https://bugzilla.novell.com/show_bug.cgi?id=426604 Summary: Regression: When sending '- - -' to /sys/class/scsi_host/hostX/scan for FC HBA the rescan does not start Product: openSUSE 10.3 Version: Final Platform: x86-64 OS/Version: Other Status: NEW Severity: Normal Priority: P5 - None Component: Kernel AssignedTo: bnc-team-screening@forge.provo.novell.com ReportedBy: konrad@virtualiron.com QAContact: kernel-maintainers@forge.provo.novell.com Found By: --- Created an attachment (id=239826) --> (https://bugzilla.novell.com/attachment.cgi?id=239826) Patch to fc_user_scan that fixed the issue. This is for SLES10 SP2 (I couldn't set the right family) When I disconnect the block device, the rescan functionality does not repopulate the SysFS. This worked with SLES 10 SP1. -sh-3.1#echo 0xffffff > /proc/sys/dev/scsi/logging_level -sh-3.1# lsscsi [1:0:0:0] disk DGC LUNZ 0220 /dev/sda [1:0:1:0] disk DGC LUNZ 0220 /dev/sdb [3:0:6:0] process ESG-SHV SCA HSBP M35 1.06 - [4:0:6:0] process ESG-SHV SCA HSBP M35 1.06 - -sh-3.1## called the routine: -sh-3.1# echo "- - -" > /sys/class/scsi_host/host1/scan -sh-3.1# host1: scsi_scan_host_selected: <4294967295:4294967295:4294967295> scsi scan: device exists on 1:0:0:0 scsi scan: Sending REPORT LUNS to host 1 channel 0 id 0 (try 0) scsi_add_timer: scmd: ffff88007758db00, time: 1500, (ffffffff88074324) sd 1:0:0:0: send 0xffff88007758db00 sd 1:0:0:0: command: Report luns: a0 00 00 00 00 00 00 00 10 00 00 00 buffer = 0xffff88007743a580, bufflen = 4096, done = 0xffffffff8807583d, queuecommand 0xffffffff881d40bb leaving scsi_dispatch_cmnd() scsi_delete_timer: scmd: ffff88007758db00, rtn: 1 sd 1:0:0:0: done 0xffff88007758db00 SUCCESS 0 sd 1:0:0:0: command: Report luns: a0 00 00 00 00 00 00 00 10 00 00 00 scsi host busy 1 failed 0 sd 1:0:0:0: Notifying upper driver of completion (result 0) scsi scan: REPORT LUNS successful (try 0) result 0x0 sd 1:0:0:0: scsi scan: REPORT LUN scan scsi scan: device exists on 1:0:0:0 scsi scan: device exists on 1:0:1:0 scsi scan: Sending REPORT LUNS to host 1 channel 0 id 1 (try 0) scsi_add_timer: scmd: ffff88007758d800, time: 1500, (ffffffff88074324) sd 1:0:1:0: send 0xffff88007758d800 sd 1:0:1:0: command: Report luns: a0 00 00 00 00 00 00 00 10 00 00 00 buffer = 0xffff88007743a480, bufflen = 4096, done = 0xffffffff8807583d, queuecommand 0xffffffff881d40bb leaving scsi_dispatch_cmnd() scsi_delete_timer: scmd: ffff88007758d800, rtn: 1 sd 1:0:1:0: done 0xffff88007758d800 SUCCESS 0 sd 1:0:1:0: command: Report luns: a0 00 00 00 00 00 00 00 10 00 00 00 scsi host busy 1 failed 0 sd 1:0:1:0: Notifying upper driver of completion (result 0) scsi scan: REPORT LUNS successful (try 0) result 0x0 sd 1:0:1:0: scsi scan: REPORT LUN scan scsi scan: device exists on 1:0:1:0 -sh-3.1## Disconnect one device. -sh-3.1# echo 1 > /sys/block/sda/device/delete sg_remove: dev=0 -sh-3.1## And second -sh-3.1#echo "- - -" > /sys/class/scsi_host/host1/scan -sh-3.1# sg_remove: dev=1 -sh-3.1# -sh-3.1## call the routine. -sh-3.1#echo "- - -" > /sys/class/scsi_host/host1/scan -sh-3.1# host1: scsi_scan_host_selected: <4294967295:4294967295:4294967295> And that is it. The FC scan code does not go any further. Looking at the fc_user_scan code and comparing it to the SLES10 SP1 and as well to Linux 2.6.27 it looks wrong and does not work correctly. I've done a backport from SLES10 SP1 and as well from 2.6.27 and in both cases the code started working again. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.