From mboxrd@z Thu Jan 1 00:00:00 1970 In-Reply-To: Date: Mon, 29 Nov 1999 10:45:24 +0100 To: Hollis R Blanchard , linuxppc-user@lists.linuxppc.org CC: linuxppc-dev@lists.linuxppc.org From: Benjamin Herrenschmidt Subject: Re: BootX-1.2b2 problem? Message-Id: <19991129104524.002317@mailhost.mipsys.com> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Sender: owner-linuxppc-dev@lists.linuxppc.org List-Id: On Mon, Nov 29, 1999, Hollis R Blanchard wrote: >MHz-specific? :) I don't think so, but certainly not all bugs may not show up >on all machines... > >I think this problem manifests before the kernel starts going, though I don't >know because I haven't seen it firsthand. At any rate, this isn't the only >problem I've heard of with 1.2b2 (want to reduce your BogoMips by a factor of >ten?), so I'd be wary of it. (Again, "not all bugs...".) The problem with BootX 1.2b2 is the new IDE reset code (for fixing the MacOS 9 driver issue). Some ATAPI devices (CDROMs/DVDs) will hang the machine during this reset. (ATAPI devices do not accept an ATA reset). I tried asking MacOS to reset the bus but this apparently doesn't work. I can use the IDE reset HW wires but I only know how to access them on some machines ("heathrow" based machines). So that's really a tricky issue... I'm trying to find a way to detect which devices are ATA and which are ATAPI and trigger the reset only for ATA. The problem is that the ATA device is likely not to be responding to identify command (because of the driver problem we are trying to work around). The BogoMIPS/Slow kernel problem is beyond my understanding for now and until I can reproduce it on one of my machines here (basically a wallstreet, an iBook and a 8500). I really so no reason why BootX would cause this and I'm thinking about a possible side effect of something in MacOS triggered by BootX, but frankly, it's weird. It's still possible that something gets wrong with the VIA which is used for calibration, can someone with this problem try reverting the calibration code not to use the VIA ? (I think It's in arch/ppc/kernel/pmac_time.c or pmac_support.c) ** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/