Hi, On Sat, Mar 02, 2002 at 06:32:01PM -0800, Andrew Morton wrote: > Matti Aarnio wrote: > > > > How much the problem lies in the gcc version (RH 7.1 2.96-97.1), and > > A dog of a compiler. Burn it. I had a machine which oopsed > mysteriously once a day somewhere down playing with swapspace ptes. > Compiling that kernel with 2.91.66 fixed the problem. <sigh> There hasn't been a compiler since as long as I can remember which hasn't broken the kernel in at least one of its incarnations. gcc-2.96-97.1 should be solid for the kernel. At least one older release of 2.96 was broken in nasty ways. If I can show you a version of 2.91 which breaks the kernel, will you forswear 2.91.66 too? :-) Bero has some notes on http://www.bero.org/gcc296.html --Stephen