Macintosh-SW/powermac-rom
1
0
Fork 0
mirror of https://github.com/elliotnunn/powermac-rom.git synced 2024-12-01 09:50:31 +00:00
Code Issues Projects Releases Wiki Activity
0bd13db3d2
powermac-rom/NanoKernel/NKThud.s

1236 lines
19 KiB
ArmAsm
Raw Normal View History

Initial commit
2017-11-19 04:11:07 +00:00
align 5
panic ; OUTSIDE REFERER
crset cr1_eq
b panic_common
panic_non_interactive
crclr cr1_eq
panic_common
mfsprg r1, 0
stmw r29, EWA.ThudSavedR29(r1)
lwz r1, EWA.PA_KDP(r1)
mflr r29
_Lock PSA.ThudLock, scratch1=r30, scratch2=r31
stw r29, KDP.ThudSavedLR(r1)
stw r0, KDP.ThudSavedR0(r1)
mfsprg r0, 1
stw r0, KDP.ThudSavedR1(r1)
stw r2, KDP.ThudSavedR2(r1)
mfsprg r2, 0
lmw r29, EWA.ThudSavedR29(r2)
stmw r3, KDP.ThudSavedR3(r1)
mfcr r0
stw r0, KDP.ThudSavedCR(r1)
; Save the silly multiply-quotient register
mfspr r0, pvr
rlwinm. r0, r0, 0, 0, 14
Remove unnecessary static branch hints These + and - characters in a branch mnemonic can cause the assembler to produce conditional branch instructions with that hint the branch predictor. The default for forward branches is -, and for backward branches is +. If a mnemonic is issued with the opposite sign, then bit 10 of the instruction (the LSB of the BO field) is set. My long-retired "ppcdisasm" script inserted these hints unconditionally, despite 98% of them not being required. The code is much cleaner now. I read in some old MPW release notes that PPCAsm and PPCLink together exhibit a quirk when linking conditional branches to imported symbols. PPCAsm always assembles these conditional branches as if they were forward branches (that is, a + hint will always set the hint bit, and a - hint will never). I hoped to use this property to divine whether the NanoKernel was linked from one or many assembly files, but I was frustrated by the lack of conditional branches between files.
2018-04-15 07:05:02 +00:00
bne @no_mq
Initial commit
2017-11-19 04:11:07 +00:00
dialect POWER
mfmq r0
dialect PowerPC
stw r0, KDP.ThudSavedMQ(r1)
@no_mq
mfxer r0
stw r0, KDP.ThudSavedXER(r1)
mfsprg r0, 2
stw r0, KDP.ThudSavedSPRG2(r1)
mfctr r0
stw r0, KDP.ThudSavedCTR(r1)
mfspr r0, pvr
stw r0, KDP.ThudSavedPVR(r1)
mfspr r0, dsisr
stw r0, KDP.ThudSavedDSISR(r1)
mfspr r0, dar
stw r0, KDP.ThudSavedDAR(r1)
; Save the time
mfpvr r0
rlwinm. r0, r0, 0, 0, 14
Remove unnecessary static branch hints These + and - characters in a branch mnemonic can cause the assembler to produce conditional branch instructions with that hint the branch predictor. The default for forward branches is -, and for backward branches is +. If a mnemonic is issued with the opposite sign, then bit 10 of the instruction (the LSB of the BO field) is set. My long-retired "ppcdisasm" script inserted these hints unconditionally, despite 98% of them not being required. The code is much cleaner now. I read in some old MPW release notes that PPCAsm and PPCLink together exhibit a quirk when linking conditional branches to imported symbols. PPCAsm always assembles these conditional branches as if they were forward branches (that is, a + hint will always set the hint bit, and a - hint will never). I hoped to use this property to divine whether the NanoKernel was linked from one or many assembly files, but I was frustrated by the lack of conditional branches between files.
2018-04-15 07:05:02 +00:00
bne @not_601
Initial commit
2017-11-19 04:11:07 +00:00
@rtcloop
dialect POWER
mfrtcu r0
mfrtcl r2
mfrtcu r3
dialect PowerPC
cmpw r0, r3
Remove unnecessary static branch hints These + and - characters in a branch mnemonic can cause the assembler to produce conditional branch instructions with that hint the branch predictor. The default for forward branches is -, and for backward branches is +. If a mnemonic is issued with the opposite sign, then bit 10 of the instruction (the LSB of the BO field) is set. My long-retired "ppcdisasm" script inserted these hints unconditionally, despite 98% of them not being required. The code is much cleaner now. I read in some old MPW release notes that PPCAsm and PPCLink together exhibit a quirk when linking conditional branches to imported symbols. PPCAsm always assembles these conditional branches as if they were forward branches (that is, a + hint will always set the hint bit, and a - hint will never). I hoped to use this property to divine whether the NanoKernel was linked from one or many assembly files, but I was frustrated by the lack of conditional branches between files.
2018-04-15 07:05:02 +00:00
bne @rtcloop
Initial commit
2017-11-19 04:11:07 +00:00
stw r0, KDP.ThudSavedTBU(r1)
stw r2, KDP.ThudSavedTB(r1)
b @end_if_601
@not_601
@tbloop
mftbu r0
mftb r2
mftbu r3
cmpw r0, r3
Remove unnecessary static branch hints These + and - characters in a branch mnemonic can cause the assembler to produce conditional branch instructions with that hint the branch predictor. The default for forward branches is -, and for backward branches is +. If a mnemonic is issued with the opposite sign, then bit 10 of the instruction (the LSB of the BO field) is set. My long-retired "ppcdisasm" script inserted these hints unconditionally, despite 98% of them not being required. The code is much cleaner now. I read in some old MPW release notes that PPCAsm and PPCLink together exhibit a quirk when linking conditional branches to imported symbols. PPCAsm always assembles these conditional branches as if they were forward branches (that is, a + hint will always set the hint bit, and a - hint will never). I hoped to use this property to divine whether the NanoKernel was linked from one or many assembly files, but I was frustrated by the lack of conditional branches between files.
2018-04-15 07:05:02 +00:00
bne @tbloop
Initial commit
2017-11-19 04:11:07 +00:00
stw r0, KDP.ThudSavedTBU(r1)
stw r2, KDP.ThudSavedTB(r1)
@end_if_601
mfspr r0, dec
stw r0, 0x07a8(r1)
mfspr r0, hid0
stw r0, 0x07ac(r1)
mfspr r0, sdr1
stw r0, 0x07b0(r1)
mfspr r0, srr0
stw r0, 0x07b4(r1)
mfspr r0, srr1
stw r0, 0x07b8(r1)
mfmsr r0
stw r0, 0x07bc(r1)
mfsr r0, 0
stw r0, 0x07c0(r1)
mfsr r0, 1
stw r0, 0x07c4(r1)
mfsr r0, 2
stw r0, 0x07c8(r1)
mfsr r0, 3
stw r0, 0x07cc(r1)
mfsr r0, 4
stw r0, 0x07d0(r1)
mfsr r0, 5
stw r0, 0x07d4(r1)
mfsr r0, 6
stw r0, 0x07d8(r1)
mfsr r0, 7
stw r0, 0x07dc(r1)
mfsr r0, 8
stw r0, 0x07e0(r1)
mfsr r0, 9
stw r0, 0x07e4(r1)
mfsr r0, 10
stw r0, 0x07e8(r1)
mfsr r0, 11
stw r0, 0x07ec(r1)
mfsr r0, 12
stw r0, 0x07f0(r1)
mfsr r0, 13
stw r0, 0x07f4(r1)
mfsr r0, 14
stw r0, 0x07f8(r1)
mfsr r0, 15
stw r0, 0x07fc(r1)
mfmsr r0
_bset r0, r0, MSR_FPbit
mtmsr r0
isync
stfd f0, 0x0800(r1)
stfd f1, 0x0808(r1)
stfd f2, 0x0810(r1)
stfd f3, 0x0818(r1)
stfd f4, 0x0820(r1)
stfd f5, 0x0828(r1)
stfd f6, 0x0830(r1)
stfd f7, 0x0838(r1)
stfd f8, 0x0840(r1)
stfd f9, 0x0848(r1)
stfd f10, 0x0850(r1)
stfd f11, 0x0858(r1)
stfd f12, 0x0860(r1)
stfd f13, 0x0868(r1)
stfd f14, 0x0870(r1)
stfd f15, 0x0878(r1)
stfd f16, 0x0880(r1)
stfd f17, 0x0888(r1)
stfd f18, 0x0890(r1)
stfd f19, 0x0898(r1)
stfd f20, 0x08a0(r1)
stfd f21, 0x08a8(r1)
stfd f22, 0x08b0(r1)
stfd f23, 0x08b8(r1)
stfd f24, 0x08c0(r1)
stfd f25, 0x08c8(r1)
stfd f26, 0x08d0(r1)
stfd f27, 0x08d8(r1)
stfd f28, 0x08e0(r1)
stfd f29, 0x08e8(r1)
stfd f30, 0x08f0(r1)
stfd f31, 0x08f8(r1)
mffs f31
lwz r0, 0x08fc(r1)
stfd f31, 0x08fc(r1)
stw r0, 0x08fc(r1)
Remove unnecessary static branch hints These + and - characters in a branch mnemonic can cause the assembler to produce conditional branch instructions with that hint the branch predictor. The default for forward branches is -, and for backward branches is +. If a mnemonic is issued with the opposite sign, then bit 10 of the instruction (the LSB of the BO field) is set. My long-retired "ppcdisasm" script inserted these hints unconditionally, despite 98% of them not being required. The code is much cleaner now. I read in some old MPW release notes that PPCAsm and PPCLink together exhibit a quirk when linking conditional branches to imported symbols. PPCAsm always assembles these conditional branches as if they were forward branches (that is, a + hint will always set the hint bit, and a - hint will never). I hoped to use this property to divine whether the NanoKernel was linked from one or many assembly files, but I was frustrated by the lack of conditional branches between files.
2018-04-15 07:05:02 +00:00
bne cr1, @0x260
Initial commit
2017-11-19 04:11:07 +00:00
if &TYPE('NKDebugShim') != 'UNDEFINED'
b @go_here_to_use_saved_debug_command
endif
@0x23c
lwz r1, 0(0)
addi r1, r1, 1
stw r1, 0(0)
li r1, 0
dcbst r1, r1
bl getchar
; gets kdp from print!!!
cmpwi r8, -0x01
Remove unnecessary static branch hints These + and - characters in a branch mnemonic can cause the assembler to produce conditional branch instructions with that hint the branch predictor. The default for forward branches is -, and for backward branches is +. If a mnemonic is issued with the opposite sign, then bit 10 of the instruction (the LSB of the BO field) is set. My long-retired "ppcdisasm" script inserted these hints unconditionally, despite 98% of them not being required. The code is much cleaner now. I read in some old MPW release notes that PPCAsm and PPCLink together exhibit a quirk when linking conditional branches to imported symbols. PPCAsm always assembles these conditional branches as if they were forward branches (that is, a + hint will always set the hint bit, and a - hint will never). I hoped to use this property to divine whether the NanoKernel was linked from one or many assembly files, but I was frustrated by the lack of conditional branches between files.
