| 00:10:44 | | Join ac_laptop [0] (~ac_laptop@186.2.247.129) |
| 00:35:52 | *** | Saving seen data "./dancer.seen" |
| 00:40:04 | | Quit Huntereb (Read error: Connection reset by peer) |
| 00:43:25 | | Quit mendelmunkis (Ping timeout: 240 seconds) |
| 02:00 |
| 02:35:55 | *** | Saving seen data "./dancer.seen" |
| 02:42:58 | | Quit massiveH (Quit: Leaving) |
| 02:56:17 | | Quit ac_laptop (Ping timeout: 260 seconds) |
| 03:00 |
| 03:05:46 | | Join Rower [0] (~Rower@78-73-72-39-no2340.tbcn.telia.com) |
| 03:20:39 | | Join lebellium [0] (~lebellium@89-92-69-66.hfc.dyn.abo.bbox.fr) |
| 03:46:38 | | Join johnb3 [0] (~johnb2@p5b3af70f.dip0.t-ipconnect.de) |
| 03:59:42 | | Join w4tchguard [0] (~w4tchguar@2001:ac8:84:33::a10d) |
| 04:00 |
| 04:05:03 | | Quit olspookishmagus (Quit: All for nothing) |
| 04:28:35 | | Quit J_Darnley (Ping timeout: 258 seconds) |
| 04:29:59 | | Join J_Darnley [0] (~J_Darnley@d51A44418.access.telenet.be) |
| 04:35:56 | *** | Saving seen data "./dancer.seen" |
| 04:56:03 | | Join olspookishmagus [0] (~pookie@snf-137798.vm.okeanos.grnet.gr) |
| 05:00 |
| 05:38:44 | | Join pamaury [0] (~pamaury@rockbox/developer/pamaury) |
| 05:58:35 | | Join hardly [0] (240c00eb@KD036012000235.au-net.ne.jp) |
| 06:00 |
| 06:05:03 | | Quit hardly (Remote host closed the connection) |
| 06:21:11 | | Join sakax [0] (~r0b0t@unaffiliated/r0b0t) |
| 06:34:09 | _bilgus_ | builtin I fixed that like a yaer ago think it was parens |
| 06:35:49 | _bilgus_ | re 2048 |
| 06:36:00 | *** | Saving seen data "./dancer.seen" |
| 06:37:59 | johnb3 | speachy : How is the feel of the buttons of the Walker? Overall I like the case of the Walker better, but the buttons seem more solid on the Eros K. The latter (Surfans) is also 15 Euros cheaper over here. |
| 06:38:12 | w4tchguard | is usb_hid_send exposed the the lua plugin? |
| 06:40:13 | _bilgus_ | very doubtful unless it has simple enough parameters to get added automagically (ie no pointers) |
| 06:41:45 | _bilgus_ | as long as the function is already in plugin.h its just a bit of glue to have it in lua |
| 06:41:46 | w4tchguard | is the a way make plugins without compiling everything? |
| 06:41:53 | w4tchguard | ah |
| 06:42:41 | w4tchguard | apps/plugin.c:740: usb_hid_send, |
| 06:42:42 | w4tchguard | apps/plugin.h:857: void (*usb_hid_send)(usage_page_t usage_page, int id); |
| 06:43:03 | _bilgus_ | I have very bad luck with trying to compile plugins without making the whole build |
| 06:43:26 | w4tchguard | that is why i wanted to use lua but I am not great at it |
| 06:43:55 | _bilgus_ | OH me either but better than when I started!\ |
| 06:43:55 | w4tchguard | How would i go about making some glue code for it |
| 06:47:20 | _bilgus_ | well usage_page_t is a struct |
| 06:48:01 | _bilgus_ | in the source the file lua/rocklib.c has most of the other glue for plugin.h |
| 06:49:22 | _bilgus_ | main thing with lua is keeping the stack ordered right you gotta push and pop stuff in proper proportions if that makes sense |
| 06:50:30 | _bilgus_ | as for the struct you'll need to fill its types with lua primatives, check_int bool string etc. |
| 06:51:16 | _bilgus_ | primitive no coffee no speel |
| 07:00 |
| 07:47:34 | | Quit johnb3 (Quit: Nettalk6 - www.ntalk.de) |
| 07:48:24 | speachy | johnb3: they're stiff. including the wheel, which IMO requires too much force. But I also bought this thing secondhand. |
| 07:48:33 | speachy | the whole thing feels _very_ solid. |
| 07:54:33 | | Quit pamaury (Ping timeout: 260 seconds) |
| 07:56:52 | | Quit w4tchguard (Ping timeout: 260 seconds) |
| 08:00 |
| 08:10:02 | bluebrother | just came across this: http://usbdm.sourceforge.net/ |
| 08:12:04 | bluebrother | could be interesting if someone wants to look into the iriver bootloader. Some ebay sellers have the hardware rather cheap. Use google to find further documentation, doesn't seem to be linked directly on the page. |
| 08:13:46 | speachy | johnb3, overall impressions −− I like it more than the x3ii, even with the wheel that doesn't quite feel right. |
| 08:15:29 | speachy | they're equivalent feature-wise, but the eros/etc effetively has two more buttons. |
| 08:18:27 | | Quit TheSphinX^ (Ping timeout: 260 seconds) |
| 08:31:30 | | Join johnb7 [0] (~johnb2@p5b3af70f.dip0.t-ipconnect.de) |
| 08:35:33 | johnb7 | speachy : I am running a battery bench on the x3ii right now to be able to calibrate the battery numbers. I thought it was in the range of 10 -12h. For the walker they claim 29h for mp3. |
| 08:36:01 | *** | Saving seen data "./dancer.seen" |
| 08:45:21 | | Join Barlow [0] (~barlow@unaffiliated/barlow) |
| 08:58:58 | speachy | the x3ii nominally has a higher quality DAC and analog output paths. |
| 08:59:47 | | Quit paulk-leonov (Excess Flood) |
| 09:00 |
| 09:00:08 | johnb7 | and that has such an impact on current consumption? |
| 09:00:25 | speachy | optimzed for "quality" not power. |
| 09:01:32 | speachy | the eros also shares that output switching/muting lag. |
| 09:02:07 | speachy | but doesn't have the click/pop that necessitates muting things routinely |
| 09:02:30 | speachy | (oh, for kernel sources to see WTF they are doing...) |
| 09:06:14 | | Join paulk-leonov [0] (~paulk-leo@vpn-0-22.aquilenet.fr) |
| 09:14:43 | | Quit johnb7 (Quit: Nettalk6 - www.ntalk.de) |
| 10:00 |
| 10:36:03 | *** | Saving seen data "./dancer.seen" |
| 10:57:07 | | Join TheSphinX^ [0] (~briehl@srv001.nosupamu.de) |
| 10:59:42 | | Join ac_laptop [0] (~ac_laptop@186.2.247.129) |
| 11:00 |
| 11:08:21 | | Quit TheSphinX^ (Ping timeout: 272 seconds) |
| 11:27:54 | | Join TheLemonMan [0] (~lemonboy@irssi/staff/TheLemonMan) |
| 11:44:28 | | Join Huntereb [0] (~Huntereb@174.226.10.229) |
| 12:00 |
| 12:18:06 | | Join efqw [0] (uid412670@gateway/web/irccloud.com/x-cjmgdgoswnanyjsg) |
| 12:18:53 | efqw | Good day speachy :) |
| 12:19:13 | efqw | I see you've made lots of progress on the code cleanup |
| 12:24:02 | | Join pamaury [0] (~pamaury@rockbox/developer/pamaury) |
| 12:36:07 | *** | Saving seen data "./dancer.seen" |
| 12:42:21 | efqw | speachy: lmk if you need help testing the m3k stuff, I still haven't glued mine back together |
| 12:43:00 | efqw | And I can flash anything without worrying about bricking the thing, since I have a full backup of the nand |
| 12:46:08 | efqw | I'll try to stay on as much as possible in the following week just in case |
| 12:46:51 | efqw | Do you have any ideas on how to construct a valid ota image? I'd rather not rely on the USB Cloner is possible. |
| 12:47:02 | efqw | *if |
| 12:49:15 | efqw | Additionally, regarding the F1C100s/F1C200s (brought up earlier this month, I just read the logs), there is a cheap Linux based handheld called PocketGo that is powered by F1C100s. That'd be a decent option as a starter hardware if you want to tinker with a bare metal port. |
| 12:50:50 | efqw | There are pads for an SPI flash in this device but it's unpopulated, I'm pretty sure you can add one in yourself if you want. |
| 13:00 |
| 13:09:11 | | Quit sakax (Read error: Connection reset by peer) |
| 13:10:01 | | Join sakax [0] (~r0b0t@unaffiliated/r0b0t) |
| 13:12:28 | | Quit Moarc (Quit: i znowu NADMUCHAŁ BALONA) |
| 13:12:58 | speachy | efqw: I havne't even started thinking about how to deal with mangling an OTA image. Indeed, I haven't even tried to dig into a running system image to determine what needs to be done |
| 13:14:20 | speachy | efqw: a couple of questions, if you don't mind. First, where does the OF mount the sd card? the vortex code claims it's directly at /mnt. |
| 13:15:13 | speachy | oh! one other thing −− do you have an umolested dump of the stock ubifs image? |
| 13:16:12 | speachy | with that I can at least experiment with integration tools. |
| 13:16:23 | | Join Moarc [0] (~chujko@a105.net128.okay.pl) |
