DEPRECATED is not set
> 99a98
> > CONFIG_GENERIC_IRQ_SHOW=y
> 101,103c100
> < # CONFIG_AUTO_IRQ_AFFINITY is not set
> < # CONFIG_IRQ_PER_CPU is not set
> < # CONFIG_HARDIRQS_SW_RESEND is not set
> ---
> > CONFIG_IRQ_FORCED_THREADING=y
> 287d283
> < CONFIG_X86_CPU=y
> 411,414d406
> < CONFIG_PM=y
> < # CONFIG_PM_DEBUG is not set
> < CONFIG_PM_SLEEP_SMP=y
> < CONFIG_PM_SLEEP=y
> 416a409
> > CONFIG_HIBERNATE_CALLBACKS=y
> 418a412,413
> > CONFIG_PM_SLEEP=y
> > CONFIG_PM_SLEEP_SMP=y
> 420c415,416
> < CONFIG_PM_OPS=y
> ---
> > CONFIG_PM=y
> > # CONFIG_PM_DEBUG is not set
> 448a445
> > # CONFIG_ACPI_APEI_PCIEAER is not set
> 553,554d549
> < CONFIG_OLPC_OPENFIRMWARE=y
> < CONFIG_OLPC_OPENFIRMWARE_DT=y
> 586a582
> > # CONFIG_RAPIDIO is not set
> 616,618c612
> < CONFIG_ASK_IP_FIB_HASH=y
> < # CONFIG_IP_FIB_TRIE is not set
> < CONFIG_IP_FIB_HASH=y
> ---
> > # CONFIG_IP_FIB_TRIE_STATS is not set
> 621a616
> > CONFIG_IP_ROUTE_CLASSID=y
> 705a701
> > # CONFIG_NF_CONNTRACK_TIMESTAMP is not set
> 713a710
> > CONFIG_NF_CONNTRACK_BROADCAST=m
> 714a712
> > # CONFIG_NF_CONNTRACK_SNMP is not set
> 731a730
> > # CONFIG_NETFILTER_XT_TARGET_AUDIT is not set
> 753a753
> > # CONFIG_NETFILTER_XT_MATCH_ADDRTYPE is not set
> 761a762
> > # CONFIG_NETFILTER_XT_MATCH_DEVGROUP is not set
> 790a792
> > # CONFIG_IP_SET is not set
> 835d836
> < CONFIG_IP_NF_MATCH_ADDRTYPE=m
> 848d848
> < CONFIG_NF_NAT_SNMP_BASIC=m
> 941d940
> < CONFIG_TIPC_NODES=255
> 994d992
> < CONFIG_PHONET_PIPECTRLR=y
> 1007a1006
> > # CONFIG_NET_SCH_SFB is not set
> 1014a1014,1015
> > # CONFIG_NET_SCH_MQPRIO is not set
> > # CONFIG_NET_SCH_CHOKE is not set
> 1024d1024
> < CONFIG_NET_CLS_ROUTE=y
> 1050a1051
> > CONFIG_RFS_ACCEL=y
> 1102a1104
> > # CONFIG_CAN_C_CAN is not set
> 1169,1177c1171,1172
> < CONFIG_BT_L2CAP=m
> < CONFIG_BT_SCO=m
> < CONFIG_BT_RFCOMM=m
> < CONFIG_BT_RFCOMM_TTY=y
> < CONFIG_BT_BNEP=m
> < CONFIG_BT_BNEP_MC_FILTER=y
> < CONFIG_BT_BNEP_PROTO_FILTER=y
> < CONFIG_BT_CMTP=m
> < CONFIG_BT_HIDP=m
> ---
> > # CONFIG_BT_L2CAP is not set
> > # CONFIG_BT_SCO is not set
> 1199a1195
> > # CONFIG_BT_WILINK is not set
> 1266a1263
> > CONFIG_ARCH_NO_SYSDEV_OPS=y
> 1272d1268
> < CONFIG_MTD_CONCAT=y
> 1296a1293
> > # CONFIG_MTD_SWAP is not set
> 1355a1353
> > # CONFIG_MTD_LATCH_ADDR is not set
> 1392a1391
> > # CONFIG_MTD_NAND_ECC_BCH is not set
> 1425,1428d1423
> < 
> < #
> < # UBI debugging options
> < #
> 1444a1440
> > CONFIG_OF_PCI=y
> 1522a1519
> > CONFIG_SENSORS_LIS3LV02D=m
> 1572a1570
> > CONFIG_SENSORS_LIS3_I2C=m
> 1698d1695
> < # CONFIG_SCSI_SAS_LIBSAS_DEBUG is not set
> 1706a1704
> > # CONFIG_SCSI_BNX2X_FCOE is not set
> 1873a1872
> > # CONFIG_PATA_ARASAN_CF is not set
> 1956c1955,1961
> < # CONFIG_TARGET_CORE is not set
> ---
> > # CONFIG_DM_FLAKEY is not set
> > CONFIG_TARGET_CORE=m
> > CONFIG_TCM_IBLOCK=m
> > CONFIG_TCM_FILEIO=m
> > CONFIG_TCM_PSCSI=m
> > CONFIG_LOOPBACK_TARGET=m
> > CONFIG_LOOPBACK_TARGET_CDB_DEBUG=y
> 2155d2159
> < CONFIG_R8169_VLAN=y
> 2178d2181
> < CONFIG_CHELSIO_T3_DEPENDS=y
> 2180d2182
> < CONFIG_CHELSIO_T4_DEPENDS=y
> 2182d2183
> < CONFIG_CHELSIO_T4VF_DEPENDS=y
> 2243a2245
> > # CONFIG_ATH5K_TRACER is not set
> 2254c2256,2258
> < # CONFIG_CARL9170 is not set
> ---
> > CONFIG_CARL9170=m
> > CONFIG_CARL9170_LEDS=y
> > CONFIG_CARL9170_WPC=y
> 2294c2298
> < CONFIG_IWLWIFI=m
> ---
> > CONFIG_IWLAGN=m
> 2301,2303c2305,2313
> < CONFIG_IWLAGN=m
> < CONFIG_IWL4965=y
> < CONFIG_IWL5000=y
> ---
> > CONFIG_IWL_P2P=y
> > CONFIG_IWLWIFI_LEGACY=m
> > 
> > #
> > # Debugging Options
> > #
> > # CONFIG_IWLWIFI_LEGACY_DEBUG is not set
> > CONFIG_IWLWIFI_LEGACY_DEVICE_TRACING=y
> > CONFIG_IWL4965=m
> 2337a2348
> > CONFIG_RT2800PCI_RT53XX=y
> 2353a2365
> > CONFIG_RTL8192CU=m
> 2354a2367
> > CONFIG_RTL8192C_COMMON=m
> 2410a2424
> > CONFIG_USB_VL600=m
> 2703a2718
> > # CONFIG_KEYBOARD_QT1070 is not set
> 2786a2802
> > # CONFIG_TOUCHSCREEN_ATMEL_MXT is not set
> 2803d2818
> < CONFIG_TOUCHSCREEN_QT602240=m
> 2806a2822
> > # CONFIG_TOUCHSCREEN_WM831X is not set
> 2829a2846
> > # CONFIG_TOUCHSCREEN_TSC2005 is not set
> 2893c2910,2913
> < CONFIG_DEVKMEM=y
> ---
> > CONFIG_UNIX98_PTYS=y
> > CONFIG_DEVPTS_MULTIPLE_INSTANCES=y
> > CONFIG_LEGACY_PTYS=y
> > CONFIG_LEGACY_PTY_COUNT=256
> 2895d2914
> < CONFIG_COMPUTONE=m
> 2899d2917
> < CONFIG_DIGIEPCA=m
> 2902d2919
> < CONFIG_ISI=m
> 2905a2923,2924
> > CONFIG_NOZOMI=m
> > CONFIG_ISI=m
> 2908,2909c2927
> < CONFIG_RISCOM8=m
> < CONFIG_SPECIALIX=m
> ---
> > CONFIG_DEVKMEM=y
> 2911,2913d2928
> < CONFIG_STALLION=m
> < CONFIG_ISTALLION=m
> < CONFIG_NOZOMI=m
> 2951d2965
> < # CONFIG_SERIAL_GRLIB_GAISLER_APBUART is not set
> 2958,2961d2971
> < CONFIG_UNIX98_PTYS=y
> < CONFIG_DEVPTS_MULTIPLE_INSTANCES=y
> < CONFIG_LEGACY_PTYS=y
> < CONFIG_LEGACY_PTY_COUNT=256
> 3063a3074,3075
> > # CONFIG_I2C_PXA is not set
> > # CONFIG_I2C_PXA_PCI is not set
> 3070a3083
> > # CONFIG_I2C_DIOLAN_U2C is not set
> 3093a3107
> > # CONFIG_SPI_ALTERA is not set
> 3097a3112
> > # CONFIG_SPI_OC_TINY is not set
> 3220a3236,3237
> > CONFIG_BATTERY_BQ27X00_I2C=y
> > CONFIG_BATTERY_BQ27X00_PLATFORM=y
> 3271a3289
> > # CONFIG_SENSORS_LINEAGE is not set
> 3284a3303
> > # CONFIG_SENSORS_LTC4151 is not set
> 3290a3310
> > # CONFIG_SENSORS_MAX6639 is not set
> 3294a3315
> > # CONFIG_PMBUS is not set
> 3304a3326,3327
> > # CONFIG_SENSORS_SCH5627 is not set
> > # CONFIG_SENSORS_ADS1015 is not set
> 3328d3350
> < CONFIG_SENSORS_LIS3_I2C=m
> 3336d3357
> < CONFIG_SENSORS_LIS3LV02D=m
> 3431a3453
> > # CONFIG_TPS6105X is not set
> 3434a3457
> > # CONFIG_TWL4030_MADC is not set
> 3442a3466
> > # CONFIG_MFD_MAX8997 is not set
> 3461a3486
> > CONFIG_AB8500_GPADC=y
> 3510a3536
> > # CONFIG_MEDIA_CONTROLLER is not set
> 3533a3560
> > # CONFIG_IR_ITE_CIR is not set
> 3572a3600,3602
> > CONFIG_VIDEOBUF2_CORE=m
> > CONFIG_VIDEOBUF2_MEMOPS=m
> > CONFIG_VIDEOBUF2_VMALLOC=m
> 3670a3701
> > # CONFIG_MEDIA_ALTERA_CI is not set
> 3679a3711
> > # CONFIG_VIDEO_NOON010PC30 is not set
> 3693a3726
> > # CONFIG_SOC_CAMERA_OV9740 is not set
> 3709a3743
> > # CONFIG_USB_GSPCA_NW80X is not set
> 3735a3770
> > # CONFIG_USB_GSPCA_VICAM is not set
> 3788a3824,3828
> > 
> > #
> > # Texas Instruments WL128x FM driver (ST based)
> > #
> > # CONFIG_RADIO_WL128X is not set
> 3839a3880
> > # CONFIG_DVB_USB_TECHNISAT_USB2 is not set
> 3871a3913,3916
> > 
> > #
> > # Supported FireWire (IEEE 1394) Adapters
> > #
> 3873,3874d3917
> < CONFIG_DVB_FIREDTV_FIREWIRE=y
> < # CONFIG_DVB_FIREDTV_IEEE1394 is not set
> 3955a3999
> > CONFIG_DVB_DIB9000=m
> 3958a4003
> > CONFIG_DVB_STV0367=m
> 4038d4082
> < CONFIG_DRM_I830=m
> 4079a4124
> > CONFIG_FB_CYBER2000_DDC=y
> 4167a4213
> > # CONFIG_LCD_LD9040 is not set
> 4174c4220
> < CONFIG_BACKLIGHT_MBP_NVIDIA=m
> ---
> > # CONFIG_BACKLIGHT_APPLE is not set
> 4401a4448,4450
> > # CONFIG_SND_USB_6FIRE is not set
> > CONFIG_SND_FIREWIRE=y
> > # CONFIG_SND_FIREWIRE_SPEAKERS is not set
> 4421a4471
> > CONFIG_SND_SOC_CS4271=m
> 4424a4475
> > CONFIG_SND_SOC_DFBMCS320=m
> 4425a4477
> > CONFIG_SND_SOC_MAX9850=m
> 4426a4479
> > CONFIG_SND_SOC_SGTL5000=m
> 4430a4484
> > CONFIG_SND_SOC_TVL320AIC32X4=m
> 4466a4521
> > CONFIG_SND_SOC_WM8991=m
> 4470a4526
> > CONFIG_SND_SOC_LM4857=m
> 4525,4526d4580
> < CONFIG_HID_EGALAX=m
> < CONFIG_HID_ELECOM=m
> 4527a4582
> > # CONFIG_HID_KEYTOUCH is not set
> 4533a4589
> > # CONFIG_HID_LCPOWER is not set
> 4539d4594
> < CONFIG_HID_MAGICMOUSE=m
> 4555a4611,4612
> > CONFIG_HID_ROCCAT_COMMON=m
> > # CONFIG_HID_ROCCAT_ARVO is not set
> 4557a4615
> > # CONFIG_HID_ROCCAT_KOVAPLUS is not set
> 4570,4571d4627
> < CONFIG_HID_WACOM=m
> < CONFIG_HID_WACOM_POWER_SUPPLY=y
