Decompiling functions with jump tables

• Functions with jump tables will load an address named jpt_XXXXXXXX or jtbl_XXXXXXXX and jump to it using jr.

Creating a scratch on decomp.me

• Copy the function assembly into the "Target Assembly" field.
• Copy the jump tables below the function assembly. Make sure you have .section .rodata before the jump tables.
• The result should look similar to the following:

.set noat      /* allow manual use of $at */
.set noreorder /* don't insert nops after branches */

glabel MyFunction
... (function asm continues here)

.section .rodata

glabel jtbl_801A7B40
... (jump table continues)

glabel jtbl_801A7B58
... (jump table continues)

• The jump tables will not be visible in the Target/Current panes of the editor, but if the decompilation shows one or more switches then it worked.

Adding to the repo

• When decompilation is complete these functions need to have the .rodata where the jump table is located ignored. The new jump table from the decompiled function will be used instead.
• This is done by modifying the splat yaml. The jump table needs to be extracted to its own file like this:

      - [0x3B560, rodata] # jpt_800E55C4
      - [0x3B720, rodata]

• We have a tool, tools/split_jpt_yaml/split_jpt_yaml.py to automate part of this process.
• It splits all of the jump tables into their own files. Usage is like this:
• It's possible the tool has already been run on the overlay you are looking at, in that case, you don't need to run it again.

# split DRA (uses jpt_ prefix)
# python3 tools/split_jpt_yaml/split_jpt_yaml.py config/splat.us.dra.yaml asm/us/dra/data/ jpt_

# split NO3 (uses jtbl_ prefix)
# python3 tools/split_jpt_yaml/split_jpt_yaml.py config/splat.us.stno3.yaml asm/us/st/no3/data/ jtbl_

• Look for the Cut below this line: print and copy the output below that.
• Paste the resulting output back into the yaml. For example in DRA:

segments:
  - name: dra
    type: code
    start: 0x00000000
    vram:  0x800A0000
    subalign: 4
    subsegments:
      (paste output here)

• The new function needs to be split into a new file. For example if our original C file is the following, and we want to decompile func_80133BDC, func_80133BDC needs to be at the start of the new file.

// 75F54.c
INCLUDE_ASM("asm/us/dra/nonmatchings/75F54", func_80133960);
INCLUDE_ASM("asm/us/dra/nonmatchings/75F54", func_80133BDC);
INCLUDE_ASM("asm/us/dra/nonmatchings/75F54", func_80133FCC);

• So we will have:

// 75F54.c
INCLUDE_ASM("asm/us/dra/nonmatchings/75F54", func_80133960);

// 93BDC.c
INCLUDE_ASM("asm/us/dra/nonmatchings/93BDC", func_80133BDC);
INCLUDE_ASM("asm/us/dra/nonmatchings/93BDC", func_80133FCC);

• The function contains jpt_80133C10 as the jump table. So we will modify the yaml like this:

# before
      - [0x4208C, rodata] # jpt_80133C10
      - [0x420B4, rodata] # jpt_80135058

# after
      - [0x4208C, .rodata, 93BDC] # jpt_80133C10
      - [0x420B4, rodata] # jpt_80135058

• The . in .rodata indicates that the segment should be ignored in the final compilation and the table from 93BDC.c should be used instead.

• Make sure to change the file paths for the INCLUDE_ASMs.