<div dir="ltr"><div dir="ltr"><div dir="ltr"><div dir="ltr"><div class="gmail_extra"><br></div><div class="gmail_extra">Jan Stancek <span dir="ltr"><<a href="mailto:jstancek@redhat.com" target="_blank">jstancek@redhat.com</a>></span> wrote:<br><div class="gmail_quote"><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"><br>
                /* wait until child1 is ready on node1, then migrate and<br>
                 * signal to check current node */<br>
-               if (read(child1_ready[0], &tmp, 1) != 1)<br>
-                       tst_brkm(TBROK | TERRNO, NULL, "read #2 failed");<br>
+               TST_CHECKPOINT_WAIT(1);<br></blockquote><div><br></div><div><div style="font-size:small" class="gmail_default">Just curious why not start from 0? TST_CHECKPOINT_WAIT(0)<br></div><div style="font-size:small" class="gmail_default"><br></div></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">
-       tst_record_childstatus(NULL, child2);<br>
-       tst_record_childstatus(NULL, child1);<br>
+       SAFE_WAITPID(child1, &status, 0);<br>
+       SAFE_WAITPID(child2, &status, 0);<br></blockquote><div><br></div><div style="font-size:small" class="gmail_default">Maybe we could set &wstatus as a NULL? we don't inspect the status information after its children terminates so why we store it into &status?<br></div><div style="font-size:small" class="gmail_default"></div></div><div style="font-size:small" class="gmail_default">Beside that two questions, patch set looks good to me.</div><br>-- <br><div class="gmail_signature"><div dir="ltr"><div>Regards,<br></div><div>Li Wang<br></div></div></div>
</div></div></div></div></div>