| 00:35:37 | | Quit Oksana (Ping timeout: 264 seconds) |
| 00:40:23 | *** | Saving seen data "./dancer.seen" |
| 00:43:40 | | Join SovietShaman [0] (quassel@024-217-039-226.res.spectrum.com) |
| 00:47:08 | | Join mendelmunkis [0] (~mendelmun@ool-ae2cb138.dyn.optonline.net) |
| 00:49:25 | | Quit mendel_munkis (Ping timeout: 240 seconds) |
| 00:50:14 | | Quit t0mato (Quit: Ping timeout (120 seconds)) |
| 01:00 |
| 01:00:32 | | Join t0mato [0] (~t0mato@193.32.127.155) |
| 01:07:42 | | Quit pixelma (Quit: .) |
| 01:07:42 | | Quit amiconn (Quit: http://quassel-irc.org - Chat comfortably. Anywhere.) |
| 01:10:33 | | Join amiconn [0] (jens@rockbox/developer/amiconn) |
| 01:10:34 | | Join pixelma [0] (marianne@rockbox/staff/pixelma) |
| 01:31:52 | | Join Oksana [0] (~Wikiwide@Maemo/community/ex-council/Wikiwide) |
| 01:37:47 | | Quit Oksana (Ping timeout: 240 seconds) |
| 02:00 |
| 02:01:19 | | Quit massiveH (Quit: Leaving) |
| 02:40:25 | *** | Saving seen data "./dancer.seen" |
| 02:48:42 | | Join petur [0] (~petur@205.153.122.11) |
| 02:48:43 | | Quit petur (Changing host) |
| 02:48:43 | | Join petur [0] (~petur@rockbox/developer/petur) |
| 03:00 |
| 03:13:52 | | Join advcomp2019_ [0] (~advcomp20@65-131-168-14.sxct.qwest.net) |
| 03:13:53 | | Quit advcomp2019_ (Changing host) |
| 03:13:53 | | Join advcomp2019_ [0] (~advcomp20@unaffiliated/advcomp2019) |
| 03:17:01 | | Quit advcomp2019 (Ping timeout: 258 seconds) |
| 04:00 |
| 04:40:26 | *** | Saving seen data "./dancer.seen" |
| 04:48:41 | | Join pamaury [0] (~pamaury@rockbox/developer/pamaury) |
| 05:00 |
| 05:02:39 | | Quit mendelmunkis (Read error: Connection reset by peer) |
| 05:02:44 | | Join mendel_munkis_ [0] (~mendelmun@ool-ae2cb138.dyn.optonline.net) |
| 05:39:46 | | Join lebellium [0] (~lebellium@2a01cb080a0f0b00059618717313e0cb.ipv6.abo.wanadoo.fr) |
| 06:00 |
| 06:40:30 | *** | Saving seen data "./dancer.seen" |
| 07:00 |
| 07:45:28 | | Join mendelmunkis [0] (~mendelmun@ool-ae2cb138.dyn.optonline.net) |
| 07:45:39 | | Quit mendel_munkis_ (Read error: Connection reset by peer) |
| 07:55:27 | | Join robertd [0] (c9d14cf7@201.209.76.247) |
| 07:57:45 | | Quit robertd (Remote host closed the connection) |
| 08:00 |
| 08:03:25 | | Join robertd [0] (c9d14cf7@201.209.76.247) |
| 08:17:08 | | Quit pamaury (Ping timeout: 256 seconds) |
| 08:19:34 | | Join Oksana_ [0] (~Wikiwide@Maemo/community/ex-council/Wikiwide) |
| 08:36:59 | | Join mendel_munkis_ [0] (~mendelmun@ool-ae2cb138.dyn.optonline.net) |
| 08:39:03 | | Quit mendelmunkis (Ping timeout: 265 seconds) |
| 08:40:34 | *** | Saving seen data "./dancer.seen" |
| 08:56:56 | | Join __Bilgus [0] (41ba23be@65.186.35.190) |
| 09:00 |
| 09:00:02 | | Nick mendel_munkis_ is now known as mendel_munkis (~mendelmun@ool-ae2cb138.dyn.optonline.net) |
| 09:00:30 | __Bilgus | speachy yes post increment there thanks, I actually got really excited when I started reading about the touchscreen controller being used for a button matrice |
| 09:01:00 | __Bilgus | unfortunately xduoo did not read the application notes |
| 09:16:20 | speachy | the x3 is quite hack-y. :/ |
| 09:21:52 | | Join pamaury [0] (~pamaury@rockbox/developer/pamaury) |
| 09:40:51 | | Nick Oksana_ is now known as Oksana (~Wikiwide@Maemo/community/ex-council/Wikiwide) |
| 10:00 |
| 10:14:11 | | Join CATTSNAPPLER [0] (aef680b0@176.sub-174-246-128.myvzw.com) |
| 10:15:50 | CATTSNAPPLER | Hello I have a question about installing rockbox on a ipod classic it seems to see my ipod but then when I go to install it it always fails I dont know what to do i also saved my log file on my pc if that helps I can jump on my pc and upload it |
| 10:16:12 | | Quit CATTSNAPPLER (Remote host closed the connection) |
| 10:16:36 | braewoods | lol |
| 10:17:01 | braewoods | 22 seconds. New record. |
| 10:17:26 | | Join CATTSNAPPLER [0] (aef680b0@176.sub-174-246-128.myvzw.com) |
| 10:18:08 | __Bilgus | CATTSNAPPLER does it boot and say firmware not found? |
| 10:18:54 | CATTSNAPPLER | No I'm not able to install anything at all |
| 10:19:08 | braewoods | CATTSNAPPLER: Did you update the original firmware first? |
| 10:19:24 | CATTSNAPPLER | How do I do that? |
