| 01:12.05 | *** join/#arm-netbook QingPei (~qingpei@124.64.125.225) |
| 01:36.34 | *** join/#arm-netbook toxicpsion (c6350927@gateway/web/freenode/ip.198.53.9.39) |
| 01:37.05 | toxicpsion | hno: ping... |
| 01:39.49 | toxicpsion | zenitraM: how about you? you around? |
| 01:44.57 | *** join/#arm-netbook toxicpsion (c6350927@gateway/web/freenode/ip.198.53.9.39) |
| 03:53.00 | A0Sheds | this project is now popular enough to attract daily trolls |
| 03:58.32 | CIA-91 | rhombus-tech: bari master * r46c6ce3ed290 /evaluated_cpus.mdwn: |
| 07:25.56 | CIA-91 | rhombus-tech: scanli master * r8370b5a4f17c /faq/discussion.mdwn: |
| 09:24.13 | *** join/#arm-netbook gordan (~gordan@31.98.186.225) |
| 09:46.45 | *** part/#arm-netbook QingPei (~qingpei@124.64.125.225) |
| 10:42.31 | *** join/#arm-netbook gordan (~gordan@31.98.186.225) |
| 10:58.43 | *** join/#arm-netbook arokux1 (~arokux@merkur178.inf.uni-konstanz.de) |
| 11:38.20 | *** join/#arm-netbook gordan (~gordan@31.98.186.225) |
| 11:42.50 | CIA-91 | rhombus-tech: Klaus master * r9dae7be6bb40 /allwinner_a10/orders/AllwinnerKKR.mdwn: |
| 12:08.41 | mnemoc | weirdly happy to not be able to boot from his 2.6.36 branch either |
| 12:25.50 | mnemoc | uhm... booting the official 2.6.36 failed too.... /me scratches his head |
| 12:32.33 | mnemoc | trying a fresh clone |
| 13:23.55 | mnemoc | zenitraM: what's the minimal stuff you copied from nanda to the SD to get the uImage there working? |
| 14:01.42 | CIA-91 | rhombus-tech: DennisF master * r3c23f9a6b263 /allwinner_a10/orders/abc.mdwn: |
| 14:30.16 | hno | mnemoc, are you creating the uimage correctly? |
| 14:31.22 | mnemoc | make ARCH=arm CROSS_COMPILE=arm-linux-gnueabi- -j16 uImage ? |
| 14:32.19 | mnemoc | and then copying arch/arm/boot/uImage to the SD |
| 14:32.40 | mnemoc | is that correct? I have zero previous experience with uboot |
| 14:37.01 | mnemoc | hno: --^ |
| 14:38.34 | hno | Ah you are using an uImage target in the make files. Have only built uImages by hand which is a bit more complex. |
| 14:38.46 | hno | should work. |
| 14:55.15 | zenitraM | mnemoc: i think i did that.. |
| 14:59.16 | mnemoc | zenitraM: what about other required files in the SD? |
| 14:59.22 | zenitraM | none |
| 14:59.25 | CIA-91 | rhombus-tech: Peter master * rc490567533b9 /allwinner_a10/orders/p.steenbergen__64__j1nx.nl.mdwn: |
| 14:59.29 | zenitraM | the kernel should boot, probably fail to find the root |
| 14:59.30 | mnemoc | uhm |
| 15:00.10 | mnemoc | I think it is booting... and autorebooting because of the empty root |
| 15:00.18 | mnemoc | but it doesn't printk anything at all |
| 15:01.32 | zenitraM | you can also build a boot.img which will contain kernel+initrd.. if the mmc uboot has the nanda command to boot them |
| 15:01.59 | zenitraM | the default initrd on the kernel tree will leave you in a shell |
| 15:02.18 | mnemoc | I only want to fix the silence problem, initrd and userspace is just fanciness for now |
| 15:02.56 | zenitraM | i'm not sure if the problem is on serial output or the parameters not being passed to the kernel |
| 15:03.12 | mnemoc | if I can't make the official 2.6.36 talk I can't expect to do so on 3.3 |
| 15:03.48 | zenitraM | you can try copying the kernel to the u-disk fat partition inside the mele |
| 15:03.55 | zenitraM | and booting from there with the built-in u-boot |
| 15:05.41 | mnemoc | i don't want to touch the nand |
| 15:06.27 | mnemoc | I'll dive into the uboot/sun4i code tomorrow to see what it expects and where does it write the params |
