Add gcc target tests #704
Merged
Add gcc target tests #704
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
skmp
force-pushed
the
skmp/add-gcc-target-tests
branch
2 times, most recently
from
2ad061f to
d0aa661
Compare
|
@Sonicadvance1 can you investigate why it fails on the runners? I suspect missing 32-bit libs from the rootfs? |
skmp
force-pushed
the
skmp/add-gcc-target-tests
branch
from
d0aa661 to
abcd338
Compare
skmp
force-pushed
the
skmp/add-gcc-target-tests
branch
from
abcd338 to
29a8fd0
Compare
skmp
force-pushed
the
skmp/add-gcc-target-tests
branch
from
0fdb887 to
b453ede
Compare
phire
approved these changes
Jan 26, 2021
Sonicadvance1
added a commit
to Sonicadvance1/FEX
that referenced
this pull request
Jul 7, 2023
Only return the particular flags that are being requested in the moment since compacting them all when requested is fairly slow. x87 fcmov in particular was requesting all the flags when it only needs a couple. This reduces a `fcmovb` instruction count blowup from 103x to 48x. Still more room to go but this one stood out as being particularly bad. Old: ```asm 0x0000000265a002bc 10ffffe0 adr x0, #-0x4 (addr 0x265a002b8) 0x0000000265a002c0 f9005f80 str x0, [x28, FEX-Emu#184] 0x0000000265a002c4 d2800014 mov x20, #0x0 0x0000000265a002c8 d2800035 mov x21, #0x1 0x0000000265a002cc d2800056 mov x22, #0x2 0x0000000265a002d0 394b0397 ldrb w23, [x28, FEX-Emu#704] 0x0000000265a002d4 d3407ef7 ubfx x23, x23, #0, FEX-Emu#32 0x0000000265a002d8 aa1702d6 orr x22, x22, x23 0x0000000265a002dc 394b0b97 ldrb w23, [x28, FEX-Emu#706] 0x0000000265a002e0 d3407ef7 ubfx x23, x23, #0, FEX-Emu#32 0x0000000265a002e4 531e76f7 lsl w23, w23, FEX-Emu#2 0x0000000265a002e8 aa1702d6 orr x22, x22, x23 0x0000000265a002ec 394b1397 ldrb w23, [x28, FEX-Emu#708] 0x0000000265a002f0 d3407ef7 ubfx x23, x23, #0, FEX-Emu#32 0x0000000265a002f4 531c6ef7 lsl w23, w23, FEX-Emu#4 0x0000000265a002f8 aa1702d6 orr x22, x22, x23 0x0000000265a002fc 394b1b97 ldrb w23, [x28, FEX-Emu#710] 0x0000000265a00300 d3407ef7 ubfx x23, x23, #0, FEX-Emu#32 0x0000000265a00304 531a66f7 lsl w23, w23, FEX-Emu#6 0x0000000265a00308 aa1702d6 orr x22, x22, x23 0x0000000265a0030c 394b1f97 ldrb w23, [x28, FEX-Emu#711] 0x0000000265a00310 d3407ef7 ubfx x23, x23, #0, FEX-Emu#32 0x0000000265a00314 531962f7 lsl w23, w23, FEX-Emu#7 0x0000000265a00318 aa1702d6 orr x22, x22, x23 0x0000000265a0031c 394b2397 ldrb w23, [x28, FEX-Emu#712] 0x0000000265a00320 d3407ef7 ubfx x23, x23, #0, FEX-Emu#32 0x0000000265a00324 53185ef7 lsl w23, w23, FEX-Emu#8 0x0000000265a00328 aa1702d6 orr x22, x22, x23 0x0000000265a0032c 394b2797 ldrb w23, [x28, FEX-Emu#713] 0x0000000265a00330 d3407ef7 ubfx x23, x23, #0, FEX-Emu#32 0x0000000265a00334 53175af7 lsl w23, w23, FEX-Emu#9 0x0000000265a00338 aa1702d6 orr x22, x22, x23 0x0000000265a0033c 394b2b97 ldrb w23, [x28, FEX-Emu#714] 0x0000000265a00340 d3407ef7 ubfx x23, x23, #0, FEX-Emu#32 0x0000000265a00344 531656f7 lsl w23, w23, FEX-Emu#10 0x0000000265a00348 aa1702d6 orr x22, x22, x23 0x0000000265a0034c 394b2f97 ldrb w23, [x28, FEX-Emu#715] 0x0000000265a00350 d3407ef7 ubfx x23, x23, #0, FEX-Emu#32 0x0000000265a00354 531552f7 lsl w23, w23, FEX-Emu#11 0x0000000265a00358 aa1702d6 orr x22, x22, x23 0x0000000265a0035c 394b3397 ldrb w23, [x28, FEX-Emu#716] 0x0000000265a00360 d3407ef7 ubfx x23, x23, #0, FEX-Emu#32 0x0000000265a00364 53144ef7 lsl w23, w23, FEX-Emu#12 0x0000000265a00368 aa1702d6 orr x22, x22, x23 0x0000000265a0036c 394b3b97 ldrb w23, [x28, FEX-Emu#718] 0x0000000265a00370 d3407ef7 ubfx x23, x23, #0, FEX-Emu#32 0x0000000265a00374 531246f7 lsl w23, w23, FEX-Emu#14 0x0000000265a00378 aa1702d6 orr x22, x22, x23 0x0000000265a0037c 394b4397 ldrb w23, [x28, FEX-Emu#720] 0x0000000265a00380 d3407ef7 ubfx x23, x23, #0, FEX-Emu#32 0x0000000265a00384 53103ef7 lsl w23, w23, FEX-Emu#16 0x0000000265a00388 aa1702d6 orr x22, x22, x23 0x0000000265a0038c 394b4797 ldrb w23, [x28, FEX-Emu#721] 0x0000000265a00390 d3407ef7 ubfx x23, x23, #0, FEX-Emu#32 0x0000000265a00394 530f3af7 lsl w23, w23, FEX-Emu#17 0x0000000265a00398 aa1702d6 orr x22, x22, x23 0x0000000265a0039c 394b4b97 ldrb w23, [x28, FEX-Emu#722] 0x0000000265a003a0 d3407ef7 ubfx x23, x23, #0, FEX-Emu#32 0x0000000265a003a4 530e36f7 lsl w23, w23, FEX-Emu#18 0x0000000265a003a8 aa1702d6 orr x22, x22, x23 0x0000000265a003ac 394b4f97 ldrb w23, [x28, FEX-Emu#723] 0x0000000265a003b0 d3407ef7 ubfx x23, x23, #0, FEX-Emu#32 0x0000000265a003b4 530d32f7 lsl w23, w23, FEX-Emu#19 0x0000000265a003b8 aa1702d6 orr x22, x22, x23 0x0000000265a003bc 394b5397 ldrb w23, [x28, FEX-Emu#724] 0x0000000265a003c0 d3407ef7 ubfx x23, x23, #0, FEX-Emu#32 0x0000000265a003c4 530c2ef7 lsl w23, w23, FEX-Emu#20 0x0000000265a003c8 aa1702d6 orr x22, x22, x23 0x0000000265a003cc 394b5797 ldrb w23, [x28, FEX-Emu#725] 0x0000000265a003d0 