> 4638a4695
> > # CONFIG_USB_STORAGE_REALTEK is not set
> 4649a4707
> > # CONFIG_USB_STORAGE_ENE_UB6250 is not set
> 4752a4811
> > # CONFIG_USB_GADGET_FUSB300 is not set
> 4847a4907
> > # CONFIG_MEMSTICK_R592 is not set
> 4854a4915
> > # CONFIG_LEDS_LM3530 is not set
> 5053a5115,5120
> > CONFIG_STALLION=m
> > CONFIG_ISTALLION=m
> > CONFIG_DIGIEPCA=m
> > CONFIG_RISCOM8=m
> > CONFIG_SPECIALIX=m
> > CONFIG_COMPUTONE=m
> 5072,5075c5139
> < CONFIG_USB_DABUSB=y
> < CONFIG_USB_SE401=m
> < CONFIG_VIDEO_USBVIDEO=m
> < CONFIG_USB_VICAM=m
> ---
> > CONFIG_DVB_CXD2099=m
> 5084,5085c5148,5150
> < CONFIG_BRCM80211_PCI=y
> < # CONFIG_BRCMFMAC is not set
> ---
> > CONFIG_BRCMSMAC=y
> > CONFIG_BRCMFMAC=y
> > # CONFIG_BRCMDBG is not set
> 5214a5280
> > # CONFIG_FB_OLPC_DCON is not set
> 5224a5291,5292
> > CONFIG_RTS_PSTOR=m
> > # CONFIG_RTS_PSTOR_DEBUG is not set
> 5229d5296
> < CONFIG_AUTOFS_FS=y
> 5273a5341
> > CONFIG_IIO_KFIFO_BUF=m
> 5286a5355,5356
> > CONFIG_LIS3L02DQ_BUF_KFIFO=y
> > # CONFIG_LIS3L02DQ_BUF_RING_SW is not set
> 5298a5369,5371
> > CONFIG_AD7606=m
> > CONFIG_AD7606_IFACE_PARALLEL=m
> > CONFIG_AD7606_IFACE_SPI=m
> 5320a5394
> > CONFIG_MAX517=m
> 5340d5413
> < CONFIG_ADIS16251=m
> 5386a5460
> > # CONFIG_IIO_SYSFS_TRIGGER is not set
> 5387a5462
> > CONFIG_XVMALLOC=y
> 5388a5464
> > # CONFIG_ZRAM_DEBUG is not set
> 5391d5466
> < CONFIG_SAMSUNG_LAPTOP=m
> 5398,5402d5472
> < 
> < #
> < # Texas Instruments shared transport line discipline
> < #
> < CONFIG_ST_BT=m
> 5408,5409d5477
> < CONFIG_LIRC_IT87=m
> < CONFIG_LIRC_ITE8709=m
> 5417,5418d5484
> < CONFIG_SMB_FS=y
> < # CONFIG_SMB_NLS_DEFAULT is not set
> 5419a5486,5488
> > CONFIG_EASYCAP_SND=y
> > # CONFIG_EASYCAP_OSS is not set
> > CONFIG_EASYCAP_DEBUG=y
> 5446a5516
> > CONFIG_FT1000_PCMCIA=m
> 5468a5539,5544
> > # CONFIG_DRM_PSB is not set
> > 
> > #
> > # Altera FPGA firmware download module
> > #
> > CONFIG_ALTERA_STAPL=m
> 5473a5550
> > CONFIG_DELL_WMI_AIO=m
> 5476a5554
> > CONFIG_HP_ACCEL=m
> 5493a5572,5573
> > CONFIG_ASUS_WMI=m
> > CONFIG_ASUS_NB_WMI=m
> 5504a5585,5586
> > # CONFIG_XO15_EBOOK is not set
> > CONFIG_SAMSUNG_LAPTOP=m
> 5515a5598
> > # CONFIG_DMI_SYSFS is not set
> 5517a5601
> > # CONFIG_SIGMA is not set
> 5689a5774
> > CONFIG_PSTORE=y
> 5826a5912
> > CONFIG_DEFAULT_MESSAGE_LOGLEVEL=4
> 5834a5921
> > # CONFIG_DEBUG_SECTION_MISMATCH is not set
> 5839d5925
> < CONFIG_BKL=y
> 5890a5977
> > # CONFIG_TEST_KSTRTOX is not set
> 6119a6207
> > CONFIG_CPU_RMAP=y
> ---
> Professional hosting for everyone - http://www.host.ru