| 13:17:51 | efqw | > where does the OF mount the sd card? the vortex code claims it's directly at /mnt |
| 13:17:51 | efqw | One sec, I'll boot the player and double check. |
| 13:21:16 | efqw | https://www.irccloud.com/pastebin/Oon8Hsaw/ |
| 13:21:36 | efqw | Yup, it's /mnt |
| 13:21:38 | speachy | huh, did not expext that. |
| 13:22:38 | efqw | About the UBI image, you can extract it from the OTA file. |
| 13:23:16 | speachy | yeah, concatenate all the bits together. was hoping to put that off for a little bit |
| 13:23:34 | speachy | wtf is with these insane markups on pocketgo hardware.. |
| 13:23:46 | efqw | I'll dig up a concatenated one and upload it to somewhere, one sec |
| 13:26:43 | efqw | https://www33.zippyshare.com/v/aWCXjEUj/file.html |
| 13:26:59 | efqw | This is from v1.4.3 |
| 13:27:41 | speachy | ok, a pocketgo v1 is ordered, adding to growing pile of rockbox dev hardware. |
| 13:27:58 | efqw | lol |
| 13:31:26 | speachy | got it, thanks. |
| 13:33:56 | efqw | Basically I have 2 ideas right now for poking with the ota stuff |
| 13:35:07 | efqw | 1. I could boot into recovery, try to kill the updater, and copy everything from that initramfs onto my MicroSD for further analysis |
| 13:35:38 | efqw | 2. I could try to construct an ota file *without* the correct signature and see if it complains |
| 13:38:25 | efqw | (I wasn't able to properly extract the initramfs on the recovery image, so I think doing it live is easier for my sanity) |
| 13:39:38 | speachy | I need to explore the OTA image to determine what changes need to be made. Making the actual changes and regenerating the ubifs image is easy. |
| 13:40:13 | speachy | in parallel, I have to write a script that deconstructs the OTA image into its constituent parts, and rebuilds a usable OTA image from those parts. |
| 13:41:10 | efqw | I was only focusing on generating a usable OTA file with the modified UBI image, I assume the other parts are relatively trivial. |
| 13:41:41 | efqw | And I highly recommend using something like Beyond Compare for comparing those two rootfs trees |
| 13:42:11 | efqw | I used 1.4.3 because that's the version that xvortex's rb release was based on |
| 13:43:20 | efqw | The unpacking and concatenating is mostly busy work as we know how those work already. |
| 13:43:35 | efqw | The real mystery is the recovery itself. |
| 13:51:46 | efqw | Well, I found a public key. This is problematic. |
| 13:52:27 | efqw | The presence of this key might indicate the updater is indeed checking the signature. |
| 13:56:21 | | Join TheSphinX^ [0] (~briehl@srv001.nosupamu.de) |
| 13:58:05 | efqw | `[ 776.061580] EXT4-fs (mmcblk0p1): couldn't mount as ext3 due to feature incompatibilities` |
| 13:58:07 | efqw | Oh wow |
| 13:58:27 | efqw | I might need to format the card as ext3 or ext2 |
| 14:00 |
| 14:19:51 | speachy | is adb always running? it appears to be.. |
| 14:24:14 | efqw | oddly enough I've never tired `adb devices` with the player connected to my computer |
| 14:24:15 | speachy | the system integration for the m3k seems to be even more of a hack than the hiby stuff. |
| 14:24:30 | speachy | looks like they turn on adb unconditionally |
| 14:25:07 | speachy | in the same script that launches the player |
| 14:25:17 | efqw | yeah, when you decompile the player binary (which is supposed to be oss since they bundled ffmpeg inside but I digress), you can find out how it's turning the dac on and off |
| 14:25:31 | efqw | the entire thing is a hacky mess tbh |
| 14:26:25 | speachy | with adb we could patch ourselves into a running system, no need to muck with binaries. |
| 14:28:28 | efqw | does adb on this device actually work? |
| 14:28:42 | speachy | no idea, haven't actually pulled mine out of the box yet. :D |
| 14:28:44 | efqw | I did a `ps` in recovery and I didn't find the adb daemon |
| 14:28:58 | speachy | but if I believe the scripts I'm reading, it's unconditionally running. |
