llvm/llvm-project

[AArch64] Add support for MSVC-style mangling for SVE

Closed

#196,170 opened on May 6, 2026

View on GitHubContributor guide
 (5 comments) (0 reactions) (0 assignees)C++ (10,782 forks)batch import
backend:AArch64clang:codegengood first issue

Repository metrics

Stars
 (26,378 stars)
PR merge metrics
 (Avg merge 1d 2h) (1,000 merged PRs in 30d)

Description

The latest MSVC release candidate has support for SVE, and with it a set of new SVE mangling patterns. We should mangle like MSVC instead of using the manglings we invented before (in #141887).

Example code:

#include <arm_sve.h>
void f(svint32_t x) {}

Latest MSVC produces the mangled symbol ?f@@YAX$$_CD@Z.

The relevant code is in MicrosoftCXXNameMangler::mangleType; replace the existing code which does #include "clang/Basic/AArch64ACLETypes.def".

(As a followup, we should also add demangling support in llvm/lib/Demangle/MicrosoftDemangle.cpp.)

Contributor guide

Research direction
Examine MicrosoftCXXNameMangler::mangleType in clang/lib/AST/MicrosoftCXXNameMangler.cpp, where SVE types are handled via #include "clang/Basic/AArch64ACLETypes.def". Replace that logic to produce MSVC's new mangling pattern (e.g., ?f@@YAX$$ CD@Z for svint32 t). Then add demangling support in llvm/lib/Demangle/MicrosoftDemangle.cpp. Refer to the linked PR #141887 for context and ensure compatibility with MSVC's RC output.
Tech stack
cpp
Domain
tooling
Issue type
Feature
DifficultyEstimated implementation difficulty for a new contributor, from 1 for very small changes to 5 for expert-level work.
4
Estimated timeA rough time range for an experienced contributor to investigate, implement, test, and prepare a pull request.
Over 1 week
Activity statusHow available the issue appears right now: fresh, active, stale, blocked, or waiting on maintainer input.
Active
ClarityHow clearly the issue explains the expected change, acceptance criteria, and next step.
Clear
Prerequisites
C++ name manglingAArch64 SVEMSVC ABICompiler backend internals
Newbie friendlinessA 1-100 score estimating how approachable this issue is for first-time contributors.
10