| 15:06.38 | zenitraM | you don't need to do anything dangerous |
| 15:06.43 | zenitraM | it's just the partition that is mounted as /sdcard |
| 15:06.45 | zenitraM | on android |
| 15:07.28 | mnemoc | and how do I remove my changes if something goes wrong? :) |
| 15:07.44 | zenitraM | no need to touch anything else :) |
| 15:07.50 | zenitraM | just use the built-in u-boot console |
| 15:07.53 | zenitraM | to boot the kernel on there |
| 15:08.03 | mnemoc | ok |
| 15:08.44 | mnemoc | zenitraM: did you boot from the external SD? or always from the internal? |
| 15:09.15 | zenitraM | mmm |
| 15:09.20 | zenitraM | i have to look at the logs |
| 15:09.35 | zenitraM | although it might be possible that it rebooted too for me on the external sd |
| 15:09.42 | zenitraM | we blamed it to the axp |
| 15:11.40 | zenitraM | mmm |
| 15:12.25 | zenitraM | the kernel tom sent me worked, it outputted this: http://pastebin.com/NbnCY25b |
| 15:13.13 | zenitraM | also tried several kernels (probably the 2.6.36 too) and they kept silent and then rebooted... |
| 15:13.56 | zenitraM | you might have to load the script.bin file |
| 15:14.05 | zenitraM | to 0x43000000 |
| 15:14.52 | mnemoc | I'm inclined to believe the official 2.6.36 has unfriendly pin (script) (defaults, fixed in the gpl-violating 3.0.8 |
| 15:15.46 | mnemoc | did you manually load script.bin in that succesful test you pasted above? |
| 15:16.10 | zenitraM | i think i didn't |
| 15:16.21 | zenitraM | but the 2.6.36 kernel works on my tablet |
| 15:16.28 | zenitraM | (bImage when being booted from boot1) |
| 15:17.01 | mnemoc | been a different bootloader the comparison is not fair |
| 15:17.38 | zenitraM | i remember also having trouble with u-boot.. i'm sure that using a boot.img the kernel booted from u-boot |
| 15:17.45 | zenitraM | but the 3.0.8 kernel didn't |
| 15:19.09 | zenitraM | https://github.com/amery/linux-allwinner/issues/3#issuecomment-4309650 <-- 2.6.36 uImage booting (on the tablet) from nand u-boot |
| 15:21.01 | mnemoc | then that bootloader must be doing something different than the open sourced *cough* one |
| 15:21.25 | mnemoc | uhm, wait |
| 15:21.31 | mnemoc | starts a 2. |
| 15:21.40 | mnemoc | that's userspace |
| 15:21.48 | mnemoc | kernel _is_ silent |
| 15:23.18 | zenitraM | mmm is the default kernel on the mele silent? |
| 15:24.08 | mnemoc | i don't remember, I can't test until tomorrow :-/ |
| 16:26.48 | CIA-91 | rhombus-tech: Ricardo master * r30118895de8b /allwinner_a10/orders/darkskull2.mdwn: |
| 16:50.47 | mnemoc | zenitraM: boot0, boot1, uboot and Linux-2.6.36-android+ are veeeeery verbose from the begining when booting the stock mele |
| 16:51.23 | zenitraM | then try booting your own 2.6.36 from nand |
| 16:52.22 | zenitraM | although probably the kernel is different, as the sources are from when linux was booted from boot1 |
| 16:52.30 | mnemoc | I'm 99% sure it's because my uboot doesn't load automatically the script file |
| 17:03.45 | zenitraM | no uboot does |
| 17:03.47 | zenitraM | you can load it manually |
| 17:04.28 | zenitraM | get the script.bin (it is on /dev/block/nanda) |
| 17:04.43 | zenitraM | copy it to the mmc, and load it from uboot by hand to 0x43000000 |
| 17:04.48 | zenitraM | using.. mmcload? |
| 17:08.44 | mnemoc | i will |
| 17:09.41 | mnemoc | if that works, next will be to give good defaults for script.bin-less boot |
| 17:10.04 | mnemoc | but life is easier with functional printk |
