fix: guard abi::__forced_unwind catch behind __GLIBCXX__

[jbachorik]

What does this PR do?:

Wraps the try { ... } catch (abi::__forced_unwind&) { ... } block in j9WallClock.cpp::timerLoop() with #if defined(__GLIBCXX__) so the file compiles on libc++ targets (macOS clang+libc++, Alpine clang+libc++).

Motivation:

abi::__forced_unwind is a libstdc++ extension — declared in libstdc++'s <cxxabi.h> regardless of which libc is in use. libc++ (LLVM's C++ stdlib, the default on macOS clang) does not declare it. Commit 25c702ba4 (#492) introduced the catch without any guard, breaking builds on libc++ targets:

j9WallClock.cpp:141:17: error: no type named '__forced_unwind' in namespace '__cxxabiv1'
} catch (abi::__forced_unwind&) {
         ~~~~~^

Why GLIBCXX and not GLIBC

The compilation constraint is "is the C++ stdlib libstdc++?", not "is libc glibc?". The two are independent:

Target	C++ stdlib	abi::__forced_unwind defined?	Catch fires?
Linux glibc + libstdc++	libstdc++	yes	yes (glibc pthread_cancel raises it)
Linux musl + libstdc++	libstdc++	yes	no (musl uses a non-exception cancellation path)
Linux musl + libc++	libc++	no	n/a
macOS clang + libc++	libc++	no	n/a

__GLIBCXX__ is the precise predicate — it's defined by libstdc++ regardless of libc, undefined under libc++. The catch is harmless dead code on musl/macOS+libstdc++ but compiles cleanly there.

The same pattern in libraryPatcher_linux.cpp is already inside #if defined(__linux__) so it compiles correctly. This change brings j9WallClock.cpp into line.

Additional Notes:

#include <cxxabi.h> is left unguarded — the header itself exists on libc++, it just doesn't declare the abi::__forced_unwind member. Including it is harmless.

OpenJ9 supports macOS, AIX, z/OS, Linux and Windows — but on every non-glibc platform J9's pthread_cancel uses a non-exception mechanism, so the catch is a no-op there regardless of whether it compiles. The guard is purely about compilation, not behaviour.

How to test the change?:

• macOS / arm64 (clang + libc++): ./gradlew :ddprof-lib:compileRelease — verified to succeed (Successfully compiled 59 files).
• Linux / glibc (libstdc++): existing forced_unwind_ut.cpp unit test still exercises the catch path; behaviour unchanged.

For Datadog employees:

• If this PR touches code that signs or publishes builds or packages, or handles credentials of any kind, I've requested a review from @DataDog/security-design-and-guidance.
• This PR doesn't touch any of that.
• JIRA: [JIRA-XXXX]

🤖 Generated with Claude Code

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 7583b8dffa

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[dd-octo-sts]

CI Test Results

Run: #25170124398 | Commit: cddde3c | Duration: 27m 12s (longest job)

✅ All 32 test jobs passed

Status Overview

JDK	glibc-aarch64/debug	glibc-amd64/debug	musl-aarch64/debug	musl-amd64/debug
8	-	✅	-	-
8-ibm	-	✅	-	-
8-j9	✅	✅	-	-
8-librca	-	-	✅	✅
8-orcl	-	✅	-	-
11	-	✅	-	-
11-j9	✅	✅	-	-
11-librca	-	-	✅	✅
17	✅	✅	-	-
17-graal	✅	✅	-	-
17-j9	✅	✅	-	-
17-librca	-	-	✅	✅
21	✅	✅	-	-
21-graal	✅	✅	-	-
21-librca	-	-	✅	✅
25	✅	✅	-	-
25-graal	✅	✅	-	-
25-librca	-	-	✅	✅

Legend: ✅ passed | ❌ failed | ⚪ skipped | 🚫 cancelled

Summary: Total: 32 | Passed: 32 | Failed: 0

---

Updated: 2026-04-30 18:28:42 UTC