danog/ir
1
0
Fork 0
mirror of https://github.com/danog/ir.git synced 2024-12-02 17:55:40 +01:00
Code Issues Packages Projects Releases Wiki Activity
976 Commits 2 Branches 0 Tags 3.2 MiB
a746ceb650
Commit Graph

166 Commits

Author SHA1 Message Date
Dmitry Stogov
c9fa8dfebd Fixed SSA deconstruction 
Previously we performed parallel copy for virtual registers, now we do
the same for the target CPU registers.
 2023-05-17 22:37:45 +03:00
Dmitry Stogov
53f4435f8e Fixed test 2023-04-26 14:24:43 +03:00
Dmitry Stogov
60802d942f Fix previous commit. We still need a temporary register for indirect calls. 2023-04-26 14:10:58 +03:00
Dmitry Stogov
9eb366698d Avoid reservaton of temporary resiser for argument passing 
We may use any scratch register that is not used for parameters
 2023-04-26 12:16:05 +03:00
Dmitry Stogov
0de0c1d0fa Improve parallel copy algorithm to support move of single source into multiple destinations 2023-04-26 10:56:55 +03:00
Dmitry Stogov
1f7a2bd243 Remove useless "AVX" tests for AArch64 2023-04-18 10:14:59 +03:00
Dmitry Stogov
f85f5fd2a8 Remove data dependency between TAILCALL and UNREACHABLE 2023-04-13 02:41:28 +03:00
Dmitry Stogov
efa8a83153 Fix spilling code for arguments passed theought stack and change RA to 
prefer reusing the same register for splitted intervals
i#	utils/
 2023-04-06 00:16:49 +03:00
Dmitry Stogov
1e5e9e08ce Re-implement instruction fusion and live-range construction 2023-04-05 19:20:43 +03:00
Dmitry Stogov
ee827ee983 Don't create two DEF UsePos 2023-03-29 17:22:49 +03:00
Dmitry Stogov
d79bd88f6f Improve x86 code generation for passing address of label to stack 
-       leal .L1, %eax
-       movl %eax, (%esp)
+       movl $.L1, (%esp)
 2023-03-29 15:48:41 +03:00
Dmitry Stogov
5bed3d73a5 Fix test 2023-03-29 15:13:17 +03:00
Dmitry Stogov
26e462fa42 Add more folding rules 2023-03-29 14:07:31 +03:00
Dmitry Stogov
e4b618ad00 Fix fusion of IF(_, CMP(AND(_, _) 0)) 2023-03-28 19:03:06 +03:00
Dmitry Stogov
6d36fb12c3 Fix example code and test 2023-03-23 00:54:47 +03:00
Dmitry Stogov
24e8e216a1 Remove a "reference" edge from LOOP_END to LOOP_BEGIN node. 2023-03-23 00:47:27 +03:00
Dmitry Stogov
ccbf3da286 Add test 2023-03-22 22:09:25 +03:00
Dmitry Stogov
1542048331 Fix TAILCALL on WIN64 2023-03-02 22:08:24 +03:00
Dmitry Stogov
29122c15c3 Add complex WIN64 tests 2023-03-02 17:57:26 +03:00
Dmitry Stogov
c89a038fd3 Fix tests with capstone 5 
Different versions of capstone may disassemble MOVD/MOVQ differentrly
 2023-03-02 17:54:50 +03:00
Dmitry Stogov
5a48805c81 Add support for Windows-64 ABI ("home space") 
Fix parameter passing code to perform sign or zero extension when pass a regiser or a constant

TODO: ARM code maight need similar changes
 2023-03-02 13:27:01 +03:00
Dmitry Stogov
ebdeba9fff Add simple tests for Windows-x86_64 
conv_004.irt and conv_010.irt fail with capstone 5 because of changes in movd/movq
 2023-03-02 00:55:20 +03:00
Dmitry Stogov
00d5e471ad Improve load fusion, register allocateion and code selection for ADD 2023-02-21 22:55:47 +03:00
Dmitry Stogov
2f2fed89bb Uze zero extended "mov" to load 64-bit register ("mov $u32, %r32") 2023-02-17 18:11:13 +03:00
Dmitry Stogov
1d7ab16c2a Allow load fuson for CALL and TAILCALL with arguments 2023-02-14 14:51:12 +03:00
Dmitry Stogov
2e31446e37 Better 'jp' elimination for IR_CMP_AND_BRANCH_FP 2023-02-07 01:57:07 +03:00
Dmitry Stogov
02104b0950 Add XFAIL-ed test for a non-efficient register allocation that should be improved 2023-02-07 00:06:53 +03:00
Dmitry Stogov
54597bc862 Clear destination regeister before INT to FP conversion to avoid partial register stall 2022-12-28 00:05:23 +03:00
Dmitry Stogov
67da9e93ea Fix register clobbering during argument passing and spill load 2022-12-26 20:25:11 +03:00
Dmitry Stogov
d26b162ffa Fix register clobbering during argument passing 2022-12-26 18:27:53 +03:00
Dmitry Stogov
95729f76bf Use IMUL instead of MUL 2022-12-16 12:57:40 +03:00
Dmitry Stogov
efbc51baaa Fixed codegeneration for TRUNC on aarch64 2022-12-07 11:56:53 +03:00
Dmitry Stogov
daf659a457 Fix incorrect conditions 2022-12-01 00:43:42 +03:00
Dmitry Stogov
c6aeb417fa Extend test suite to support XFAIL section 2022-11-22 10:43:29 +03:00
Dmitry Stogov
3e3746d5cb Refactor API that expose target CPU register constraints for register allocator 2022-11-17 23:30:35 +03:00
Dmitry Stogov
05127b1b13 Remove duplicate code and allow load fusion of IR_SHIFT.op2 2022-11-16 13:20:58 +03:00
Dmitry Stogov
673779ba6a Use IR_COPY_INT/FP rule instead of IR_COPY op 2022-11-16 12:55:40 +03:00
Dmitry Stogov
4c536aae20 Extend SCCP to perform Dead Load Elimination 2022-11-08 15:39:00 +03:00
Dmitry Stogov
37dececa71 Add more tests (8 tests ara failed on 32-bit x86) 2022-11-08 11:56:22 +03:00
Dmitry Stogov
2ff0617db6 Perform iterative folding and DCE as a final pass of SCCP 2022-11-08 00:41:08 +03:00
Dmitry Stogov
924f5949f2 Fixed SSE operands alignment and 32-bit support 2022-09-27 20:36:34 +03:00
Dmitry Stogov
05fd1f971d Better LOAD fusion 2022-09-21 23:54:45 +03:00
Dmitry Stogov
69a3d6fd27 Verify type compatibility 2022-09-02 09:50:38 +03:00
Dmitry Stogov
32198c00b7 Reimplement JMP optimization 2022-08-30 23:15:20 +03:00
Dmitry Stogov
0596de2291 Fuse LOAD into IMULL/3 2022-08-30 11:26:38 +03:00
Dmitry Stogov
fd8539e17d Eliminate TEST after ADD/SUB/AND/OR/XOR 2022-08-29 22:22:30 +03:00
Dmitry Stogov
47083e0f9f Improve LOAD fusion 2022-08-25 18:16:17 +03:00
Dmitry Stogov
65e1619de8 Fuse address calculation into LOAD/STORE 2022-08-24 16:11:04 +03:00
Dmitry Stogov
32e045d93e typo 2022-08-23 17:02:34 +03:00
Dmitry Stogov
88b8731c16 Fix incorrect condition codes 2022-08-02 13:04:03 +03:00
Dmitry Stogov
9b25587eb6 Compound assignment instruction fusion 2022-06-21 17:33:57 +03:00
Dmitry Stogov
5fb115ab11 Remove LOOP_EXIT 2022-06-15 17:27:31 +03:00
Dmitry Stogov
c28fe2734d Validate operand types 2022-06-03 11:23:05 +03:00
Dmitry Stogov
ab8019e0cd Aarch64 back-end (incomplete) 2022-06-02 15:12:56 +03:00
Dmitry Stogov
bb842b489c Aarch64 backend support & unification 2022-06-01 18:16:32 +03:00
Dmitry Stogov
f5bbdeea27 Fix buffer overflow 2022-05-26 17:19:43 +03:00
Dmitry Stogov
7e782a291a Extend disassembler to support .rodata section and IP relative data labels 2022-05-26 01:17:02 +03:00
Dmitry Stogov
19e93fd3f6 Allow multi-target test suite 2022-05-25 17:38:22 +03:00
Dmitry Stogov
463002107a Rename "gcm_blocks" into "cfg_map" 2022-05-25 09:33:47 +03:00
Dmitry Stogov
04667faf22 Reorder blocks according to branch probability 2022-05-24 12:47:39 +03:00
Dmitry Stogov
d3c1e4a02f Reorder basic blocks to reduce number of jumps and improve code locality 2022-05-24 00:43:35 +03:00
Dmitry Stogov
6f7f7b1268 Implement code generation for type conversion instructions 
Register constraints might need to be tweeked.
 2022-05-20 13:07:41 +03:00
Dmitry Stogov
c464b123cb Add test for TRUNC 2022-05-20 09:09:52 +03:00
Dmitry Stogov
d250f77713 Improve type conversion nodes 2022-05-20 09:00:13 +03:00
Dmitry Stogov
c6b0e95d6b Add type conversion nodes (no code generation yet) 2022-05-20 01:01:48 +03:00
Dmitry Stogov
911219493d Implement IJMP instruction (indirect jump or computed goto) 2022-05-19 18:56:48 +03:00
Dmitry Stogov
bae7df6a5f Implement code generation for MIN and MAX instructions 2022-05-19 17:03:00 +03:00
Dmitry Stogov
bf369d0eac Swap operands for better load fusion 2022-05-19 13:17:50 +03:00
Dmitry Stogov
c9bb858e50 Fuse loads without register allocation when this makes sense. 
Make oarameters passed through stack to reuse the same stack slot for spilling.
 2022-05-19 10:53:08 +03:00
Dmitry Stogov
88ab04a3c2 New tests 2022-05-18 23:53:16 +03:00
Dmitry Stogov
cdd39f22b0 Merge spills for VSTORE with -O0 2022-05-18 23:12:20 +03:00
Dmitry Stogov
c5a24ff734 Add support for instructions that modify result directly in memory 2022-05-18 21:49:08 +03:00
Dmitry Stogov
5319951060 Align stack once 2022-05-17 23:01:37 +03:00
Dmitry Stogov
e794451451 Preallocate call stack 2022-05-17 22:37:13 +03:00
Dmitry Stogov
445dd65c78 Improve argument passing 2022-05-17 17:30:04 +03:00
Dmitry Stogov
4e917faaba Fix stack parameters loading 2022-05-17 15:00:58 +03:00
Dmitry Stogov
1e7059d7e0 Pass arguments through stack in reverse order 2022-05-17 12:34:31 +03:00
Dmitry Stogov
6fb5380906 Take into account spill slot size and alignment 2022-05-16 22:16:29 +03:00
Dmitry Stogov
8496780ece Fix temporary register usage for parralel arguments passing 2022-05-16 15:34:36 +03:00
Dmitry Stogov
f086da2550 Clenaup (remove unnecessary SHIFT case) 2022-05-16 14:36:27 +03:00
Dmitry Stogov
a3b597feef Use different interval for registers clobbered by CALL 2022-05-13 15:53:54 +03:00
Dmitry Stogov
896ddb9e77 Flexable scratch register constraints (allow MUL %edx) 2022-05-13 15:10:15 +03:00
Dmitry Stogov
1f673ebfda Better temporary register usage for SSA deconstruction 2022-05-13 00:32:37 +03:00
Dmitry Stogov
69b5a852e5 Make DESSA API use "ir_ref" instead of "virtual register number" 
(0 - is still a temporary register)
 2022-05-06 16:19:57 +03:00
Dmitry Stogov
2403fa1edc Fix spill loads during argument passing 2022-05-06 12:55:07 +03:00
Dmitry Stogov
b580c926e6 Avoid need for temporary register for parameters loading 2022-05-06 11:27:24 +03:00
Dmitry Stogov
89f320d7b7 Add SWITCH support for temporary registers 2022-05-06 10:00:19 +03:00
Dmitry Stogov
dd5a3a3b72 Add flexible support for temporary registers. 
Get rid of hardcoded temporary registers (incomplete)
 2022-05-05 22:35:39 +03:00
Dmitry Stogov
7de4566498 Add tests for 64-bit constants 2022-05-04 15:37:07 +03:00
Dmitry Stogov
310f605d6c Fix register clobbering 2022-04-26 22:49:41 +03:00
Dmitry Stogov
4a6c8d60a6 Fix ALLOCA to align stack frame 2022-04-22 12:55:38 +03:00
Dmitry Stogov
549ac2efd9 Add test 2022-04-22 11:32:59 +03:00
Dmitry Stogov
c47de38bab Merge spill slots for VAR, VLOAD and VSTORE (this may be unsafe) 2022-04-22 11:30:33 +03:00
Dmitry Stogov
034ef95e07 Allow memory update instructions (without loading into register) 2022-04-22 01:40:10 +03:00
Dmitry Stogov
84b2bac02c Add more tests 2022-04-22 00:11:34 +03:00
Dmitry Stogov
6f3cc3052c Implement ABS for C code generator 
Remove POW
 2022-04-21 01:00:46 +03:00
Dmitry Stogov
506e7b658f Implement ABS and NEG 2022-04-21 00:31:28 +03:00
Dmitry Stogov
ffdb53821d Refactor constraint model 
Each instruction consist from 4 sub positions LOAD, USE, DEF, SAVE.
Hardware constraints are modeled conectiong live intervals and fixed
intervals to different sub-positions.
 2022-04-20 18:53:15 +03:00
Dmitry Stogov
e449345514 Fix few CSSP bugs 2022-04-19 16:45:03 +03:00
Dmitry Stogov
0922b7cd7f Add vreg hints 2022-04-15 16:02:23 +03:00