--- Log for 13.02.117 Server: adams.freenode.net Channel: #rockbox --- Nick: logbot Version: Dancer V4.16 Started: 9 days and 5 hours ago 00.01.14 # ohh builtin is there a way to run donkey kong mame on rockbox? 00.01.23 # kfogel: most ereaders are based on linux 00.01.34 # jailbreaking seems like a better solution 00.01.43 # and there are plenty of jailbreaks for ereaders 00.03.21 # pamaury: *nod* A jailbreak isn't quite the same as running an top-to-bottom free OS, but it's a place to start anyway. Thanks. 00.04.54 # kfogel: if the devices run a linux, for which you have the source and that you can jailbreak, I don't see what non-free. You can just remove all the non-free part. Of course you still have the problem of replacing all the binary modules but that's a huge work anyway, independently of the approach 00.06.17 # pamaury: Yes. I suspect that just such a binary module is what drives the screens, so the device would be pretty useless without that work being done. I'll look into it some more; I don't have time to do much of the hacking myself, unfortuately. 00.07.28 # <__builtin> ComputingComet: not as far as I know 00.08.02 # <__builtin> we don't have a MAME port, first of all 00.08.48 # <__builtin> though I suppose I could port SDLMAME if it weren't for the licensing issue 00.09.49 Part kfogel ("Ex-Chat") 00.11.05 # <__builtin> wait a second... 00.11.09 # <__builtin> they recently relicensed 00.12.29 # so its possible? 00.13.02 # <__builtin> no, not yet at least 00.13.57 # <__builtin> it's all GPLv2 or later now 00.14.35 # <__builtin> it might be possible though 00.16.36 # <__builtin> hmm, it's all C++ though 00.19.08 # how does the packbox/pacman work 00.19.09 # then 00.19.13 # pacbox* 00.19.34 # <__builtin> pacbox is specialized to emulate only a pacman machine 00.21.18 # and its been done for dk on ipodlinux put it so old i dont use it on my ipod 00.21.46 # <__builtin> ipodlinux has nothing to do with rockbox 00.21.56 # ik 00.22.02 # im just saying 00.22.56 # <__builtin> ComputingComet: what is it that you want to do again? 00.23.03 # <__builtin> play Donkey Kong or run MAME? 00.23.15 # Donkey Kong 00.26.13 # <__builtin> well, you can probably find a Gameboy version of it and dump the ROM yourself ;) 00.30.15 Quit skapazzo (Quit: leaving) 00.33.28 # i tried it lags to all hell 00.35.32 # <__builtin> what device do you have? 00.35.47 # ipod 5 00.35.51 # ipod video 00.36.35 # and and a fuze 00.36.36 # <__builtin> that's one of the slowest devices we support, at least in terms of # of cores*CPU frequency/screen size 00.36.40 # <__builtin> (ipodvideo) 00.37.06 # <__builtin> you'll have much more luck with the fuze, I'd wager 00.38.53 Quit bertrik (Quit: leaving) 00.47.08 Quit pamaury (Ping timeout: 264 seconds) 00.50.46 # <__builtin> argh great 00.50.58 # <__builtin> now xworld freezes up randomly 00.52.42 # <__builtin> and of course it only happens on target, so I can't debug without scattering a million splash()es everywhere :( 00.56.23 # whats xworld 00.57.08 # <__builtin> it's a plugin that emulates the game "Another World" 00.57.40 # ohh dont have it 00.58.03 # <__builtin> what version of rockbox are you running? 00.59.05 # newest 00.59.18 Quit xorly (Ping timeout: 255 seconds) 00.59.56 # <__builtin> sorry, did you mean that you don't have the 'xworld' plugin, or the game 'Another World'? 01.00.05 # xworld plugin 01.00.25 # <__builtin> hmm, that doesn't make sense then 01.00.34 # <__builtin> are you running the release version? 01.00.41 # yes 01.00.55 # <__builtin> that explains it... our last "release" is horribly out of date 01.01.00 # XD 01.01.02 # really 01.01.09 # i did it cuz stable 01.01.12 *** Saving seen data "./dancer.seen" 01.02.05 # <__builtin> there's no reason for you not to run the development builds 01.02.20 Quit ender` (Quit: History is a set of lies agreed upon. — Napoleon Bonaparte) 01.02.24 # <__builtin> heck, they're probably more "stable" than the release 01.02.49 # <__builtin> there's a bazillion bug fixes and new features that you're missing out on, too 01.03.13 # <__builtin> and like 30 new game plugins 01.03.42 # jesus im installing now 01.03.46 Join pamaury [0] (~pamaury@rockbox/developer/pamaury) 01.04.01 # and installed 01.04.33 # so many games 01.04.35 # so many 01.05.01 # * __builtin apologizes 01.05.10 # now pacbox does not work incompatible version 01.05.17 # <__builtin> you need to reboot 01.05.25 # i did 01.06.09 # ill do it again 01.07.02 # <__builtin> are you sure? that shouldn't happen if you did 01.07.32 # ok it works now i cuz i fucked up reboot somehow 01.08.39 # <__builtin> most of the new games are my fault 01.09.23 # fault? 01.09.50 # <__builtin> actually, all of them are, I think 01.10.10 # so what dos files do i need for xworld 01.10.56 # <__builtin> would you mind helping me test if this bug I'm seeing is reproducible? 01.11.04 # sure 01.12.31 # what do i do i have the dos files for another world and out of this world 01.13.10 # <__builtin> extract them to /.rockbox/xworld/ 01.13.21 # <__builtin> create that folder if it doesn't exist 01.13.29 # all of them even exe 01.13.42 # <__builtin> you only need the MEMLIST.BIN and BANK01-BANK0D 01.14.37 # do i use out of this world or another world files 01.14.50 # <__builtin> either one will work 01.15.02 # <__builtin> as long as it has those files 01.15.17 # ok now about this error what do i need to do 01.15.28 # <__builtin> what error? 01.15.37 # i mean bug 01.15.39 # what bug 01.15.55 # and i got a string error 01.15.55 # <__builtin> start xworld and choose "Start New Game" 01.16.04 # <__builtin> that's just a warning, don't worry about it 01.16.38 # when i clicked start new game i get these symbols and some text 01.18.00 # i did the codes now its starting 01.18.17 # i got to title screen 01.18.22 # and now it breaks 01.18.28 # *PANIC* 01.18.40 # <__builtin> tell me what else it says 01.19.26 # incorrect CPI mode in mutex_lock (0x12!=0x1f) 01.19.54 # pc:00074464 sp:000007b0 01.19.59 # <__builtin> ohh...... 01.20.03 # <__builtin> that explains a lot 01.20.15 # <__builtin> what device is this? 01.20.22 # a: 03f83318 01.20.34 # ipod 01.21.23 # a: 03f82914 01.21.38 # a: 03f929dc 01.22.00 # a: 03f83188 01.22.01 # bt end 01.22.04 # thats all it says 01.22.11 # <__builtin> ok, thanks 01.22.15 # im gonna reboot now 01.22.42 # is there a way to make it work i wanted to platy 01.22.43 # play* 01.23.16 # <__builtin> I think I know what the problem is 01.23.58 # <__builtin> a recent commit added CPU mode checks to the kernel (3e73866) 01.25.43 # <__builtin> jhMikeS (logs): see above ^ 01.25.57 # ? 01.26.12 # <__builtin> jhMikeS is the author of that commit 01.26.18 Quit pamaury (Ping timeout: 255 seconds) 01.28.01 # <__builtin> that commit doesn't really "cause" the crash you're seeing, though 01.28.26 # <__builtin> but it just puts in sanity checks for the real issue 01.28.53 # whats all this sgt stuff 01.29.24 # <__builtin> the initials of the author of that game collection 01.29.41 # <__builtin> (S.G. Tatham) 01.29.52 # <__builtin> see http://www.chiark.greenend.org.uk/~sgtatham/puzzles/ 01.30.01 # i see this game has one level 01.30.16 # ohh its random 01.30.17 # k 01.30.19 # im stupid 01.31.29 # <__builtin> ok, just checked: that commit does indeed break xworld 01.32.06 # when does it get into dev patch 01.32.13 # <__builtin> hmm? 01.32.29 # when will i be able to install it 2 play 01.32.52 # <__builtin> I can build you a version right that should work 01.33.02 # that would be amazing 01.33.23 # <__builtin> keep in mind that it's going to be slow as heck on the video 01.34.12 # ik 01.34.23 # when i get my fuze back ill be on that 01.35.36 # <__builtin> I actually don't have to build it... http://download.rockbox.org/daily/ipodvideo/rockbox-ipodvideo-20170120.zip should be what you need 01.36.31 # and just take the xworld.rock 01.36.40 # <__builtin> no, you need to install the whole zip 01.36.44 # idk what it is why 01.36.58 # nvm what i said 01.37.00 # i mean why 01.37.10 # <__builtin> .rocks are closely tied to the core 01.37.24 # <__builtin> they won't work if there's a version mismatch 01.37.58 # ok rebooting 01.37.59 # and ill brb 01.38.25 # idk