d3407ef7 ubfx x23, x23, #0, FEX-Emu#32 0x0000000265a003d4 530b2af7 lsl w23, w23, FEX-Emu#21 0x0000000265a003d8 aa1702d6 orr x22, x22, x23 0x0000000265a003dc 924002d6 and x22, x22, #0x1 0x0000000265a003e0 93400294 sbfx x20, x20, #0, FEX-Emu#1 0x0000000265a003e4 934002b5 sbfx x21, x21, #0, FEX-Emu#1 0x0000000265a003e8 f10002df cmp x22, #0x0 (0) 0x0000000265a003ec 9a950294 csel x20, x20, x21, eq 0x0000000265a003f0 4e080e84 dup v4.2d, x20 0x0000000265a003f4 394baf94 ldrb w20, [x28, FEX-Emu#747] 0x0000000265a003f8 91000695 add x21, x20, #0x1 (1) 0x0000000265a003fc 92400ab5 and x21, x21, #0x7 0x0000000265a00400 d2800200 mov x0, #0x10 0x0000000265a00404 9b007e80 mul x0, x20, x0 0x0000000265a00408 8b000380 add x0, x28, x0 0x0000000265a0040c 3dc0bc05 ldr q5, [x0, FEX-Emu#752] 0x0000000265a00410 d2800200 mov x0, #0x10 0x0000000265a00414 9b007ea0 mul x0, x21, x0 0x0000000265a00418 8b000380 add x0, x28, x0 0x0000000265a0041c 3dc0bc06 ldr q6, [x0, FEX-Emu#752] 0x0000000265a00420 4ea41c80 mov v0.16b, v4.16b 0x0000000265a00424 6e651cc0 bsl v0.16b, v6.16b, v5.16b 0x0000000265a00428 4ea01c04 mov v4.16b, v0.16b 0x0000000265a0042c d2800200 mov x0, #0x10 0x0000000265a00430 9b007e80 mul x0, x20, x0 0x0000000265a00434 8b000380 add x0, x28, x0 0x0000000265a00438 3d80bc04 str q4, [x0, FEX-Emu#752] 0x0000000265a0043c 58000040 ldr x0, pc+8 (addr 0x265a00444) 0x0000000265a00440 d63f0000 blr x0 ``` New: ```asm 0x0000000265a002bc 10ffffe0 adr x0, #-0x4 (addr 0x265a002b8) 0x0000000265a002c0 f9005f80 str x0, [x28, FEX-Emu#184] 0x0000000265a002c4 d2800014 mov x20, #0x0 0x0000000265a002c8 d2800035 mov x21, #0x1 0x0000000265a002cc d2800056 mov x22, #0x2 0x0000000265a002d0 394b1f97 ldrb w23, [x28, FEX-Emu#711] 0x0000000265a002d4 331902f6 bfi w22, w23, FEX-Emu#7, FEX-Emu#1 0x0000000265a002d8 394b2797 ldrb w23, [x28, FEX-Emu#713] 0x0000000265a002dc 331702f6 bfi w22, w23, FEX-Emu#9, FEX-Emu#1 0x0000000265a002e0 394b2f97 ldrb w23, [x28, FEX-Emu#715] 0x0000000265a002e4 331502f6 bfi w22, w23, FEX-Emu#11, FEX-Emu#1 0x0000000265a002e8 394b4797 ldrb w23, [x28, FEX-Emu#721] 0x0000000265a002ec 330f02f6 bfi w22, w23, FEX-Emu#17, FEX-Emu#1 0x0000000265a002f0 394b4f97 ldrb w23, [x28, FEX-Emu#723] 0x0000000265a002f4 330d02f6 bfi w22, w23, FEX-Emu#19, FEX-Emu#1 0x0000000265a002f8 394b5797 ldrb w23, [x28, FEX-Emu#725] 0x0000000265a002fc 330b02f6 bfi w22, w23, FEX-Emu#21, FEX-Emu#1 0x0000000265a00300 924002d6 and x22, x22, #0x1 0x0000000265a00304 93400294 sbfx x20, x20, #0, FEX-Emu#1 0x0000000265a00308 934002b5 sbfx x21, x21, #0, FEX-Emu#1 0x0000000265a0030c f10002df cmp x22, #0x0 (0) 0x0000000265a00310 9a950294 csel x20, x20, x21, eq 0x0000000265a00314 4e080e84 dup v4.2d, x20 0x0000000265a00318 394baf94 ldrb w20, [x28, FEX-Emu#747] 0x0000000265a0031c 91000695 add x21, x20, #0x1 (1) 0x0000000265a00320 92400ab5 and x21, x21, #0x7 0x0000000265a00324 d2800200 mov x0, #0x10 0x0000000265a00328 9b007e80 mul x0, x20, x0 0x0000000265a0032c 8b000380 add x0, x28, x0 0x0000000265a00330 3dc0bc05 ldr q5, [x0, FEX-Emu#752] 0x0000000265a00334 d2800200 mov x0, #0x10 0x0000000265a00338 9b007ea0 mul x0, x21, x0 0x0000000265a0033c 8b000380 add x0, x28, x0 0x0000000265a00340 3dc0bc06 ldr q6, [x0, FEX-Emu#752] 0x0000000265a00344 4ea41c80 mov v0.16b, v4.16b 0x0000000265a00348 6e651cc0 bsl v0.16b, v6.16b, v5.16b 0x0000000265a0034c 4ea01c04 mov v4.16b, v0.16b 0x0000000265a00350 d2800200 mov x0, #0x10 0x0000000265a00354 9b007e80 mul x0, x20, x0 0x0000000265a00358 8b000380 add x0, x28, x0 0x0000000265a0035c 3d80bc04 str q4, [x0, FEX-Emu#752] 0x0000000265a00360 58000040 ldr x0, pc+8 (addr 0x265a00368) 0x0000000265a00364 d63f0000 blr x0 ```
Sonicadvance1
added a commit
to Sonicadvance1/FEX
that referenced
this pull request
Jul 7, 2023
Only return the particular flags that are being requested in the moment since compacting them all when requested is fairly slow. x87 fcmov in particular was requesting all the flags when it only needs a couple. This reduces a `fcmovb` instruction count blowup from 103x to 38x. Still more room to go but this one stood out as being particularly bad. Old: ```asm 0x0000000265a002bc 10ffffe0 adr x0, #-0x4 (addr 0x265a002b8) 0x0000000265a002c0 f9005f80 str x0, [x28, FEX-Emu#184] 0x0000000265a002c4 d2800014 mov x20, #0x0 0x0000000265a002c8 d2800035 mov x21, #0x1 0x0000000265a002cc d2800056 mov x22, #0x2 0x0000000265a002d0 394b0397 ldrb w23, [x28, FEX-Emu#704] 0x0000000265a002d4 d3407ef7 ubfx x23, x23, #0, FEX-Emu#32 0x0000000265a002d8 aa1702d6 orr x22, x22, x23 0x0000000265a002dc 394b0b97 ldrb w23, [x28, FEX-Emu#706] 0x0000000265a002e0 d3407ef7 ubfx x23, x23, #0, FEX-Emu#32 0x0000000265a002e4 531e76f7 lsl w23, w23, FEX-Emu#2 0x0000000265a002e8 aa1702d6 orr x22, x22, x23 0x0000000265a002ec 394b1397 ldrb w23, [x28, FEX-Emu#708] 0x0000000265a002f0 d3407ef7 