fix: reduce debug level for rpc/evo.cpp to avoid clang crash

[UdjinM6]

Issue being fixed or feature implemented

The LLVM 18.1.8 compiler is crashing when generating debug information for rpc/evo.cpp:
https://github.com/dashpay/dash/actions/runs/19477052188/job/55739582899
https://github.com/UdjinM6/dash/actions/runs/19470819712/job/55717447032

2699740 from #6966 triggered a bug in clang - no crash after reverting it:
https://github.com/UdjinM6/dash/actions/runs/19495040944

What was done?

Implement a workaround.

How Has This Been Tested?

https://github.com/UdjinM6/dash/actions/runs/19495218479

Breaking Changes

Not exactly breaking but the tradeoff is that rpc/evo.cpp won't have debug symbols which is acceptable imo.

Checklist:

• I have performed a self-review of my own code
• I have commented my code, particularly in hard-to-understand areas
• I have added or updated relevant unit/integration/functional/e2e tests
• I have made corresponding changes to the documentation
• I have assigned this pull request to a milestone (for repository code-owners and collaborators only)

[github-actions]

✅ No Merge Conflicts Detected

This PR currently has no conflicts with other open PRs.

[coderabbitai]

Walkthrough

The pull request adds a targeted per-object CXXFLAGS override in the build configuration. When compiling with Clang, it sets -g0 for the rpc/evo.cpp object (rpc/libbitcoin_node_a-evo.$(OBJEXT)) to avoid an LLVM 18.1.8 debug-info generation crash. The change applies only to that single object and does not modify other build targets or exported/public declarations.

Estimated code review effort

🎯 1 (Trivial) | ⏱️ ~3–5 minutes

• Inspect src/Makefile.am to confirm the object target name rpc/libbitcoin_node_a-evo.$(OBJEXT) matches the actual generated object for rpc/evo.cpp.
• Verify the conditional for Clang is correct and that -g0 is the intended flag.
• Check for Makefile syntax correctness and ensure no unintended side effects on other targets.

Pre-merge checks and finishing touches

✅ Passed checks (3 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title clearly and specifically describes the main change: reducing debug level for rpc/evo.cpp to avoid a clang crash, which matches the core objective of the PR.
Description check	✅ Passed	The description is directly related to the changeset, explaining the compiler crash issue, the specific problem triggered by a previous commit, the workaround implemented, and testing evidence.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.

✨ Finishing touches

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Post copyable unit tests in a comment

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[coderabbitai]

Actionable comments posted: 0

🧹 Nitpick comments (1)

src/Makefile.am (1)

619-622: Per-object CXXFLAGS override is correct; consider clarifying scope of the workaround

The object name rpc/libbitcoin_node_a-evo.$(OBJEXT) and CXXFLAGS += -g0 usage look correct for targeting just rpc/evo.cpp, and appending -g0 should reliably strip debug info for that one TU while leaving other flags intact. This is a minimal, well-scoped workaround.

One thing to be aware of: this rule applies unconditionally, so all builds using this Makefile (not just Guix / Clang 18.1.8) will now compile rpc/evo.cpp without debug symbols. If that’s intentional, it might be worth making the comment explicit (e.g., mention that the workaround is applied globally and can be revisited once the upstream LLVM/Clang bug is fixed, maybe with a bug/issue reference).

Please double‑check that:

• The generated object file name from rpc/evo.cpp actually matches rpc/libbitcoin_node_a-evo.$(OBJEXT) in your build (e.g., by running make V=1 and inspecting the compile line).
• Applying -g0 unconditionally is acceptable for non-Guix, non-Clang-18 builds, or whether you’d prefer to gate this via a configure option or environment flag in a follow-up, to avoid scope creep here. Based on learnings.

📜 Review details

Configuration used: CodeRabbit UI

Review profile: CHILL

Plan: Pro

📥 Commits

Reviewing files that changed from the base of the PR and between 97bd9fa and 2227859.

📒 Files selected for processing (1)

• src/Makefile.am (1 hunks)

🧰 Additional context used

🧠 Learnings (1)

📓 Common learnings

Learnt from: knst
Repo: dashpay/dash PR: 6692
File: src/llmq/blockprocessor.cpp:217-224
Timestamp: 2025-08-19T14:57:31.801Z
Learning: In PR #6692, knst acknowledged a null pointer dereference issue in ProcessBlock() method where LookupBlockIndex may return nullptr but is passed to gsl::not_null, and created follow-up PR #6789 to address it, consistent with avoiding scope creep in performance-focused PRs.

Learnt from: UdjinM6
Repo: dashpay/dash PR: 6786
File: ci/test/04_install.sh:99-101
Timestamp: 2025-08-01T07:46:37.840Z
Learning: In backport PRs like #6786, UdjinM6 prefers to defer non-critical fixes (such as shell command expansion issues) to separate commits/PRs to maintain focus on the primary backport objectives, consistent with the project's pattern of avoiding scope creep.

