[LoongArch] Minor refinement to monotonic atomic semantics. #112681
Merged
Conversation
This file contains hidden or 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
|
Thank you for submitting a Pull Request (PR) to the LLVM Project! This PR will be automatically labeled and the relevant teams will be notified. If you wish to, you can add reviewers by using the "Reviewers" section on this page. If this is not working for you, it is probably because you do not have write permissions for the repository. In which case you can instead tag reviewers by name in a comment by using If you have received no comments on your PR for a week, you can request a review by "ping"ing the PR by adding a comment “Ping”. The common courtesy "ping" rate is once a week. Please remember that you are asking for valuable time from other developers. If you have further questions, they may be answered by the LLVM GitHub User Guide. You can also ask questions in a comment on this PR, on the LLVM Discord or on the forums. |
|
@llvm/pr-subscribers-backend-loongarch Author: None (tangaac) ChangesSupport LoongArch atomic memory operations with monotonic semantics. Full diff: https://github.com/llvm/llvm-project/pull/112681.diff 3 Files Affected:
diff --git a/llvm/lib/Target/LoongArch/LoongArchInstrInfo.td b/llvm/lib/Target/LoongArch/LoongArchInstrInfo.td
index 15a8f4e3c0755d..a93f552b5a8fe8 100644
--- a/llvm/lib/Target/LoongArch/LoongArchInstrInfo.td
+++ b/llvm/lib/Target/LoongArch/LoongArchInstrInfo.td
@@ -2076,6 +2076,48 @@ defm atomic_cmp_swap_i32 : ternary_atomic_op_failure_ord;
defm atomic_cmp_swap_i64 : ternary_atomic_op_failure_ord;
let Predicates = [IsLA64] in {
+
+def : Pat<(atomic_swap_i32_monotonic GPR:$addr, GPR:$incr),
+ (AMSWAP_W GPR:$incr, GPR:$addr)>;
+def : Pat<(atomic_swap_i64_monotonic GPR:$addr, GPR:$incr),
+ (AMSWAP_D GPR:$incr, GPR:$addr)>;
+def : Pat<(atomic_load_add_i32_monotonic GPR:$rj, GPR:$rk),
+ (AMADD_W GPR:$rk, GPR:$rj)>;
+def : Pat<(atomic_load_add_i64_monotonic GPR:$rj, GPR:$rk),
+ (AMADD_D GPR:$rk, GPR:$rj)>;
+def : Pat<(atomic_load_sub_i32_monotonic GPR:$rj, GPR:$rk),
+ (AMADD_W (SUB_W R0, GPR:$rk), GPR:$rj)>;
+def : Pat<(atomic_load_sub_i64_monotonic GPR:$rj, GPR:$rk),
+ (AMADD_D (SUB_D R0, GPR:$rk), GPR:$rj)>;
+def : Pat<(atomic_load_and_i32_monotonic GPR:$rj, GPR:$rk),
+ (AMAND_W GPR:$rk, GPR:$rj)>;
+def : Pat<(atomic_load_and_i64_monotonic GPR:$rj, GPR:$rk),
+ (AMAND_D GPR:$rk, GPR:$rj)>;
+def : Pat<(atomic_load_or_i32_monotonic GPR:$rj, GPR:$rk),
+ (AMOR_W GPR:$rk, GPR:$rj)>;
+def : Pat<(atomic_load_or_i64_monotonic GPR:$rj, GPR:$rk),
+ (AMOR_D GPR:$rk, GPR:$rj)>;
+def : Pat<(atomic_load_xor_i32_monotonic GPR:$rj, GPR:$rk),
+ (AMXOR_W GPR:$rk, GPR:$rj)>;
+def : Pat<(atomic_load_xor_i64_monotonic GPR:$rj, GPR:$rk),
+ (AMXOR_D GPR:$rk, GPR:$rj)>;
+def : Pat<(atomic_load_umin_i32_monotonic GPR:$rj, GPR:$rk),
+ (AMMIN_WU GPR:$rk, GPR:$rj)>;
+def : Pat<(atomic_load_umin_i64_monotonic GPR:$rj, GPR:$rk),
+ (AMMIN_DU GPR:$rk, GPR:$rj)>;
+def : Pat<(atomic_load_umax_i32_monotonic GPR:$rj, GPR:$rk),
+ (AMMAX_WU GPR:$rk, GPR:$rj)>;
+def : Pat<(atomic_load_umax_i64_monotonic GPR:$rj, GPR:$rk),
+ (AMMAX_DU GPR:$rk, GPR:$rj)>;
+def : Pat<(atomic_load_min_i32_monotonic GPR:$rj, GPR:$rk),
+ (AMMIN_W GPR:$rk, GPR:$rj)>;
+def : Pat<(atomic_load_min_i64_monotonic GPR:$rj, GPR:$rk),
+ (AMMIN_D GPR:$rk, GPR:$rj)>;
+def : Pat<(atomic_load_max_i32_monotonic GPR:$rj, GPR:$rk),
+ (AMMAX_W GPR:$rk, GPR:$rj)>;
+def : Pat<(atomic_load_max_i64_monotonic GPR:$rj, GPR:$rk),
+ (AMMAX_D GPR:$rk, GPR:$rj)>;
+
def : AtomicPat<int_loongarch_masked_atomicrmw_xchg_i64,
PseudoMaskedAtomicSwap32>;
def : Pat<(atomic_swap_i32 GPR:$addr, GPR:$incr),
diff --git a/llvm/test/CodeGen/LoongArch/ir-instruction/atomicrmw-minmax.ll b/llvm/test/CodeGen/LoongArch/ir-instruction/atomicrmw-minmax.ll
index 2bd29c2670a680..03386514a72c80 100644
--- a/llvm/test/CodeGen/LoongArch/ir-instruction/atomicrmw-minmax.ll
+++ b/llvm/test/CodeGen/LoongArch/ir-instruction/atomicrmw-minmax.ll
@@ -1368,7 +1368,7 @@ define i16 @atomicrmw_umax_i16_monotonic(ptr %a, i16 %b) nounwind {
define i32 @atomicrmw_umax_i32_monotonic(ptr %a, i32 %b) nounwind {
; LA64-LABEL: atomicrmw_umax_i32_monotonic:
; LA64: # %bb.0:
-; LA64-NEXT: ammax_db.wu $a2, $a1, $a0
+; LA64-NEXT: ammax.wu $a2, $a1, $a0
; LA64-NEXT: move $a0, $a2
; LA64-NEXT: ret
%1 = atomicrmw umax ptr %a, i32 %b monotonic
@@ -1378,7 +1378,7 @@ define i32 @atomicrmw_umax_i32_monotonic(ptr %a, i32 %b) nounwind {
define i64 @atomicrmw_umax_i64_monotonic(ptr %a, i64 %b) nounwind {
; LA64-LABEL: atomicrmw_umax_i64_monotonic:
; LA64: # %bb.0:
-; LA64-NEXT: ammax_db.du $a2, $a1, $a0
+; LA64-NEXT: ammax.du $a2, $a1, $a0
; LA64-NEXT: move $a0, $a2
; LA64-NEXT: ret
%1 = atomicrmw umax ptr %a, i64 %b monotonic
@@ -1445,7 +1445,7 @@ define i16 @atomicrmw_umin_i16_monotonic(ptr %a, i16 %b) nounwind {
define i32 @atomicrmw_umin_i32_monotonic(ptr %a, i32 %b) nounwind {
; LA64-LABEL: atomicrmw_umin_i32_monotonic:
; LA64: # %bb.0:
-; LA64-NEXT: ammin_db.wu $a2, $a1, $a0
+; LA64-NEXT: ammin.wu $a2, $a1, $a0
; LA64-NEXT: move $a0, $a2
; LA64-NEXT: ret
%1 = atomicrmw umin ptr %a, i32 %b monotonic
@@ -1455,7 +1455,7 @@ define i32 @atomicrmw_umin_i32_monotonic(ptr %a, i32 %b) nounwind {
define i64 @atomicrmw_umin_i64_monotonic(ptr %a, i64 %b) nounwind {
; LA64-LABEL: atomicrmw_umin_i64_monotonic:
; LA64: # %bb.0:
-; LA64-NEXT: ammin_db.du $a2, $a1, $a0
+; LA64-NEXT: ammin.du $a2, $a1, $a0
; LA64-NEXT: move $a0, $a2
; LA64-NEXT: ret
%1 = atomicrmw umin ptr %a, i64 %b monotonic
@@ -1531,7 +1531,7 @@ define i16 @atomicrmw_max_i16_monotonic(ptr %a, i16 %b) nounwind {
define i32 @atomicrmw_max_i32_monotonic(ptr %a, i32 %b) nounwind {
; LA64-LABEL: atomicrmw_max_i32_monotonic:
; LA64: # %bb.0:
-; LA64-NEXT: ammax_db.w $a2, $a1, $a0
+; LA64-NEXT: ammax.w $a2, $a1, $a0
; LA64-NEXT: move $a0, $a2
; LA64-NEXT: ret
%1 = atomicrmw max ptr %a, i32 %b monotonic
@@ -1541,7 +1541,7 @@ define i32 @atomicrmw_max_i32_monotonic(ptr %a, i32 %b) nounwind {
define i64 @atomicrmw_max_i64_monotonic(ptr %a, i64 %b) nounwind {
; LA64-LABEL: atomicrmw_max_i64_monotonic:
; LA64: # %bb.0:
-; LA64-NEXT: ammax_db.d $a2, $a1, $a0
+; LA64-NEXT: ammax.d $a2, $a1, $a0
; LA64-NEXT: move $a0, $a2
; LA64-NEXT: ret
%1 = atomicrmw max ptr %a, i64 %b monotonic
@@ -1617,7 +1617,7 @@ define i16 @atomicrmw_min_i16_monotonic(ptr %a, i16 %b) nounwind {
define i32 @atomicrmw_min_i32_monotonic(ptr %a, i32 %b) nounwind {
; LA64-LABEL: atomicrmw_min_i32_monotonic:
; LA64: # %bb.0:
-; LA64-NEXT: ammin_db.w $a2, $a1, $a0
+; LA64-NEXT: ammin.w $a2, $a1, $a0
; LA64-NEXT: move $a0, $a2
; LA64-NEXT: ret
%1 = atomicrmw min ptr %a, i32 %b monotonic
@@ -1627,7 +1627,7 @@ define i32 @atomicrmw_min_i32_monotonic(ptr %a, i32 %b) nounwind {
define i64 @atomicrmw_min_i64_monotonic(ptr %a, i64 %b) nounwind {
; LA64-LABEL: atomicrmw_min_i64_monotonic:
; LA64: # %bb.0:
-; LA64-NEXT: ammin_db.d $a2, $a1, $a0
+; LA64-NEXT: ammin.d $a2, $a1, $a0
; LA64-NEXT: move $a0, $a2
; LA64-NEXT: ret
%1 = atomicrmw min ptr %a, i64 %b monotonic
diff --git a/llvm/test/CodeGen/LoongArch/ir-instruction/atomicrmw.ll b/llvm/test/CodeGen/LoongArch/ir-instruction/atomicrmw.ll
index f2f459ecaa2eca..b1af9c17b60183 100644
--- a/llvm/test/CodeGen/LoongArch/ir-instruction/atomicrmw.ll
+++ b/llvm/test/CodeGen/LoongArch/ir-instruction/atomicrmw.ll
@@ -3982,7 +3982,7 @@ define i8 @atomicrmw_xchg_0_i8_monotonic(ptr %a) nounwind {
; LA64-NEXT: ori $a2, $zero, 255
; LA64-NEXT: sll.w $a2, $a2, $a1
; LA64-NEXT: nor $a2, $a2, $zero
-; LA64-NEXT: amand_db.w $a3, $a2, $a0
+; LA64-NEXT: amand.w $a3, $a2, $a0
; LA64-NEXT: srl.w $a0, $a3, $a1
; LA64-NEXT: ret
%1 = atomicrmw xchg ptr %a, i8 0 monotonic
@@ -4011,7 +4011,7 @@ define i8 @atomicrmw_xchg_minus_1_i8_monotonic(ptr %a) nounwind {
; LA64-NEXT: bstrins.d $a0, $zero, 1, 0
; LA64-NEXT: ori $a2, $zero, 255
; LA64-NEXT: sll.w $a2, $a2, $a1
-; LA64-NEXT: amor_db.w $a3, $a2, $a0
+; LA64-NEXT: amor.w $a3, $a2, $a0
; LA64-NEXT: srl.w $a0, $a3, $a1
; LA64-NEXT: ret
%1 = atomicrmw xchg ptr %a, i8 -1 monotonic
@@ -4090,7 +4090,7 @@ define i16 @atomicrmw_xchg_0_i16_monotonic(ptr %a) nounwind {
; LA64-NEXT: ori $a2, $a2, 4095
; LA64-NEXT: sll.w $a2, $a2, $a1
; LA64-NEXT: nor $a2, $a2, $zero
-; LA64-NEXT: amand_db.w $a3, $a2, $a0
+; LA64-NEXT: amand.w $a3, $a2, $a0
; LA64-NEXT: srl.w $a0, $a3, $a1
; LA64-NEXT: ret
%1 = atomicrmw xchg ptr %a, i16 0 monotonic
@@ -4121,7 +4121,7 @@ define i16 @atomicrmw_xchg_minus_1_i16_monotonic(ptr %a) nounwind {
; LA64-NEXT: lu12i.w $a2, 15
; LA64-NEXT: ori $a2, $a2, 4095
; LA64-NEXT: sll.w $a2, $a2, $a1
-; LA64-NEXT: amor_db.w $a3, $a2, $a0
+; LA64-NEXT: amor.w $a3, $a2, $a0
; LA64-NEXT: srl.w $a0, $a3, $a1
; LA64-NEXT: ret
%1 = atomicrmw xchg ptr %a, i16 -1 monotonic
@@ -4142,7 +4142,7 @@ define i32 @atomicrmw_xchg_i32_monotonic(ptr %a, i32 %b) nounwind {
;
; LA64-LABEL: atomicrmw_xchg_i32_monotonic:
; LA64: # %bb.0:
-; LA64-NEXT: amswap_db.w $a2, $a1, $a0
+; LA64-NEXT: amswap.w $a2, $a1, $a0
; LA64-NEXT: move $a0, $a2
; LA64-NEXT: ret
%1 = atomicrmw xchg ptr %a, i32 %b monotonic
@@ -4162,7 +4162,7 @@ define i64 @atomicrmw_xchg_i64_monotonic(ptr %a, i64 %b) nounwind {
;
; LA64-LABEL: atomicrmw_xchg_i64_monotonic:
; LA64: # %bb.0:
-; LA64-NEXT: amswap_db.d $a2, $a1, $a0
+; LA64-NEXT: amswap.d $a2, $a1, $a0
; LA64-NEXT: move $a0, $a2
; LA64-NEXT: ret
%1 = atomicrmw xchg ptr %a, i64 %b monotonic
@@ -4273,7 +4273,7 @@ define i32 @atomicrmw_add_i32_monotonic(ptr %a, i32 %b) nounwind {
;
; LA64-LABEL: atomicrmw_add_i32_monotonic:
; LA64: # %bb.0:
-; LA64-NEXT: amadd_db.w $a2, $a1, $a0
+; LA64-NEXT: amadd.w $a2, $a1, $a0
; LA64-NEXT: move $a0, $a2
; LA64-NEXT: ret
%1 = atomicrmw add ptr %a, i32 %b monotonic
@@ -4293,7 +4293,7 @@ define i64 @atomicrmw_add_i64_monotonic(ptr %a, i64 %b) nounwind {
;
; LA64-LABEL: atomicrmw_add_i64_monotonic:
; LA64: # %bb.0:
-; LA64-NEXT: amadd_db.d $a2, $a1, $a0
+; LA64-NEXT: amadd.d $a2, $a1, $a0
; LA64-NEXT: move $a0, $a2
; LA64-NEXT: ret
%1 = atomicrmw add ptr %a, i64 %b monotonic
@@ -4405,7 +4405,7 @@ define i32 @atomicrmw_sub_i32_monotonic(ptr %a, i32 %b) nounwind {
; LA64-LABEL: atomicrmw_sub_i32_monotonic:
; LA64: # %bb.0:
; LA64-NEXT: sub.w $a2, $zero, $a1
-; LA64-NEXT: amadd_db.w $a1, $a2, $a0
+; LA64-NEXT: amadd.w $a1, $a2, $a0
; LA64-NEXT: move $a0, $a1
; LA64-NEXT: ret
%1 = atomicrmw sub ptr %a, i32 %b monotonic
@@ -4426,7 +4426,7 @@ define i64 @atomicrmw_sub_i64_monotonic(ptr %a, i64 %b) nounwind {
; LA64-LABEL: atomicrmw_sub_i64_monotonic:
; LA64: # %bb.0:
; LA64-NEXT: sub.d $a2, $zero, $a1
-; LA64-NEXT: amadd_db.d $a1, $a2, $a0
+; LA64-NEXT: amadd.d $a1, $a2, $a0
; LA64-NEXT: move $a0, $a1
; LA64-NEXT: ret
%1 = atomicrmw sub ptr %a, i64 %b monotonic
@@ -4609,7 +4609,7 @@ define i8 @atomicrmw_and_i8_monotonic(ptr %a, i8 %b) nounwind {
; LA64-NEXT: andi $a1, $a1, 255
; LA64-NEXT: sll.w $a1, $a1, $a2
; LA64-NEXT: orn $a1, $a1, $a3
-; LA64-NEXT: amand_db.w $a3, $a1, $a0
+; LA64-NEXT: amand.w $a3, $a1, $a0
; LA64-NEXT: srl.w $a0, $a3, $a2
; LA64-NEXT: ret
%1 = atomicrmw and ptr %a, i8 %b monotonic
@@ -4646,7 +4646,7 @@ define i16 @atomicrmw_and_i16_monotonic(ptr %a, i16 %b) nounwind {
; LA64-NEXT: bstrpick.d $a1, $a1, 15, 0
; LA64-NEXT: sll.w $a1, $a1, $a2
; LA64-NEXT: orn $a1, $a1, $a3
-; LA64-NEXT: amand_db.w $a3, $a1, $a0
+; LA64-NEXT: amand.w $a3, $a1, $a0
; LA64-NEXT: srl.w $a0, $a3, $a2
; LA64-NEXT: ret
%1 = atomicrmw and ptr %a, i16 %b monotonic
@@ -4667,7 +4667,7 @@ define i32 @atomicrmw_and_i32_monotonic(ptr %a, i32 %b) nounwind {
;
; LA64-LABEL: atomicrmw_and_i32_monotonic:
; LA64: # %bb.0:
-; LA64-NEXT: amand_db.w $a2, $a1, $a0
+; LA64-NEXT: amand.w $a2, $a1, $a0
; LA64-NEXT: move $a0, $a2
; LA64-NEXT: ret
%1 = atomicrmw and ptr %a, i32 %b monotonic
@@ -4687,7 +4687,7 @@ define i64 @atomicrmw_and_i64_monotonic(ptr %a, i64 %b) nounwind {
;
; LA64-LABEL: atomicrmw_and_i64_monotonic:
; LA64: # %bb.0:
-; LA64-NEXT: amand_db.d $a2, $a1, $a0
+; LA64-NEXT: amand.d $a2, $a1, $a0
; LA64-NEXT: move $a0, $a2
; LA64-NEXT: ret
%1 = atomicrmw and ptr %a, i64 %b monotonic
@@ -4716,7 +4716,7 @@ define i8 @atomicrmw_or_i8_monotonic(ptr %a, i8 %b) nounwind {
; LA64-NEXT: bstrins.d $a0, $zero, 1, 0
; LA64-NEXT: andi $a1, $a1, 255
; LA64-NEXT: sll.w $a1, $a1, $a2
-; LA64-NEXT: amor_db.w $a3, $a1, $a0
+; LA64-NEXT: amor.w $a3, $a1, $a0
; LA64-NEXT: srl.w $a0, $a3, $a2
; LA64-NEXT: ret
%1 = atomicrmw or ptr %a, i8 %b monotonic
@@ -4745,7 +4745,7 @@ define i16 @atomicrmw_or_i16_monotonic(ptr %a, i16 %b) nounwind {
; LA64-NEXT: bstrins.d $a0, $zero, 1, 0
; LA64-NEXT: bstrpick.d $a1, $a1, 15, 0
; LA64-NEXT: sll.w $a1, $a1, $a2
-; LA64-NEXT: amor_db.w $a3, $a1, $a0
+; LA64-NEXT: amor.w $a3, $a1, $a0
; LA64-NEXT: srl.w $a0, $a3, $a2
; LA64-NEXT: ret
%1 = atomicrmw or ptr %a, i16 %b monotonic
@@ -4766,7 +4766,7 @@ define i32 @atomicrmw_or_i32_monotonic(ptr %a, i32 %b) nounwind {
;
; LA64-LABEL: atomicrmw_or_i32_monotonic:
; LA64: # %bb.0:
-; LA64-NEXT: amor_db.w $a2, $a1, $a0
+; LA64-NEXT: amor.w $a2, $a1, $a0
; LA64-NEXT: move $a0, $a2
; LA64-NEXT: ret
%1 = atomicrmw or ptr %a, i32 %b monotonic
@@ -4786,7 +4786,7 @@ define i64 @atomicrmw_or_i64_monotonic(ptr %a, i64 %b) nounwind {
;
; LA64-LABEL: atomicrmw_or_i64_monotonic:
; LA64: # %bb.0:
-; LA64-NEXT: amor_db.d $a2, $a1, $a0
+; LA64-NEXT: amor.d $a2, $a1, $a0
; LA64-NEXT: move $a0, $a2
; LA64-NEXT: ret
%1 = atomicrmw or ptr %a, i64 %b monotonic
@@ -4815,7 +4815,7 @@ define i8 @atomicrmw_xor_i8_monotonic(ptr %a, i8 %b) nounwind {
; LA64-NEXT: bstrins.d $a0, $zero, 1, 0
; LA64-NEXT: andi $a1, $a1, 255
; LA64-NEXT: sll.w $a1, $a1, $a2
-; LA64-NEXT: amxor_db.w $a3, $a1, $a0
+; LA64-NEXT: amxor.w $a3, $a1, $a0
; LA64-NEXT: srl.w $a0, $a3, $a2
; LA64-NEXT: ret
%1 = atomicrmw xor ptr %a, i8 %b monotonic
@@ -4844,7 +4844,7 @@ define i16 @atomicrmw_xor_i16_monotonic(ptr %a, i16 %b) nounwind {
; LA64-NEXT: bstrins.d $a0, $zero, 1, 0
; LA64-NEXT: bstrpick.d $a1, $a1, 15, 0
; LA64-NEXT: sll.w $a1, $a1, $a2
-; LA64-NEXT: amxor_db.w $a3, $a1, $a0
+; LA64-NEXT: amxor.w $a3, $a1, $a0
; LA64-NEXT: srl.w $a0, $a3, $a2
; LA64-NEXT: ret
%1 = atomicrmw xor ptr %a, i16 %b monotonic
@@ -4865,7 +4865,7 @@ define i32 @atomicrmw_xor_i32_monotonic(ptr %a, i32 %b) nounwind {
;
; LA64-LABEL: atomicrmw_xor_i32_monotonic:
; LA64: # %bb.0:
-; LA64-NEXT: amxor_db.w $a2, $a1, $a0
+; LA64-NEXT: amxor.w $a2, $a1, $a0
; LA64-NEXT: move $a0, $a2
; LA64-NEXT: ret
%1 = atomicrmw xor ptr %a, i32 %b monotonic
@@ -4885,7 +4885,7 @@ define i64 @atomicrmw_xor_i64_monotonic(ptr %a, i64 %b) nounwind {
;
; LA64-LABEL: atomicrmw_xor_i64_monotonic:
; LA64: # %bb.0:
-; LA64-NEXT: amxor_db.d $a2, $a1, $a0
+; LA64-NEXT: amxor.d $a2, $a1, $a0
; LA64-NEXT: move $a0, $a2
; LA64-NEXT: ret
%1 = atomicrmw xor ptr %a, i64 %b monotonic
|
|
Need some suggestions to simplify this td file. |
tangaac
changed the title
tangaac
changed the title
llvm-project/llvm/lib/Target/AArch64/AArch64InstrAtomics.td Lines 528 to 536 in 2f0b4f4 llvm-project/llvm/lib/Target/AArch64/AArch64InstrFormats.td Lines 11919 to 11951 in 2f0b4f4 |
tangaac
force-pushed
the
fix/atomic-insts
branch
from
8343f08 to
329cac3
Compare
heiher
reviewed
Oct 18, 2024
tangaac
force-pushed
the
fix/atomic-insts
branch
from
329cac3 to
a9452c1
Compare
|
About the title and commit message: I think this is not a "fix" but a fine-grained optimization or tuning. |
|
About testcase: gcc: clang: But this can be done in a separate PR. |
tangaac
changed the title
tangaac
changed the title
|
@tangaac Congratulations on having your first Pull Request (PR) merged into the LLVM Project! Your changes will be combined with recent changes from other authors, then tested by our build bots. If there is a problem with a build, you may receive a report in an email or a comment on this PR. Please check whether problems have been caused by your change specifically, as the builds can include changes from many authors. It is not uncommon for your change to be included in a build that fails due to someone else's changes, or infrastructure issues. How to do this, and the rest of the post-merge process, is covered in detail here. If your change does cause a problem, it may be reverted, or you can revert it yourself. This is a normal part of LLVM development. You can fix your changes and open a new PR to merge them again. If you don't get any reports, no action is required from you. Your changes are working as expected, well done! |
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.
Don't use "_db" version AM instructions for LoongArch atomic memory operations with monotonic semantics.