sym2 probs in bk-scsi tree

sym2 probs in bk-scsi tree

[Andrew Morton]

This is on the power4 machine, which has previously worked quite happily.

During boot the system spends maybe one minute dicking with the SCSI
controller and emitting scary new reset and abort messages.


scsi(0:0:10:0): Beginning Domain Validation                       
sym0:10: wide asynchronous.                
sym0:10: FAST-80 WIDE SCSI 160.0 MB/s DT (12.5 ns, offset 31)
scsi(0:0:10:0): Ending Domain Validation                     
  Vendor: IBM       Model: IC35L036UCD210-0  Rev: S5BS
  Type:   Direct-Access                      ANSI SCSI revision: 03
sym0:11:0: tagged command queuing enabled, command queue depth 16. 
scsi(0:0:11:0): Beginning Domain Validation                       
sym0:11: wide asynchronous.                
sym0:11: FAST-80 WIDE SCSI 160.0 MB/s DT (12.5 ns, offset 31)
scsi(0:0:11:0): Ending Domain Validation                     
  Vendor: IBM       Model: HSBPD4E  PU3SCSI  Rev: 0013
  Type:   Enclosure                          ANSI SCSI revision: 02
scsi(0:0:15:0): Beginning Domain Validation                        
sym0:15:0:phase change 6-7 6@28050390 resid=4.
sym0:15:0:phase change 6-7 5@28050390 resid=3.
sym0:15:0:phase change 6-7 9@28050390 resid=7.
sym0:15:0:phase change 6-7 6@28050390 resid=4.
sym0:15:0: DEVICE RESET operation started.    
sym0:15:0: DEVICE RESET operation complete.
sym0:15:control msgout: c.                 
sym0: TARGET 15 has been reset.
sym0:15:0: ABORT operation started.
sym0:15:0: ABORT operation complete.
sym0:15:0: BUS RESET operation started.
sym0:15:0: BUS RESET operation complete.
sym0: SCSI BUS reset detected.          
sym0: SCSI BUS has been reset.
sym0:15:0: DEVICE RESET operation started.
sym0:15:0: DEVICE RESET operation complete.
sym0:15:control msgout: c.                 
sym0: TARGET 15 has been reset.
sym0:15:0: ABORT operation started.
sym0:15:0: ABORT operation complete.
sym0:15:0: BUS RESET operation started.
sym0:15:0: BUS RESET operation complete.
sym0: SCSI BUS reset detected.          
sym0: SCSI BUS has been reset.
sym0:15:0: DEVICE RESET operation started.
sym0:15:0: DEVICE RESET operation complete.
sym0:15:control msgout: c.                 
sym0: TARGET 15 has been reset.
sym0:15:0: ABORT operation started.
sym0:15:0: ABORT operation complete.
sym0:15:0: BUS RESET operation started.
sym0:15:0: BUS RESET operation complete.
sym0: SCSI BUS reset detected.          
sym0: SCSI BUS has been reset.
scsi(0:0:15:0): Domain Validation skipping write tests
scsi(0:0:15:0): Ending Domain Validation              
sym1: <1010-66> rev 0x1 at pci 0000:41:01.1 irq 104
sym1: No NVRAM, ID 7, Fast-80, LVD, parity checking
sym1: SCSI BUS has been reset.                     
scsi1 : sym-2.1.18j           
sym2: <875> rev 0x3 at pci 0001:21:01.0 irq 117
sym2: No NVRAM, ID 7, Fast-20, SE, parity checking
sym2: SCSI BUS has been reset.                    
scsi2 : sym-2.1.18j           
st: Version 20040318, fixed bufsize 32768, s/g segs 256
SCSI device sda: 71096640 512-byte hdwr sectors (36401 MB)
SCSI device sda: drive cache: write through               
 sda: sda1 sda2 sda3 sda4 < sda5 >         

Re: sym2 probs in bk-scsi tree

[Matthew Wilcox]

On Wed, Apr 07, 2004 at 11:18:20PM -0700, Andrew Morton wrote:
> This is on the power4 machine, which has previously worked quite happily.
> 
> During boot the system spends maybe one minute dicking with the SCSI
> controller and emitting scary new reset and abort messages.

They certainly are scary.  What is device 15 on this scsi bus?  This
looks like domain validation being applied to a device that can't do it
to me.  Did I get that right, James?

> scsi(0:0:10:0): Beginning Domain Validation                       
> sym0:10: wide asynchronous.                
> sym0:10: FAST-80 WIDE SCSI 160.0 MB/s DT (12.5 ns, offset 31)
> scsi(0:0:10:0): Ending Domain Validation                     
>   Vendor: IBM       Model: IC35L036UCD210-0  Rev: S5BS
>   Type:   Direct-Access                      ANSI SCSI revision: 03
> sym0:11:0: tagged command queuing enabled, command queue depth 16. 
> scsi(0:0:11:0): Beginning Domain Validation                       
> sym0:11: wide asynchronous.                
> sym0:11: FAST-80 WIDE SCSI 160.0 MB/s DT (12.5 ns, offset 31)
> scsi(0:0:11:0): Ending Domain Validation                     
>   Vendor: IBM       Model: HSBPD4E  PU3SCSI  Rev: 0013
>   Type:   Enclosure                          ANSI SCSI revision: 02
> scsi(0:0:15:0): Beginning Domain Validation                        
> sym0:15:0:phase change 6-7 6@28050390 resid=4.
> sym0:15:0:phase change 6-7 5@28050390 resid=3.
> sym0:15:0:phase change 6-7 9@28050390 resid=7.
> sym0:15:0:phase change 6-7 6@28050390 resid=4.
> sym0:15:0: DEVICE RESET operation started.    
> sym0:15:0: DEVICE RESET operation complete.
> sym0:15:control msgout: c.                 
> sym0: TARGET 15 has been reset.
> sym0:15:0: ABORT operation started.
> sym0:15:0: ABORT operation complete.
> sym0:15:0: BUS RESET operation started.
> sym0:15:0: BUS RESET operation complete.
> sym0: SCSI BUS reset detected.          
> sym0: SCSI BUS has been reset.
> sym0:15:0: DEVICE RESET operation started.
> sym0:15:0: DEVICE RESET operation complete.
> sym0:15:control msgout: c.                 
> sym0: TARGET 15 has been reset.
> sym0:15:0: ABORT operation started.
> sym0:15:0: ABORT operation complete.
> sym0:15:0: BUS RESET operation started.
> sym0:15:0: BUS RESET operation complete.
> sym0: SCSI BUS reset detected.          
> sym0: SCSI BUS has been reset.
> sym0:15:0: DEVICE RESET operation started.
> sym0:15:0: DEVICE RESET operation complete.
> sym0:15:control msgout: c.                 
> sym0: TARGET 15 has been reset.
> sym0:15:0: ABORT operation started.
> sym0:15:0: ABORT operation complete.
> sym0:15:0: BUS RESET operation started.
> sym0:15:0: BUS RESET operation complete.
> sym0: SCSI BUS reset detected.          
> sym0: SCSI BUS has been reset.
> scsi(0:0:15:0): Domain Validation skipping write tests
> scsi(0:0:15:0): Ending Domain Validation              
> sym1: <1010-66> rev 0x1 at pci 0000:41:01.1 irq 104
> sym1: No NVRAM, ID 7, Fast-80, LVD, parity checking
> sym1: SCSI BUS has been reset.                     
> scsi1 : sym-2.1.18j           
> sym2: <875> rev 0x3 at pci 0001:21:01.0 irq 117
> sym2: No NVRAM, ID 7, Fast-20, SE, parity checking
> sym2: SCSI BUS has been reset.                    
> scsi2 : sym-2.1.18j           
> st: Version 20040318, fixed bufsize 32768, s/g segs 256
> SCSI device sda: 71096640 512-byte hdwr sectors (36401 MB)
> SCSI device sda: drive cache: write through               
>  sda: sda1 sda2 sda3 sda4 < sda5 >         

-- 
"Next the statesmen will invent cheap lies, putting the blame upon 
the nation that is attacked, and every man will be glad of those
conscience-soothing falsities, and will diligently study them, and refuse
to examine any refutations of them; and thus he will by and by convince 
himself that the war is just, and will thank God for the better sleep 
he enjoys after this process of grotesque self-deception." -- Mark Twain

Re: sym2 probs in bk-scsi tree

[Brian King]

Matthew Wilcox wrote:
> On Wed, Apr 07, 2004 at 11:18:20PM -0700, Andrew Morton wrote:
> 
>>This is on the power4 machine, which has previously worked quite happily.
>>
>>During boot the system spends maybe one minute dicking with the SCSI
>>controller and emitting scary new reset and abort messages.
> 
> 
> They certainly are scary.  What is device 15 on this scsi bus?  This
> looks like domain validation being applied to a device that can't do it
> to me.  Did I get that right, James?

Looks like a SES to me. Would it make sense to not log errors during 
domain validation?

-Brian

Re: sym2 probs in bk-scsi tree

[Matthew Wilcox]

On Thu, Apr 08, 2004 at 10:16:03AM -0500, Brian King wrote:
> >They certainly are scary.  What is device 15 on this scsi bus?  This
> >looks like domain validation being applied to a device that can't do it
> >to me.  Did I get that right, James?
> 
> Looks like a SES to me. Would it make sense to not log errors during 
> domain validation?

I was thinking more about disabling domain validation towards these
kinds of devices ...

-- 
"Next the statesmen will invent cheap lies, putting the blame upon 
the nation that is attacked, and every man will be glad of those
conscience-soothing falsities, and will diligently study them, and refuse
to examine any refutations of them; and thus he will by and by convince 
himself that the war is just, and will thank God for the better sleep 
he enjoys after this process of grotesque self-deception." -- Mark Twain

Re: sym2 probs in bk-scsi tree

[Brian King]

Matthew Wilcox wrote:
> On Thu, Apr 08, 2004 at 10:16:03AM -0500, Brian King wrote:
> 
>>>They certainly are scary.  What is device 15 on this scsi bus?  This
>>>looks like domain validation being applied to a device that can't do it
>>>to me.  Did I get that right, James?
>>
>>Looks like a SES to me. Would it make sense to not log errors during 
>>domain validation?
> 
> 
> I was thinking more about disabling domain validation towards these
> kinds of devices ...
> 

That would be a resonable thing to do as well, performance to SES 
devices is really not an issue.

-Brian

Re: sym2 probs in bk-scsi tree

[James Bottomley]

On Thu, 2004-04-08 at 09:57, Matthew Wilcox wrote:
> They certainly are scary.  What is device 15 on this scsi bus?  This
> looks like domain validation being applied to a device that can't do it
> to me.  Did I get that right, James?

Actually, it might be a driver issue; I'll have to look. I suspect the
SES device probably said it didn't support either wide or sync in its
inquiry return; we went ahead and tried negotiating anyway and it got
confused.

James

Re: sym2 probs in bk-scsi tree

[James Bottomley]

On Thu, 2004-04-08 at 09:57, Matthew Wilcox wrote:
> They certainly are scary.  What is device 15 on this scsi bus?  This
> looks like domain validation being applied to a device that can't do it
> to me.  Did I get that right, James?

OK, try the attached, it should predicate DV on what we already know
from the inquiry data.

James

===== drivers/scsi/scsi_transport_spi.c 1.9 vs edited =====
--- 1.9/drivers/scsi/scsi_transport_spi.c	Thu Apr  8 08:39:09 2004
+++ edited/drivers/scsi/scsi_transport_spi.c	Thu Apr  8 12:35:07 2004
@@ -490,7 +490,7 @@
 	}
 
 	/* test width */
-	if (i->f->set_width) {
+	if (i->f->set_width && sdev->wdtr) {
 		i->f->set_width(sdev, 1);
 
 		if (!spi_dv_device_compare_inquiry(sreq, buffer,
@@ -502,6 +502,10 @@
 	}
 
 	if (!i->f->set_period)
+		return;
+
+	/* device can't handle synchronous */
+	if(!sdev->ppr && !sdev->sdtr)
 		return;
 
 	/* now set up to the maximum */

Re: sym2 probs in bk-scsi tree

[Andrew Morton]

James Bottomley <James.Bottomley@SteelEye.com> wrote:
>
> On Thu, 2004-04-08 at 09:57, Matthew Wilcox wrote:
> > They certainly are scary.  What is device 15 on this scsi bus?  This
> > looks like domain validation being applied to a device that can't do it
> > to me.  Did I get that right, James?
> 
> OK, try the attached, it should predicate DV on what we already know
> from the inquiry data.
> 
> James
> 
> ===== drivers/scsi/scsi_transport_spi.c 1.9 vs edited =====
> --- 1.9/drivers/scsi/scsi_transport_spi.c	Thu Apr  8 08:39:09 2004
> +++ edited/drivers/scsi/scsi_transport_spi.c	Thu Apr  8 12:35:07 2004
> @@ -490,7 +490,7 @@
>  	}
>  
>  	/* test width */
> -	if (i->f->set_width) {
> +	if (i->f->set_width && sdev->wdtr) {
>  		i->f->set_width(sdev, 1);
>  
>  		if (!spi_dv_device_compare_inquiry(sreq, buffer,
> @@ -502,6 +502,10 @@
>  	}
>  
>  	if (!i->f->set_period)
> +		return;
> +
> +	/* device can't handle synchronous */
> +	if(!sdev->ppr && !sdev->sdtr)
>  		return;
>  
>  	/* now set up to the maximum */

Yup, that makes the pauses and scary messages go away, thanks.

fwiw, here's the scsi-related part of dmesg:

sym0: <1010-66> rev 0x1 at pci 0000:41:01.0 irq 103
sym0: No NVRAM, ID 7, Fast-80, LVD, parity checking
sym0: SCSI BUS has been reset.
scsi0 : sym-2.1.18j
DEV: registering device: ID = 'host0'
CLASS: registering class device: ID = 'host0'
class_hotplug - name = host0
  Vendor: IBM       Model: IC35L036UCD210-0  Rev: S5BS
  Type:   Direct-Access                      ANSI SCSI revision: 03
sym0:8:0: tagged command queuing enabled, command queue depth 16.
scsi(0:0:8:0): Beginning Domain Validation
sym0:8: wide asynchronous.
sym0:8: FAST-80 WIDE SCSI 160.0 MB/s DT (12.5 ns, offset 31)
scsi(0:0:8:0): Ending Domain Validation
DEV: registering device: ID = '0:0:8:0'
bus scsi: add device 0:0:8:0
CLASS: registering class device: ID = '0:0:8:0'
class_hotplug - name = 0:0:8:0
CLASS: registering class device: ID = '0:0:8:0'
class_hotplug - name = 0:0:8:0
  Vendor: IBM       Model: IC35L036UCD210-0  Rev: S5BS
  Type:   Direct-Access                      ANSI SCSI revision: 03
sym0:9:0: tagged command queuing enabled, command queue depth 16.
scsi(0:0:9:0): Beginning Domain Validation
sym0:9: wide asynchronous.
sym0:9: FAST-80 WIDE SCSI 160.0 MB/s DT (12.5 ns, offset 31)
scsi(0:0:9:0): Ending Domain Validation
DEV: registering device: ID = '0:0:9:0'
bus scsi: add device 0:0:9:0
CLASS: registering class device: ID = '0:0:9:0'
class_hotplug - name = 0:0:9:0
CLASS: registering class device: ID = '0:0:9:0'
class_hotplug - name = 0:0:9:0
  Vendor: IBM       Model: IC35L036UCD210-0  Rev: S5BS
  Type:   Direct-Access                      ANSI SCSI revision: 03
sym0:10:0: tagged command queuing enabled, command queue depth 16.
scsi(0:0:10:0): Beginning Domain Validation
sym0:10: wide asynchronous.
sym0:10: FAST-80 WIDE SCSI 160.0 MB/s DT (12.5 ns, offset 31)
scsi(0:0:10:0): Ending Domain Validation
DEV: registering device: ID = '0:0:10:0'
bus scsi: add device 0:0:10:0
CLASS: registering class device: ID = '0:0:10:0'
class_hotplug - name = 0:0:10:0
CLASS: registering class device: ID = '0:0:10:0'
class_hotplug - name = 0:0:10:0
  Vendor: IBM       Model: IC35L036UCD210-0  Rev: S5BS
  Type:   Direct-Access                      ANSI SCSI revision: 03
sym0:11:0: tagged command queuing enabled, command queue depth 16.
scsi(0:0:11:0): Beginning Domain Validation
sym0:11: wide asynchronous.
sym0:11: FAST-80 WIDE SCSI 160.0 MB/s DT (12.5 ns, offset 31)
scsi(0:0:11:0): Ending Domain Validation
DEV: registering device: ID = '0:0:11:0'
bus scsi: add device 0:0:11:0
CLASS: registering class device: ID = '0:0:11:0'
class_hotplug - name = 0:0:11:0
CLASS: registering class device: ID = '0:0:11:0'
class_hotplug - name = 0:0:11:0
  Vendor: IBM       Model: HSBPD4E  PU3SCSI  Rev: 0013
  Type:   Enclosure                          ANSI SCSI revision: 02
scsi(0:0:15:0): Beginning Domain Validation
sym0:15:0:phase change 6-7 6@28050390 resid=4.
scsi(0:0:15:0): Ending Domain Validation
DEV: registering device: ID = '0:0:15:0'
bus scsi: add device 0:0:15:0
CLASS: registering class device: ID = '0:0:15:0'
class_hotplug - name = 0:0:15:0
CLASS: registering class device: ID = '0:0:15:0'
class_hotplug - name = 0:0:15:0
bound device '0000:41:01.0' to driver 'sym53c8xx'
PCI: Enabling device: (0000:41:01.1), cmd 143
sym1: <1010-66> rev 0x1 at pci 0000:41:01.1 irq 104
sym1: No NVRAM, ID 7, Fast-80, LVD, parity checking
sym1: SCSI BUS has been reset.
scsi1 : sym-2.1.18j
DEV: registering device: ID = 'host1'
CLASS: registering class device: ID = 'host1'
class_hotplug - name = host1
bound device '0000:41:01.1' to driver 'sym53c8xx'
PCI: Enabling device: (0001:21:01.0), cmd 143
sym2: <875> rev 0x3 at pci 0001:21:01.0 irq 117
sym2: No NVRAM, ID 7, Fast-20, SE, parity checking
sym2: SCSI BUS has been reset.
scsi2 : sym-2.1.18j
DEV: registering device: ID = 'host2'
CLASS: registering class device: ID = 'host2'
...
device class 'scsi_tape': registering
bus scsi: add driver st
bus scsi: add driver sd
SCSI device sda: 71096640 512-byte hdwr sectors (36401 MB)
SCSI device sda: drive cache: write through
 sda: sda1 sda2 sda3 sda4 < sda5 >
Attached scsi disk sda at scsi0, channel 0, id 8, lun 0
bound device '0:0:8:0' to driver 'sd'
SCSI device sdb: 71096640 512-byte hdwr sectors (36401 MB)
SCSI device sdb: drive cache: write through
 sdb: sdb1 < sdb5 >
Attached scsi disk sdb at scsi0, channel 0, id 9, lun 0
bound device '0:0:9:0' to driver 'sd'
SCSI device sdc: 71096640 512-byte hdwr sectors (36401 MB)
SCSI device sdc: drive cache: write through
 sdc: sdc1 < sdc5 >
Attached scsi disk sdc at scsi0, channel 0, id 10, lun 0
bound device '0:0:10:0' to driver 'sd'
SCSI device sdd: 71096640 512-byte hdwr sectors (36401 MB)
SCSI device sdd: drive cache: write through
 sdd: sdd1 sdd2 sdd3 sdd4 < sdd5 >
Attached scsi disk sdd at scsi0, channel 0, id 11, lun 0
bound device '0:0:11:0' to driver 'sd'
bus scsi: add driver sr
device class 'scsi_generic': registering
CLASS: registering class device: ID = 'sg0'
class_hotplug - name = sg0
Attached scsi generic sg0 at scsi0, channel 0, id 8, lun 0,  type 0
CLASS: registering class device: ID = 'sg1'
class_hotplug - name = sg1
Attached scsi generic sg1 at scsi0, channel 0, id 9, lun 0,  type 0
CLASS: registering class device: ID = 'sg2'
class_hotplug - name = sg2
Attached scsi generic sg2 at scsi0, channel 0, id 10, lun 0,  type 0
CLASS: registering class device: ID = 'sg3'
class_hotplug - name = sg3
Attached scsi generic sg3 at scsi0, channel 0, id 11, lun 0,  type 0
CLASS: registering class device: ID = 'sg4'
class_hotplug - name = sg4
Attached scsi generic sg4 at scsi0, channel 0, id 15, lun 0,  type 13