Learnt from: kwvg
Repo: dashpay/dash PR: 6543
File: src/wallet/receive.cpp:240-251
Timestamp: 2025-02-06T14:34:30.466Z
Learning: Pull request #6543 is focused on move-only changes and refactoring, specifically backporting from Bitcoin. Behavior changes should be proposed in separate PRs.

⏰ Context from checks skipped due to timeout of 90000ms. You can increase the timeout in your CodeRabbit configuration to a maximum of 15 minutes (900000ms). (10)

• GitHub Check: linux64_tsan-build / Build source
• GitHub Check: arm-linux-build / Build source
• GitHub Check: mac-build / Build source
• GitHub Check: linux64_nowallet-build / Build source
• GitHub Check: linux64_fuzz-build / Build source
• GitHub Check: linux64_ubsan-build / Build source
• GitHub Check: linux64-build / Build source
• GitHub Check: linux64_sqlite-build / Build source
• GitHub Check: win64-build / Build source
• GitHub Check: Lint / Run linters

[UdjinM6]

@coderabbitai how do I only apply this fix when using Clang?

[coderabbitai]

Tip

For best results, initiate chat on the files or code changes.

🧩 Analysis chain

🏁 Script executed:

#!/bin/bash
# Check the current change in src/Makefile.am
cat src/Makefile.am | grep -A5 -B5 "rpc/libbitcoin_node_a-evo"

Length of output: 445

---

🏁 Script executed:

#!/bin/bash
# Look for existing compiler detection patterns in configure.ac or Makefile files
rg -n "CLANG|clang\+\+|__clang__|CXX.*clang" configure.ac src/Makefile.am build-aux/ || echo "No existing Clang detection found"

Length of output: 1644

---

🏁 Script executed:

#!/bin/bash
# Check if there are any existing conditional CXXFLAGS based on compiler
rg -n "if.*CXX|ifeq.*CXX|ifneq.*compiler" src/Makefile.am Makefile.am configure.ac | head -20

Length of output: 784

---

🏁 Script executed:

#!/bin/bash
# Look for how the build system detects or handles different compilers
cat configure.ac | grep -i "compiler\|clang\|gcc" | head -30

Length of output: 2250

---

@UdjinM6, there are two main approaches to conditionally apply the -g0 flag only when using Clang:

Option 1: Detect in configure.ac (Recommended)

Add compiler detection in configure.ac:

AC_MSG_CHECKING([if compiler is clang])
AC_COMPILE_IFELSE([AC_LANG_SOURCE([[
  #ifndef __clang__
  #error "not clang"
  #endif
]])], [is_clang=yes], [is_clang=no])
AC_MSG_RESULT([$is_clang])
AM_CONDITIONAL([IS_CLANG], [test x$is_clang = xyes])

Then in src/Makefile.am, wrap the workaround:

# Workaround for LLVM 18.1.8 bug: reduce debug level for rpc/evo.cpp to avoid crash
# in debug info generation for complex std::function template
if IS_CLANG
rpc/libbitcoin_node_a-evo.$(OBJEXT): CXXFLAGS += -g0
endif

After modifying configure.ac, you'll need to regenerate the build system with ./autogen.sh.

Option 2: Detect in Makefile.am

Add a check directly in src/Makefile.am before the per-object rule:

# Workaround for LLVM 18.1.8 bug: reduce debug level for rpc/evo.cpp to avoid crash
# in debug info generation for complex std::function template
ifeq ($(shell $(CXX) -dM -E -x c++ /dev/null 2>/dev/null | grep -q __clang__ && echo yes),yes)
rpc/libbitcoin_node_a-evo.$(OBJEXT): CXXFLAGS += -g0
endif

Option 1 is cleaner and more consistent with the existing autotools patterns in your codebase (I can see similar compiler detection in configure.ac). Option 2 is simpler if you want to avoid touching configure.ac, but runs the detection test every time make is invoked.

Which approach would you prefer?

---

🧠 Learnings used

Learnt from: kwvg
Repo: dashpay/dash PR: 6529
File: src/rpc/governance.cpp:1074-1089
Timestamp: 2025-02-14T15:15:58.165Z
Learning: Code blocks marked with `// clang-format off` and `// clang-format on` directives should be excluded from clang-format suggestions as they are intentionally formatted manually for better readability.

---

Learnt from: kwvg
Repo: dashpay/dash PR: 6729
File: src/rpc/evo.cpp:1273-1273
Timestamp: 2025-07-09T15:05:36.250Z
Learning: When clang-format suggestions significantly harm readability (like splitting logical parameter groups across multiple lines), it's acceptable to use `// clang-format off` and `// clang-format on` directives to exclude the problematic section from automatic formatting, prioritizing code readability over strict line length compliance.

