envbvs@epb6.lbl.gov (Brian V. Smith) (01/10/90)
In article <9001061811.AA07230@expo.lcs.mit.edu>, keith@EXPO.LCS.MIT.EDU (Keith Packard) writes: < < > OS: Ultrix 2.0 < > C compiler: gcc 1.35 < > Machine: VaxStation II. < < You've got two strikes here (other than using a VaxStation II). First: < Ultrix 2.0 is pretty old and we've never built R4 using it (I've been < using Ultrix 3.1 on my VS2000). Second; I suspect you've not < used the 'fixincludes' shell script which is included with gcc to < make the ioctl's work correctly (this from the error message < 'ENOTTY' from xterm which indicates an ioctl which the terminal < driver doesn't understand). < < The second is certainly a problem; you'll have to run fixincludes and < rebuild quite a bit of the distribution (anything using ioctl needs to get < recompiled). This is why compiling ddx/dec/qvss with /bin/cc caused the < server to start working; it can use the old ioctl mechanism. < I have a similar problem although I'm running Ultrix 3.0 and I used gcc 1.34 instead of 1.35. I swear that I ran 'fixincludes', but I also get the 'Not a typewriter' problem, so I compiled main.c, screen.c and cursor.c with cc. That fixed the 'typewriter' problem. But, the server itself is having a major problem. When I start any R4 client that wants to create a window I get the error: XIO: fatal IO error 49 (Can't assign requested address) on X server "epb2:0.0" after 29 requests (15 known processed) with 0 events remaining. after which the client dies. Clients like xdpyinfo or xrdb which don't open a window or grab the pointer work fine. What does this error mean? I get the same error if I try to run R4 clients (xterm, xclock) on an R3 server. I am compiling the whole server with cc now, to see what happens. _____________________________________ Brian V. Smith (bvsmith@lbl.gov) Lawrence Berkeley Laboratory I don't speak for LBL, these non-opinions are all mine.
envbvs@epb2.lbl.gov (Brian V. Smith) (01/13/90)
In article <4595@helios.ee.lbl.gov>, I wrote: < In article <9001061811.AA07230@expo.lcs.mit.edu>, keith@EXPO.LCS.MIT.EDU < (Keith Packard) writes: < < [ some deleted ] < < The second is certainly a problem; you'll have to run fixincludes and < < rebuild quite a bit of the distribution (anything using ioctl needs to get < < recompiled). This is why compiling ddx/dec/qvss with /bin/cc caused the < < server to start working; it can use the old ioctl mechanism. < < < < I have a similar problem although I'm running Ultrix 3.0 and I used < gcc 1.34 instead of 1.35. < The problem was that the fixincludes script in the gcc-1.34 distribution was broken so it wasn't creating any of the fixed include files it was supposed to. I have gcc-1.36 now and it has repaired (into copies) many include files and everything works fine now! All (server and clients) are compiled with gcc and they work. Thanks to all who responded. _____________________________________ Brian V. Smith (bvsmith@lbl.gov) Lawrence Berkeley Laboratory I don't speak for LBL, these non-opinions are all mine.