mandrel: update to 23.0.1.2; update revdeps

[icp1994]

Testing the changes

• I tested the changes in this PR: briefly (built jet, included in PR)

Local build testing

• I built this PR locally for my native architecture: x86_64

cc @leahneukirchen

[icp1994]

I'm not sure the patch is necessary. The source file it patched was completely removed with it's parent dir. Found some similar code at substratevm/src/com.oracle.svm.hosted/src/com/oracle/svm/hosted/c/libc/HostedMuslLibC.java

[leahneukirchen]

The patch is needed if it links musl statically else (easy to verify).

Please also rebuild Babashka.

[icp1994]

The patch is needed if it links musl statically else (easy to verify).

Does that require a musl system to verify?

[leahneukirchen]

2023-07-16T11:06:37.4448375Z => babashka-1.3.181_1: running post-install hook: 06-strip-and-debug-pkgs ...
2023-07-16T11:06:37.8129728Z    Stripped static executable: /usr/bin/bb

We want this dynamically linked.

[leahneukirchen]

Something is wrong, bb and jet are still static binaries on x86_64-musl.

[icp1994]

I'm a bit stumped. There's a BABASHKA_STATIC envvar that's being exported in these templates which add's --static to the native-image build args. See Jet for example. If I patch that arg out from the compile scripts, I get:

=> jet-0.6.26_1: running post-install hook: 06-strip-and-debug-pkgs ...
   Stripped executable: /usr/bin/jet
=> ERROR: jet-0.6.26_1: non-PIE executable found in PIE build: /usr/bin/jet
=> ERROR: jet-0.6.26_1: post-install_06-strip-and-debug-pkgs: 'find ${PKGDESTDIR} -type f' exited with 1

[leahneukirchen]

Hmm, looks like the linker doesn't get passed fPIE or so.

[leahneukirchen]

You need to patch /mandrel-mandrel-23.0.0.0-Final/substratevm/src/com.oracle.svm.hosted/src/com/oracle/svm/hosted/c/libc/HostedMuslLibC.java getAdditionalLinkerOptions which hard-codes -no-pie.

[leahneukirchen]

ugh, I just now saw the TEXTREL comment there -.- perhaps we actually wanna static link. sorry.

[leahneukirchen]

Indeed, bb builds fine as PIE executable but segfaults on startup with musl.

So let's scrap that, sorry for the distraction.

[leahneukirchen]

Then this is good to go, I think!

[icp1994]

Do you want me to remove the musl.patch file again then? I don't think it's doing anything we need.

[leahneukirchen]

Okay, but please keep the line that non-static is broken on musl.