| 17:19.50 | mnemoc | if it's not uboot, it can be the .axf or boot1... from sd I'm skipping both |
| 17:23.11 | mnemoc | http://dpaste.com/722318/ <- stock |
| 17:23.19 | mnemoc | http://dpaste.com/722319/ <- mmc |
| 17:23.29 | mnemoc | even the loading address difers |
| 17:26.07 | zenitraM | try loading from udisk using the stock... it *might* make a difference, and then you have the whole chain |
| 17:26.20 | zenitraM | script.bin is still loaded by boot1 so manually loading wouldn't be needed |
| 17:28.39 | mnemoc | fatload mmc 0 script.bin; boot made the trick |
| 17:33.08 | zenitraM | you didn't specify the loading address? |
| 17:33.25 | mnemoc | err, yes, sorry |
| 17:33.48 | mnemoc | fatboot mmc 0 0x43000000 script.bin |
| 17:34.15 | zenitraM | ah |
| 17:34.18 | zenitraM | and.. did it work then? |
| 17:34.43 | mnemoc | the official 2.6.36 booted verbosely from mmc |
| 17:35.28 | mnemoc | now that my mind is in peace I can switch back to $work$ :) |
| 17:36.22 | mnemoc | I'll compare the output with my other 2.6.36 branches, and then try with 3.0 |
| 17:36.45 | mnemoc | s/^/tomorrow/ |
| 17:38.34 | zenitraM | nice :) |
| 18:53.03 | CIA-91 | rhombus-tech: lkcl master * r3692850a9511 /allwinner_a10/orders/abc.mdwn: order processing cleanup |
| 19:11.26 | CIA-91 | rhombus-tech: lkcl master * r32b51505cc59 /allwinner_a10/orders/homeconnect.mdwn: cleanup |
| 19:11.27 | CIA-91 | rhombus-tech: lkcl master * r3abe753a17b6 /allwinner_a10/orders/ (5 files in 2 dirs): invalid contact details |
| 19:11.27 | CIA-91 | rhombus-tech: lkcl master * r291f42240c61 /allwinner_a10/orders/all78pat.mdwn: really messy.... |
| 19:11.27 | CIA-91 | rhombus-tech: lkcl master * rf0aff46662a1 /allwinner_a10/orders/ (alpha.mdwn not_valid/alpha.mdwn): invalid contact details |
| 19:11.27 | CIA-91 | rhombus-tech: lkcl master * r9c242ab68e09 /allwinner_a10/orders/andry-i.mdwn: cleanup |
| 19:11.27 | CIA-91 | rhombus-tech: lkcl master * rd3201762df3a /allwinner_a10/orders/ (quantity:2.html quantity:2.mdwn): rename to mdwn from html |
| 19:11.38 | CIA-91 | rhombus-tech: lkcl master * rda9c458dd244 /allwinner_a10/orders/legendinis.mdwn: invalid contact details |
| 19:11.38 | CIA-91 | rhombus-tech: lkcl master * r77cbda3764f4 /allwinner_a10/orders/all78pat3.mdwn: really messy.... |
| 19:11.38 | CIA-91 | rhombus-tech: lkcl master * r57c5018d3408 /allwinner_a10/orders/jonathancom.mdwn: cleanup |
| 19:11.38 | CIA-91 | rhombus-tech: lkcl master * r5cbce299570c /allwinner_a10/orders/ (geolomia.html geolomia.mdwn pald3.html pald3.mdwn): rename to mdwn from html |
| 19:11.47 | CIA-91 | rhombus-tech: lkcl master * r5b8bbf8cd72b /allwinner_a10/orders/pald3.mdwn: cleanup |
| 19:13.51 | CIA-91 | rhombus-tech: lkcl master * r3742fdcccced /allwinner_a10/orders/ (2 files): rename crazy name |
| 19:14.54 | hno | the kernel needs a script.bin loaded at suitable address. Won't work otherwise. |
| 19:15.16 | hno | boot1 loads it by default, but with uboot you need to do it all. |
| 19:15.58 | CIA-91 | rhombus-tech: lkcl master * r46264b10dcf3 /allwinner_a10/orders/ (2 files): rename crazy name |
| 19:15.59 | hno | and depending on kernel & uboot version kernel boot parameters may need to be loaded as well. (if uboot and kernel do not agree on where parameters are) |
| 19:16.38 | CIA-91 | rhombus-tech: lkcl master * r6fba2055302f /allwinner_a10/orders/Hello.mdwn: empty preorder |
| 19:16.39 | hno | mnemoc & zenitraM ^^ |