---

Learnt from: knst
Repo: dashpay/dash PR: 6692
File: src/llmq/blockprocessor.cpp:217-224
Timestamp: 2025-08-19T14:57:31.801Z
Learning: In PR #6692, knst acknowledged a null pointer dereference issue in ProcessBlock() method where LookupBlockIndex may return nullptr but is passed to gsl::not_null, and created follow-up PR #6789 to address it, consistent with avoiding scope creep in performance-focused PRs.

---

Learnt from: UdjinM6
Repo: dashpay/dash PR: 6786
File: ci/test/04_install.sh:99-101
Timestamp: 2025-08-01T07:46:37.840Z
Learning: In backport PRs like #6786, UdjinM6 prefers to defer non-critical fixes (such as shell command expansion issues) to separate commits/PRs to maintain focus on the primary backport objectives, consistent with the project's pattern of avoiding scope creep.

---

Learnt from: kwvg
Repo: dashpay/dash PR: 6543
File: src/wallet/receive.cpp:240-251
Timestamp: 2025-02-06T14:34:30.466Z
Learning: Pull request #6543 is focused on move-only changes and refactoring, specifically backporting from Bitcoin. Behavior changes should be proposed in separate PRs.

[coderabbitai]

Actionable comments posted: 0

🧹 Nitpick comments (1)

src/Makefile.am (1)

619-622: Effective workaround for the Clang compiler crash.

The implementation correctly applies -g0 only when Clang is detected, using a per-object CXXFLAGS override that follows Make conventions. The shell-based detection is straightforward and addresses the immediate issue.

Known tradeoffs:

• The shell command executes on every Make invocation (characteristic of the Makefile-based detection approach).
• Debug symbols are disabled for rpc/evo.cpp across all Clang versions, not just 18.1.8. This is conservative but ensures the crash is avoided.

Optional refinement: If minimizing the impact on other Clang versions is desired, you could add version-specific detection:

rpc/libbitcoin_node_a-evo.$(OBJEXT): CXXFLAGS += $(shell $(CXX) --version | grep -E 'clang version (18\.1\.[0-9]+|18\.1\.8)' -q && echo "-g0")

However, the current broad approach is simpler and safer for a workaround.

📜 Review details

Configuration used: CodeRabbit UI

Review profile: CHILL

Plan: Pro

📥 Commits

Reviewing files that changed from the base of the PR and between 2227859 and 387d5bf.

📒 Files selected for processing (1)

• src/Makefile.am (1 hunks)

🧰 Additional context used

🧠 Learnings (1)

📓 Common learnings

Learnt from: UdjinM6
Repo: dashpay/dash PR: 6786
File: ci/test/04_install.sh:99-101
Timestamp: 2025-08-01T07:46:37.840Z
Learning: In backport PRs like #6786, UdjinM6 prefers to defer non-critical fixes (such as shell command expansion issues) to separate commits/PRs to maintain focus on the primary backport objectives, consistent with the project's pattern of avoiding scope creep.

Learnt from: kwvg
Repo: dashpay/dash PR: 6543
File: src/wallet/receive.cpp:240-251
Timestamp: 2025-02-06T14:34:30.466Z
Learning: Pull request #6543 is focused on move-only changes and refactoring, specifically backporting from Bitcoin. Behavior changes should be proposed in separate PRs.

Learnt from: knst
Repo: dashpay/dash PR: 6692
File: src/llmq/blockprocessor.cpp:217-224
Timestamp: 2025-08-19T14:57:31.801Z
Learning: In PR #6692, knst acknowledged a null pointer dereference issue in ProcessBlock() method where LookupBlockIndex may return nullptr but is passed to gsl::not_null, and created follow-up PR #6789 to address it, consistent with avoiding scope creep in performance-focused PRs.

⏰ Context from checks skipped due to timeout of 90000ms. You can increase the timeout in your CodeRabbit configuration to a maximum of 15 minutes (900000ms). (9)

• GitHub Check: linux64_fuzz-build / Build source
• GitHub Check: linux64_sqlite-build / Build source
• GitHub Check: linux64-build / Build source
• GitHub Check: linux64_ubsan-build / Build source
• GitHub Check: arm-linux-build / Build source
• GitHub Check: win64-build / Build source
• GitHub Check: linux64_nowallet-build / Build source
• GitHub Check: mac-build / Build source
• GitHub Check: linux64_tsan-build / Build source

[PastaPastaPasta]

Strange; but utACK 387d5bf

[kwvg]

@UdjinM6 can you try #6988? It bumps to Clang/LLVM 19, if it resolves the crash, we should be able to avoid the workaround.

[UdjinM6]

#6988 doesn't have this issue it seems https://github.com/UdjinM6/dash/actions/runs/19511128262/job/55850873237 but is it ready? I think we could merge this and revert it in #6988. This way we could make sure we don't miss another clang build issue while waiting for #6988.

[kwvg]

@UdjinM6 the Clang 19 bump has been extracted to #6995, will post Guix build hashes soon

[UdjinM6]

Closing in fav of #6995