Previous day | Jump to hour: 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 | Next day

Seconds: Show Hide | Joins: Show Hide | View raw
Font: Serif Sans-Serif Monospace | Size: Small Medium Large

Click in the nick column to highlight everything a person has said.
The Logo icon identifies that the person is a core developer (has commit access).

Notice: Only Gecko based browsers prior to FF4 support the multipart/mixed "server push" method used by this log reader to auto-update. Since you do not appear to use such a browser, this page will simply show the current log, and not automatically update.

#rockbox log for 2016-08-06

00:00:26 Quit edhelas (Ping timeout: 258 seconds)
00:18:54 Quit ender` (Quit: The early bird may get the worm, but the second mouse gets the cheese.)
00:19:45 Join ender` [0] (
00:27:37 Quit kapplejacks (Quit: Page closed)
00:28:13*[Saint] wonders what is vague about "MPEG-1/MPEG-2, resolution can't be greater than target resolution and must be a factor of two"
00:29:00[Saint]Reddit'll be a whole barrel of laughs.
00:29:13[Saint]"Hey, Reddit, how do I Rockbox?"
00:29:20[Saint]>Shut up, fag.
00:32:43[Saint]Seems like that place would be a barren wilderness if not for saratoga.
00:32:59[Saint]So there's at least one person there who actually knows something. WHich is pleasing.
00:33:44[Saint]Pity that his comments seem to get downvoted a lot in favor of outright misinformation or guesses.
00:35:32[Saint]well, one presumes saratoga3 is saratoga.
01:05:02 Quit ender` (Quit: The early bird may get the worm, but the second mouse gets the cheese.)
01:08:55 Quit ZincAlloy (Quit: Leaving.)
01:23:39***Saving seen data "./dancer.seen"
01:28:24 Join Bray90820_ [0] (
01:30:54 Quit Bray90820 (Ping timeout: 252 seconds)
01:54:29 Quit Tirifto (Quit: Tirifto)
02:00:01 Join [Sinner] [0] (~hayden@rockbox/staff/saint)
02:00:02 Quit [Saint] (Ping timeout: 240 seconds)
02:03:54 Join [Saint] [0] (~hayden@rockbox/staff/saint)
02:04:41 Quit [Sinner] (Read error: Connection reset by peer)
02:05:24 Quit [Saint] (Read error: Connection reset by peer)
02:06:23 Join [Saint] [0] (~hayden@rockbox/staff/saint)
02:39:38 Quit pamaury (Ping timeout: 260 seconds)
03:23:43***Saving seen data "./dancer.seen"
03:25:53 Join krabador [0] (~krabador@unaffiliated/krabador)
03:33:44 Join robertd [0] (c9f2b970@gateway/web/freenode/ip.
03:42:09 Join bluebrother [0] (~dom@rockbox/developer/bluebrother)
03:42:30 Join fs-bluebot_ [0] (
03:44:31 Quit fs-bluebot (Ping timeout: 240 seconds)
03:45:11 Quit bluebrother^ (Ping timeout: 250 seconds)
03:53:06__builtinwhy are mutexes SWCODEC only?
04:07:36 Quit prof_wolfff (Ping timeout: 250 seconds)
04:20:52 Join prof_wolfff [0] (
05:07:56 Join treaki_ [0] (
05:11:27 Quit treaki (Ping timeout: 250 seconds)
05:15:38 Join musickflgj [0] (6cdcee76@gateway/web/freenode/ip.
05:22:27[Saint]Ask your question.
05:22:56[Saint](and/or glance at the topic next time :))
05:23:47***Saving seen data "./dancer.seen"
05:25:31musickflgjThe question is, where is the about page for the rockbox? I can't find it, and I don't really understand what it is.
05:27:18[Saint]hiding in plain sight.
05:27:55[Saint]That page is hilariously out of date, though.
05:28:50[Saint]It can basically be summarized with "Why Rockbox? Your DAP firmware is a piece of shit, you know it, I know it, we all know it - that's why you went looking for a third party firmware for it - We are that firmware"
05:30:03musickflgjAh, I see. Thank you. Is the quality of the music the same or is it worse/better?
05:31:15[Saint]Better and worse is very subjective.
05:31:26[Saint]It is infinitely more flexible, I can say that.
05:32:09[Saint]Some people like the artificially coloured audio from their DAPs and think that audio when faithfully reproduced sounds "bad" or "wrong".
05:32:23[Saint]WHen in fact they're just actually hearing it reproduced as intended.