[-- Attachment #2: crash.jpg --]
[-- Type: image/jpeg, Size: 183796 bytes --]

Re: 2.6.39-rc3-git3 crashs after booting

[James Bottomley]

On Thu, 2011-04-14 at 10:40 -0700, Andrew Morton wrote:
> (cc's added.  Please retain them)
> 
> The crash appears to be due to either the scsi or block code.
> 
> On Thu, 14 Apr 2011 04:17:51 -0400 "werner" <w.landgraf@ru.ru> wrote:
> 
> > The problem with the crashs after booting which started 
> > with 2.6.39-rc1 continues with 2.6.39-rc3-git3
> > 
> > Pls see enclosed a screen foto

It's a bit hard to tell from the jpg, but it looks like it could be
another instance of the steve whitehouse crash that was discovered a day
or so ago.  We're still preparing the final fix, but this is what we
have as an interim if you could try it out.

James

---
diff --git a/block/blk.h b/block/blk.h
index c8db371..11d0d25 100644
--- a/block/blk.h
+++ b/block/blk.h
@@ -62,7 +62,8 @@ static inline struct request *__elv_next_request(struct request_queue *q)
 			return rq;
 		}
 
-		if (!q->elevator->ops->elevator_dispatch_fn(q, 0))
+		if (test_bit(QUEUE_FLAG_DEAD, &q->queue_flags) ||
+		    !q->elevator->ops->elevator_dispatch_fn(q, 0))
 			return NULL;
 	}
 }
diff --git a/drivers/scsi/scsi_sysfs.c b/drivers/scsi/scsi_sysfs.c
index e44ff64..5aa4246 100644
--- a/drivers/scsi/scsi_sysfs.c
+++ b/drivers/scsi/scsi_sysfs.c
@@ -322,14 +322,9 @@ static void scsi_device_dev_release_usercontext(struct work_struct *work)
 		kfree(evt);
 	}
 