| 10:19:24 | braewoods | or ensure it was installed properly? some people modify their ipods with more capacity. |
| 10:19:49 | braewoods | CATTSNAPPLER: iTunes somehow. I'm not familiar with Apple products that much. |
| 10:20:07 | CATTSNAPPLER | My ipod is a stock classic nothin changed |
| 10:20:32 | braewoods | Ok. |
| 10:21:07 | braewoods | those that do need to reinstall the firmware usually since it's stored on the internal storage alongside the user files. |
| 10:21:38 | braewoods | i would try reinstalling the original firmware to latest release perhaps even reformatting it and try again |
| 10:22:02 | __Bilgus | CATTSNAPPLER (un?)fortunately I don't have any experience with the ipods we have users that are but it could take em a while to come out of the wood work, you might check out the forums though |
| 10:22:04 | braewoods | a lot of times stuff expects latest original stuff and can create compatibility issues if it isn't |
| 10:22:09 | CATTSNAPPLER | I had already reformatted it |
| 10:22:21 | braewoods | No idea then. |
| 10:22:44 | speachy | so itunes sees it as an ipod, and lets you interact with it? |
| 10:22:55 | CATTSNAPPLER | Yep |
| 10:23:00 | speachy | and it boots into the ipod firmware cleanly |
| 10:23:15 | CATTSNAPPLER | Yep X2 |
| 10:23:16 | braewoods | is this your rockbox install? |
| 10:23:21 | braewoods | your first |
| 10:23:34 | CATTSNAPPLER | yea it's my first time doin the tango with rockbox |
| 10:23:43 | braewoods | normal behavior of a successful install is that it boots into rockbox not the stock firmware |
| 10:23:52 | braewoods | you need to do something special if it worked to boot OF |
| 10:24:02 | braewoods | at least for dual boot targets |
| 10:24:05 | CATTSNAPPLER | Thatd be great if I could even get rockbox installed |
| 10:24:36 | CATTSNAPPLER | I dont want to dual boot I just wanna turn my ipod into a drag and drop thing |
| 10:24:36 | braewoods | i've used iriver h10 and sandisk sansa clip+ targets before |
| 10:25:12 | braewoods | i see. rockbox's USB mode exposes everything through UMS. no MTP support. |
| 10:25:25 | CATTSNAPPLER | Huh? |
| 10:25:42 | braewoods | meaning... you can't just unplug it whenever you feel like. |
| 10:25:53 | braewoods | you need to properly eject it like any other removeable drive |
| 10:26:12 | braewoods | MTP allows that but it has to be supported. |
| 10:26:45 | braewoods | CATTSNAPPLER: sorry, it just effects how the internal storage of the device is exposed to the host computer. |
| 10:26:46 | CATTSNAPPLER | Oooohhhh that doesnt really bug me I just want this to be done I've been trying to figure this out all night and well to no avail |
| 10:27:05 | braewoods | UMS is basically, it exposes all internal storage as removeable drives |
| 10:27:11 | braewoods | it's USB Mass Storage |
| 10:27:18 | CATTSNAPPLER | Ah ok |
| 10:27:32 | braewoods | not the safest option but it works the same. |
| 10:27:40 | braewoods | MTP is safer but rockbox doesn't implement it. |
| 10:27:47 | CATTSNAPPLER | Is anyone familiar with rockbox's log files? Bc I cant read it |
| 10:28:07 | braewoods | you can try pasting it to something like dpaste.com |
| 10:28:35 | CATTSNAPPLER | I'll put it up on paste bin gimme a secound |
| 10:29:00 | braewoods | ugh. why do people always want to use pastebin.com? it's full of distracting ads and other nuisances. lol |
| 10:30:29 | CATTSNAPPLER | It's the most approachable for me ig? |
| 10:30:36 | * | braewoods shrugs. |
| 10:30:40 | __Bilgus | gotta support it somehowe |
| 10:30:42 | braewoods | there's tons of just as easy to use ones. |
| 10:30:50 | braewoods | for free and less non-sense. |
| 10:31:18 | braewoods | just something i've observed from moderating #learnprogramming |
| 10:31:28 | __Bilgus | that haven't gotten to the scale of needing to add ad support |
| 10:31:32 | braewoods | seems to be a common problem in any channel where pastebins are used |
| 10:31:45 | CATTSNAPPLER | pastebin.com/nxDRL8mL |
| 10:32:11 | __Bilgus | I did click on some BS pastebin the other day that wanted me to do a captcha then failed to actually post my paste |
