On Nov 1, dillema@acm.org wrote
> 
> My intel '486 fails to netboot with a current kernel (supped
> yesterday). This machine loads a kernel from disk and then
> continues to boot over NFS. It fails with the following
> messages and ending up in the kernel debugger:
> 
> nfs_boot: trying RARP (and RPC/bootparam)
> uvm_fault(0xf022d31c, 0x0, 0, 1) -> 5
> kernel: page fault trap, code=0
> stopped in swapper at  _we_write_mbuf+0x2ab: movl    0x1c(%edx),%ebx
> 
> Two other kernels build from the same sources work fine, although
> those are not used for netbooting nor use the we ethernet driver.
> 
> Any ideas?
> 

Did you use egcs (what does gcc -v report ? )
There's a codegen bug in egcs which is triggered by the we driver.
You need to recompile if_we.o and/or (I can't remember which one) dp8390.o
without -O2.

--
Manuel Bouyer, LIP6, Universite Paris VI.           Manuel.Bouyer@lip6.fr
--