<div dir="ltr"><div>Thank you Erik!</div><div><br></div><div>I expected that. I guess I better rebuild it for every bisect I do.<br></div><div><br></div><div>Cheers, Thomas<br></div><div><br></div><div><br></div></div><br><div class="gmail_quote"><div dir="ltr" class="gmail_attr">On Fri, Mar 31, 2023 at 3:21 PM <<a href="mailto:erik.joelsson@oracle.com">erik.joelsson@oracle.com</a>> wrote:<br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">Hello Thomas,<br>
<br>
On 3/31/23 05:43, Thomas Stüfe wrote:<br>
> Hi,<br>
><br>
> quick question, I'm doing arm crossbuilds, trying to find an errornous <br>
> patch with bisecting. With crossbuild, I am specifying build-jdk.<br>
><br>
> How current does the build-jdk have to be? If the build jdk does not <br>
> fit the source I am trying to build, would that be immediately obvious <br>
> or would I get strange errors? I am gauging whether to rebuild the <br>
> build jdk for every bisect point.<br>
><br>
As far as I know, we expect the BUILD_JDK to be an exact match, but in <br>
practice it probably works with something a bit older most of the time. <br>
It depends on what differences there are between the current source and <br>
the BUILD_JDK. I haven't tried using an out of date BUILD_JDK but I <br>
would imagine the errors to be strange and unpredictable. In your case, <br>
if you want to avoid strange surprises, letting the BUILD_JDK be rebuilt <br>
automatically is preferred, but if the bisect range is small, maybe it's <br>
worth chancing it to save build time.<br>
<br>
/Erik<br>
<br>
<br>
</blockquote></div>