Re: port-amd64/57153: 10.0 Beta kernel panics immediately after detecting CPUs

["ali via gnats" <gnats-admin%NetBSD.org@localhost>]

The following reply was made to PR port-amd64/57153; it has been noted by GNATS.

From: ali <ali.lindgren%gmail.com@localhost>
To: gnats-bugs%gnats.netbsd.org@localhost
Cc: 
Subject: Re: port-amd64/57153: 10.0 Beta kernel panics immediately after
 detecting CPUs
Date: Mon, 13 Oct 2025 19:25:54 +0200

 --000000000000f2509906410d8f34
 Content-Type: text/plain; charset="UTF-8"
 
 Sorry for the late feedback; I'm not currently on this particular list, so
 I'm not in Cc, so excuse me if I respond with some manual quoting from the
 gnats page here...
 
 I'll try the original kernel and see if any warning is printed before the
 crash, and then set up a build tree and try to add some more prints, but
 it's going to take a while before I have time to do any serious debugging
 I'm afraid. The original server died, so I've switched to an other
 identical one in the same rack. It hung without any traces of why after 72h
 of "memtester 20G", but has now been running fine (albeit without any load
 too) for five days, using the patched kernel.
 
 Best regards,
 Anders Ali Lindgren
 
 Christos wrote:
 > Something is wrong finding the numa region for that page> Don't you see =
 > an error from that printf:
 >
 >         if (!warned) {
 >                 printf("uvm_page_numa_lookup: failed, first pg=3D%p =
 > pa=3D%#"
 >                     PRIxPADDR "\n", pg, VM_PAGE_TO_PHYS(pg));
 >                 warned =3D true;
 >         }
 >
 > If not, perhaps print some information about the address that failed =
 > when it fails...
 >
 > christos=
 
 --000000000000f2509906410d8f34
 Content-Type: text/html; charset="UTF-8"
 Content-Transfer-Encoding: quoted-printable
 
 <div dir=3D"ltr"><div>Sorry for the late feedback; I&#39;m not currently on=
  this particular list, so I&#39;m not in Cc, so excuse me if I respond with=
  some manual quoting from the gnats page here...</div><div><br></div><div>I=
 &#39;ll try the original kernel and see if any warning is printed before th=
 e crash, and then set up a build tree and try to add some more prints, but =
 it&#39;s going to take a while before I have time to do any serious debuggi=
 ng I&#39;m afraid. The original server died, so I&#39;ve switched to an oth=
 er identical one in the same rack. It hung without any traces of why after =
 72h of &quot;memtester 20G&quot;, but has now been running fine (albeit wit=
 hout any load too) for five days, using the patched kernel.</div><div></div=
 ><div><br></div><div>Best regards,</div><div>Anders Ali Lindgren</div><div>=
 <br></div><div>Christos wrote:</div>&gt; Something is wrong finding the num=
 a region for that page&gt; Don&#39;t you see =3D<br>&gt; an error from that=
  printf:<br>&gt;<br>&gt;=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0if (!warned) {<br=
 >&gt;=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0printf(&=
 quot;uvm_page_numa_lookup: failed, first pg=3D3D%p =3D<br>&gt; pa=3D3D%#&qu=
 ot;<br>&gt;=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =
 =C2=A0 =C2=A0PRIxPADDR &quot;\n&quot;, pg, VM_PAGE_TO_PHYS(pg));<br>&gt;=C2=
 =A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0warned =3D3D tru=
 e;<br>&gt;=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0}<br>&gt;<br>&gt; If not, perha=
 ps print some information about the address that failed =3D<br>&gt; when it=
  fails...<br>&gt;<br>&gt; christos=3D<br><div><br></div><div><br></div></di=
 v>
 
 --000000000000f2509906410d8f34--