On Thu, Jul 12, 2018 at 10:41:37PM +0100, ael wrote: > You will all love this! I just reported a checksum failure downloading
I have just got back to this. Somehow, it seems that the iso file that I downloaded (twice?) was truncated: $ ls -lh devuan_ascii_2.0.0_i386_desktop-live.iso* -rw-r--r-- 1 ael ael 1005M Jun 6 23:44 devuan_ascii_2.0.0_i386_desktop-live.iso -rw-r--r-- 1 ael ael 977M Jul 12 21:37 devuan_ascii_2.0.0_i386_desktop-live.isofail 997M rather than 1005M. As you will gather, I have now successfully downloaded a complete file. I used wget: I have never known that to fail, so I have no idea what happened. Maybe some proxy/cache had a truncated copy which might explain why the same thing happened twice. I noted a kernal bug/message. I have now found how that was triggered: it is reproducable. I usually loop-mount iso images as a check before using them. That is what triggers the dmesg dump: # mount -o loop -t iso9660 /store1/ael/devuan_images/devuan_ascii_2.0.0_i386_desktop-live.iso /testmnt/ The contents of /testmnt look OK. The bug seems to be around the kernel hung_task_timeout. But you will still like the fact that systemd is involved. But I have to admit that I have seen a kernel timing bug on this Clevo board before (reported to both debian and the proper KML, but with no response). When I have a little more time, I expect to report this problem, maybe to the same bugs. But the main point is that it seems not to be devuan specific. ael _______________________________________________ Dng mailing list Dng@lists.dyne.org https://mailinglists.dyne.org/cgi-bin/mailman/listinfo/dng
