5152d0c9f7
Foundation for replacing fold's dead/abstract binary encoder with a real
SSA P-Code frontend. Verified root-cause gap: serializeBinaryIR() emits only
data-discovery nodes (sections/strings/imports/exports); the control-flow
BinaryExprKind values (BasicBlock/CallSite/...) are never constructed, so the
Z3-BB encoder's encodable filter is always empty and encodeBinaryBBs never
runs — hence 0 binary Z3 findings. The P-Code pipeline supersedes that path.
Phase 0 — docs/internal/ghidra-pcode-integration-research.md: P-Code opcode
core, the decompiler C++ embedding API (LoadImage/SleighArchitecture/Funcdata),
the CORE+DECCORE+SLEIGH source sets, no-bison finding, and pypcode as build
reference.
Phase 1 — vendored Ghidra decompiler at pinned commit (Ghidra_12.1_build,
7e89d94e34...) under third_party/ghidra-decompiler/ (cpp + AArch64 specs +
Apache-2.0 LICENSE + VENDOR.md + .gitignore for build artifacts). Validated by
hand on this clang-21/macOS toolchain:
- full SSA engine (CORE+DECCORE+SLEIGH, 80 files incl. heritage/funcdata/
action) compiles clean, -std=c++11 -O2, NO BFD, NO bison (generated
parsers are committed upstream; ruleparse isn't in the lib set);
- sleigh compiler builds; AArch64 .slaspec -> .sla compiles;
- tools/pcode_dump.cpp (custom LoadImage, no BFD) lifts the real arm64
sym._div_by_zero_direct to correct P-Code: `mov w9,#0 -> COPY(const 0)`,
`sdiv -> INT_SDIV(w8,w9)` — divisor zero-ness explicit in the data flow,
so the Phase-3 encoder proves div-by-zero trivially, multi-arch.
Not yet meson-wired (next step) and no fold-side IR/encoder yet (Phase 2-3);
this commit is the validated vendor + smoke tool.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
174 lines
7.9 KiB
C++
174 lines
7.9 KiB
C++
/* ###
|
|
* IP: GHIDRA
|
|
*
|
|
* Licensed under the Apache License, Version 2.0 (the "License");
|
|
* you may not use this file except in compliance with the License.
|
|
* You may obtain a copy of the License at
|
|
*
|
|
* http://www.apache.org/licenses/LICENSE-2.0
|
|
*
|
|
* Unless required by applicable law or agreed to in writing, software
|
|
* distributed under the License is distributed on an "AS IS" BASIS,
|
|
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
|
* See the License for the specific language governing permissions and
|
|
* limitations under the License.
|
|
*/
|
|
#ifndef __SLGHPATTERN_HH__
|
|
#define __SLGHPATTERN_HH__
|
|
|
|
#include "context.hh"
|
|
|
|
namespace ghidra {
|
|
|
|
// A mask/value pair viewed as two bitstreams
|
|
class PatternBlock {
|
|
int4 offset; // Offset to non-zero byte of mask
|
|
int4 nonzerosize; // Last byte(+1) containing nonzero mask
|
|
vector<uintm> maskvec; // Mask
|
|
vector<uintm> valvec; // Value
|
|
void normalize(void);
|
|
public:
|
|
PatternBlock(int4 off,uintm msk,uintm val);
|
|
PatternBlock(bool tf);
|
|
PatternBlock(const PatternBlock *a,const PatternBlock *b);
|
|
PatternBlock(vector<PatternBlock *> &list);
|
|
PatternBlock *commonSubPattern(const PatternBlock *b) const;
|
|
PatternBlock *intersect(const PatternBlock *b) const;
|
|
bool specializes(const PatternBlock *op2) const;
|
|
bool identical(const PatternBlock *op2) const;
|
|
PatternBlock *clone(void) const;
|
|
void shift(int4 sa) { offset += sa; normalize(); }
|
|
int4 getLength(void) const { return offset+nonzerosize; }
|
|
uintm getMask(int4 startbit,int4 size) const;
|
|
uintm getValue(int4 startbit,int4 size) const;
|
|
bool alwaysTrue(void) const { return (nonzerosize==0); }
|
|
bool alwaysFalse(void) const { return (nonzerosize==-1); }
|
|
bool isInstructionMatch(ParserWalker &walker) const;
|
|
bool isContextMatch(ParserWalker &walker) const;
|
|
void encode(Encoder &encoder) const;
|
|
void decode(Decoder &decoder);
|
|
};
|
|
|
|
class DisjointPattern;
|
|
class Pattern {
|
|
public:
|
|
virtual ~Pattern(void) {}
|
|
virtual Pattern *simplifyClone(void) const=0;
|
|
virtual void shiftInstruction(int4 sa)=0;
|
|
virtual Pattern *doOr(const Pattern *b,int4 sa) const=0;
|
|
virtual Pattern *doAnd(const Pattern *b,int4 sa) const=0;
|
|
virtual Pattern *commonSubPattern(const Pattern *b,int4 sa) const=0;
|
|
virtual bool isMatch(ParserWalker &walker) const=0; // Does this pattern match context
|
|
virtual int4 numDisjoint(void) const=0;
|
|
virtual DisjointPattern *getDisjoint(int4 i) const=0;
|
|
virtual bool alwaysTrue(void) const=0;
|
|
virtual bool alwaysFalse(void) const=0;
|
|
virtual bool alwaysInstructionTrue(void) const=0;
|
|
virtual void encode(Encoder &encoder) const=0;
|
|
virtual void decode(Decoder &decoder)=0;
|
|
};
|
|
|
|
class DisjointPattern : public Pattern { // A pattern with no ORs in it
|
|
virtual PatternBlock *getBlock(bool context) const=0;
|
|
public:
|
|
virtual int4 numDisjoint(void) const { return 0; }
|
|
virtual DisjointPattern *getDisjoint(int4 i) const { return (DisjointPattern *)0; }
|
|
uintm getMask(int4 startbit,int4 size,bool context) const;
|
|
uintm getValue(int4 startbit,int4 size,bool context) const;
|
|
int4 getLength(bool context) const;
|
|
bool specializes(const DisjointPattern *op2) const;
|
|
bool identical(const DisjointPattern *op2) const;
|
|
bool resolvesIntersect(const DisjointPattern *op1,const DisjointPattern *op2) const;
|
|
static DisjointPattern *decodeDisjoint(Decoder &decoder);
|
|
};
|
|
|
|
class InstructionPattern : public DisjointPattern { // Matches the instruction bitstream
|
|
PatternBlock *maskvalue;
|
|
virtual PatternBlock *getBlock(bool context) const { return context ? (PatternBlock *)0 : maskvalue; }
|
|
public:
|
|
InstructionPattern(void) { maskvalue = (PatternBlock *)0; } // For use with decode
|
|
InstructionPattern(PatternBlock *mv) { maskvalue = mv; }
|
|
InstructionPattern(bool tf) { maskvalue = new PatternBlock(tf); }
|
|
PatternBlock *getBlock(void) { return maskvalue; }
|
|
virtual ~InstructionPattern(void) { if (maskvalue != (PatternBlock *)0) delete maskvalue; }
|
|
virtual Pattern *simplifyClone(void) const { return new InstructionPattern(maskvalue->clone()); }
|
|
virtual void shiftInstruction(int4 sa) { maskvalue->shift(sa); }
|
|
virtual Pattern *doOr(const Pattern *b,int4 sa) const;
|
|
virtual Pattern *doAnd(const Pattern *b,int4 sa) const;
|
|
virtual Pattern *commonSubPattern(const Pattern *b,int4 sa) const;
|
|
virtual bool isMatch(ParserWalker &walker) const { return maskvalue->isInstructionMatch(walker); }
|
|
virtual bool alwaysTrue(void) const { return maskvalue->alwaysTrue(); }
|
|
virtual bool alwaysFalse(void) const { return maskvalue->alwaysFalse(); }
|
|
virtual bool alwaysInstructionTrue(void) const { return maskvalue->alwaysTrue(); }
|
|
virtual void encode(Encoder &encoder) const;
|
|
virtual void decode(Decoder &decoder);
|
|
};
|
|
|
|
class ContextPattern : public DisjointPattern { // Matches the context bitstream
|
|
PatternBlock *maskvalue;
|
|
virtual PatternBlock *getBlock(bool context) const { return context ? maskvalue : (PatternBlock *)0; }
|
|
public:
|
|
ContextPattern(void) { maskvalue = (PatternBlock *)0; } // For use with decode
|
|
ContextPattern(PatternBlock *mv) { maskvalue = mv; }
|
|
PatternBlock *getBlock(void) { return maskvalue; }
|
|
virtual ~ContextPattern(void) { if (maskvalue != (PatternBlock *)0) delete maskvalue; }
|
|
virtual Pattern *simplifyClone(void) const { return new ContextPattern(maskvalue->clone()); }
|
|
virtual void shiftInstruction(int4 sa) { } // do nothing
|
|
virtual Pattern *doOr(const Pattern *b,int4 sa) const;
|
|
virtual Pattern *doAnd(const Pattern *b,int4 sa) const;
|
|
virtual Pattern *commonSubPattern(const Pattern *b,int4 sa) const;
|
|
virtual bool isMatch(ParserWalker &walker) const { return maskvalue->isContextMatch(walker); }
|
|
virtual bool alwaysTrue(void) const { return maskvalue->alwaysTrue(); }
|
|
virtual bool alwaysFalse(void) const { return maskvalue->alwaysFalse(); }
|
|
virtual bool alwaysInstructionTrue(void) const { return true; }
|
|
virtual void encode(Encoder &encoder) const;
|
|
virtual void decode(Decoder &decoder);
|
|
};
|
|
|
|
// A pattern with a context piece and an instruction piece
|
|
class CombinePattern : public DisjointPattern {
|
|
ContextPattern *context; // Context piece
|
|
InstructionPattern *instr; // Instruction piece
|
|
virtual PatternBlock *getBlock(bool cont) const { return cont ? context->getBlock() : instr->getBlock(); }
|
|
public:
|
|
CombinePattern(void) { context = (ContextPattern *)0; instr = (InstructionPattern *)0; }
|
|
CombinePattern(ContextPattern *con,InstructionPattern *in) {
|
|
context = con; instr = in; }
|
|
virtual ~CombinePattern(void);
|
|
virtual Pattern *simplifyClone(void) const;
|
|
virtual void shiftInstruction(int4 sa) { instr->shiftInstruction(sa); }
|
|
virtual bool isMatch(ParserWalker &walker) const;
|
|
virtual bool alwaysTrue(void) const;
|
|
virtual bool alwaysFalse(void) const;
|
|
virtual bool alwaysInstructionTrue(void) const { return instr->alwaysInstructionTrue(); }
|
|
virtual Pattern *doOr(const Pattern *b,int4 sa) const;
|
|
virtual Pattern *doAnd(const Pattern *b,int4 sa) const;
|
|
virtual Pattern *commonSubPattern(const Pattern *b,int4 sa) const;
|
|
virtual void encode(Encoder &encoder) const;
|
|
virtual void decode(Decoder &decoder);
|
|
};
|
|
|
|
class OrPattern : public Pattern {
|
|
vector<DisjointPattern *> orlist;
|
|
public:
|
|
OrPattern(void) {} // For use with decode
|
|
OrPattern(DisjointPattern *a,DisjointPattern *b);
|
|
OrPattern(const vector<DisjointPattern *> &list);
|
|
virtual ~OrPattern(void);
|
|
virtual Pattern *simplifyClone(void) const;
|
|
virtual void shiftInstruction(int4 sa);
|
|
virtual bool isMatch(ParserWalker &walker) const;
|
|
virtual int4 numDisjoint(void) const { return orlist.size(); }
|
|
virtual DisjointPattern *getDisjoint(int4 i) const { return orlist[i]; }
|
|
virtual bool alwaysTrue(void) const;
|
|
virtual bool alwaysFalse(void) const;
|
|
virtual bool alwaysInstructionTrue(void) const;
|
|
virtual Pattern *doOr(const Pattern *b,int4 sa) const;
|
|
virtual Pattern *doAnd(const Pattern *b,int4 sa) const;
|
|
virtual Pattern *commonSubPattern(const Pattern *b,int4 sa) const;
|
|
virtual void encode(Encoder &encoder) const;
|
|
virtual void decode(Decoder &decoder);
|
|
};
|
|
|
|
} // End namespace ghidra
|
|
#endif
|