ubfx x23, x23, #0, FEX-Emu#32 0x0000000265a002f4 531c6ef7 lsl w23, w23, FEX-Emu#4 0x0000000265a002f8 aa1702d6 orr x22, x22, x23 0x0000000265a002fc 394b1b97 ldrb w23, [x28, FEX-Emu#710] 0x0000000265a00300 d3407ef7 ubfx x23, x23, #0, FEX-Emu#32 0x0000000265a00304 531a66f7 lsl w23, w23, FEX-Emu#6 0x0000000265a00308 aa1702d6 orr x22, x22, x23 0x0000000265a0030c 394b1f97 ldrb w23, [x28, FEX-Emu#711] 0x0000000265a00310 d3407ef7 ubfx x23, x23, #0, FEX-Emu#32 0x0000000265a00314 531962f7 lsl w23, w23, FEX-Emu#7 0x0000000265a00318 aa1702d6 orr x22, x22, x23 0x0000000265a0031c 394b2397 ldrb w23, [x28, FEX-Emu#712] 0x0000000265a00320 d3407ef7 ubfx x23, x23, #0, FEX-Emu#32 0x0000000265a00324 53185ef7 lsl w23, w23, FEX-Emu#8 0x0000000265a00328 aa1702d6 orr x22, x22, x23 0x0000000265a0032c 394b2797 ldrb w23, [x28, FEX-Emu#713] 0x0000000265a00330 d3407ef7 ubfx x23, x23, #0, FEX-Emu#32 0x0000000265a00334 53175af7 lsl w23, w23, FEX-Emu#9 0x0000000265a00338 aa1702d6 orr x22, x22, x23 0x0000000265a0033c 394b2b97 ldrb w23, [x28, FEX-Emu#714] 0x0000000265a00340 d3407ef7 ubfx x23, x23, #0, FEX-Emu#32 0x0000000265a00344 531656f7 lsl w23, w23, FEX-Emu#10 0x0000000265a00348 aa1702d6 orr x22, x22, x23 0x0000000265a0034c 394b2f97 ldrb w23, [x28, FEX-Emu#715] 0x0000000265a00350 d3407ef7 ubfx x23, x23, #0, FEX-Emu#32 0x0000000265a00354 531552f7 lsl w23, w23, FEX-Emu#11 0x0000000265a00358 aa1702d6 orr x22, x22, x23 0x0000000265a0035c 394b3397 ldrb w23, [x28, FEX-Emu#716] 0x0000000265a00360 d3407ef7 ubfx x23, x23, #0, FEX-Emu#32 0x0000000265a00364 53144ef7 lsl w23, w23, FEX-Emu#12 0x0000000265a00368 aa1702d6 orr x22, x22, x23 0x0000000265a0036c 394b3b97 ldrb w23, [x28, FEX-Emu#718] 0x0000000265a00370 d3407ef7 ubfx x23, x23, #0, FEX-Emu#32 0x0000000265a00374 531246f7 lsl w23, w23, FEX-Emu#14 0x0000000265a00378 aa1702d6 orr x22, x22, x23 0x0000000265a0037c 394b4397 ldrb w23, [x28, FEX-Emu#720] 0x0000000265a00380 d3407ef7 ubfx x23, x23, #0, FEX-Emu#32 0x0000000265a00384 53103ef7 lsl w23, w23, FEX-Emu#16 0x0000000265a00388 aa1702d6 orr x22, x22, x23 0x0000000265a0038c 394b4797 ldrb w23, [x28, FEX-Emu#721] 0x0000000265a00390 d3407ef7 ubfx x23, x23, #0, FEX-Emu#32 0x0000000265a00394 530f3af7 lsl w23, w23, FEX-Emu#17 0x0000000265a00398 aa1702d6 orr x22, x22, x23 0x0000000265a0039c 394b4b97 ldrb w23, [x28, FEX-Emu#722] 0x0000000265a003a0 d3407ef7 ubfx x23, x23, #0, FEX-Emu#32 0x0000000265a003a4 530e36f7 lsl w23, w23, FEX-Emu#18 0x0000000265a003a8 aa1702d6 orr x22, x22, x23 0x0000000265a003ac 394b4f97 ldrb w23, [x28, FEX-Emu#723] 0x0000000265a003b0 d3407ef7 ubfx x23, x23, #0, FEX-Emu#32 0x0000000265a003b4 530d32f7 lsl w23, w23, FEX-Emu#19 0x0000000265a003b8 aa1702d6 orr x22, x22, x23 0x0000000265a003bc 394b5397 ldrb w23, [x28, FEX-Emu#724] 0x0000000265a003c0 d3407ef7 ubfx x23, x23, #0, FEX-Emu#32 0x0000000265a003c4 530c2ef7 lsl w23, w23, FEX-Emu#20 0x0000000265a003c8 aa1702d6 orr x22, x22, x23 0x0000000265a003cc 394b5797 ldrb w23, [x28, FEX-Emu#725] 0x0000000265a003d0 d3407ef7 ubfx x23, x23, #0, FEX-Emu#32 0x0000000265a003d4 530b2af7 lsl w23, w23, FEX-Emu#21 0x0000000265a003d8 aa1702d6 orr x22, x22, x23 0x0000000265a003dc 924002d6 and x22, x22, #0x1 0x0000000265a003e0 93400294 sbfx x20, x20, #0, FEX-Emu#1 0x0000000265a003e4 934002b5 sbfx x21, x21, #0, FEX-Emu#1 0x0000000265a003e8 f10002df cmp x22, #0x0 (0) 0x0000000265a003ec 9a950294 csel x20, x20, x21, eq 0x0000000265a003f0 4e080e84 dup v4.2d, x20 0x0000000265a003f4 394baf94 ldrb w20, [x28, FEX-Emu#747] 0x0000000265a003f8 91000695 add x21, x20, #0x1 (1) 0x0000000265a003fc 92400ab5 and x21, x21, #0x7 0x0000000265a00400 d2800200 mov x0, #0x10 0x0000000265a00404 9b007e80 mul x0, x20, x0 0x0000000265a00408 8b000380 add x0, x28, x0 0x0000000265a0040c 3dc0bc05 ldr q5, [x0, FEX-Emu#752] 0x0000000265a00410 d2800200 mov x0, #0x10 0x0000000265a00414 9b007ea0 mul x0, x21, x0 0x0000000265a00418 8b000380 add x0, x28, x0 0x0000000265a0041c 3dc0bc06 ldr q6, [x0, FEX-Emu#752] 0x0000000265a00420 4ea41c80 mov v0.16b, v4.16b 0x0000000265a00424 6e651cc0 bsl v0.16b, v6.16b, v5.16b 0x0000000265a00428 4ea01c04 mov v4.16b, v0.16b 0x0000000265a0042c d2800200 mov x0, #0x10 0x0000000265a00430 9b007e80 mul x0, x20, x0 0x0000000265a00434 8b000380 add x0, x28, x0 0x0000000265a00438 3d80bc04 str q4, [x0, FEX-Emu#752] 0x0000000265a0043c 58000040 ldr x0, pc+8 (addr 0x265a00444) 0x0000000265a00440 d63f0000 blr x0 ``` New: ```asm 0x0000000265a002bc 10ffffe0 adr x0, #-0x4 (addr 0x265a002b8) 0x0000000265a002c0 f9005f80 str x0, [x28, FEX-Emu#184] 0x0000000265a002c4 d2800014 mov x20, #0x0 0x0000000265a002c8 d2800035 mov x21, #0x1 0x0000000265a002cc d2800056 mov x22, #0x2 0x0000000265a002d0 394b0397 ldrb w23, [x28, FEX-Emu#704] 0x0000000265a002d4 330002f6 bfxil w22, w23, #0, FEX-Emu#1 0x0000000265a002d8 924002d6 and