farmhash intentionally (?!?!) produces different results in debug builds than non-debug builds

[jherico]

smhasher/farmhash-c.c

Line 287 in 1099e2c

static inline uint32_t debug_tweak32(uint32_t x) {

This drove me nuts when I was trying to debug a downstream library. Testing on Ubuntu vs Windows and Release vs Debug produced 4 different results (because the endian test was ALSO broken on Windows) and made me think I was taking crazy pills.

There's no commentary on why the code does this and it's part of the original commit for the file. This verges on gas-lighting. @uxcn what the actual hell?

[rurban]

I think this problem deserves a notice. Like "Instable on DEBUG"

[uxcn]

The code is strictly meant to be a faithful translation of Google's C++ implementation, to C (with some intrinsics replaced) to increase portability. I genuinely couldn't tell anyone why the algorithms are written to generate different results based on NDEBUG.

…

-Jason

On Sat, Apr 25, 2020, 10:00 PM Reini Urban ***@***.***> wrote: I think this problem deserves a notice. Like "Instable on DEBUG" — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#124 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAYALHV2PUKQRQNE4O2HOF3ROO5XLANCNFSM4MQ4HFEQ> .