Well, as I said: this is not a new problem, I have been experiencing it for quite some time now.
It occurs quite irregular, sometimes 4 times in a row, sometimes not at all for a whole evening... Now that Myth can cope with it without a hard hang, it is bearable, but it would be better if it would not happen at all. It seems to occur less frequently since I moved my mysql files to a different set of disks than the main filestore used by Myth and after re-shuffling all my disks so the load is more evenly spread over all controllers. I'm using software RAID: 2 disks as RAID1 for root, 4 disks as RAID5 for datastore, using a Promise 20269 controller on the PCI bus to create enough busses for all the disks. Possibly the traffic from the PVRs and to the IDE controller sometimes clash and create a timing problem... Thinking of moving either the disks or the PVRs to a separate (EPIA?) board-based miniature system or doing something with Xen to separate functionality within the box (it is simply doing too much at the moment to determine the root cause of problems). Possibly even this weekend... Stanley. > Hmm, that's decoder stuff. Unfortunately, that's not my area of > expertise. Perhaps John Harvey can help here? > > Anyway, I think it is unrelated to the frequency change fix. > > Hans > > On Friday 16 June 2006 23:53, Stanley Merkx wrote: >> Well, perhaps the hang-while-watching-recording is not completely >> Myth's fault: It just occured while I was looking for something else >> with dmesg. >> >> I get a large amount of the following messages: >> >> ivtv0 warning: DEC: Sched Buffer end reached 0x0488efe4 >> ivtv0 warning: DEC: Mailbox 10: 0x00000000 0x0487efe4 0x0487efe4 >> 0x0487efe4 ivtv0 warning: DEC: Sched Buffer end reached 0x0488efe4 >> ivtv0 warning: DEC: Mailbox 10: 0x00000000 0x0487efe4 0x0487efe4 >> 0x0487efe4 ivtv0 warning: DEC: Sched Buffer end reached 0x0488efe4 >> ivtv0 warning: DEC: Mailbox 10: 0x00000000 0x0487efe4 0x0487efe4 >> 0x0487efe4 ivtv0 warning: DEC: Sched Buffer end reached 0x0488efe4 >> ivtv0 warning: DEC: Mailbox 10: 0x00000000 0x0487efe4 0x0487efe4 >> 0x0487efe4 >> >> Regards, >> Stanley. >> >> > Hi Hans, >> > >> > I have been running 0.4.6 from almost the moment you asked us to >> > test it. >> > >> > I'm having no problems at all with a P4, Hyperthreading enabled, on >> > Debian with a self compiled 2.6.15 kernel, 2 PVR350's and MythTV >> > 0.19-fixes. I do occasionally experience a lockup while watching a >> > recording, but this looks like it happens when a Myth recording >> > ends or a new one starts while watching another. >> > This is most likely a Myth problem, as currently I can fix this by >> > pressing Esc and then Enter to carry on playing, Myth 0.18 did a >> > real hard lockup (power cycle required). Do not have sufficient >> > logs to report to the MythTV list(s)... >> > >> > One clear difference I do notice compared to 0.4.5 when changing >> > channels in Myth LiveTV: >> > >> > (sequence of events:) >> > 0.4.5: >> > press key to change channel >> > display freezes for ~2 seconds >> > display runs for ~.5 second >> > display freezes >> > channel changes >> > >> > 0.4.6: >> > press key to change channel >> > display freezes for ~2 seconds >> > display runs for ~.5 seconds >> > display freezes >> > display runs for ~.5 seconds >> > channel changes >> > >> > So with 0.4.6 it seems to take slightly longer to change the >> > channel and the channel actually changes while video is running. In >> > 0.4.5 the channel change only occurs when the video is frozen. >> > >> > I suspect this is (partly?) caused by Myth, as the behaviour >> > experienced in 0.4.5 was introduced by a MythTV update, not by an >> > ivtv update. But perhaps some (subtle) change in ivtv >> > activated/worsened this behaviour in Myth? >> > >> > I do recall a thread not so long ago about not needing to stop the >> > decoder when changing channels... Could this be related?? >> > >> > Thanks for the good work! (how's the budget for the new monitor >> > coming along? still need donations? ;-) >> > >> > Regards, >> > Stanley. >> > >> >> Hi all, >> >> >> >> I'd like people to have a go with the 0.4.6 and 0.6.3 development >> >> snapshots available here: >> >> >> >> 0.4.6: >> >> http://ivtvdriver.org/viewcvs/ivtv/branches/0.4.tar.gz?view=tar >> >> >> >> 0.6.3: >> >> http://ivtvdriver.org/viewcvs/ivtv/branches/0.6.tar.gz?view=tar >> >> >> >> The channel changing problem should be fixed now (and should be a >> >> bit smoother too as the encoder is no longer paused/resumed). >> >> Please let me know if there are still problems with it. Also test >> >> if you didn't have problems before: it's done differently now >> >> (similar to what Hauppauge does) so if I'm unlucky it might break >> >> something that used to work :-( >> >> >> >> There are two pending issues that I'd like to solve before I make >> >> another release: 0.4.6 contains support for cx25840 sliced VBI for >> >> NTSC: I get conflicting reports on whether it works or not. >> >> >> >> The second issue is a problem with the i2c bus and tveeprom. This >> >> is the blocking bug that I'd really like to fix before doing a >> >> release. >> >> >> >> Thanks, >> >> >> >> Hans >> >> >> >> _______________________________________________ >> >> ivtv-devel mailing list >> >> [email protected] >> >> http://ivtvdriver.org/mailman/listinfo/ivtv-devel >> > >> > _______________________________________________ >> > ivtv-devel mailing list >> > [email protected] >> > http://ivtvdriver.org/mailman/listinfo/ivtv-devel >> >> _______________________________________________ >> ivtv-devel mailing list >> [email protected] >> http://ivtvdriver.org/mailman/listinfo/ivtv-devel > > _______________________________________________ > ivtv-devel mailing list > [email protected] > http://ivtvdriver.org/mailman/listinfo/ivtv-devel > _______________________________________________ ivtv-devel mailing list [email protected] http://ivtvdriver.org/mailman/listinfo/ivtv-devel