what it is about this i have a phone i can do all this theres just something about doing this on a mp3 player thats fun not being connected to everything 01.41.45 # <__builtin> by the way, the real gems of sgt-puzzles, in my opinion at least, are Cube, Net, Map, and Untangle 01.42.57 Quit MrZeus1 (Ping timeout: 255 seconds) 01.43.00 Join Bilgus_ph [0] (4cf32773@gateway/web/freenode/ip.76.243.39.115) 01.43.02 # it runs good actually im after the cutscene and its fine 01.43.16 # <__builtin> hmm, that's good 01.43.59 # nvm 01.44.28 # johnb2: yes it is ready, though only a firmware with boot data will boot from the sd card and for now you will still need to disable the rooot with Mihails second patch ftm 01.44.30 # now it just get to the code ldkd and it just keeps doing the same thing over and over again 01.45.15 # <__builtin> you need to swim upwards 01.45.18 # i am 01.45.32 # atleast im trying 01.45.52 # <__builtin> this game isn't supposed to be easy ;) 01.46.17 # do i scroll or click up im trying both 01.46.37 # <__builtin> hold MENU 01.46.53 # still just float down 01.47.25 # whats the max frame rate? 01.47.58 Join _mt_ [0] (~MT@2601:482:4402:7b60:4e2:fd42:809:89e5) 01.48.05 # <__builtin> it varies, usually around 12 or so 01.48.19 # yeah im just a solid 12 01.48.25 # have not gone to 11 01.51.54 # jesus to many bug guys 01.52.19 # <__builtin> enjoy! :P 01.53.18 Quit cc___ (Ping timeout: 255 seconds) 01.53.49 # since release have you guys updated imgviewer 01.57.31 # nvm a 01.57.38 # and why does the mpegplayer not work 01.59.45 # last I knew it did 02.00.09 # i tried in release and it did not work it just stayed black forever 02.00.20 # did you rename your old install and copy the new one to your device? 02.00.50 # ? 02.01.16 # no i deleted the old install 02.01.39 # or rename .rockbox to .rockbox.old then copy .rockbox from the zip file to device 02.01.50 # oh ok same difference 02.02.43 # every update ive tried on the release on the dev and on the dev you gave me when i try and play it i keep black screen 02.02.47 # confused you tried it in release or dev? 02.04.14 # I'm not builtin but idk either way you said It didn't work in your original install either? 02.05.57 # sorry was not paying attention to names ive tried it on every version dev and release im currently on and install builtin gave me that runs xworld properly 02.07.34 # i'm not sure of the particualrs of mpeg player i'd have a look at the manual it probably has limited file support (or format support) 02.09.42 Join jhMikeS [0] (~jethead71@d192-24-173-177.try.wideopenwest.com) 02.11.31 # jhMikeS I have a patch up to remove some unused fuctionality from bootloaders (O_CREAT) specifically I'm not particularly fond of the implementation atm any ideas to make it cleaner? 02.16.50 # beep boop 02.17.04 # Bilgus_ph: it's on gerrit? 02.17.42 # yea #g1560 02.17.44 # 3Gerrit review #1560 at http://gerrit.rockbox.org/r/1560 : 3Remove unneeded fuctionality from bootloaders by William Wilgus 02.18.03 # ill be back to my other acct in a bit 02.18.07 # Bilgus_ph: there is at least one bootloader that writes new files 02.19.17 # oh yea? maybe I should just focus on bootloader that won't compile anymore like the clipV1 02.19.26 Quit TorC (Ping timeout: 240 seconds) 02.19.28 Quit __builtin (Ping timeout: 245 seconds) 02.19.56 Join __builtin [0] (~xray@rockbox/developer/builtin) 02.20.34 # gigabeat_s can extract tar files (it's functionality that comes in handy upon first installs sometimes) 02.21.38 # I kinda wish to separate the functions to make it easier for the unused functions to be stripped I think, Without removing write It knocks off like 1k so might be enough 02.21.44 # the truth is though, there isn't a huge amount of code in general dedicated to creating a new file 02.22.09 # i think it was 2.5k compiled 02.22.22 # <__builtin> jhMikeS: I take it that certain blocking kernel functions shouldn't be called in an IRQ context, right? 02.23.16 Join TorC [0] (~TorC@fsf/member/TorC) 02.23.17 # you must've taken out way more than what I see on gerrit 02.23.26 # nope 02.23.44 # __builtin: never. you can poll semaphores and queues with TIMEOUT_NOBLOCK though 02.24.30 # <__builtin> ok, I'll fix xworld then 02.24.42 # __builtin: if it blocks, no ISR because a thread is the only thing that blocks 02.25.17 # fat_time and is_diacritic hits 1kb and ocreate calls create_stream_internal( which is another 2.5k 02.26.52 # thats after compression btw 02.27.19 # bbiab 02.27.24 Quit Bilgus_ph (Quit: Page closed) 02.30.12 # this game is bs 02.31.09 Quit amiconn (Quit: No Ping reply in 64 seconds.) 02.31.29 # __builtin: also fiq on arm is totally forbidden from interacting with any thread stuff. audio callbacks should never call anything thread related 02.32.18 Join amiconn [0] (~amiconn@rockbox/developer/amiconn) 02.32.32 # Bilgus_ph: create_stream_internal is 2.5K? wtf 02.34.11 # * jhMikeS checked an elf for fuze v2 and it's 204 bytes 02.36.50 # Build Server message: 3New build round started. Revision 0a9f717, 255 builds, 17 clients. 02.37.06 # <__builtin> alright, this last commit *should* fix xworld 02.37.42 # i thought you fixed it already 02.38.46 # <__builtin> well, that wasn't a "fix" that I gave, just an earlier version without the sanity checks 02.38.49 # Bilgus: I guess that + fat_create_file is 772 bytes 02.38.58 # <__builtin> so in a sense it was still broken, but it just didn't crash because of it 02.39.16 # so do i update 02.39.27 # Bilgus: I'd suggest reducing the number of cache buffers is you want low-hanging fruit for a bootloader 02.39.49 # <__builtin> ComputingComet: there won't be any visible benefit 02.42.25 # its having draw errors and i did not even update 02.44.38 # i just started new game its fine 02.46.12 # Build Server message: 3Build round completed after 562 seconds. 02.46.13 # Build Server message: 3Revision 0a9f717 result: 2 errors 0 warnings 02.48.35 # I'd assume they would compress pretty well though wouldn't they? 02.48.48 # @jhMikes 02.51.10 # on clip+ with the patch on gerrit [INFO] Bootloader size: 77508 bytes and with only the conditional on create file removed [INFO] Bootloader size: 81844 bytes 02.51.21 # (prior to compression) 02.51.24 Join PurlingNayuki [0] (~Thunderbi@218.18.200.2) 02.52.25 # packed: [INFO] Packed bootloader size: 43153 bytes vs [INFO] Packed bootloader size: 45721 bytes 02.52.50 # <__builtin> ComputingComet: what do you mean by "draw errors"? 02.52.53 # yes i got a laser gim 02.52.55 # gun* 02.53.13 # is there something else funky going on with AMS? the clipV1 is even more 02.53.30 # <__builtin> amiconn: http://build.rockbox.org/shownewlog.cgi?rev=0a9f717;type=sansac200v2sim 02.55.52 Quit PurlingNayuki (Ping timeout: 260 seconds) 02.56.58 # Bilgus: it's just the image size that matters? not the bss section too? 02.58.52 # Well I assume we have the whole ram minus the dual-boot function & UCL unpack function 02.58.52 # s which is like 300 bytes so no its just the image size that is the issue 02.58.59 # btw, they'll compress to 0 because bss isn't actually part of the image itself 02.59.12 # figure we compress the OF and what is left is the room we have for RB 03.01.16 *** Saving seen data "./dancer.seen" 03.02.36 # figure its not just Create file its, also create_stream_internal and what ever ancillary functions those two call as well 03.03.07 # perhaps just direct the bootloader config to not define O_CREAT in fcntl.h and then it would error on any attempt to use it 03.03.31 # I'm thinking maybe a separate file handling implementation for boot loader would be the cleanest 03.04.30 # sure that would be inline with what I have atm undef O_CREATE but as you were saying there are targets that do create files 03.04.34 # nontrivial and a pita to keep maintained in step 03.04.47 # Ikr lol 03.05.00 # Bilgus: keep it in the target's config. most aren't that tight. some don't matter at all 03.05.50 # ok so do something like NO_O_CREAT_BOOTLOADER 03.05.55 # if some other thing is defined, then block #define O_CREAT 03.06.40 # just blocking O_CREAT won't do it ill have to still ifdef out the actual routine but sounds good 03.07.21 # or maybe FILESYS_CONFIG_LEAN ? block rename(), all create and write functionality, and some other functions that are likely unneeded 03.07.54 # oh, mkdir(). maybe even don't include dir functions at all? 03.08.03 # <__builtin> "ROCKBOX_LEAN_AND_MEAN" 03.08.21 # a la WIN32_LEAN_AND_MEAN ? 