2018-04-15 07:05:02 +00:00
bne @0x260
Initial commit
2017-11-19 04:11:07 +00:00
b @0x23c
@0x260
lwz r8, 0x0edc(r1)
ori r8, r8, 0x02
stw r8, 0x0edc(r1)
_log '<EFBFBD> NanoKernel debugger^n'
@prompt
if &TYPE('NKDebugShim') != 'UNDEFINED'
b @NKDebugShimCode
endif
_log '<EFBFBD> ' ; thats an omega, btw
li r17, 0x00 ; r17 = charcount
stw r17, -0x08fc(r1)
@input_busywait
bl getchar
cmpwi r8, -1
Remove unnecessary static branch hints These + and - characters in a branch mnemonic can cause the assembler to produce conditional branch instructions with that hint the branch predictor. The default for forward branches is -, and for backward branches is +. If a mnemonic is issued with the opposite sign, then bit 10 of the instruction (the LSB of the BO field) is set. My long-retired "ppcdisasm" script inserted these hints unconditionally, despite 98% of them not being required. The code is much cleaner now. I read in some old MPW release notes that PPCAsm and PPCLink together exhibit a quirk when linking conditional branches to imported symbols. PPCAsm always assembles these conditional branches as if they were forward branches (that is, a + hint will always set the hint bit, and a - hint will never). I hoped to use this property to divine whether the NanoKernel was linked from one or many assembly files, but I was frustrated by the lack of conditional branches between files.
2018-04-15 07:05:02 +00:00
beq @input_busywait
Initial commit
2017-11-19 04:11:07 +00:00
mr r16, r8
cmpwi r16, 8 ; backspace
cmpwi cr1, r17, 0
Remove unnecessary static branch hints These + and - characters in a branch mnemonic can cause the assembler to produce conditional branch instructions with that hint the branch predictor. The default for forward branches is -, and for backward branches is +. If a mnemonic is issued with the opposite sign, then bit 10 of the instruction (the LSB of the BO field) is set. My long-retired "ppcdisasm" script inserted these hints unconditionally, despite 98% of them not being required. The code is much cleaner now. I read in some old MPW release notes that PPCAsm and PPCLink together exhibit a quirk when linking conditional branches to imported symbols. PPCAsm always assembles these conditional branches as if they were forward branches (that is, a + hint will always set the hint bit, and a - hint will never). I hoped to use this property to divine whether the NanoKernel was linked from one or many assembly files, but I was frustrated by the lack of conditional branches between files.
2018-04-15 07:05:02 +00:00
bne @not_backspace
ble cr1, @input_busywait
Initial commit
2017-11-19 04:11:07 +00:00
; Backspace, wipe position, then backspace again!
subi r17, r17, 1
li r8, 8
bl Printc
li r8, ' '
bl Printc
li r8, 8
bl Printc
b @input_busywait
@not_backspace
; If
cmpwi cr2, r17, 95
addi r18, r1, -0x960 ; prepare to copy the line!
Remove unnecessary static branch hints These + and - characters in a branch mnemonic can cause the assembler to produce conditional branch instructions with that hint the branch predictor. The default for forward branches is -, and for backward branches is +. If a mnemonic is issued with the opposite sign, then bit 10 of the instruction (the LSB of the BO field) is set. My long-retired "ppcdisasm" script inserted these hints unconditionally, despite 98% of them not being required. The code is much cleaner now. I read in some old MPW release notes that PPCAsm and PPCLink together exhibit a quirk when linking conditional branches to imported symbols. PPCAsm always assembles these conditional branches as if they were forward branches (that is, a + hint will always set the hint bit, and a - hint will never). I hoped to use this property to divine whether the NanoKernel was linked from one or many assembly files, but I was frustrated by the lack of conditional branches between files.
2018-04-15 07:05:02 +00:00
blt cr2, @short_line
Initial commit
2017-11-19 04:11:07 +00:00
_log '^b'
b @input_busywait
@0x30c
addi r17, r17, 1 ; accept the character as an addition to the line
mr r8, r16
bl Printc
b @input_busywait
@short_line
cmpwi r16, 13
stbx r16, r17, r18
Remove unnecessary static branch hints These + and - characters in a branch mnemonic can cause the assembler to produce conditional branch instructions with that hint the branch predictor. The default for forward branches is -, and for backward branches is +. If a mnemonic is issued with the opposite sign, then bit 10 of the instruction (the LSB of the BO field) is set. My long-retired "ppcdisasm" script inserted these hints unconditionally, despite 98% of them not being required. The code is much cleaner now. I read in some old MPW release notes that PPCAsm and PPCLink together exhibit a quirk when linking conditional branches to imported symbols. PPCAsm always assembles these conditional branches as if they were forward branches (that is, a + hint will always set the hint bit, and a - hint will never). I hoped to use this property to divine whether the NanoKernel was linked from one or many assembly files, but I was frustrated by the lack of conditional branches between files.
2018-04-15 07:05:02 +00:00
bne @0x30c
Initial commit
2017-11-19 04:11:07 +00:00
li r16, 0x00
stbx r16, r17, r18
_log '^n'
@go_here_to_use_saved_debug_command
; Now a line is expected to be committed:
addi r15, r1, -0x960
; r15 = start
bl next_cmd_word
; r15 = ptr
; r16 = char
cmpwi r16, 0x00
Remove unnecessary static branch hints These + and - characters in a branch mnemonic can cause the assembler to produce conditional branch instructions with that hint the branch predictor. The default for forward branches is -, and for backward branches is +. If a mnemonic is issued with the opposite sign, then bit 10 of the instruction (the LSB of the BO field) is set. My long-retired "ppcdisasm" script inserted these hints unconditionally, despite 98% of them not being required. The code is much cleaner now. I read in some old MPW release notes that PPCAsm and PPCLink together exhibit a quirk when linking conditional branches to imported symbols. PPCAsm always assembles these conditional branches as if they were forward branches (that is, a + hint will always set the hint bit, and a - hint will never). I hoped to use this property to divine whether the NanoKernel was linked from one or many assembly files, but I was frustrated by the lack of conditional branches between files.
2018-04-15 07:05:02 +00:00
beq @prompt
Initial commit
2017-11-19 04:11:07 +00:00
bl @load_commands
mflr r16
bl @load_tbl
mflr r17
; r16 = command strings
; r17 = lut
bl cmd_lookup
; cr0 = found
; r17 = ptr to lut entry
Remove unnecessary static branch hints These + and - characters in a branch mnemonic can cause the assembler to produce conditional branch instructions with that hint the branch predictor. The default for forward branches is -, and for backward branches is +. If a mnemonic is issued with the opposite sign, then bit 10 of the instruction (the LSB of the BO field) is set. My long-retired "ppcdisasm" script inserted these hints unconditionally, despite 98% of them not being required. The code is much cleaner now. I read in some old MPW release notes that PPCAsm and PPCLink together exhibit a quirk when linking conditional branches to imported symbols. PPCAsm always assembles these conditional branches as if they were forward branches (that is, a + hint will always set the hint bit, and a - hint will never). I hoped to use this property to divine whether the NanoKernel was linked from one or many assembly files, but I was frustrated by the lack of conditional branches between files.
2018-04-15 07:05:02 +00:00
bne @bad_command
Initial commit
2017-11-19 04:11:07 +00:00
bl @load_tbl
mflr r16
lwz r17, 0x0000(r17)
add r16, r16, r17
mtlr r16
blr
@bad_command
_log '???^n'
b @prompt
@load_commands
blrl
string CString
dc.b 'dm'
dc.b 'dml'
dc.b 'g'
dc.b 'id'
dc.b 'kd'
dc.b 'td'
dc.b '?'
dc.b 'help'
dc.b 0xff
align 2
@load_tbl
blrl
@tbl
dc.l @cmd_dumpmem_physical - @tbl
dc.l @cmd_dumpmem_logical - @tbl
dc.l @cmd_goto - @tbl
dc.l @cmd_opaque_id_info - @tbl
dc.l @cmd_display_kern_data - @tbl
dc.l @cmd_dump_registers - @tbl
dc.l @cmd_help - @tbl
dc.l @cmd_help - @tbl
dc.l 0
@cmd_help
_log 'Commands:^n'
_log ' dm address [length] -- Display physical^n'
_log ' dml address [length] -- Display logical^n'
_log ' g [address] -- Go resume^n'
_log ' id [-all -p -t -tm -q -s -r -c -sp -e -cg -a -n -nc]^n'
_log ' id idvalue -- Obtain opaque ID info^n'
_log ' kd -- Display kernel data^n'
_log ' td -- Dump registers^n'
b @prompt
@cmd_dumpmem_physical
; r15 = start
bl next_cmd_word
; r15 = ptr
; r16 = char
cmpwi r16, 0x00
Remove unnecessary static branch hints These + and - characters in a branch mnemonic can cause the assembler to produce conditional branch instructions with that hint the branch predictor. The default for forward branches is -, and for backward branches is +. If a mnemonic is issued with the opposite sign, then bit 10 of the instruction (the LSB of the BO field) is set. My long-retired "ppcdisasm" script inserted these hints unconditionally, despite 98% of them not being required. The code is much cleaner now. I read in some old MPW release notes that PPCAsm and PPCLink together exhibit a quirk when linking conditional branches to imported symbols. PPCAsm always assembles these conditional branches as if they were forward branches (that is, a + hint will always set the hint bit, and a - hint will never). I hoped to use this property to divine whether the NanoKernel was linked from one or many assembly files, but I was frustrated by the lack of conditional branches between files.
2018-04-15 07:05:02 +00:00
beq @missing_physical_addr
Initial commit
2017-11-19 04:11:07 +00:00
bl major_0x187b0
Remove unnecessary static branch hints These + and - characters in a branch mnemonic can cause the assembler to produce conditional branch instructions with that hint the branch predictor. The default for forward branches is -, and for backward branches is +. If a mnemonic is issued with the opposite sign, then bit 10 of the instruction (the LSB of the BO field) is set. My long-retired "ppcdisasm" script inserted these hints unconditionally, despite 98% of them not being required. The code is much cleaner now. I read in some old MPW release notes that PPCAsm and PPCLink together exhibit a quirk when linking conditional branches to imported symbols. PPCAsm always assembles these conditional branches as if they were forward branches (that is, a + hint will always set the hint bit, and a - hint will never). I hoped to use this property to divine whether the NanoKernel was linked from one or many assembly files, but I was frustrated by the lack of conditional branches between files.
2018-04-15 07:05:02 +00:00
bne @bad_length_1
Initial commit
2017-11-19 04:11:07 +00:00
mr r30, r16
li r31, 0x10
; r15 = start
bl next_cmd_word
; r15 = ptr
; r16 = char
cmpwi r16, 0x00
Remove unnecessary static branch hints These + and - characters in a branch mnemonic can cause the assembler to produce conditional branch instructions with that hint the branch predictor. The default for forward branches is -, and for backward branches is +. If a mnemonic is issued with the opposite sign, then bit 10 of the instruction (the LSB of the BO field) is set. My long-retired "ppcdisasm" script inserted these hints unconditionally, despite 98% of them not being required. The code is much cleaner now. I read in some old MPW release notes that PPCAsm and PPCLink together exhibit a quirk when linking conditional branches to imported symbols. PPCAsm always assembles these conditional branches as if they were forward branches (that is, a + hint will always set the hint bit, and a - hint will never). I hoped to use this property to divine whether the NanoKernel was linked from one or many assembly files, but I was frustrated by the lack of conditional branches between files.
2018-04-15 07:05:02 +00:00
beq @0x5e0
Initial commit
2017-11-19 04:11:07 +00:00
bl major_0x187b0
Remove unnecessary static branch hints These + and - characters in a branch mnemonic can cause the assembler to produce conditional branch instructions with that hint the branch predictor. The default for forward branches is -, and for backward branches is +. If a mnemonic is issued with the opposite sign, then bit 10 of the instruction (the LSB of the BO field) is set. My long-retired "ppcdisasm" script inserted these hints unconditionally, despite 98% of them not being required. The code is much cleaner now. I read in some old MPW release notes that PPCAsm and PPCLink together exhibit a quirk when linking conditional branches to imported symbols. PPCAsm always assembles these conditional branches as if they were forward branches (that is, a + hint will always set the hint bit, and a - hint will never). I hoped to use this property to divine whether the NanoKernel was linked from one or many assembly files, but I was frustrated by the lack of conditional branches between files.
2018-04-15 07:05:02 +00:00
bne @bad_length_1
Initial commit
2017-11-19 04:11:07 +00:00
mr r31, r16
@0x5e0
addi r31, r31, 0x03
rlwinm r31, r31, 0, 0, 29
mr r16, r30
mr r17, r31
bl print_memory
b @prompt
@missing_physical_addr
_log 'Need a physical address^n'
b @prompt
@bad_length_1
_log 'Length must be a hexadecimal value^n'
b @prompt
@cmd_dumpmem_logical
_log 'Logical memory^n'
; r15 = start
bl next_cmd_word
; r15 = ptr
; r16 = char
cmpwi r16, 0x00
Remove unnecessary static branch hints These + and - characters in a branch mnemonic can cause the assembler to produce conditional branch instructions with that hint the branch predictor. The default for forward branches is -, and for backward branches is +. If a mnemonic is issued with the opposite sign, then bit 10 of the instruction (the LSB of the BO field) is set. My long-retired "ppcdisasm" script inserted these hints unconditionally, despite 98% of them not being required. The code is much cleaner now. I read in some old MPW release notes that PPCAsm and PPCLink together exhibit a quirk when linking conditional branches to imported symbols. PPCAsm always assembles these conditional branches as if they were forward branches (that is, a + hint will always set the hint bit, and a - hint will never). I hoped to use this property to divine whether the NanoKernel was linked from one or many assembly files, but I was frustrated by the lack of conditional branches between files.
2018-04-15 07:05:02 +00:00
beq @missing_logical_addr
Initial commit
2017-11-19 04:11:07 +00:00
bl major_0x187b0
Remove unnecessary static branch hints These + and - characters in a branch mnemonic can cause the assembler to produce conditional branch instructions with that hint the branch predictor. The default for forward branches is -, and for backward branches is +. If a mnemonic is issued with the opposite sign, then bit 10 of the instruction (the LSB of the BO field) is set. My long-retired "ppcdisasm" script inserted these hints unconditionally, despite 98% of them not being required. The code is much cleaner now. I read in some old MPW release notes that PPCAsm and PPCLink together exhibit a quirk when linking conditional branches to imported symbols. PPCAsm always assembles these conditional branches as if they were forward branches (that is, a + hint will always set the hint bit, and a - hint will never). I hoped to use this property to divine whether the NanoKernel was linked from one or many assembly files, but I was frustrated by the lack of conditional branches between files.
2018-04-15 07:05:02 +00:00
bne @bad_length_2
Initial commit
2017-11-19 04:11:07 +00:00
mr r30, r16
li r31, 0x10
; r15 = start
bl next_cmd_word
; r15 = ptr
; r16 = char
cmpwi r16, 0x00
Remove unnecessary static branch hints These + and - characters in a branch mnemonic can cause the assembler to produce conditional branch instructions with that hint the branch predictor. The default for forward branches is -, and for backward branches is +. If a mnemonic is issued with the opposite sign, then bit 10 of the instruction (the LSB of the BO field) is set. My long-retired "ppcdisasm" script inserted these hints unconditionally, despite 98% of them not being required. The code is much cleaner now. I read in some old MPW release notes that PPCAsm and PPCLink together exhibit a quirk when linking conditional branches to imported symbols. PPCAsm always assembles these conditional branches as if they were forward branches (that is, a + hint will always set the hint bit, and a - hint will never). I hoped to use this property to divine whether the NanoKernel was linked from one or many assembly files, but I was frustrated by the lack of conditional branches between files.
2018-04-15 07:05:02 +00:00
beq @0x6b0
Initial commit
2017-11-19 04:11:07 +00:00
bl major_0x187b0
Remove unnecessary static branch hints These + and - characters in a branch mnemonic can cause the assembler to produce conditional branch instructions with that hint the branch predictor. The default for forward branches is -, and for backward branches is +. If a mnemonic is issued with the opposite sign, then bit 10 of the instruction (the LSB of the BO field) is set. My long-retired "ppcdisasm" script inserted these hints unconditionally, despite 98% of them not being required. The code is much cleaner now. I read in some old MPW release notes that PPCAsm and PPCLink together exhibit a quirk when linking conditional branches to imported symbols. PPCAsm always assembles these conditional branches as if they were forward branches (that is, a + hint will always set the hint bit, and a - hint will never). I hoped to use this property to divine whether the NanoKernel was linked from one or many assembly files, but I was frustrated by the lack of conditional branches between files.
2018-04-15 07:05:02 +00:00
bne @bad_length_2
Initial commit
2017-11-19 04:11:07 +00:00
mr r31, r16
@0x6b0
addi r31, r31, 0x03
rlwinm r31, r31, 0, 0, 29
mr r16, r30
mr r17, r31
bl print_memory_logical
b @prompt
@missing_logical_addr
_log 'Need a logical address^n'
b @prompt
@bad_length_2
_log 'Length must be a hexadecimal value^n'
b @prompt
@cmd_goto
; r15 = start
bl next_cmd_word
; r15 = ptr
; r16 = char
cmpwi r16, 0x00
lwz r31, 0x0904(r1)
Remove unnecessary static branch hints These + and - characters in a branch mnemonic can cause the assembler to produce conditional branch instructions with that hint the branch predictor. The default for forward branches is -, and for backward branches is +. If a mnemonic is issued with the opposite sign, then bit 10 of the instruction (the LSB of the BO field) is set. My long-retired "ppcdisasm" script inserted these hints unconditionally, despite 98% of them not being required. The code is much cleaner now. I read in some old MPW release notes that PPCAsm and PPCLink together exhibit a quirk when linking conditional branches to imported symbols. PPCAsm always assembles these conditional branches as if they were forward branches (that is, a + hint will always set the hint bit, and a - hint will never). I hoped to use this property to divine whether the NanoKernel was linked from one or many assembly files, but I was frustrated by the lack of conditional branches between files.
2018-04-15 07:05:02 +00:00
beq @0x748
Initial commit
2017-11-19 04:11:07 +00:00
bl major_0x187b0
Remove unnecessary static branch hints These + and - characters in a branch mnemonic can cause the assembler to produce conditional branch instructions with that hint the branch predictor. The default for forward branches is -, and for backward branches is +. If a mnemonic is issued with the opposite sign, then bit 10 of the instruction (the LSB of the BO field) is set. My long-retired "ppcdisasm" script inserted these hints unconditionally, despite 98% of them not being required. The code is much cleaner now. I read in some old MPW release notes that PPCAsm and PPCLink together exhibit a quirk when linking conditional branches to imported symbols. PPCAsm always assembles these conditional branches as if they were forward branches (that is, a + hint will always set the hint bit, and a - hint will never). I hoped to use this property to divine whether the NanoKernel was linked from one or many assembly files, but I was frustrated by the lack of conditional branches between files.
2018-04-15 07:05:02 +00:00
bne @bad_resume_address
Initial commit
2017-11-19 04:11:07 +00:00
stw r16, 0x0904(r1)
@0x748
@NKDebugShimCode
_log 'Resuming at '
lwz r31, 0x0904(r1)
mr r8, r31
bl Printw
_log ' - wish me luck.^n'
bl prereturn
lwz r8, 0x0904(r1)
Use new _AssertAndRelease macro This is the usual way to release a lock. Most MP calls employ it indirectly, through their return path.
2017-12-12 07:54:19 +00:00
_AssertAndRelease PSA.ThudLock, scratch=r9
Initial commit
2017-11-19 04:11:07 +00:00
mtlr r8
blr
@bad_resume_address
_log 'Need hexadecimal value for resume address^n'
b @prompt
@cmd_opaque_id_info
; r15 = start
bl next_cmd_word
; r15 = ptr
; r16 = char
cmpwi r16, 0x00
Remove unnecessary static branch hints These + and - characters in a branch mnemonic can cause the assembler to produce conditional branch instructions with that hint the branch predictor. The default for forward branches is -, and for backward branches is +. If a mnemonic is issued with the opposite sign, then bit 10 of the instruction (the LSB of the BO field) is set. My long-retired "ppcdisasm" script inserted these hints unconditionally, despite 98% of them not being required. The code is much cleaner now. I read in some old MPW release notes that PPCAsm and PPCLink together exhibit a quirk when linking conditional branches to imported symbols. PPCAsm always assembles these conditional branches as if they were forward branches (that is, a + hint will always set the hint bit, and a - hint will never). I hoped to use this property to divine whether the NanoKernel was linked from one or many assembly files, but I was frustrated by the lack of conditional branches between files.
2018-04-15 07:05:02 +00:00
beq @missing_opaque_id
Initial commit
2017-11-19 04:11:07 +00:00
bl @load_id_args
mflr r16
li r17, 0x00
; r16 = command strings
; r17 = lut
bl cmd_lookup
; cr0 = found
; r17 = ptr to lut entry
Remove unnecessary static branch hints These + and - characters in a branch mnemonic can cause the assembler to produce conditional branch instructions with that hint the branch predictor. The default for forward branches is -, and for backward branches is +. If a mnemonic is issued with the opposite sign, then bit 10 of the instruction (the LSB of the BO field) is set. My long-retired "ppcdisasm" script inserted these hints unconditionally, despite 98% of them not being required. The code is much cleaner now. I read in some old MPW release notes that PPCAsm and PPCLink together exhibit a quirk when linking conditional branches to imported symbols. PPCAsm always assembles these conditional branches as if they were forward branches (that is, a + hint will always set the hint bit, and a - hint will never). I hoped to use this property to divine whether the NanoKernel was linked from one or many assembly files, but I was frustrated by the lack of conditional branches between files.
