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: Fri, 3 Oct 2025 20:36:37 +0200

 --00000000000078ed30064045624a
 Content-Type: text/plain; charset="UTF-8"
 
 This week I tried to migrate my personal vanity server to a new rack server
 at the local computer club. The "new" server hardware is some 2013:ish
 vintage (whose exact make escapes me as it comes with no apparent vendor
 markings) with two discrete CPU sockets each sporting an Opteron 6220 eight
 core CPUs at 3.0 GHz, 64GB of DDR3, and three SATA3 slots.. I transplanted
 my system ssd with an up to date 10.1 GENERIC install into it to see how it
 liked the hardware, but shortly into boot, it panics
 in uvm_page_redim()+0x2e9.
 
 Thanks to logix on #NetBSD @ ircnet, I found
 https://mail-index.netbsd.org/netbsd-bugs/2023/01/02/msg076860.html which
 looked like the exact same error. He kindly provided me with a quick build
 of GENERIC + Nicholas' patch from
 https://mail-index.netbsd.org/netbsd-bugs/2024/04/06/msg082372.html, and
 with that patch I today got it running flawlessly. It's only been running
 for a couple of hours so far, but seems stable.
 
 So as another datapoint, this patch is necessary to boot 10.1 on dual
 Opteron 6220, otherwise the kernel panics just after cpu detect just like
 in the original PR.
 
 Can someone consider this patch, or an appropriate derivate, for pullup?
 
 Best regards,
 Anders Ali Lindgren
 
 --00000000000078ed30064045624a
 Content-Type: text/html; charset="UTF-8"
 Content-Transfer-Encoding: quoted-printable
 
 <div dir=3D"ltr"><div>This week I tried to migrate my personal vanity serve=
 r to a new rack server at the local computer club. The &quot;new&quot; serv=
 er hardware is some 2013:ish vintage (whose exact make escapes me as it com=
 es with no apparent vendor markings) with two discrete CPU sockets each spo=
 rting an Opteron 6220 eight core CPUs at 3.0 GHz, 64GB of DDR3, and three S=
 ATA3 slots.. I transplanted my system ssd with an up to date 10.1 GENERIC i=
 nstall into it to see how it liked the hardware, but shortly into boot, it =
 panics in=C2=A0uvm_page_redim()+0x2e9.</div><div><br></div><div>Thanks to l=
 ogix on #NetBSD=C2=A0@ ircnet, I found=C2=A0<a href=3D"https://mail-index.n=
 etbsd.org/netbsd-bugs/2023/01/02/msg076860.html">https://mail-index.netbsd.=
 org/netbsd-bugs/2023/01/02/msg076860.html</a> which looked like the exact s=
 ame error. He kindly provided me with a quick build of GENERIC=C2=A0+=C2=A0=
 Nicholas&#39; patch from <a href=3D"https://mail-index.netbsd.org/netbsd-bu=
 gs/2024/04/06/msg082372.html">https://mail-index.netbsd.org/netbsd-bugs/202=
 4/04/06/msg082372.html</a>, and with that patch I today got it running flaw=
 lessly. It&#39;s only been running for a couple of hours so far, but seems =
 stable.</div><div><br></div><div>So as another datapoint, this patch is nec=
 essary to boot 10.1 on dual Opteron 6220, otherwise the kernel panics just =
 after cpu detect just like in the original PR.</div><div><br></div><div>Can=
  someone consider this patch, or an appropriate derivate, for pullup?</div>=
 <div><br></div><div>Best regards,</div><div>Anders Ali Lindgren</div><div><=
 br></div><div><br></div><div><br></div></div>
 
 --00000000000078ed30064045624a--