x22, x22, #0x1 0x0000000265a002dc 93400294 sbfx x20, x20, #0, FEX-Emu#1 0x0000000265a002e0 934002b5 sbfx x21, x21, #0, FEX-Emu#1 0x0000000265a002e4 f10002df cmp x22, #0x0 (0) 0x0000000265a002e8 9a950294 csel x20, x20, x21, eq 0x0000000265a002ec 4e080e84 dup v4.2d, x20 0x0000000265a002f0 394baf94 ldrb w20, [x28, FEX-Emu#747] 0x0000000265a002f4 91000695 add x21, x20, #0x1 (1) 0x0000000265a002f8 92400ab5 and x21, x21, #0x7 0x0000000265a002fc d2800200 mov x0, #0x10 0x0000000265a00300 9b007e80 mul x0, x20, x0 0x0000000265a00304 8b000380 add x0, x28, x0 0x0000000265a00308 3dc0bc05 ldr q5, [x0, FEX-Emu#752] 0x0000000265a0030c d2800200 mov x0, #0x10 0x0000000265a00310 9b007ea0 mul x0, x21, x0 0x0000000265a00314 8b000380 add x0, x28, x0 0x0000000265a00318 3dc0bc06 ldr q6, [x0, FEX-Emu#752] 0x0000000265a0031c 4ea41c80 mov v0.16b, v4.16b 0x0000000265a00320 6e651cc0 bsl v0.16b, v6.16b, v5.16b 0x0000000265a00324 4ea01c04 mov v4.16b, v0.16b 0x0000000265a00328 d2800200 mov x0, #0x10 0x0000000265a0032c 9b007e80 mul x0, x20, x0 0x0000000265a00330 8b000380 add x0, x28, x0 0x0000000265a00334 3d80bc04 str q4, [x0, FEX-Emu#752] 0x0000000265a00338 58000040 ldr x0, pc+8 (addr 0x265a00340) 0x0000000265a0033c d63f0000 blr x0 ```
Sonicadvance1
added a commit
to Sonicadvance1/FEX
that referenced
this pull request
Jul 8, 2023
Only return the particular flags that are being requested in the moment since compacting them all when requested is fairly slow. x87 fcmov in particular was requesting all the flags when it only needs a couple. This reduces a `fcmovb` instruction count blowup from 103x to 38x. Still more room to go but this one stood out as being particularly bad. Old: ```asm 0x0000000265a002bc 10ffffe0 adr x0, #-0x4 (addr 0x265a002b8) 0x0000000265a002c0 f9005f80 str x0, [x28, FEX-Emu#184] 0x0000000265a002c4 d2800014 mov x20, #0x0 0x0000000265a002c8 d2800035 mov x21, #0x1 0x0000000265a002cc d2800056 mov x22, #0x2 0x0000000265a002d0 394b0397 ldrb w23, [x28, FEX-Emu#704] 0x0000000265a002d4 d3407ef7 ubfx x23, x23, #0, FEX-Emu#32 0x0000000265a002d8 aa1702d6 orr x22, x22, x23 0x0000000265a002dc 394b0b97 ldrb w23, [x28, FEX-Emu#706] 0x0000000265a002e0 d3407ef7 ubfx x23, x23, #0, FEX-Emu#32 0x0000000265a002e4 531e76f7 lsl w23, w23, FEX-Emu#2 0x0000000265a002e8 aa1702d6 orr x22, x22, x23 0x0000000265a002ec 394b1397 ldrb w23, [x28, FEX-Emu#708] 0x0000000265a002f0 d3407ef7 ubfx x23, x23, #0, FEX-Emu#32 0x0000000265a002f4 531c6ef7 lsl w23, w23, FEX-Emu#4 0x0000000265a002f8 aa1702d6 orr x22, x22, x23 0x0000000265a002fc 394b1b97 ldrb w23, [x28, FEX-Emu#710] 0x0000000265a00300 d3407ef7 ubfx x23, x23, #0, FEX-Emu#32 0x0000000265a00304 531a66f7 lsl w23, w23, FEX-Emu#6 0x0000000265a00308 aa1702d6 orr x22, x22, x23 0x0000000265a0030c 394b1f97 ldrb w23, [x28, FEX-Emu#711] 0x0000000265a00310 d3407ef7 ubfx x23, x23, #0, FEX-Emu#32 0x0000000265a00314 531962f7 lsl w23, w23, FEX-Emu#7 0x0000000265a00318 aa1702d6 orr x22, x22, x23 0x0000000265a0031c 394b2397 ldrb w23, [x28, FEX-Emu#712] 0x0000000265a00320 d3407ef7 ubfx x23, x23, #0, FEX-Emu#32 0x0000000265a00324 53185ef7 lsl w23, w23, FEX-Emu#8 0x0000000265a00328 aa1702d6 orr x22, x22, x23 0x0000000265a0032c 394b2797 ldrb w23, [x28, FEX-Emu#713] 0x0000000265a00330 d3407ef7 ubfx x23, x23, #0, FEX-Emu#32 0x0000000265a00334 53175af7 lsl w23, w23, FEX-Emu#9 0x0000000265a00338 aa1702d6 orr x22, x22, x23 0x0000000265a0033c 394b2b97 ldrb w23, [x28, FEX-Emu#714] 0x0000000265a00340 d3407ef7 ubfx x23, x23, #0, FEX-Emu#32 0x0000000265a00344 531656f7 lsl w23, w23, FEX-Emu#10 0x0000000265a00348 aa1702d6 orr x22, x22, x23 0x0000000265a0034c 394b2f97 ldrb w23, [x28, FEX-Emu#715] 0x0000000265a00350 d3407ef7 ubfx x23, x23, #0, FEX-Emu#32 0x0000000265a00354 531552f7 lsl w23, w23, FEX-Emu#11 0x0000000265a00358 aa1702d6 orr x22, x22, x23 0x0000000265a0035c 394b3397 ldrb w23, [x28, FEX-Emu#716] 0x0000000265a00360 d3407ef7 ubfx x23, x23, #0, FEX-Emu#32 0x0000000265a00364 53144ef7 lsl w23, w23, FEX-Emu#12 0x0000000265a00368 aa1702d6 orr x22, x22, x23 0x0000000265a0036c 394b3b97 ldrb w23, [x28, FEX-Emu#718] 0x0000000265a00370 d3407ef7 ubfx x23, x23, #0, FEX-Emu#32 0x0000000265a00374 531246f7 lsl w23, w23, FEX-Emu#14 0x0000000265a00378 aa1702d6 orr x22, x22, x23 0x0000000265a0037c 394b4397 ldrb w23, [x28, FEX-Emu#720] 0x0000000265a00380 d3407ef7 ubfx x23, x23, #0, FEX-Emu#32 0x0000000265a00384 53103ef7 lsl w23, w23, FEX-Emu#16 0x0000000265a00388 aa1702d6 orr x22, x22, x23 0x0000000265a0038c 394b4797 ldrb w23, [x28, FEX-Emu#721] 0x0000000265a00390 d3407ef7 ubfx x23, x23, #0, FEX-Emu#32 0x0000000265a00394 530f3af7 lsl w23, w23, FEX-Emu#17 0x0000000265a00398 aa1702d6 orr x22, x22, x23 0x0000000265a0039c 394b4b97 ldrb w23, [x28, FEX-Emu#722] 0x0000000265a003a0 d3407ef7 ubfx x23, x23, #0, FEX-Emu#32 0x0000000265a003a4 530e36f7 