Open psifertex opened 11 months ago
This is fixed in the next
and future v6
release:
./cstool -d ppc64be "7c 00 00 ac"
0 7c 00 00 ac dcbf 0, r0
ID: 8 (dcbf)
op_count: 1
operands[0].type: MEM
operands[0].mem.base: REG = 0
operands[0].mem.offset: REG = r0
operands[0].access: READ
We cannot give a date yet for the v6
release. But we highly recommend to use the next
branch if you can. It provides enormously better disassembly for PPC
, ARM
, AArch64
.
Also see the release guide.
Unfortunately, I don't have time currently to check it myself, but the instructions in: https://github.com/Vector35/binaryninja-api/issues/3983 https://github.com/Vector35/binaryninja-api/issues/3989 https://github.com/Vector35/binaryninja-api/issues/4023 https://github.com/Vector35/binaryninja-api/issues/3968 are probably fixed in next
as well.
By the way, if you are curious about details of the long-running auto-sync project that will be a part of Capstone 6.0, @Rot127 wrote a blog post about why, how, and what was done to modernize Capstone: Auto-Sync - Generating disassembler plugins.
From a slack report:
It's a very low impact bug since it's strictly visual, but for the PPC disassembly (which I believe rides on top of capstone?) the disassembly text for dcbf and dcbz instructions gets a bit mangled.
For example, the opcode 7c 00 00 ac should disassemble as dcbf 0, r0 but the disassembly text shows as dcbf , r0 Checking with capstone outside of binja, the output is correct, which leads me to believe this is an issue within binja itself.