LLVM 11.0-RC2 Released For This Widely-Used, Open-Source Compiler Stack

The second and possibly final release candidate of LLVM 11.0 is now available for testing.

LLVM 11.0 after being under development for a half-year is preparing to ship with build speed improvements around pre-compiled headers, AMD Radeon "Navi 2" support, C++20 improvements, usage of C17 by default if no other C standard is specified, parsing but no handling yet for the GNU "asm inline" C extension, Radeon GCN offload capabilities for OpenMP, load hardening mitigation work and SESES as the latest on the compiler-based mitigation front, support for new Arm CPUs, and much more as previously covered in our LLVM 11.0 feature overview.

Since LLVM 11.0-RC1 at the end of July, there have been more bug fixes pouring in. LLVM 11.0 has been running behind schedule for their release plans, but they hope to be able to ship in the coming weeks for LLVM 11.0 and the associated sub-projects like Clang 11.0.

As it stands now, most of the blocker bugs have been resolved but a few release blockers remain so far. LLVM release manager Hans Wennborg in the RC2 announcement relayed that there is hope "we can still wrap up fairly soon" while LLVM 11.0 is already open for development on the main Git branch.