std: bad_alloc #2715
Comments
|
I renamed the .PrusaSlicer directory to something else, and start PrusaSlicer again and was able to slicer this object and larger ones without any problems. |
|
@maxminhtron Are you able to reproduce this bug yourself in any way? |
|
Yes. I changed the settings and would get the same message. It may be the setting size. |
|
It may be the setting size.
You mean scaling of the object?
po 5. 8. 2019 v 16:22 odesílatel maxminhtron <notifications@github.com>
napsal:
… Yes. I changed the settings and would get the same message. It may be the
setting size.
—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub
<#2715?email_source=notifications&email_token=ABMPSI5TDHGSZC4GEXXJDV3QDAZSXA5CNFSM4IJDFZL2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOD3R7EVQ#issuecomment-518255190>,
or mute the thread
<https://github.com/notifications/unsubscribe-auth/ABMPSIYZKPS6HKQXBI7GNUDQDAZSXANCNFSM4IJDFZLQ>
.
|
|
I am not sure. I tried different print settings, and would get that error message if "clock" is choosen. |
|
@maxminhtron Thanks, providing the config helped. The cause of the crash is (probably) that you set infill extrusion width to 0% (Print Settings->Advanced), which is not the expected way how to turn off infill (I assume that's what you intended to do). Apparently, PrusaSlicer cannot cope and I get an ASAN crash during slicing (and the bad_alloc in case ASAN is disabled). PrusaSlicer should either not have allowed you to set it that way, or correctly understand the intent. We should do something about it, I'm not sure what. What do you think @bubnikv ? Running in debug mode crashes even sooner (on loading the config), because of a failed assert in Flow::mm3_per_mm(). Variables at the time of the crash contain: The log output and backtrace at the ASAN crash in Release mode:
|
|
The problem of this config is |
|
This is quite a common problem, that one enters either a tiny number as a percent, or vice versa, enters an absolute value, but he things that the number was accepted as a percent. I know @YuSanka already implemented some safety checks in this regard. We should likely add more of these. Also the application should not crash, that is for sure. |
|
I did not changed the infil setting on purpose. It must have happen by accident. Thank you my friends. Now that the problem is resolved I will close the thread. |
|
The upcoming PrusaSlicer 2.1.0-alpha will produce some error message. Most likely you will get an error message just when you enter the value. If not, then you will get the following message when the configuration set is validated by the back end:
|
or too big). Fixes std: bad_alloc #2715
|
Yeah... thanks to this, I'm not able to print 0.85 extrusion width with my 0.4 mm nozzle. |
Same here. Trying to print the collapsing lightsaber: https://www.thingiverse.com/thing:3606120 The comments indicate that you can print the vase blade files using an 0.85 mm extrusion width on an 0.4 mm nozzle in vase mode. However PrusaSlicer 2.1.0-alpha warns that this is an excessive extrusion width and will not slice. Going to try PrusaSlicer 2.0. |
|
Crazy thought... Use a larger nozzle? |
|
Woooow... Why I wasn't thinking about it before??? 😑 I really don't want to change nozzle and calibrate the first layer every time when I need/want to print something with bigger extrusion width... |
We need to set some reasonable threshold. I don't know how you can print such wide extrusions. Maybe your nozzle has a large flat area? What limit do you recommend? What multiple of nozzle diameter? |
|
BTW leaving the parameters wide open potentially triggers some weird stuff inside slicer, producing weird bug reports to occupy us from writing useful stuff. That's why we added the check. |
|
I have a profile for "Vase mode" and there I have these parameters: @bubnikv I understand. I'm using stock MK3 with 0.4 nozzle. |
|
@Zemistr Have you measured your vase mode wall to see if it was actually 0.85 mm? I had done a quick test print using settings similar to yours, but my vase mode wall had only reached 0.78 mm. |
Good point... I will have a look and I will tell you 😉 |
|
@kazibole |
I answered there ;) |
|
I understand that there should be a limit but it could be lower. There is no problem printing an extrusion width equal to the layer height. I have tried things like printing 0.15 extrusion width with a layer height of 0.20 with a 0.20 nozzle and 1.2 extrusion multiplier and having the plate higher of 0.1 then the normal printing level (to have thiner width line then 0.20 before I found a 0.15 nozzle). |
|
The limit was changed from 200% to 300% of the nozzle diameter. |
I opened #3079 about this same issue, but it never received any attention. The max should be 400% or a little over that (~410%) as that is the recommended extrusion width for maximal inter-layer bonding. This is based on information from one of my engineering textbooks, but I can't remember which one right now. Would it be possible to bump it up from 300% to 400% so I can close my bug too? Pro-tip: you can leave the outer perimeter at 100% to get parts that print dramatically faster, significantly stronger, and noticeably faster. I'm sure it will eventually become the accepted norm once it trickles down to the average user. Edit: I have a completely stock 2.5s |
Version
PrusaSlicer 2.0.0+linux64-201905201652
Operating system type + version
Linux 4.15.0-54-generic x86_64
3D printer brand / version + firmware version (if known)
Prusa I3 MK3S 3D PRINTER
Behavior
std: bad_alloc
I imported a 22 kbyte stl file and when I tried to slice it I get the error message :
std: bad_alloc.
Once I clicked ok on the error message box, the object mesh appears.
I tried to export the gcode, but would also get the same error message.
The error message appears on the bottom of the slicer window.
below is the stl file zip.
Clock_by_A26_Remix_-_Minute_Hand.stl.zip
Is this a new feature request?
No
The text was updated successfully, but these errors were encountered: