-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathoddities.txt
57 lines (44 loc) · 1.87 KB
/
oddities.txt
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
avr-objdump -j .data -t _BUILD/LCDdirectLVDS.elf:
SYMBOL TABLE:
00800060 l d .data 00000000 .data
008000d4 l O .data 0000000c CSWTCH.86
008000c4 l O .data 00000010 spriteQ
00800064 l O .data 00000010 spriteLEAF
00800084 l O .data 00000010 spriteGOOMBA
008000a4 l O .data 00000010 spriteDEADGOOMBA
008000b4 l O .data 00000010 spriteSOLID
00800094 l O .data 00000010 spriteFLOWER
00800074 l O .data 00000010 spriteCOIN
008000e0 g O .data 00000006 LeafCamMotion
008000e6 g O .data 00000006 LeafMotion
00800062 g O .data 00000002 p_nowSprite
008000f2 g O .data 00000006 DefaultMotion
00800060 g O .data 00000001 updateFrame
008000fe g .data 00000000 __data_end
008000ec g O .data 00000006 DefaultCamMotion
00800061 g O .data 00000001 nowIsReward
008000f8 g O .data 00000006 NadaMotion
008000fe g .data 00000000 _edata
00800060 g .data 00000000 __data_start
What about Biggie, and 1up?!
was static sprite_t
now const __flash sprite_t
Changing *only* (Biggie, 1up, and) Question
No errors while building
Though not functional (Q)
Moving to const __flash increases codesize by roughly the same amount as
the decrease in memory requirement (?!)
Odd, no? Since, if it's a loaded-value in memory, then wouldn't it require
code-space to store the value to load?
Run into this before... duh. It is in flash, either way, but not listed in
.text since flash-usage is the sum of .data and .text
BUT
Changing over remaining (three) sprites should've had no effect on
flash-usage, yet it jumped from 6260 to 6322 or 62 bytes...
while .data decreased from 100 to 52 or 48 bytes...
Changing *one* sprite inbetween seemed to be +16B and -16B respectively
But changing one *earlier* seemed to be +18 and -16...
Dunno.
Moved motion_t's to flash:
Now 6340Bytes burnt, only 18B increase (!?)
.data decreased from 52 to 22 or 30B, as expected