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).

#rockbox log for 2017-10-30

00:15:33 Quit _meg (Ping timeout: 240 seconds)
00:18:10 Join _meg [0] (~notsure@
00:19:30 Quit Bilgus_PH (Quit: Page closed)
00:23:10 Quit ender` (Quit: True terror is to wake up one morning and discover that your high school class is running the country. — Kurt Vonnegut)
00:41:19 Quit pamaury (Ping timeout: 248 seconds)
00:57:16***Saving seen data "./dancer.seen"
01:06:03 Join jhMikeS [0] (
01:08:54 Join chrisb [0] (
01:27:46 Quit chrisb (Read error: Connection reset by peer)
02:08:35 Join JdGordon_ [0] (~jonno@rockbox/developer/JdGordon)
02:11:38 Quit JdGordon (Ping timeout: 246 seconds)
02:14:33 Quit _meg (Ping timeout: 240 seconds)
02:18:50 Join _meg [0] (~notsure@
02:29:39 Quit GeekShadow (Ping timeout: 240 seconds)
02:52:58 Join GeekShadow [0] (~antoine@reactos/tester/GeekShadow)
02:57:18***Saving seen data "./dancer.seen"
03:02:31 Quit Moarc (Ping timeout: 255 seconds)
03:04:35 Join Moarc [0] (
03:21:04 Join chrisb [0] (
04:25:07 Quit _meg (Ping timeout: 240 seconds)
04:28:00 Join _meg [0] (~notsure@
04:33:46 Join JdGordon [0] (~jonno@rockbox/developer/JdGordon)
04:37:05 Quit JdGordon_ (Ping timeout: 248 seconds)
04:57:21***Saving seen data "./dancer.seen"
05:20:00 Quit chrisb (Read error: Connection reset by peer)
05:32:00 Quit JdGordon (Ping timeout: 248 seconds)
05:33:46 Join JdGordon [0] (~jonno@rockbox/developer/JdGordon)
06:06:43 Quit TheSeven (Ping timeout: 258 seconds)
06:09:52 Join TheSeven [0] (~quassel@rockbox/developer/TheSeven)
06:18:47 Join PurlingNayuki [0] (~Thunderbi@
06:19:38 Quit PurlingNayuki (Remote host closed the connection)
06:22:36 Quit _meg (Ping timeout: 260 seconds)
06:24:37 Join _meg [0] (~notsure@
06:30:27 Quit jhMikeS (Ping timeout: 240 seconds)
06:47:05 Quit TheSeven (Ping timeout: 255 seconds)
06:47:30 Join TheSeven [0] (~quassel@rockbox/developer/TheSeven)
06:57:23***Saving seen data "./dancer.seen"
07:00:20 Join jhMikeS [0] (
07:04:48 Quit jhMikeS (Ping timeout: 248 seconds)
07:11:16 Join JdGordon_ [0] (~jonno@rockbox/developer/JdGordon)
07:13:52 Quit JdGordon (Ping timeout: 248 seconds)
07:25:37 Join PurlingNayuki [0] (~purlingna@
07:27:20 Join ychan [0] (~purlingna@
07:29:40 Quit PurlingNayuki (Ping timeout: 240 seconds)
07:32:55 Quit ychan (Remote host closed the connection)
07:33:22 Join PurlingNayuki [0] (~purlingna@
07:34:24 Quit PurlingNayuki (Remote host closed the connection)
07:34:50 Join PurlingNayuki [0] (~purlingna@
07:39:53 Join ychan [0] (~purlingna@
07:43:57 Quit Ruhan (Quit: Connection closed for inactivity)
07:44:51 Quit JdGordon_ (Ping timeout: 258 seconds)
07:49:10 Quit PurlingNayuki (Remote host closed the connection)
07:49:10 Quit ychan (Remote host closed the connection)
07:49:47 Join PurlingNayuki [0] (~Thunderbi@
07:50:37 Quit PurlingNayuki (Remote host closed the connection)
07:52:54 Join PurlingNayuki [0] (~purlingna@
07:54:04 Join JdGordon [0] (~jonno@rockbox/developer/JdGordon)
07:54:24 Join deevious [0] (~Thunderbi@
07:57:00 Quit pixelma (Quit: No Ping reply in 120 seconds.)
