aboutsummaryrefslogtreecommitdiffstats
path: root/pkg/ifuzz/x86/decode.go
Commit message (Collapse)AuthorAgeFilesLines
* all: apply linter auto fixesTaras Madan2025-07-171-5/+7
| | | | ./tools/syz-env bin/golangci-lint run ./... --fix
* pkg/ifuzz: fix instruction decoding on x86Alexander Potapenko2024-04-301-0/+4
| | | | | | | | | | | | | | | | | | Decode() was only checking full opcode byte(s), whereas certain instructions are encoded in a way that some bits of the opcode are stored in the ModR/M byte. In particular, e.g. there is a variation of MUL encoded as: F7 /4 (which means the opcode byte is F7, and MODRM.reg is 4), and a variation of TEST encoded as: F7 /0 (opcode byte is also F7, and MODRM.reg is 0), which were previously indistinguishable (the decoder would incorrectly treat the MUL instruction as a TEST instruction if there were at least four extra bytes following it). Make sure to calculate and check the MODRM.reg value if insn.Reg is set to a non-negative value.
* pkg/ifuzz/iset: rename ifuzzimpl to isetDmitry Vyukov2020-11-211-8/+8
| | | | | | | | | ifuzzimpl is too lenghty and too clumsy on my taste (nm/vmimpl worked better b/c it's shorter and used less). I've tried to come up with something shorter and nicer. We could use "insn" as a common name for "instruction" in ifuzz, but it's a commonly used name already so would cause lots of conflicts. "iset" is somewhat descriptive, short and nice.
* pkg/ifuzz/x86: don't use X86 suffix for typesDmitry Vyukov2020-11-211-2/+2
| | | | x86.InsnSetX86 is excessive. Everything in x86 package is x86-ish already.
* pkg/ifuzz: invert ifuzz and ifuzzimplDmitry Vyukov2020-11-211-8/+9
| | | | | | | | | ifuzzimpl imports the public interface package ifuzz and prog package needs to import ifuzzimpl (implementation guts that nobody outside of ifuzz should care about). This is not right. Invert everything so that prog package only needs to import ifuzz and ifuzz imports ifuzzimpl.
* pkg/ifuzz: reorganize files to allow other architecturesAlexey Kardashevskiy2020-11-201-0/+236
At the moment ifuzz only generates x86 instructions. In order to support instruction fuzzing for others (ARM, POWERPC), some separation of the common and arch layers is needed. This adds 2 packages: 1. "x86" where x86 instruction generator goes to 2. "ifuzzimpl which contains some common code. The goal was to keep changes to the rand.go to the minimum. The next patch will use this when adding PPC64. This should cause no behavioural change. Signed-off-by: Alexey Kardashevskiy <aik@linux.ibm.com>