| 10:32:36 | braewoods | there's some pastebins i don't think will ever go ad supported |
| 10:32:39 | speachy | CATTSNAPPLER: ht looks like it couldn't actually write to the ipod. |
| 10:32:46 | braewoods | hm. |
| 10:32:53 | CATTSNAPPLER | How do I fix it not writing? |
| 10:33:01 | braewoods | rockbox usually needs the original firmware booted into UMS mode |
| 10:33:14 | CATTSNAPPLER | How does one do that? |
| 10:33:16 | braewoods | did you do anything special for that? |
| 10:33:53 | CATTSNAPPLER | I just plugged it into my pc and it saw it as a ipod / external storage thingy right off the bat |
| 10:34:33 | braewoods | https://download.rockbox.org/daily/manual/rockbox-ipod6g.pdf |
| 10:34:37 | braewoods | did you check the manual? |
| 10:34:52 | speachy | [ziputil.cpp:126 INFO] creating output path failed for: "D://.rockbox/tagnavi.config" |
| 10:34:54 | CATTSNAPPLER | Yea I have |
| 10:35:08 | CATTSNAPPLER | What does that mean? |
| 10:35:08 | braewoods | did you follow the steps for usb connection? |
| 10:35:18 | CATTSNAPPLER | I thought I did |
| 10:35:23 | braewoods | hm. |
| 10:35:27 | braewoods | how full is it? |
| 10:35:42 | CATTSNAPPLER | I had formatted it so its empty |
| 10:36:00 | __Bilgus | You also need to ensure the “Enable use as disk” option is enabled for your playerin Itunes << |
| 10:36:15 | CATTSNAPPLER | How do I do that? |
| 10:36:19 | __Bilgus | if its not that then you need to make sure its actually mounted at D: |
| 10:36:32 | | Quit robertd (Remote host closed the connection) |
| 10:36:36 | braewoods | are you actually using itunes currently? |
| 10:36:43 | __Bilgus | i'm guessing you go fumble in itunes for the proper setting |
| 10:36:46 | CATTSNAPPLER | I would assume I am |
| 10:36:52 | braewoods | ... |
| 10:37:01 | braewoods | either it is or it isn't installed |
| 10:37:09 | braewoods | and it's not a default application so we wouldn't know. |
| 10:37:40 | CATTSNAPPLER | I have itunes installed and all that jazz along with the ipod utilities |
| 10:37:44 | braewoods | so, what are you using to set it up? |
| 10:37:49 | braewoods | windows? |
| 10:38:13 | CATTSNAPPLER | Yep |
| 10:38:16 | __Bilgus | the D: mountpoint points to windows |
| 10:38:21 | __Bilgus | d : |
| 10:38:24 | braewoods | ok. just confirming |
| 10:38:26 | __Bilgus | lol |
| 10:38:32 | CATTSNAPPLER | My ipod was the d drive |
| 10:38:43 | CATTSNAPPLER | My windows is on c |
| 10:39:39 | braewoods | https://www.copytrans.net/support/how-to-use-the-ipod-as-a-storage-drive-enable-disk-use/ |
| 10:40:37 | *** | Saving seen data "./dancer.seen" |
| 10:41:05 | braewoods | speachy: i noticed archos was "retired". what determines when devices are "retired"? |
| 10:41:48 | CATTSNAPPLER | When theres no support for em anymore? |
| 10:41:54 | speachy | braewoods: "stable but will not be supported going forward" |
| 10:42:43 | braewoods | speachy: i'm guessing it was removed due to how limited of a platform it was. |
| 10:42:49 | braewoods | and probably ancient |
| 10:43:05 | speachy | ie no further builds. Code's been removed from git. |
| 10:43:39 | CATTSNAPPLER | I'm gonna try that one thing braewoods sent in see if that works if not I'll be back |
| 10:44:12 | braewoods | either way need to make sure it's exposed correctly |
| 10:44:21 | speachy | braewoods: yes, ancient and very limited. though ironically more capable than the current generation of low-end DAPs. |
| 10:44:26 | CATTSNAPPLER | Wish me luck |
| 10:45:02 | braewoods | speachy: how much RAM do most portalplayer targets have? i can't usually find much about that in the docs. |
| 10:45:59 | braewoods | i expect it to be similar to the amount found in most consumer routers (64 MB or lessish) |
| 10:46:08 | speachy | IIRC all portalplayers are 32 (or maybe 64?) MB. |
| 10:46:17 | braewoods | ah. |
| 10:46:31 | braewoods | the spec sheet suggested it could handle up to 256 or so |
| 10:46:33 | braewoods | so i wasn't sure |
| 10:46:37 | braewoods | the platform |
| 10:46:39 | braewoods | at least |
| 10:49:02 | speachy | all of the iPods had 32MB, even the non-PP ones |
| 10:49:40 | speachy | no wait, the samsung-based ones are 64MB. |
| 10:52:26 | CATTSNAPPLER | So I'm back |
| 10:52:28 | CATTSNAPPLER | Back* |
| 10:52:38 | CATTSNAPPLER | And no dice it's still wanting to be an ass |
| 10:54:49 | __Bilgus | speachy whereas _battery_voltage is called a bunch but it is only called once a minute once everything is setup |
| 10:55:24 | __Bilgus | so I think we could go to a poll on the adc with a timeout like the Sansas do |
| 10:55:43 | speachy | polling is fine, as long as it doesn't happen in interrupt context... |
| 10:56:18 | CATTSNAPPLER | Welp I dont know what to do anymore |
| 11:00 |
| 11:01:04 | braewoods | CATTSNAPPLER: does the utility ask to run as Admin? |
| 11:01:10 | braewoods | when you first run it. |
| 11:01:49 | | Quit CATTSNAPPLER (Remote host closed the connection) |
| 11:17:10 | __Bilgus | https://gerrit.rockbox.org/r/#/c/2721 |
| 11:17:15 | | Quit __Bilgus (Remote host closed the connection) |
| 11:18:11 | | Join _Bilgus_ [0] (41ba23be@65.186.35.190) |
| 11:48:30 | | Quit beencubed (Quit: Leaving) |
| 12:00 |
| 12:00:58 | | Quit _Bilgus_ (Remote host closed the connection) |
| 12:40:39 | *** | Saving seen data "./dancer.seen" |
| 12:45:01 | | Join __Bilgus [0] (41ba23be@65.186.35.190) |
| 12:46:36 | __Bilgus | speachy per your gerrit comment, I just copied the first one from the code further down but the second one the manual states: VBAT Enable Control. No matter TCHEN is 1 or 0, VBATEN can be set to 1 to sample the voltage of battery, and when the value of voltage is ready, PBATEN will be cleared by hardware auto. |
| 12:47:35 | __Bilgus | is there a reason to use the state value instead? I figured this way there wasn't another register getting clobbered |
| 12:51:11 | fs-bluebot_ | Build Server message: New build round started. Revision fa330c1, 280 builds, 9 clients. |
| 12:52:50 | bluebrother^ | since it came up in the forums ... I'm not planning to do actual work on it, but I made the theme editor build with Qt5. |
| 13:00 |
| 13:07:46 | fs-bluebot_ | Build Server message: Build round completed after 995 seconds. |
| 13:07:49 | fs-bluebot_ | Build Server message: Revision fa330c1 result: All green |
| 13:10:30 | fs-bluebot_ | Build Server message: New build round started. Revision 93bb14c, 280 builds, 9 clients. |
| 13:18:40 | __Bilgus | Haha I was reading the manual ' 13nm CMOS process' and I was like wow! |
| 13:19:40 | | Quit JanC (Remote host closed the connection) |
| 13:20:07 | | Join JanC [0] (~janc@lugwv/member/JanC) |
| 13:20:42 | speachy | yeah, I think they left out a digit. :D |
| 13:22:49 | bluebrother^ | why do people post Rockbox questions on reddit instead of the Rockbox forum? |
| 13:24:05 | __Bilgus | probably because someone keeps answering |
| 13:24:35 | speachy | hey, it's arguably better there than on head-fi or somesuch |
| 13:25:00 | * | bluebrother^ is surprised how much question he missed by simply not using reddit |
| 13:28:28 | | Join johnb2 [0] (~johnb2@p5b3af327.dip0.t-ipconnect.de) |
| 13:32:12 | fs-bluebot_ | Build Server message: Build round completed after 1302 seconds. |
| 13:32:14 | fs-bluebot_ | Build Server message: Revision 93bb14c result: All green |
| 13:32:15 | fs-bluebot_ | Build Server message: New build round started. Revision 6810d83, 280 builds, 9 clients. |
| 13:32:24 | speachy | __Bilgus, did you confirm what context _battery_voltage() runs under? tick context, or via an async thread? |
| 13:32:48 | __Bilgus | its an async thread run by the kernel |
| 13:33:16 | __Bilgus | well the repeated caller the other places its called are user driven I believe |
| 13:33:17 | speachy | excellent. |
| 13:35:15 | speachy | so the patch changes it to measure at most once every 30s. |
| 13:35:32 | speachy | a shame we can't power down the ADC. |
| 13:36:43 | __Bilgus | well really we could since we now check for a button press in interrupt |
| 13:37:35 | __Bilgus | 70ms to boot back up |
| 13:37:42 | speachy | yeah |
| 13:37:52 | speachy | that's a bit long for checking keypresses. |
| 13:38:24 | __Bilgus | yeah thats out of full power down though the sleep mode might be faster |
