On Mon, Sep 28, 1998 at 02:12:39PM -0400, Adam D. McKenna wrote:
> Linux version 2.0.33 (root@nose) (gcc version egcs-2.90.23 980102
> (egcs-1.0.1 release)) #1 Mon Apr 20 12:19:16 EDT 1998
>
> Besides one power outage, this box has been up for over 120 days without
> crashing. It's not an SMP box, but still, I have never seen any issues with
> compiling the kernel with egcs (on the other hand, I've never used funky
> optimization like -O99 either.)
A reply to all the success reports,
I didn't write the wait loops in the kernel, nor the compiler optimizer (surprise!).
But I reported a 2.0.3x hard lockup problem to Alan, and he told me egcs or gcc-2.8.x
was a no-go with 2.0.x.
Ofcourse, I'm cannot be sure he's right, but this guy often is, so I took his word for
it, and I haven't had problems since.
One _can_ be very lucky. Or perhaps the problem only exists in some driver which I used
and you people who report no problems don't. I don't know. Luck is such an
unpredictable
factor isn't it ? :)
I didn't get lockups on a ``heavily loaded web server'', I got lockups when doing
socket I/O
with 50+ threads creating and destroying sockets all the time and transferring approx.
80-90 MBit/s
on a fast ethernet.
React on this as you see fit. If you do not experience problems, I guess there's no
reason to
react at all. If you actually have a problem, try a gcc-2.7.x compile.
................................................................
: [EMAIL PROTECTED] : And I see the elder races, :
:.........................: putrid forms of man :
: Jakob �stergaard : See him rise and claim the earth, :
: OZ9ABN : his downfall is at hand. :
:.........................:............{Konkhra}...............:
