Refactor SMS_EmitSinkInPollutionEffect vector normalization

[FranCarnovale]

Improved match on SMS_EmitSinkInPollutionEffect.
Removed temps, and just normalized C and B.

[decomp-dev]

Report for GMSJ01 (f1ccc3e - 3e705ba)

📉 Matched code: 27.88% (-0.04%, -1548 bytes)

🥀 5 broken matches

Unit	Item	Bytes	Before	After
mario/JSystem/JKernel/JKRDvdAramRipper	JKRDvdAramRipper::loadToAram(char*, unsigned long, JKRExpandSwitch, unsigned long, unsigned long)	-148	100.00%	0.00%
mario/JSystem/JKernel/JKRDvdAramRipper	nextSrcData(unsigned char*)	-11	100.00%	95.92%
mario/JSystem/JKernel/JKRDvdAramRipper	JKRDecompressFromDVDToAram(JKRDvdFile*, unsigned long, unsigned long, unsigned long, unsigned long, unsigned long)	-2	100.00%	99.55%
mario/JSystem/JKernel/JKRDvdAramRipper	decompSZS_subroutine(unsigned char*, unsigned long)	-2	100.00%	99.81%
mario/JSystem/JKernel/JKRDvdAramRipper	JKRDvdAramRipper::syncAram(JKRADCommand*, int)	-1	100.00%	99.83%

📈 1 improvement in an unmatched item

Unit	Item	Bytes	Before	After
mario/MarioUtil/EffectUtil	SMS_EmitSinkInPollutionEffect(const JGeometry::TVec3<float>&, const JGeometry::TVec3<float>&, bool)	+10	96.66%	98.75%

[Mrkol]

I think you screwed up git -- can you do a git rebase and ensure that you have a clean diff of your changes?

[FranCarnovale]

Should be good now, sorry about that.