[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: New test kernel - second attempt



Hi,

further follow-up, last tests for today. The Netra T1 was still open for the tests, the one which booted

Loading Linux 6.17.0-rc5+ ...
Loading initial ramdisk ...

/dev/sda2: clean, 33785/1011840 files, 724818/4047104 blocks
3008;start=0e6a730b368e44c897ac90aa211eb72a;user=root;hostname=debian-sparc64;machineid=8b9a152319f14cd59
23fb8a80457b56a;bootid=63ef412630eb42fca7ad2d38d1539e29;pid=1;pidfdid=2;comm=systemd;type=boot[ 45.6843
88] Fixing recursive fault but reboot is needed!
[   45.929220] systemd[1]: segfault at 32 ip fffff80100a224ac (rpc fffff80100458d3c) sp 000007fefff41cd1
error 1 in libsystemd-shared-258.so[2224ac,fffff80100800000+3fa000]
[   46.462211] systemd[1]: segfault at 28 ip fffff801011b08e8 (rpc fffff801011881f0) sp 000007fefff2d591
error 1 in libc.so.6[b08e8,fffff80101100000+194000]
[   46.654294] systemd[1]: segfault at 1008034bdd0 ip fffff801011a53ec (rpc fffff801011a5348) sp 000007fe
fff2b931 error 1 in libc.so.6[a53ec,fffff80101100000+194000]
000000000000001500000000004200000000000000000000ffffffff80300000000000000005ce18000000000000000e000000000
01f20b3000000000058003b0000000000000877000500340000000000405ec80000000000405ecc000400340000000000405ec800
00000000405ecc000300340000000000405ec80000000000405ecc000200340000000000404868000000000040486c0001004e000
000000085845c0000000000858460
Watchdog Reset
Externally Initiated Reset


the output is dirty since some color codes come out.

I did reboot as proposed... but it fails for I/O errors:
/dev/sda2: recovering journal
/dev/sda2: clean, 33795/1011840 files, 725650/4047104 blocks
3008;start=49e9dcbbe6e8415ea64c096f811f6f61;user=root;hostname=debian-sparc64;machineid=8b9a152319f14cd59
23fb8a80457b56a;bootid=ba10d5eaddab44f5bc7a3d6afd4f21b6;pid=1;pidfdid=2;comm=systemd;type=boot[ 228.4900
60] sd 0:0:0:0: [sda] tag#297 timing out command, waited 180s
[  228.574948] I/O error, dev sda, sector 13901518 op 0x1:(WRITE) flags 0x9800 phys_seg 8 prio class 2
[  228.702627] Aborting journal on device sda2-8.
[  253.583514] sd 0:0:0:0: [sda] tag#297 timing out command, waited 180s
[  253.668339] I/O error, dev sda, sector 13992638 op 0x1:(WRITE) flags 0x103000 phys_seg 1 prio class 2 [  253.789648] Buffer I/O error on dev sda2, logical block 1624066, lost async page write
[  253.894046] sd 0:0:0:0: [sda] tag#300 timing out command, waited 180s
[  253.978800] I/O error, dev sda, sector 13583030 op 0x1:(WRITE) flags 0x103000 phys_seg 1 prio class 2 [  254.100123] Buffer I/O error on dev sda2, logical block 1572865, lost async page write [  254.207606] EXT4-fs error (device sda2): ext4_journal_check_start:87: comm kworker/u4:4: Detected abor
ted journal
[  256.841511] EXT4-fs error (device sda2): ext4_journal_check_start:87: comm kworker/u4:1: Detected abor
ted journal
[  256.994388] EXT4-fs (sda2): Remounting filesystem read-only
[  257.067800] EXT4-fs (sda2): ext4_do_writepages: jbd2_start: 3067 pages, ino 393758; err -30


it is, by definition, the same disk that worked & booted on the Fire 120.

I then attempted to boot the older 6.16:

Loading Linux 6.16.7+deb14-sparc64 ...
Loading initial ramdisk ...

/dev/sda2: recovering journal
/dev/sda2 contains a file system with errors, check forced.
/dev/sda2: 33796/1011840 files (0.2% non-contiguous), 727699/4047104 blocks
fsck exited with status code 1
3008;start=ff9f4e8bd6514200ab7c4d915e3c4a75;user=root;hostname=debian-sparc64;machineid=8b9a152319f14cd59
3008;start=9057abc423f1498d9018f5cd9643795e;user=root;hostname=debian-sparc64;machineid=8b9a152319f14cd59
23fb8a80457b56a;bootid=1af80057945f45d6a4cea5ccffed75b5;pid=515;pidfdid=516;comm=(agetty);servicename=ser
ial-getty@ttyS0.service;invocationid=0ee7aa1c7d74494fbc15f021438ab905;type=service
Debian GNU/Linux forky/sid debian-sparc64 ttyS0

debian-sparc64 login: root
Password:

Login incorrect
p104 583.036741] systemd[1]: segfault at 67effd43fd9 ip fffff80101146618 (rpc fffff8010114663c) sp 000007
feffd422c1 error 1 in libc.so.6[46618,fffff80101100000+194000]
[1G



so this one completes fsck / file recovery (no scsi read errors)
I mistyped my password and got a segfault..... yay!

If I reboot and do not get my password wrong....
debian-sparc64 login: root
Password:
[  211.600247] systemd[1]: Caught <BUS>.

L

and now the lock is so bad I am unable to return to LOM with a break! argh!

At this point I try the historic 6.1.0-9 kernel that came from the CD. Boots, fsck/journal recover, login... and hangs! no errors displayed. No break...

At this point I am worried, but the NetBSD 10.1 Hard disk back in the tray, perform a cold start... and get a login prompt: Console is smooth and no funny chars.


Interesting is that the Netra T1 and Fire 120 are quite similar in hardware... the latter is a bit higher spec'd (e.g. accepts double memory) but essentially differs  with the lower-spec'd embedded UltraSparc IIe vs high-performance IIi. (the advantage is that the IIe just needs a simple CPU cooler with fan, no air ducts and is so much quieter... almost like a workstaiton).

Riccardo


Reply to: