5.0.x release suggestions

[XVilka]

Would be nice to have fixes for these in the patch release:

5.0.1

• macOS python wheel for 5.0.0rc4 claims to be universal2 but is actually just amd64 #2066
• [SuperH]: fix IMM format specifier in test #2080
• Update Python bindings and test them in CI #2086
• Add Python bindings for WASM #2095
• Add Python bindings for SH #2096
• Update Python binding constants #2097
• Fixing TriCore disasm instructions #2088
• [ARM] Fix VFP feature check #2090
• Sync Python bindings for x86, m68k, and mos65xx #2100
• Restore the ARM register naming from v4. #2108
• allow absolute CMAKE_INSTALL_*DIR #2134

5.0.2

• Universal wheel for macOS contains only x86_64 build #2154
• 5.0.1 has more broken 'NOREGNAME' syntax on ARM32. #2145
• 5.0.1 has broken ARM operand information (in Python 2) #2144
• 5.0.0 & 5.0.1 do not install in python 3 on macOS (cannot load dynamic library) #2147
• [x86] regression on rcl with capstone v5 #2079
• cstool -d x64 "4c 85 7d 30" doesn't give the correct register reads after sync with LLVM 7.0.1 #2128
• cstool has dependency in static library but there is logical inconsistency in options #2025
• Deprecated modules #2005
• Using CAPSTONE_X86_REDUCE breaks the build #1955 (if it still happens)
• TMS320-c64x disassembler differs when using CS_OPT_DETAIL ON or OFF #1893
• fix compile errors in capstone_static - unresolved external symbols #2148
• Windows precompiled binaries #2153
• Publish news on http://www.capstone-engine.org/ for the 5.0.2 release

After they are fixed in the next branch, fixes could be cherry-picked to the v5 branch one by one for the release.

[XVilka]

There are also many PPC regressions found by @wargio but they will be fixed in the #2013 so it doesn't make sense trying to fix them in the patch release.

[XVilka]

@kabeor you need to cherry-pick corresponding commits from next to v5: https://github.com/capstone-engine/capstone/tree/v5

[kabeor]

Well looks like there are some misunderstood. v5 should merge something only if we are ready to release a version. so I think we needn't cherry-pick.

[XVilka]

@kabeor but you already cherry-picked some of these fixes - SH and ARM VFP fix. Idea is to cherry-pick these fixes and tag 5.0.1

[kabeor]

@XVilka Aha @aquynh did this. I should talk with him about this tmw😂

[XVilka]

Looks like only x86 regression and deprecated modules left in this list, the rest was merged

[wargio]

also the ppc regressions

[XVilka]

@wargio there are way too many, not sure if it makes sense to devote time instead of investing time in auto-sync PPC

[Rot127]

@kabeor Could we add a note to the 5.0.1 release, that the v6 release will bring some big changes (possibly linking to #2124)? In this case people have the chance to decide if they might want to skip v5 altogether.