03.08.36 # Mkdir gets removed as unused code so not a big deal there 03.08.55 # do you use ffunction-sections? 03.09.20 # I didn't look in the default make file 03.10.27 # ill try it and see IDK though since its technically in a function not standalone 03.11.24 # we should probably use that anyway. I think the reason it wasn't was before eabi, it would generate long calls. that may no longer be the case 03.12.52 # hmmm...looks like it is there for a bootloader build 03.13.58 # yep EXTRA_DEFINES 03.14.48 # the problem is that create is in with open so its a valid codepath 03.16.08 # define O_CREAT as 0 and it will be optimized out 03.16.19 # really? 03.17.39 # I guess FF_CREAT too. if the code is unreachable the compiler will remove it 03.20.08 # yep nice that makes it cleaner :) 03.20.59 # so maybe activate it with BOOTLOADER_READ_ONLY_FS 03.22.26 # well I guess thats not true 03.22.57 # should probably dump FF_CREAT and FF_EXCL anyway. they don't get stored and they aren't used in deep code. they were just put there to pass flags in a single parameter. doing that would give me a couple bits back for other functions 03.22.59 # If we removed all file write functions it would be 03.24.04 # if you aren't calling the write functions I guess the -ffunction-sections will dump them anyway 03.24.27 # yeah as long as they are separate.. 03.26.02 # I was going to change it for you so there's only O_CREAT to worry about. I don't need the other junk. 03.26.57 # sounds logical 03.30.28 # in that case i'll do separate commits for each 03.31.03 # * jhMikeS was just going to edit it up to pass oflag straight to the inner function 03.33.21 Quit Bray90820 (Ping timeout: 276 seconds) 03.34.16 # sounds good but I'll put it on your commit and the other on main 03.47.52 Quit ZincAlloy (Quit: Leaving.) 03.48.19 Quit ComputingComet (Read error: Connection reset by peer) 04.00.23 # Bilgus: pushed it to g1561 for the moment 04.00.24 # 3Gerrit review #1561 at http://gerrit.rockbox.org/r/1561 : 3Remove FF_CREAT and FF_EXCL flags in from file code. by Michael Sevakis 04.00.36 # cool thanks 04.14.28 # worked perfectly 04.15.17 # clip v1 fits in the fw file again and should be enough left over to allow me to do external boot + bootdata in it 05.01.18 *** Saving seen data "./dancer.seen" 05.29.21 Join [Saint] [0] (~sinner@rockbox/staff/saint) 05.48.44 Quit alexweissman (Remote host closed the connection) 06.15.56 Join alexweissman [0] (~alexweiss@c-68-51-123-75.hsd1.in.comcast.net) 06.17.19 Quit _mt_ (Ping timeout: 255 seconds) 06.42.45 Join TheEaterOfSouls [0] (~chatzilla@unaffiliated/theeaterofsouls) 06.42.54 # Hey guys, not sure what commit might've caused this as I haven't updated since August when the new USB driver was introduced but I just installed the latest build on my Clip Zip and both FLAC and MP3 randomly skip when playing (both on internal and external storage). Anyone else having this issue? 06.51.38 # Build Server message: 3New build round started. Revision dc22522, 255 builds, 17 clients. 06.52.23 # <[Saint]> TheEaterOfSouls: can you verify this with a clean config? 06.53.03 # I will do so 06.53.26 # <[Saint]> Please backup your config to config.cfg.bak and then remove the original, and tell me if you can reproduce. 06.56.45 Quit [7] (Ping timeout: 256 seconds) 06.57.02 Join TheSeven [0] (~quassel@rockbox/developer/TheSeven) 06.59.54 Quit furrywolf (Ping timeout: 245 seconds) 07.00.29 # Didn't bother backing it up haha but I am no longer having the issue, sorry. 07.01.22 *** Saving seen data "./dancer.seen" 07.01.42 # Build Server message: 3Build round completed after 605 seconds. 07.01.43 # Build Server message: 3Revision dc22522 result: 4 errors 0 warnings 07.10.41 # <[Saint]> TheEaterOfSouls: damn, that's mildly annoying. 07.10.56 # <[Saint]> If you can get it to happen again...I would appreciate seeing the config. 