<table border="1" cellspacing="0" cellpadding="8">
    <tr>
        <th>Issue</th>
        <td>
            <a href=https://github.com/llvm/llvm-project/issues/133262>133262</a>
        </td>
    </tr>

    <tr>
        <th>Summary</th>
        <td>
            [LLVM] Suboptimal code for rounding shift right by constant
        </td>
    </tr>

    <tr>
      <th>Labels</th>
      <td>
            new issue
      </td>
    </tr>

    <tr>
      <th>Assignees</th>
      <td>
      </td>
    </tr>

    <tr>
      <th>Reporter</th>
      <td>
          johnplatts
      </td>
    </tr>
</table>

<pre>
    LLVM generates extra code for `((x & round_incr) + ((x ^ round_incr) >> 1)) >> 4` than for `(int32_t)(((int64_t)x + (int64_t)round_incr) >> 5)` where `x` is an int32_t value and `round_incr` is equivalent to `15 + (int32_t)((x & 32) != 0)` according to a snippet over at https://alive2.llvm.org/ce/z/Ar8vv9.


</pre>
<img width="1" height="1" alt="" src="http://email.email.llvm.org/o/eJxsUkFvpDwM_TXmYnUUHGDgwGGqllO_0yfttQoQIFUmYRNDp_vrVzCz7XS1EhKyn_2eeTwVoxmd1jXkj5A_JWrhyYf6zU9utoo5Jq3vP-qXlx__4aidDop1RH3hoLDzvcbBB4RCAJVA5QWBCgx-cf2rcV0AqhDoET_R_PlvVD6DfMYUqLorMygE8qTcHb1xLOmV98Hy-hjHRba3Ln9kvlr_1Mm39ULg-6SD3ogvW2UiKoc3AVyVXTQq12_4Hct1UP9czKqsdozst4k0v9P-duHVDUlXF1KQTyhu-qrrfOiNGzcShdGZedaMftUBFePEPEeQJ6AGqFHWrJoO1q7ngw8jUNNpoOYXUHMK5bpWBxAnEKekr2VfyUoluk6PmRRlUckymeq-y2UhRCuo7Is0HdJSHcuBKhqORZVmMjE1CcqFpGOaZWVeHsRQ9FWlj8dBaMoGBZnQZ2Xs5xGJiXHRdSolFZRY1Wob9wwROf2OOwpEW6RCvS09tMsYIRPWRI5fNGzY7uHbEgb5E_6_tH5mc1b2K1_7T9jMipMZGIMZJ8b2AzvvIivHyRJs_d2z0fC0tIfOn4GaTe32epiDf9MdAzX7jRGouX3EWtPvAAAA___a6etp">