Date: Mon, 1 Oct 2001 13:29:00 -0500 (CDT) From: Chris Dillon <cdillon@wolves.k12.mo.us> To: Keith Mitchell <kmitch@weenix.guru.org> Cc: <freebsd-stable@freebsd.org> Subject: Re: fs corruption (ATA / 4.4-REL) Message-ID: <Pine.BSF.4.32.0110011319240.11244-100000@mail.wolves.k12.mo.us> In-Reply-To: <20011001141127.A4415@weenix.guru.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On Mon, 1 Oct 2001, Keith Mitchell wrote: > > The VIA chipset is probably your culprit, though I'm not exactly sure > > if the MVP3 has the PCI bus corruption problem or if it is only the > > later KT133/KT266 chipsets. If so, your motherboard vendor is > > supposed to fix the PCI bus corruption problems in the BIOS, so try a > > BIOS update. If that doesn't fix it, you'll have to install > > 4.4-STABLE, since Soren just recently put a fix into the ata driver to > > work around a problem with those chipsets that the motherboard vendors > > should be fixing but aren't. > > That only affects DMA, right? I'm going to try and install with I don't know. I believe it can affect just about any PCI bus transfer when under heavy load, if I understand correctly. > hw.ata.wc=0 > hw.ata.ata_dma=0 > hw.ata.atapi_dma=0 Turning off wc probably won't fix the problem (unless the drive's write cache is bad?) though it will slow things down which could possibly mask the real problem. If turning off DMA fixes the problem you might not know if the problem is really DMA related or if turning off DMA just slowed things down enough to mask the real problem. > And see if that works. I don't see any of the IDMA errors that > other people are reporting (maybe because I'm only uses UDMA33 and > not UDMA66/100). All I see is data and/or filesystem corruption. > Sometimes the fs gets blaster beyond repair, other times I see > files get merged together, etc. A pain in the butt. Sounds much more like the PCI bus corruption problem than a drive that doesn't support DMA properly. Drives that don't support DMA properly generally trigger the kernel to emit the error message mentioned previously. I'm not sure if there are cases where that would not occur. -- Chris Dillon - cdillon@wolves.k12.mo.us - cdillon@inter-linc.net FreeBSD: The fastest and most stable server OS on the planet - Available for IA32 (Intel x86) and Alpha architectures - IA64, PowerPC, UltraSPARC, and ARM architectures under development - http://www.freebsd.org To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-stable" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Pine.BSF.4.32.0110011319240.11244-100000>