07:58:08 Join pixelma [0] (~pixelma@rockbox/staff/pixelma)
08:02:46 Join PurlingNayuki1 [0] (~Thunderbi@
08:03:36 Quit PurlingNayuki1 (Remote host closed the connection)
08:11:54 Join ychan [0] (~purlingna@
08:18:00 Quit PurlingNayuki (Remote host closed the connection)
08:18:00 Quit ychan (Remote host closed the connection)
08:18:59 Join PurlingNayuki [0] (~purlingna@
08:19:42 Join ychan [0] (~purlingna@
08:19:46 Join johnb4 [0] (
08:20:20 Quit ychan (Remote host closed the connection)
08:20:20 Quit PurlingNayuki (Remote host closed the connection)
08:20:46 Join PurlingNayuki [0] (~purlingna@
08:27:47 Join cfxxbkxqxt [0] (~hi@
08:27:55 Quit cfxxbkxqxt (Remote host closed the connection)
08:29:44 Join yicnriqfsa [0] (~hi@
08:29:52 Quit yicnriqfsa (Remote host closed the connection)
08:31:58 Join vchmlwrtve [0] (~hi@
08:31:58 Quit vchmlwrtve (Remote host closed the connection)
08:36:20 Join TheLemonMan [0] (~lemonboy@irssi/staff/TheLemonMan)
08:49:56 Join ender` [0] (
08:54:06 Join wodz [0] (
08:57:24***Saving seen data "./dancer.seen"
08:57:57 Quit johnb4 (Ping timeout: 264 seconds)
09:08:03 Join johnb4 [0] (
09:17:23Bilgusjohnb4 I moved the disk underclocking to g#1709 its still a WIP but I'll probably move the I2C underclocking there as well
09:17:25fs-bluebotGerrit review #1709 at : As3525 v1/v2 Add power savings menu by William Wilgus
09:19:40BilgusIt should work in its current form but it still needs some safety code
09:27:04 Quit TheLemonMan (Ping timeout: 240 seconds)
09:28:26 Quit PurlingNayuki (Remote host closed the connection)
09:28:51 Join PurlingNayuki [0] (~purlingna@
09:29:13 Quit PurlingNayuki (Remote host closed the connection)
09:29:43 Join PurlingNayuki [0] (~purlingna@
09:29:52 Quit PurlingNayuki (Remote host closed the connection)
09:30:16 Join PurlingNayuki [0] (~purlingna@
09:32:41 Quit johnb4 (Ping timeout: 255 seconds)
09:44:27 Join TheLemonMan [0] (~lemonboy@irssi/staff/TheLemonMan)
09:46:15 Quit PurlingNayuki (Remote host closed the connection)
09:46:43 Join PurlingNayuki [0] (~purlingna@
09:48:02 Quit PurlingNayuki (Remote host closed the connection)
09:48:34 Join PurlingNayuki [0] (~purlingna@
09:51:34 Quit TheLemonMan (Ping timeout: 240 seconds)
09:52:32 Join johnb3 [0] (
09:54:29 Join TheLemonMan [0] (~lemonboy@irssi/staff/TheLemonMan)
10:01:29 Quit PurlingNayuki (Remote host closed the connection)
10:04:32 Join PurlingNayuki [0] (~purlingna@
10:04:58 Quit PurlingNayuki (Remote host closed the connection)
10:05:22 Join PurlingNayuki [0] (~purlingna@
10:08:29 Quit PurlingNayuki (Remote host closed the connection)
10:08:56 Join PurlingNayuki [0] (~purlingna@
10:14:59 Join pamaury [0] (~pamaury@rockbox/developer/pamaury)
10:31:11 Quit johnb3 (Ping timeout: 248 seconds)
10:33:21 Quit pamaury (Ping timeout: 264 seconds)
10:40:30 Join johnb4 [0] (
10:57:27***Saving seen data "./dancer.seen"
11:11:00 Join dys [0] (~dys@2003:5b:203b:100:6af7:28ff:fe06:801)
11:26:48 Join pamaury [0] (~pamaury@rockbox/developer/pamaury)
11:30:14wodzpamaury: ping
11:32:10pamaurywodz: pong
11:36:57wodzpamaury: how far did you get in exploring usbip?
11:37:49 Quit kugel (Ping timeout: 252 seconds)
11:38:19wodzpamaury: I see fundamental problem that you need to know vid:pid before you export device. It is done either by setting in configFS (to create virtual gadget) or with usbipd in case of real hardware.
11:38:37pamaurywodz: I've never used usbip, I only ever used vhci
11:39:07 Join kugel [0] (
11:39:07 Quit kugel (Changing host)
11:39:07 Join kugel [0] (~kugel@rockbox/developer/kugel)
11:39:32pamauryis that a problem though? That sounds like a rather small detail if it's just the VID:PID
11:41:16wodzpamaury: I am not sure. From the scarce documentation I can't figure out how enumeration is actually performed.
11:42:45pamauryit was my impression that usbip is not very well documented
11:43:48pamaurywodz: some googling gives
11:50:05wodzpamaury: Not very verbose description what it does and how to use it :P
12:01:53wodzAnyway the interesting thing is enumeration process which seems to be skipped in this case (statically allocated descriptors).
12:03:21pamauryindeed, it looks like you have to emulate the enumeration level on the client side
12:04:50wodzthis is not something which cannot be done but it complicates the things considerably :/
12:06:09pamauryI am not sure it's so complicated, basically you need the client to first make a bunch of control request to get all descriptors, cache them, and then attach with usbip no?
12:09:38wodzbasically yes
12:57:31***Saving seen data "./dancer.seen"
13:09:45 Join Rower [0] (
13:15:28 Quit JanC (Ping timeout: 248 seconds)
13:18:40 Quit johnb4 (Ping timeout: 248 seconds)
13:41:43 Join amayer [0] (
13:46:04 Quit paulk-gagarine (Excess Flood)
13:46:20 Join paulk-gagarine [0] (
14:16:21 Join Ruhan [0] (uid76353@gateway/web/
14:23:33 Quit Moarc (Quit: i znowu NADMUCHAŁ BALONA)
14:23:38wodzpamaury: The link you found is the best source of usbip documentation I've seen
14:25:36wodzpamaury: Can you remind me what is the sequence of requestes during enumeration?
14:27:32 Join Moarc [0] (
14:29:45pamaurywodz: iirc, get dev desc, reset, set address, get config desc
14:29:46 Quit Moarc (Client Quit)
14:31:07pamauryoh and set config after that
14:31:56pamaurythe device is usually reset at the very beginning even before the first device request I think
14:33:30 Join jhMikeS [0] (
14:34:08wodzwhy double GET_DESCRIPTOR (DEVICE) ?
14:34:12pamaurythen SET_CONFIGURATION, but I'm not sure if that counts as part of the enumeration
14:34:54pamaurywodz: historical reasons mostly, usually the very first GET_DESCRIPTOR (DEVICE) will only ask for a small amount of data (ie not the whole descriptor), then reset, then ask the full descriptor
14:35:23pamauryWindows does it like that and thus most devices expect it and most OSes copy that to make sure device work. Usually windows hell
14:35:24wodzIsn't driver loaded based on enumeration responsible for SET_CONFIGURATION ?