2018-04-15 07:05:02 +00:00
bne @0x884
Initial commit
2017-11-19 04:11:07 +00:00
li r29, 0x00
li r30, 0x00
srwi r31, r17, 2
@0x82c
mr r8, r30
mr r9, r31
bl GetNextIDOfClass
mr. r30, r8
Remove unnecessary static branch hints These + and - characters in a branch mnemonic can cause the assembler to produce conditional branch instructions with that hint the branch predictor. The default for forward branches is -, and for backward branches is +. If a mnemonic is issued with the opposite sign, then bit 10 of the instruction (the LSB of the BO field) is set. My long-retired "ppcdisasm" script inserted these hints unconditionally, despite 98% of them not being required. The code is much cleaner now. I read in some old MPW release notes that PPCAsm and PPCLink together exhibit a quirk when linking conditional branches to imported symbols. PPCAsm always assembles these conditional branches as if they were forward branches (that is, a + hint will always set the hint bit, and a - hint will never). I hoped to use this property to divine whether the NanoKernel was linked from one or many assembly files, but I was frustrated by the lack of conditional branches between files.
2018-04-15 07:05:02 +00:00
beq @0x868
Initial commit
2017-11-19 04:11:07 +00:00
mr r8, r8
bl Printw
addi r29, r29, 0x01
andi. r29, r29, 0x07
Remove unnecessary static branch hints These + and - characters in a branch mnemonic can cause the assembler to produce conditional branch instructions with that hint the branch predictor. The default for forward branches is -, and for backward branches is +. If a mnemonic is issued with the opposite sign, then bit 10 of the instruction (the LSB of the BO field) is set. My long-retired "ppcdisasm" script inserted these hints unconditionally, despite 98% of them not being required. The code is much cleaner now. I read in some old MPW release notes that PPCAsm and PPCLink together exhibit a quirk when linking conditional branches to imported symbols. PPCAsm always assembles these conditional branches as if they were forward branches (that is, a + hint will always set the hint bit, and a - hint will never). I hoped to use this property to divine whether the NanoKernel was linked from one or many assembly files, but I was frustrated by the lack of conditional branches between files.
2018-04-15 07:05:02 +00:00
bne @0x82c
Initial commit
2017-11-19 04:11:07 +00:00
_log '^n'
b @0x82c
@0x868
cmpwi r29, 0x00
Remove unnecessary static branch hints These + and - characters in a branch mnemonic can cause the assembler to produce conditional branch instructions with that hint the branch predictor. The default for forward branches is -, and for backward branches is +. If a mnemonic is issued with the opposite sign, then bit 10 of the instruction (the LSB of the BO field) is set. My long-retired "ppcdisasm" script inserted these hints unconditionally, despite 98% of them not being required. The code is much cleaner now. I read in some old MPW release notes that PPCAsm and PPCLink together exhibit a quirk when linking conditional branches to imported symbols. PPCAsm always assembles these conditional branches as if they were forward branches (that is, a + hint will always set the hint bit, and a - hint will never). I hoped to use this property to divine whether the NanoKernel was linked from one or many assembly files, but I was frustrated by the lack of conditional branches between files.
2018-04-15 07:05:02 +00:00
beq @prompt
Initial commit
2017-11-19 04:11:07 +00:00
_log '^n'
b @prompt
@0x884
bl major_0x187b0
Remove unnecessary static branch hints These + and - characters in a branch mnemonic can cause the assembler to produce conditional branch instructions with that hint the branch predictor. The default for forward branches is -, and for backward branches is +. If a mnemonic is issued with the opposite sign, then bit 10 of the instruction (the LSB of the BO field) is set. My long-retired "ppcdisasm" script inserted these hints unconditionally, despite 98% of them not being required. The code is much cleaner now. I read in some old MPW release notes that PPCAsm and PPCLink together exhibit a quirk when linking conditional branches to imported symbols. PPCAsm always assembles these conditional branches as if they were forward branches (that is, a + hint will always set the hint bit, and a - hint will never). I hoped to use this property to divine whether the NanoKernel was linked from one or many assembly files, but I was frustrated by the lack of conditional branches between files.
2018-04-15 07:05:02 +00:00
bne @bad_opaque_id
Initial commit
2017-11-19 04:11:07 +00:00
mr r30, r16
mr r8, r16
; r8 = id
bl LookupID
; r8 = something not sure what
; r9 = 0:inval, 1:proc, 2:task, 3:timer, 4:q, 5:sema, 6:cr, 7:cpu, 8:addrspc, 9:evtg, 10:cohg, 11:area, 12:not, 13:log
mr r31, r8
_log 'ID '
mr r8, r30
bl Printw
bl @load_id_kind_strings
mflr r17
slwi r18, r9, 4
add r8, r17, r18
bl PrintS
cmpwi r9, 0x00
Remove unnecessary static branch hints These + and - characters in a branch mnemonic can cause the assembler to produce conditional branch instructions with that hint the branch predictor. The default for forward branches is -, and for backward branches is +. If a mnemonic is issued with the opposite sign, then bit 10 of the instruction (the LSB of the BO field) is set. My long-retired "ppcdisasm" script inserted these hints unconditionally, despite 98% of them not being required. The code is much cleaner now. I read in some old MPW release notes that PPCAsm and PPCLink together exhibit a quirk when linking conditional branches to imported symbols. PPCAsm always assembles these conditional branches as if they were forward branches (that is, a + hint will always set the hint bit, and a - hint will never). I hoped to use this property to divine whether the NanoKernel was linked from one or many assembly files, but I was frustrated by the lack of conditional branches between files.
2018-04-15 07:05:02 +00:00
beq @0x978
Initial commit
2017-11-19 04:11:07 +00:00
_log ' at '
mr r8, r31
bl Printw
_log '^n'
mr r16, r31
bl @load_more_jumps
mflr r17
slwi r18, r9, 2
lwzx r17, r17, r18
bl print_memory
b @prompt
@missing_opaque_id
_log 'Need an opaque ID^n'
b @prompt
@bad_opaque_id
_log 'ID must be a hexadecimal value^n'
b @prompt
@0x978
_log '^n'
b @prompt
@load_id_kind_strings
blrl
string CString
dc.b 'is invalid. '
dc.b 'Process '
dc.b 'Task '
dc.b 'Timer '
dc.b 'Queue '
dc.b 'Semaphore '
dc.b 'Critical Region'
dc.b 'Cpu '
dc.b 'Address Space '
dc.b 'Event Group '
dc.b 'Coherence Group'
dc.b 'Area '
dc.b 'Notification '
dc.b 'Console Log '
align 2
@load_more_jumps
blrl
dc.l 0
dc.l Process.Size
dc.l Task.Size
dc.l Timer.Size
dc.l Queue.Size
dc.l Semaphore.Size
dc.l CriticalRegion.Size
dc.l CPU.Size
dc.l AddressSpace.Size
dc.l EventGroup.Size
dc.l CoherenceGroup.Size
dc.l Area.Size
dc.l Notification.Size
dc.l ConsoleLog.Size
@load_id_args
blrl
dc.b '-all'
dc.b '-p'
dc.b '-t'
dc.b '-tm'
dc.b '-q'
dc.b '-s'
dc.b '-r'
dc.b '-c'
dc.b '-sp'
dc.b '-e'
dc.b '-cg'
dc.b '-a'
dc.b '-n'
dc.b '-nc'
dc.b 0xff
align 2
@cmd_display_kern_data
mfsprg r17, 0
_log 'Kernel version '
lhz r8, KDP.InfoRecord + InfoRecord.NKNanoKernelInfoVer(r1)
bl Printh
_log 'Code base '
lwz r8, KDP.PA_NanoKernelCode(r1)
bl Printw
_log 'PSA '
addi r8, r17, PSA.Base
bl Printw
_log 'KDP '
mr r8, r1
bl Printw
_log 'EDP '
lwz r8, KDP.PA_EmulatorData(r1)
bl Printw
_log '^nCurrent EWA '
mr r8, r17
bl Printw
_log ' is CPU '
Reverse userspace sync primitives Namely queues, semaphores, critical regions, event groups and "notifications". The MP calls implementing these services have been named after their MPLibrary wrapper functions. This convention will be followed in the future (no more NKCreateEvent).
2018-02-17 16:57:01 +00:00
lhz r8, EWA.CPUIndex(r17)
Initial commit
2017-11-19 04:11:07 +00:00
bl Printh
_log ' ID-'
lwz r8, -0x0340(r17)
bl Printw
lwz r18, EWA.PA_CurTask(r17)
_log '^nCurrent task '
mr r8, r18
bl Printw
_log 'ID-'
lwz r8, Task.ID(r18)
bl Printw
_log 'name \"'
lwz r8, Task.Name(r18)
rotlwi r8, r8, 8
bl Printc
rotlwi r8, r8, 8
bl Printc
rotlwi r8, r8, 8
bl Printc
rotlwi r8, r8, 8
bl Printc
_log '\" Owning process '
lwz r8, 0x006c(r18)
bl Printw
_log ' ID-'
lwz r8, 0x0060(r18)
bl Printw
_log '^nAddress Space '
lwz r18, -0x001c(r17)
mr r8, r18
bl Printw
_log ' ID-'
lwz r8, 0x0000(r18)
bl Printw
_log '^n'
bl print_xpt_info
b @prompt
@cmd_dump_registers
_log 'Kernel registers:^n'
bl print_sprgs
bl print_sprs
_log '^n'
bl print_segment_registers
_log '^n'
bl print_gprs
b @prompt
mflr r16
lwz r17, 0x07b4(r1)
rlwinm r17, r17, 16, 16, 27
cmpwi r17, 0x6800
Remove unnecessary static branch hints These + and - characters in a branch mnemonic can cause the assembler to produce conditional branch instructions with that hint the branch predictor. The default for forward branches is -, and for backward branches is +. If a mnemonic is issued with the opposite sign, then bit 10 of the instruction (the LSB of the BO field) is set. My long-retired "ppcdisasm" script inserted these hints unconditionally, despite 98% of them not being required. The code is much cleaner now. I read in some old MPW release notes that PPCAsm and PPCLink together exhibit a quirk when linking conditional branches to imported symbols. PPCAsm always assembles these conditional branches as if they were forward branches (that is, a + hint will always set the hint bit, and a - hint will never). I hoped to use this property to divine whether the NanoKernel was linked from one or many assembly files, but I was frustrated by the lack of conditional branches between files.
2018-04-15 07:05:02 +00:00
bne major_0x18040_0x100
Initial commit
2017-11-19 04:11:07 +00:00
lwz r17, 0x071c(r1)
srwi r17, r17, 16
andi. r17, r17, 0xffa0
cmpwi r17, 0x2a0
Remove unnecessary static branch hints These + and - characters in a branch mnemonic can cause the assembler to produce conditional branch instructions with that hint the branch predictor. The default for forward branches is -, and for backward branches is +. If a mnemonic is issued with the opposite sign, then bit 10 of the instruction (the LSB of the BO field) is set. My long-retired "ppcdisasm" script inserted these hints unconditionally, despite 98% of them not being required. The code is much cleaner now. I read in some old MPW release notes that PPCAsm and PPCLink together exhibit a quirk when linking conditional branches to imported symbols. PPCAsm always assembles these conditional branches as if they were forward branches (that is, a + hint will always set the hint bit, and a - hint will never). I hoped to use this property to divine whether the NanoKernel was linked from one or many assembly files, but I was frustrated by the lack of conditional branches between files.