| 19:16.52 | CIA-91 | rhombus-tech: lkcl master * rf8e0dafe9981 /allwinner_a10/orders/Castrol.mdwn: empty preorder |
| 19:17.24 | hno | but early printk should work on our kernel even without script.bin. But you need suitable boot arguments to enable early printk. |
| 19:17.57 | CIA-91 | rhombus-tech: lkcl master * r99e42550caee /allwinner_a10/orders/ (2 files): rename crazy name |
| 19:18.07 | mnemoc | hno: I think script.bin to (re)enable the ttyS0 at soc level |
| 19:18.25 | mnemoc | s/to/is required to/ |
| 19:18.27 | hno | if you want a ttyS0 then it's needed yes. |
| 19:18.59 | CIA-91 | rhombus-tech: lkcl master * r5178c9aa13ef /allwinner_a10/orders/ (not_valid/orescek.mdwn orescek.mdwn): invalid contact details |
| 19:19.26 | mnemoc | hno: I was our uboot needs boot.scr or uEnv.txt support |
| 19:19.27 | hno | it's even likely the kernel will crash and burn if there is no script.bin loaded. I do not think it's designed to run without it. |
| 19:19.33 | mnemoc | s/was/think/ |
| 19:19.42 | mnemoc | damn, need more coffee desperately |
| 19:20.16 | hno | not sure where the sdcard uboot stores it's environment. |
| 19:20.29 | hno | but that's a separate issue. |
| 19:20.29 | mnemoc | hardcoded on the binary |
| 19:21.06 | hno | boot.scr is just a uboot script. Not strictly needed. |
| 19:21.26 | mnemoc | sure, but it's a way to do whatever loads are needed |
| 19:21.32 | mnemoc | without touching the bin |
| 19:21.33 | hno | yes. |
| 19:21.49 | hno | first step is getting something booting from uboot console. |
| 19:22.53 | hno | there is a hell of a lot difference between u-boot lichee-dev and lichee-dev-mmc. Need to check with Tom what the plan there is. |
| 19:24.04 | hno | lichee-dev is androidified, with usb gadget fastboot support, nand support and much more. lichee-dev-mmc is direct booting with sdcard support but none of the other stuff. |
| 19:25.20 | hno | but I would be pretty glad if we could do the nand via mtd instead. But need to document the nand & dma controllers first.. |
| 19:29.31 | mnemoc | :) |
| 19:30.47 | mnemoc | the less weirdness the better |
| 19:31.30 | hno | their wear leveling layer itsn't too wierd, but it's different. |
| 19:33.02 | hno | and we can do better than plain block level emulation. |
| 19:33.50 | hno | not too interested in having mainly VFAT filesystems. |
| 19:34.33 | mnemoc | I'll push the nand driver to the 3.3 tree as next thing after printk() |
| 19:34.57 | mnemoc | possible before |
| 19:35.01 | mnemoc | possibly* |
| 19:35.58 | hno | Have you rebased your 3.0 tree? Getting some unexpected conflicts. |
| 19:36.10 | mnemoc | an script did :( |
| 19:36.19 | mnemoc | disabled it already |
| 19:37.10 | mnemoc | I had defined it as rebased upon android-common-3.0, they updated, rebase... happened :-/ |
| 19:37.31 | mnemoc | now it's stand alone, 3.3 is also standalone |
| 19:39.35 | mnemoc | really sorry. won't happen again. |
| 19:39.43 | hno | no problem. |
| 19:44.12 | *** join/#arm-netbook Gumboot (~gumboot@rev.bovine.muck.net.nz) |
| 19:49.05 | mnemoc | location of the u-boot env: https://github.com/amery/uboot-allwinner/commit/cf54463fd782c690cf790ca35b5a15504b57c287 |
| 19:59.36 | hno | mnemoc, which allwinner uboot is that? |
| 20:00.22 | hno | ah, the mmc one. |
| 20:01.42 | hno | so it should store on mmc? |
| 20:01.48 | mnemoc | yes, the mmc one |
| 20:02.05 | mnemoc | somewhere between uboot itself and the boot partition |
| 20:02.38 | hno | @544KByte size 128KByte. |