05:33:03musickflgjUnderstood. So it would get the exact audio, rather then filtered audio, for better or worse?
05:34:30[Saint]As close to it as possible, yes.
05:35:08[Saint]If it turns out you do prefer artificial colouring/shaping, we have a full parametric EQ and bass/treble high/low pass shelf filters.
05:35:43musickflgjI see. I am debating buying a DAP, and am factoring these into my considerations. Thank you very much for your time.
05:36:44[Saint]These days I really don't think there is much of a market for a dedicated audio player.
05:36:45[Saint]Most people carry around a smartphone with enough computational power that inefficient routines really don't matter too much.
05:37:03[Saint]And Joe Average has very firmly adopted MP3 as the defacto codec.
05:37:54[Saint]If you're looking at it as an audio enthusiast, then, absolutely - dedicated DAP wins.
05:38:12[Saint]If you just want to play some music occasionally, you probably already own that device.
05:39:01[Saint]Oh - one other use case I guess:
05:39:19[Saint]"Essentially disposable device you don't mind getting soaked in body fluids at the gym"
05:39:33[Saint]That's a definite use case for a dedicated DAP.
05:40:25[Saint]Powerful mobile phones and the advent of fast unlimited mobile data pretty much killed the DAP scene.
05:40:38musickflgjI would like to play high quality classical FLAC's on a device. I do not have a fantastic budget, so I was looking towards the Sansa models. My phone, I believe, doesn't have the same quality as my computer, and so I'd like something that can play them right. The rockbox, then, fits well, because I don't want filtered music. I want to hear it all.
05:41:08 Part Galois
05:41:09 Join Galois [0] (
05:41:34musickflgjThe gym thing, maybe, but I'm not going to listen to classical anyways while working out. Haha, no data, maybe I'm wrong, would particularly like flac.
05:41:51[Saint]flac is a really really really low overhead codec. It is likely that your soundcard is the part providing the terrible experience.
05:42:09[Saint]an external USB DAC would probably solve that.
05:42:43Galoisone problem I have with classical (notably opera) is that smartphones even today mess up gapless playback
05:42:46[Saint]and perhaps even the speakers themselves.
05:43:08[Saint]PC speakers are hardly precision monitors.
05:43:26[Saint](and many of them do their own high/low shelf filtering in hardware - annoyingly)
05:45:02[Saint]Galois: hate to be a pedant (ha! right...), but, wouldn't that be a problem with all media equally, regardless of genre?
05:45:43GaloisI think most of my pop music consists of individual songs amalgamated into albums where a gap isn't really noticeable
05:45:55musickflgjIt would be excusable to most people. They dont really care much about sound quality.
05:46:43 Quit krabador (Remote host closed the connection)
05:47:14[Saint]honestly, unless the source material has VERY high dynamic range - with modern LAME, a vast vast vast percentage of the population can't pick from flac8 or LAME 320CBR
05:47:47[Saint]Even someone trained has a pretty difficult time with this unless there is veru high dynamic range and with it, expected artifacts.
05:48:03Galoisfor me a good pair of in-ear headphones does more for sound quality than whatever device I use, except for REALLY low-end devices
05:48:38[Saint]Right. Most people are frankly just kidding themselves with lossless vs. lossy.
05:49:00[Saint]We have more than enough data to prove rather categorically that "golden ear" audiophiles don't exist.
05:49:23[Saint]They'll claim they do, though, very loudly.
05:49:25[Saint]And often.
05:49:32[Saint]To anyone even remotely interested.
05:49:53musickflgjI just like music. I also like having the best for as little as possible. But, I have no claims on anything fortunately.
05:50:43[Saint]I can discern between mp3 and lossless in some very specific cases, most obviously the case of tracks with extremely high dynamic range.
05:51:05[Saint]But I know what I'm listening for, and I need to have specific samples to do it from.
05:51:25[Saint]I fail ABX on unknown lossless/lossy content reliably.
05:51:27musickflgjBut I like the idea of FLAC itself. Perhaps I can hear nothing different. I do not know. I just know that it's harder to hear someone breathing in on an MP3 then in a FLAC.
05:51:52musickflgjBut perhaps that is a placebo?
05:52:00[Saint]almost certainly.
05:52:48Galoiseveryone can abx the hard stuff like fatboy with experience. Normal music is much harder.
05:52:54[Saint]Though there's certainly a chance as you specified classical genre, which often has a very wide sweep for dynamic range.
05:53:04 Quit michaelni (Read error: Connection reset by peer)
05:54:13[Saint]Modern music where everything has been condensed and packed suitably for the "loudness wars" contest popular music has going on, and...yeah, I would think reliably picking out encoding artifacts to be much much harder.
05:54:54[Saint]Though the current joke isn't lossy or lossless in the deluded audipphile world.
05:55:02[Saint]The current joke is "HD" audio.
05:55:24[Saint]96bit/192KHz "HD" audio.
05:55:37[Saint]Because apparently we're bats and here at sub and supersonic frequencies...
05:55:54[Saint](hint: we don't)
05:56:35[Saint]16/48 is plenty to faithfully reproduce the entire human audible range.
05:56:46[Saint]we can actually do it in a tiny bit over 15 bits.
05:57:08[Saint]stacking on more numbers there does absolutely nothing in consumer audio.
05:57:27[Saint]it is useful in recording and mastering, but not at the consumer level.
05:57:45Galoiswell, it lets the consumer do recording and mastering ... that's not nothing
05:58:01Galoisre-recording, re-mastering, whatever
05:58:10[Saint]Yes but it isn't marketed as such.
05:58:30[Saint]They very clearly market it in most circumstances as an improvement in SQ.
05:58:42[Saint]Which is just batshit.
05:59:09[Saint]There's no more audio in there. Lossless is lossless.
05:59:17[Saint]There's just a bunch of empty space.
05:59:34musickflgjHow is it useful in recording and mastering though?
05:59:43musickflgjIf it's just empty space I mean.
06:00:04Galoisthink of it as a photograph with too many megapixels for you to see at normal size
06:00:14Galoisthe extra pixels are still useful if you zoom in
06:01:20[Saint]well done, good analogy.
06:01:23Galoisso you know maybe you want to re-sample something at quarter speed. Suddenly your 192khz becomes 48khz. That's zooming in.
06:01:43[Saint]That's a damn good analogy, actually.
06:02:00[Saint]Couldn't have said it better myself.
06:02:11[Saint]I'll steal that for future arguments.
06:02:17Galoisyer welcome
06:04:20[Saint]There's actually a very good argument in favor of not using 48KHz samplerate, but honestly I don't think anyone really notices the largely imperceptible pitch shift between 44.1 and 48KHz.
06:04:48[Saint]People find it very difficult to detect very minor pitch shiting.
06:04:52[Saint]Arrr, shifting.
06:05:37[Saint]You can speed up content as much as ~8% and have people just totally not notice it unless they have a frame of reference.
06:05:56[Saint]So a .0* difference isn't going to ruffle any feathers.
06:06:34[Saint](Television speeds up or slows down shows to make up for overtime/undertime _all the time_)
06:07:04[Saint]If they're running a minute or so early, or late, they'll just increment the playback speed by a largely imperceptible amount.
06:08:07musickflgjI see. Good analogy btw @Galois
06:08:11[Saint]Speed up a 40 minute show for a 60 minute time slot by ~8%, and bam, that's three more minutes of commercials.
06:08:21[Saint]And no one notices.
06:09:08musickflgjIm not understanding that one.
06:09:27 Join michaelni [0] (
06:09:31musickflgjDo you mean that it can be played faster then the original recording and no one notices?
06:09:39[Saint]Yes. Absolutely.
06:10:07[Saint]Humans are only really good at picking up differences in pitch or tone when they have a direct comparison.
06:10:30Galoisin specialized contexts, people will notice. I used to do fansubs. I noticed even the slightest deviations from standard timing.
06:10:39Galoisnormally, people won't notice.
06:10:55[Saint]Yeah, granted.
06:11:25musickflgjI see. Kind of like how cartoon characters can move their lips and no one notices unless they look at it and try to match it with the word?
06:11:43[Saint]I kinda went on a tangent there, sorry - I went off the deep end as to why there's a reason some people won't use 48KHz samplerate, but how that reason is largely inconsequential for the masses.
06:12:21[Saint]musickflgj: that's usually a result of the original audio stream, being replaced with a translation.
06:12:32[Saint]But I guess you could say it was similar.
06:12:37musickflgjIt's fine. It's better then the place was before I started asking questions, "dead".
06:12:52musickflgjI see.
06:15:04Galoislike you record something off the air. Then you time it. Then you get the DVD release, and you think your timings from before will work, but no, it's off. And sometimes it's off by different amounts because they varied the amount and locations of nonstandard timing.