2018-04-15 07:05:02 +00:00
bne major_0x18040_0x64
Initial commit
2017-11-19 04:11:07 +00:00
_log 'Caused by emulator termination request^n'
b major_0x18040_0x9c
major_0x18040_0x64
_log 'Caused by unhandled emulator exception^n'
major_0x18040_0x9c
lwz r17, 0x079c(r1)
lwz r18, 0x0704(r1)
subf r17, r18, r17
cmpwi r17, 0x100
cmpwi cr1, r17, -0x100
Remove unnecessary static branch hints These + and - characters in a branch mnemonic can cause the assembler to produce conditional branch instructions with that hint the branch predictor. The default for forward branches is -, and for backward branches is +. If a mnemonic is issued with the opposite sign, then bit 10 of the instruction (the LSB of the BO field) is set. My long-retired "ppcdisasm" script inserted these hints unconditionally, despite 98% of them not being required. The code is much cleaner now. I read in some old MPW release notes that PPCAsm and PPCLink together exhibit a quirk when linking conditional branches to imported symbols. PPCAsm always assembles these conditional branches as if they were forward branches (that is, a + hint will always set the hint bit, and a - hint will never). I hoped to use this property to divine whether the NanoKernel was linked from one or many assembly files, but I was frustrated by the lack of conditional branches between files.
2018-04-15 07:05:02 +00:00
bgt major_0x18040_0x100
blt cr1, major_0x18040_0x100
Initial commit
2017-11-19 04:11:07 +00:00
_log 'Looks like interrupt stack overflow by os or application^n'
major_0x18040_0x100
mtlr r16
blr
print_xpt_info ; OUTSIDE REFERER
mflr r16
lwz r18, 0x064c(r1)
llabel r18, NKBtm
add r19, r18, r19
_log 'Termination caller '
lwz r20, 0x0904(r1)
mr r8, r20
bl Printw
subf. r21, r18, r20
cmplw cr1, r20, r19
Remove unnecessary static branch hints These + and - characters in a branch mnemonic can cause the assembler to produce conditional branch instructions with that hint the branch predictor. The default for forward branches is -, and for backward branches is +. If a mnemonic is issued with the opposite sign, then bit 10 of the instruction (the LSB of the BO field) is set. My long-retired "ppcdisasm" script inserted these hints unconditionally, despite 98% of them not being required. The code is much cleaner now. I read in some old MPW release notes that PPCAsm and PPCLink together exhibit a quirk when linking conditional branches to imported symbols. PPCAsm always assembles these conditional branches as if they were forward branches (that is, a + hint will always set the hint bit, and a - hint will never). I hoped to use this property to divine whether the NanoKernel was linked from one or many assembly files, but I was frustrated by the lack of conditional branches between files.
2018-04-15 07:05:02 +00:00
blt print_xpt_info_0x84
bge cr1, print_xpt_info_0x84
Initial commit
2017-11-19 04:11:07 +00:00
_log '( NK+'
mr r8, r21
bl Printw
_log ')^n'
print_xpt_info_0x84
_log ' Last exception at '
mfspr r8, srr1
bl Printw
mfspr r8, srr0
bl Printw
mfspr r8, srr0
subf. r21, r18, r8
cmplw cr1, r8, r19
Remove unnecessary static branch hints These + and - characters in a branch mnemonic can cause the assembler to produce conditional branch instructions with that hint the branch predictor. The default for forward branches is -, and for backward branches is +. If a mnemonic is issued with the opposite sign, then bit 10 of the instruction (the LSB of the BO field) is set. My long-retired "ppcdisasm" script inserted these hints unconditionally, despite 98% of them not being required. The code is much cleaner now. I read in some old MPW release notes that PPCAsm and PPCLink together exhibit a quirk when linking conditional branches to imported symbols. PPCAsm always assembles these conditional branches as if they were forward branches (that is, a + hint will always set the hint bit, and a - hint will never). I hoped to use this property to divine whether the NanoKernel was linked from one or many assembly files, but I was frustrated by the lack of conditional branches between files.
2018-04-15 07:05:02 +00:00
blt print_xpt_info_0xf8
bge cr1, print_xpt_info_0xf8
Initial commit
2017-11-19 04:11:07 +00:00
_log '( NK+'
mr r8, r21
bl Printw
_log ')'
print_xpt_info_0xf8
_log '^n'
mtlr r16
blr
; print_sprgs
; Goldmine. Tells me what the SPRGs do!
print_sprgs ; OUTSIDE REFERER
mflr r16
_log 'SPRGs ewa: '
mfsprg r8, 0
bl Printw
_log ' r1: '
mfsprg r8, 1
bl Printw
_log ' lr: '
mfsprg r8, 2
bl Printw
_log ' vecBase: '
mfsprg r8, 3
bl Printw
_log '^n'
mtlr r16
blr
; print_sprs
; Both user-mode and supervisor-only
print_sprs ; OUTSIDE REFERER
mflr r16
_log ' cr: '
lwz r8, 0x0780(r1)
bl Printw
_log 'xer: '
lwz r8, 0x0788(r1)
bl Printw
_log 'ctr: '
lwz r8, 0x0790(r1)
bl Printw
_log 'lr: '
lwz r8, 0x078c(r1)
bl Printw
_log '^n dsisr: '
lwz r8, 0x0798(r1)
bl Printw
_log 'dar:'
lwz r8, 0x079c(r1)
bl Printw
_log 'pvr: '
lwz r8, 0x0794(r1)
bl Printw
_log '^n'
mtlr r16
blr
print_segment_registers ; OUTSIDE REFERER
mflr r16
_log ' sr0-sr7 '
li r17, 0x08
mtctr r17
li r18, 0x00
print_segment_registers_0x28
mfsrin r8, r18
addis r18, r18, 0x1000
bl Printw
Remove unnecessary static branch hints These + and - characters in a branch mnemonic can cause the assembler to produce conditional branch instructions with that hint the branch predictor. The default for forward branches is -, and for backward branches is +. If a mnemonic is issued with the opposite sign, then bit 10 of the instruction (the LSB of the BO field) is set. My long-retired "ppcdisasm" script inserted these hints unconditionally, despite 98% of them not being required. The code is much cleaner now. I read in some old MPW release notes that PPCAsm and PPCLink together exhibit a quirk when linking conditional branches to imported symbols. PPCAsm always assembles these conditional branches as if they were forward branches (that is, a + hint will always set the hint bit, and a - hint will never). I hoped to use this property to divine whether the NanoKernel was linked from one or many assembly files, but I was frustrated by the lack of conditional branches between files.
2018-04-15 07:05:02 +00:00
bdnz print_segment_registers_0x28
Initial commit
2017-11-19 04:11:07 +00:00
_log '^n sr8-sr15 '
li r17, 0x08
mtctr r17
print_segment_registers_0x5c
mfsrin r8, r18
addis r18, r18, 0x1000
bl Printw
Remove unnecessary static branch hints These + and - characters in a branch mnemonic can cause the assembler to produce conditional branch instructions with that hint the branch predictor. The default for forward branches is -, and for backward branches is +. If a mnemonic is issued with the opposite sign, then bit 10 of the instruction (the LSB of the BO field) is set. My long-retired "ppcdisasm" script inserted these hints unconditionally, despite 98% of them not being required. The code is much cleaner now. I read in some old MPW release notes that PPCAsm and PPCLink together exhibit a quirk when linking conditional branches to imported symbols. PPCAsm always assembles these conditional branches as if they were forward branches (that is, a + hint will always set the hint bit, and a - hint will never). I hoped to use this property to divine whether the NanoKernel was linked from one or many assembly files, but I was frustrated by the lack of conditional branches between files.
2018-04-15 07:05:02 +00:00
bdnz print_segment_registers_0x5c
Initial commit
2017-11-19 04:11:07 +00:00
_log '^n'
mtlr r16
blr
print_gprs ; OUTSIDE REFERER
mflr r16
addi r17, r1, 0x6fc
_log ' r0-r7 '
li r18, 0x08
mtctr r18
print_gprs_0x28
lwzu r8, 0x0004(r17)
bl Printw
Remove unnecessary static branch hints These + and - characters in a branch mnemonic can cause the assembler to produce conditional branch instructions with that hint the branch predictor. The default for forward branches is -, and for backward branches is +. If a mnemonic is issued with the opposite sign, then bit 10 of the instruction (the LSB of the BO field) is set. My long-retired "ppcdisasm" script inserted these hints unconditionally, despite 98% of them not being required. The code is much cleaner now. I read in some old MPW release notes that PPCAsm and PPCLink together exhibit a quirk when linking conditional branches to imported symbols. PPCAsm always assembles these conditional branches as if they were forward branches (that is, a + hint will always set the hint bit, and a - hint will never). I hoped to use this property to divine whether the NanoKernel was linked from one or many assembly files, but I was frustrated by the lack of conditional branches between files.
2018-04-15 07:05:02 +00:00
bdnz print_gprs_0x28
Initial commit
2017-11-19 04:11:07 +00:00
_log '^n r8-r15 '
li r18, 0x08
mtctr r18
print_gprs_0x58
lwzu r8, 0x0004(r17)
bl Printw
Remove unnecessary static branch hints These + and - characters in a branch mnemonic can cause the assembler to produce conditional branch instructions with that hint the branch predictor. The default for forward branches is -, and for backward branches is +. If a mnemonic is issued with the opposite sign, then bit 10 of the instruction (the LSB of the BO field) is set. My long-retired "ppcdisasm" script inserted these hints unconditionally, despite 98% of them not being required. The code is much cleaner now. I read in some old MPW release notes that PPCAsm and PPCLink together exhibit a quirk when linking conditional branches to imported symbols. PPCAsm always assembles these conditional branches as if they were forward branches (that is, a + hint will always set the hint bit, and a - hint will never). I hoped to use this property to divine whether the NanoKernel was linked from one or many assembly files, but I was frustrated by the lack of conditional branches between files.
2018-04-15 07:05:02 +00:00
bdnz print_gprs_0x58
Initial commit
2017-11-19 04:11:07 +00:00
_log '^n r16-r23 '
li r18, 0x08
mtctr r18
print_gprs_0x88
lwzu r8, 0x0004(r17)
bl Printw
Remove unnecessary static branch hints These + and - characters in a branch mnemonic can cause the assembler to produce conditional branch instructions with that hint the branch predictor. The default for forward branches is -, and for backward branches is +. If a mnemonic is issued with the opposite sign, then bit 10 of the instruction (the LSB of the BO field) is set. My long-retired "ppcdisasm" script inserted these hints unconditionally, despite 98% of them not being required. The code is much cleaner now. I read in some old MPW release notes that PPCAsm and PPCLink together exhibit a quirk when linking conditional branches to imported symbols. PPCAsm always assembles these conditional branches as if they were forward branches (that is, a + hint will always set the hint bit, and a - hint will never). I hoped to use this property to divine whether the NanoKernel was linked from one or many assembly files, but I was frustrated by the lack of conditional branches between files.