| 20:02.49 | mnemoc | yes, there :) |
| 20:03.36 | hno | wonder why 128Kbyte is allocated for the env. It's only some KB in size. |
| 20:03.51 | hno | most devices have like 4K env area |
| 20:04.05 | hno | not that it matters. |
| 20:04.12 | mnemoc | http://rhombus-tech.net/a10_mmc_boot/ asks to put the SPL at 8KB and uboot at 32KB |
| 20:05.14 | mnemoc | I suppose it's just for the sake of using the available space for something |
| 20:05.21 | hno | BROM looks at 8KB for a boot sector, max 24KByte. |
| 20:06.22 | mnemoc | that gives 512KB for uboot until the env |
| 20:06.52 | hno | well, BROM itself only looks for something much smaller, but the A10 initial bootloader is at most 24KByte. |
| 20:07.38 | mnemoc | the partition table lives in the first 8KB? |
| 20:07.57 | hno | Normally partition table lives in first sector. |
| 20:09.13 | hno | Some excotic systems have partition tables elsewhere, but none that we care about. |
| 20:09.31 | mnemoc | EFI puts it at the end of the disk iirc |
| 20:10.28 | hno | No. The GPT partition table is in the second sector. |
| 20:10.57 | hno | with a backup at the end of the disk. |
| 20:11.49 | mnemoc | i see |
| 20:12.15 | hno | http://en.wikipedia.org/wiki/GUID_Partition_Table |
| 20:13.27 | hno | I wish they had placed the GPT partition table @128K or something, would make hybrid booting much saner. |
| 20:21.33 | hno | mnemoc, which tree is the current one you work on? |
| 20:22.37 | mnemoc | none :p |
| 20:23.14 | hno | which kernel tree does it make sense I start using? |
| 20:23.29 | hno | or continue in 3.0? |
| 20:23.30 | mnemoc | linux-3.3-sun4i includes only the core of sun4i support |
| 20:24.02 | mnemoc | allwinner-v3.0-android stills needs video and sound fixed (api changes) |
| 20:24.24 | hno | which tree is linux-3.3-sun4i based on? |
| 20:24.45 | mnemoc | linus' 3.3 + cherry-picks from allwinner-v3.0-android |
| 20:24.48 | mnemoc | + fixes |
| 20:24.54 | mnemoc | no merge involved |
| 20:25.38 | hno | Ok. |
| 20:26.01 | mnemoc | by core of sun4i I mean arch/arm/mach-sun4i + driver/tty/serial |
| 20:27.06 | mnemoc | planned to add the nand driver first thing after the printk mistery gets solved |
| 20:27.23 | mnemoc | but I can I do before if you want to work on that part |
| 20:28.03 | mnemoc | obviusly you can also work on the 3.0 tree. I can pick your commits from there with the rest |
| 20:28.36 | hno | let's solve the printk mystery first of all. Hard to do anything without printk. |
| 20:29.05 | hno | did you also add the early-printk support? |
| 20:29.20 | mnemoc | yes |
| 20:52.17 | mnemoc | it feels so weird when changing branches takes longer than the actual kernel build... |
| 20:53.40 | CIA-91 | rhombus-tech: Lunchbox master * r67e6ac9f2d72 /allwinner_a10/orders/Lord__95__Lunchbox.mdwn: |
| 20:54.17 | mnemoc | should start using tmpfs when playing kernel hacker wannabe |
| 20:54.29 | hno | Hmm.. seem to have lost my sd-card reader, It's probably on my desk somewhere... |
| 21:01.24 | mnemoc | is really happy his laptop has a builtin sdhc reader |
| 21:01.27 | hno | nice. MMC u-boot running. |
| 21:03.10 | hno | saveenv works fine. |
| 21:08.16 | hno | and it's built with hush parser so boot.scr makes sense. |
| 21:23.01 | mnemoc | Starting kernel ... |
| 21:31.03 | hno | will try that tomorrow night. Now I need to prepare my wifes birthday a bit and get some hours sleep. |
| 21:33.08 | mnemoc | enjoy, good night :) |
| 21:52.19 | CIA-91 | rhombus-tech: Giampiero master * r6a289305dced /Pippo_Kennedy.mdwn: |
