> This clearly points at a compiler bug, with the VIA cpu, but since
> world/kernel build cleanly and anything else bugs out quite early, I would
> suspect an optimization bug.
This may have nothing to do with it, but little story...
We used to buy whiteboxed machines, and as a test we would build
world and compile perl. One time world was fine, perl failed, and would always
seem to fail in the same spot. The company replaced every part and narrowed it
down to the power supply. Once they changed that out, perl compiled fine. The
next box had the same problems, perl stopped compiling in the same place,
and found out they re-used the power supply. Changed it out, perl compiled.
We convinced them to throw away the power supply (I'm sure someone else ended
up with it).
May be nothing of the sort, but just wanted to relay a weird
experience we ran into.
firstname.lastname@example.org mailing list
To unsubscribe, send any mail to "[EMAIL PROTECTED]"