Add native AOT runtime-deps and sdk variants #4822
Conversation
ghost
added
area-dockerfiles
lbussell
force-pushed
the
aot
branch
2 times, most recently
from
16ad9a5 to
4a505fe
Compare
src/runtime-deps/8.0/cbl-mariner2.0-distroless/arm64v8/Dockerfile
Outdated
Show resolved
Hide resolved
mthalman
left a comment
mthalman
left a comment
There was a problem hiding this comment.
My meta-review comment is "let's be sure to avoid including changes in this PR that we want in main" because AOT won't be getting merged to main.
src/runtime-deps/8.0/cbl-mariner2.0-distroless/arm64v8/Dockerfile
Outdated
Show resolved
Hide resolved
This reverts commit 3990cbd.
| ARG REPO=mcr.microsoft.com/dotnet/sdk | ||
| FROM $REPO:8.0.100-rc.1-jammy-arm64v8 | ||
|
|
||
| COPY <<EOF /etc/apt/sources.list.d/amd64.list |
There was a problem hiding this comment.
Do we have a general agreement that HereDocs should be used? If so what is the thought on when it should be used? I would expect consistent usage of it - e.g. in RUN instructions as well. Making consistent use is beyond the scope of this PR but given the introduction of it here we should establish an agreed upon direction and log follow-ups tracking this work.
There was a problem hiding this comment.
Thanks for bringing this up. I tried heredocs for RUN instructions but they don't play nice with line endings with Git for Windows' default behavior (checkout as CRLF and commit as LF). It creates some really strange errors.
Because of that I'd probably lean towards switching this command to using echo instead, to keep everything consistent.
There was a problem hiding this comment.
I agree that sounds best for now.
There was a problem hiding this comment.
I noticed the same thing. You have to do unix-style to use heredocs.
Line endings are the most annoying (at least obvious) difference between Windows and unix-style. More so than slash/backslash.
|
This PR and #4847 have significant overlap in the templates. I'd like to wait until that is merged first before merging this one so that the merge of that PR to main is less complicated. |
4 participants
Please see #4129 and #4804
I am currently working on the tests for these images.After discussion, tests will be included in a separate PR.