Project

General

Profile

Actions

Bug #8511

open

mpt_sas driver no longer supports CD/DVD/BD (SCSI type 5) SATA devices

Added by Marcel Kroontje about 6 years ago. Updated about 6 years ago.

Status:
New
Priority:
Normal
Assignee:
-
Category:
driver - device drivers
Start date:
2017-07-18
Due date:
% Done:

0%

Estimated time:
Difficulty:
Medium
Tags:
needs-triage
Gerrit CR:
External Bug:

Description

The log contains the following errors for all CD/DVD SATA devices:

Jul 17 16:49:12 omnios scsi: [ID 243001 kern.warning] WARNING: /pci@0,0/pci8086,101@1/pci1000,30b0@0 (mpt_sas0):
Jul 17 16:49:12 omnios mptsas request inquiry page 0x83 for target:9, lun:0 failed!
Jul 17 16:49:12 omnios scsi: [ID 243001 kern.warning] WARNING: /pci@0,0/pci8086,101@1/pci1000,30b0@0 (mpt_sas0):
Jul 17 16:49:12 omnios mptsas request inquiry page 0x83 for target:9, lun:0 failed!
Jul 17 16:49:12 omnios scsi: [ID 107833 kern.notice] /pci@0,0/pci8086,101@1/pci1000,30b0@0 (mpt_sas0):
Jul 17 16:49:12 omnios mptsas_handle_topo_change: could not get SATA GUID for target 9.

Since OmniOS r151016 the following statement is present in mpt_sas.c in mptsas_get_sata_guid():

sata_guid = -1;

This causes the "mptsas_handle_topo_change: could not get SATA GUID for target 9" error and the device is skipped.
If I uncomment this line, the SATA BD drive is working again.

The whole page 0x83 inquiry should not be issued for type 5 SCSI devices, because they do not support this.
This is correctly implemented in mptsas_create_lun(), where the inq83 call is skipped for type 5 devices (and optical
and enclosure devices):

/*
 * For DVD/CD ROM and tape devices and optical
 * devices, we won't try to enumerate them under
 * scsi_vhci, so no need to try page83
*/
if (sd_inq && (sd_inq->inq_dtype DTYPE_RODIRECT ||
sd_inq->inq_dtype DTYPE_OPTICAL ||
sd_inq->inq_dtype == DTYPE_ESI))
goto create_lun;

Something similar should be done when obtaining the SATA GUID.
This would also get rid of the inq83 errors (which were already there before r151016, but it would not cause the
devices to be skipped).

Actions #1

Updated by Robert Mustacchi about 6 years ago

Hi Marcel, sorry about the breakage here. This is my fault. If I put together a patch for this would you be able to test it?

Actions #2

Updated by Marcel Kroontje about 6 years ago

Hi Robert, thanks for the follow-up. Of course I will test your patch.

Actions

Also available in: Atom PDF