[llvm-dev] [LLVM RFC] Add llvm.typeid.for intrinsic
Wang Nan via llvm-dev
llvm-dev at lists.llvm.org
Fri Aug 14 03:05:12 PDT 2015
This is for BPF output. BPF program output bytes to perf through a
tracepoint. For decoding such data, we need a way to describe the format
of the buffer. This patch is a try which gives each variable a unique
number by introducing a new intrinsic 'llvm.typeid.for'.
At the bottom is an example of using that intrinsic and the result
of
$ clang -target bpf -O2 -c -S ./test_typeid.c
There is a limitation of the newly introduced intrinsic that, I can't
find a way to make the intrinsic to accept all types without name
mangling. Therefore, we have to define different intrinsics for
different type. See the example below, by using macro trick, we define
llvm.typeid.for.p0struct.mystr and llvm.typeid.for.p0struct.mystr2, and
also the different output functions.
Another problem is that I'm still unable to find a way to insert dwarf
information in this stage. After clang, debug information are already
isolated, and debug information entries are linked together. Adjusting
debug information requires me to create new metadata and new debug info
entries, link them properly then insert into correct place. Which is
possible, but makes code ugly.
Because of the above two problems, I decided to try clang builtin
again. I think that should be the last try. If still not work, then
I'd like to stop working on it until I have any better idea (BCC
rewriter should be a considerable solution). Let patch series
'Make eBPF programs output data to perf' be merged into upstream
without the 'typeid' change. Before the decoding problem solved, we
have to let user decode the BPF output themself manually or use
perf script or babeltrace script.
Thank you.
----------------- EXAMPLE -----------------
extern void output(int id, void *ptr, int size);
#define OUTPUT_STR(name) \
struct name {
#define OUTPUT_STR_END(name) \
}; \
unsigned long long __get_typeid_##name(struct name *) asm ("llvm.typeid.for.p0struct."#name); \
static inline void output_##name(struct name *str) \
{\
output(__get_typeid_##name(str), str, sizeof(struct name));\
};\
static struct name __g_##name;
OUTPUT_STR(mystr)
int x;
int y;
int z;
OUTPUT_STR_END(mystr);
OUTPUT_STR(mystr2)
int x;
int y;
OUTPUT_STR_END(mystr2);
--------------- RESULT -------------
int func(void)
{
int x = 123;
struct mystr myvar;
struct mystr2 myvar2;
output_mystr(&myvar);
output_mystr2(&myvar2);
output_mystr(&myvar);
return 0;
}
int func2(void)
{
int x = 123;
struct mystr myvar;
struct mystr2 myvar2;
output_mystr2(&myvar2);
output_mystr(&myvar);
output_mystr2(&myvar2);
return 0;
}
.text
.globl func
.align 8
func: # @func
# BB#0: # %entry
mov r6, r10
addi r6, -16
mov r1, 1
mov r2, r6
mov r3, 12
call output
mov r2, r10
addi r2, -24
mov r1, 2
mov r3, 8
call output
mov r1, 1
mov r2, r6
mov r3, 12
call output
mov r0, 0
ret
.globl func2
.align 8
func2: # @func2
# BB#0: # %entry
mov r6, r10
addi r6, -24
mov r1, 2
mov r2, r6
mov r3, 8
call output
mov r2, r10
addi r2, -16
mov r1, 1
mov r3, 12
call output
mov r1, 2
mov r2, r6
mov r3, 8
call output
mov r0, 0
ret
Signed-off-by: Wang Nan <wangnan0 at huawei.com>
---
include/llvm/IR/Intrinsics.td | 1 +
lib/CodeGen/SelectionDAG/SelectionDAGBuilder.cpp | 27 ++++++++++++++++++++++++
lib/CodeGen/SelectionDAG/SelectionDAGBuilder.h | 1 +
3 files changed, 29 insertions(+)
diff --git a/include/llvm/IR/Intrinsics.td b/include/llvm/IR/Intrinsics.td
index 83cfebe..8ebeb24 100644
--- a/include/llvm/IR/Intrinsics.td
+++ b/include/llvm/IR/Intrinsics.td
@@ -640,6 +640,7 @@ def int_masked_scatter: Intrinsic<[],
def int_bitset_test : Intrinsic<[llvm_i1_ty], [llvm_ptr_ty, llvm_metadata_ty],
[IntrNoMem]>;
+def int_typeid_for : Intrinsic<[llvm_i64_ty], [llvm_any_ty], [IntrNoMem]>;
//===----------------------------------------------------------------------===//
// Target-specific intrinsics
//===----------------------------------------------------------------------===//
diff --git a/lib/CodeGen/SelectionDAG/SelectionDAGBuilder.cpp b/lib/CodeGen/SelectionDAG/SelectionDAGBuilder.cpp
index ce4912f..ff453cd 100644
--- a/lib/CodeGen/SelectionDAG/SelectionDAGBuilder.cpp
+++ b/lib/CodeGen/SelectionDAG/SelectionDAGBuilder.cpp
@@ -5077,6 +5077,10 @@ SelectionDAGBuilder::visitIntrinsicCall(const CallInst &I, unsigned Intrinsic) {
setValue(&I, N);
return nullptr;
}
+ case Intrinsic::typeid_for: {
+ visitTypeidfor(I);
+ return nullptr;
+ }
}
}
@@ -6769,6 +6773,29 @@ void SelectionDAGBuilder::visitPatchpoint(ImmutableCallSite CS,
FuncInfo.MF->getFrameInfo()->setHasPatchPoint();
}
+void SelectionDAGBuilder::visitTypeidfor(const CallInst &CI) {
+ SDValue Res;
+ static std::vector<const StructType *> StructTypes;
+ int ID = -1;
+ Value *PtrArg = CI.getArgOperand(0);
+ PointerType *PTy = cast<PointerType>(PtrArg->getType());
+ if (PTy) {
+ StructType *STy = cast<StructType>(PTy->getElementType());
+ if (STy) {
+ for (unsigned i = 0, N = StructTypes.size(); i != N; ++i)
+ if (StructTypes[i] == STy)
+ ID = i + 1;
+ if (ID == -1) {
+ StructTypes.push_back(STy);
+ ID = StructTypes.size();
+ }
+ }
+ }
+
+ Res = DAG.getConstant(ID, getCurSDLoc(), MVT::i32);
+ setValue(&CI, Res);
+}
+
/// Returns an AttributeSet representing the attributes applied to the return
/// value of the given call.
static AttributeSet getReturnAttrs(TargetLowering::CallLoweringInfo &CLI) {
diff --git a/lib/CodeGen/SelectionDAG/SelectionDAGBuilder.h b/lib/CodeGen/SelectionDAG/SelectionDAGBuilder.h
index f71190d..f037689 100644
--- a/lib/CodeGen/SelectionDAG/SelectionDAGBuilder.h
+++ b/lib/CodeGen/SelectionDAG/SelectionDAGBuilder.h
@@ -855,6 +855,7 @@ private:
void visitStatepoint(const CallInst &I);
void visitGCRelocate(const CallInst &I);
void visitGCResult(const CallInst &I);
+ void visitTypeidfor(const CallInst &I);
void visitUserOp1(const Instruction &I) {
llvm_unreachable("UserOp1 should not exist at instruction selection time!");
--
1.8.3.4
More information about the llvm-dev
mailing list