2018-04-15 07:05:02 +00:00
bdnz print_gprs_0x88
Initial commit
2017-11-19 04:11:07 +00:00
_log '^n r24-r31 '
li r18, 0x08
mtctr r18
print_gprs_0xb8
lwzu r8, 0x0004(r17)
bl Printw
Remove unnecessary static branch hints These + and - characters in a branch mnemonic can cause the assembler to produce conditional branch instructions with that hint the branch predictor. The default for forward branches is -, and for backward branches is +. If a mnemonic is issued with the opposite sign, then bit 10 of the instruction (the LSB of the BO field) is set. My long-retired "ppcdisasm" script inserted these hints unconditionally, despite 98% of them not being required. The code is much cleaner now. I read in some old MPW release notes that PPCAsm and PPCLink together exhibit a quirk when linking conditional branches to imported symbols. PPCAsm always assembles these conditional branches as if they were forward branches (that is, a + hint will always set the hint bit, and a - hint will never). I hoped to use this property to divine whether the NanoKernel was linked from one or many assembly files, but I was frustrated by the lack of conditional branches between files.
2018-04-15 07:05:02 +00:00
bdnz print_gprs_0xb8
Initial commit
2017-11-19 04:11:07 +00:00
_log '^n'
mtlr r16
blr
print_memory ; OUTSIDE REFERER
mflr r18
srwi r17, r17, 4
print_memory_0x8
mr r8, r16
bl Printw
_log ' '
lwz r8, 0x0000(r16)
bl Printw
lwz r8, 0x0004(r16)
bl Printw
lwz r8, 0x0008(r16)
bl Printw
lwz r8, 0x000c(r16)
bl Printw
_log ' *'
li r8, 0x10
addi r16, r16, -0x01
mtctr r8
print_memory_0x60
lbzu r8, 0x0001(r16)
cmpwi r8, 0xff
Remove unnecessary static branch hints These + and - characters in a branch mnemonic can cause the assembler to produce conditional branch instructions with that hint the branch predictor. The default for forward branches is -, and for backward branches is +. If a mnemonic is issued with the opposite sign, then bit 10 of the instruction (the LSB of the BO field) is set. My long-retired "ppcdisasm" script inserted these hints unconditionally, despite 98% of them not being required. The code is much cleaner now. I read in some old MPW release notes that PPCAsm and PPCLink together exhibit a quirk when linking conditional branches to imported symbols. PPCAsm always assembles these conditional branches as if they were forward branches (that is, a + hint will always set the hint bit, and a - hint will never). I hoped to use this property to divine whether the NanoKernel was linked from one or many assembly files, but I was frustrated by the lack of conditional branches between files.
2018-04-15 07:05:02 +00:00
beq print_memory_0x74
Initial commit
2017-11-19 04:11:07 +00:00
cmpwi r8, 0x20
Remove unnecessary static branch hints These + and - characters in a branch mnemonic can cause the assembler to produce conditional branch instructions with that hint the branch predictor. The default for forward branches is -, and for backward branches is +. If a mnemonic is issued with the opposite sign, then bit 10 of the instruction (the LSB of the BO field) is set. My long-retired "ppcdisasm" script inserted these hints unconditionally, despite 98% of them not being required. The code is much cleaner now. I read in some old MPW release notes that PPCAsm and PPCLink together exhibit a quirk when linking conditional branches to imported symbols. PPCAsm always assembles these conditional branches as if they were forward branches (that is, a + hint will always set the hint bit, and a - hint will never). I hoped to use this property to divine whether the NanoKernel was linked from one or many assembly files, but I was frustrated by the lack of conditional branches between files.
2018-04-15 07:05:02 +00:00
bgt print_memory_0x78
Initial commit
2017-11-19 04:11:07 +00:00
print_memory_0x74
li r8, 0x20
print_memory_0x78
bl Printc
Remove unnecessary static branch hints These + and - characters in a branch mnemonic can cause the assembler to produce conditional branch instructions with that hint the branch predictor. The default for forward branches is -, and for backward branches is +. If a mnemonic is issued with the opposite sign, then bit 10 of the instruction (the LSB of the BO field) is set. My long-retired "ppcdisasm" script inserted these hints unconditionally, despite 98% of them not being required. The code is much cleaner now. I read in some old MPW release notes that PPCAsm and PPCLink together exhibit a quirk when linking conditional branches to imported symbols. PPCAsm always assembles these conditional branches as if they were forward branches (that is, a + hint will always set the hint bit, and a - hint will never). I hoped to use this property to divine whether the NanoKernel was linked from one or many assembly files, but I was frustrated by the lack of conditional branches between files.
2018-04-15 07:05:02 +00:00
bdnz print_memory_0x60
Initial commit
2017-11-19 04:11:07 +00:00
_log '*^n'
addi r16, r16, 0x01
addi r17, r17, -0x01
bl getchar
cmpwi r8, -0x01
Remove unnecessary static branch hints These + and - characters in a branch mnemonic can cause the assembler to produce conditional branch instructions with that hint the branch predictor. The default for forward branches is -, and for backward branches is +. If a mnemonic is issued with the opposite sign, then bit 10 of the instruction (the LSB of the BO field) is set. My long-retired "ppcdisasm" script inserted these hints unconditionally, despite 98% of them not being required. The code is much cleaner now. I read in some old MPW release notes that PPCAsm and PPCLink together exhibit a quirk when linking conditional branches to imported symbols. PPCAsm always assembles these conditional branches as if they were forward branches (that is, a + hint will always set the hint bit, and a - hint will never). I hoped to use this property to divine whether the NanoKernel was linked from one or many assembly files, but I was frustrated by the lack of conditional branches between files.
2018-04-15 07:05:02 +00:00
bne print_memory_0xb0
Initial commit
2017-11-19 04:11:07 +00:00
cmpwi r17, 0x00
Remove unnecessary static branch hints These + and - characters in a branch mnemonic can cause the assembler to produce conditional branch instructions with that hint the branch predictor. The default for forward branches is -, and for backward branches is +. If a mnemonic is issued with the opposite sign, then bit 10 of the instruction (the LSB of the BO field) is set. My long-retired "ppcdisasm" script inserted these hints unconditionally, despite 98% of them not being required. The code is much cleaner now. I read in some old MPW release notes that PPCAsm and PPCLink together exhibit a quirk when linking conditional branches to imported symbols. PPCAsm always assembles these conditional branches as if they were forward branches (that is, a + hint will always set the hint bit, and a - hint will never). I hoped to use this property to divine whether the NanoKernel was linked from one or many assembly files, but I was frustrated by the lack of conditional branches between files.
2018-04-15 07:05:02 +00:00
bne print_memory_0x8
Initial commit
2017-11-19 04:11:07 +00:00
print_memory_0xb0
_log '^n'
mtlr r18
blr
print_memory_logical ; OUTSIDE REFERER
mflr r18
srwi r17, r17, 4
print_memory_logical_0x8
mr r8, r16
bl Printw
_log ' '
li r19, 0x10
print_memory_logical_0x24
mr r27, r16
bl PagingFunc1
Remove unnecessary static branch hints These + and - characters in a branch mnemonic can cause the assembler to produce conditional branch instructions with that hint the branch predictor. The default for forward branches is -, and for backward branches is +. If a mnemonic is issued with the opposite sign, then bit 10 of the instruction (the LSB of the BO field) is set. My long-retired "ppcdisasm" script inserted these hints unconditionally, despite 98% of them not being required. The code is much cleaner now. I read in some old MPW release notes that PPCAsm and PPCLink together exhibit a quirk when linking conditional branches to imported symbols. PPCAsm always assembles these conditional branches as if they were forward branches (that is, a + hint will always set the hint bit, and a - hint will never). I hoped to use this property to divine whether the NanoKernel was linked from one or many assembly files, but I was frustrated by the lack of conditional branches between files.
2018-04-15 07:05:02 +00:00
beq print_memory_logical_0x5c
blt print_memory_logical_0x48
Initial commit
2017-11-19 04:11:07 +00:00
_log '..'
b print_memory_logical_0x6c
print_memory_logical_0x48
_log '--'
b print_memory_logical_0x6c
print_memory_logical_0x5c
Reverse PPC-68k context switching (incomplete)
2018-03-31 08:17:23 +00:00
bl PagingL2PWithoutBATs
Initial commit
2017-11-19 04:11:07 +00:00
rlwimi r31, r27, 0, 20, 31
lbz r8, 0x0000(r31)
bl print_unknown
print_memory_logical_0x6c
addi r16, r16, 0x01
addi r19, r19, -0x01
andi. r8, r19, 0x03
Remove unnecessary static branch hints These + and - characters in a branch mnemonic can cause the assembler to produce conditional branch instructions with that hint the branch predictor. The default for forward branches is -, and for backward branches is +. If a mnemonic is issued with the opposite sign, then bit 10 of the instruction (the LSB of the BO field) is set. My long-retired "ppcdisasm" script inserted these hints unconditionally, despite 98% of them not being required. The code is much cleaner now. I read in some old MPW release notes that PPCAsm and PPCLink together exhibit a quirk when linking conditional branches to imported symbols. PPCAsm always assembles these conditional branches as if they were forward branches (that is, a + hint will always set the hint bit, and a - hint will never). I hoped to use this property to divine whether the NanoKernel was linked from one or many assembly files, but I was frustrated by the lack of conditional branches between files.
2018-04-15 07:05:02 +00:00
bne print_memory_logical_0x84
Initial commit
2017-11-19 04:11:07 +00:00
li r8, 0x20
bl Printc
print_memory_logical_0x84
cmpwi r19, 0x00
Remove unnecessary static branch hints These + and - characters in a branch mnemonic can cause the assembler to produce conditional branch instructions with that hint the branch predictor. The default for forward branches is -, and for backward branches is +. If a mnemonic is issued with the opposite sign, then bit 10 of the instruction (the LSB of the BO field) is set. My long-retired "ppcdisasm" script inserted these hints unconditionally, despite 98% of them not being required. The code is much cleaner now. I read in some old MPW release notes that PPCAsm and PPCLink together exhibit a quirk when linking conditional branches to imported symbols. PPCAsm always assembles these conditional branches as if they were forward branches (that is, a + hint will always set the hint bit, and a - hint will never). I hoped to use this property to divine whether the NanoKernel was linked from one or many assembly files, but I was frustrated by the lack of conditional branches between files.
