pat@vitsun.UUCP (Patrick Turley) (08/03/88)
In article <2126@ssc-vax.UUCP> dmg@ssc-vax.UUCP (David Geary) writes: >"Window display lock broken because process xxxx blocked" In SunView, processes obtain locks from the windowing system before they write to the screen. These locks time out after 1 second of virtual process time (I think that's right - check the SunView manual if the precise number interests you). If the lock is broken, that is the message you get. I don't have NeWS here so I don't know a lot about it but it seems logical to infer from what I've heard that it is going through the usual SunView arbitration mechanisms (NeWS will run in a SunView window, won't it?) but taking too long on something. I was once interested in increasing the timeout value and called up Sun software support to find out if it was a system parameter that I could fiddle with in some configuration file or something. It turns out that it doesn't appear in a any configuration file but you can increase it by poking a particular location in the kernel. Of course, that means the new value is only good until the system goes down. Perhaps this could be automated by feeding adb's stdin with a script in rc.local. Pat +-------------------------+---------------------------------------+ | | | * * | Patrick Turley | | | | | * * * * | Senior Software Engineer | | | | | * * * * * * | Visual Information Technologies | | | | | * * * * * * | 3460 Lotus Drive | | | | | * * * * | Plano, TX 75075 | | | | * * | (214) 596-5600 convex!vitsun!pat | +-------------------------+---------------------------------------+
deke@socrates.ee.rochester.edu (Deke Kassabian) (08/03/88)
In article <8808022253.AA13019@vitsun.com> pat@vitsun.UUCP (Patrick Turley) writes: >In article <2126@ssc-vax.UUCP> dmg@ssc-vax.UUCP (David Geary) writes: >>"Window display lock broken because process xxxx blocked" > >In SunView, processes obtain locks from the windowing system before they >write to the screen. These locks time out after 1 second of virtual process >time (I think that's right - check the SunView manual if the precise number >interests you). If the lock is broken, that is the message you get. That sounds about right. I would agree that this suntools behaviour is probably the equivalent of what David is seeing in NeWS. I do have NeWS, but haven't seen this happen yet. > I was once interested in increasing the timeout value >and called up Sun software support to find out if it was a system parameter >that I could fiddle with in some configuration file or something. It turns >out that it doesn't appear in a any configuration file but you can >increase it by poking a particular location in the kernel. Of course, that >means the new value is only good until the system goes down. Perhaps this >could be automated by feeding adb's stdin with a script in rc.local. The better way, or at least the more common handling, is to "re-acquire" the lock *before* the timeout occurs. The timeout is necessary, by the way, to prevent a lock from remaining after a process has unexpectedly died. But this is digressing and becoming a Suntools conversation. Are we correct in assuming that this NeWS behaviour is analogous to the Suntools windowlock? ------------------------------------------------------------------------ \\\ Deke Kassabian, URochester Department of Electrical Engineering \\\ \\\ deke@ee.rochester.edu "I never metacharacter \\\ \\\ or ...!rochester!ur-valhalla!deke I didn't like......" \\\ ------------------------------------------------------------------------- "Isn't fun the BEST thing to have ?"