Red Hat Bugzilla – Bug 168325
IrDA Services (initscript) brocken in fc4
Last modified: 2007-11-30 17:11:13 EST
Description of problem: (Not sure if I've chosen the correct component - perhaps 'initscripts'?) Start IrDA Service, irda appears to start but fails Stop IrDA Service, fails Same when invoked at Startup and Shutdown This worked in fc3 without issue Also, see additional info Version-Release number of selected component (if applicable): irda-utils-0.9.16-7 (initscripts-8.11.1-1) How reproducible: Always Steps to Reproduce: 1./etc/rc.d/init.d/irda start 2./etc/rc.d/init.d/irda stop 3. Actual results: No working irda service Expected results: Working irda service Additional info: - On '/etc/rc.d/init.d/irda start' there is a lot of disk activity and then no service: [root@morgansmachine ~]# irdadump 0 packets received by filter - So, nothing. - And: [root@morgansmachine ~]# /etc/rc.d/init.d/irda stop Shutting down IrDA: [FAILED] - Here's the end of dmesg and lsmod: [root@morgansmachine ~]# dmesg ... slamr: slamr0 is SiS630 card. CSLIP: code copyright 1989 Regents of the University of California PPP generic driver version 2.4.2 PPP Deflate Compression module registered NET: Registered protocol family 23 IrCOMM protocol (Dag Brattli) [root@morgansmachine ~]# lsmod | grep ir irnet 27397 0 irtty_sir 6209 0 sir_dev 20813 1 irtty_sir ircomm_tty 30537 0 ircomm 14021 1 ircomm_tty irda 130425 4 irnet,sir_dev,ircomm_tty,ircomm crc_ccitt 2241 2 irda,ppp_async ppp_generic 39892 7 irnet,ppp_deflate,ppp_async dm_mirror 26157 0 dm_mod 58101 8 dm_snapshot,dm_zero,dm_mirror - But if I follow the instructions for starting IrDA from the howto at <http://tuxmobil.org/Infrared-HOWTO/infrared-howto-s-configuration.html#id2878405> IrDA seems to start (with swapping irtty for irtty_sir): [root@morgansmachine ~]# modprobe irda irtty_sir [root@morgansmachine ~]# lsmod | grep ir irda 130425 0 crc_ccitt 2241 2 irda,ppp_async dm_mirror 26157 0 dm_mod 58101 8 dm_snapshot,dm_zero,dm_mirror [root@morgansmachine ~]# irattach /dev/ttyS1 -s [root@morgansmachine ~]# irdadump 23:33:20.447554 xid:cmd abca3c67 > ffffffff S=6 s=0 (14) 23:33:20.537533 xid:cmd abca3c67 > ffffffff S=6 s=1 (14) 23:33:20.627517 xid:cmd abca3c67 > ffffffff S=6 s=2 (14) 23:33:20.717506 xid:cmd abca3c67 > ffffffff S=6 s=3 (14) 23:33:20.807493 xid:cmd abca3c67 > ffffffff S=6 s=4 (14) 23:33:20.897477 xid:cmd abca3c67 > ffffffff S=6 s=5 (14) 23:33:20.987467 xid:cmd abca3c67 > ffffffff S=6 s=* morgansmachine hint=0400 [ Computer ] (30) 23:33:23.447100 xid:cmd abca3c67 > ffffffff S=6 s=0 (14) 23:33:23.537077 xid:cmd abca3c67 > ffffffff S=6 s=1 (14) 23:33:23.627064 xid:cmd abca3c67 > ffffffff S=6 s=2 (14) 23:33:23.717048 xid:cmd abca3c67 > ffffffff S=6 s=3 (14) 23:33:23.807038 xid:cmd abca3c67 > ffffffff S=6 s=4 (14) 23:33:23.897023 xid:cmd abca3c67 > ffffffff S=6 s=5 (14) 23:33:23.987010 xid:cmd abca3c67 > ffffffff S=6 s=* morgansmachine hint=0400 [ Computer ] (30) 14 packets received by filter [root@morgansmachine ~]# dmesg ... slamr: slamr0 is SiS630 card. CSLIP: code copyright 1989 Regents of the University of California PPP generic driver version 2.4.2 PPP Deflate Compression module registered NET: Registered protocol family 23 IrCOMM protocol (Dag Brattli) NET: Unregistered protocol family 23 irda: Unknown parameter `irtty' NET: Registered protocol family 23 NET: Unregistered protocol family 23 irda: Unknown parameter `irtty_sir' NET: Registered protocol family 23 [root@morgansmachine ~]# - Also, kIrDAd seems to stay 'running' after '/etc/rc.d/init.d/irda stop' and won't be killed.
I also ran into this. I even "patched" /etc/init.d/irda (and /etc/rc.d/init.d/functions) to correct this. Sadly, the machine I have this patch on, is now in repair. I'll try to attach my patch shortly. The problem is caused by daemon /usr/sbin/irattach ${DEVICE} ${ARGS} in start(). My analysis: irattach will fork quite quickly and the old process will exit(0) directly after forking. This has as a side effect that the initscript will always return success. All subsequent errors in daemon mode (for instance errors in [irattach/irattach.c:] start_tty()) will not be noticed by this initscript. (A less important issue is that the initscript uses /var/lock/subsys/irda. I think it would be more elegant to use irattachs pidfile: /var/run/irattach.pid. I corrected that also in my patch.) Please prod me if I haven't attached my patch in a few days.
Created attachment 119282 [details] patch to /etc/rc.d/init.d/irda (and to /etc/rc.d/init.d/functions) As described in my earlier comment: - adds --quiet option to daemon() in /etc/rc.d/init.d/functions (that file is part of the initscripts rpm) - uses that --quiet option in /etc/rc.d/init.d/irda - uses /var/run/irattach.pid and not /var/lock/subsys/irda I'd appreciate any comments.
I've added your patch to the irda initscript, but didn't add --quiet to functions. You need to propose this in a new bugzilla filed against initscripts