DISCLAIMER: These notes are from the defunct k8 project which precedes SquirrelJME. The notes for SquirrelJME start on 2016/02/26! The k8 project was effectively a Java SE 8 operating system and as such all of the notes are in the context of that scope. That project is no longer my goal as SquirrelJME is the spiritual successor to it.
10:34
I should be spitting out machine code very soon. My current only instruction to be spit out is lwsync. Which current appears that it gets assembled as: [FINEST] ASM: 0 0 0 0.
10:46
Some more fixing and lwsync is [FINEST] ASM: 0 4a c0 2. Although it should be 0: 7c 20 04 ac lwsync according to binutils.
10:52
The assembler wants to assemble AsmOp: (inst=SYNC addr=-1 len=4 args=[OPCD[[(0-6)]]:31, X_XO[[(21-10)]]:598, SYNC_L[[(9-2)]]:1, SYNC_E[[(12-4)]]:0]). And 31 | (598 << 21) | (1 << 9) is 1254097439 or 0x4A C0 02 1F.
10:36
I desire 0b01111100 00100000 00000100 10101100 But I get instead 0b01001010
11000000 00000010 00011111 Which if you look at it appears rather swapped,
sort of.
11:08
Printing the fields gives me stuff such as 0b111110.
11:17
That is printing in the incorrect order. The printed field order is [FINEST] 0b011111 [FINEST] 0b01 [FINEST] 0b0000 [FINEST] 0b1001010110.
11:29
My reverse bit ordering gives me 0b-------- 11111000 01000000 00110000 which
is not correct.
12:00
Seems like one of my fields it not being printed.
12:10
Ok, now my instruction is:
[FINEST] ASM: f8 40 03 52 0b11111000 01000000 00000011 01010010
need 0b01111100 00100000 00000100 10101100
12:19
Yipee! I believe I have it now!
[FINEST] ASM: 7c 20 04 ac 0b01111100 00100000 00000100 10101100
need 0b01111100 00100000 00000100 10101100
7c 20 04 ac lwsync according to binutils.
So for PowerPC I needed to swap the instruction bits AND the field bits. Normally big endian is on the byte level so it looks like: [8 9 10 11 12 13 14 15] [0 1 2 3 4 5 6 7], however PowerPC instructions are [15 14 13 12 11 10 9 8] [7 6 5 4 3 2 1 0].
15:58
Going to need to work on variable flowthrough (which gets what) and register allocation.
18:14
One thing I can do though is that when it comes time to write a Java compiler I can reuse the compiler system I am writing right now.
22:09
The SSA processing will be quite simple, it is just moving around of variables and such to determine how good they are and such.