fix(ci): compile rust docs in release and deny warnings

[giladchase]

• No reason to run cargo on release in the ci, and this busts cache
• Cargo doc doesn't pick up the usual place we pass -Dwarnings, which is
  RUSTFLAGS, it needs a special flag RUSTDOCFLAGS (it cannot reference
  the usual RUSTFLAGS so we have to duplicate)

[reviewable-StarkWare]

This change is 

[github-actions]

Artifacts upload workflows:

• Started at 2024-12-19T13:16:38Z
• Started at 2024-12-19T13:25:28Z
• Started at 2024-12-19T13:34:11Z
• Started at 2024-12-20T09:52:13Z
• Started at 2024-12-20T10:04:51Z
• Started at 2024-12-20T11:53:40Z
• Started at 2024-12-20T11:55:56Z
• Started at 2024-12-20T12:28:11Z
• Started at 2024-12-20T12:51:23Z
• Started at 2024-12-22T08:15:34Z
• Started at 2024-12-22T08:21:00Z
• Started at 2024-12-25T10:59:46Z
• Started at 2024-12-25T11:01:29Z
• Started at 2024-12-29T19:30:14Z

[dorimedini-starkware]

Reviewed 5 of 5 files at r1, all commit messages.
Reviewable status: all files reviewed, 2 unresolved discussions (waiting on @giladchase)

---

.github/workflows/main_nightly.yml line 66 at r1 (raw file):

        - name: Build Documentation
          run: cargo doc --all --no-deps

the command in run_tests.py is cargo doc -r --document-private-items --no-deps (disregarding the --workspace vs. -p crate list).
why not just remove the -r and keep it in run_tests.py`?

Code quote:

 run: cargo doc --all --no-deps

---

.github/workflows/main_nightly.yml line 69 at r1 (raw file):

          env:
            # Note that cargo doc requires RUSTDOCFLAGS, not RUSTFLAGS.
            RUSTDOCFLAGS: "-D warnings -C link-arg=-fuse-ld=lld"

won't this work?

Suggestion:

RUSTDOCFLAGS: ${{ env.RUSTFLAGS }}

[giladchase]

Reviewable status: 4 of 5 files reviewed, 2 unresolved discussions (waiting on @dorimedini-starkware)

---

.github/workflows/main_nightly.yml line 66 at r1 (raw file):

Previously, dorimedini-starkware wrote…

the command in run_tests.py is cargo doc -r --document-private-items --no-deps (disregarding the --workspace vs. -p crate list).
why not just remove the -r and keep it in run_tests.py`?

Whoops, added.

RE run_tests, you mean remove release so we can keep it in the PR flow?
I'm not sure why release was there in the first place, Ill reopen the PR as draft then and bring it up in the infra sync.

Regardless, I suggest moving it to nightly unless someone is actually looking at them on a regular basis / when reviewing PRs.
I see other ppl only generating them in a "release" workflow, and then I assume fixing all the issues on the spot.
The only issues that arise are broken links in the docs, which are trivial to fix all at once before release.

---

.github/workflows/main_nightly.yml line 69 at r1 (raw file):

Previously, dorimedini-starkware wrote…

won't this work?

Ohh nice.
Done

[dorimedini-starkware]

Reviewed 1 of 1 files at r2, all commit messages.
Reviewable status: all files reviewed, 1 unresolved discussion (waiting on @giladchase)

---

.github/workflows/main_nightly.yml line 66 at r1 (raw file):

Regardless, I suggest moving it to nightly unless someone is actually looking at them on a regular basis / when reviewing PRs.

I'm saying: if it doesn't slow down the CI (i.e. we get a cache hit on compilation) it's nice to fix doc issues in the PR that introduces them

---

.github/workflows/main_nightly.yml line 69 at r1 (raw file):

Previously, giladchase wrote…

Ohh nice.
Done

I'm actually asking if this works or not 😅

[dorimedini-starkware]

Reviewable status: all files reviewed, 2 unresolved discussions (waiting on @giladchase)

---

a discussion (no related file):
rebase over main-v0.13.4

[dorimedini-starkware]

Reviewed 3 of 3 files at r3, 1 of 1 files at r4, all commit messages.
Reviewable status: all files reviewed, 1 unresolved discussion (waiting on @giladchase)

[giladchase]

Reviewable status: 5 of 6 files reviewed, 1 unresolved discussion (waiting on @dorimedini-starkware)

---

a discussion (no related file):

Previously, dorimedini-starkware wrote…

rebase over main-v0.13.4

Done.

---

.github/workflows/main_nightly.yml line 66 at r1 (raw file):

Previously, dorimedini-starkware wrote…

Regardless, I suggest moving it to nightly unless someone is actually looking at them on a regular basis / when reviewing PRs.

I'm saying: if it doesn't slow down the CI (i.e. we get a cache hit on compilation) it's nice to fix doc issues in the PR that introduces them

Added in run_tests, file change is reverted.

---

.github/workflows/main_nightly.yml line 69 at r1 (raw file):

Previously, dorimedini-starkware wrote…

I'm actually asking if this works or not 😅

Stale, file change reverted.

[dorimedini-starkware]

Reviewed 1 of 1 files at r5, all commit messages.
Reviewable status: complete! all files reviewed, all discussions resolved (waiting on @giladchase)