| 14:29:12 | speachy | (granted the player binary could be disabling it afterwards) |
| 14:29:44 | efqw | (in the recovery initramfs there's an inittab entry to enable adb but the binaries are missing) |
| 14:30:08 | efqw | I'll reboot and check the regular partition |
| 14:30:34 | speachy | it's definitely present in the main filesystem |
| 14:30:43 | efqw | https://www.irccloud.com/pastebin/2BiycGwr/ |
| 14:31:12 | efqw | Oh wait I messed up my UBI partition |
| 14:31:18 | speachy | sweeeet |
| 14:31:19 | efqw | Have to reflash it first |
| 14:31:44 | speachy | I have the eros (and all of the other hiby-based platforms) running with external USB storage auto-mounted. |
| 14:32:31 | efqw | Search for `modprobe` in the disassembled player binary :) |
| 14:35:15 | efqw | # /etc/init.d/S90adb start |
| 14:35:15 | efqw | Starting mass_storage... |
| 14:35:15 | efqw | notice : mass storage and adb don't use, kernel config error |
| 14:36:11 | *** | Saving seen data "./dancer.seen" |
| 14:42:25 | efqw | Found the problem. `/sys/class/android_usb/android0` does not exist. |
| 14:42:32 | efqw | So adb doesn't work. |
| 14:43:10 | speachy | ok, so it's not an android base kernel. that's ... odd, as the standard ingenic platform is. FiiO might have just disabled it, I guess |
| 14:44:19 | efqw | There is another thing that is worth noting: if you don't have an SD card inserted and attach the device over USB, the player binary will load the serial gadget modules instead. |
| 14:45:22 | efqw | There is no shell on it obviously, it's reserved for writing SN (/data/userfs/sn.txt) |
| 14:45:43 | efqw | If we can exploit this somehow and get a shell from it, the rest should be trivial, |
| 14:46:33 | speachy | to launch into recovery, does the same /proc/jz/reset trick work? |
| 14:47:10 | efqw | echo 1 to /proc/jz/reset/reset or? |
| 14:47:23 | speachy | echo recovery > /proc/jz/reset/reset |
| 14:47:42 | speachy | 'cat /proc/jz/reset/reset' should give you the legal options. |
| 14:47:50 | efqw | yes |
| 14:47:59 | speachy | excellent. |
| 14:54:08 | braewoods | speachy: my parts finally arrived and i have modded the ihp-120 with CF card and new battery... interestingly i had to do some snipping on the adapter so it would fit. |
| 14:54:36 | braewoods | including cutting off the jumper for slave / master |
| 14:54:41 | braewoods | since it was too tall |
| 14:54:43 | braewoods | :o |
| 14:54:52 | braewoods | seems to work in spite of it being in slave mode though |
| 14:55:46 | braewoods | fortunately i have other options if i need the jumper circuit shorted... |
| 14:55:53 | braewoods | like permanently soldering it |
| 14:55:58 | braewoods | but seems to work open |
| 14:56:14 | braewoods | so i see no reason to modify it further |
| 14:56:36 | braewoods | i'll work on the spare unit i acquired too |
| 14:56:39 | braewoods | another day |
| 14:56:51 | braewoods | at least now i know what to expect |
| 15:00 |
| 15:03:58 | | Join johnb7 [0] (~johnb2@p5b3af70f.dip0.t-ipconnect.de) |
| 15:05:28 | braewoods | i suspect hard drives are going to be phased out in newer laptops |
| 15:05:44 | braewoods | with m.2 being so much faster and flash prices continuing to drop... |
| 15:05:58 | speachy | sata? yes. emmc on the low end, nvme on the rest |
| 15:06:00 | braewoods | it's only a matter of time before the 2.5" hard drive goes extinct. |
| 15:06:19 | speachy | just enough local storage to run the OS, "the cloud" for everything else |
| 15:06:20 | braewoods | for newer devices anyway |
| 15:06:36 | braewoods | speachy: bleh. |
| 15:06:42 | braewoods | fuck the chromeos model |
| 15:06:44 | braewoods | lol |
| 15:07:07 | speachy | the sad thing is that the chromeos model is a huge net improvement |
| 15:07:19 | braewoods | for whom? |
| 15:07:55 | braewoods | i prefer computers that are still semi-usable if the external network goes down |
