mirror of
https://github.com/Druthulu/BFM-decomp
synced 2026-09-30 07:31:59 -04:00
Drew T
f5f8dec5ee
- COLD-START RESULT (verified independently): 4763/4763 ins, 45 mismatched = 99.06% byte / 99.94% structural, exact frame, exact opcode histogram. NOT a match; nothing banked (45 != 0, the byte-gate is the sole arbiter). The residual is 3 register-grant ties, 0 structural divergence. Named next move: variable REUSE across c0..c3/a0v..a3v, the one §76 lever class the pass never reached. - MY BRIEF'S PREMISE WAS WRONG BY CONSTRUCTION -> §79. I chose this target partly because §71's callee-set fingerprint returned 0.00 against every matched giant = "a genuine cold start". But the function makes ZERO jal calls, so its callee fingerprint is EMPTY and §71 CANNOT FIRE: 0.00 meant "cannot answer", not "no relative". Grepping the target's DATA symbol D_800A5E60 found the matched func_8017BEBC at once -- func_8017BF14 is the 4-light-box member of the same renderer family whose 3-box sibling func_8017D960 was matched hours earlier. RULE: when §71 returns an empty/zero-overlap callee set, fall back to DATA-symbol fingerprinting; an empty fingerprint must never become a cold-start brief. - NEW LEVER (§79): THE FRAME LAYOUT IS A DECLARATION-ORDER ORACLE. gcc-2.7.2 assigns stack slots to spilled pseudos in pseudo-number order, and pseudo numbers follow first use ~ declaration order -- so the target's frame map reads back its source's declaration order. Moving ONE line took 73% -> 84% structural and brought all 127 slots into exact correspondence. - §76 CONFIRMED AT SCALE: the entire -62 length residual was ONE allocno-class decision (c0..c3 declared inside the cull blocks -> 1-death local allocnos -> global.c:668-671 removes those regs from the global pool -> r1lo spills), 52% -> 93%. An __asm__ ref-dial reached the same spill and scored WORSE -- declaration scope beat the ref dial again. - PIN NUANCE: pins are safe on a 0-jal function (§74's hazard cannot arise), 4 pins took 94% -> 99%; but §72 held -- pins 5 and 6 made it worse. - EFFORT ANSWER, HONEST: xHigh from a genuine cold start on a 4,763-ins giant bought the decode, the exact length, the exact frame and 99.06%, and did NOT close. Budget a SECOND pass at this size: the first buys structure, the last ~1% is register grants. - FULL R22 DISCHARGED: make clean + extract-all + check-all -> 140 passed, 0 failed of 140 (run after the agent finished, per the deferral recorded in the pool commit). tools-health OK.
58 lines
2.8 KiB
Python
58 lines
2.8 KiB
Python
"""Structural aligner for func_8017BF14: masks registers AND sp-relative offsets,
|
|
so a frame-size drift does not hide real structural divergence.
|
|
usage: bf14_shape.py <obj.o> [nhunks] [--min N]"""
|
|
import re, subprocess, difflib, shutil, sys
|
|
OBJ = sys.argv[1]
|
|
TGT = '/home/musashi/bfm-decomp/asm/ov_SC03_116/nonmatchings/ov_SC03_116_jr_8017AE2C/func_8017BF14.s'
|
|
NH = int(sys.argv[2]) if len(sys.argv) > 2 and not sys.argv[2].startswith('-') else 40
|
|
MIN = int(sys.argv[sys.argv.index('--min')+1]) if '--min' in sys.argv else 1
|
|
OD = [c for c in ["mips-linux-gnu-objdump", "mipsel-linux-gnu-objdump"] if shutil.which(c)][0]
|
|
out = subprocess.run([OD, "-drz", OBJ], capture_output=True, text=True).stdout
|
|
def words_mine():
|
|
w=[]
|
|
for line in out.splitlines():
|
|
m=re.match(r'\s*[0-9a-f]+:\s+([0-9a-f]{8})\s',line)
|
|
if m: w.append(int(m.group(1),16))
|
|
elif 'R_MIPS_' in line and w and not isinstance(w[-1],tuple) and (w[-1]>>26) not in (2,3):
|
|
w[-1]=('R',w[-1])
|
|
return w
|
|
def words_tgt():
|
|
w=[]
|
|
for line in open(TGT):
|
|
m=re.match(r'\s*/\* \w+ [0-9A-F]{8} ([0-9A-F]{8}) \*/\s+(\S+)\s*(.*)',line)
|
|
if m:
|
|
v=int.from_bytes(bytes.fromhex(m.group(1)),'little')
|
|
w.append(('R',v) if ('%hi(' in line or '%lo(' in line) else v)
|
|
return w
|
|
MEM = set(range(0x20,0x30)) | {0x32,0x3A} # lb..sw, lwc2, swc2
|
|
def shape(x):
|
|
rel=isinstance(x,tuple); v=x[1] if rel else x
|
|
op=(v>>26)&0x3F
|
|
if op==0 or op==0x1C: k=(op<<12)|(v&0x3F)|(((v>>6)&0x1F)<<7)
|
|
elif op in (2,3): k=(op<<26)
|
|
elif op==0x12: k=v&0xFC1F07FF
|
|
elif op in (1,4,5,6,7): k=(op<<20)|(((v>>16)&0x1F)<<8)
|
|
else:
|
|
imm = 0 if rel else (v & 0xFFFF)
|
|
base=(v>>21)&0x1F
|
|
if op in MEM and base==29: imm=0 # mask sp offsets
|
|
k=(op<<20)|imm
|
|
return (k,'R' if rel else '')
|
|
mt=[m.group(1).strip() for m in (re.match(r'\s*[0-9a-f]+:\s+[0-9a-f]{8}\s+(.*)',l) for l in out.splitlines()) if m]
|
|
tt=[m.group(1).strip() for m in (re.match(r'\s*/\* \w+ [0-9A-F]{8} [0-9A-F]{8} \*/\s+(.*)',l) for l in open(TGT)) if m]
|
|
wm,wt=words_mine(),words_tgt()
|
|
a=[shape(x) for x in wm]; b=[shape(x) for x in wt]
|
|
sm=difflib.SequenceMatcher(None,a,b,autojunk=False); ops=sm.get_opcodes()
|
|
eq=sum(i2-i1 for t,i1,i2,j1,j2 in ops if t=='equal')
|
|
bad=[o for o in ops if o[0]!='equal']
|
|
print("mine %d target %d SHAPE-aligned %d/%d = %.1f%% hunks=%d div-tgt=%d"
|
|
% (len(wm),len(wt),eq,len(b),100.0*eq/len(b),len(bad),sum(o[4]-o[3] for o in bad)))
|
|
n=0
|
|
for t,i1,i2,j1,j2 in bad:
|
|
if max(j2-j1,i2-i1)<MIN: continue
|
|
n+=1
|
|
if n>NH: break
|
|
print(" %-8s tgt[%d:%d](%d) mine[%d:%d](%d)"%(t,j1,j2,j2-j1,i1,i2,i2-i1))
|
|
for k in range(j1,min(j2,j1+8)): print(" T %4d %s"%(k,tt[k] if k<len(tt) else '??'))
|
|
for k in range(i1,min(i2,i1+8)): print(" M %4d %s"%(k,mt[k] if k<len(mt) else '??'))
|