erectus has quit [Remote host closed the connection]
erectus has joined #openscad
erectus has quit [Remote host closed the connection]
erectus has joined #openscad
arebil has joined #openscad
erectus has quit [Remote host closed the connection]
erectus has joined #openscad
guso78k has joined #openscad
<guso78k>
kintel, would prefer here to have the previous tbb find solution.
guso78k9 has joined #openscad
<Zr40>
kintel: also for simpler examples. Differencing two cubes shows the same symptom
knielsen has quit [Ping timeout: 268 seconds]
knielsen has joined #openscad
guso78k9 has quit [Ping timeout: 250 seconds]
guso78k has quit [Ping timeout: 250 seconds]
califax has quit [Remote host closed the connection]
califax has joined #openscad
califax has quit [Remote host closed the connection]
califax has joined #openscad
fling has quit [Remote host closed the connection]
fling has joined #openscad
aiyion2 has quit [Remote host closed the connection]
aiyion2 has joined #openscad
aiyion2 has quit [Remote host closed the connection]
aiyion2 has joined #openscad
kow__ has joined #openscad
drkow has quit [Ping timeout: 260 seconds]
mmu_man has joined #openscad
teepee_ has joined #openscad
teepee has quit [Ping timeout: 255 seconds]
teepee_ is now known as teepee
califax has quit [Remote host closed the connection]
califax has joined #openscad
teepee has quit [Remote host closed the connection]
teepee has joined #openscad
guso78k has joined #openscad
fling has quit [Remote host closed the connection]
teepee has quit [Remote host closed the connection]
teepee has joined #openscad
fling has joined #openscad
<mmu_man>
anyone has an account on the OpenCASCADE Mantis? Want to ask them to ship a .pc file for pkg-config but too lazy to create an account just for that
<guso78k>
OpenCASCADE is a company name ? thought it is the name of the CadQuery kernel ??
<guso78k>
prefer previous solution because it worked in my cmake, this is the only reason. this TBB::tbb stuff is not understood and i always replace it manually ;)
<guso78k>
I have noticed that there is yet another version. maybe this works without any change :)
<kintel>
mmu_man sorry, was confusing with OBJ.
<guso78k>
mmu_man yes: i'll also try to compile it here
<mmu_man>
Project MESSAGE: VRML reader disabled because OpenCascade < v7.7
<guso78k>
kintel, thats interesting. I believed, its a permanent error and never even considered deleting cached files ...
<kintel>
guso78k I just discovered that last night :)
<guso78k>
https://bpa.st/6GXQ BUMMER. I was even curious, that 'apt' is installable her in my Fedora :')
<mmu_man>
kintel: ohhhhh 😍
<guso78k>
s/her/here/
<mmu_man>
love that APT recommends cargo when I tell it to install cargo :D
<mmu_man>
and then cargo downloads the whole of Internet… :-(
voxpelli has quit [Ping timeout: 256 seconds]
drkow has joined #openscad
<guso78k>
kintel, for my texture stuff I have " glUniform1f(renderer_shader.data.csg_rendering.texturefactor , 0.0);
<guso78k>
- glUniform1i(renderer_shader.data.csg_rendering.tex1 , 0); " in my code. With that included, when I press "F6" and then "F5" , openSCAD sometimes crashes deep somwhere in draw_tri -> memcpy
<guso78k>
any instant clue ?
kow__ has quit [Ping timeout: 268 seconds]
drkow has quit [Remote host closed the connection]
drfff has joined #openscad
drfff has quit [Remote host closed the connection]
drfff has joined #openscad
guso78k5 has joined #openscad
gimzmoe has quit [Ping timeout: 246 seconds]
guso78k5 has quit [Client Quit]
gwillen has quit [Ping timeout: 256 seconds]
<mmu_man>
kintel: wow, generated code is ugly but it works fine :-)=
gwillen has joined #openscad
snakedGT has joined #openscad
Virindi has quit [Ping timeout: 246 seconds]
kintel has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
Virindi has joined #openscad
snaked has quit [Ping timeout: 264 seconds]
voxpelli has joined #openscad
<mmu_man>
ideally I could run kicad as CLI to export the wrl from the Makefile but eh
<mmu_man>
KiKit has some export commands but only for 2D it seems
<guso78k>
my question is if by just adding glUniform1i and glUniform1f you have to adapt other metrics/allocation sizes/configurations in order not to risk a Segfault
<guso78k>
of course it can be result of other nasty things I do for my texture feature
voxpelli has quit [Ping timeout: 256 seconds]
<guso78k>
could this be an artefact of the "Voodoo warning ... " documented in QGLView.cc ?