| 15:08:07 | speachy | low-end shovelware laptops that are barely usable out of the box, and by the time their clueless users have used 'em for more than an hour, infested with even more malware. |
| 15:08:21 | braewoods | fair enough. |
| 15:08:38 | braewoods | just i can't see chromeos being viable for serious work... |
| 15:08:47 | braewoods | i guess it works for schools though. |
| 15:09:13 | braewoods | much of what i do i can't see happening with chrome os |
| 15:09:37 | braewoods | speachy: incidently i addressed that problem for my home users by switching them to linux mint |
| 15:09:40 | braewoods | just as good for now |
| 15:10:12 | braewoods | never been hit by ransomware... |
| 15:10:27 | braewoods | but sheesh it's nasty from what i've read |
| 15:11:12 | braewoods | speachy: quick question. how many rockbox targets have IO LEDs? |
| 15:11:17 | braewoods | i've found one that does. |
| 15:11:42 | braewoods | the iriver h1xx series has LEDs when connected to power and one for disk activity |
| 15:13:32 | efqw | speachy: I tried renaming /res/key/key.pub and the recovery binary wouldn't even launch anymore |
| 15:13:44 | efqw | so I think the pubkey is indeed being checked |
| 15:14:30 | braewoods | efqw: what target? |
| 15:14:43 | efqw | we're working on the fiio m3k |
| 15:14:56 | braewoods | and the bootloaders or so are signed? |
| 15:15:16 | braewoods | interesting. |
| 15:15:37 | braewoods | i can't see why an MP3 player or so would need that... |
| 15:15:43 | efqw | bootloader is not signed, you can replace it if you want |
| 15:15:50 | braewoods | ah. |
| 15:16:13 | speachy | braewoods: ...but the update image is. :/ |
| 15:16:39 | efqw | but the recovery partition (kernel+tiny initramfs) contains a public key that the recovery binary (/sbin/recovery) checks |
| 15:16:57 | braewoods | well, if you can figure out how the signing works and can replace the public key... |
| 15:17:33 | efqw | We *can* already flash literally everything (including the bootloader) with the help of ingenic's USB cloner tool. |
| 15:18:09 | efqw | What we're trying to find out is whether it's possible to skip Ingenic's USB cloner tool or not |
| 15:18:34 | braewoods | no idea... |
| 15:18:44 | braewoods | it's a shame you can't completely replace the OF here |
| 15:18:56 | braewoods | it's possible to do that on a few targets |
| 15:19:20 | braewoods | i learned the iriver h1xx port can be fully replaced... |
| 15:19:39 | braewoods | i replaced the OF on mine with rockbox stored in ROM |
| 15:19:47 | efqw | You totally can strip away a lot of the stuff and only leave in the bare minimum required for hosted rockbox to run. |
| 15:20:16 | braewoods | what's a "USB cloner" exactly? |
| 15:20:27 | braewoods | it sounds like what dd can do when used on a device in UMS mode |
| 15:20:31 | efqw | It's ingenic's proprietary usb flashing tool |
| 15:20:37 | efqw | It talks to the boot rom |
| 15:20:37 | braewoods | Oh. |
| 15:20:43 | braewoods | different from UMS |
| 15:21:39 | efqw | Well, technically not to the bootrom, it uploads a precompiled u-boot SPL, jumps to that, and then talks to the u-boot |
| 15:22:11 | efqw | the u-boot is the actual software on the target doing flash operations |
| 15:22:37 | braewoods | one thing i miss about the older devices that relied on firmware stored on their removable storage |
| 15:22:43 | braewoods | wasn't really possible to brick them |
| 15:23:07 | efqw | FiiO didn't even use u-boot on this device. |
| 15:23:34 | braewoods | interesting... there's a lot of devices that use u-boot but you can't get the source for it |
| 15:23:40 | braewoods | nor the kernel for that matter |
| 15:23:48 | efqw | This one is running X-Loader |
| 15:24:09 | braewoods | openwrt usually works by leaving the u-boot alone and just installing a new firmware in place of the old |
| 15:24:45 | efqw | and adding lzma-loader if the stock bootloader is trash |