-	if (sdev->request_queue) {
-		sdev->request_queue->queuedata = NULL;
-		/* user context needed to free queue */
-		scsi_free_queue(sdev->request_queue);
-		/* temporary expedient, try to catch use of queue lock
-		 * after free of sdev */
-		sdev->request_queue = NULL;
-	}
+	/* temporary expedient, try to catch use of queue lock after
+	 * free of sdev */
+	sdev->request_queue = NULL;
 
 	scsi_target_reap(scsi_target(sdev));
 
@@ -937,6 +932,11 @@ void __scsi_remove_device(struct scsi_device *sdev)
 	if (sdev->host->hostt->slave_destroy)
 		sdev->host->hostt->slave_destroy(sdev);
 	transport_destroy_device(dev);
+	/* Setting this to NULL causes the request function to reject
+	 * any I/O requests */
+	sdev->request_queue->queuedata = NULL;
+	/* Freeing the queue signals to block that we're done */
+	scsi_free_queue(sdev->request_queue);
 	put_device(dev);
 }
 

Re: 2.6.39-rc3-git3 crashs after booting

[Ingo Molnar]

* James Bottomley <James.Bottomley@HansenPartnership.com> wrote:

> +	/* temporary expedient, try to catch use of queue lock after
> +	 * free of sdev */
> +	sdev->request_queue = NULL;

Couldnt help to notice this small ugliness while skimming the patch - please 
use the customary (multi-line) comment style:

  /*
   * Comment .....
   * ...... goes here.
   */

specified in Documentation/CodingStyle.

Thanks,

        Ingo

Re: 2.6.39-rc3-git3 crashs after booting

[James Bottomley]

On Thu, 2011-04-14 at 20:15 +0200, Ingo Molnar wrote:
> * James Bottomley <James.Bottomley@HansenPartnership.com> wrote:
> 
> > +	/* temporary expedient, try to catch use of queue lock after
> > +	 * free of sdev */
> > +	sdev->request_queue = NULL;
> 
> Couldnt help to notice this small ugliness while skimming the patch - please 
> use the customary (multi-line) comment style:
> 
>   /*
>    * Comment .....
>    * ...... goes here.
>    */

It's just code motion, so it was cut and paste.  The actual file is a
nasty mix of comment styles, so it doesn't really matter one way or the
other.

James

Re: 2.6.39-rc3-git3 crashs after booting

[Ingo Molnar]

* James Bottomley <James.Bottomley@HansenPartnership.com> wrote:

> On Thu, 2011-04-14 at 20:15 +0200, Ingo Molnar wrote:
> > * James Bottomley <James.Bottomley@HansenPartnership.com> wrote:
> > 
> > > +	/* temporary expedient, try to catch use of queue lock after
> > > +	 * free of sdev */
> > > +	sdev->request_queue = NULL;
> > 
> > Couldnt help to notice this small ugliness while skimming the patch - please 
> > use the customary (multi-line) comment style:
> > 
> >   /*
> >    * Comment .....
> >    * ...... goes here.
> >    */
> 
> It's just code motion, so it was cut and paste. The actual file is a nasty 
> mix of comment styles, so it doesn't really matter one way or the other.

That's not really true - most other multi-line comments are fine in that file - 
i checked that before sending my feedback.

Really, why are you even arguing about it? You have not noticed it during 
review - that can happen - but you should really fix small details if someone 
points them out during review. Instead you are forcing reviewers though an 
unnecessary and time-wasting email exchange.

Thanks,

	Ingo

Re: 2.6.39-rc3-git3 crashs after booting

[Bart Van Assche]

On Thu, Apr 14, 2011 at 7:40 PM, Andrew Morton
<akpm@linux-foundation.org> wrote:
> On Thu, 14 Apr 2011 04:17:51 -0400 "werner" <w.landgraf@ru.ru> wrote:
>
>> The problem with the crashs after booting which started
>> with 2.6.39-rc1 continues with 2.6.39-rc3-git3

Although I'm not sure whether it is related, another bug report about
a lockup introduced between 2.6.38 and 2.6.39-rc1 can be found here:
https://bugzilla.kernel.org/show_bug.cgi?id=32982.

Bart.