elliott@shiva.trl.oz.au (Paul Elliott) (09/16/90)
Has anybody noticed any problems with screenblank causing SS1's running SUNOS 4.1 to panic. This problem occurs (not every time) at boot up time when screenblank is executed in rc.local, and has happened to several of our SS1's and an SS1+. The following messages are reported: screenblank started BAD TRAP pid 92, `screenblank': Data fault kernel read fault at addr=0xf81a4fb4, pme=0x0 Sync Error Reg 80<INVALID> Begin traceback ......... End traceback panic: Data fault When the machine is rebooted, all is ok. This problem did not occur when running SUNOS 4.0.3. ACSnet:paul@trlamct.trl.oz FAX: 61 3 543 8863 ARPA:paul%trlamct.trl.oz@uunet.uu.net PH: 61 3 541 6705 UUCP: ...!uunet!munnari!trlamct.trl.oz!paul Paul Elliott. AI Systems Telecom Australia Research Labs
mikulska%odin@ucsd.edu (Margaret Mikulska) (10/08/90)
In article <1990Sep16.130430.15551@rice.edu> elliott@shiva.trl.oz.au (Paul Elliott) writes: >Has anybody noticed any problems with screenblank causing SS1's running >SUNOS 4.1 to panic. This problem occurs (not every time) at boot up time >when screenblank is executed in rc.local, and has happened to several of >our SS1's and an SS1+. > >When the machine is rebooted, all is ok. This problem did not occur when >running SUNOS 4.0.3. Indeed yes, I came across this one, with the same, if I recall correctly, error mesages. What cured the problem for me was to specify time interval as float, not integer. We have several Suns here with screenblanks, some running 4.0, some 4.0.3, and a few 4.1. It seems that at least under 4.0 you could say "screenblank -d NNN" (for instance), but under 4.1 you have to say "screenblank -d NNN.000". Weird. (Our screenblanks are also in /etc/rc.local.) We certainly didn't have any more screenblank-related panic since that small change. Margaret Mikulska UC San Diego Dept. of Electrical and Computer Engineering mikulska@mephisto.ucsd.edu