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 2017-10-18

00:01:38_Bilguswhich ever MBOOT bl you end up using you'll want to use these rather than the one from the other day since I tested them
00:02:15_Bilgusalso if NO sd card is inserted on boot it displays ATA err -3
00:04:22 Join smbgaiden [0] (~smbgaiden@cpe-70-95-76-131.san.res.rr.com)
00:05:53 Quit pamaury (Ping timeout: 255 seconds)
00:08:49 Quit maffe (Quit: IRC ist obsolet!)
00:12:33 Quit _meg (Ping timeout: 240 seconds)
00:14:40 Join _meg [0] (~notsure@211.25.203.45)
00:46:23 Quit smbgaiden (Ping timeout: 255 seconds)
00:47:19 Quit ender` (Quit: My door mechanisms will be designed so that blasting the control panel on the outside seals the door and blasting the control panel on the inside opens the door, not vice versa. — Evil Overlord List #96)
01:00
01:00:44 Quit Aldem (Read error: Connection reset by peer)
01:04:11 Join Aldem [0] (~Aldem@unaffiliated/aldem)
01:07:37 Join smbgaiden [0] (~smbgaiden@cpe-70-95-76-131.san.res.rr.com)
01:33:09 Quit smbgaiden (Read error: Connection reset by peer)
01:33:33 Quit _meg (Ping timeout: 240 seconds)
01:33:57 Join _meg [0] (~notsure@211.25.203.45)
01:46:25 Join CrashBash-Kun [0] (~CrashBash@unaffiliated/crashbash-kun)
01:51:05***Saving seen data "./dancer.seen"
01:55:53 Quit petur (Quit: Leaving)
01:56:00 Quit Ruhan (Quit: Connection closed for inactivity)
02:00
02:02:23 Quit CrashBash-Kun (Quit: Leaving)
02:07:33 Join jhMikeS [0] (~jethead71@d192-24-173-177.try.wideopenwest.com)
02:10:32 Quit pystar89 (Ping timeout: 255 seconds)
02:31:37 Join Ruhan [0] (uid76353@gateway/web/irccloud.com/x-lhjucdztkiuwyrrw)
02:36:48 Quit Aldem (Read error: Connection reset by peer)
02:38:26 Quit __builtin (Ping timeout: 255 seconds)
02:39:22 Join __builtin [0] (~xray@rockbox/developer/builtin)
02:49:55 Quit Strife1989 (Ping timeout: 240 seconds)
02:50:36 Join Strife89 [0] (~quassel@adsl-98-80-187-177.mcn.bellsouth.net)
03:00
03:09:57 Join smbgaiden [0] (~smbgaiden@cpe-70-95-76-131.san.res.rr.com)
03:51:09***Saving seen data "./dancer.seen"
04:00
04:30:26 Quit smbgaiden (Ping timeout: 258 seconds)
04:31:01 Join smbgaiden [0] (~smbgaiden@cpe-70-95-76-131.san.res.rr.com)
04:59:46 Quit michaelni (Ping timeout: 252 seconds)
05:00
05:12:57 Join michaelni [0] (~michael@213-47-41-20.cable.dynamic.surfer.at)
05:24:48 Quit snw (Ping timeout: 246 seconds)
05:31:19 Quit Tony (Ping timeout: 260 seconds)
05:33:49 Join snw [0] (~snow_bcks@ganon.dot-server.net)
05:51:13***Saving seen data "./dancer.seen"
06:00
06:07:48 Quit [7] (Ping timeout: 258 seconds)
06:08:16 Join TheSeven [0] (~quassel@rockbox/developer/TheSeven)
06:20:27 Quit TheSeven (Ping timeout: 258 seconds)
06:20:56 Join TheSeven [0] (~quassel@rockbox/developer/TheSeven)
06:22:09 Join scorche [0] (~scorche@rockbox/administrator/scorche)
06:30:02 Quit dys (Ping timeout: 258 seconds)
06:39:35 Quit smbgaiden (Ping timeout: 240 seconds)
06:51:15 Quit Ruhan (Quit: Connection closed for inactivity)
07:00
07:04:46 Join johnb2 [0] (~johnb2@p5B3AFC38.dip0.t-ipconnect.de)
07:10:57johnb2_Bilgus: Thanks. I will use the latest BL bin file for dd. However, I feel I was not precise in my ask: I am pretty sure it is the FW that shuts down on low voltage, not the BL. So the MBOOT FW /w Recovery & NoVoltShutdown was what I wanted, i.e. if I get it to boot I get right into the recovery to dd the BL. I can try to build that myself tonight. Thanks for testing the bin files.
07:12:05 Quit johnb2 (Quit: Nettalk6 - www.ntalk.de)
07:47:27 Quit _meg (Ping timeout: 240 seconds)
07:48:46 Join _meg [0] (~notsure@211.25.203.45)
07:51:15***Saving seen data "./dancer.seen"
08:00
08:19:26 Join ender` [0] (krneki@foo.eternallybored.org)
09:00
09:00:05 Join wodz [0] (~wodz@iwl138.internetdsl.tpnet.pl)
09:24:27 Quit _meg (Ping timeout: 260 seconds)
09:27:16 Join _meg [0] (~notsure@211.25.203.45)
09:51:17***Saving seen data "./dancer.seen"
10:00
10:01:57 Join pamaury [0] (~pamaury@rockbox/developer/pamaury)
10:25:27 Join puck1pedia [0] (~puck@puckipedia.com)
10:25:33 Quit ArneB (Disconnected by services)
10:25:34 Join ps-auxw [0] (~arneb@p5B0FBD7D.dip0.t-ipconnect.de)
10:25:38 Join fs-bluebot_ [0] (~fs-bluebo@port-92-196-67-22.dynamic.qsc.de)
10:25:54 Join uwe__ [0] (~uwe_@dslb-088-066-247-118.088.066.pools.vodafone-ip.de)
10:26:01 Join Huntereb_ [0] (~Huntereb@d-209-42-136-23.cpe.metrocast.net)
10:26:14 Join petur [0] (~petur@213.224.21.58)
10:26:14 Quit petur (Changing host)
10:26:14 Join petur [0] (~petur@rockbox/developer/petur)
10:26:50 Quit _meg (Ping timeout: 255 seconds)
10:27:37 Join Piece_Maker [0] (~Acou_Bass@host-78-144-149-178.as13285.net)
10:28:35 Quit Huntereb (Ping timeout: 248 seconds)
10:28:35 Quit puckipedia (Ping timeout: 248 seconds)
10:29:04 Quit Acou_Bass (Ping timeout: 248 seconds)
10:29:04 Quit uwe_ (Ping timeout: 248 seconds)
10:29:04 Quit fs-bluebot (Ping timeout: 248 seconds)
10:29:04 Nick Piece_Maker is now known as Acou_Bass (~Acou_Bass@host-78-144-149-178.as13285.net)
10:29:20 Join TheLemonMan [0] (~lemonboy@irssi/staff/TheLemonMan)
10:29:35 Join MrZeus_ [0] (~MrZeus@81.144.218.162)
10:29:51 Join _meg [0] (~notsure@211.25.203.45)
10:30:53 Quit MrZeus (Read error: Connection reset by peer)
10:34:17 Quit _meg (Ping timeout: 246 seconds)
10:34:27 Quit pamaury (Ping timeout: 260 seconds)
10:39:26 Quit alucryd (Ping timeout: 255 seconds)
10:39:53 Quit Acou_Bass (Ping timeout: 246 seconds)
10:41:18 Join alucryd [0] (~quassel@archlinux/developer/alucryd)
10:42:29 Join _meg [0] (~notsure@211.25.203.45)
10:45:03 Join Acou_Bass [0] (~Acou_Bass@host-78-144-149-178.as13285.net)
11:00
11:37:02 Join pamaury [0] (~pamaury@rockbox/developer/pamaury)
11:51:21***Saving seen data "./dancer.seen"
11:57:45 Join JdGordon [0] (~jonno@rockbox/developer/JdGordon)
11:59:25 Quit JdGordon_ (Ping timeout: 252 seconds)
12:00
12:20:02 Nick puck1pedia is now known as puckipedia (~puck@puckipedia.com)
12:53:55 Join robertd1 [0] (~root@201.242.176.168)
12:56:34 Quit _meg (Ping timeout: 240 seconds)
12:59:59 Join _meg [0] (~notsure@211.25.203.45)
13:00
13:14:58_Bilgusjohnb2 I assume you mean the recovery fw with no low batt its in the same folder
13:35:09_Bilgusalso thinking about it you should be able to use regular firmware on that your sd card with that bootloader because I redirected all writes to internal back to the sdcard
13:38:15_Bilgussorry let me rephrase with those nointernal bootloaders you should be able to start any FW from your sd card you'll still need to edit the fw to block internal drives in them but it won't need rockbox_main.clip+ nor to be a multiboot FW
13:51:23***Saving seen data "./dancer.seen"
14:00
14:02:09pamaurydoes someone have the lsubs result of the ROM recovery? I'd like to see the difference between the two modes (30MB and the other)
14:08:31 Quit JdGordon (Ping timeout: 248 seconds)
14:15:30_Bilgusmaybe brb
14:18:08 Join JdGordon [0] (~jonno@rockbox/developer/JdGordon)
14:22:17_Bilgusnope can't find them
14:27:46_BilgusI can get you one of the other mode ~960 mb
14:36:23wodzC amazes me at times! http://www.lysator.liu.se/c/duffs-device.html This is the most mind blowing C construct I stepped in a year or so
14:41:31 Join Ruhan [0] (uid76353@gateway/web/irccloud.com/x-hlsxlhoaiqiwzupp)
14:41:49pamaurywodz: wow, I'm amazed this is valid C code
14:48:50wodzpamaury: I checked and compiler neatly unwinds this loop.
14:49:22wodzpamaury: Another use of such peculiar case placement is this https://www.chiark.greenend.org.uk/~sgtatham/coroutines.html
14:53:03 Join amayer [0] (~amayer@107-1-97-172-ip-static.hfc.comcastbusiness.net)
15:00
15:06:47_Bilgushmm I thought for sure when the device was started by shorting out the chip enable it came up as 960 mb but it shows the full drive
15:07:16_Bilgusif I leave it shorted it doesn't show any drive but lsusb still returns the same info
15:07:21_Bilgushttps://pastebin.com/31JFaVKp
15:08:30_Bilguscomes up as M200Plus[402] Bus 001 Device 036: ID 0781:6200 SanDisk Corp.
15:12:57 Quit EmanueleSorce[m] (Ping timeout: 264 seconds)
15:15:05 Join EmanueleSorce[m] [0] (emanueleso@gateway/shell/matrix.org/x-cfrvdnigiciagvot)
15:44:10 Quit wodz (Ping timeout: 255 seconds)
15:50:36 Join dys [0] (~dys@2003:5b:203b:100:6af7:28ff:fe06:801)
15:51:08 Quit LjL (Quit: You can't break me, for I have no spine)
15:51:27***Saving seen data "./dancer.seen"
15:54:04 Quit prof_wolfff (Ping timeout: 255 seconds)
15:54:36 Join LjL [0] (~ljl@unaffiliated/ljl)
16:00
16:16:41 Join CH23 [0] (3e48c15d@gateway/web/freenode/ip.62.72.193.93)
16:17:21CH23can it be that there's different recovery modes for the NAND and the SOC?
