|
|
|
|
|
|
|
|
|
|
|
|
| home | help | back | first | fref | pref | prev | next | nref | lref | last | post |
To: bugs@MIT.EDU Cc: wdc@MIT.EDU Date: Thu, 21 Dec 1995 15:29:52 EST From: Marc Horowitz <marc@MIT.EDU> [Rules: You must tell me how you fixed it. You must fix the bug; workarounds and kludges do not get you free ice cream. If it's fixed in a newer version of Solaris, this offer is null and void. If you can prove to me that it's a hardware bug, that's acceptable, you don't need to cut a new chip. Bill, if you can get Sun to do the above, you get the Tosci's.] Four times on three different sparc 5's in w20, I've experienced the following problem: While running a very disk-intensive application, if I move my cursor between windows, I see a failure mode where the cursor colors change, but the cursor bitmap and mask do not. I am able to notice this normally unnoticeable state because between these two operations, the sun hangs. Hard. Stop-a doesn't do a thing. Unplugging and reseating the keyboard gives me a monitor prompt. I can do simple things, including a scsi-probe. Whatever I do or don't do, "go" just leaves me in the same hung state. Whatever I can do, typing "sync" is very bad. The machine hangs without writing out any blocks of memory. The keyboard cable thing causes the keyboard to beep, but that's *it*. At this point, it's really, truly hung to the point of requiring a power-cycle. I suspect that there's an interrupt bug somewhere, when the video and disk devices are frobbed just so with just the right timing, something close to the hardware just deadlocks. Have fun :-) Marc
|
|
|
|
|
|
|
|
|
|
|
|
| home | help | back | first | fref | pref | prev | next | nref | lref | last | post |