| 22:25.28 | hno | Hmm.. arch/arm/mach-sun4i/pm/standby/dram/./../common.h:21:29: error: redefinition of typedef '__s8' |
| 22:25.28 | hno | include/asm-generic/int-ll64.h:19:25: note: previous declaration of '__s8' was here |
| 22:26.04 | mnemoc | uhm... why I never got that? |
| 22:28.31 | hno | different .config perhaps? |
| 22:28.40 | hno | fixed. |
| 22:28.49 | mnemoc | sun4i_defconfig |
| 22:37.45 | hno | trivial fix available in my github tree. |
| 22:38.18 | hno | github pull requests broken at the moment. https://github.com/hno/linux-allwinner/commit/a7b338b79af107c056ef7dc46f3a9ed2a2608e7a |
| 22:47.46 | mnemoc | thanks, cherry-picked it to linux-3.3-sun4i and allwinner-v3.0-android |
| 22:48.10 | CIA-91 | rhombus-tech: hno master * r884922b67704 /allwinner_a10/kernel_compile.mdwn: Clean up notes on binary files |
| 22:49.07 | CIA-91 | rhombus-tech: hno master * rc7899df1bf24 /allwinner_a10/kernel_compile.mdwn: |
| 22:49.19 | mnemoc | fyi I made a non-android 2.6.37 branch with the "core" set of commits in linux-3.3-sun4i straight from allwinner-2.6.36-android and didn't work :-/ |
| 22:49.30 | mnemoc | allwinner-2.6.36-android does work |
| 22:49.37 | hno | ok. |
| 22:49.52 | mnemoc | I'll make the same but using 2.6.36 |
| 22:50.12 | hno | probably something stupid. |
| 22:51.57 | hno | did you load a sys_config1.bin? |
| 22:52.24 | mnemoc | Starting kernel ... |
| 22:52.27 | mnemoc | err |
| 22:52.29 | mnemoc | only script.bin |
| 22:52.47 | mnemoc | my paste button is also the wheel of the mouse :-/ |
| 22:53.10 | hno | script.bin is sys_config1.bin I think. |
| 22:53.34 | mnemoc | it was good enough to make allwinner-2.6.36-android talk |
| 23:11.14 | hno | how do you load script.bin? |
| 23:17.34 | hno | Mine prints some data |
| 23:18.09 | hno | oops.. wrong kernel. |
| 23:20.46 | hno | hangs. |
| 23:27.56 | hno | u-boot is missing source command. |
| 23:28.03 | hno | so no boot.scr support yet. |
| 23:28.09 | mnemoc | autosrc? |
| 23:28.46 | mnemoc | iirc that how source was called some time ago |
| 23:29.22 | hno | no autosrc either. |
| 23:31.45 | hno | only missing CONFIG_CMD_SOURCE from config. |
| 23:37.20 | mnemoc | good, amery/linux-2.6.36-sun4i doesn't boot either |
| 23:39.02 | hno | good or bad? |
| 23:39.16 | mnemoc | the only "core" element I left outside was 9aae53f907709010497ac6db7338786cb0c1fbe6 |
| 23:40.02 | mnemoc | good because I can tell now the problem was not "bad porting" of some commits |
| 23:40.14 | mnemoc | but in that chunk of evilness |
| 23:40.32 | mnemoc | which I simply couldn't port |
| 23:40.57 | mnemoc | + add r0, r4, #((0x3000000 + KERNEL_START) & 0xff000000) >> 18 |
| 23:41.00 | mnemoc | + add r6, r3, #48 << 20 |
| 23:41.04 | mnemoc | + str r6, [r0, #((0x3000000 + KERNEL_START) & 0x00f00000) >> 18] |
| 23:41.07 | mnemoc | ^--- this part |
| 23:41.41 | mnemoc | but I really need to go to sleep now :( |
| 23:45.36 | hno | Why did it not port? |
| 23:46.18 | mnemoc | because the asm of that area has changed a lot |
| 23:46.33 | mnemoc | http://dpaste.com/722478/ |
| 23:46.41 | mnemoc | in 2.6.36 |
| 23:47.40 | mnemoc | testing 2.6.36 with it applied now |
| 23:49.34 | mnemoc | meh |
| 23:49.40 | mnemoc | didn't make a difference |
| 23:50.29 | mnemoc | will continue at amery/linux-2.6.36-sun4i until it decides to talk |
| 23:50.35 | mnemoc | but tomorrow |
| 23:51.17 | mnemoc | good night |
| 23:51.27 | mnemoc | happy wife-bday |