Hi Lance, Lance Fredrickson wrote,

Glad to see an aarch64 TLS fix! From what I remember, the TLS bug basically made all c++ applications pretty much unusable. Does this make c++ applications viable as well?

I think so. I never could reproduce the C++ issues in a good way, but this bug effects all applications using TLS including C++. best regards Waldemar

Lance

On Thu, Sep 18, 2025, 5:46 PM Waldemar Brodkorb <wbx@openadk.org> wrote:

Hi uClibc-ng hackers,

a new release is out.

Finally the aarch64 tls bug is fixed. Unfortunately it is not covered by any testcase of uClibc-ng-test.

Here is the git shortlog for the release:

Marius Melzer (1):       malloc-standard: Fix truncation problem in malloc

Waldemar Brodkorb (5):       Fix Sysvipc for ARM, AARCH64, RISCV64, KVX and m68k       remove uClibc specific gettext_printf       aarch64: do not force tls-model in CPU_CFLAGS, this isn't required       aarch64: respect r_addend for TLSDESC relocations       bump version for 1.0.55 release

yliu (5):       mips32: Preventing FORCE_FRAME_POINTER from being optimized away       mips32: Define __IPC_64 according to kernel version       time64: fixed msgctl/semctl/shmctl result errors for for MIPS32/ RISCV32       time64: fix *ctl functions in mips32eb/mips64       time64: fix *ctl functions in mips32eb

Thanks yliu for the SysVIPC fixes!

best regards  Waldemar _______________________________________________ devel mailing list -- devel@uclibc-ng.org To unsubscribe send an email to devel-leave@uclibc-ng.org