06:15:53Galoisin that context it was very easy for me to see even a one-frame difference, which isn't that hard (0.03 seconds), but still, I would never notice it if I weren't doing subtitling
06:18:28*[Saint] nods
06:44:37 Quit nlogex (Ping timeout: 250 seconds)
06:58:45musickflgjAgreed. Thank you fine people for the discussion, I will be on my way now.
06:58:50 Quit musickflgj (Quit: Page closed)
07:23:51***Saving seen data "./dancer.seen"
08:01:35 Join JdGordon_ [0] (~jonno@rockbox/developer/JdGordon)
08:04:43 Quit JdGordon (Ping timeout: 252 seconds)
08:38:08 Join edhelas [0] (
08:40:16 Quit pixelma (Remote host closed the connection)
08:40:16 Quit amiconn (Remote host closed the connection)
08:40:38 Join pixelma [0] (~pixelma@rockbox/staff/pixelma)
08:40:38 Join amiconn [0] (~amiconn@rockbox/developer/amiconn)
09:23:52***Saving seen data "./dancer.seen"
09:26:57 Join Tirifto [0] (~tirifto@
09:27:19 Quit Tirifto (Changing host)
09:27:19 Join Tirifto [0] (~tirifto@unaffiliated/tirifto)
09:35:47 Join NathanV [0] (68dc3297@gateway/web/freenode/ip.
09:50:47 Quit TheSeven (Ping timeout: 258 seconds)
09:51:04 Join [7] [0] (~quassel@rockbox/developer/TheSeven)
10:02:44 Quit Tirifto (Read error: Connection reset by peer)
10:05:36 Join ender` [0] (
10:06:03 Join Tirifto [0] (~tirifto@
10:28:04 Quit edhelas (Ping timeout: 260 seconds)
10:35:31 Join athidhep [0] (~afoakf@unaffiliated/athidhep)
11:23:54***Saving seen data "./dancer.seen"
11:39:39 Join edhelas [0] (
11:44:45 Quit edhelas (Quit: Leaving.)
12:23:59 Join pamaury [0] (~pamaury@rockbox/developer/pamaury)
12:58:25 Join einhirn [0] (
13:01:17 Join einhirn_ [0] (
13:02:15pamaury__builtin: re mutex and SWCODEC. Probably a relic from the past to reduce code size because mutexes were only used on SWCODEC platform
13:02:33 Quit einhirn (Ping timeout: 240 seconds)
13:02:59 Join Rower [0] (
13:03:02 Join einhirn [0] (
13:05:47 Quit einhirn_ (Ping timeout: 250 seconds)
13:05:56 Quit einhirn (Client Quit)
13:09:59 Quit robertd (Quit: Page closed)
13:12:24 Join robertd [0] (c9f2b970@gateway/web/freenode/ip.
13:13:10 Join einhirn [0] (
13:19:17 Quit einhirn (Quit: Miranda IM! Smaller, Faster, Easier.
13:19:28robertdhi, im trying to make a uboot image and receive this error
13:22:34pamauryrobertd: why do you want to make a uboot image ?
13:23:27pamauryI am not sure I would attempt to reflash uboot on the sony players, if you fail, you will brick the device
13:23:55***Saving seen data "./dancer.seen"
13:25:22robertd<pamaury> I was trying to draft from the sony sources. I wrote an update script and working on the kernel
13:25:40 Quit ender` (Quit: Sufficiently advanced stupidity is indistinguishable from malice.)
13:27:10 Join ender` [0] (
13:27:17robertd<pamaury> I also found an obscure device the PanasonicMV100 which uses the emma mobile. The update is strange it comes as a zip file
13:27:45pamaurythe thing you have to be aware of is that (assuming it works like on mine), if you break any of uboot or the kernel, you will brick the device, just saying ;) Because when the device boots, if loads uboot. Then uboot looks at the NVP, if it see the update flags, it loads the kernel with the initrd that runs the update script. Otherwise it runs the kernel normally
13:28:16pamauryrobertd: the upgrade method on the sony players is very specific to sony
13:28:53 Join ZincAlloy [0] (~Adium@2a02:8108:8b80:1700:d814:6ae0:2908:e867)
13:29:36robertd<pamaury> Yes, i wanted to know how the update process works
13:32:15pamauryit's very easy to brick a Sony player: if the regular kernel does not boot or usb does not work, you cannot change the NVP update flag and thus cannot upgrade => bricked. If you upgrade does not clear the NVP flag, the device will be stuck in an upgrade loop => brick. If uboot fails boot, there is no usb recovery mode => brick :(
13:32:49pamaury(unless you are willing to remove the emmc from the board, which is huge pain)
13:33:57pamauryideally you would want to flash a different uboot that has a recovery mode, but unless you are ready to unbrick, there is 0% chance you will get it right the first time I would say
13:34:22robertd<pamaury> I need to add the entry in the upgrade file. the kernel 2.6.23 is the source for a several devices
13:34:41pamauryrobertd: what do you mean "add the entry"
13:37:25robertd<pamaury> To give the instructions to the kernel to remove the upgrade flag
13:38:01pamauryrobertd: yes, like this: pamaury/1bd21f04388132ce2c13e6092f508b35">
13:38:28pamauryif I were you that you be the first instruction I run. Also check before-hand that you file passes syntax check
13:42:36robertd<pamaury> thank you. I am having a lot of trouble trying to dump the memory. So far I am obtaining info from the firmware extraction
13:43:51pamauryrobertd: what are you trying to dump ?
