SBUS controller burnt on SPARC CPU-5V ?

Hi,

after hotplugging a SPARC CPU-5V onto a VME backplane, we couldn't boot from the SCSI disk anymore. Boot from SCSI CD-ROM didn't work either. At the OPB prompt, diag tests speak about DMA failure.

ok boot cdrom -s

Boot device: /iommu/sbus/<a href="mailto:espdma&#64;5" target="_blank">espdma@5</a>,8400000/<a href="mailto:esp&#64;5" target="_blank">esp@5</a>,8800000/<a href="mailto:sd&#64;6" target="_blank">sd@6</a>,0:d File and argss

**aborted after 2 minutes**

ok test scsi

Dma register test -- succeeded.

Esp register test -- succeeded.

Dma read test -- failed.

Dma write test -- succeeded.

scsi selftest failed. Return code = -1

ok probe-scsi

**aborted after 2 minutes**

When we do not abort the probe-scsi, we receive a message telling us that one device on the "scsi bus may be hung" , but actually we removed all SCSI devices.

We think that some SBUS controller chip may be burnt, because when we try a network boot, we see the RARP packets flowing form the SPARC CPU-5V to our ARP server ( via cross-connect cable) but the SPARC CPU-5V do not see the answer. He keeps on waiting for RARP packets.

Seems like a cmos gate on the SBUS has been burnt during hotplug.

Does someone already have seen something similar ?

This is the last chance before giving up, any clue will be appreciated.

Thanks

[1504 byte] By [yerik] at [2007-11-25 23:07:55]