| 15:25:12 | braewoods | for more space? |
| 15:25:18 | braewoods | i can't see why else anyone would use lzma |
| 15:25:30 | braewoods | not sure how practical it is with a low RAM system |
| 15:25:48 | braewoods | xz / lzma are both RAM hungry |
| 15:25:53 | efqw | The vast majority of the kernel images on openwrt are lzma compressed. |
| 15:25:54 | braewoods | although that's more on the compressor side |
| 15:26:05 | braewoods | ah. i was thinking rockbox. |
| 15:26:38 | efqw | I'm not familiar with bare metal rb targets. |
| 15:26:49 | efqw | I'm mostly here for hosted ones |
| 15:28:06 | efqw | Wanted to try the sony stuff but I really dislike proprietary connectors |
| 15:28:18 | efqw | So Ingenic it is. |
| 15:28:22 | braewoods | so basically this tool is like heimdall or odin for samsung devices |
| 15:28:26 | efqw | yes |
| 15:28:43 | efqw | or sunxi-fel/rkdeveloptool |
| 15:29:36 | braewoods | i think most boards should have a rescue ROM that can't be overwritten so you can rescue it over USB or so |
| 15:29:53 | braewoods | a much easier alternative to JTAG |
| 15:32:35 | efqw | speachy: tag me if you need something else, or if you want to send me a build to play with, I got some other stuff to attend to. |
| 15:33:02 | efqw | Try the serial thing, load it in ghidra and see if you can find any exploits. |
| 15:33:30 | efqw | I'm not an expert when it comes to finding potential vulns |
| 15:34:26 | braewoods | i can think of a way to get some people at it ;p |
| 15:34:45 | braewoods | fake a press release that this device is unhackable |
| 15:34:47 | braewoods | lol |
| 15:48:24 | _bilgus_ | braewoods, recommend against removing the OF completely just in case something happens |
| 15:48:40 | _bilgus_ | granted it wouldn't matter so much on some targets |
| 15:50:04 | | Join w4tchguard [0] (~w4tchguar@2001:ac8:84:33::a10d) |
| 15:53:55 | | Join TheSphinX_ [0] (~briehl@srv001.nosupamu.de) |
| 15:53:59 | | Quit TheSphinX^ (Ping timeout: 258 seconds) |
| 15:54:07 | | Nick TheSphinX_ is now known as TheSphinX^ (~briehl@srv001.nosupamu.de) |
| 16:00 |
| 16:05:19 | braewoods | _bilgus_: in this case the OF is entirely incompatible with the CF cards |
| 16:05:22 | braewoods | so |
| 16:05:27 | braewoods | not much point... |
| 16:05:40 | braewoods | otherwise i'd just leave it probably |
| 16:12:12 | | Quit johnb7 (Quit: Nettalk6 - www.ntalk.de) |
| 16:21:29 | speachy | braewoods: I'm hardly the expert on that.. but I'd wager "not many". especially once we moved to solid-state storage. |
| 16:27:43 | | Quit Rower (Ping timeout: 258 seconds) |
| 16:34:56 | | Quit w4tchguard (Remote host closed the connection) |
| 16:35:16 | | Join w4tchguard [0] (~w4tchguar@2001:ac8:84:33::a10d) |
| 16:36:12 | *** | Saving seen data "./dancer.seen" |
| 17:00 |
| 17:00:01 | lebellium | pamaury: is it possible to add DMP-Z1 support in upgtool? |
| 17:02:17 | lebellium | can you guess the keys without the device? |
| 17:19:15 | | Quit TheLemonMan (Quit: "It's now safe to turn off your computer.") |
| 17:54:12 | | Quit w4tchguard (Ping timeout: 260 seconds) |
| 18:00 |
| 18:30:23 | | Quit ac_laptop (Ping timeout: 260 seconds) |
| 18:31:27 | efqw | Now I'm starting to look for CVEs in the libs they use and in the old kernel. :( |
| 18:31:28 | | Quit Xeha (Ping timeout: 246 seconds) |
| 18:36:16 | *** | Saving seen data "./dancer.seen" |
| 18:38:45 | efqw | Didn't find anything of value in the libs. |
| 18:40:34 | | Join ac_laptop [0] (~ac_laptop@186.2.247.129) |
| 19:00 |
| 19:12:56 | | Quit pamaury (Ping timeout: 258 seconds) |
| 19:13:01 | speachy | I wonder if fiio could be talked into sharing this stuff with us. |
| 19:19:25 | speachy | geez, the 3.10.14 kernel all of the ingenic stuff is based on was released just over seven years ago. and is missing at least four years of fixes (3.10.108, in November 2017) |
