mirror of
https://github.com/ZDoom/gzdoom-gles.git
synced 2024-11-28 06:53:40 +00:00
- Changed the four-byte fillers in asm_x86_64/tmap3.s from 0x88888888 to 0x44444444 because
newer version of gas complained about it. SVN r2481 (trunk)
This commit is contained in:
Randy Heit
parent
677d07f837
commit
8ce9e178d4
1 changed files with 42 additions and 42 deletions
|
|
@ -1,28 +1,28 @@
|
|||
#%include "valgrind.inc"
|
||||
#%include "valgrind.inc"
|
||||
|
||||
.section .text
|
||||
.section .text
|
||||
|
||||
.globl ASM_PatchPitch
|
||||
ASM_PatchPitch:
|
||||
.globl ASM_PatchPitch
|
||||
ASM_PatchPitch:
|
||||
movl dc_pitch(%rip), %ecx
|
||||
movl %ecx, pm+3(%rip)
|
||||
movl %ecx, vltpitch+3(%rip)
|
||||
# selfmod pm, vltpitch+6
|
||||
ret
|
||||
.align 16
|
||||
movl %ecx, vltpitch+3(%rip)
|
||||
# selfmod pm, vltpitch+6
|
||||
ret
|
||||
.align 16
|
||||
|
||||
.globl setupvlinetallasm
|
||||
.globl setupvlinetallasm
|
||||
setupvlinetallasm:
|
||||
movb %dil, shifter1+2(%rip)
|
||||
movb %dil, shifter2+2(%rip)
|
||||
movb %dil, shifter3+2(%rip)
|
||||
movb %dil, shifter4+2(%rip)
|
||||
# selfmod shifter1, shifter4+3
|
||||
movb %dil, shifter4+2(%rip)
|
||||
# selfmod shifter1, shifter4+3
|
||||
ret
|
||||
.align 16
|
||||
|
||||
.section .rtext,"awx"
|
||||
|
||||
.align 16
|
||||
|
||||
.section .rtext,"awx"
|
||||
|
||||
.globl vlinetallasm4
|
||||
.type vlinetallasm4,@function
|
||||
vlinetallasm4:
|
||||
|
|
@ -38,18 +38,18 @@ vlinetallasm4:
|
|||
subq $8, %rsp # Does the stack need to be 16-byte aligned for Linux?
|
||||
.cfi_adjust_cfa_offset 8
|
||||
|
||||
# rax = bufplce base address
|
||||
# rbx =
|
||||
# rcx = offset from rdi/count (negative)
|
||||
# edx/rdx = scratch
|
||||
# rdi = bottom of columns to write to
|
||||
# r8d-r11d = column offsets
|
||||
# r12-r15 = palookupoffse[0] - palookupoffse[4]
|
||||
# rax = bufplce base address
|
||||
# rbx =
|
||||
# rcx = offset from rdi/count (negative)
|
||||
# edx/rdx = scratch
|
||||
# rdi = bottom of columns to write to
|
||||
# r8d-r11d = column offsets
|
||||
# r12-r15 = palookupoffse[0] - palookupoffse[4]
|
||||
|
||||
movl dc_count(%rip), %ecx
|
||||
movq dc_dest(%rip), %rdi
|
||||
testl %ecx, %ecx
|
||||
jle vltepilog # count must be positive
|
||||
jle vltepilog # count must be positive
|
||||
|
||||
movq bufplce(%rip), %rax
|
||||
movq bufplce+8(%rip), %r8
|
||||
|
|
@ -60,14 +60,14 @@ vlinetallasm4:
|
|||
subq %rax, %r10
|
||||
movl %r8d, source2+4(%rip)
|
||||
movl %r9d, source3+4(%rip)
|
||||
movl %r10d, source4+4(%rip)
