Search found 164 matches

by movax20h
Tue Jun 23, 2020 9:07 am
Forum: Pending
Topic: [0.18.32] Crash on save: FluidBox::getFluidManager()
Replies: 5
Views: 1630

Re: [0.18.32] Crash on save: FluidBox::getFluidManager()

What makes you think the crashes are random? Correction: I am unaware why I am getting SIGSEGV errors. Sometimes autosaving works, sometimes autosave is the last event before the SIGSEGV. You should share the save to test it. Try decreasing autosave interval to 1 to capture something that is close ...
by movax20h
Tue Jun 23, 2020 2:18 am
Forum: Resolved Problems and Bugs
Topic: Minimal Server Latency of 6 ticks
Replies: 5
Views: 10393

Re: Minimal Server Latency of 6 ticks

I noticed the same issue. When using on the LAN or running multiple instances on the same machine connected to headless server, using loopback (127.0.0.1 / ::1), I see 7, or 8 ticks of latency. Sometimes it does jump higher, to 11, 13, even 19. All on an empty freshly generated map with 0 entities t...
by movax20h
Thu Jun 18, 2020 6:24 pm
Forum: Not a bug
Topic: [0.18.31] Searching production for Uranium shows things unrelated to Uranium
Replies: 11
Views: 2654

Re: [0.18.31] Searching production for Uranium shows things unrelated to Uranium

Levenshtein itself is only suitable for a fuzzy string comparison, not for a fuzzy substring search. None of the on-line (meaning that only the search pattern can be preprocessed, there's no prior indexing of the data to be searched) fuzzy search algorithms are really performance wonders. However, ...
by movax20h
Wed Jun 17, 2020 7:26 pm
Forum: Resolved Problems and Bugs
Topic: [wheybags] [0.18.4] Unoptimal code generation of Linux binary (too many redundant stores/loads)
Replies: 11
Views: 6020

Re: [wheybags] [0.18.4] Unoptimal code generation of Linux binary (too many redundant stores/loads)

The binary size is still somehow bigger on Linux, not sure why: factorio_0.18.32$ ls -al -rw-rw-r-- 1 user user 28299264 Jun 16 10:40 factorio.exe -rwxr-xr-x 1 user user 36932256 Jun 17 19:20 factorio.stripped -rwxr-xr-x 1 user user 194706448 Jun 16 08:50 factorio.orig Compared to previous windows b...
by movax20h
Wed Jun 17, 2020 7:16 pm
Forum: Not a bug
Topic: [0.18.31] Searching production for Uranium shows things unrelated to Uranium
Replies: 11
Views: 2654

Re: [0.18.31] Searching production for Uranium shows things unrelated to Uranium

Heh, i looked into why it happens: 44.916 Info StringMatcher.cpp:71: can also be used to manually connect and disconnect electric poles and power switches with [font=default-semibold][color=#80cef0]left mouse button[/color][/font]. | uranium | true 44.916 Info StringMatcher.cpp:71: copper cable | u...
by movax20h
Wed Jun 17, 2020 7:12 pm
Forum: 1 / 0 magic
Topic: [0.18.31] Ongoing unspecific crashes, generally related to Fluids or Electric Circuits
Replies: 6
Views: 1864

Re: [0.18.31] Ongoing unspecific crashes, generally related to Fluids or Electric Circuits

Thanks all, I can confirm that I have double checked all the BIOS settings and I have no overclocking (CPU or RAM) enabled. Then I guess, you need to find a way to make it crash reliably. Does it crash when you interact with the game (i.e. placing / removing an entity by you or bots)? If so, try to...
by movax20h
Wed Jun 17, 2020 7:08 pm
Forum: Resolved Problems and Bugs
Topic: [wheybags] [0.18.4] Unoptimal code generation of Linux binary (too many redundant stores/loads)
Replies: 11
Views: 6020

Re: [wheybags] [0.18.4] Unoptimal code generation of Linux binary (too many redundant stores/loads)

