| | <<<Back 1 day (to 2020/06/17) | Fwd 1 day (to 2020/06/19)>>> | 20200618 |
| sebras | ator: if I run "./build/sanitize/mutool draw -F png -o /dev/null -B 10 -T 4 -s t ./pdfref17.pdf N-1" I rather consistently get "warning: assert: overwrite hash slot with different value!" and "warning: assert: overwrite hash slot with different value!" | 16:42.05 |
| | twice, but twice only. | 16:42.17 |
| | I'm not quite sure that we want this to happen. | 16:43.05 |
| | I haven't nailed down the reason for this, but it is related to the color convertor some how. | 16:43.23 |
| | I'd like your input before spending more time on it. it might be threading-related as I can't repeat it without -T 4 | 16:43.49 |
| hisacro | I built mupdf from source (debian) I cannot get mupdf-gl to work, it outputs blank window | 20:52.37 |
| | rebuilt with "make sanitize" here's log http://ix.io/2pxo | 20:52.54 |
| | did a search on the web, I believe it has to do with memory allocation but I not sure any pointers in right direction is much appreciated | 20:54.39 |
| malc_ | hisacro: i would guess that it's asan error you cite (where it fails to allocate shadow memory) | 20:57.23 |
| | ie its an error within asan itself not mupdf | 20:57.59 |
| hisacro | malc_, version of asan present http://ix.io/2pxw | 21:00.06 |
| malc_ | hisacro: i doubt that even if it worked it would have helped you to pinpoint the issue, what graphics card/library do you have there? | 21:01.10 |
| hisacro | malc_, Intel UHD Graphics 620 & NVIDIA GeForce 930MX maybe I'll try reinstalling nvidia driver | 21:05.43 |
| malc_ | hisacro: does mupdf | 21:06.30 |
| | hisacro: does mupdf-x11 work? is what i was trying to say | 21:06.52 |
| hisacro | it does | 21:07.05 |
| | without any errors | 21:07.11 |
| malc_ | and you need the gl version cause.... | 21:07.37 |
| hisacro | to see contents menu I suppose and have an interface? | 21:08.14 |
| malc_ | hisacro: okay, if you build things from source anyway might i suggest https://repo.or.cz/llpp.git ? | 21:09.49 |
| hisacro | malc_, thanks for the suggestion I didn't know that exist | 21:17.59 |
| malc_ | hisacro: i would be interested to know whether that works with this hybrid nvintl thing you've got there | 21:18.55 |
| hisacro | malc_, nope it didn't give segmentation fault | 21:41.06 |
| | s/give/gives | 21:41.15 |
| malc_ | hisacro: not entirely sure what you mean by it here | 21:42.32 |
| hisacro | llpp produces sementation fault when run | 21:44.25 |
| malc_ | hisacro: you just gave two contradicting statements, (1) nope it didn't give segmentation fault (2) llpp produces sementation fault when run | 21:46.54 |
| | so which is it? | 21:46.58 |
| hisacro | I was answering to this "i would be interested to know whether that works" no it didn't work | 21:48.02 |
| | it gives "segmentation fault" | 21:48.16 |
| malc_ | okay, not quite sure how to interpret the "nope it didn't give segmentation fault" but okay, can you run it under gdb and get a backtrace at segfault site? | 21:49.18 |
| hisacro | sorry, I left a comma | 21:49.58 |
| | malc_, http://ix.io/2pxH | 21:56.04 |
| malc_ | hisacro: it's ocaml runtime.. not llpp nor mupdf code | 21:59.02 |
| | nope sorry! | 21:59.44 |
| | it's in iris_gl which was built without symbols | 22:00.15 |
| | try this: LIBGL_ALWAYS_SOFTWARE=1 mupdf-gl file | 22:00.47 |
| hisacro | malc_, wow worked | 22:07.56 |
| | hearty thanks! | 22:08.08 |
| malc_ | hisacro: nah, you just instructed mesa to use software rendering, ie your nvidia and intel are not utilized | 22:09.31 |
| | which sucks | 22:09.35 |
| hisacro | I'm not using non-free drivers, I might load those modules and try later (does it make any change?) | 22:12.11 |
| malc_ | hisacro: using nouveau? also i have no idea what "those modules" mean | 22:13.35 |
| hisacro | yes, I mean I'll reboot and load non-free kernal modules | 22:15.30 |
| malc_ | hisacro: good luck | 22:16.06 |
| hisacro | http://ix.io/2pxO | 22:16.12 |
| malc_ | hisacro: i would suggest discussing this with ator, sebras or Robin_Watts_ | 22:17.37 |
| | anywho time to get some sleep | 22:17.48 |
| hisacro | thanks, appreciate your time figuring.. I'll try and get back later. | 22:18.12 |
| | <<<Back 1 day (to 2020/06/17) | Forward 1 day (to 2020/06/19)>>> | |
Log of #mupdf at irc.freenode.net.