i386/74008: IBM eServer x225 cannot boot any v5.x - endless dump scrolling in bootloader.

[Kostik Belousov]

The following reply was made to PR i386/74008; it has been noted by GNATS.

From: Kostik Belousov <kostikbel at gmail.com>
To: "Bjoern A. Zeeb" <bzeeb-lists at lists.zabbadoz.net>
Cc: bug-followup at freebsd.org, kib at freebsd.org
Subject: Re: i386/74008: IBM eServer x225 cannot boot any v5.x - endless dump scrolling in bootloader.
Date: Thu, 25 Oct 2007 21:03:31 +0300

 --xsILJLaVOAbiPmXa
 Content-Type: text/plain; charset=us-ascii
 Content-Disposition: inline
 Content-Transfer-Encoding: quoted-printable
 
 On Wed, Oct 24, 2007 at 03:36:11PM +0000, Bjoern A. Zeeb wrote:
 > On Wed, 24 Oct 2007, Bjoern A. Zeeb wrote:
 >=20
 > Hi,
 >=20
 > ok, as I was told this looks like gibberish I dumped 6 lines:
 >=20
 >=20
 > FreeBSD/i386 bootstrap loader, Revision 1.1
 > (bz at xxxxxxxx.xxxxxx.xx, Wed Oct 24 15:08:42 UTC 2007)
 >=20
 > int=3D00000000  err=3D00000000  efl=3D00030a07  eip=3D00008044
 > eax=3D0000e2c3  ebx=3D0000e401  ecx=3D000003f0  edx=3D0000f000
 > esi=3D00000000  edi=3D00006459  ebp=3D000003e2  esp=3D000003be
 > cs=3De69b  ds=3D0040  es=3D0000    fs=3D0000  gs=3D0000  ss=3D9cf8
 > cs:eip=3Df7 f1 a3 45 00 8b c2 59-f6 f5 0a 06 ca 00 a2 47
 >        00 fe c4 88 26 44 00 eb-b5 f8 c3 57 0f b6 46 18
 >        c1 e0 03 bf 2e 7d 03 f8-66 2e 8b 55 04 66 2e 8b
 >        05 5f c3 1e 8e 5e 00 8b-76 06 8b 0c 83 f9 1a 72
 >        27 83 f9 1e 72 0f 83 f9-4a 72 05 b9 4a 00 eb 08
 >        b9 1e 00 eb 03 b9 1a 00-51 56 32 c0 88 04 46 e2
 > ss:esp=3D10 3f 59 64 01 00 e2 03-d0 03 01 e4 80 00 01 00
 >        01 20 37 08 9c 08 80 00-01 00 4f 05 1e 65 e2 64
 >        01 02 00 00 00 00 40 9d-04 00 00 00 f6 03 f6 03
 >        04 00 80 00 01 00 01 02-00 00 03 00 00 80 01 e4
 >        00 f0 00 00 00 00 c4 8c-9b e6 16 02 4e 91 00 00
 >        16 02 00 00 00 00 00 00-00 00 00 00 00 00 00 00
 > BTX halted
 >=20
 >=20
 > eip=3D... (roughly) translates to .. with inval offsets...
 >=20
 > 0xffffd8a0:     div    %ecx
 > 0xffffd8a2:     mov    %eax,0xc28b0045
 > 0xffffd8a7:     pop    %ecx
 > 0xffffd8a8:     div    %ch
 > 0xffffd8aa:     or     (%esi),%al
 > 0xffffd8ac:     lret   $0xa200
 > 0xffffd8af:     inc    %edi
 > 0xffffd8b0:     add    %bh,%dh
 > 0xffffd8b2:     les    0xeb004426(%eax),%ecx
 > 0xffffd8b8:     mov    $0xf8,%ch
 > 0xffffd8ba:     ret=20
 > 0xffffd8bb:     push   %edi
 > 0xffffd8bc:     movzbl 0x18(%esi),%eax
 > 0xffffd8c0:     shl    $0x3,%eax
 > 0xffffd8c3:     mov    $0xf8037d2e,%edi
 > 0xffffd8c8:     mov    %cs:0x4(%ebp),%dx
 > 0xffffd8cd:     mov    %cs:0x8e1ec35f,%ax
 > 0xffffd8d5:     pop    %esi
 > 0xffffd8d6:     add    %cl,0xc8b0676(%ebx)
 > 0xffffd8dc:     cmp    $0x1a,%ecx
 > 0xffffd8df:     jb     0xffffd908
 > 0xffffd8e1:     cmp    $0x1e,%ecx
 > 0xffffd8e4:     jb     0xffffd8f5
 > 0xffffd8e6:     cmp    $0x4a,%ecx
 > 0xffffd8e9:     jb     0xffffd8f0
 > 0xffffd8eb:     mov    $0x8eb004a,%ecx
 > 0xffffd8f0:     mov    $0x3eb001e,%ecx
 > 0xffffd8f5:     mov    $0x5651001a,%ecx
 > 0xffffd8fa:     xor    %al,%al
 > 0xffffd8fc:     mov    %al,(%esi,%eax,2)
 
 I said this looks like garbage due to _very_ strange combination of the
 cs and eip values. But actual code is not quite a garbage. Could you check
 the CF card geometry ?
 
 --xsILJLaVOAbiPmXa
 Content-Type: application/pgp-signature
 Content-Disposition: inline
 
 -----BEGIN PGP SIGNATURE-----
 Version: GnuPG v2.0.4 (FreeBSD)
 
 iD8DBQFHINpyC3+MBN1Mb4gRAqLwAKC1yCTf5WGa0d8wUjmfPLTZjw4FrACg6h5K
 q3m/xPBLJdJBE5XFv8Jn4aE=
 =LuvA
 -----END PGP SIGNATURE-----
 
 --xsILJLaVOAbiPmXa--