Breakthrough, entire TLB assembly dumped.

[retrobenny]

I managed to extract the lot of it with a GS code and a lot of game-play to populate most of it naturally, brute-forcing the last few spots that refused to populate memory before finally dumping it all.
With the help of AIO and theboy181, this was achieved with great success and the benefit of boosted emulation performance, a nice jump on interpreter speeds.

While I expect GE Decompressor already allowed extracting all of this, this still may be useful, as a patch and a custom PJ64 RDB configuration allows keeping the entire TLB loaded at all times for visibility and debugging via edits to equivalent cart data of it without needing to chase fleeting dynamic pointer address offsets.
15018000=B4018000 for example.

Enclosed will be the raw data and disasm of the entire TLB asm data and an updated/complete raw file as well.
(small overdump mistake at the end of earlier raw/disasm but nothing missing)

Add this info to Project64.rdb before/after patching a copy of your US version of Conker with the included bps file.

[30C7AFAE-F7DF4888-C:45]
Good Name=Conker's Bad Fur Day (TLB HACK)
Internal Name=CONKER BFD TLB
Status=Compatible
Clear Frame=2
Culling=1
Emulate Clear=1
Rom In Memory=1
FuncFind=2
RDRAM Size=8
Save Type=16kbit Eeprom
SMM-Cache=0
SMM-FUNC=0
SMM-PI DMA=0
SMM-TLB=0
ViRefresh=1500
Randomize SI/PI Interrupts=0
TLB: PAddr Start=335544320
TLB: Vaddr Len=2076672
TLB: Vaddr Start=352321536

Conker TLB dumps and patch.zip

[retrobenny]

Leaving a note here that 151FA130-151FAFFF actually appears to be duplicate code from 15041130-15041FFF for some strange reason.

I don't know why it lumped that extra data there as it doesn't appear when first populating 151F9000 and 151FA000 into expansion pak's 805F9000 and 805FA000 offsets.