14:36:19pamaurywodz: usually yes, although the number of devices that actually support several configuration must be close to 0
14:36:30pamaurybut technically yes
14:36:49 Join Moarc [0] (
14:38:17 Quit [Saint] (Remote host closed the connection)
14:38:57 Join dfkt [0] (~dfkt@unaffiliated/dfkt)
14:39:22 Join [Saint] [0] (~sinner@rockbox/staff/saint)
14:39:44pamauryyou probably can use a different enumeration, the minimum is: reset, SET_ADDRESS
14:40:37pamauryyou are allowed to read the device descriptor before SET_ADDRESS, and read the remaining descriptors before the SET_CONFIG
14:57:33***Saving seen data "./dancer.seen"
15:23:40 Quit JdGordon (Ping timeout: 255 seconds)
15:32:50 Join JdGordon [0] (~jonno@rockbox/developer/JdGordon)
15:33:05 Quit knittl (Ping timeout: 252 seconds)
15:33:54 Join knittl [0] (~knittl@unaffiliated/knittl)
15:51:44 Quit LjL (Quit: You can't break me, for I have no spine)
15:53:00 Quit wodz (Ping timeout: 240 seconds)
16:18:17 Quit JdGordon (Ping timeout: 246 seconds)
16:25:57 Quit Ruhan (Quit: Connection closed for inactivity)
16:27:48 Join JdGordon [0] (~jonno@rockbox/developer/JdGordon)
16:40:40 Join LjL [0] (~ljl@unaffiliated/ljl)
16:46:06 Quit JdGordon (Ping timeout: 248 seconds)
16:49:30 Join Ruhan [0] (uid76353@gateway/web/
16:55:16 Join JdGordon [0] (~jonno@rockbox/developer/JdGordon)
16:57:35***Saving seen data "./dancer.seen"
17:14:02 Join JdGordon_ [0] (~jonno@rockbox/developer/JdGordon)
17:15:30 Quit JdGordon (Ping timeout: 252 seconds)
17:39:20 Quit pamaury (Ping timeout: 240 seconds)
17:58:46 Quit TheLemonMan (Quit: "It's now safe to turn off your computer.")
18:04:42 Quit JdGordon_ (Ping timeout: 258 seconds)
18:19:09 Join xorly [0] (
18:19:25 Join JdGordon [0] (
18:19:40 Quit JdGordon (Changing host)
18:19:40 Join JdGordon [0] (~jonno@rockbox/developer/JdGordon)
18:23:32 Join pamaury [0] (~pamaury@rockbox/developer/pamaury)
18:29:18 Join johnb4 [0] (
18:35:54 Quit johnb4 (Ping timeout: 255 seconds)
18:39:26 Quit JdGordon (Ping timeout: 240 seconds)
18:47:31 Quit dys (Ping timeout: 255 seconds)
18:49:00 Join JdGordon [0] (~jonno@rockbox/developer/JdGordon)
18:52:19BilgusjhMikeS, when you get a chance could you look at g#1709 specifically the sd drivers line 972 & 966 I was wondering about the position of the mutex locks if it would work as I expect or do I need it after the wait for interface disable or if it would even work at all
18:52:21fs-bluebotGerrit review #1709 at : As3525 v1/v2 Add power savings menu by William Wilgus
18:57:40***Saving seen data "./dancer.seen"
19:14:56 Join johnb3 [0] (
19:22:19 Quit JdGordon (Ping timeout: 240 seconds)
19:26:17 Join johnb4 [0] (
19:26:41 Join lebellium [0] (
19:26:49 Quit lebellium (Client Quit)
19:31:26 Quit paulk-gagarine (Excess Flood)
19:31:26 Join lebellium [0] (
19:31:49 Join JdGordon [0] (~jonno@rockbox/developer/JdGordon)
19:38:33 Join paulk-gagarine [0] (
19:44:23 Quit johnb4 (Quit: Nettalk6 -
20:05:40 Quit johnb3 (Ping timeout: 240 seconds)
20:11:57jhMikeSBilgus: For v1, just lock the mutex. It appears that if it's not locked for a transfer, then controller isn't enabled.
20:15:41jhMikeSBilgus: I guess it's the same for v2. My only concern is sd_enable (on both) which could turn it on and then your code would get stuck. If you own the mutex, just disable it and restore the state.
