From mboxrd@z Thu Jan 1 00:00:00 1970 Subject: Re: benh performance problem From: Benjamin Herrenschmidt To: Hollis Blanchard Cc: Mikolaj Krzewicki , linuxppc-dev list In-Reply-To: References: <401FE138.8080807@lycos.nl> <1075847083.17327.31.camel@gaston> <40221664.70208@lycos.nl> <1076044476.885.9.camel@gaston> Content-Type: text/plain Message-Id: <1076188588.27930.57.camel@gaston> Mime-Version: 1.0 Date: Sun, 08 Feb 2004 08:16:43 +1100 Sender: owner-linuxppc-dev@lists.linuxppc.org List-Id: > pbbuttons is loading the machine? What about all those extra > decrementer ticks? I assume benchmarks showed no performance > degradation on some class of ppc32; what was the slowest system tested? > For that matter, couldn't this hurt 4xx, 8xx(x), as well? Heh, well, I did see any obvious degradation on my powerbook but I didn't benchmark much actually. Maybe I should, or make kernel HZ a config option. > Perhaps HZ should be user-configurable? Although that doesn't help > distributions who'd like one kernel to work everywhere... How does > Linux handle increased HZ on old i386 anyways? I don't know :) > > Not sure if I can fix any of this at this point without doing a > > major rewrite of the exception handling code, I suspect those > > CPUs don't like running in real mode and our exception handling > > happens mostly in that mode in ppc32... > > I didn't realize CPUs go slower in real mode..? It is the case with 970's and I think with IBM G3s, I'm not sure about Motorola G4s and other CPUs. In real mode, all memory is treated like it has the G bit set, preventing speculation (among others, I suppose prefetch gets killed too). I's very visible on lmbench with a G5: The null syscall overhead of the ppc64 kernel is lower than the one of the ppc32 kernel despite actually running more code and saving more & bigger registers. Ben. ** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/