[PATCH] Enable invoking the patchpoint intrinsic

[Kevin Modzelewski]

Ok I will update the patch, but I'm not sure how you mean to not thread
through the ImmutableCallSite, since I thought InvokeInst is not a subclass
of CallInst?
On Oct 15, 2014 10:49 AM, "Juergen Ributzka" <juergen at apple.com> wrote:

> I still think not threading through ImmutableCallSite would be nicer and
> reduce unnecessary changes. Otherwise LGTM.
>
> ================
> Comment at: lib/CodeGen/SelectionDAG/SelectionDAGBuilder.h:637-643
> @@ -636,7 +636,9 @@
>
> -  std::pair<SDValue, SDValue> LowerCallOperands(const CallInst &CI,
> -                                                unsigned ArgIdx,
> -                                                unsigned NumArgs,
> -                                                SDValue Callee,
> -                                                bool useVoidTy = false);
> +  std::pair<SDValue, SDValue> LowerCallOperands(
> +          ImmutableCallSite CS,
> +          unsigned ArgIdx,
> +          unsigned NumArgs,
> +          SDValue Callee,
> +          bool useVoidTy = false,
> +          MachineBasicBlock *LandingPad = nullptr);
>
> ----------------
> While you are already at it, could you please clean up the sins I
> committed and update it to the current coding standard (camelCase for
> functions and CamelCase for variables)?
>
> ================
> Comment at: lib/CodeGen/SelectionDAG/SelectionDAGBuilder.h:783
> @@ -780,3 +782,3 @@
>    void visitStackmap(const CallInst &I);
> -  void visitPatchpoint(const CallInst &I);
> +  void visitPatchpoint(ImmutableCallSite CS, MachineBasicBlock
> *LandingPad);
>
> ----------------
> I would add a default value for LandingPad = nullptr and not switch to
> ImmutableCallSite for the first argument.
>
> http://reviews.llvm.org/D5634
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20141015/07644efa/attachment.html>