13:45:13robertd<pamaury> a part of the roofts
13:46:06pamauryyou need to execute code on the device for this
13:48:57pamauryor remove the emmc from the board
13:49:23robertd<pamaury> Exactly, i am trying to draft with that purpouse. For me removing the emmc from the board is beyond my capabilities
13:50:31pamauryon my device, the emmc is on a separate board that can be easily removed. Accessing it is a not exactly easy but I managed to unbrick (twice), so if you are afraid of bricking, I can run it on mine before. In the worst case, I can unbrick
13:51:28robertd<pamaury> which device do you own ?
13:54:21 Join paulk-collins [0] (
13:57:36robertd<pamaury> The e460 is from the same generation of the nwz a60 and the S760
13:58:38pamaurywhat is your device again ?
13:59:06pamauryall those devices are very similar anyway, the cpu is the same, one would need to check but even the kernel and uboot are all incredibly similar
13:59:07robertd<pamaury> the nwza867
14:00:39 Quit paulk-collins (Ping timeout: 244 seconds)
14:03:12pamauryrobertd: also my "quick" analysis of uboot suggest that it has a recovery console of some kind, but I never understood how to activate it, maybe there are pins on the board
14:04:21robertd<pamaury> I have been searching and reading a lot of info about those devices . Let me see if i can find something about the recovery console
14:04:36 Join paulk-collins [0] (
14:05:38pamaurywell technically it is a uboot console, but I just don't understand from the uboot code how it's activated. I have to admit that was some time ago that I looked at it, I don't remember the details
14:07:40robertd<pamaury> there is an old post from japan in which they ported android to the em1
14:09:28pamaurywhat kind of port are you aiming at ? native port or hosted port ?
14:12:00robertd<pamaury> I believe that the hosted port would be easier
14:12:28pamauryyes, then you don't need to touch either uboot or the kernel
14:14:40robertd<pamaury> Now I see why porting rockbox to new devices it is so difficult and takes a lot of effort.
14:19:15pamauryyes.... and at on hosted port you are not starting from scratch, but you have to play nice with the manufacturer's kernel/code. If I remember correcty, on the sony, all the UI and all is just a single executable that is run after the device boots. You could simply rename it, and replace it with another one that would be rockbox (and have the possibility to run the OF by pushing a button on boot)
14:22:49robertd<pamaury> That is just what I am aiming for.
14:23:57robertd<pamaury> There is another method that was used with the ereaders from sony. Crafting an upgrade as if it where an android device
14:25:40pamauryI am not sure that's a good idea in this case
14:26:18pamauryI doubt it would be able to run android, not enough RAM and also you would an android port for this specific device, which you don't have
14:27:48robertd<pamaury> I agree. In this case it is an old device
14:40:29 Quit athidhep (Quit: athidhep)
14:40:45pamauryrobertd: I have another suggestion to help bootstrap the whole thing. I would pack an upgrade with two file: first the shell script, second a liux executable that would want to test-run. The script would clear the upgrade flag, load a few modules, mount file systems if necessary and then run the second file
14:41:02pamaurythis way you can test an executable without risking (too much) to brick
14:41:42pamaurythe executable would run with the upgrade kernel/initrd so it's not exactly the same environement as the real thing but it let's you test and debug a few helpful thing
14:42:16pamaurymost notably, you can understand how keys work (there is a special icx_keys module on my device) and then design a bootloader that either exec the OF or RB based on keys
14:42:38pamauryI can try it on mine, see if I can go anywhere with this
14:42:48pamaurythe "hardest" part is to manage to compile the sony cross compiler :-p
14:42:57pamaury(which I haven't tried yet)
14:44:50pamauryideally I would like not to really on this cross compiler but rather have an equivalent crosstool config file that works, because otherwise toolchain rot and you end up with the yp-r0 situation
14:45:42robertd<pamaury> That is a great suggestion. I will use that aproach. Yes, the sony tool triplet is a nightmare they give you the gcc but not the croos binutils
14:46:14pamauryyeah, but I except their toolchain has nothing special
14:46:30pamauryalso it would be a good idea to document all of this on the wiki, unlike what I did :-/
14:48:31robertd<pamaury> Indeed, even the kernel requests for the arm-sony-linux cross compiler
14:49:14pamaurymaybe we should start working on the cross compiler, get it to work or get an equivalent cross tool config
14:51:29pamauryon the nwz-e463, sony provides a package named "sony-cross-gcc-for-dev-4.1.2-05000204.src"
14:51:53robertd<pamaury> I managed to find the sony cross binutils 2.23 but the poison patch included is giving me a hard time
14:53:16robertd<pamaury> it is the same for the A-series the cross gcc 4.1.2
14:56:06pamaurywhere did you find the binutils ?