| 13:39:11 | __Bilgus | it claims 1ma worst case I doubt there is too much saving there |
| 13:39:26 | speachy | that's more than I thought |
| 13:39:38 | __Bilgus | unless its actually 10ma |
| 13:40:08 | speachy | heh, it's funny, the gig before this one we agonized over shaving off nanowatts in sleep. |
| 13:42:45 | __Bilgus | I'm more concerned about proper operation first |
| 13:43:35 | speachy | I recently found reports from a few years ago that claimed ~12h battery life on the X3 with rockbox −− well prior to all of our efforts to improve things.. and your battery bench run was essentially unchanged from that. |
| 13:43:44 | | Quit pamaury (Ping timeout: 240 seconds) |
| 13:44:40 | __Bilgus | well with LPM disabled its probably less ATM |
| 13:45:51 | speachy | possibly, but tbh I'd be surprised if the sum total of our improvements even make a 5% overall difference. |
| 13:46:08 | speachy | judging by the USB power meter, the base consumption of this thing is embarrasingly high. |
| 13:46:50 | fs-bluebot_ | Build Server message: Build round completed after 876 seconds. |
| 13:46:51 | fs-bluebot_ | Build Server message: Revision 6810d83 result: All green |
| 13:47:05 | __Bilgus | explains the comforting warmth |
| 13:48:15 | __Bilgus | I just try to keep in mind the changes that SOmeone did to the AS3525 core in the name of 10 minutes more battery and the number of times Ive traced issues to them and reverted |
| 13:48:43 | speachy | yeah. the X3's reclocking is a good example of that, eh? |
| 13:48:59 | __Bilgus | nah those issues were already there |
| 13:50:00 | speachy | at least there don't seem to be any remaining crashes/etc but there's still that barely-audible artifact when the clocks switch. (introduced via the analog path, I believe −− we're not seeing a buffer underrun) |
| 13:50:30 | | Join sakax [0] (~r0b0t@unaffiliated/r0b0t) |
| 13:50:57 | __Bilgus | you mean without running plugins? |
| 13:51:02 | speachy | yeah |
| 13:51:07 | johnb2 | speachy : what are the conditions where you think the constant 192MHz were not enough? EQ, wavpack,... ? For me, just playing songs it seemed fine. |
| 13:51:16 | __Bilgus | AH ok |
| 13:52:13 | johnb2 | and, I don't use plugins (except for batteryBench). |
| 13:52:39 | __Bilgus | So I noticed something about the AHB scheduler but I don't see the defines in the sources |
| 13:54:46 | speachy | hunting down those writes was a fun exercise |
| 13:56:27 | speachy | here's the funny thing −− the PM says only the first two bits are used; the rest are reserved. |
| 13:57:32 | __Bilgus | makes me wonder if they didn't finish the damn thing |
| 13:57:35 | speachy | but there are clearly some documentation um, inconsistencies |
| 13:58:13 | speachy | bits defined/used in the header that don't exist (or are "set to 0") in the docs |
| 13:58:47 | | Quit johnb2 (Quit: Nettalk6 - www.ntalk.de) |
| 14:00 |
| 14:00:06 | speachy | __Bilgus: 12.2.1 of the PM defines HARBx_PRIOR |
| 14:01:04 | speachy | and table 41-2 shows the hole in the address map where that missing write goes |
| 14:01:31 | | Quit petur (Quit: Connection reset by beer) |
| 14:01:57 | __Bilgus | nice thanks |
| 14:02:55 | | Join johnb4 [0] (~johnb2@p5b3af327.dip0.t-ipconnect.de) |
| 14:02:57 | speachy | s/missing/unexplained/ |
| 14:14:23 | | Quit johnb4 (Ping timeout: 260 seconds) |
| 14:19:48 | fs-bluebot_ | Build Server message: New build round started. Revision 1ae8213, 280 builds, 8 clients. |
| 14:36:16 | fs-bluebot_ | Build Server message: Build round completed after 988 seconds. |
| 14:36:17 | fs-bluebot_ | Build Server message: Revision 1ae8213 result: All green |
| 14:39:24 | | Join pamaury [0] (~pamaury@rockbox/developer/pamaury) |
| 14:40:42 | *** | Saving seen data "./dancer.seen" |
| 14:41:59 | | Join johnb4 [0] (~johnb2@p5b3af327.dip0.t-ipconnect.de) |
| 14:43:10 | | Nick Guest60328 is now known as alexbobp (~alex@meowface.org) |
| 14:47:31 | __Bilgus | speachy https://github.com/Rockbox/rockbox/blob/master/firmware/target/mips/ingenic_jz47xx/system-jz4760.c#L694 |
| 14:47:45 | __Bilgus | whats the provenance on that init? |
| 14:48:22 | | Quit johnb4 (Ping timeout: 246 seconds) |
