<html>
    <head>
      <base href="https://bugs.llvm.org/">
    </head>
    <body><table border="1" cellspacing="0" cellpadding="8">
        <tr>
          <th>Bug ID</th>
          <td><a class="bz_bug_link 
          bz_status_NEW "
   title="NEW - [DI] LLVM loses track of induction variables at -O1 and beyond"
   href="https://bugs.llvm.org/show_bug.cgi?id=35530">35530</a>
          </td>
        </tr>

        <tr>
          <th>Summary</th>
          <td>[DI] LLVM loses track of induction variables at -O1 and beyond
          </td>
        </tr>

        <tr>
          <th>Product</th>
          <td>libraries
          </td>
        </tr>

        <tr>
          <th>Version</th>
          <td>trunk
          </td>
        </tr>

        <tr>
          <th>Hardware</th>
          <td>PC
          </td>
        </tr>

        <tr>
          <th>OS</th>
          <td>All
          </td>
        </tr>

        <tr>
          <th>Status</th>
          <td>NEW
          </td>
        </tr>

        <tr>
          <th>Severity</th>
          <td>enhancement
          </td>
        </tr>

        <tr>
          <th>Priority</th>
          <td>P
          </td>
        </tr>

        <tr>
          <th>Component</th>
          <td>DebugInfo
          </td>
        </tr>

        <tr>
          <th>Assignee</th>
          <td>unassignedbugs@nondot.org
          </td>
        </tr>

        <tr>
          <th>Reporter</th>
          <td>davide@freebsd.org
          </td>
        </tr>

        <tr>
          <th>CC</th>
          <td>aprantl@apple.com, friss@apple.com, jdevlieghere@apple.com, jingham@apple.com, jmolenda@apple.com, llvm-bugs@lists.llvm.org, vsk@apple.com
          </td>
        </tr></table>
      <p>
        <div>
        <pre>Testcase:

$ cat missing.cpp
#include <atomic>
#include <limits.h>
#include <stdint.h>

int
main(void)
{
        std::atomic<uint64_t> limit(0);
        std::atomic<uint64_t> vr_ac(0);
        uint64_t vr_index;

        limit.store(UINT64_MAX, std::memory_order_seq_cst);
        for (vr_index = 0; vr_index < limit; vr_index++)
                vr_ac.store(vr_index, std::memory_order_seq_cst);
        return vr_ac;
}


The quality of DI generates varies a bit depending on the optimization level.

-O0 -g:
(lldb) p vr_ac
(std::__1::atomic<unsigned long long>) $0 = 519251459
(lldb) p vr_index
(uint64_t) $1 = 519251460
(lldb) p limit
(std::__1::atomic<unsigned long long>) $2 = 18446744073709551615

-O1 -g:
(lldb) p vr_index
(uint64_t) $1 = 12232201402
(lldb) p vr_ac
error: Couldn't materialize: couldn't get the value of variable vr_ac: no
location, value may have been optimized out
error: errored out in DoExecute, couldn't PrepareToExecuteJITExpression
(lldb) p limit
error: Couldn't materialize: couldn't get the value of variable limit: no
location, value may have been optimized out
error: errored out in DoExecute, couldn't PrepareToExecuteJITExpression

-O2 -g (also -O3 -g):
(lldb) p vr_index
(uint64_t) $0 = 0
(lldb) p limit
error: Couldn't materialize: couldn't get the value of variable limit: no
location, value may have been optimized out
error: errored out in DoExecute, couldn't PrepareToExecuteJITExpression
(lldb) p vr_ac
error: Couldn't materialize: couldn't get the value of variable vr_ac: no
location, value may have been optimized out

At -O1, it's not able to retrieve the values for vr_ac and limit (although
they're atomic and therefore not optimized out). At -O2, it's slightly worse as
the value of vr_index is wrong (always zero, no matter when I break).

I'm actually not on top-of-tree so the situation might have improved a bit, but
writing this down so I won't forget (and I'll test with a new compiler as soon
as I can).</pre>
        </div>
      </p>


      <hr>
      <span>You are receiving this mail because:</span>

      <ul>
          <li>You are on the CC list for the bug.</li>
      </ul>
    </body>
</html>