-
Notifications
You must be signed in to change notification settings - Fork 14.3k
[CoroSplit][DebugInfo] Adjust heuristic for moving DIScope of funclets #108611
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
[CoroSplit][DebugInfo] Adjust heuristic for moving DIScope of funclets #108611
Conversation
CoroSplit has a heuristic where the scope line for funclets is adjusted to match the line of the suspend intrinsic that caused the split. This is useful as it avoids a jump on the line table from the original function declaration to the line where the split happens. However, very often using the line of the split is not ideal: if we can avoid it, we should not have a line entry for the split location, as this would cause breakpoints by line to match against two functions: the funclet before and the funclet after the split. This patch adjusts the heuristics to look for the first instruction with a non-zero line number after the split. In other words, this patch makes breakpoints on `await foo()` lines behave much more like a regular function call.
@llvm/pr-subscribers-llvm-transforms @llvm/pr-subscribers-coroutines Author: Felipe de Azevedo Piovezan (felipepiovezan) ChangesCoroSplit has a heuristic where the scope line for funclets is adjusted to match the line of the suspend intrinsic that caused the split. This is useful as it avoids a jump on the line table from the original function declaration to the line where the split happens. However, very often using the line of the split is not ideal: if we can avoid it, we should not have a line entry for the split location, as this would cause breakpoints by line to match against two functions: the funclet before and the funclet after the split. This patch adjusts the heuristics to look for the first instruction with a non-zero line number after the split. In other words, this patch makes breakpoints on Full diff: https://github.com/llvm/llvm-project/pull/108611.diff 2 Files Affected:
diff --git a/llvm/lib/Transforms/Coroutines/CoroSplit.cpp b/llvm/lib/Transforms/Coroutines/CoroSplit.cpp
index 8ea460badaad5d..2f43c82b1a1d83 100644
--- a/llvm/lib/Transforms/Coroutines/CoroSplit.cpp
+++ b/llvm/lib/Transforms/Coroutines/CoroSplit.cpp
@@ -887,6 +887,45 @@ Value *CoroCloner::deriveNewFramePointer() {
llvm_unreachable("bad ABI");
}
+// Adjust the scope line of the funclet to the first line number after the
+// suspend point. This avoids a jump in the line table from the function
+// declaration (where prologue instructions are attributed to) to the suspend
+// point.
+// Only adjust the scope line when the files are the same.
+// If no candidate line number is found, fallback to the line of ActiveSuspend.
+static void updateScopeLine(Instruction *ActiveSuspend,
+ DISubprogram &SPToUpdate) {
+ if (!ActiveSuspend)
+ return;
+
+ auto *Successor = ActiveSuspend->getNextNonDebugInstruction();
+ // Corosplit splits the BB around ActiveSuspend, so the meaningful
+ // instructions are not in the same BB.
+ if (auto *Branch = dyn_cast_or_null<BranchInst>(Successor);
+ Branch && Branch->isUnconditional())
+ Successor = Branch->getSuccessor(0)->getFirstNonPHIOrDbg();
+
+ // Find the first successor of ActiveSuspend with a non-zero line location.
+ // If that matches the file of ActiveSuspend, use it.
+ for (; Successor; Successor = Successor->getNextNonDebugInstruction()) {
+ auto DL = Successor->getDebugLoc();
+ if (!DL || DL.getLine() == 0)
+ continue;
+
+ if (SPToUpdate.getFile() == DL->getFile()) {
+ SPToUpdate.setScopeLine(DL.getLine());
+ return;
+ }
+
+ break;
+ }
+
+ // If the search above failed, fallback to the location of ActiveSuspend.
+ if (auto DL = ActiveSuspend->getDebugLoc())
+ if (SPToUpdate.getFile() == DL->getFile())
+ SPToUpdate.setScopeLine(DL->getLine());
+}
+
static void addFramePointerAttrs(AttributeList &Attrs, LLVMContext &Context,
unsigned ParamIndex, uint64_t Size,
Align Alignment, bool NoAlias) {
@@ -955,18 +994,10 @@ void CoroCloner::create() {
auto &Context = NewF->getContext();
- // For async functions / continuations, adjust the scope line of the
- // clone to the line number of the suspend point. However, only
- // adjust the scope line when the files are the same. This ensures
- // line number and file name belong together. The scope line is
- // associated with all pre-prologue instructions. This avoids a jump
- // in the linetable from the function declaration to the suspend point.
if (DISubprogram *SP = NewF->getSubprogram()) {
assert(SP != OrigF.getSubprogram() && SP->isDistinct());
- if (ActiveSuspend)
- if (auto DL = ActiveSuspend->getDebugLoc())
- if (SP->getFile() == DL->getFile())
- SP->setScopeLine(DL->getLine());
+ updateScopeLine(ActiveSuspend, *SP);
+
// Update the linkage name to reflect the modified symbol name. It
// is necessary to update the linkage name in Swift, since the
// mangling changes for resume functions. It might also be the
diff --git a/llvm/test/Transforms/Coroutines/swift-async-dbg.ll b/llvm/test/Transforms/Coroutines/swift-async-dbg.ll
index 2cd0f1c42b730b..c0b7247ff49ada 100644
--- a/llvm/test/Transforms/Coroutines/swift-async-dbg.ll
+++ b/llvm/test/Transforms/Coroutines/swift-async-dbg.ll
@@ -37,9 +37,9 @@ define swifttailcc void @coroutineA(ptr swiftasync %arg) !dbg !48 {
%i7 = call ptr @llvm.coro.async.resume(), !dbg !54
%i10 = call { ptr } (i32, ptr, ptr, ...) @llvm.coro.suspend.async.sl_p0s(i32 0, ptr %i7, ptr nonnull @__swift_async_resume_get_context, ptr nonnull @coroutineA.1, ptr %i7, i64 0, i64 0, ptr %arg), !dbg !54
- %i11 = extractvalue { ptr } %i10, 0, !dbg !54
- %i12 = call ptr @__swift_async_resume_get_context(ptr %i11), !dbg !54
- call void @dont_optimize(ptr %var_with_dbg_value, ptr %var_with_dbg_declare)
+ %i11 = extractvalue { ptr } %i10, 0, !dbg !55
+ %i12 = call ptr @__swift_async_resume_get_context(ptr %i11), !dbg !55
+ call void @dont_optimize(ptr %var_with_dbg_value, ptr %var_with_dbg_declare), !dbg !100
call void @llvm.dbg.value(metadata ptr %var_with_dbg_value, metadata !50, metadata !DIExpression(DW_OP_deref)), !dbg !54
%i17 = load i32, ptr getelementptr inbounds (<{i32, i32}>, ptr @coroutineBTu, i64 0, i32 1), align 8, !dbg !54
call void @llvm.dbg.value(metadata !DIArgList(ptr %var_with_dbg_value, i32 %i17), metadata !501, metadata !DIExpression(DW_OP_LLVM_arg, 0, DW_OP_LLVM_arg, 1, DW_OP_plus, DW_OP_deref)), !dbg !54
@@ -47,7 +47,7 @@ define swifttailcc void @coroutineA(ptr swiftasync %arg) !dbg !48 {
%i19 = call swiftcc ptr @swift_task_alloc(i64 %i18), !dbg !54
; CHECK-NOT: define
; CHECK-LABEL: define {{.*}} @coroutineATY0_(
-; CHECK-SAME: ptr swiftasync %[[frame_ptr:.*]])
+; CHECK-SAME: ptr swiftasync %[[frame_ptr:.*]]) !dbg ![[ATY0:[0-9]*]]
; CHECK: #dbg_declare(ptr %[[frame_ptr]], {{.*}} !DIExpression(
; CHECK-SAME: DW_OP_LLVM_entry_value, 1, DW_OP_plus_uconst, 24)
; CHECK: #dbg_value(ptr %[[frame_ptr]], {{.*}} !DIExpression(
@@ -88,6 +88,8 @@ define swifttailcc void @coroutineA(ptr swiftasync %arg) !dbg !48 {
; CHECK-SAME: DW_OP_LLVM_entry_value, 1, DW_OP_plus_uconst, 24)
}
+; CHECK: ![[ATY0]] = {{.*}}DISubprogram(linkageName: "coroutineATY0_", {{.*}} scopeLine: 42
+
; Everything from here on is just support code for the coroutines.
@coroutineBTu = global <{i32, i32}> <{ i32 trunc (i64 sub (i64 ptrtoint (ptr @"coroutineB" to i64), i64 ptrtoint (ptr @"coroutineBTu" to i64)) to i32), i32 16 }>, align 8
@@ -172,5 +174,7 @@ declare { ptr } @llvm.coro.suspend.async.sl_p0s(i32, ptr, ptr, ...)
!71 = !DILocation(line: 0, scope: !70)
!73 = !DILocation(line: 0, scope: !72)
!54 = !DILocation(line: 6, scope: !48)
+!55 = !DILocation(line: 0, scope: !48)
!42 = !DILocation(line: 3, scope: !37)
!47 = !DILocation(line: 0, scope: !44)
+!100 = !DILocation(line: 42, scope: !48)
|
CoroSplit has a heuristic where the scope line for funclets is adjusted to match the line of the suspend intrinsic that caused the split. This is useful as it avoids a jump on the line table from the original function declaration to the line where the split happens.
However, very often using the line of the split is not ideal: if we can avoid it, we should not have a line entry for the split location, as this would cause breakpoints by line to match against two functions: the funclet before and the funclet after the split.
This patch adjusts the heuristics to look for the first instruction with a non-zero line number after the split. In other words, this patch makes breakpoints on
await foo()
lines behave much more like a regular function call.