| 14:51:39 | speachy | __Bilgus: no clue |
| 14:52:00 | speachy | I'm actually looking at other codebases to see if I can suss something out |
| 14:52:05 | __Bilgus | I assume if you start cube your player still crashes |
| 14:52:09 | speachy | nope |
| 14:52:21 | __Bilgus | no? |
| 14:53:04 | speachy | I even have a revised caching patch in place |
| 14:53:50 | speachy | a few audio artifacts but it's spinning madly |
| 14:55:20 | speachy | master + updated cache patch (not yet committed) and that WIP tweak set |
| 14:55:46 | speachy | oh, reclocking too |
| 14:56:24 | speachy | switched to fire |
| 14:56:43 | | Quit __Bilgus (Remote host closed the connection) |
| 15:00 |
| 15:03:19 | speachy | ok, g#2709 is updated, and seems stable so far. |
| 15:03:51 | | Join johnb2 [0] (~johnb2@p5b3af327.dip0.t-ipconnect.de) |
| 15:03:57 | speachy | I think the reason it was making things worse was due to hyper-aggressive dumping of the entire cache |
| 15:04:17 | | Join __Bilgus [0] (41ba23be@65.186.35.190) |
| 15:05:27 | __Bilgus | speachy I haven't crashed at head either sofar |
| 15:06:29 | __Bilgus | but that init per the namual says the bits we do know are set 3: {p0 bridge, gps, uhc}, {eth, dma, p1 bridge, otg} |
| 15:07:11 | speachy | yeah, I copied that out of the manual. because I didn't like undocumented magic numbers |
| 15:07:41 | __Bilgus | wouldn't it be better to put everyone on the same footing with {otg, gps, uhc}, {eth, dma, p1 bridge, p0 bridge} |
| 15:08:30 | __Bilgus | unless uhc handles something else besides uart |
| 15:10:20 | speachy | what I want is to give DMA the highest priority |
| 15:11:29 | __Bilgus | none of those give dma preffered priority |
| 15:12:36 | speachy | but we're not bus-starved |
| 15:13:19 | __Bilgus | no? I was under the impression that why the glitches happen under high load |
| 15:14:49 | speachy | I'm only considering audio glitches for the moment; I think that's entirely due to interrupt processing delays (from heavy loads) |
| 15:15:34 | speachy | (for reasons as yet to be determined..) |
| 15:16:02 | speachy | now audio dma dropout shouldn't cause the hangs/crashes we were also seeing. I don't have a theory for that. |
| 15:17:04 | __Bilgus | unfortunately there is no specific thing that made it better either it seems oh just got a crash' |
| 15:17:32 | johnb2 | Just installed current HEAD. The clicks are gone for me! |
| 15:17:35 | speachy | for me the crashes were nearly always following audio artifacts |
| 15:18:06 | __Bilgus | it was after 5 mins of cube with audio artifacts constant |
| 15:18:46 | speachy | __Bilgus: for me the crashes tended to be pretty rapid |
| 15:19:00 | speachy | but −− I have another theory about that now, in hindsight |
| 15:19:14 | speachy | relating to the caching patch I just updated |
| 15:19:37 | __Bilgus | oh you haven't pushed that yet have you |
| 15:19:48 | __Bilgus | so lets see if I can repro then switch to that |
| 15:20:02 | speachy | not yet, but it's already outlasted its previous iteration by a factor of 10, under heavy load too. |
| 15:21:25 | speachy | I think the critical change was changing the inline ASM definition of SYNC_WB from "sync" to "sync":::"memory" |
| 15:22:10 | speachy | to tell GCC that it can clobber memory |
| 15:23:00 | speachy | and therefore to treat it as a proper memory barrier |
| 15:23:29 | __Bilgus | oh I didn't catch that change |
| 15:23:36 | speachy | just added |
| 15:24:34 | __Bilgus | nice that could be our smoking gun |
| 15:24:42 | speachy | yeah, I think it very well could be |
| 15:27:14 | | Quit Moarc_ (Quit: i znowu NADMUCHAŁ BALONA) |
| 15:29:38 | | Join Moarc [0] (~chujko@a105.net128.okay.pl) |
| 15:33:41 | | Quit johnb2 (Quit: Nettalk6 - www.ntalk.de) |
| 15:35:35 | fs-bluebot_ | Build Server message: New build round started. Revision 0cb162a, 280 builds, 9 clients. |
| 15:35:47 | __Bilgus | ok I can get a crash at head eventually seemingly randomly i'll try 2709 now |
| 15:35:51 | speachy | ok, it's in. after some tab excisions |
| 15:40:29 | speachy | ... SD-over-USB writes seem to be a bit faster with this too. |