| 19:23:58 | | Join lebellium_ [0] (~lebellium@89-92-69-66.hfc.dyn.abo.bbox.fr) |
| 19:24:05 | | Quit lebellium (Ping timeout: 240 seconds) |
| 19:25:07 | | Quit lebellium_ (Client Quit) |
| 19:28:41 | | Quit bluebrother (Disconnected by services) |
| 19:28:46 | | Join bluebrother^ [0] (~dom@rockbox/developer/bluebrother) |
| 19:29:33 | | Join fs-bluebot_ [0] (~fs-bluebo@55d415e4.access.ecotel.net) |
| 19:32:13 | | Quit fs-bluebot (Ping timeout: 260 seconds) |
| 19:54:45 | | Quit sakax (Remote host closed the connection) |
| 20:00 |
| 20:20:00 | | Quit Soap_ (Read error: Connection reset by peer) |
| 20:36:19 | *** | Saving seen data "./dancer.seen" |
| 20:58:19 | speachy | at the end of the day, the only meaningful change in the vortex image was deleting the stock init script and replacing it with one that invokes our launcher |
| 21:00 |
| 21:04:55 | | Join massiveH [0] (~massiveH@ool-18e4e82f.dyn.optonline.net) |
| 21:13:52 | | Join fmlatghor [0] (~lcoogan@2601:5cd:8100:2890:9220:3aff:fe1a:350d) |
| 21:15:53 | fmlatghor | So I recently stumbled upon this project from a reddit thread about hardware audio players that run opensource firmware/software −− do you guys recommend any device(s)? |
| 21:34:53 | massiveH | fmlatghor, all devices running rockbox work basically the same from device to device, so the things to shop for would be space of the drive and availability |
| 21:35:39 | fmlatghor | Oh ok, thanks. |
| 21:51:29 | | Join bluebrother [0] (~dom@rockbox/developer/bluebrother) |
| 21:51:36 | | Join fs-bluebot [0] (~fs-bluebo@55d444eb.access.ecotel.net) |
| 21:51:45 | | Quit fs-bluebot_ (Ping timeout: 240 seconds) |
| 21:52:45 | | Quit bluebrother^ (Ping timeout: 240 seconds) |
| 22:00 |
| 22:02:24 | _bilgus_ | Well the frame buffer code compiles cleanly for targets without remote I haven't even gotten to the remote stuff yet |
| 22:03:54 | _bilgus_ | I decided to just let the plugins that manipulate the framebuffer directly do so for now with a fucntion that can be removed as soon as all are converted and verified |
| 22:04:04 | _bilgus_ | looking at you doom |
| 22:04:39 | _bilgus_ | at this point I don't know if its cleaner but it sure is more flexible |
| 22:05:45 | | Quit fmlatghor (Quit: WeeChat 2.8) |
| 22:10:29 | __builtin | is SDL an offender |
| 22:10:30 | __builtin | ? |
| 22:14:28 | _bilgus_ | give me a minute to finish this round and I'll check |
| 22:16:56 | speachy | there is just so much extra crap on these ingenic linux targets |
| 22:17:15 | speachy | for example, full wifi (ap and client) tools. |
| 22:18:50 | speachy | and the fiio rootfs is even worse. airkiss, ncurses, alsa and tinyalsa (both libraries and cmdline) |
| 22:19:44 | speachy | apple airplay support |
| 22:20:00 | speachy | the airkiss weechat client |
| 22:20:52 | speachy | sshd |
| 22:21:02 | speachy | (and a lot of other network-related tools) |
| 22:21:34 | speachy | ntfs-3g and exfat support −− including mkfs for fat and exfat. |
| 22:21:40 | _bilgus_ | __builtin, I think this is the exhaustive list |
| 22:21:41 | speachy | no fsck though. |
| 22:21:41 | _bilgus_ | doom, fire, imageviewer, invadrox, osd, xlcd, pacbox, pictureflow, plasma, puzzles, rockboy, sdl, test_resize, xworld, zxbox |
| 22:22:19 | _bilgus_ | pretty much anywhere someone implemented their own screen code I imagine |
| 22:22:39 | _bilgus_ | some are extremely straight forward some not so much |
| 22:24:19 | _bilgus_ | I just blanket string replaced them for now as this patch is getting really hard to follow through the 50 different files its spread across |
| 22:24:51 | _bilgus_ | exageration? maybe but probably close to the truth! |
| 22:27:20 | __builtin | right, I can help for puzzles, sdl and xworld |
| 22:27:33 | __builtin | I haven't quite been following your changes - what are you doing exactly? |
| 22:29:26 | _bilgus_ | g#2811 |
