[llvm-dev] Removed a call to EmitXRayTable() from ARMAsmPrinter

Renato Golin via llvm-dev llvm-dev at lists.llvm.org
Mon Jan 9 12:50:39 PST 2017

On 9 January 2017 at 20:47, Serge Rogatch via llvm-dev
<llvm-dev at lists.llvm.org> wrote:
>> Without this call, xray_instr_map gets empty in the executable, so that
>> XRay doesn't patch anything. While implementing the tail call handling, I am
>> reverting this change (I mean, just returning the call to emitXRayTable()
>> where it was), or are there reasons I shouldn't?

Any reason why no tests were harmed in the process? Maybe we're not
testing as thorough as we thought?


More information about the llvm-dev mailing list