20:16:16 Quit JdGordon (Ping timeout: 248 seconds)
20:16:37Bilgusexplain that a bit more stupidly for me please
20:17:40jhMikeSit turns the interface on and off for every transfer, yeah?
20:19:32Bilgusyes and in v2 at least it holds a mutex before the call to enable
20:19:50jhMikeSso, take the lock and if the interface is enabled, just disable it and restore the state to what it was. sd_enable can turn it on externally.
20:19:56jhMikeSso does v1
20:21:06jhMikeShow could the clocking thing help if it's really off most of the time anyway?
20:21:53BilgusI'm not sure it will on v2 as of yet on v1 though its never truly disabled
20:22:53 Join TheLemonMan [0] (~lemonboy@irssi/staff/TheLemonMan)
20:23:17Bilgusby setting MCI_POWERSAVE it turns off the clock when bus is inactive on the v1 devices
20:24:01jhMikeSI guess it doesn't do quite the same action in enable_controller(). shouldn't really matter
20:25:09BilgusI don't have a v1 device to test but johnb tells me its a 4hr runtime increase
20:25:14 Join dys [0] (
20:25:46 Join JdGordon [0] (~jonno@rockbox/developer/JdGordon)
20:26:04jhMikeSBilgus: I guess that's significant
20:27:01Bilgusyeah the patch he pulled from flyspray that shits all over the clocks nets him 8hrs but I think I can get close to that without destabilizing the whole system
20:27:41BilgusI just happen to have v2 devices to test so I figured I'd test some of the ideas there as well
20:29:24jhMikeSthe caller has to acquire the mutex to really guarantee things happen at the right time
20:29:31BilgusSo am I correct in assuming with the v1 code I can just hold the mutex and not wait for the enable==false and in v2 hold the mutex and set enable = false and set to its previous state when I'm done?
20:32:03jhMikeSsd_enabled being non-static is tweaking my interest. that should be nuked.
20:34:11Bilgusits defined extern in the debug menu
20:34:57Bilgusmakes sense I suppose when disabled the registers are all 0
20:34:58jhMikeSdebug should call a function
20:35:23jhMikeSso, really that variable doesn't do much
20:35:45jhMikeSI guess just shut the interface off and leave it off. if anything needs to happen it'll get turned on again.
20:36:29Bilgusafter holding the mutex..
20:38:38jhMikeSof course
20:38:51jhMikeSactually, the debug menu can conceivably mess things up
20:39:20jhMikeSthat's implemented completely wrong
20:39:32BilgusI'll work on fixing that next I need to do something similar with CPSR_0 as well
20:40:03BilgusThe I2c Registers only display for a fraction of a second when the interface is enabled
20:40:17jhMikeSor, maybe....(got other things on my mind) . I guess it can't touch it if it's transferring something.
20:41:01jhMikeSbetter to just have a debug info function that handles the details, really
20:41:26jhMikeSan internal one
20:41:34pamauryI think I finally manage to connect the usb stack code to the mmc code in the clip+ ROM \o/
20:45:10Bilgusok I'll work on the I2c one first since its the most broken and do something similar for the sd driver as well
20:46:14Bilguspamaury did you find anything interesting as far as the boot prommer?
20:46:31jhMikeSon further thought, the interface could end up off again by the time sd_enable returns because unlock a mutex can task switch
20:47:24pamauryBilgus: not yet, I still have many methods to reverse before I can tell what is going on
20:48:50Bilgusso I should still store the state and restore it then
20:50:11BilgusI really can't depend on it getting disabled once I hold the mutex
20:51:26 Join johnb4 [0] (
20:52:42pamauryit's kind of scary to think that after reversing over a hundred classes, I think I'm actually started to understand how the designer of the code thought
20:54:32Bilgushmm, in the v2 code sd_enable is never read so I think its just kinda hanging out for no reason
20:55:07johnb4Bilgus: I am still running the build before you did the latest changes (on my Fuze), i.e. MCI&SD&i2C. No problems found so far. I have the impression that the device takes a little longer to register on USB on windows. I will try recording next. Anything else I should test?