lsl w23, w23, FEX-Emu#18 0x0000000265a003a8 aa1702d6 orr x22, x22, x23 0x0000000265a003ac 394b4f97 ldrb w23, [x28, FEX-Emu#723] 0x0000000265a003b0 d3407ef7 ubfx x23, x23, #0, FEX-Emu#32 0x0000000265a003b4 530d32f7 lsl w23, w23, FEX-Emu#19 0x0000000265a003b8 aa1702d6 orr x22, x22, x23 0x0000000265a003bc 394b5397 ldrb w23, [x28, FEX-Emu#724] 0x0000000265a003c0 d3407ef7 ubfx x23, x23, #0, FEX-Emu#32 0x0000000265a003c4 530c2ef7 lsl w23, w23, FEX-Emu#20 0x0000000265a003c8 aa1702d6 orr x22, x22, x23 0x0000000265a003cc 394b5797 ldrb w23, [x28, FEX-Emu#725] 0x0000000265a003d0 d3407ef7 ubfx x23, x23, #0, FEX-Emu#32 0x0000000265a003d4 530b2af7 lsl w23, w23, FEX-Emu#21 0x0000000265a003d8 aa1702d6 orr x22, x22, x23 0x0000000265a003dc 924002d6 and x22, x22, #0x1 0x0000000265a003e0 93400294 sbfx x20, x20, #0, FEX-Emu#1 0x0000000265a003e4 934002b5 sbfx x21, x21, #0, FEX-Emu#1 0x0000000265a003e8 f10002df cmp x22, #0x0 (0) 0x0000000265a003ec 9a950294 csel x20, x20, x21, eq 0x0000000265a003f0 4e080e84 dup v4.2d, x20 0x0000000265a003f4 394baf94 ldrb w20, [x28, FEX-Emu#747] 0x0000000265a003f8 91000695 add x21, x20, #0x1 (1) 0x0000000265a003fc 92400ab5 and x21, x21, #0x7 0x0000000265a00400 d2800200 mov x0, #0x10 0x0000000265a00404 9b007e80 mul x0, x20, x0 0x0000000265a00408 8b000380 add x0, x28, x0 0x0000000265a0040c 3dc0bc05 ldr q5, [x0, FEX-Emu#752] 0x0000000265a00410 d2800200 mov x0, #0x10 0x0000000265a00414 9b007ea0 mul x0, x21, x0 0x0000000265a00418 8b000380 add x0, x28, x0 0x0000000265a0041c 3dc0bc06 ldr q6, [x0, FEX-Emu#752] 0x0000000265a00420 4ea41c80 mov v0.16b, v4.16b 0x0000000265a00424 6e651cc0 bsl v0.16b, v6.16b, v5.16b 0x0000000265a00428 4ea01c04 mov v4.16b, v0.16b 0x0000000265a0042c d2800200 mov x0, #0x10 0x0000000265a00430 9b007e80 mul x0, x20, x0 0x0000000265a00434 8b000380 add x0, x28, x0 0x0000000265a00438 3d80bc04 str q4, [x0, FEX-Emu#752] 0x0000000265a0043c 58000040 ldr x0, pc+8 (addr 0x265a00444) 0x0000000265a00440 d63f0000 blr x0 ``` New: ```asm 0x0000000265a002bc 10ffffe0 adr x0, #-0x4 (addr 0x265a002b8) 0x0000000265a002c0 f9005f80 str x0, [x28, FEX-Emu#184] 0x0000000265a002c4 d2800014 mov x20, #0x0 0x0000000265a002c8 d2800035 mov x21, #0x1 0x0000000265a002cc d2800056 mov x22, #0x2 0x0000000265a002d0 394b0397 ldrb w23, [x28, FEX-Emu#704] 0x0000000265a002d4 330002f6 bfxil w22, w23, #0, FEX-Emu#1 0x0000000265a002d8 924002d6 and x22, x22, #0x1 0x0000000265a002dc 93400294 sbfx x20, x20, #0, FEX-Emu#1 0x0000000265a002e0 934002b5 sbfx x21, x21, #0, FEX-Emu#1 0x0000000265a002e4 f10002df cmp x22, #0x0 (0) 0x0000000265a002e8 9a950294 csel x20, x20, x21, eq 0x0000000265a002ec 4e080e84 dup v4.2d, x20 0x0000000265a002f0 394baf94 ldrb w20, [x28, FEX-Emu#747] 0x0000000265a002f4 91000695 add x21, x20, #0x1 (1) 0x0000000265a002f8 92400ab5 and x21, x21, #0x7 0x0000000265a002fc d2800200 mov x0, #0x10 0x0000000265a00300 9b007e80 mul x0, x20, x0 0x0000000265a00304 8b000380 add x0, x28, x0 0x0000000265a00308 3dc0bc05 ldr q5, [x0, FEX-Emu#752] 0x0000000265a0030c d2800200 mov x0, #0x10 0x0000000265a00310 9b007ea0 mul x0, x21, x0 0x0000000265a00314 8b000380 add x0, x28, x0 0x0000000265a00318 3dc0bc06 ldr q6, [x0, FEX-Emu#752] 0x0000000265a0031c 4ea41c80 mov v0.16b, v4.16b 0x0000000265a00320 6e651cc0 bsl v0.16b, v6.16b, v5.16b 0x0000000265a00324 4ea01c04 mov v4.16b, v0.16b 0x0000000265a00328 d2800200 mov x0, #0x10 0x0000000265a0032c 9b007e80 mul x0, x20, x0 0x0000000265a00330 8b000380 add x0, x28, x0 0x0000000265a00334 3d80bc04 str q4, [x0, FEX-Emu#752] 0x0000000265a00338 58000040 ldr x0, pc+8 (addr 0x265a00340) 0x0000000265a0033c d63f0000 blr x0 ```
Sonicadvance1
added a commit
to Sonicadvance1/FEX
that referenced
this pull request
Jul 16, 2023
The BFI cascades in this particular instruction weren't optimal. Biggest improvement is the 8-bit version, while the 16-bit version gets a minor improvement. 8-bit instruction count reduced from 38 to 29. 16-bit instruction count reduced from 34 to 28. RCL can have a similar optimization done to it. ```asm Before 16-bit: 0x0000ffff80a801e0 10ffffe0 adr x0, #-0x4 (addr 0xffff80a801dc) 0x0000ffff80a801e4 f9005f80 str x0, [x28, FEX-Emu#184] 0x0000ffff80a801e8 d3403cb4 uxth x20, w5 0x0000ffff80a801ec d3403cf5 uxth x21, w7 0x0000ffff80a801f0 394b0396 ldrb w22, [x28, FEX-Emu#704] 0x0000ffff80a801f4 12001294 and w20, w20, #0x1f 0x0000ffff80a801f8 d2800017 mov x23, #0x0 0x0000ffff80a801fc b3403eb7 bfxil x23, x21, #0, FEX-Emu#16 0x0000ffff80a80200 b37002d7 bfi x23, x22, FEX-Emu#16, FEX-Emu#1 0x0000ffff80a80204 b36f3eb7 bfi x23, x21, FEX-Emu#17, FEX-Emu#16 0x0000ffff80a80208 b35f02d7 bfi x23, x22, FEX-Emu#33, FEX-Emu#1 0x0000ffff80a8020c aa1703e0 mov x0, x23 0x0000ffff80a80210 b35e3ea0 bfi x0, x21, FEX-Emu#34, FEX-Emu#16 0x0000ffff80a80214 aa0003f5 mov x21, x0 0x0000ffff80a80218 b34e02d5 