14:56:15pamauryit's also annoying that they provide rpm package
14:56:48pamauryI had a look at the patches in the RPM for gcc 4.1.2, most (if not almost all) are simply backports of bugs
14:57:06pamaurywe can simply use a more recent gcc (like 4.9) to get rid of all of them I think
14:57:15robertd<pamaury> In the sony global distribution. It is under linux audio not the devices source code page
14:58:00pamauryhum, I can't see it, am I here:
14:59:30robertd<pamaury> my mistake sorry,
15:01:57robertd<pamaury> you are most welcome it is the only binutils available from the global source page
15:02:05pamauryrobertd: I am not familar with RPM, what "%{sony_vendor}" means ?
15:02:10pamauryoh wow
15:02:22pamauryis it an environment variable ?
15:03:42pamauryrobertd: the binutils patches also look like bugfixes, except maybe for the poison patch, which might be a sony thing, but I would need to check this
15:04:14pamauryalso they use things like %sony_cross_dev_setup
15:04:57robertd<pamaury> Yes The packages had to be converted
15:05:01robertdone sec
15:05:52 Join thum [0] (
15:07:11pamauryI might be wrong but my feeling is that any reasonable arm linux cross binutils and gcc will do, AS LONG AS we have the properly glibc/whatever-libc-they-use includes and linker file right
15:07:21pamaurythat means matching exactly the one on the device
15:13:02pamauryhonestly the whole thing look at a f**** nightmare ^^
15:13:33 Quit NathanV (Ping timeout: 250 seconds)
15:13:45robertd<pamaury> I had to use the alien package to convert the rpm files. Yes it is a mess of a process to do the crosscompiler
15:13:56pamaurywhat is "alien" ?
15:15:24robertdit is a program that converts rpm files to .deb
15:16:05pamauryoh, so you are trying to install it this way, that's not a bad idea actually, to start
15:16:59robertd<pamaury> thanks, they include the glibc 2.7 for the device
15:17:41pamauryyes, but glibc 2.7 seems to depend on another one, libopt
15:19:30robertd<pamaury> it is inside the same package sony-devel-libopt-build-libc
15:23:47robertd<pamaury> I changed the line for the em1-config which refers to arm-sony-linux-gcc for the cross compiler that I use. Some programs run but u-boot fails
15:23:59***Saving seen data "./dancer.seen"
15:24:15 Join timemob [0] (
15:25:40timemob<robertd> do you realize how freaking annoying your way of quoting the nickname is. it's absolutely infuriating because normally you do "pamaury: text here". FYI.
15:26:00timemobyour shit looks like you're pasting what pamaury says.
15:26:36robertdSorry I did not mean to be annoying. Wont happen again.
15:27:14pamaurytimemob: no need to be rude
15:27:16robertdpamaury the problem with the kernel from the souce code is that I am getting mixed implicit and normal rules: deprecated syntax warning
15:27:42timemobpamaury, I know. just watched it for the last few hours and it's super buggy :)
15:28:41pamauryrobertd: what is the build dependency order ? do you really need to build the kernel ?
15:28:53 Quit timemob (Client Quit)
15:31:02robertdpamaury Indeed, I was trying to compare which modules are included with the one I manage to extract from the update
15:31:24pamauryrobertd: the sony source contains almost nothing iirc
15:32:30pamauryall the interesting modules are out of tree I am afraid
15:33:47robertdpamaury then all comes to the using the suggested procedure. Thank you, you saved me from a lot of work
15:36:29pamauryrobertd: could you do something for me ? dump all the gcc and binutils config ?
15:36:47pamaurysomething like sony-arm-gcc -v
15:37:02pamauryand put on pastebin
15:37:24pamauryhum, I think there is more to dump, let me find the exact command
15:39:17pamauryalso gcc -dumpspecs
15:39:56pamaurythere must be an equivalent for the binutils but I can't find it
15:40:53robertdpamaury ok I am into it. let me give it a try
15:43:19pamaurybasically I'd like to come up with a crosstool config that gives the same result, I only want what is necessary to build an executable for linux, so probably gcc, binutils and maybe glibc (I don't know how it works w.r.t includes and linker files for cross compilers)
15:59:17robertdpamaury do I need to extract from this ? Sorry for the delay I am really slow with this
15:59:49pamaurynot sure what you mean
16:00:01pamauryare you trying to do the cross tool config ?
16:01:11robertdpamaury, yes It will take me a while... sorry
16:03:52pamauryno that's fine, take your time, but please do pastebin the gcc -v output, just so that it's somewhere and can serve as a reference
16:04:12 Quit ArneB (Ping timeout: 240 seconds)
16:04:42 Join ps-auxw [0] (
16:21:00__builtinpamaury: thanks!
16:22:05 Nick duo8 is now known as henkaku (~ZNC-SRV-H@
16:30:25 Join kugel_ [0] (
16:30:25 Quit kugel_ (Changing host)
16:30:25 Join kugel_ [0] (~kugel@rockbox/developer/kugel)
16:31:26 Quit kugel (Ping timeout: 276 seconds)
17:09:35pamaurywodz (logs): apparently I made a mistake, I can access TCSM0 by its virtual "physical address": 0x932b000c and 0xb32b000c, I don't know why it didn't work last time I tried
17:24:00***Saving seen data "./dancer.seen"
17:38:11 Quit krnlyng (Ping timeout: 244 seconds)
17:39:22pamaurythat means that I can use EBASE, also it means that I can keep the RAM at 0x80000000 and have a single wired entry to map for example 0x88000000-0x8fffffff uncached to 0xb0000000-0xb7ffffff which includes TCSM0, TCSM1 and SRAM :)
17:39:58pamaurymaybe not such a big range, a single entry 16MB is enough to cover all TCSM and SRAM
17:40:27pamauryactually we several entries I could map them so that they look like they are adjacent but I am not sure it's really worth it
17:50:45 Join krnlyng [0] (
18:45:16 Join zoktar [0] (
18:45:16 Quit zoktar (Changing host)
18:45:16 Join zoktar [0] (~zoktar@unaffiliated/zoktar)
18:55:30 Join Strife89 [0] (
19:12:32 Quit paulk-collins (Quit: Leaving)
19:24:01***Saving seen data "./dancer.seen"
19:31:59 Join krabador [0] (~krabador@unaffiliated/krabador)
19:59:10 Quit Tirifto (Quit: Tirifto)
20:01:05 Join Tirifto [0] (~tirifto@
20:01:45 Quit Tirifto (Changing host)
20:01:45 Join Tirifto [0] (~tirifto@unaffiliated/tirifto)
20:19:10 Quit gluytium (Quit: IRC (Quit: Not that there is anything wrong with that))
20:30:27 Join gluytium [0] (~g@
21:05:03 Quit shadows (Ping timeout: 258 seconds)
21:24:04***Saving seen data "./dancer.seen"
21:42:10 Join Jason__ [0] (4d638c8d@gateway/web/freenode/ip.
21:42:41Jason__I wonder if there is still a market for Rockbox with Android and iPhones and Spotify etc
21:43:48Jason__amyone home
21:44:41 Join wodz [0] (
21:44:42 Quit krabador (Remote host closed the connection)
21:44:51Jason__hi Wodz
21:45:03wodzJason__: hi
21:45:17Jason__Is there still a market for Rockbox?
21:45:26Jason__with the advent of music streaming from Spotify or Deezer?
21:45:37Jason__and everyone using their phones for music playback?
21:45:50wodzpamaury: It is really great gift from ingenic that they decided to support EBASE
21:46:17wodzJason__: Technically speaking there were never market for rockbox. It is not commercial product
21:46:29Jason__yea but you know what I am getting at
21:46:34Jason__Do people still use it?
21:47:19wodzJason__: I am not the right person to ask. For me rockbox is hobby. I like to tinker with hardware/software. Reverse engineering is pure fun for me :p-)
21:47:34Jason__I used to use Rockbox many years ago
21:47:59Jason__on an Archos Jukebox recorder, then an iRiver device, then an iPod
21:48:14Jason__but I now have an iPhone which does everything
21:48:46wodzJason__: If that service your needs, good for you
21:49:18Jason__It does
21:49:21gevaertsBut does it do anything *well*? :)
21:49:43Jason__I think Rockbox plugins spawned the ideas for 'apps'
21:50:44wodzactually plugins are probably the weakest part of rockbox. At least from the maintainer/dev point of view
21:51:06Jason__But they do provide stuff like games and whatnot
21:51:36gevaertsGiven that java applets on phones predate rockbox, I'd doubt that
21:52:11Jason__I do miss my Archos Recorder 10
21:52:15Jason__my first ever MP3 player
21:54:27Jason__It was built like a tank with its metal body and rubber corners
21:55:44gevaertsTo be fair, it had to be, otherwise it'd have collapsed under its own weight :)
21:56:51Jason__I think the standalone MP3 player is dying
21:56:56Jason__as everybody now uses phones
21:58:19gevaertsThat's definitely true
21:58:53gevaertsAlthough there seems to be a remaining niche for small, light and cheap players, which people seem to like to use while exercising
21:58:59Jason__Its extremely rare to find a device that supports line-in recording (whether recording to WAV or MP3)
21:59:04gevaertsBut that "cheap" bit usually means "no rockbox"
21:59:23Jason__I haven't bought a CD for the last few years as I have a Spotify account
22:36:25 Quit Jason__ (Ping timeout: 250 seconds)
22:37:16*pamaury has a weird bug with usb on jz4760B, again
22:38:39wodzpamaury: which one this time?
22:39:00pamauryI wanted to upload by test rockbox bootloader using hwstub_load
22:39:07pamauryand it fails with "unspecified usb error
22:39:41pamauryI pinned this down the size of transfers right now. If I transfer ~3900 bytes per request it works, but ~4000 fails
22:40:14pamauryI have no idea of what's going one currently, maybe it's hitting the time limit for control transfers, I need to fire the usb analyzer again
22:51:15pamaurywodz: I saw your comment on gerrit how coprocessor hazard
22:52:13pamauryI am award of the potential problem but on mips32r1 I don't what else I can do, the jump+nop gives two cycle, is there any other option ?
22:53:42pamauryalso the jz4760 manual says that mfc0 has a 3 interlock cycle dependency, so if you try to use v0, the hazard will be cleared by stalling
22:53:52pamaury(at least that's my understanding of interlock)
22:54:58pamaurybut yeah I see your point that some cores might have a different behaviour
22:57:20wodzpamaury: if jz is interlocked we are safe
22:58:12wodzpamaury: barrier instructions in r2 where invented just because of implementation differences. There is no standard way on mips to know how many nops/snops are safe
23:02:56wodzbtw. 4k core in atj is not interlocked
23:07:43 Quit pamaury (Ping timeout: 258 seconds)
23:08:03 Join pamaury [0] (~pamaury@rockbox/developer/pamaury)
23:08:30pamauryoh ok, I thought most recent mips were interlocked
23:08:58pamaurywell then I guess the code might need some fixing, perhaps we need a define in target-config.h that gives the mips revision, or the existence of barrier
23:12:09wodzpamaury: apparently I remembered wrong m4k IS interlocked by introducing 'slips' instead of stalls
23:12:23pamauryslips ?
23:13:37wodz see page 25 and later
23:16:04pamauryah, a slip is a bubble :)
23:20:55 Join krabador [0] (~krabador@unaffiliated/krabador)
23:23:02 Quit Tirifto (Quit: Tirifto)
23:24:05***Saving seen data "./dancer.seen"
23:28:55 Join Tirifto [0] (~tirifto@
23:29:35 Join Strife89|Quassel [0] (
23:29:45 Quit Tirifto (Changing host)
23:29:45 Join Tirifto [0] (~tirifto@unaffiliated/tirifto)
23:33:11 Quit Strife89 (Ping timeout: 240 seconds)
23:48:52[Saint]Bloody hell cron.
23:49:21[Saint]Y u no export user PATH explicitly when the user crontab is built?
23:49:28[Saint]I swear it used to...
23:50:17[Saint]I wondered what the hell was going on there for a second. Couldn't execute from crontab to ~/bin* without exporting the path specifically.

Previous day | Next day