07.11.15 # <[Saint]> I mean, I'm happy your issue is fixed, but I would love to have known the trigger. 07.11.45 # I will change settings back one by one and let you know if it happens. 07.12.37 # I guess I just assumed something to do with config handling was changed, I haven't looked at all of the commits since August lol. 07.16.10 # <[Saint]> That would be several hundred. 07.18.28 # Yes. 07.54.04 Quit alexweissman (Read error: Connection reset by peer) 07.54.26 Join alexweissman [0] (~alexweiss@c-68-51-123-75.hsd1.in.comcast.net) 07.59.00 Quit alexweissman (Ping timeout: 255 seconds) 07.59.28 Join alexweissman [0] (~alexweiss@c-68-51-123-75.hsd1.in.comcast.net) 08.01.40 # __builtin: Fixed (hopefully) 08.11.11 Quit TheEaterOfSouls (Quit: ChatZilla 0.9.93 [Firefox 51.0.1/20170125094131]) 08.39.27 Join ender` [0] (krneki@foo.eternallybored.org) 08.59.42 Quit Horrorcat (Quit: Gateway shutdown) 09.01.23 *** Saving seen data "./dancer.seen" 09.01.50 Join xorly [0] (~xorly@ip-89-176-102-19.net.upcbroadband.cz) 09.02.20 Join Horrorcat [0] (129994c4c6@unaffiliated/horrorcat) 09.43.21 Quit xorly (Ping timeout: 256 seconds) 09.52.39 Join PurlingNayuki [0] (~Thunderbi@218.18.200.2) 09.57.21 Quit PurlingNayuki (Ping timeout: 264 seconds) 10.10.56 Join elensil [0] (~edhelas@2001:1c02:1903:d800:c4af:b9a9:eafa:cc75) 10.13.49 Join MrZeus [0] (~MrZeus@81.144.218.162) 10.55.16 Join clipzip [0] (51f32cb7@gateway/web/freenode/ip.81.243.44.183) 10.55.21 # http://themes.rockbox.org is down? 10.56.34 # <[Saint]> If you can't get to it...yes. 11.01.25 *** Saving seen data "./dancer.seen" 11.03.50 Join skapazzo [0] (~skapazzo@151.9.205.1) 11.06.17 # strange, because https://www.rockbox.org/ is working 11.06.34 # <[Saint]> Not particular. 11.06.39 # <[Saint]> Different servers. 11.07.00 # <[Saint]> You'll likely find the forums down too. 11.07.13 # <[Saint]> If they're not, then there's some weird shit going on. 11.07.30 # they are down too 11.07.33 # thanks for the information 11.07.43 # <[Saint]> de nada. 11.20.02 Join xorly [0] (~xorly@193.85.203.185) 11.34.54 Join ZincAlloy [0] (~Adium@2a02:8108:8b80:1700:1d24:5e5:1418:9a28) 11.39.18 # themes.rockbox.org took too long to respond 11.39.20 # failured. 11.52.18 Join Rower [0] (husvagn@d83-183-134-99.cust.tele2.se) 11.56.08 Join robertd1 [0] (~root@186-90-12-124.genericrev.cantv.net) 12.02.03 Join pamaury [0] (~pamaury@rockbox/developer/pamaury) 12.05.56 # Bilgus: I've seen you work to reduce bootloader size. I am not sure I'm a big fan of it, especially removing nontrivial functionality like file writing. It's quite error-prone and obscure 12.11.33 # <[Saint]> yeah, in my opinion you need a _damn_ good reason to touch bootloaders. 12.11.59 # <[Saint]> especially when neither size nor boot time is an active concern. 12.13.47 # well size is a concern on the clipv1 apparently 12.13.49 Quit ZincAlloy (Quit: Leaving.) 12.14.53 Quit [Saint] (Read error: Connection reset by peer) 12.15.05 # the question is: is it worth adding to the ifdef hell even more just to add this functionality to the clipv1 ? It's an old target now, I know people love it but that's a race that will eventually be lost, the bootloader size is only ever increasing over time and we are already a few bytes away from the maximum on those targets 12.19.53 Join [Saint] [0] (~sinner@rockbox/staff/saint) 12.35.37 Quit jhMikeS (Ping timeout: 256 seconds) 12.46.27 Join cc___ [0] (~ac@2001:910:113f:1:6a05:caff:fe1c:1627) 13.01.29 *** Saving seen data "./dancer.seen" 13.38.40 Quit xorly (Ping timeout: 252 seconds) 13.46.39 Quit prof_wolfff (Ping timeout: 240 seconds) 13.55.19 Join mutnai [0] (6db90a3e@gateway/web/freenode/ip.109.185.10.62) 13.59.41 Join prof_wolfff [0] (~prof_wolf@82.159.0.123.dyn.user.ono.com) 14.03.36 Join xorly [0] (~xorly@193.85.203.185) 14.03.56 Join paulk-collins [0] (~paulk@gagarine.paulk.fr) 14.21.41 # actually the clipV1 won't compile bootloader at head 14.30.23 Join dongs_ [0] (~dongs@bcas.tv) 14.30.55 # oh and a slow AF dump of the themes site https://www-drv.com/site/0omecwmaks86mo8wk4eh0w/Rockbox_theme_dump/themes.rockbox.org/ 14.32.08 Quit dongs (Write error: Broken pipe) 14.32.09 Quit Cu5tosLimen (Excess Flood) 14.32.15 Nick dongs_ is now known as dongs (~dongs@bcas.tv) 14.32.17 Quit munch` (Quit: Let's blast this shit and get naked) 14.32.17 Quit pixelma (Quit: No Ping reply in 120 seconds.) 