voxpelli has joined #openscad
<mmu_man>
aw, kicad-cli v8 will support vrml export directly so I should be able to script that all with a makefile
<kintel>
guso78k I feel it's a simpler explanation - this is really a CPU memory crash, so the GPU is likely not involved
<kintel>
I haven't been able to reproduce, but if you're able to figure out why it crashes, e.g. what wrong data was passed to memcpy(), that would be very helpful
<kintel>
Your log seem to indicate a memory alignment issue
drfff has quit [Remote host closed the connection]
drfff has joined #openscad
<guso78k>
kintel i'd had a hard time to find out, why its crashing. the crash is far away from the code which triggers crash and the crash is i an area where i don't have influence nor I am aware how it should work correctly.
<guso78k>
My feeling is that somwhere , foreign data is overwritten with wrong content and well esablished code simply fails to work correctly then.
<kintel>
wrong content shouldn't crash memcpu though
<kintel>
Perhaps printf the arguments to memcpu may give a good hint
<kintel>
*memcpy
<guso78k>
how could i influence good/bad memory alignment by typing .cc code ?
<pca006132>
I think you should try memory sanitizer first...
<pca006132>
I don't think memmove will fail due to alignment issue
<pca006132>
the name is "__memmove_avx_unaligned_erms"
<guso78k>
yes, i can perform your tests, with printf and with the sanitizier. anybody has a quick reminder for the compile option ?
<pca006132>
-fsanitize=address
<guso78k>
is it a cmake option ?
<pca006132>
no, for cmake you probably need -DCMAKE_CXX_FLAGS="-fsanitize=address"
<guso78k>
great. i can use that
<pca006132>
remember to build with debug info as well...
<pca006132>
-DCMAKE_BUILD_TYPE=RelWithDebInfo
<guso78k>
:)
<guso78k>
will take some time to prepare
TylerTork has joined #openscad
<TylerTork>
I was surprised to find that the built-in editor, while it provides content assist for system functions, doesn't use the comments or module prototype to give any typeahead or popup help for user-written and library modules. There doesn't even seem to be a feature request for this in github. Am I missing something?
<guso78k>
kintel, when I press F6, and then F5, VertexArray.h: addPositionData, addNormalData and addColorData is beeng called with std::shared_ptr<IAttributeData> data, whereas data->toBytes() is nullptr
<guso78k>
I dont have any idea why this could happen. maybe OOM ?
<guso78k>
i use a very simple design and not actively any textures here
<pca006132>
are you using the current master?
<pca006132>
ah I guess it is a nullptr when the data is empty
<pca006132>
although it is still null after setting color, it works though, no idea :/
<kintel>
In openscad, use cmake -DSANITIZE_ADDRESS=1 to enable asan
<kintel>
guso78k Yes, I think adding empty data arrays to those functions is normal as the data is just empty containers to be filled in later
<kintel>
The VBO code is still pretty convoluted, even after I did significant cleanup. More work to be done.. It's complex because in OpenCSG mode, we use a single VBO to hold all data for an entire CSG product, which translates into many render calls using different subsets of the VBO.
<guso78k>
there is done some pointer arithetitc later so nullptr -4 = 0xffffffffffffffc which is a bad thing
<guso78k>
is it possible that std::make_shared<AttributeData<GLfloat, 3, GL_FLOAT>>()); does not initialize /allocate the array ptr ?
<guso78k>
here is the ptr arithmetic: const GLbyte *src = data->toBytes() + data->sizeInBytes() - data->sizeofAttribute();
<guso78k>
pca006132 nope, i am on my own branch
<kintel>
sizeInBytes should always be larger than sizeOfAttribute, so I think that's fine
<TylerTork>
No answers about content assist?
<kintel>
..but an underlying bug could trigger something odd. If you can intercept a crash and print the data, that would work
<kintel>
TylerTork It's probably a missing feature for which nobody opened a ticket yet
mmu_man has quit [Ping timeout: 256 seconds]
J24k84 has quit [Quit: Client closed]
J24k84 has joined #openscad
<guso78k>
kintel nope, this is not valid in my case, i have here: