Parse subregister information from Ghidra (#110)

cwe_checker_rs/src/pcode/expressions.rs

@@ -315,6 +315,14 @@ impl From<ExpressionType> for IrCastOpType {
     }
 }
 
+#[derive(Serialize, Deserialize, Debug, PartialEq, Eq, Hash, Clone)]
+pub struct RegisterProperties {
+    pub register: String,
+    pub base_register: String,
+    pub lsb: ByteSize,
+    pub size: ByteSize,
+}
+
 #[cfg(test)]
 mod tests {
     use super::*;
@@ -354,4 +362,19 @@ mod tests {
         )
         .unwrap();
     }
+
+    #[test]
+    fn register_properties_deserialization() {
+        let _: RegisterProperties = serde_json::from_str(
+            r#"
+            {
+                "register": "AH",
+                "base_register": "EAX",
+                "lsb": 2,
+                "size": 1
+            }
+            "#,
+        )
+        .unwrap();
+    }
 }

cwe_checker_rs/src/pcode/term.rs

-use super::{Expression, ExpressionType, Variable};
+use super::{Expression, ExpressionType, RegisterProperties, Variable};
 use crate::intermediate_representation::Arg as IrArg;
 use crate::intermediate_representation::Blk as IrBlk;
 use crate::intermediate_representation::ByteSize;
@@ -403,6 +403,7 @@ pub struct Project {
     pub program: Term<Program>,
     pub cpu_architecture: String,
     pub stack_pointer_register: Variable,
+    pub register_properties: Vec<RegisterProperties>,
     pub register_calling_convention: Vec<CallingConvention>,
 }
 
@@ -733,6 +734,14 @@ mod tests {
                 "is_virtual": false
             },
             "cpu_architecture": "x86_64",
+            "register_properties": [
+                {
+                    "register": "AH",
+                    "base_register": "EAX",
+                    "lsb": 2,
+                    "size": 1
+                }
+            ],
             "register_calling_convention": [
                 {
                     "calling_convention": "default",

ghidra/p_code_extractor/PcodeExtractor.java

@@ -24,7 +24,6 @@ import ghidra.program.model.lang.PrototypeModel;
 import ghidra.program.model.lang.Register;
 import ghidra.program.model.listing.Function;
 import ghidra.program.model.listing.FunctionIterator;
-import ghidra.program.model.listing.FunctionManager;
 import ghidra.program.model.listing.Instruction;
 import ghidra.program.model.listing.InstructionIterator;
 import ghidra.program.model.listing.Listing;
@@ -250,6 +249,7 @@ public class PcodeExtractor extends GhidraScript {
         } catch (FileNotFoundException e) {
             System.out.println(e);
         }
+        project.setRegisterProperties(HelperFunctions.getRegisterList());
 
         return project;
     }

ghidra/p_code_extractor/bil/RegisterProperties.java

+package bil;
+
+import com.google.gson.annotations.SerializedName;
+
+public class RegisterProperties {
+
+    @SerializedName("register")
+    private String register;
+    @SerializedName("base_register")
+    private String baseRegister;
+    @SerializedName("lsb")
+    private int lsb;
+    @SerializedName("size")
+    private int size;
+
+
+    public RegisterProperties(String register, String baseRegister, int lsb, int size) {
+        this.setRegister(register);
+        this.setBaseRegister(baseRegister);
+        this.setLsb(lsb);
+        this.setSize(size);
+    }
+
+    public String getRegister() {
+        return register;
+    }
+
+    public void setRegister(String register) {
+        this.register = register;
+    }
+
+    public String getBaseRegister() {
+        return baseRegister;
+    }
+
+    public void setBaseRegister(String baseRegister) {
+        this.baseRegister = baseRegister;
+    }
+
+    public int getLsb() {
+        return lsb;
+    }
+
+    public void setLsb(int lsb) {
+        this.lsb = lsb;
+    }
+
+    public int getSize() {
+        return size;
+    }
+
+    public void setSize(int size) {
+        this.size = size;
+    }
+    
+}

ghidra/p_code_extractor/internal/HelperFunctions.java

@@ -5,9 +5,11 @@ import java.util.HashMap;
 import java.util.regex.Matcher;
 import java.util.regex.Pattern;
 
+import bil.RegisterProperties;
 import bil.Variable;
 import ghidra.program.model.address.Address;
 import ghidra.program.model.address.AddressIterator;
+import ghidra.program.model.lang.Language;
 import ghidra.program.model.lang.Register;
 import ghidra.program.model.listing.Function;
 import ghidra.program.model.listing.FunctionManager;
@@ -221,4 +223,28 @@ public final class HelperFunctions {
         String[] arch = langId.split(":");
         return arch[0] + "_" + arch[2];
     }
+
+
+    /**
+     * Returns a list of register properties including the register name itself,
+     * the name of the base register, the lsb of the register in the base register
+     * and the size of the register.
+     * 
+     * @return list of register properties
+     */
+    public static ArrayList<RegisterProperties> getRegisterList() {
+        ArrayList<RegisterProperties> regProps = new ArrayList<RegisterProperties>();
+        Language language = ghidraProgram.getLanguage();
+        int archSizeInBytes = (int)(language.getLanguageDescription().getSize() / 8);
+        for(Register reg : language.getRegisters()) {
+            regProps.add(
+                new RegisterProperties(reg.getName(), 
+                                       reg.getBaseRegister().getName(), 
+                                       (int)(reg.getLeastSignificatBitInBaseRegister() / archSizeInBytes),
+                                       context.getRegisterVarnode(reg).getSize())
+            );
+        }
+
+        return regProps;
+    }
 }

ghidra/p_code_extractor/term/Project.java

 package term;
 
+import bil.RegisterProperties;
 import bil.Variable;
 import internal.RegisterConvention;
 
@@ -12,6 +13,8 @@ public class Project {
     private Term<Program> program;
     @SerializedName("stack_pointer_register")
     private Variable stackPointerRegister;
+    @SerializedName("register_properties")
+    private ArrayList<RegisterProperties> registerProperties;
     @SerializedName("cpu_architecture")
     private String cpuArch;
     @SerializedName("register_calling_convention")
@@ -58,4 +61,12 @@ public class Project {
     public void setRegisterConvention(ArrayList<RegisterConvention> conventions) {
         this.conventions = conventions;
     }
+
+    public ArrayList<RegisterProperties> getRegisterProperties() {
+        return registerProperties;
+    }
+
+    public void setRegisterProperties(ArrayList<RegisterProperties> registerProperties) {
+        this.registerProperties = registerProperties;
+    }
 }