2018-04-15 07:05:02 +00:00
bgt print_memory_logical_0x24
Initial commit
2017-11-19 04:11:07 +00:00
_log ' *'
li r8, 0x10
addi r16, r16, -0x10
mtctr r8
print_memory_logical_0xac
mr r27, r16
bl PagingFunc1
li r8, 0x20
Remove unnecessary static branch hints These + and - characters in a branch mnemonic can cause the assembler to produce conditional branch instructions with that hint the branch predictor. The default for forward branches is -, and for backward branches is +. If a mnemonic is issued with the opposite sign, then bit 10 of the instruction (the LSB of the BO field) is set. My long-retired "ppcdisasm" script inserted these hints unconditionally, despite 98% of them not being required. The code is much cleaner now. I read in some old MPW release notes that PPCAsm and PPCLink together exhibit a quirk when linking conditional branches to imported symbols. PPCAsm always assembles these conditional branches as if they were forward branches (that is, a + hint will always set the hint bit, and a - hint will never). I hoped to use this property to divine whether the NanoKernel was linked from one or many assembly files, but I was frustrated by the lack of conditional branches between files.
2018-04-15 07:05:02 +00:00
bne print_memory_logical_0xdc
Reverse PPC-68k context switching (incomplete)
2018-03-31 08:17:23 +00:00
bl PagingL2PWithoutBATs
Initial commit
2017-11-19 04:11:07 +00:00
rlwimi r31, r27, 0, 20, 31
lbz r8, 0x0000(r31)
cmpwi r8, 0xff
Remove unnecessary static branch hints These + and - characters in a branch mnemonic can cause the assembler to produce conditional branch instructions with that hint the branch predictor. The default for forward branches is -, and for backward branches is +. If a mnemonic is issued with the opposite sign, then bit 10 of the instruction (the LSB of the BO field) is set. My long-retired "ppcdisasm" script inserted these hints unconditionally, despite 98% of them not being required. The code is much cleaner now. I read in some old MPW release notes that PPCAsm and PPCLink together exhibit a quirk when linking conditional branches to imported symbols. PPCAsm always assembles these conditional branches as if they were forward branches (that is, a + hint will always set the hint bit, and a - hint will never). I hoped to use this property to divine whether the NanoKernel was linked from one or many assembly files, but I was frustrated by the lack of conditional branches between files.
2018-04-15 07:05:02 +00:00
beq print_memory_logical_0xd8
Initial commit
2017-11-19 04:11:07 +00:00
cmpwi r8, 0x20
Remove unnecessary static branch hints These + and - characters in a branch mnemonic can cause the assembler to produce conditional branch instructions with that hint the branch predictor. The default for forward branches is -, and for backward branches is +. If a mnemonic is issued with the opposite sign, then bit 10 of the instruction (the LSB of the BO field) is set. My long-retired "ppcdisasm" script inserted these hints unconditionally, despite 98% of them not being required. The code is much cleaner now. I read in some old MPW release notes that PPCAsm and PPCLink together exhibit a quirk when linking conditional branches to imported symbols. PPCAsm always assembles these conditional branches as if they were forward branches (that is, a + hint will always set the hint bit, and a - hint will never). I hoped to use this property to divine whether the NanoKernel was linked from one or many assembly files, but I was frustrated by the lack of conditional branches between files.
2018-04-15 07:05:02 +00:00
bgt print_memory_logical_0xdc
Initial commit
2017-11-19 04:11:07 +00:00
print_memory_logical_0xd8
li r8, 0x20
print_memory_logical_0xdc
bl Printc
addi r16, r16, 0x01
Remove unnecessary static branch hints These + and - characters in a branch mnemonic can cause the assembler to produce conditional branch instructions with that hint the branch predictor. The default for forward branches is -, and for backward branches is +. If a mnemonic is issued with the opposite sign, then bit 10 of the instruction (the LSB of the BO field) is set. My long-retired "ppcdisasm" script inserted these hints unconditionally, despite 98% of them not being required. The code is much cleaner now. I read in some old MPW release notes that PPCAsm and PPCLink together exhibit a quirk when linking conditional branches to imported symbols. PPCAsm always assembles these conditional branches as if they were forward branches (that is, a + hint will always set the hint bit, and a - hint will never). I hoped to use this property to divine whether the NanoKernel was linked from one or many assembly files, but I was frustrated by the lack of conditional branches between files.
2018-04-15 07:05:02 +00:00
bdnz print_memory_logical_0xac
Initial commit
2017-11-19 04:11:07 +00:00
_log '*^n'
addi r17, r17, -0x01
bl getchar
cmpwi r8, -0x01
Remove unnecessary static branch hints These + and - characters in a branch mnemonic can cause the assembler to produce conditional branch instructions with that hint the branch predictor. The default for forward branches is -, and for backward branches is +. If a mnemonic is issued with the opposite sign, then bit 10 of the instruction (the LSB of the BO field) is set. My long-retired "ppcdisasm" script inserted these hints unconditionally, despite 98% of them not being required. The code is much cleaner now. I read in some old MPW release notes that PPCAsm and PPCLink together exhibit a quirk when linking conditional branches to imported symbols. PPCAsm always assembles these conditional branches as if they were forward branches (that is, a + hint will always set the hint bit, and a - hint will never). I hoped to use this property to divine whether the NanoKernel was linked from one or many assembly files, but I was frustrated by the lack of conditional branches between files.
2018-04-15 07:05:02 +00:00
bne print_memory_logical_0x114
Initial commit
2017-11-19 04:11:07 +00:00
cmpwi r17, 0x00
Remove unnecessary static branch hints These + and - characters in a branch mnemonic can cause the assembler to produce conditional branch instructions with that hint the branch predictor. The default for forward branches is -, and for backward branches is +. If a mnemonic is issued with the opposite sign, then bit 10 of the instruction (the LSB of the BO field) is set. My long-retired "ppcdisasm" script inserted these hints unconditionally, despite 98% of them not being required. The code is much cleaner now. I read in some old MPW release notes that PPCAsm and PPCLink together exhibit a quirk when linking conditional branches to imported symbols. PPCAsm always assembles these conditional branches as if they were forward branches (that is, a + hint will always set the hint bit, and a - hint will never). I hoped to use this property to divine whether the NanoKernel was linked from one or many assembly files, but I was frustrated by the lack of conditional branches between files.
2018-04-15 07:05:02 +00:00
bne print_memory_logical_0x8
Initial commit
2017-11-19 04:11:07 +00:00
print_memory_logical_0x114
_log '^n'
mtlr r18
blr
; > r16 = command strings
; > r17 = lut
; < cr0 = found
; < r17 = ptr to lut entry
cmd_lookup ; OUTSIDE REFERER
addi r15, r15, -0x01
addi r16, r16, -0x01
mr r18, r15
cmd_lookup_0xc
lbzu r21, 0x0001(r16)
lbzu r20, 0x0001(r15)
cmpwi r21, 0xff
cmpwi cr1, r21, 0x00
Remove unnecessary static branch hints These + and - characters in a branch mnemonic can cause the assembler to produce conditional branch instructions with that hint the branch predictor. The default for forward branches is -, and for backward branches is +. If a mnemonic is issued with the opposite sign, then bit 10 of the instruction (the LSB of the BO field) is set. My long-retired "ppcdisasm" script inserted these hints unconditionally, despite 98% of them not being required. The code is much cleaner now. I read in some old MPW release notes that PPCAsm and PPCLink together exhibit a quirk when linking conditional branches to imported symbols. PPCAsm always assembles these conditional branches as if they were forward branches (that is, a + hint will always set the hint bit, and a - hint will never). I hoped to use this property to divine whether the NanoKernel was linked from one or many assembly files, but I was frustrated by the lack of conditional branches between files.
2018-04-15 07:05:02 +00:00
beq cmd_lookup_0x44
beq cr1, cmd_lookup_0x50
Initial commit
2017-11-19 04:11:07 +00:00
cmpw r20, r21
Remove unnecessary static branch hints These + and - characters in a branch mnemonic can cause the assembler to produce conditional branch instructions with that hint the branch predictor. The default for forward branches is -, and for backward branches is +. If a mnemonic is issued with the opposite sign, then bit 10 of the instruction (the LSB of the BO field) is set. My long-retired "ppcdisasm" script inserted these hints unconditionally, despite 98% of them not being required. The code is much cleaner now. I read in some old MPW release notes that PPCAsm and PPCLink together exhibit a quirk when linking conditional branches to imported symbols. PPCAsm always assembles these conditional branches as if they were forward branches (that is, a + hint will always set the hint bit, and a - hint will never). I hoped to use this property to divine whether the NanoKernel was linked from one or many assembly files, but I was frustrated by the lack of conditional branches between files.
2018-04-15 07:05:02 +00:00
beq cmd_lookup_0xc
Initial commit
2017-11-19 04:11:07 +00:00
cmd_lookup_0x2c
lbzu r21, 0x0001(r16)
cmpwi r21, 0x00
Remove unnecessary static branch hints These + and - characters in a branch mnemonic can cause the assembler to produce conditional branch instructions with that hint the branch predictor. The default for forward branches is -, and for backward branches is +. If a mnemonic is issued with the opposite sign, then bit 10 of the instruction (the LSB of the BO field) is set. My long-retired "ppcdisasm" script inserted these hints unconditionally, despite 98% of them not being required. The code is much cleaner now. I read in some old MPW release notes that PPCAsm and PPCLink together exhibit a quirk when linking conditional branches to imported symbols. PPCAsm always assembles these conditional branches as if they were forward branches (that is, a + hint will always set the hint bit, and a - hint will never). I hoped to use this property to divine whether the NanoKernel was linked from one or many assembly files, but I was frustrated by the lack of conditional branches between files.
2018-04-15 07:05:02 +00:00
bne cmd_lookup_0x2c
Initial commit
2017-11-19 04:11:07 +00:00
cmd_lookup_0x38
addi r17, r17, 0x04
mr r15, r18
b cmd_lookup_0xc
cmd_lookup_0x44
addi r15, r18, 0x01
cmpw r15, r18
blr
cmd_lookup_0x50
cmpwi r20, 0x20
Remove unnecessary static branch hints These + and - characters in a branch mnemonic can cause the assembler to produce conditional branch instructions with that hint the branch predictor. The default for forward branches is -, and for backward branches is +. If a mnemonic is issued with the opposite sign, then bit 10 of the instruction (the LSB of the BO field) is set. My long-retired "ppcdisasm" script inserted these hints unconditionally, despite 98% of them not being required. The code is much cleaner now. I read in some old MPW release notes that PPCAsm and PPCLink together exhibit a quirk when linking conditional branches to imported symbols. PPCAsm always assembles these conditional branches as if they were forward branches (that is, a + hint will always set the hint bit, and a - hint will never). I hoped to use this property to divine whether the NanoKernel was linked from one or many assembly files, but I was frustrated by the lack of conditional branches between files.
2018-04-15 07:05:02 +00:00
beqlr
Initial commit
2017-11-19 04:11:07 +00:00
cmpwi r20, 0x00
Remove unnecessary static branch hints These + and - characters in a branch mnemonic can cause the assembler to produce conditional branch instructions with that hint the branch predictor. The default for forward branches is -, and for backward branches is +. If a mnemonic is issued with the opposite sign, then bit 10 of the instruction (the LSB of the BO field) is set. My long-retired "ppcdisasm" script inserted these hints unconditionally, despite 98% of them not being required. The code is much cleaner now. I read in some old MPW release notes that PPCAsm and PPCLink together exhibit a quirk when linking conditional branches to imported symbols. PPCAsm always assembles these conditional branches as if they were forward branches (that is, a + hint will always set the hint bit, and a - hint will never). I hoped to use this property to divine whether the NanoKernel was linked from one or many assembly files, but I was frustrated by the lack of conditional branches between files.