I wonder if gcc should just ignore this flag when it doesn't affect results. I talked to some GCC devs about it, and they kind of agree that -ffloat-store maybe could be ignored on x86_64 (and some other flags), but the thing is even on x86_64 one can still use x87 (because some people port old stu...
by movax20h
Wed Jun 17, 2020 6:27 pm
Forum: Resolved Problems and Bugs
Topic: [wheybags] [0.18.4] Unoptimal code generation of Linux binary (too many redundant stores/loads)
Replies: 11
Views: 6020

Re: [wheybags] [0.18.4] Unoptimal code generation of Linux binary (too many redundant stores/loads)

Thank you for the fix. Can't wait to try my new 35% faster factorio (35% if I did the math right). BTW, have the removal of -ffloat-store changed the result of floating point operations? Shortly after I reported this bug I did some fuzzing of gcc arithmetic expressions and I haven't found any insta...
by movax20h
Wed Jun 17, 2020 6:24 pm
Forum: Resolved Problems and Bugs
Topic: [wheybags] [0.18.4] Unoptimal code generation of Linux binary (too many redundant stores/loads)
Replies: 11
Views: 6020

Re: [wheybags] [0.18.4] Unoptimal code generation of Linux binary (too many redundant stores/loads)

I can confirm this is fixed now. Example: 000000000062b120 <_ZNK15GhostController16getMovementSpeedEv>: 62b120: 80 bf 98 00 00 00 00 cmpb $0x0,0x98(%rdi) 62b127: 66 0f ef c0 pxor %xmm0,%xmm0 62b12b: 74 37 je 62b164 <_ZNK15GhostController16getMovementSpeedEv+0x44> 62b12d: 48 8b 47 28 mov 0x28(%rdi),%...
by movax20h
Tue Jun 16, 2020 5:28 pm
Forum: 1 / 0 magic
Topic: [0.18.31] Ongoing unspecific crashes, generally related to Fluids or Electric Circuits
Replies: 6
Views: 1864

Re: [0.18.31] Ongoing unspecific crashes, generally related to Fluids or Electric Circuits

AFAIK AMD Ryzen 3600X should not be affected by the infamous (rare) Ryzen SMT issues under Linux, they only affected Ryzen 1800X manufactured before 25th week of 2017. To me the fact that it does crash in the same spot usually, is indicator of a bug, but I would suggest excluding a hardware fault fi...
by movax20h
Tue Jun 16, 2020 5:15 pm
Forum: Technical Help
Topic: [0.17.79] Demo crashes immediately (Ubuntu 20.04, Mesa driver 20.0.4, Intel Iris)
Replies: 10
Views: 1712

Re: [0.17.79] Demo crashes immediately

Hi majortal. Yes, as suspected this is a compiler and mesa issue. It only happens with iris driver and gcc 10 with some specific flags and link time optimization enabled. The issue should be fixed in Mesa 20.0.8 and 20.1.0, which were released about a week ago. I am pretty sure Ubuntu will backports...
by movax20h
Mon Jun 15, 2020 7:07 pm
Forum: 1 / 0 magic
Topic: [0.18.31] Ongoing unspecific crashes, generally related to Fluids or Electric Circuits
Replies: 6
Views: 1864

Re: [0.18.31] Ongoing unspecific crashes, generally related to Fluids or Electric Circuits

You have a number of mods installed. Would be good to reduce number of mods to easy troubleshooting or include a zip file with all mods in this save.
by movax20h
Mon Jun 15, 2020 6:57 pm
Forum: Technical Help
Topic: [0.17.79] Demo crashes immediately (Ubuntu 20.04, Mesa driver 20.0.4, Intel Iris)
Replies: 10
Views: 1712

Re: [0.17.79] Demo crashes immediately

Thanks for the report however: * When I test the demo it works perfectly * You seem to be the only person reporting having an issue * The stack trace failed to symbolize (typical on Linux machines) * The stack trace is 2 entries long (that typically means it crashed before it even gets to executing...
by movax20h
Sun Jun 14, 2020 8:22 pm
Forum: Resolved Problems and Bugs
Topic: [kovarex][0.18.31] Incorrectly sized train view dialog on 1920x1920 display
Replies: 4
Views: 4252

[kovarex][0.18.31] Incorrectly sized train view dialog on 1920x1920 display

This is with 0.18.31 from Steam on Linux. Screenshot at 2020-06-14 03-09-50.png This is a default sizing of this dialog on my system, and it has the right side cut off. I in fact never realized there are some extra options for train on the right size, because I never realized the dialog is cropped o...
by movax20h
Sat Feb 22, 2020 3:36 pm
Forum: General discussion
Topic: Parrallel processing in games & applications
Replies: 433
Views: 202751

Re: Parrallel processing in games & applications

Each entity would also have an old-state and next-state. In each phase the next-state is computed from old-state and at the end you switch the two for the next phase. I have toyed with this idea as a thought experiment ... and concluded it doesn't work out in practice. e.g. 2 mobs. Each has a threa...
by movax20h
Fri Feb 14, 2020 11:34 pm
Forum: Resolved Problems and Bugs
Topic: [0.18.6] IPv6 address in square brackets without port is claimed to be incorrect.
Replies: 16
Views: 4184

Re: [0.18.6] IPv6 address in square brackets without port is claimed to be incorrect.

The parsing is done by calling getaddrinfo() which is a function outside of our control. That is not true. getaddrinfo and getaddrinfo_a DOES NOT receive a single string containing both address and service. They are separate arguments, thus you must parse them to separate them out. My guess is you ...
by movax20h
Fri Feb 14, 2020 11:06 pm
Forum: Resolved Problems and Bugs
Topic: [0.18.6] IPv6 address in square brackets without port is claimed to be incorrect.
Replies: 16
Views: 4184

Re: [0.18.6] IPv6 address in square brackets without port is claimed to be incorrect.

My Win 10 test shows that my mods are not correct for the server. But it sees the server properly and communicates with it. Yeah, I just noticed your attempt in the logs. Not sure why would it say mod mismatch, as I don't run any mods. Maybe you are not running 0.18.6? Also, just to confirm. You di...
by movax20h
Fri Feb 14, 2020 9:15 pm
Forum: Resolved Problems and Bugs
Topic: [0.18.6] IPv6 address in square brackets without port is claimed to be incorrect.
Replies: 16
Views: 4184

Re: [0.18.6] IPv6 address in square brackets without port is claimed to be incorrect.

My Win 10 test shows that my mods are not correct for the server. But it sees the server properly and communicates with it. Yeah, I just noticed your attempt in the logs. Not sure why would it say mod mismatch, as I don't run any mods. Maybe you are not running 0.18.6? Also, just to confirm. You di...
by movax20h
Fri Feb 14, 2020 8:39 pm
Forum: Ideas and Suggestions
Topic: Post-game wishlist
Replies: 6
Views: 1725

Re: Post-game wishlist

Supplying the outposts I supplied defence outposts with 6 types of items (turret, wall, substation, repair pack, artillery missiles and construction bots). Some other items could also be useful. I packed all the stuff in the same carriage, using slot filters. I ran into following issues: I tried to...
by movax20h
Fri Feb 14, 2020 8:14 pm
Forum: Resolved Problems and Bugs
Topic: [0.18.6] Warning ChildProcessAgent.cpp:79: We were not waiting for that child
Replies: 12
Views: 2428

Re: [0.18.6] Warning ChildProcessAgent.cpp:79: We were not waiting for that child

According to my checks with proc/tasks/X/wchan, it is the main thread doing it, and somewhere in mainLoopStepHeadless, just after the sleep where it sleeps for 30 seconds. (gdb) #0 0x00007fb3f6791720 in nanosleep () from /lib/x86_64-linux-gnu/libc.so.6 #1 0x00007fb3f67bc874 in usleep () from /lib/x8...

Go to advanced search