|
||||
movl %r10d, source4+4(%rip)
|
||||
|
||||
pm: imulq $320, %rcx
|
||||
pm: imulq $320, %rcx
|
||||
|
||||
movq palookupoffse(%rip), %r12
|
||||
movq palookupoffse+8(%rip), %r13
|
||||
movq palookupoffse+16(%rip), %r14
|
||||
movq palookupoffse+24(%rip), %r15
|
||||
movq palookupoffse+24(%rip), %r15
|
||||
|
||||
movl vince(%rip), %r8d
|
||||
movl vince+4(%rip), %r9d
|
||||
|
|
@ -76,53 +76,53 @@ pm: imulq $320, %rcx
|
|||
movl %r8d, step1+3(%rip)
|
||||
movl %r9d, step2+3(%rip)
|
||||
movl %r10d, step3+3(%rip)
|
||||
movl %r11d, step4+3(%rip)
|
||||
movl %r11d, step4+3(%rip)
|
||||
|
||||
addq %rcx, %rdi
|
||||
negq %rcx
|
||||
negq %rcx
|
||||
|
||||
movl vplce(%rip), %r8d
|
||||
movl vplce+4(%rip), %r9d
|
||||
movl vplce+8(%rip), %r10d
|
||||
movl vplce+12(%rip), %r11d
|
||||
# selfmod loopit, vltepilog
|
||||
jmp loopit
|
||||
jmp loopit
|
||||
|
||||
.align 16
|
||||
.align 16
|
||||
loopit:
|
||||
movl %r8d, %edx
|
||||
shifter1: shrl $24, %edx
|
||||
step1: addl $0x88888888, %r8d
|
||||
step1: addl $0x44444444, %r8d
|
||||
movzbl (%rax,%rdx), %edx
|
||||
movl %r9d, %ebx
|
||||
movb (%r12,%rdx), %dl
|
||||
shifter2: shrl $24, %ebx
|
||||
step2: addl $0x88888888, %r9d
|
||||
source2: movzbl 0x88888888(%rax,%rbx), %ebx
|
||||
step2: addl $0x44444444, %r9d
|
||||
source2: movzbl 0x44444444(%rax,%rbx), %ebx
|
||||
movl %r10d, %ebp
|
||||
movb (%r13,%rbx), %bl
|
||||
shifter3: shr $24, %ebp
|
||||
step3: addl $0x88888888, %r10d
|
||||
source3: movzbl 0x88888888(%rax,%rbp), %ebp
|
||||
step3: addl $0x44444444, %r10d
|
||||
source3: movzbl 0x44444444(%rax,%rbp), %ebp
|
||||
movl %r11d, %esi
|
||||
movb (%r14,%rbp), %bpl
|
||||
shifter4: shr $24, %esi
|
||||
step4: add $0x88888888, %r11d
|
||||
source4: movzbl 0x88888888(%rax,%rsi), %esi
|
||||
step4: add $0x44444444, %r11d
|
||||
source4: movzbl 0x44444444(%rax,%rsi), %esi
|
||||
movb %dl, (%rdi,%rcx)
|
||||
movb %bl, 1(%rdi,%rcx)
|
||||
movb (%r15,%rsi), %sil
|
||||
movb %bpl, 2(%rdi,%rcx)
|
||||
movb %sil, 3(%rdi,%rcx)
|
||||
|
||||
vltpitch: addq $320, %rcx
|
||||
jl loopit
|
||||
vltpitch: addq $320, %rcx
|
||||
jl loopit
|
||||
|
||||
movl %r8d, vplce(%rip)
|
||||
movl %r9d, vplce+4(%rip)
|
||||
movl %r10d, vplce+8(%rip)
|
||||
movl %r11d, vplce+12(%rip)
|
||||
|
||||
movl %r11d, vplce+12(%rip)
|
||||
|
||||
vltepilog:
|
||||
addq $8, %rsp
|
||||
.cfi_adjust_cfa_offset -8
|
||||
|
|
@ -137,5 +137,5 @@ vltepilog:
|
|||
ret
|
||||
.cfi_endproc
|
||||
.align 16
|
||||
|
||||
|
||||
|
||||
|
|
|
|||
Loading…
Reference in a new issue