something is weird about llvm submodule handling

[matthiaskrgr]

I just observed the following:
I did a git pull and it downloaded a bunch of llvm submodule data.
Afterwards, I did an x.py install, I'm not sure what happened with the llvm submodule but llvm did NOT get rebuilt, so probably nothing touched it?
Then, I did another x.py install afterwards, without touching anything on the repo and THAT did start rebuilding llvm again.

I'm having a gut feeling that we should have had rebuilt llvm already on the first x.py install but I don't know what the underlying issue is.

cfg:

[llvm]
download-ci-llvm = false
assertions = true
ninja = true
targets = "X86"
link-jobs = 16
cflags = "-march=native" # -D_GLIBCXX_DEBUG
cxxflags = "-march=native" # -D_GLIBCXX_DEBUG
ldflags = "-march=native" # -D_GLIBCXX_DEBUG
polly = false
[build]
dist-stage = 2
install-stage = 2
extended = true
tools = ["cargo", "clippy", "rustfmt", "analysis", "src", "miri"] # + "rust-demangler" if `profiler`
low-priority = true
[install]
prefix = "/home/gh-matthiaskrgr/.rustup/toolchains/local-debug-assertions"
sysconfdir = "etc"
[rust]
debug = true
debug-assertions = true
debug-assertions-std = true
overflow-checks = true
overflow-checks-std = true
debuginfo-level = 1
codegen-backends = ["llvm", "cranelift"]
backtrace-on-ice = true
verify-llvm-ir = true
lto = "thin"
[target.x86_64-unknown-linux-gnu]
[dist]
compression-formats = ["xz"]
```

[onur-ozkan]

Was the HEAD older than #122632 ?

[matthiaskrgr]

I guess that could have been the case 🤔

[onur-ozkan]

I guess that could have been the case 🤔

Then let's close the issue for now. If the problem recurs, we can reopen it.

[matthiaskrgr]

Just happened again 🫠
edit: as I updated form 8387315 to ed172db then did install , ctrl+c , install

[matthiaskrgr]

still an issue 😿

[onur-ozkan]

Hmm..

Just happened again 🫠 edit: as I updated form 8387315 to ed172db then did install , ctrl+c , install

I tried this with download-ci-llvm = false but it didn't happen.