chet@cwns1.INS.CWRU.Edu (Chet Ramey) (01/13/90)
I just got my IBM/4.3 source distribution, and I'd like to draw on the experience of others who have rebuilt kernels from source. Is it `safe' to use hc - I think my version is 2.1o - to compile the 4.3 kernel? I've seen hc generate bad code, generate incorrect code, and generate code the assembler will reject. I don't really want to use PCC because of it's total lack of optimization, but I will if I have to. Thanks in advance. -- Chet Ramey Network Services Group "Help! Help! I'm being Case Western Reserve University repressed!" chet@ins.CWRU.Edu
jfc@athena.mit.edu (John F Carr) (01/13/90)
We build with hc2 (.1s) and have found only one compiler related problem in the kernel (this was in the Andrew File System code, which you probably don't run). I advise that you not mix compilers. I don't know exactly what the problem was, but I was able to reliably make a kernel built with a mix of hc1 and hc2 crash hard (the only fix was CTRL-ALT-PAUSE or power). --John Carr (jfc@athena.mit.edu) Project Athena Systems Development
wlm@archet.UUCP (William L. Moran Jr.) (01/14/90)
Yes, it should be safe to build a kernel with 2.1o. I've built many with hc (mostly 2.1n I think). BTW, I think the current hc version is 2.1s, but I can't remember how you go about getting it. Bill Moran -- arpa: moran-william@cs.yale.edu or wlm@ibm.com uucp: uunet!bywater!acheron!archet!wlm or decvax!yale!moran-william ------------------------------------------------------------------------------- ...the trouble with an alarm clock is that what seems sensible when you set it seems absurd when it goes off. Archie Goodwin