In article <cajgzzooiovfba3zo19woy+t7p8r_iqjgrbgpfvvtugdtu0b...@mail.gmail.com>,
enh <e...@google.com> wrote:
>this was found by fuzzing the LLVM __cxa_demangle on an ld128 Android
>system using hwasan, but it turns out no to simply be a buffer
>overflow --- the results are just wrong. (which shows how much anyone
>uses ld128 in conjunction with %a!)
Thanks a lot for the heads up. NetBSD uses 2x64 bit words instead
of 4x32 bit words to represent long double so we avoided this bug
by chance (we don't have EXT_FRAC{H,L}MBITS. Nevertheless I will
commit something like it for completeness.
https://nxr.netbsd.org/xref/src/sys/sys/ieee754.h#126
Best,
christos