16:24:10_BilgusI don't think we can touch the NAND directly and pretty sure it is interfaced by the SOC since i see no supporting chips
16:25:11_BilgusM200Plus seems to be the recovery mode it is burnt into the chip
16:26:31_Bilgusthe ams docs mention a usb prommer but this has been mentioned before and the spec document has never been found
16:28:48pamaurythe recovery mode (M200Plus) is clearly the ROM
16:33:35CH23you mentioned finding different recovery modes in the firmware iirc
16:36:23pamauryyeah, I guess they correspond to the 900MB mode and the 30MB mode
16:36:48pamaurybut I can't tell at this point
16:40:34 Quit CH23 (Ping timeout: 260 seconds)
16:42:10_BilgusI was thinking when I forced my device in to recovery mode it came back as 900 MB but it just exposes the whole drive
16:42:40_BilgusAre you trying to look for the device identifiers of the 900 and 30 mb modes?
16:43:35pamauryI just wanted to know if there is a difference in the interface, that would slightly help reverse engineering of the ROM. Otherwise I'll make do without
16:55:16_Bilguson the fuze+ hbus being 200 decreased runtime by 2.3%
16:55:25_Bilgusover normal
16:56:02_Bilgusnext i'm running hbus 200 with autoslow disabled at cpu max
17:00
17:00:56 Quit Ruhan (Quit: Connection closed for inactivity)
17:05:45 Quit pamaury (Ping timeout: 264 seconds)
17:19:26 Quit jhMikeS (Ping timeout: 240 seconds)
17:32:43 Join CH23 [0] (56580a02@gateway/web/freenode/ip.86.88.10.2)
17:40:43 Join jhMikeS [0] (~jethead71@d192-24-173-177.try.wideopenwest.com)
17:41:05CH23i just tried to get into recovery mode, normally working sansa clip+, by connecting the 2 empty points next to the menu button
17:42:01CH23the way to do this is to short those 2 points, then connect to USB, or was it to power it on while shorting?
17:45:04 Quit jhMikeS (Ping timeout: 258 seconds)
17:45:20 Quit petur (Quit: Connection reset by beer)
17:46:29 Quit CH23 (Ping timeout: 260 seconds)
17:51:29***Saving seen data "./dancer.seen"
17:55:36 Join maffe [0] (~Miranda@p200300CD03E33500088044973D39F966.dip0.t-ipconnect.de)
17:57:15_Bilgusiirc its not the menu button I think its near the down button
17:58:07_Bilgusyeah just to the left of the down button
17:58:50_Bilgusunit is off to start you hold a flathead across the two points and finally plug in usb screen doesn't turn on but usb is active on PC
18:00
18:00:56_Bilgusif you just want to access the full drive though you can just use the 'recovery' firmware I put on the forum
18:05:34 Quit JdGordon (Ping timeout: 240 seconds)
18:07:15 Join smbgaiden [0] (~smbgaiden@cpe-70-95-76-131.san.res.rr.com)
18:18:52 Quit dys (Ping timeout: 255 seconds)
18:19:15 Join prof_wolfff [0] (~prof_wolf@120.red-83-54-195.dynamicip.rima-tde.net)
18:22:29 Join JdGordon [0] (~jonno@rockbox/developer/JdGordon)
18:58:12 Join pamaury [0] (~pamaury@rockbox/developer/pamaury)
19:00
19:07:32 Join almog1006 [0] (4d8bf186@gateway/web/freenode/ip.77.139.241.134)
19:18:33 Join lebellium [0] (~chatzilla@89-93-177-206.hfc.dyn.abo.bbox.fr)
19:23:12 Join petur [0] (~petur@rockbox/developer/petur)
19:36:01 Part robertd1
19:51:33***Saving seen data "./dancer.seen"
19:51:48 Join CH23 [0] (56580a02@gateway/web/freenode/ip.86.88.10.2)
19:53:47CH23_Bilgus: the icon on the down key looks like a 'menu' icon, that's what i meant :) do you still want a 8gb sansa clip+ image? it's not a clean image though, multibootloader is installed
19:54:57_Bilgusneed a clipzip 8GB image made on a windows machine, won't matter for a linux user as I discovered we can just copy the bootloader.img directly
19:55:43_BilgusI have a 4gb clip+ and 8 gb clip+ but just a bunch of 4gb clipzips
19:56:17_Bilgusalso I put the multibootloader in all of the images anyways because why not
19:58:22CH23sorry i only have clip+'s
19:58:42CH23also is the image different when made on windows?
19:59:15_Bilguson windows the only way to do it is to pull the whole drive image
19:59:43_Bilgusjust a limitation because of the only free software I found that can actulally image a raw drive
20:00
20:04:39 Join johnb2 [0] (~johnb2@p5B3AFC38.dip0.t-ipconnect.de)
20:12:20 Quit johnb2 (Read error: No route to host)
20:13:18 Quit _meg (Ping timeout: 248 seconds)
20:13:23CH23what's the difference between dd-ing /dev/sdX, or the windows equivalent? they should both be the full raw disk image, right?
20:15:44 Join _meg [0] (~notsure@211.25.203.45)
20:28:18_Bilgusdd has the ability to set a count of how much you want to pull
20:28:44_Bilgusif you set dd without a count then yes itd be the full disk image
20:29:38_Bilgusbut we don't rally need to worry about anything past the bootloader since that can be formatted after if need be
20:29:42_Bilgusreally*
20:34:19 Join advcomp2019_ [0] (~advcomp20@65-131-171-37.sxct.qwest.net)
20:34:39 Quit advcomp2019_ (Changing host)
20:34:39 Join advcomp2019_ [0] (~advcomp20@unaffiliated/advcomp2019)
20:37:26 Quit advcomp2019 (Ping timeout: 240 seconds)
20:37:41CH23knowing this could have saved me about 2 hours of my life
20:37:53CH23better late than never :P
20:39:04CH23is it possible to use a 4gb image on an 8gb nand?
20:41:13_BilgusI doubt it
20:42:25_Bilgusbut see we are only seeing the emulated part of the Nand everything you do is filtered through a controller that has bad blocks mapped out of the memory space and a bunch of who know s what going on
20:43:06_Bilgusthats why its doubtful even putting the same chip back will help
20:44:17_Bilgusand without information on how to access that controller there is no way to do anything with the actual NAND
20:45:48 Join advcomp2019__ [0] (~advcomp20@65-131-174-193.sxct.qwest.net)
20:45:48 Quit advcomp2019__ (Changing host)
20:45:48 Join advcomp2019__ [0] (~advcomp20@unaffiliated/advcomp2019)
20:46:27 Part almog1006
20:49:10 Quit advcomp2019_ (Ping timeout: 255 seconds)
20:51:04CH23so when we write an image to the NAND in recovery, it's not really directly written to the NAND?
20:54:34_Bilguswell sure it is its just interpreted by the controller and really its recovery in as much as you have access to the full drive to do with as you wish
20:55:36_Bilgusyour flash data sheet mentions how the bad blocks are handled by the controller on page 38 NOTE the controller is external to the NAND
20:58:13_Bilgus'The bad column address information is stored in the controller which maps each bad column address to a spare column address. There are a total of 1,024 extra bytes available per 9,192 bytes per plane. There can be up to 48 bytes of bad columns per plane at time of shipment. The remaining extra bytes can be used as EEC columns or as spare columns'
20:59:20_Bilgusoops thats 8192 sorry
20:59:31pamauryNAND flash never do bad block handling, that's always the responsability of the the layer above
20:59:58pamauryeither you put a flash aware file system or a FTL to get a block interface (like SD <-> flash)
21:00
21:01:29 Join johnb2 [0] (5b3afc38@gateway/web/freenode/ip.91.58.252.56)
21:04:20 Join JanC_ [0] (~janc@lugwv/member/JanC)
21:05:37 Quit JanC (Killed (orwell.freenode.net (Nickname regained by services)))
21:05:37 Nick JanC_ is now known as JanC (~janc@lugwv/member/JanC)
21:08:27_Bilgusand in our case AFAICT that Flash Translation Layer is in the main chip itsself mapped at a reserved memory location 0xC607-0000
21:09:02_Bilgusso a complete black box
21:10:44pamauryyes
21:11:43pamauryI mean not sure about that address but it looks like it's a black box on the SoC
21:12:25pamaurythis address is the SD interface
21:12:48pamauryit's more likely that the SD to flash block is a separate entity that is connected to this SD interface
21:14:00_Bilguseither way we only see the tail end of the whole shebang
21:14:49_Bilgusand I'm guessing the sd address was found by reverse engineering because the thing is never mentioned in any data sheet I can find
21:15:46 Quit lebellium (Read error: Connection reset by peer)
21:16:10 Join lebellium [0] (~chatzilla@89-93-177-206.hfc.dyn.abo.bbox.fr)
21:16:26pamauryyeah this is specific to the amsv2 for which we don't have the datasheet
21:30:39 Join jhMikeS [0] (~jethead71@d192-24-173-177.try.wideopenwest.com)
21:32:20johnb2jhMikeS: Is there anything particular one should test right now with G#1702 ?
21:32:22fs-bluebot_Gerrit review #1702 at http://gerrit.rockbox.org/r/1702 : Globally convert PCM to use frame counts instead of size by Michael Sevakis
21:33:42 Join ZincAlloy [0] (~Adium@ip1f12fa3f.dynamic.kabel-deutschland.de)
21:35:31_Bilgusjohnb2 how did your ATA fail clip+ do?
21:37:03 Join johnb3 [0] (~johnb2@p5B3AFC38.dip0.t-ipconnect.de)
21:37:45CH23i thought that rockbox was on a very good standing with Austria Micro Systems?
21:39:13 Quit johnb2 (Ping timeout: 260 seconds)
21:39:26pamauryafaik, they wanted to help and suddenly went radio silent
21:39:42CH23ah that's sad
21:39:47johnb3I compiled MBOOT FW + your EXPOSE gerrit patch + NoVOltageShutdown. Twice the clip got stuck on your init screen. Once I got it connected to USB, fdkisk showed the drive, but dd aborted after 6MB writing. However at this point, the screen showed the USB logo which I had not seen before, so I am not sure whether it was really in recovery mode.
21:40:30pamauryCH23: I might be wrong though, at least I think they ave us the datasheets of AS35{14,25,31,34}
21:40:53johnb3Does your gerrit task include "disabling of internal init"?
21:41:05pamauryoh it's on the wiki: https://www.rockbox.org/wiki/AustriaMicrosystems
21:41:37johnb3_Bilgus : what would I have to change to disable the internal init?
21:42:15CH23yeah i read that, that's why i'm half surprised we don't really know anything about the clip+ apart from what you reverse engineered
21:43:09_Bilgusjohnb just use the fw in the media fire folder it doesn't need to be multiboot fw to expose the drive
21:44:00_BilgusI disabled low voltage shutdown in one of them
21:44:14johnb3The problem is I don't get the player to boot if I don't use a MBOOT fw. And then I am not sure if I can ROLO your files.
21:44:17pamauryCH23: actually I didn't do the initial reverse engineering of the OF, just the ROM
21:44:39_Bilguswhy can't you rolo the files?
21:44:41pamaurymaybe because the amsv2 doesn't even have a datasheet? it seems like a monster chip made from the AS3525, remove this, add that
21:45:27_Bilgusas far as disabling the internal drive I'll push that to gerrit for you
21:45:54johnb31. boot from internal doesn't work. 2. If I use the regular MBooot fw, I get the voltage shut down and can never browse to your binaries.
21:46:30johnb3So only if I had your BL already installed I could move on.
21:46:36johnb3Thanks that would help.
21:47:25CH23pamaury: while we're at it, i have my clip+ in recovery, is there any more info you want from it that i could provide?
21:48:07pamauryCH23: not really, at that point I don't know what special commands it may have or how to interact with it
21:49:48johnb3_Bilgus : right now it is again in a state where there is no sign of life at all (warm and battery charged to a certain level, because I had it connected to usb). So I need it to run empty and do the fridge trick again ;-) which will only by tomorrow. But then I will try your gerrit update.
21:50:07CH23has anyone ever tried using a resistive value across those 2 pads used to get it into recovery?
21:50:37_Bilgusi'll just make a diff for you instead
21:50:52johnb3fine.
21:51:37***Saving seen data "./dancer.seen"
21:53:30jhMikeSjohnb3: first, make sure it compiles. next run playback and recording and check it. I've got AMS 1&2, imx31, s3c2440, Coldfire, pp502x, pp5002 (and SDL app/sim) on hand so can verify those myself.
21:59:42pamauryjhMikeS: I'll check imx233, but the code looks okay
21:59:53johnb3Checking your list, I only have additional players that run RB as an app: Samsung YP-R0, and 3 Sony Linux players.
22:00
22:00:12johnb3Would that help?
22:01:14johnb3Ah, on the sonys we don't have recording yet, do we?
22:01:19_BilgusJOHNB https://pastebin.com/8M2K1zvr
22:01:57johnb3That is against the lates on gerrit?
22:02:03johnb3*latest
22:02:37_BilgusI never pushed the NOINTERNAL stuff
22:03:03_Bilgusbut yes its the same way I disabled the internal drive in the Bootloader I posted
22:03:25johnb3I meant the pastebin is a diff against the latest gerrit?
22:04:08_Bilgusmight be a bit behind but nothing has changed there
22:04:24johnb3ok, thanks.
22:04:47_Bilgusnp
22:06:53jhMikeSpamaury: I think I tweaked that one the most of any
22:07:50pamauryjhMikeS: yeah, you added some irq guards around get_peak_buf and others
22:08:31jhMikeSpamaury: also, post-increment the buffers (as it's supposed to be)
22:10:03jhMikeSThere were extra calls to pcm_play/rec_lock/unlock that didn't appear required since the callback locked-out from above in pcm.c
22:10:28 Join p3tur [0] (~petur@78-23-23-252.access.telenet.be)
22:10:28 Quit p3tur (Changing host)
22:10:28 Join p3tur [0] (~petur@rockbox/developer/petur)
22:11:00jhMikeSpamaury: does PHYSICAL_ADDR give an uncached address as well?
22:11:06pamauryjhMikeS: speaking of post-increment, I don't really see the differece
22:11:10pamauryyes
22:11:50pamauryis it a problem to return an uncached address? it can be converted to cache if necessary, I don't know if the other drivers convert
22:12:25 Quit ender` (Ping timeout: 255 seconds)
22:12:51jhMikeSno, it should do that because scanning the record buffer while DMA is writing RAM can cause invalid data to be cached
22:13:29 Quit ender| (Ping timeout: 246 seconds)
22:13:42pamauryright that's a good point, and what about this post-increment thing?
22:14:07jhMikeSpost-increment is for accuracy in reporting samples still available from the source. though, perhaps I could use pre-increment and subtract the remaining size.
22:14:16 Quit johnb3 (Quit: Nettalk6 - www.ntalk.de)
22:15:26pamauryjhMikeS: my original code did not use dac_size for reporting
22:15:46pamaurybut now looking at the code, it appears pcm_get_bytes_waiting might have reported the wrong value
22:15:55jhMikeSI know, but it reported on the current tranfer, not on the state relative to the entire buffer
22:16:33pamauryah that's true, I forgot to change this when I modified my code to split the input buffer into smaller chunks
22:17:12pamaurybasically my driver was crap :-/
22:18:21 Join ender` [0] (krneki@foo.eternallybored.org)
22:18:55 Join ender| [0] (krneki@2a01:260:4094:1:42:42:42:42)
22:25:07jhMikeSpamaury: heh. anyway, imx233_dma_get_info() reads bar before ahb_bytes, which shouldn't change. it's better to underreport the count a little than to overshoot if the core registers change during readout.
22:32:13CH23pamaury: you need the AS3525v2 datasheet, correct?
22:33:28jhMikeSCH23: aren't we using the AS3543 one for that?
22:34:19pamauryCH23: no one has it, we only have AS3525 datasheet afaik
22:34:23CH23i'm not sure. i don't know much about NAND, or SOC, but i'm usually good at finding stuff
22:34:27jhMikeSI've got it
22:35:03pamauryjhMikeS: the datasheet of AS3525v2 ?
22:36:15jhMikeSthe as3543, which was the reference. there's even a new one (v1.11): https://ams.com/kor/content/download/125308/723179/file/AS3543_Datasheet_EN_v2.pdf
22:37:27pamaurybut it's just the codec part right? AS3543 is a codec not a soc
22:37:59jhMikeSyeah, that was the audio reference.
22:39:43CH23i did find an 'AS3525-A/-B C22O22' datasheet
22:40:38CH23is that the one you already have?
22:41:19jhMikeSI have a slew of as3525 ones
22:41:47jhMikeSalso, iirc there was one other funman brought up
22:46:25jhMikeSbah, so much datasheet. was it 3527?
22:47:09_Bilgusas3527 is very similar to the 3525
22:48:40jhMikeSone thing I never got an answer on was where info about the i2c came from (reversed or something else)
22:49:41pamauryjhMikeS: it looks like reversed
22:50:41jhMikeSI have some vague recollection of actually seeing a linux patch reference and then failing to save it :\
22:51:31pamauryjhMikeS: actually you are right, there is a linux patch, I may still have it somewhere, I don't remember how complete it was though
22:53:35jhMikeSIt no longer hits on Google. we're the only hit for anything concrete for that
22:54:43pamauryI have it
22:55:27funmanthere used to be a linux as3525 patch
22:55:50funmanis that what you're after?
22:57:45_Bilguslinux patch https://lists.denx.de/pipermail/u-boot/2006-December/018300.html
22:58:10funmanu-boot https://lists.denx.de/pipermail/u-boot/2006-December/018299.html
22:58:30funmanah next to each other :)
22:58:59_BilgusI had found it last week
23:00
23:05:27pamaurythere is something i never really thought about, is the nand flash wired to the nand block of the AS353x? In other words, could we use the nand interface to talk directly to the flash, bypassing the sd<->flash black box?
23:06:22_BilgusI tried it
23:06:43pamauryand?
23:07:36_BilgusI didn't get anywhere with it
23:07:37_Bilguslol
23:09:01 Quit lebellium (Quit: ChatZilla 0.9.93 [Firefox 56.0/20170926190823])
23:09:19pamaurythere would be some magic involved though, assuming it's even possible (which it should), the nand pins must be multiplexed between the nand interface and the sd to flash block, so some register controls that
23:10:33pamaurydo we even have the pinout of the AS3525v2? is it the same as the v1?
23:16:12 Quit TheLemonMan (Quit: "It's now safe to turn off your computer.")
23:18:31 Quit ZincAlloy (Quit: Leaving.)
23:24:28 Quit amayer (Quit: Leaving)
23:29:28 Join dys [0] (~dys@tmo-107-127.customers.d1-online.com)
23:36:50 Quit maffe (Quit: IRC ist obsolet!)
23:50:47 Quit ender| (Ping timeout: 246 seconds)
23:51:20 Quit ender` (Ping timeout: 264 seconds)
23:51:38***Saving seen data "./dancer.seen"
23:56:13 Quit p3tur (Quit: Leaving)

Previous day | Next day