On 11-12-15 12:53 AM, Desai, Kashyap wrote:
-----Original Message-----
From: Douglas Gilbert [mailto:dgilbert@xxxxxxxxxxxx]
Sent: Monday, December 12, 2011 10:04 PM
To: Desai, Kashyap
Cc: linux-scsi@xxxxxxxxxxxxxxx
Subject: Re: sg_requests for descriptor sense data
On 11-12-12 10:43 AM, Desai, Kashyap wrote:
Douglas,
I am trying to generate Descriptor format sense data from driver
attached to LSI controller.
I have tried few sets of commands on my linux machine. Please help me
if anything is wrong here.
Which version of sg3_utils are you using? 1.32 is the current
stable version. From the output of #3 it seems that a very
old Linux kernel or old version of sg3_utils (or both) is being
used.
It is a RHEL6 and 1.32 version of sg util.
1. Output of " sg_inq -d /dev/sg0"
[root@dhcp-135-24-192-117 ~]# sg_inq -d /dev/sg0
standard INQUIRY:
PQual=0 Device_type=0 RMB=0 version=0x06 [SPC-4]
[AERC=0] [TrmTsk=0] NormACA=0 HiSUP=1 Resp_data_format=2
SCCS=0 ACC=0 TPGS=0 3PC=0 Protect=0 BQue=0
EncServ=0 MultiP=0 [MChngr=0] [ACKREQQ=0] Addr16=0
[RelAdr=0] WBus16=0 Sync=0 Linked=0 [TranDis=0] CmdQue=1
[SPI: Clocking=0x0 QAS=0 IUS=0]
length=74 (0x4a) Peripheral device type: disk
Vendor identification: ATA
Product identification: WDC WD30EZRS-00J
Product revision level: 0A80
Unit serial number: WD-WCAWZ0036672
Version descriptors:
SAM-4 (no version claimed)
SAT-2 (no version claimed)
SPC-4 (no version claimed)
SBC-3 (no version claimed)
SAS ANSI INCITS 376-2003
ATA/ATAPI-8 ATA-ACS ATA/ATAPI command set (no version claimed)
That is an ATA disk so there is a SATL somewhere. This output
looks good (and looks like libata and _not_ the better
SATL found in the firmware of LSI's MPT SAS HBAs).
I agree.! I am not sure about what is a theory behind it, but yes your input clears my doubt
Thanks.
2. Output of " sg_requests -v /dev/sg0 -d"
[root@dhcp-135-24-192-117 ~]# sg_requests -v /dev/sg0 -d
Request Sense cmd: 03 01 00 00 fc 00
request sense: Fixed format, current; Sense key: Illegal Request
Additional sense: Invalid field in cdb
bad field in Request Sense cdb
Ah, that looks like libata's SATL! Whoever programmed
that didn't understand the SCSI REQUEST SENSE command.
I think libata's SATL has been broken with respect
to REQUEST SENSE since day one.
3. Reading beyond the range of drive to get sense data.
Output of "sg_read if=/dev/sg0 cdbsz=16 skip=0x15d50a3af00 bs=512
verbose=3 count=512"
[root@dhcp-135-24-192-117 ~]# sg_read if=/dev/sg0 cdbsz=16
skip=0x15d50a3af00 bs=512 verbose=3 count=512
Opened /dev/sg0 for SG_IO with flags=0x2
SG_GET_RESERVED_SIZE yields: 32768
read cdb: 88 00 00 00 01 5d 50 a3 af 00 00 00 00 80 00 00
duration=123 ms
reading: SCSI status: Check Condition
Fixed format, current; Sense key: Illegal Request
Additional sense: Logical block address out of range
Raw sense data (in hex):
70 00 05 00 00 00 00 0a 00 00 00 00 21 00 00 00
00 00
sg_read: SCSI READ failed
Some error occurred, remaining block count=512
0+0 records in
Could you try that again with ddpt? The "Logical block
address out of range" is correct but I haven't seen
a SG_GET_RESERVED_SIZE error in a very long time.
I am trying this. Do you have any default "ddpt" command to be executed ?
I am seeing some issue while running ddpt command related to ": undefined symbol: set_scsi_pt_flags"
That implies there is an old version of the sgutils dev
library on your system. On Debian/Ubuntu it would have
a name like libsgutils2-dev_1.33-0.1_i386.deb but for
that error it will be version 1.28 [20091002] or earlier.
You need to delete that old sgutils dev library package.
Once it is gone ddpt will build what it needs and link it
in statically.
Another approach is to fetch a recent sg3_utils tarball,
build it and install all three packages the build generates.
After that ddpt will run properly.
Advice to others: if you have pressure from distribution
maintainers to share common code between packages in
shared libraries; politely ignore them.
Doug Gilbert
[root@dhcp-135-24-192-117 ddpt-0.92]# ddpt if=/dev/sg3 of=/dev/null count=10 verbose=2
Assume block size of 512 bytes for both input and output
>> Input file type: pass-through [pt] device
open /dev/sg3 with flags=0x802
inquiry cdb: 12 00 00 00 24 00
/dev/sg3: ATA WDC WD30EZRS-00J 0A80 [pdt=0]
>> Output file type: null device
read capacity (10) cdb: 25 00 00 00 00 00 00 00 00 00
READ CAPACITY (10) response cannot represent this capacity
read capacity (16) cdb: 9e 10 00 00 00 00 00 00 00 00 00 00 00 20 00 00
/dev/sg3 [pt]: blocks=5860533168 [0x15d50a3b0], _bs=512, 3.00 TB
skip=0 (blocks on input), seek=0 (blocks on output)
ibs=512 bytes, obs=512 bytes, OBPC=0
initial count=10 (blocks of input), blocks_per_transfer=128
ddpt: symbol lookup error: ddpt: undefined symbol: set_scsi_pt_flags
[root@dhcp-135-24-192-117 ddpt-0.92]#
My Drive is SPC-4 complaint and sg_requests suggested to have SPC-3 or
higher version to enable descriptor format.
_But_ for me it is not working.
Hmmm, hopefully your drive is compliant with some ATA standard,
perhaps ACS-1. The issues we are looking at here are SAT (or
SAT-2) compliance. Not sure how libata handles the D_SENSE
field in Control mode page but I wouldn't be surprised if it
is broken.
A recent Seagate 3 TB SAS disk with 512 byte blocks comes from
the factory like this:
D_SENSE 1 [cha: y, def: 1, sav: 1] Descriptor format sense
data
and it does return descriptor sense data when appropriate.
I tried with another 3TB SAS drive and yes things are as expected there. I am able to see Descriptor sense data.
Eventually, When I try to run command as mentioned in #3, I want to
see Descriptor format sense instead of Fixed one.
Well when a disk is over 2 TB in size then descriptor
sense format is required to represent LBAs that are
bigger than 32 bits (assuming a LB size of 512 bytes).
Doug Gilbert
--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
--
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