| 15:40:35 | speachy | hmm, this is freaky. |
| 15:40:43 | speachy | I've lost all buttons but power |
| 15:42:26 | __Bilgus | rolo |
| 15:42:34 | __Bilgus | reboot and it'll be back |
| 15:42:38 | speachy | the button rework might have a missing corner case. |
| 15:42:52 | __Bilgus | thats what I mentioned yesterday but when i recompiled it was gone |
| 15:43:04 | __Bilgus | nah I got it prior to that |
| 15:43:13 | speachy | first time I ever had something like that happen |
| 15:43:14 | __Bilgus | its random as well |
| 15:43:30 | __Bilgus | 1 in 10 probably |
| 15:43:55 | __Bilgus | but only from rolo' |
| 15:44:22 | speachy | this happened when paused on the main menu with the screen off |
| 15:44:36 | speachy | hit power, it resumed, screen, everything working, just no buttons |
| 15:45:12 | __Bilgus | hmm thats new |
| 15:45:54 | __Bilgus | hopefully its reproducible |
| 15:46:19 | | Join livvy [0] (~livvy@gateway/tor-sasl/livvy) |
| 15:46:27 | __Bilgus | maybe it missed the IRQ reint? |
| 15:50:15 | __Bilgus | oh its my fault I see it |
| 15:50:20 | speachy | I think the problem is L133 |
| 15:51:58 | fs-bluebot_ | Build Server message: Build round completed after 982 seconds. |
| 15:52:00 | fs-bluebot_ | Build Server message: Revision 0cb162a result: 5 errors 2 warnings |
| 15:52:01 | __Bilgus | I think its 285 |
| 15:52:20 | speachy | ah yes |
| 15:52:27 | __Bilgus | when I messed with the pins during adc read it messed it up |
| 15:52:43 | __Bilgus | and being that its set to input atm that fits |
| 15:55:36 | | Join cockroach [0] (~blattodea@pdpc/supporter/active/cockroach) |
| 15:57:24 | speachy | so about L133 (ish), if KEY_INT is asserted, only BUTTON_POWER or BUTTON_NONE is possible? (KEY_INT is asserted for all keys?) |
| 15:57:36 | speachy | (all keys otehr than POWER, that is) |
| 15:58:21 | __Bilgus | yes INT does everything but hold and power |
| 15:58:53 | __Bilgus | BOP only does back option Play but it supplies then through the divider that gets read at INT |
| 15:59:45 | __Bilgus | thats why I thought they might have it hooked up to the ADC as a toch interface but no |
| 16:00 |
| 16:03:02 | __Bilgus | hmm not able to reproduce with either setting but i did start getting glitches in the data |
| 16:04:08 | __Bilgus | static sounds as cpu load goes up and now a TLB crash |
| 16:07:58 | __Bilgus | something something memset |
| 16:08:07 | __Bilgus | but whos calling you |
| 16:08:47 | speachy | you might want to try pulling in g#2729. |
| 16:09:41 | fs-bluebot_ | Build Server message: New build round started. Revision 48a2b80, 280 builds, 9 clients. |
| 16:09:42 | speachy | I doubt it has any bearing on the crash but it might help with the audio breakups. It's what I'm using now |
| 16:10:10 | __Bilgus | i'm adding in CPu load in the sound menu with the EQ dither etc. |
| 16:10:25 | speachy | seems prudent |
| 16:11:34 | speachy | yeah, SD writes are definitely faster now; they're scrolling by faster than before |
| 16:11:56 | speachy | (unzipping the file over USB..) |
| 16:12:45 | __Bilgus | Ah gotcha HASS surround with timestretch ON |
| 16:14:31 | __Bilgus | Oh turn on timestretch first then hass |
| 16:14:31 | speachy | well, that would do it! |
| 16:15:07 | __Bilgus | hmm theres something with the menu for timestretch too |
| 16:15:27 | __Bilgus | it has static so its probably one of those menus that sets the value every refresh |
| 16:23:21 | fs-bluebot_ | Build Server message: Build round completed after 820 seconds. |
| 16:23:23 | fs-bluebot_ | Build Server message: Revision 48a2b80 result: 5 errors 2 warnings |
| 16:24:11 | speachy | whoops, I missed that red. |
| 16:26:27 | fs-bluebot_ | Build Server message: New build round started. Revision 9a28912, 280 builds, 9 clients. |
| 16:39:39 | fs-bluebot_ | Build Server message: Build round completed after 793 seconds. |
| 16:39:41 | fs-bluebot_ | Build Server message: Revision 9a28912 result: All green |
| 16:40:46 | *** | Saving seen data "./dancer.seen" |
| 17:00 |
| 17:07:47 | __Bilgus | speachy on the IRQ handler patch I see where you short circuited the priority for the dma etc but have you looked at the asm? |