14.32.19 Join Cu5tosLimen [0] (~CustosLim@unaffiliated/cust0slim3n) 14.32.34 Join pixelma [0] (~pixelma@rockbox/staff/pixelma) 14.34.49 Join munch [0] (pls@gateway/shell/elitebnc/x-glmhadwjvxnlvcvx) 14.42.27 # @ Pamaury the other options I thought of were a separate file handling routine for bootloader or splitting open( into standalone functions, the former as jhMikeS pointed out makes upkeep a pita, and the latter strays away from posix. Plus I'm sure it isn't just the clipV1 that won't compile bootloader at head 14.52.50 Quit mutnai (Quit: Page closed) 15.01.30 *** Saving seen data "./dancer.seen" 15.04.25 Quit derf (Ping timeout: 240 seconds) 15.06.22 Join derf [0] (~derf@static-108-18-126-14.washdc.fios.verizon.net) 15.33.03 Join ZincAlloy [0] (~Adium@2a02:8108:8b80:1700:c4f6:1633:841c:9e4e) 15.45.22 Join JanC_ [0] (~janc@lugwv/member/JanC) 15.46.19 Quit JanC (Read error: Connection reset by peer) 15.46.39 Nick JanC_ is now known as JanC (~janc@lugwv/member/JanC) 15.56.45 Quit [Saint] (Ping timeout: 264 seconds) 16.02.49 Join amayer [0] (~amayer@mail.weberadvertising.com) 16.19.13 Quit fs-bluebot (Ping timeout: 240 seconds) 16.19.33 Quit bluebrother (Ping timeout: 264 seconds) 16.20.51 Join bluebrother [0] (~dom@rockbox/developer/bluebrother) 16.29.02 Join [Saint] [0] (~sinner@rockbox/staff/saint) 16.31.57 Join _mt_ [0] (~MT@129.59.122.20) 16.33.29 Join fs-bluebot [0] (~fs-bluebo@xd9beea3a.dyn.telefonica.de) 16.36.18 Join PurlingNayuki [0] (~Thunderbi@218.18.200.2) 17.00.11 Join furrywolf [0] (~randyg@172.56.39.243) 17.01.32 *** Saving seen data "./dancer.seen" 17.10.19 Quit Moarc (Quit: i znowu NADMUCHAŁ BALONA) 17.13.43 Join Moarc [0] (~chujko@a105.net128.okay.pl) 17.22.08 Quit clipzip (Ping timeout: 260 seconds) 17.24.09 Quit _mt_ (Ping timeout: 240 seconds) 17.51.10 Join athidhep [0] (~afoakf@unaffiliated/athidhep) 17.58.53 Quit PurlingNayuki (Read error: Connection reset by peer) 17.58.58 Join PurlingNayuki1 [0] (~Thunderbi@218.18.200.2) 18.03.25 Quit PurlingNayuki1 (Ping timeout: 258 seconds) 18.03.55 Join mutnai [0] (6db91733@gateway/web/freenode/ip.109.185.23.51) 18.09.38 Join girafe [0] (~girafe@LFbn-1-11729-221.w2-7.abo.wanadoo.fr) 18.44.36 Join _mt_ [0] (~MT@129.59.122.20) 18.47.57 Join billyogat [0] (43ec586d@gateway/web/freenode/ip.67.236.88.109) 18.51.21 Quit pamaury (Ping timeout: 264 seconds) 18.52.39 # im getting an ERR_CONNECTION_TIMED_OUT on both themes.rockbox.org and forums.rockbox.org. has anyone else noticed this? 18.58.40 # well, they are working now 19.01.36 *** Saving seen data "./dancer.seen" 19.05.43 Join MrZeus1 [0] (~MrZeus@2a02:c7f:7008:3400:5130:2eda:2c23:ccdf) 19.08.52 Join johnb2 [0] (~johnb2@p57B45262.dip0.t-ipconnect.de) 19.10.37 Quit _mt_ (Ping timeout: 260 seconds) 19.11.46 Quit billyogat (Quit: Page closed) 19.19.22 Quit xorly (Ping timeout: 260 seconds) 19.20.46 Join lebellium [0] (~chatzilla@89-93-177-91.hfc.dyn.abo.bbox.fr) 19.39.26 Join foolsh [0] (~starchase@162-204-199-234.lightspeed.sbndin.sbcglobal.net) 19.42.48 Quit johnb2 (Ping timeout: 255 seconds) 19.50.37 Join TheEaterOfSouls [0] (~chatzilla@unaffiliated/theeaterofsouls) 20.00.47 Join chrisb [0] (~chrisb@pool-71-175-246-111.phlapa.east.verizon.net) 20.07.20 Join jhMikeS [0] (~jethead71@d192-24-173-177.try.wideopenwest.com) 20.20.21 Quit dfkt (Quit: SIC GORGIAMVS ALLOS SVBJECTATOS NVNC.) 20.26.28 Quit mutnai (Ping timeout: 260 seconds) 20.38.09 Quit TheEaterOfSouls (Quit: ChatZilla 0.9.93 [Firefox 51.0.1/20170125094131]) 20.39.11 Join xorly [0] (~xorly@ip-89-176-102-19.net.upcbroadband.cz) 20.45.46 Join Senji [0] (~Senji@85.187.103.250) 21.00.53 Join johnb2 [0] (~johnb2@p57B45262.dip0.t-ipconnect.de) 21.01.39 *** Saving seen data "./dancer.seen" 21.03.04 Join _mt_ [0] (~MT@129.59.122.20) 21.09.54 Quit johnb2 (Ping