I have a question regarding the DBG_VALUE instruction. IIUC the DBG_VALUE appears in two different forms: 1. DBG_VALUE( %Reg, %Reg, metadata) 2. DBG_VALUE(%Reg, offset, metadata) In the first version, the second operand is always a %noreg in the form of: DBG_VALUE(%Reg, %noreg, metadata). Is it possible to have a different version of DBG_VALUE that only has 2 operands which will have only %Reg and metadata, because the second operand appears to be redundant? Thanks, -Aditya
On Thu, Sep 12, 2013 at 2:36 PM, Aditya Kumar <hiraditya at codeaurora.org>wrote:> I have a question regarding the DBG_VALUE instruction. > > IIUC the DBG_VALUE appears in two different forms: > > 1. DBG_VALUE( %Reg, %Reg, metadata) > 2. DBG_VALUE(%Reg, offset, metadata) > > In the first version, the second operand is always a %noreg in the form of: > DBG_VALUE(%Reg, %noreg, metadata). > > Is it possible to have a different version of DBG_VALUE that only has 2 > operands which will have only %Reg and metadata, because the second operand > appears to be redundant?Possible, but it simplifies the code to only have one representation for the schema rather than two. -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20130912/26dcb33e/attachment.html>
Possibly Parallel Threads
- [RFC] A value-tracking LiveDebugValues implementation
- [Debuginfo] Changing llvm.dbg.value and DBG_VALUE to support multiple location operands
- GSoC 2020 Project "Improve MegreFunctions to incorporate MergeSimilarFunctions patches and ThinLTO Support"
- [LLVMdev] Improving the quality of debug locations / DbgValueHistoryCalculator
- [LLVMdev] liveness assertion problem in llc