| 17:08:41 | __Bilgus | it seems like unless the compiler is good you should probably skip the switch and just return directly, then you don't need to copy IRQ either |
| 17:10:01 | __Bilgus | if (ipl0 & 1<<IRQ_DMAC0) { ipl0 &= ~(1<<IRQ_DMAC0); return __dmac_get_irq() + IRQ_DMA_0; |
| 17:13:07 | | Quit Oksana (Read error: Connection reset by peer) |
| 17:27:56 | | Quit pamaury (Ping timeout: 256 seconds) |
| 17:29:53 | speachy | I didn't try too hard to optimize cycles out of there. |
| 17:30:46 | speachy | I have some pretty heavy IRQ rework started, and part of that code is going to get pulled into asm anyway |
| 17:34:13 | __Bilgus | ok I was wondering if you had an idea yet it sounds like you do |
| 17:59:40 | __Bilgus | speachy I've been listening to it for a while now and I still hear a bit of crackle at track change but I tell you what no underruns till I did plugins and no crashes so far |
| 17:59:59 | * | speachy nods. |
| 18:00 |
| 18:02:26 | __Bilgus | hmm new one stopping a song with the power button leaves the audio going for a moment with a 40 hz tick or so |
| 18:02:58 | | Quit sakax (Remote host closed the connection) |
| 18:03:09 | speachy | are you using that 2729 patch? it ups the playback buffer size a bit |
| 18:03:40 | __Bilgus | yes oh am I hearing uninitd audio data? |
| 18:04:21 | speachy | I think ultimately we'll either have to put in double-buffering at the PCM HW API layer or completely rework the IRQ stuff. |
| 18:05:14 | __Bilgus | And. |
| 18:05:20 | speachy | Possibly.. I don't know if the pcmbuf is initialized fully. that buffer size bump was an experiment. |
| 18:05:20 | __Bilgus | not or lol |
| 18:07:08 | __Bilgus | uh oh the adc didn't stabalize before the battery log picked up 0xFFF as the voltage level 9+ v lol Really full battery |
| 18:07:20 | speachy | whoops. :) |
| 18:07:31 | __Bilgus | I was setting the initial value to 6DD but |
| 18:08:40 | __Bilgus | yep wiped it out in between patches |
| 18:12:37 | __Bilgus | weird maybe that initial read takes longer than the timeout |
| 18:29:08 | speachy | I saw something like this when I cut out a bit that wasn't documented |
| 18:38:18 | | Quit __Bilgus (Remote host closed the connection) |
| 18:40:50 | *** | Saving seen data "./dancer.seen" |
| 19:00 |
| 19:02:09 | | Quit lebellium (Quit: Leaving) |
| 19:07:24 | | Join Oksana [0] (~Wikiwide@Maemo/community/ex-council/Wikiwide) |
| 19:10:35 | | Join ac_laptop [0] (~ac_laptop@186.2.247.129) |
| 19:21:00 | | Join beencubed [0] (~beencubed@209.131.238.248) |
| 19:22:28 | | Join bluebrother [0] (~dom@rockbox/developer/bluebrother) |
| 19:22:56 | | Join fs-bluebot [0] (~fs-bluebo@55d4b206.access.ecotel.net) |
| 19:25:16 | | Quit fs-bluebot_ (Ping timeout: 265 seconds) |
| 19:25:52 | | Quit bluebrother^ (Ping timeout: 256 seconds) |
| 19:55:07 | | Join massiveH [0] (~massiveH@ool-18e4e82f.dyn.optonline.net) |
| 20:00 |
| 20:40:51 | *** | Saving seen data "./dancer.seen" |
| 20:43:17 | | Quit cockroach (Quit: leaving) |
| 21:00 |
| 21:18:46 | | Join __Bilgus [0] (41ba23be@65.186.35.190) |
| 22:00 |
| 22:12:54 | speachy | ...so the linux jz47xx pcm driver does a double-buffered DMA approach with >= 256 bytes/buffer. |
| 22:17:30 | | Quit livvy (Remote host closed the connection) |
| 22:25:09 | __Bilgus | I'll admit I didn't pay very close attention to the specs are we that tight on ram? |
| 22:40:54 | *** | Saving seen data "./dancer.seen" |
| 22:41:06 | | Quit ac_laptop (Quit: WeeChat 2.9) |
| 22:41:35 | | Join ac_laptop [0] (~ac_laptop@186.2.247.129) |
| 22:58:49 | __Bilgus | well I finally tracked down what was causing the adc to fail on ROLO |
| 22:59:37 | __Bilgus | the KEY_INT pin was never set explicitly to input |
| 23:00 |
| 23:00:18 | __Bilgus | maybe the gpios glitch? |
| 23:11:37 | | Quit TheSeven (Ping timeout: 260 seconds) |
| 23:11:56 | | Join TheSeven [0] (~quassel@rockbox/developer/TheSeven) |
| 23:16:53 | | Quit ac_laptop (Ping timeout: 260 seconds) |
| 23:51:39 | | Quit __Bilgus (Remote host closed the connection) |
icon identifies that the person is a core developer (has commit access).