NEW: Add libnvjpeg2k and libnvtiff #28142
Conversation
|
Hi! This is the staged-recipes linter and your PR looks excellent! 🚀 |
|
Hi! This is the friendly automated conda-forge-linting service. I just wanted to let you know that I linted all conda-recipes in your PR ( |
recipes/nvjpeg2k/build.sh
Outdated
| ln -s ${PREFIX}/${targetsDir}/$j ${PREFIX}/$j | ||
|
|
||
| if [[ $j =~ \.so\. ]]; then | ||
| patchelf --set-rpath '$ORIGIN' --force-rpath ${PREFIX}/${targetsDir}/$j |
There was a problem hiding this comment.
|
|
Thanks Daniel! 🙏 Likely we need the same RPATH fixes that we have added for the CTK: conda-forge/cuda-feedstock#10 Perhaps the changes from this PR ( conda-forge/libnvjpeg-feedstock#8 ) would be a good starting point for fixing the issues in libraries If we have executables that need RPATH fixes, we can discuss those as well. Sometimes these differ depending on the layout |
|
I used the nvjpeg-feedstock as a template to create this recipe. Those fixes had already been merged before I started this recipe, so they are already incorporated. The same overlinking errors occur and are suppressed in the libnjpeg feedstock. This is why I have added the conda-forge.yml to this recipe which supressed that error. Unfortunately, staged-recipes is not a perfect replica of the end feedstock, so it is ignored for now. |
|
Hi! This is the friendly automated conda-forge-linting service. I wanted to let you know that I linted all conda-recipes in your PR ( Here's what I've got... For recipes/nvjpeg2k/meta.yaml:
For recipes/nvtiff/meta.yaml:
For recipes/nvtiff/meta.yaml:
|
|
@adibbley, please review |
There was a problem hiding this comment.
Is using the targetsDir layout necessary for these packages? This is generally only used by packages included in cuda. The rpms/debs for these libraries install to system standard locations, such as:
$ rpm -qpl libnvjpeg2k0-cuda-12-0.8.0.38-1.x86_64.rpm
warning: libnvjpeg2k0-cuda-12-0.8.0.38-1.x86_64.rpm: Header V4 RSA/SHA512 Signature, key ID d42d0685: NOKEY
/usr/lib64/libnvjpeg2k
/usr/lib64/libnvjpeg2k/12
/usr/lib64/libnvjpeg2k/12/libnvjpeg2k.so.0
/usr/lib64/libnvjpeg2k/12/libnvjpeg2k.so.0.8.0.38
/usr/share/licenses/libnvjpeg2k0-cuda-12
/usr/share/licenses/libnvjpeg2k0-cuda-12/LICENSE
Note: libnvjpeg2k/12 subdir is used as those formats support side by side installs of the cuda variants, which is not needed here.
There was a problem hiding this comment.
I assume by "cuda" you mean the "cuda toolkit"?
Is using the targetsDir layout necessary for these packages?
You tell me! I believe the goal is to be consistent with layout on other platforms, so this is the feedback I'm looking for.
|
When you have a moment, could you please add a os_version:
linux_64: cos7
linux_aarch64: cos7It looks like This will help up track which OS version these binaries support and help us update to new OS versions/images when ready |
|
Hi! This is the staged-recipes linter and I found some lint. It looks like some changes were made outside the If these changes are intentional (and you aren't submitting a recipe), please add a File-specific lints and/or hints:
|
carterbox
force-pushed
the
nvjpeg2k
branch
2 times, most recently
from
649517b
to
c1f444c
Compare
|
The builds will not succeed because nvtiff requires a later version of glibc than the containers in staged recipes provide. |
|
Hi! This is the staged-recipes linter and your PR looks excellent! 🚀 |
There was a problem hiding this comment.
Wonder if you can really use alma8 at this point ? My understanding is conda-forge/conda-forge.github.io#1941 and conda-forge/conda-forge-pinning-feedstock#6626 need to be merged first
There was a problem hiding this comment.
This config parameter doesn't work yet which is why the docker_image key is being used in the conda_build_config.yml. However, John wants it here so that when the new pinnings go live, there is no accidental building against to alma9.
Checklist
url) rather than a repo (e.g.git_url) is used in your recipe (see here for more details).Closes #28216
Closes #28217