| 22:29:28 | fs-bluebot | Gerrit review #2811 at http://gerrit.rockbox.org/r/2811 : LCD core move buf ptr and address look up function viewport struct WIP by William Wilgus |
| 22:30:06 | _bilgus_ | basically viewports carry around their stride so their buffers don't need to be the size of framebuffer |
| 22:30:46 | _bilgus_ | amognst other things it allows me to use lcd primitives on arbritrary buffers |
| 22:30:57 | fs-bluebot | Build Server message: New build round started. Revision 4f87369, 287 builds, 9 clients. |
| 22:33:00 | _bilgus_ | on top of that the image buffer is tied to a struct framebuffer_t that carries buf, elems, and fn adress to a address translation function which would enable flip rotate and some other simple blits |
| 22:33:54 | _bilgus_ | if you wanted to wire it up I should say I havent gotten that far yet |
| 22:36:20 | *** | Saving seen data "./dancer.seen" |
| 22:45:28 | __builtin | hmm, I wish something like that had existed when I first built the puzzles port |
| 22:45:50 | fs-bluebot | Build Server message: Build round completed after 894 seconds. |
| 22:45:54 | fs-bluebot | Build Server message: Revision 4f87369 result: All green |
| 22:45:57 | __builtin | I built something like that for drawing into a scaled-up framebuffer for the "zoom" feature |
| 22:46:17 | __builtin | https://git.rockbox.org/cgit/rockbox.git/tree/apps/plugins/puzzles/rockbox.c#n42 |
| 22:48:04 | _bilgus_ | I did the same but that was a few years ago I feel confident enough with the codebase now to actually do something about it |
| 22:50:53 | _bilgus_ | oh __builtin is that is nice the font part was what I was about to embark on lol I might steal some of the zoom stuff and rewrite in lua |
| 22:51:33 | _bilgus_ | right now it has a bicubic resize taht is subpar to say the least |
| 22:59:47 | speachy | __builtin: so.. feeling up for porting scummvm over? :D |
| 23:00 |
| 23:01:11 | __builtin | _bilgus_: the font pack thing? |
| 23:01:36 | __builtin | speachy: it's C++, no? |
| 23:01:42 | speachy | hmm, true |
| 23:01:55 | __builtin | so "possible" but it'd be less about scumm and more about getting c++ to play nice with rockbox |
| 23:02:06 | __builtin | not that it couldn't be done, of course... |
| 23:02:53 | _bilgus_ | yeah allowing easier access to fonts would be nice even if its implemented in a include script |
| 23:03:45 | __builtin | the problem I ran into with puzzles was that it assumes that fonts of arbitrary size are available |
| 23:04:00 | __builtin | which is not a bad assumption if you have freetype and friends |
| 23:04:19 | __builtin | but since we're bitmap-only that assumption breaks down |
| 23:04:55 | _bilgus_ | I went as far as to make vector numbers but it was too much of a pain to do unicode characters as vectors by hand |
| 23:05:21 | _bilgus_ | well I made a drawing program to do it for me but I still had to draw them |
| 23:06:28 | _bilgus_ | I suppose technically the fonts have vectors inside them maybe the scaling routine just needs to be a bit lower (like in the font engine) |
| 23:23:40 | __builtin | if you're going to do that you may as well port freetype |
| 23:23:55 | __builtin | in fact mendelmunkis already started that |
| 23:24:01 | __builtin | g#2451 |
| 23:24:03 | fs-bluebot | Gerrit review #2451 at http://gerrit.rockbox.org/r/2451 : Hacky freetype by Moshe Piekarski |
| 23:24:15 | __builtin | err, g#2450 |
| 23:24:16 | fs-bluebot | Gerrit review #2450 at http://gerrit.rockbox.org/r/2450 : Freetype compiles by Moshe Piekarski |
| 23:41:43 | speachy | I'd like to land the gcc 4.9.4 toolchain update this weekend. |
| 23:42:40 | speachy | commit the rockboxdev.sh and configure scripts, allow a week for builders to update, then flip the switch on the builders too. |
| 23:58:07 | | Quit TheSeven (Disconnected by services) |
| 23:58:17 | | Join [7] [0] (~quassel@rockbox/developer/TheSeven) |
icon identifies that the person is a core developer (has commit access).