[PATCH] D37311: [DebugInfo] Lower dbg.declare to DBG_VALUE with DW_OP_deref

Adrian Prantl via llvm-commits llvm-commits at lists.llvm.org
Tue Sep 5 09:46:08 PDT 2017


> On Sep 4, 2017, at 12:10 PM, David Blaikie <dblaikie at gmail.com> wrote:
> 
> 
> 
> On Thu, Aug 31, 2017 at 2:18 PM Reid Kleckner via Phabricator via llvm-commits <llvm-commits at lists.llvm.org <mailto:llvm-commits at lists.llvm.org>> wrote:
> rnk added a comment.
> 
> Running debuginfo-tests was enlightening. My understanding of these opcodes is a lot worse than I thought it was. :( They are really impossible to understand without staring at the LLDB or GDB expression evaluation code. I'll have to think about this some more.
> 
> Happy to chat about them in person if it helps. There was also some historic misunderstanding of some of the opcodes that ended up baked into LLDB and LLVM (they both agreed on something but it wasn't the thing the DWARF Spec said, then we changed it somehow (this was related to DW_OP_bit_piece))

Yes, I fixed this in both LLVM and LLDB. The debuginfo-tests predate DW_OP_bit_piece support in LLVM significantly, so the locations in there are a lot simpler. Since the tests are run with both GDB and LLDB, they should be a little better guarded against symmetric errors like that, although I am not sure whether there is a current bot running debuginfo-tests with GDB at the moment.

-- adrian

>  
> 
> 
> https://reviews.llvm.org/D37311 <https://reviews.llvm.org/D37311>
> 
> 
> 
> _______________________________________________
> llvm-commits mailing list
> llvm-commits at lists.llvm.org <mailto:llvm-commits at lists.llvm.org>
> http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-commits <http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-commits>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20170905/ac4266e9/attachment.html>


More information about the llvm-commits mailing list