On Saturday, April 28, 2012 18:31 CEST, Miod Vallat <m...@online.fr> wrote: > > But then later, while installing the packages, I got this: > > > > base51.tgz 98% |************************************ | 56481 KB 00:06 > > ETAgzip: stdout: Broken pipe > > ftp: -: short write > > base51.tgz 98% |************************************ | 56481 KB 09:19 > > > > Segmentation fault (core dumped) > > 'base51.tgz' did not install correctly. > > [...] > > When the installer was done with all the sets, I just went ahead, trying to > > reinstall both again, but they failed the same way again. > > If you insist this will eventually suceed (I had base install correctly > after 2 or 3 failures). > > > Is this due to the processor bugs you mentioned, one of those instabilities? > > I think so. I have only seen this on R4000 version 2.2 on Indigo. An > R4000 revision 3.0 in an Indy does not exhibit such problems. > > This hints that this is either a processor-related issue or an > Indigo-specific issue. As I only have one R4000 version 2.2 and only one > Indigo, I can't tell more. > > > going on, I had again boot from tftp due to the light: > > bootp()bootecoff bootp()bsd.IP22 --c > > You could boot from disk, a simple `boot --c' would do. The double dash > tells the PROM the argument is not for the boot command but for the boot > loader, which (in the case of the OpenBSD boot loader) passes it to the > kernel. > > > While in the installer, at the point where I configure the disks, it showed > > me sd0, sd1, and sd2. sd0 and sd1, were easily configured, until it came to > > sd2, where it did hang for a long time, unsuccessfully trying to read the > > disk layout. Only afterwards I recognized, this sd2 is not a disk, but is > > the > > Floptical drive installed in the box. I don't know whether there is an easy > > way to prevent the installer from trying to install on such device, or > > maybe its just easier to mention that as a limitation in the INSTALL.sgi > > file. > > This reminds me of very long probe time for Zip drives on macppc, but I > thought this had been fixed. This might depend on the actual SCSI > controller driver being used. Anyway, floptical drives on SGI are so > rare I don't think it is worth mentioning (although it would be nice to > get the issue fixed).
Do you have a floptical drive? I guess I should be able to remove it, without harm to the system, and could bring it with me for you to g2k12. thanks, Sebastian > > Miod
