On Thu, 28 Sep 2023 11:20:52 +1000 David Gibson <david(a)gibson.dropbear.id.au> wrote:While working on unifying the hashing for the flow table, I noticed some awkwardness in the siphash functions. While looking into that I noticed some bugs. So.. here we are. Changes since v1: * Don't accidentally increase the alignment of union inany_addr David Gibson (10): siphash: Make siphash functions consistently return 64-bit results siphash: Make sip round calculations an inline function rather than macro siphash: Add siphash_feed() helper siphash: Clean up hash finalisation with posthash_final() function siphash: Fix bug in state initialisation siphash: Use more hygienic state initialiser siphash: Use specific structure for internal state siphash: Make internal helpers public siphash, checksum: Move TBAA explanation to checksum.c siphash: Use incremental rather than all-at-once siphash functionsApplied, thanks. -- Stefano