bfi x21, x22, FEX-Emu#50, FEX-Emu#1 0x0000ffff80a8021c 9ad426b7 lsr x23, x21, x20 0x0000ffff80a80220 b3403ee7 bfxil x7, x23, #0, FEX-Emu#16 0x0000ffff80a80224 51000698 sub w24, w20, #0x1 (1) 0x0000ffff80a80228 9ad826b5 lsr x21, x21, x24 0x0000ffff80a8022c d34002b5 ubfx x21, x21, #0, FEX-Emu#1 0x0000ffff80a80230 7100069f cmp w20, #0x1 (1) 0x0000ffff80a80234 9a9622b4 csel x20, x21, x22, hs 0x0000ffff80a80238 390b0394 strb w20, [x28, FEX-Emu#704] 0x0000ffff80a8023c d34f3ef4 ubfx x20, x23, FEX-Emu#15, FEX-Emu#1 0x0000ffff80a80240 d34e3af5 ubfx x21, x23, FEX-Emu#14, FEX-Emu#1 0x0000ffff80a80244 ca150294 eor x20, x20, x21 0x0000ffff80a80248 390b2f94 strb w20, [x28, FEX-Emu#715] 0x0000ffff80a8024c 58000040 ldr x0, pc+8 (addr 0xffff80a80254) 0x0000ffff80a80250 d63f0000 blr x0 0x0000ffff80a80254 967da128 bl #-0x6097b60 (addr 0xffff7a9e86f4) 0x0000ffff80a80258 0000ffff udf #0xffff 0x0000ffff80a8025c 00010023 unallocated (Unallocated) 0x0000ffff80a80260 00000000 udf #0x0 [DEBUG] RIP: 0x10020 [DEBUG] Guest Code instructions: 1 [DEBUG] Host Code instructions: 34 [DEBUG] Blow-up Amt: 34x After 16-bit: 0x0000ffffa7c801e0 10ffffe0 adr x0, #-0x4 (addr 0xffffa7c801dc) 0x0000ffffa7c801e4 f9005f80 str x0, [x28, FEX-Emu#184] 0x0000ffffa7c801e8 d3403cb4 uxth x20, w5 0x0000ffffa7c801ec d3403cf5 uxth x21, w7 0x0000ffffa7c801f0 394b0396 ldrb w22, [x28, FEX-Emu#704] 0x0000ffffa7c801f4 12001294 and w20, w20, #0x1f 0x0000ffffa7c801f8 b37002d5 bfi x21, x22, FEX-Emu#16, FEX-Emu#1 0x0000ffffa7c801fc b36f42b5 bfi x21, x21, FEX-Emu#17, FEX-Emu#17 0x0000ffffa7c80200 b35e42b5 bfi x21, x21, FEX-Emu#34, FEX-Emu#17 0x0000ffffa7c80204 9ad426b7 lsr x23, x21, x20 0x0000ffffa7c80208 b3403ee7 bfxil x7, x23, #0, FEX-Emu#16 0x0000ffffa7c8020c 51000698 sub w24, w20, #0x1 (1) 0x0000ffffa7c80210 9ad826b5 lsr x21, x21, x24 0x0000ffffa7c80214 d34002b5 ubfx x21, x21, #0, FEX-Emu#1 0x0000ffffa7c80218 7100069f cmp w20, #0x1 (1) 0x0000ffffa7c8021c 9a9622b4 csel x20, x21, x22, hs 0x0000ffffa7c80220 390b0394 strb w20, [x28, FEX-Emu#704] 0x0000ffffa7c80224 d34f3ef4 ubfx x20, x23, FEX-Emu#15, FEX-Emu#1 0x0000ffffa7c80228 d34e3af5 ubfx x21, x23, FEX-Emu#14, FEX-Emu#1 0x0000ffffa7c8022c ca150294 eor x20, x20, x21 0x0000ffffa7c80230 390b2f94 strb w20, [x28, FEX-Emu#715] 0x0000ffffa7c80234 58000040 ldr x0, pc+8 (addr 0xffffa7c8023c) 0x0000ffffa7c80238 d63f0000 blr x0 0x0000ffffa7c8023c bd9cc128 unallocated (Unallocated) 0x0000ffffa7c80240 0000ffff udf #0xffff 0x0000ffffa7c80244 00010023 unallocated (Unallocated) 0x0000ffffa7c80248 00000000 udf #0x0 [DEBUG] RIP: 0x10020 [DEBUG] Guest Code instructions: 1 [DEBUG] Host Code instructions: 28 [DEBUG] Blow-up Amt: 28x Before 8-bit: 0x0000ffffa92801e0 10ffffe0 adr x0, #-0x4 (addr 0xffffa92801dc) 0x0000ffffa92801e4 f9005f80 str x0, [x28, FEX-Emu#184] 0x0000ffffa92801e8 d3401cb4 uxtb x20, w5 0x0000ffffa92801ec d3401cf5 uxtb x21, w7 0x0000ffffa92801f0 394b0396 ldrb w22, [x28, FEX-Emu#704] 0x0000ffffa92801f4 12001294 and w20, w20, #0x1f 0x0000ffffa92801f8 d2800017 mov x23, #0x0 0x0000ffffa92801fc b3401eb7 bfxil x23, x21, #0, FEX-Emu#8 0x0000ffffa9280200 b37802d7 bfi x23, x22, FEX-Emu#8, FEX-Emu#1 0x0000ffffa9280204 b3771eb7 bfi x23, x21, FEX-Emu#9, FEX-Emu#8 0x0000ffffa9280208 b36f02d7 bfi x23, x22, FEX-Emu#17, FEX-Emu#1 0x0000ffffa928020c b36e1eb7 bfi x23, x21, FEX-Emu#18, FEX-Emu#8 0x0000ffffa9280210 b36602d7 bfi x23, x22, FEX-Emu#26, FEX-Emu#1 0x0000ffffa9280214 b3651eb7 bfi x23, x21, FEX-Emu#27, FEX-Emu#8 0x0000ffffa9280218 b35d02d7 bfi x23, x22, FEX-Emu#35, FEX-Emu#1 0x0000ffffa928021c aa1703e0 mov x0, x23 0x0000ffffa9280220 b35c1ea0 bfi x0, x21, FEX-Emu#36, FEX-Emu#8 0x0000ffffa9280224 aa0003f5 mov x21, x0 0x0000ffffa9280228 b35402d5 bfi x21, x22, FEX-Emu#44, FEX-Emu#1 0x0000ffffa928022c 9ad426b7 lsr x23, x21, x20 0x0000ffffa9280230 b3401ee7 bfxil x7, x23, #0, FEX-Emu#8 0x0000ffffa9280234 51000698 sub w24, w20, #0x1 (1) 0x0000ffffa9280238 9ad826b5 lsr x21, x21, x24 0x0000ffffa928023c d34002b5 ubfx x21, x21, #0, FEX-Emu#1 0x0000ffffa9280240 7100069f cmp w20, #0x1 (1) 0x0000ffffa9280244 9a9622b4 csel x20, x21, x22, hs 0x0000ffffa9280248 390b0394 strb w20, [x28, FEX-Emu#704] 0x0000ffffa928024c d3471ef4 ubfx x20, x23, FEX-Emu#7, FEX-Emu#1 0x0000ffffa9280250 d3461af5 ubfx x21, x23, FEX-Emu#6, FEX-Emu#1 0x0000ffffa9280254 ca150294 eor x20, x20, x21 0x0000ffffa9280258 390b2f94 strb w20, [x28, FEX-Emu#715] 0x0000ffffa928025c 58000040 ldr x0, pc+8 (addr 0xffffa9280264) 0x0000ffffa9280260 d63f0000 blr x0 0x0000ffffa9280264 bf062128 unallocated (Unallocated) 0x0000ffffa9280268 0000ffff udf #0xffff 0x0000ffffa928026c 00010022 unallocated (Unallocated) 0x0000ffffa9280270 00000000 udf #0x0 [DEBUG] RIP: 0x10020 [DEBUG] Guest Code instructions: 1 [DEBUG] Host Code instructions: 38 [DEBUG] Blow-up Amt: 38x After 8-bit: 0x0000ffff9cc801e0 10ffffe0 adr x0, #-0x4 (addr 0xffff9cc801dc) 0x0000ffff9cc801e4 f9005f80 str x0, [x28, FEX-Emu#184] 0x0000ffff9cc801e8 d3401cb4 uxtb x20, w5 0x0000ffff9cc801ec d3401cf5 uxtb x21, w7 0x0000ffff9cc801f0 394b0396 ldrb w22, [x28, FEX-Emu#704] 0x0000ffff9cc801f4 12001294 and w20, w20, #0x1f 0x0000ffff9cc801f8 b37802d5 bfi x21, x22, FEX-Emu#8, FEX-Emu#1 0x0000ffff9cc801fc b37722b5 bfi x21, x21, FEX-Emu#9, FEX-Emu#9 0x0000ffff9cc80200 b36e46b5 bfi x21, x21, FEX-Emu#18, FEX-Emu#18 0x0000ffff9cc80204 b3778eb5 bfi x21, x21, FEX-Emu#9, FEX-Emu#36 0x0000ffff9cc80208 9ad426b7 lsr x23, x21, x20 0x0000ffff9cc8020c b3401ee7 bfxil x7, x23, #0, FEX-Emu#8 0x0000ffff9cc80210 51000698 sub w24, w20, #0x1 (1) 0x0000ffff9cc80214 9ad826b5 lsr x21, x21, x24 0x0000ffff9cc80218 d34002b5 ubfx x21, x21, #0, FEX-Emu#1 0x0000ffff9cc8021c 7100069f cmp w20, #0x1 (1) 0x0000ffff9cc80220 9a9622b4 csel x20, x21, x22, hs 0x0000ffff9cc80224 390b0394 strb w20, [x28, FEX-Emu#704] 0x0000ffff9cc80228 d3471ef4 ubfx x20, x23, FEX-Emu#7, FEX-Emu#1 0x0000ffff9cc8022c d3461af5 ubfx x21, x23, FEX-Emu#6, FEX-Emu#1 0x0000ffff9cc80230 ca150294 eor x20, x20, x21 0x0000ffff9cc80234 390b2f94 strb w20, [x28, FEX-Emu#715] 0x0000ffff9cc80238 58000040 ldr x0, pc+8 (addr 0xffff9cc80240) 0x0000ffff9cc8023c d63f0000 blr x0 0x0000ffff9cc80240 b2a75128 unallocated (Unallocated) 0x0000ffff9cc80244 0000ffff udf #0xffff 0x0000ffff9cc80248 00010022 unallocated (Unallocated) 0x0000ffff9cc8024c 00000000 udf #0x0 [DEBUG] RIP: 0x10020 [DEBUG] Guest Code instructions: 1 [DEBUG] Host Code instructions: 29 [DEBUG] Blow-up Amt: 29x ```
Sonicadvance1
added a commit
to Sonicadvance1/FEX
that referenced
this pull request
Jul 18, 2023
The BFI cascades in this particular instruction weren't optimal. Biggest improvement is the 8-bit version, while the 16-bit version gets a minor improvement. 8-bit instruction count reduced from 38 to 29. 16-bit instruction count reduced from 34 to 28. RCL can have a similar optimization done to it. ```asm Before 16-bit: 0x0000ffff80a801e0 10ffffe0 adr x0, #-0x4 (addr 0xffff80a801dc) 0x0000ffff80a801e4 f9005f80 str x0, [x28, FEX-Emu#184] 0x0000ffff80a801e8 d3403cb4 uxth x20, w5 0x0000ffff80a801ec d3403cf5 uxth x21, w7 0x0000ffff80a801f0 394b0396 ldrb w22, [x28, FEX-Emu#704] 0x0000ffff80a801f4 12001294 and w20, w20, #0x1f 0x0000ffff80a801f8 d2800017 mov x23, #0x0 0x0000ffff80a801fc b3403eb7 bfxil x23, x21, #0, FEX-Emu#16 0x0000ffff80a80200 b37002d7 bfi x23, x22, FEX-Emu#16, FEX-Emu#1 0x0000ffff80a80204 b36f3eb7 bfi x23, x21, FEX-Emu#17, FEX-Emu#16 0x0000ffff80a80208 b35f02d7 bfi x23, x22, FEX-Emu#33, FEX-Emu#1 0x0000ffff80a8020c aa1703e0 mov x0, x23 0x0000ffff80a80210 b35e3ea0 bfi x0, x21, FEX-Emu#34, FEX-Emu#16 0x0000ffff80a80214 aa0003f5 mov x21, x0 0x0000ffff80a80218 b34e02d5 bfi x21, x22, FEX-Emu#50, FEX-Emu#1 0x0000ffff80a8021c 9ad426b7 lsr x23, x21, x20 0x0000ffff80a80220 b3403ee7 bfxil x7, x23, #0, FEX-Emu#16 0x0000ffff80a80224 51000698 sub w24, w20, #0x1 (1) 0x0000ffff80a80228 9ad826b5 lsr x21, x21, x24 0x0000ffff80a8022c d34002b5 ubfx x21, x21, #0, FEX-Emu#1 0x0000ffff80a80230 7100069f cmp w20, #0x1 (1) 0x0000ffff80a80234 9a9622b4 csel x20, x21, x22, hs 0x0000ffff80a80238 390b0394 strb w20, [x28, FEX-Emu#704] 0x0000ffff80a8023c d34f3ef4 ubfx x20, x23, FEX-Emu#15, FEX-Emu#1 0x0000ffff80a80240 d34e3af5 ubfx x21, x23, FEX-Emu#14, FEX-Emu#1 0x0000ffff80a80244 ca150294 eor x20, x20, x21 0x0000ffff80a80248 390b2f94 strb w20, [x28, FEX-Emu#715] 0x0000ffff80a8024c 58000040 ldr x0, pc+8 (addr 0xffff80a80254) 0x0000ffff80a80250 d63f0000 blr x0 0x0000ffff80a80254 967da128 bl #-0x6097b60 (addr 0xffff7a9e86f4) 0x0000ffff80a80258 0000ffff udf #0xffff 0x0000ffff80a8025c 00010023 unallocated (Unallocated) 0x0000ffff80a80260 00000000 udf #0x0 [DEBUG] RIP: 0x10020 [DEBUG] Guest Code instructions: 1 [DEBUG] Host Code instructions: 34 [DEBUG] Blow-up Amt: 34x After 16-bit: 0x0000ffffa7c801e0 10ffffe0 adr x0, #-0x4 (addr 0xffffa7c801dc) 0x0000ffffa7c801e4 f9005f80 str x0, [x28, FEX-Emu#184] 0x0000ffffa7c801e8 d3403cb4 uxth x20, w5 0x0000ffffa7c801ec d3403cf5 uxth x21, w7 0x0000ffffa7c801f0 394b0396 ldrb w22, [x28, FEX-Emu#704] 0x0000ffffa7c801f4 12001294 and w20, w20, #0x1f 0x0000ffffa7c801f8 b37002d5 bfi x21, x22, FEX-Emu#16, FEX-Emu#1 0x0000ffffa7c801fc b36f42b5 bfi x21, x21, FEX-Emu#17, FEX-Emu#17 0x0000ffffa7c80200 b35e42b5 bfi x21, x21, FEX-Emu#34, FEX-Emu#17 0x0000ffffa7c80204 9ad426b7 lsr x23, x21, x20 0x0000ffffa7c80208 b3403ee7 bfxil x7, x23, #0, FEX-Emu#16 0x0000ffffa7c8020c 51000698 sub w24, w20, #0x1 (1) 0x0000ffffa7c80210 9ad826b5 lsr x21, x21, x24 0x0000ffffa7c80214 d34002b5 ubfx x21, x21, #0, FEX-Emu#1 0x0000ffffa7c80218 7100069f cmp w20, #0x1 (1) 0x0000ffffa7c8021c 9a9622b4 csel x20, x21, x22, hs 0x0000ffffa7c80220 390b0394 strb w20, [x28, FEX-Emu#704] 0x0000ffffa7c80224 d34f3ef4 ubfx x20, x23, FEX-Emu#15, FEX-Emu#1 0x0000ffffa7c80228 d34e3af5 ubfx x21, x23, FEX-Emu#14, FEX-Emu#1 0x0000ffffa7c8022c ca150294 eor x20, x20, x21 0x0000ffffa7c80230 390b2f94 strb w20, [x28, FEX-Emu#715] 0x0000ffffa7c80234 58000040 ldr x0, pc+8 (addr 0xffffa7c8023c) 0x0000ffffa7c80238 d63f0000 blr x0 0x0000ffffa7c8023c bd9cc128 unallocated (Unallocated) 0x0000ffffa7c80240 0000ffff udf #0xffff 0x0000ffffa7c80244 00010023 unallocated (Unallocated) 0x0000ffffa7c80248 00000000 udf #0x0 [DEBUG] RIP: 0x10020 [DEBUG] Guest Code instructions: 1 [DEBUG] Host Code instructions: 28 [DEBUG] Blow-up Amt: 28x Before 8-bit: 0x0000ffffa92801e0 10ffffe0 adr x0, #-0x4 (addr 0xffffa92801dc) 0x0000ffffa92801e4 f9005f80 str x0, [x28, FEX-Emu#184] 0x0000ffffa92801e8 d3401cb4 uxtb x20, w5 0x0000ffffa92801ec d3401cf5 uxtb x21, w7 0x0000ffffa92801f0 394b0396 ldrb w22, [x28, FEX-Emu#704] 0x0000ffffa92801f4 12001294 and w20, w20, #0x1f 0x0000ffffa92801f8 d2800017 mov x23, #0x0 0x0000ffffa92801fc b3401eb7 bfxil x23, x21, #0, FEX-Emu#8 0x0000ffffa9280200 b37802d7 bfi x23, x22, FEX-Emu#8, FEX-Emu#1 0x0000ffffa9280204 b3771eb7 bfi x23, x21, FEX-Emu#9, FEX-Emu#8 0x0000ffffa9280208 b36f02d7 bfi x23, x22, FEX-Emu#17, FEX-Emu#1 0x0000ffffa928020c b36e1eb7 bfi x23, x21, FEX-Emu#18, FEX-Emu#8 0x0000ffffa9280210 b36602d7 bfi x23, x22, FEX-Emu#26, FEX-Emu#1 0x0000ffffa9280214 b3651eb7 bfi x23, x21, FEX-Emu#27, FEX-Emu#8 0x0000ffffa9280218 b35d02d7 bfi x23, x22, FEX-Emu#35, FEX-Emu#1 0x0000ffffa928021c aa1703e0 mov x0, x23 0x0000ffffa9280220 b35c1ea0 bfi x0, x21, FEX-Emu#36, FEX-Emu#8 0x0000ffffa9280224 aa0003f5 mov x21, x0 0x0000ffffa9280228 b35402d5 bfi x21, x22, FEX-Emu#44, FEX-Emu#1 0x0000ffffa928022c 9ad426b7 lsr x23, x21, x20 0x0000ffffa9280230 b3401ee7 bfxil x7, x23, #0, FEX-Emu#8 0x0000ffffa9280234 51000698 sub w24, w20, #0x1 (1) 0x0000ffffa9280238 9ad826b5 lsr x21, x21, x24 0x0000ffffa928023c d34002b5 ubfx x21, x21, #0, FEX-Emu#1 0x0000ffffa9280240 7100069f cmp w20, #0x1 (1) 0x0000ffffa9280244 9a9622b4 csel x20, x21, x22, hs 0x0000ffffa9280248 390b0394 strb w20, [x28, FEX-Emu#704] 0x0000ffffa928024c d3471ef4 ubfx x20, x23, FEX-Emu#7, FEX-Emu#1 0x0000ffffa9280250 d3461af5 ubfx x21, x23, FEX-Emu#6, FEX-Emu#1 0x0000ffffa9280254 ca150294 eor x20, x20, x21 0x0000ffffa9280258 390b2f94 strb w20, [x28, FEX-Emu#715] 0x0000ffffa928025c 58000040 ldr x0, pc+8 (addr 0xffffa9280264) 0x0000ffffa9280260 d63f0000 blr x0 0x0000ffffa9280264 bf062128 unallocated (Unallocated) 0x0000ffffa9280268 0000ffff udf #0xffff 0x0000ffffa928026c 00010022 unallocated (Unallocated) 0x0000ffffa9280270 00000000 udf #0x0 [DEBUG] RIP: 0x10020 [DEBUG] Guest Code instructions: 1 [DEBUG] Host Code instructions: 38 [DEBUG] Blow-up Amt: 38x After 8-bit: 0x0000ffff9cc801e0 10ffffe0 adr x0, #-0x4 (addr 0xffff9cc801dc) 0x0000ffff9cc801e4 f9005f80 str x0, [x28, FEX-Emu#184] 0x0000ffff9cc801e8 d3401cb4 uxtb x20, w5 0x0000ffff9cc801ec d3401cf5 uxtb x21, w7 0x0000ffff9cc801f0 394b0396 ldrb w22, [x28, FEX-Emu#704] 0x0000ffff9cc801f4 12001294 and w20, w20, #0x1f 0x0000ffff9cc801f8 b37802d5 bfi x21, x22, FEX-Emu#8, FEX-Emu#1 0x0000ffff9cc801fc b37722b5 bfi x21, x21, FEX-Emu#9, FEX-Emu#9 0x0000ffff9cc80200 b36e46b5 bfi x21, x21, FEX-Emu#18, FEX-Emu#18 0x0000ffff9cc80204 b3778eb5 bfi x21, x21, FEX-Emu#9, FEX-Emu#36 0x0000ffff9cc80208 9ad426b7 lsr x23, x21, x20 0x0000ffff9cc8020c b3401ee7 bfxil x7, x23, #0, FEX-Emu#8 0x0000ffff9cc80210 51000698 sub w24, w20, #0x1 (1) 0x0000ffff9cc80214 9ad826b5 lsr x21, x21, x24 0x0000ffff9cc80218 d34002b5 ubfx x21, x21, #0, FEX-Emu#1 0x0000ffff9cc8021c 7100069f cmp w20, #0x1 (1) 0x0000ffff9cc80220 9a9622b4 csel x20, x21, x22, hs 0x0000ffff9cc80224 390b0394 strb w20, [x28, FEX-Emu#704] 0x0000ffff9cc80228 d3471ef4 ubfx x20, x23, FEX-Emu#7, FEX-Emu#1 0x0000ffff9cc8022c d3461af5 ubfx x21, x23, FEX-Emu#6, FEX-Emu#1 0x0000ffff9cc80230 ca150294 eor x20, x20, x21 0x0000ffff9cc80234 390b2f94 strb w20, [x28, FEX-Emu#715] 0x0000ffff9cc80238 58000040 ldr x0, pc+8 (addr 0xffff9cc80240) 0x0000ffff9cc8023c d63f0000 blr x0 0x0000ffff9cc80240 b2a75128 unallocated (Unallocated) 0x0000ffff9cc80244 0000ffff udf #0xffff 0x0000ffff9cc80248 00010022 unallocated (Unallocated) 0x0000ffff9cc8024c 00000000 udf #0x0 [DEBUG] RIP: 0x10020 [DEBUG] Guest Code instructions: 1 [DEBUG] Host Code instructions: 29 [DEBUG] Blow-up Amt: 29x ```
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Overview
This adds gcc's "gcc.target" tests for i386 and amd64. We mostly pass these, however there are a few failures that should be investigated (#713). The 32-bit tests don't work on CI but work for me locally, so that also needs to be investigated (#714).
There's test cases all the way up to avx512, though I haven't included anything avx*