Hi, i stumbled over btrace(8) which lists the SCSI commands by blktrace, blkparse, and /sys/kernel/debug.
The only traffic is every 2 seconds a 0x4A GET EVENT STATUS NOTIFICATION command in "Polled" operational mode, and notification class request "Media". I did not find out yet how to get to view the drive's replies. 11,1 0 2 2.047952134 0 C R (4a 01 00 00 10 00 00 00 08 00 ..) [0] 11,1 1 4 2.047580736 19046 G N [kworker/1:3] 11,1 1 5 2.047582462 19046 I R 8 (4a 01 00 00 10 00 00 00 08 00 ..) [kworker/1:3] 11,1 1 6 2.047582648 19046 D R 8 (4a 01 00 00 10 00 00 00 08 00 ..) [kworker/1:3] (I wonder why C with the older timestamp is listed before G, I, and D.) The other drives get the same treatment. I tested that e.g. xorriso's SCSI commands are reported. So either the drive really pulls in the tray on its own, or it is annoyed by a hundred 0x4A commands and finally wants to give the caller an answer. https://lwn.net/Articles/423619/ and Michael Biebl's previous proposal brought me to # cat /sys/block/sr1/events_poll_msecs 2000 # echo 0 >/sys/block/sr1/events_poll_msecs # cat /sys/block/sr1/events_poll_msecs 0 This silences the 0x4A commands. But the drive tray still gets pulled in. Shortly after, btrace reports 11,1 1 0 0.000000000 0 m N cfq20698S / put_queue A second experiment pulled in the tray without any btrace message. In summary there seems to be really no initiator in the operating system for the tray movements. Very astonishing. Have a nice day :) Thomas -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: https://lists.debian.org/950555510799137...@scdbackup.webframe.org
