| --- Log opened Wed Jan 03 00:00:27 2018 | ||
| blueCmd | shorne_: I had a dream tonight that I got an email from you where you solved the copyright issue | 08:57 |
|---|---|---|
| blueCmd | it was pretty sureal | 08:58 |
| shorne_ | blueCmd: I think I might have been having the same dream :), its not solved yet. What was the solution? | 09:43 |
| blueCmd | shorne_: I woke up before I read it :( | 09:44 |
| shorne_ | mafm[m]: the 64-bit mode is just a spec as far as I know. no implementions. I am not sure I have really seen 64-bit in the spec either | 09:44 |
| nooga | haha, that's what happens in dreams :D | 09:44 |
| shorne_ | nooga: how did you generate this trace '[<c01991b4>] ude/asm-generic/dma-mapping-common.hmnao delete nr=%ld' in your gist? | 09:46 |
| shorne_ | Current kernel rc version has pretty reliable stack traces if you turn on KALLSYMS and FRAMEPOINTERs | 09:47 |
| nooga | i tried turning KALLSYMS on | 09:48 |
| shorne_ | or FRAME_POINTER | 09:48 |
| -!- shorne_ is now known as shorne | 09:48 | |
| nooga | also as complains about unaligned opcode when I try to compile the kernel with dbeug info | 09:48 |
| nooga | debug | 09:49 |
| shorne | I think I fixed that a while ago | 09:49 |
| shorne | which kernel version? | 09:49 |
| shorne | Try to use current upstream linux stable | 09:51 |
| nooga | 4.4.0 | 09:51 |
| shorne | The master branch here https://github.com/openrisc/linux is not so good | 09:51 |
| shorne | I have just now switched it to point to for-next instead of master | 09:52 |
| shorne | I should work to pull the feature branches out to another branch and keep them up to date | 09:53 |
| shorne | Things not upstream now include drivers for : ps2, spi, gpio and usb opencores hardware | 09:55 |
| shorne | 'for-next' branch is a current linux rc, with patched queued to to upstream during the next kernel merge window | 09:56 |
| nooga | I see | 09:56 |
| nooga | I'm cloning https://github.com/openrisc/linux.git right now | 09:57 |
| nooga | will try to build a more verbose version | 09:57 |
| shorne | you can clone : git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git | 09:58 |
| nooga | but I don't think it will help with the crash itself, I tried running jor1k kernel binaries and they crash in the same way at the same point | 09:58 |
| nooga | and that would indicate a bug in my emu | 09:59 |
| nooga | but it's almost impossible to debug without more details | 09:59 |
| nooga | ah, so the torvalds' kernel contains openrisc patches :O | 10:01 |
| shorne | but... interesting on your crash GPR02: ffff9aa0 | 10:01 |
| nooga | cool | 10:01 |
| nooga | yeah, that caught my eye too | 10:01 |
| shorne | so r2 is frame pointer | 10:02 |
| nooga | no idea where it's coming from yet | 10:02 |
| shorne | if you have frame pointers enabled... but its strange that its jumping to there | 10:03 |
| shorne | check instructuions where the strack trace points, maybe some load from r2->r9 and return | 10:04 |
| shorne | and backtrace from there | 10:05 |
| nooga | hm, I can try and figure out when r2 gets this value | 10:12 |
| nooga | There's something weird going on, looks as if format strings were printed out instead of the formatted text | 10:23 |
| nooga | and it crashes earlier than before https://gist.github.com/nooga/681965dc59f1a4f6207803a8908a4c31 | 10:24 |
| nooga | :D | 10:24 |
| nooga | (this is upstream kernel built with defconfig and KALLSYMS) | 10:25 |
| nooga | brb | 10:29 |
| nooga | and again this 0xffff.... pattern appears | 10:55 |
| nooga | I can't imagine how is it possible that I see format strings instead of the formatted text, it's almost as if printk was somehow broken | 13:40 |
| -!- [X-Scale] is now known as X-Scale | 20:29 | |
| --- Log closed Thu Jan 04 00:00:29 2018 | ||
Generated by irclog2html.py 2.15.2 by Marius Gedminas - find it at mg.pov.lt!