Joseph Huber

I've gone through a bit of the code and fixed the errors from the incorrect bit shifting instructions from the src/simd.h file for Altivec. This fixed the infinite looping found...

I found another Altivec bug. The `simdi8_set(x)` macro wasn't calling the byte version of `vec_splats` because the cast to unsigned char wasn't being applied to the full expansion of the...

That should be correct. I checked the PPC disassembly just to make sure and it basically just places an array made from brace list in static memory and does a...

Is it possible for you to verify that the patch in https://reviews.llvm.org/D131639 solves your issue?

> Yeah that's the only usage I know of, was wondering if it work work regardless. One solution is to just use `_Bool` directly.

I'll update it real quick, let me know if it works because I didn't have any errors on my end.

If this were a simple workaround to make someone happy I'd be fine, but since this seems to be a rather annoying work-around only required for broken user code I'm...

As discussed in [D131639](https://reviews.llvm.org/D131639), we will not work around this. There is no guarantee that the system or toolchain will not include headers like `` so we shouldn't be expected...

Can you provide the original source input? It'll probably be easier to work with than 300,000 lines of post-preprocessed C++. LLVM-IR would also be very helpful. If you run with...

> ``` > (gdb) bt > #0 __GI_raise (sig=sig@entry=6) at ../sysdeps/unix/sysv/linux/raise.c:51 > #1 0x00007ffff693e7f1 in __GI_abort () at abort.c:79 > #2 0x00007ffff692e3fa in __assert_fail_base (fmt=0x7ffff6ab56c0 "%s%s%s:%u: %s%sAssertion `%s' failed.\n%n", >...