20:56:41Bilgusno not for the moment I have a few things I'm switching around And I would like to try lowering the IDE clock for your v1 device once I get it to a decent state
20:57:43***Saving seen data "./dancer.seen"
20:58:36Bilgusthough on your device when you go into the debug menu if you go down to the actual registers in HWinfo does I2C2_CSPR stay 0000000 most of the time and only flash values once every 30 seconds or so?
20:59:44johnb4saratoga has a Sansa E2xx if I am not mistaken. That one suffered from the FS patch. Maybe he is willing to try your stuff on his device.
21:00:03johnb4let me check
21:05:57 Quit Ruhan (Quit: Connection closed for inactivity)
21:06:11 Join ZincAlloy [0] (~Adium@2a02:8108:8b80:1700:6076:462f:3ccf:eb54)
21:06:17 Part ZincAlloy
21:06:52jhMikeSBilgus: agreed. the debug screen and grep confirm that
21:07:31BilgusI'll add it back for my use and see if it breaks anything
21:07:35johnb4Bilgus: from what I can tell it stays at zero all the time
21:08:33jhMikeSBilgus: I see no use for it whatsoever :)
21:09:34Bilgushow else would I know its status to restore it when I'm done?
21:10:13Bilguscheck CGU_SD_SLOT?
21:11:08Bilgusor are we back to don't worry about it when I get the mutex it is of no consequence to disable it?
21:11:27 Quit LjL (Quit: You can't break me, for I have no spine)
21:11:35Bilgusjohnb yeah the i2c controller needs to be enabled to get the value of the registers the debug menu misses that
21:16:19jhMikeSBilgus: If you make the debug info stuff right, then just disable it and leave it disabled.
21:16:49Bilgusok will do
21:19:35pamauryurg, the sd/mmc code on that thing is just horrible :(
21:21:06jhMikeSBilgus: I just did a patch just now :)
21:21:58BilgusIs it already pushed?
21:25:57jhMikeSnot quite, checking that it compiles
21:26:31Bilgusah ok thanks :)
21:30:29 Quit johnb4 (Quit: Nettalk6 -
21:31:28 Quit prof_wolfff (Ping timeout: 248 seconds)
21:35:13jhMikeSI only compiled it but it's (maybe, probably?) fine
21:36:38BilgusI'll test it when I get the 12C code worked out
21:38:34 Join LjL [0] (~ljl@unaffiliated/ljl)
21:39:00Bilguslooks better though
21:48:45 Join smoke_fumus [0] (~smoke_fum@
22:08:46 Join Ruhan [0] (uid76353@gateway/web/
22:15:40 Quit TheLemonMan (Quit: "It's now safe to turn off your computer.")
22:17:52 Quit amayer (Quit: Leaving)
22:39:20 Join zajebani [0] (6d5c41bb@gateway/web/freenode/ip.
22:40:44 Part zajebani
22:45:35 Quit xorly (Ping timeout: 248 seconds)
22:49:01 Quit lebellium (Quit: ChatZilla 0.9.93 [Firefox 56.0.2/20171024165158])
22:57:45***Saving seen data "./dancer.seen"
23:02:49 Quit ender` (Quit: Please help Conserve Gravity - Play Chess, not Basketball.)
23:03:59 Quit JdGordon (Ping timeout: 240 seconds)
23:20:12 Quit minus (Quit: Bye)
23:20:15 Join prof_wolfff [0] (
23:25:58 Join JdGordon [0] (~jonno@rockbox/developer/JdGordon)
23:29:23 Join minus [0] (
23:35:11 Quit JdGordon (Ping timeout: 248 seconds)
23:41:30 Quit anonus (Remote host closed the connection)
23:43:12 Join anonus [0] (
23:44:20 Join JdGordon [0] (~jonno@rockbox/developer/JdGordon)

Previous day | Next day