Crash when using "fill bed with instances"

[napter]

Version

2.3.0

Operating system type + version

Windows 64

3D printer brand / version + firmware version (if known)

N/A

Behavior

When using fill bed with instances with various STLs PrusaSlicer crashes. I've attached an STL as an example.

Steps to reproduct:
Import STL
Right click on object, click fill bed with instances
PrusaSlicer simply disappears

I have my arrange distance set at 0.44 - not sure if that contributes at all.

Project File (.3MF) where problem occurs

Cutter - 2021-01-11T153013.119.zip

[bubnikv]

What printer profile do you have selected? Is the crash reproducible? út 12. 1. 2021 v 1:38 odesílatel napter <[email protected]> napsal:

…

Version 2.3.0 Operating system type + version Windows 64 3D printer brand / version + firmware version (if known) N/A Behavior When using fill bed with instances with various STLs PrusaSlicer crashes. I've attached an STL as an example. Steps to reproduct: Import STL Right click on object, click fill bed with instances PrusaSlicer simply disappears I have my arrange distance set at 0.44 - not sure if that contributes at all. Project File (.3MF) where problem occurs Cutter - 2021-01-11T153013.119.zip <https://github.com/prusa3d/PrusaSlicer/files/5799082/Cutter.-.2021-01-11T153013.119.zip> — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#5749>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ABMPSI4GTBUE5PB4PGOIF5DSZOKZDANCNFSM4V6NOYJQ> .

[napter]

@bubnikv it happens no matter what printer profile you have selected. I've tried it with my Prusa and with the system Prusa mini profile.

Yes it is reproducible - it happens every time just follow my steps to reproduce.

[bubnikv]

I wonder whether it may be a duplicate of #5745

[napter]

In this case it slices without crashing so it doesn't seem the same - that bug said it crashes when slicing. Another odd thing to note - it will not crash if you set your arrange spacing to 0 and then fill bed with instances. But if arrange depth is larger than 0 and you select fill bed with instances, it crashes. This behavior has happened to me with multiple STLs - there is one sample attached to the issue to reproduce.

[lukasmatena]

it happens every time just follow my steps to reproduce.

It is not this simple, unfortunately. I for example could not reproduce it on my system (Ubuntu 16.04).

[bubnikv]

So what is the spacing that we should use to reproduce the crash? út 12. 1. 2021 v 9:47 odesílatel lukasmatena <[email protected]> napsal:

…

it happens every time just follow my steps to reproduce. It is not this simple, unfortunately. I for example could not reproduce it on my system (Ubuntu 16.04). — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#5749 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ABMPSI4KFNARHLL5AUSWWQTSZQEBZANCNFSM4V6NOYJQ> .

[napter]

@lukasmatena Just to confirm: Did you use my STL? Is your arrange distance greater than 0? (You can try 0.44). The crash is happening on 2.3.0 with the windows 64 build. I can't speak to the ubuntu release.

[lukasmatena]

@napter Yes, I tried your STL and 0.44 arrange distance. But I fully trust you that you see the crash. I just wanted to say that it may depend on many things. Yes, it could be platform-specific (or video driver specific or whatever). A backtrace would narrow it down, but I'm not sure what is the easiest way to get one on Windows.
What if you run the application with --sw-renderer command-line argument? Would it still crash?

[tamasmeszaros]

I can reproduce the crash on Manjaro Linux. The fix is on the way.

[foreachthing]

Maybe someone finds this useful:
STL: Cutter from above.
OS: Windows 10 x64
PS: 2.3.0

• Spacing (NO rotation!):
  1. 6 mm: OK.
  2. 3.1 mm: OK.
  3. 1.77 mm: OK.
  4. 0.88 mm: OK.
  5. 0.44 mm: OK.
• Spacing with rotation:
  1. 6 mm: OK.
  2. 3.1 mm: CRASH.
  3. 1.77 mm: OK.
  4. 0.88 mm: CRASH.
  5. 0.44 mm: CRASH.

EDIT: Added two log files (--loglevel=5):

1. 0.88 mm, rotation: PS-lv5---088-rot.log
2. 1.77 mm, rotation: PS-lv5--177-rot.log

Also added 3.1 mm to list above.

[napter]

@tamasmeszaros Any plan to merge this fix? I've been waiting for it since I have struggled to compile builds for windows, but it doesn't seem to be added to PrusaSlicer 2.3.1-rc.

[bubnikv]

2.3.1 is a minimal patch release. Your fix is in the current master, which is the upcoming 2.4.0