2018-04-15 07:05:02 +00:00
beqlr
Initial commit
2017-11-19 04:11:07 +00:00
b cmd_lookup_0x38
; > r15 = start
; < r15 = ptr
; < r16 = char
next_cmd_word ; OUTSIDE REFERER
addi r15, r15, -0x01
next_cmd_word_0x4
lbzu r16, 0x0001(r15)
cmpwi r16, 0x20
Remove unnecessary static branch hints These + and - characters in a branch mnemonic can cause the assembler to produce conditional branch instructions with that hint the branch predictor. The default for forward branches is -, and for backward branches is +. If a mnemonic is issued with the opposite sign, then bit 10 of the instruction (the LSB of the BO field) is set. My long-retired "ppcdisasm" script inserted these hints unconditionally, despite 98% of them not being required. The code is much cleaner now. I read in some old MPW release notes that PPCAsm and PPCLink together exhibit a quirk when linking conditional branches to imported symbols. PPCAsm always assembles these conditional branches as if they were forward branches (that is, a + hint will always set the hint bit, and a - hint will never). I hoped to use this property to divine whether the NanoKernel was linked from one or many assembly files, but I was frustrated by the lack of conditional branches between files.
2018-04-15 07:05:02 +00:00
beq next_cmd_word_0x4
Initial commit
2017-11-19 04:11:07 +00:00
blr
major_0x187b0 ; OUTSIDE REFERER
addi r15, r15, -0x01
li r16, 0x00
major_0x187b0_0x8
lbzu r17, 0x0001(r15)
cmplwi r17, 0x30
cmplwi cr1, r17, 0x39
Remove unnecessary static branch hints These + and - characters in a branch mnemonic can cause the assembler to produce conditional branch instructions with that hint the branch predictor. The default for forward branches is -, and for backward branches is +. If a mnemonic is issued with the opposite sign, then bit 10 of the instruction (the LSB of the BO field) is set. My long-retired "ppcdisasm" script inserted these hints unconditionally, despite 98% of them not being required. The code is much cleaner now. I read in some old MPW release notes that PPCAsm and PPCLink together exhibit a quirk when linking conditional branches to imported symbols. PPCAsm always assembles these conditional branches as if they were forward branches (that is, a + hint will always set the hint bit, and a - hint will never). I hoped to use this property to divine whether the NanoKernel was linked from one or many assembly files, but I was frustrated by the lack of conditional branches between files.
2018-04-15 07:05:02 +00:00
blt major_0x187b0_0x28
bgt cr1, major_0x187b0_0x28
Initial commit
2017-11-19 04:11:07 +00:00
slwi r16, r16, 4
rlwimi r16, r17, 0, 28, 31
b major_0x187b0_0x8
major_0x187b0_0x28
cmplwi r17, 0x61
cmplwi cr1, r17, 0x66
Remove unnecessary static branch hints These + and - characters in a branch mnemonic can cause the assembler to produce conditional branch instructions with that hint the branch predictor. The default for forward branches is -, and for backward branches is +. If a mnemonic is issued with the opposite sign, then bit 10 of the instruction (the LSB of the BO field) is set. My long-retired "ppcdisasm" script inserted these hints unconditionally, despite 98% of them not being required. The code is much cleaner now. I read in some old MPW release notes that PPCAsm and PPCLink together exhibit a quirk when linking conditional branches to imported symbols. PPCAsm always assembles these conditional branches as if they were forward branches (that is, a + hint will always set the hint bit, and a - hint will never). I hoped to use this property to divine whether the NanoKernel was linked from one or many assembly files, but I was frustrated by the lack of conditional branches between files.
2018-04-15 07:05:02 +00:00
blt major_0x187b0_0x48
bgt cr1, major_0x187b0_0x48
Initial commit
2017-11-19 04:11:07 +00:00
addi r17, r17, -0x57
slwi r16, r16, 4
rlwimi r16, r17, 0, 28, 31
b major_0x187b0_0x8
major_0x187b0_0x48
cmplwi r17, 0x41
cmplwi cr1, r17, 0x46
Remove unnecessary static branch hints These + and - characters in a branch mnemonic can cause the assembler to produce conditional branch instructions with that hint the branch predictor. The default for forward branches is -, and for backward branches is +. If a mnemonic is issued with the opposite sign, then bit 10 of the instruction (the LSB of the BO field) is set. My long-retired "ppcdisasm" script inserted these hints unconditionally, despite 98% of them not being required. The code is much cleaner now. I read in some old MPW release notes that PPCAsm and PPCLink together exhibit a quirk when linking conditional branches to imported symbols. PPCAsm always assembles these conditional branches as if they were forward branches (that is, a + hint will always set the hint bit, and a - hint will never). I hoped to use this property to divine whether the NanoKernel was linked from one or many assembly files, but I was frustrated by the lack of conditional branches between files.
2018-04-15 07:05:02 +00:00
blt major_0x187b0_0x68
bgt cr1, major_0x187b0_0x68
Initial commit
2017-11-19 04:11:07 +00:00
addi r17, r17, -0x37
slwi r16, r16, 4
rlwimi r16, r17, 0, 28, 31
b major_0x187b0_0x8
major_0x187b0_0x68
cmpwi r17, 0x00
Remove unnecessary static branch hints These + and - characters in a branch mnemonic can cause the assembler to produce conditional branch instructions with that hint the branch predictor. The default for forward branches is -, and for backward branches is +. If a mnemonic is issued with the opposite sign, then bit 10 of the instruction (the LSB of the BO field) is set. My long-retired "ppcdisasm" script inserted these hints unconditionally, despite 98% of them not being required. The code is much cleaner now. I read in some old MPW release notes that PPCAsm and PPCLink together exhibit a quirk when linking conditional branches to imported symbols. PPCAsm always assembles these conditional branches as if they were forward branches (that is, a + hint will always set the hint bit, and a - hint will never). I hoped to use this property to divine whether the NanoKernel was linked from one or many assembly files, but I was frustrated by the lack of conditional branches between files.
2018-04-15 07:05:02 +00:00
beqlr
Initial commit
2017-11-19 04:11:07 +00:00
cmpwi r17, 0x20
blr
prereturn ; OUTSIDE REFERER
lwz r1, EWA.PA_KDP(r1)
mfmsr r0
_bset r0, r0, MSR_FPbit
mtmsr r0
isync
lfd f31, 0x08fc(r1)
mtfsf 0xff, f31
lfd f0, 0x0800(r1)
lfd f1, 0x0808(r1)
lfd f2, 0x0810(r1)
lfd f3, 0x0818(r1)
lfd f4, 0x0820(r1)
lfd f5, 0x0828(r1)
lfd f6, 0x0830(r1)
lfd f7, 0x0838(r1)
lfd f8, 0x0840(r1)
lfd f9, 0x0848(r1)
lfd f10, 0x0850(r1)
lfd f11, 0x0858(r1)
lfd f12, 0x0860(r1)
lfd f13, 0x0868(r1)
lfd f14, 0x0870(r1)
lfd f15, 0x0878(r1)
lfd f16, 0x0880(r1)
lfd f17, 0x0888(r1)
lfd f18, 0x0890(r1)
lfd f19, 0x0898(r1)
lfd f20, 0x08a0(r1)
lfd f21, 0x08a8(r1)
lfd f22, 0x08b0(r1)
lfd f23, 0x08b8(r1)
lfd f24, 0x08c0(r1)
lfd f25, 0x08c8(r1)
lfd f26, 0x08d0(r1)
lfd f27, 0x08d8(r1)
lfd f28, 0x08e0(r1)
lfd f29, 0x08e8(r1)
lfd f30, 0x08f0(r1)
lfd f31, 0x08f8(r1)
lwz r0, 0x07c0(r1)
mtsr 0x00, r0
lwz r0, 0x07c4(r1)
mtsr 0x01, r0
lwz r0, 0x07c8(r1)
mtsr 0x02, r0
lwz r0, 0x07cc(r1)
mtsr 0x03, r0
lwz r0, 0x07d0(r1)
mtsr 0x04, r0
lwz r0, 0x07d4(r1)
mtsr 0x05, r0
lwz r0, 0x07d8(r1)
mtsr 0x06, r0
lwz r0, 0x07dc(r1)
mtsr 0x07, r0
lwz r0, 0x07e0(r1)
mtsr 0x08, r0
lwz r0, 0x07e4(r1)
mtsr 0x09, r0
lwz r0, 0x07e8(r1)
mtsr 0x0a, r0
lwz r0, 0x07ec(r1)
mtsr 0x0b, r0
lwz r0, 0x07f0(r1)
mtsr 0x0c, r0
lwz r0, 0x07f4(r1)
mtsr 0x0d, r0
lwz r0, 0x07f8(r1)
mtsr 0x0e, r0
lwz r0, 0x07fc(r1)
mtsr 0x0f, r0
lwz r0, 0x07a8(r1)
mtspr dec, r0
lwz r0, 0x07b4(r1)
mtspr srr0, r0
lwz r0, 0x07b8(r1)
mtspr srr1, r0
lwz r0, 0x07bc(r1)
mtmsr r0
mfpvr r0
rlwinm. r0, r0, 0, 0, 14
Remove unnecessary static branch hints These + and - characters in a branch mnemonic can cause the assembler to produce conditional branch instructions with that hint the branch predictor. The default for forward branches is -, and for backward branches is +. If a mnemonic is issued with the opposite sign, then bit 10 of the instruction (the LSB of the BO field) is set. My long-retired "ppcdisasm" script inserted these hints unconditionally, despite 98% of them not being required. The code is much cleaner now. I read in some old MPW release notes that PPCAsm and PPCLink together exhibit a quirk when linking conditional branches to imported symbols. PPCAsm always assembles these conditional branches as if they were forward branches (that is, a + hint will always set the hint bit, and a - hint will never). I hoped to use this property to divine whether the NanoKernel was linked from one or many assembly files, but I was frustrated by the lack of conditional branches between files.
2018-04-15 07:05:02 +00:00
bne @not_601
Initial commit
2017-11-19 04:11:07 +00:00
lwz r0, 0x0784(r1)
mtspr mq, r0
@not_601
lwz r0, 0x0788(r1)
mtxer r0
lwz r0, 0x078c(r1)
mtsprg 2, r0
lwz r0, 0x0790(r1)
mtctr r0
; Only because this crashes QEMU
if &TYPE('NKDebugShim') = 'UNDEFINED'
lwz r0, 0x0794(r1)
mtspr pvr, r0
endif
lwz r0, 0x0798(r1)
mtspr dsisr, r0
lwz r0, 0x079c(r1)
mtspr dar, r0
lwz r0, 0x0780(r1)
mtcr r0
lwz r0, 0x0700(r1)
lwz r2, 0x0704(r1)
mtsprg 1, r2
lmw r2, 0x0708(r1)
blr
align 5
Reference in New Issue Copy Permalink