More AMD Zen 4 Compiler Code Merged For GCC 13

Following the recent Zen 4 tuning patches that were merged to GCC 13 (Git) just ahead of Christmas, today an AMD patch adding the Zen 4 automatons have been merged ahead of this next open-source compiler release.

The znver4 target being introduced in GCC 13 continues to be revised ahead of the GCC 13.1 stable release due out in March~April.

The latest addition is adding the Zen 4 modeling automatons for the Zen decoders, integer execution pipes, AGU pipes, branch, floating point execution and fp store units. This is the znver4 scheduling/modeling information to help the GCC compiler make accurate decisions to benefit the execution on Zen 4 processors. This is also the first time these automata units and reservation information is being added completely separate from the "znver" entries that have built up in GCC and shared across Zen 1/2/3.

See this commit for more details on this latest AMD work for the GCC 13 compiler to benefit the AMD Ryzen 7000 series and EPYC 9004 "Genoa" processors. It's better late than never but still unfortunate that AMD continues to lag behind Intel in their punctuality for upstreaming new processor optimizations into the GNU Compiler Collection.