Matthew Dillon schrieb: > > :With NFS v3 there seem still to be some open issues. > :Im running the latest (4.0)-current with the new vm/NFS changes. > :While I haven't found any problems with NFSv2 so far, v3 still seems to make > :trouble. > : > :I noticed the error some months ago, while my /usr/obj was NFS mounted, and > :a build failed while making termcap.db. Today, I gave it another try. > :I copied /usr/src/share/termcap into an NFS mounted directory and did > :a "make". I compared the output of termcap.db with the one build on the local > :drive. > :While the NFS mounted one was only 1077760 bytes in size, the correct > :size (from the local build) should be 1245184 bytes. I did the build > :several times, everytime I got the same values. I then remounted the > :direcory NFSv2. Now the build produced the right file (in size and content). > : > :The NFS Server is a Solaris 7 machine. > : > :Can anyone else confirm this error? > : > :Daniel > > I can't help you here, but I want to make sure: The problems you are > having are the same problems you were having a few months ago? ( I > want to make sure I haven't introduced new problems in -4.x, and if I > have to fix them ASAP! ). I am not sure if they are the same, but it seems so: A make world with /usr/obj NFS mounted worked up to the point where the termcap was built. Now only the error is slightly different: A few months ago, cap_mkdb exited with SIG 11, now it succeeds but generates the wrong file. I think it's the same bug. The SEGV maybe have gone away because of some library reordering over the months.
Daniel To Unsubscribe: send mail to majord...@freebsd.org with "unsubscribe freebsd-current" in the body of the message
