Package Details: luxcorerender 2:2.7.beta1-2

Git Clone URL: https://aur.archlinux.org/luxcorerender.git (read-only, click to copy)
Package Base: luxcorerender
Description: Physically correct, unbiased rendering engine.
Upstream URL: https://www.luxcorerender.org/
Licenses: Apache
Provides: luxrays
Submitter: bartus
Maintainer: bartus (howetuft)
Last Packager: bartus
Votes: 15
Popularity: 0.000077
First Submitted: 2018-05-11 21:03 (UTC)
Last Updated: 2024-11-25 20:46 (UTC)

Dependencies (21)

Sources (15)

Pinned Comments

bartus commented on 2020-06-11 15:32 (UTC) (edited on 2020-08-22 09:39 (UTC) by bartus)

This package is also hosted on GitHub.
Use env vars to control build process:
  • DISABLE_OPENCL=1 to skip opencl kernel build (yields DISABLE_CUDA=1)
  • DISABLE_CUDA=1 to skip cuda kernel build.
Usage cases:
  • export DISABLE_CUDA=1 before build
  • DISABLE_CUDA=1 ~your-aur-helper~
  • makepkg DISABLE_CUDA=1
  • yay -S blender-2.8-git --mflags "DISABLE_CUDA=1"

bartus commented on 2019-04-10 11:42 (UTC)

Please report issues and patches to luxcorerender@github.com

Latest Comments

« First ‹ Previous 1 2 3 4 5 6 7 .. 9 Next › Last »

howetuft commented on 2023-08-21 05:16 (UTC)

You see me absolutely confused. I've just relaunched a makepkg -fis against the package, after having cleaned build dirs: everything worked fine for me. Again, my previous build in a clean chroot with devtools (extra-x86_64-build -- -I fmt9-9.1.0-5-x86_64.pkg.tar.xz -I pyside6-tools-wrappers-20230711-1-any.pkg.tar.xz) was also successful. So I cannot reproduce the issue (not denying it did not happen on your system, obviously). I would propose: let other users report the same anomaly if they encounter it - and, if someone is in this case, please pastebin build log.

DarkShadow44 commented on 2023-08-20 22:41 (UTC) (edited on 2023-08-20 22:41 (UTC) by DarkShadow44)

The first time I run makepkg is a fresh git clone from the AUR. That breaks.

But running makepkg again works for some weird reason.

Btw, command is "makepkg -fis" in both cases

howetuft commented on 2023-08-20 17:58 (UTC)

The first time you run makepkg, do you clean beforehand? I mean, do you clean/erase the previous build directory? Otherwise, there can be remnants from the previous build, especially in CMake files, that interfere with the new build...

DarkShadow44 commented on 2023-08-20 17:19 (UTC)

So, I just retried, and the results are weird. When I run makepkg, I get the same fmt error in spdlog. But when I run makepkg again without cleaning, it suddenly works...

Not sure why that is, but it's very reproducible.

howetuft commented on 2023-08-19 14:59 (UTC)

@DarkShadow44 I fixed fmt9 dependency, so this should not be an issue any more. I also slightly modified the last patch, to reinforce fmt9 include paths over fmt10 ones.

I rebuilt the package in a clean chroot (with devtools) and everything worked fine for me.

Could you please retry to build on your side?

howetuft commented on 2023-08-19 10:47 (UTC)

@DarkShadow44 OK, thank you. At least, it points out one thing: I forgot to add fmt9 in depends, in lieu of fmt :-( I shall quickly fix that. For the second issue, once fmt9 is installed, I'm a bit disconcerted: that is exactly what fmt9 should fix. I still have to investigate...

DarkShadow44 commented on 2023-08-19 10:21 (UTC)

@MarsSeed

Log: https://pastebin.com/vRYaLS64

CMake Error at samples/luxcoredemo/CMakeLists.txt:31 (find_package):
  Could not find a configuration file for package "fmt" that is compatible
  with requested version range "9.0.0...<10.0.0".

  The following configuration files were considered but not accepted:

    /usr/lib/cmake/fmt/fmt-config.cmake, version: 10.1.0
    /lib/cmake/fmt/fmt-config.cmake, version: 10.1.0

After installing fmt9: https://pastebin.com/vraUcwiY

/tmp/ramdisk/luxcorerender/src/LuxCore-luxcorerender_v2.6/src/luxcore/kernelcachefill.cpp:316:2:   required from here
/usr/include/spdlog/logger.h:374:75: error: cannot bind non-const lvalue reference of type ‘double&’ to an rvalue of type ‘double’
  374 |             fmt::vformat_to(fmt::appender(buf), fmt, fmt::make_format_args(std::forward<Args>(args)...));

howetuft commented on 2023-08-19 10:10 (UTC)

@DarkShadow Just to check something: Could you please manually install fmt9 from AUR and retry to build? Thanks!

MarsSeed commented on 2023-08-19 09:54 (UTC) (edited on 2023-08-19 09:54 (UTC) by MarsSeed)

@DarkShadow44, could you please share the build log on pastebin, or at least excerpts from that, with the initial configuration part and the part where it shows failures in a comment here?

@howetuft, I did not yet have a chance to try the build again due to other engagements. But I wi try again sometime in the next few days to see if it builds for me.

I see your point about the -git package. Originally I thought that that would be easier to fix, but now I understand that this stable one is more viable, as per your experience.

DarkShadow44 commented on 2023-08-19 09:43 (UTC)

Doesn't build, since fmt is at 10.0 now.

« First ‹ Previous 1 2 3 4 5 6 7 .. 9 Next › Last »