timeout: 240 seconds) 21.39.52 Join TheLemonMan [0] (~root@irssi/staff/TheLemonMan) 21.59.24 Quit _mt_ (Ping timeout: 240 seconds) 22.07.56 Quit prof_wolfff (Ping timeout: 260 seconds) 22.14.15 Join _mt_ [0] (~MT@129.59.122.20) 22.20.19 Join prof_wolfff [0] (~prof_wolf@82.159.0.123.dyn.user.ono.com) 22.21.55 Join pamaury [0] (~pamaury@rockbox/developer/pamaury) 22.33.07 # jhMikeS, Pamaury: see what you think about this g#1563 22.33.08 # 3Gerrit review #1563 at http://gerrit.rockbox.org/r/1563 : 3Remove O_CREAT codepath when O_RDONLY is the specified open argument by William Wilgus 22.34.43 # Bilgus: see with jhMikeS, does that actually save anything ? the compiler would need to do global optimization to notice 22.36.48 Part chrisb ("rcirc on GNU Emacs 26.0.50.1") 22.49.44 # pamaury: trying to get FWIMAGE --descramble-> --bfin_boot-> ELF --adi-ldr-> --scramble-> FWIMAGE to work 22.50.08 # looks like the ldr tool expects symbols dxe_init_start and dxe_init_end, otherwise it won't generate INIT blocks 22.51.59 # dys: it almost certainly won't work 22.52.20 # an ELF file cannot really encode a DXE perfectly (at least not in general) 22.52.40 # i guess that's where the magic symbols come in? 22.52.57 # what we usually do is that we write a patching tool that takes a fwimage and modifies it directly 22.53.09 # so that you don't take the risk of loosing any information 22.53.20 # I see 22.55.38 # how do you plan to patch the firmware ? 22.55.44 Quit _mt_ (Ping timeout: 245 seconds) 22.56.24 # there's a TEST/CALIBRATION tool you can boot into with a secret key combo 22.56.48 # I could clobber it with my own code, I guess 22.57.04 # is it in it's own elf ? ie is it a separate DXE ? 22.57.13 # no, it's part of the big elf 22.58.03 # ah, that makes it more tricky, especially since the address can change with every firmware 23.00.09 # you could also simply add more code to the binary (at the end of the big DXE) and patch one jump in the test/calibration tool to jump to it ? That can work if the OF jumps early to this code and doesn't overwrite this added code 23.01.43 *** Saving seen data "./dancer.seen" 23.02.58 # injecting a second init DXE between the original init DXE and the application DXE might also be an option 23.03.33 # that way we get DRAM and stack initialized and can boot into the application by simply returning to the boot rom 23.03.54 # yeah but you take the risk of bricking if your code does not return to boot rom for some reason 23.04.20 # do you have a simple way to overwrite the flash ? 23.05.25 # my only attempt at it resulted in a bricked unit… maybe i'll take the noleads flash off with hot air on the bricked unit and trying to flash it out-of-circuit 23.06.22 Join _mt_ [0] (~MT@129.59.122.20) 23.06.28 Quit amiconn (Quit: No Ping reply in 64 seconds.) 23.07.36 Join amiconn [0] (~amiconn@rockbox/developer/amiconn) 23.11.05 Quit paulk-collins (Remote host closed the connection) 23.11.23 Nick SovietShaman is now known as CommunistWitchDr (quasselcor@97-87-177-85.dhcp.stls.mo.charter.com) 23.11.36 # If I can resurrect it, it might make a great prototyping jig 23.21.57 Quit pamaury (Ping timeout: 258 seconds) 23.24.05 # http://forums.rockbox.org/index.php/topic,51664.msg239123/topicseen.html#msg239123 23.24.10 # thank you Vortex... 23.27.08 Quit TheLemonMan (Quit: "It's now safe to turn off your computer.") 23.27.49 # He wanted his own paypal button 23.28.05 Quit robertd1 (Ping timeout: 258 seconds) 23.30.50 Join Bilgus_ph [0] (~Bilgus_ph@108.100.144.152) 23.32.43 # <__builtin> lebellium: hmm? 23.33.08 # Pamaury & jhMikeS it allows the compiler to optimize it out but unfortunately is not in compliance with Posix so disregard :/ 23.36.20 # __builtin: if you didn't follow the story, you should start reading http://forums.rockbox.org/index.php/topic,51139.0.html 23.36.57 # they guy understood the whole spirit of rockbox 23.37.02 # ... 23.37.47 Quit Bilgus_ph (Read error: Connection reset by peer) 23.45.24 Quit lebellium (Quit: ChatZilla 0.9.93 [Firefox 51.0.1/20170125094131]) 23.46.24 Quit MrZeus1 (Ping timeout: 255 seconds)