Date: Tue, 27 Nov 2007 16:39:08 -0500 From: John Baldwin <jhb@freebsd.org> To: freebsd-current@freebsd.org Cc: Barney Cordoba <barney_cordoba@yahoo.com>, current@freebsd.org, =?iso-8859-15?q?S=F8ren_Schmidt?= <sos@deepcore.dk> Subject: Re: Any successful installs on a Broadcom HT1000 chipset? Message-ID: <200711271639.09601.jhb@freebsd.org> In-Reply-To: <474C71B1.1080104@deepcore.dk> References: <73807.10710.qm@web63912.mail.re1.yahoo.com> <474C71B1.1080104@deepcore.dk>
next in thread | previous in thread | raw e-mail | index | archive | help
On Tuesday 27 November 2007 02:36:17 pm S=F8ren Schmidt wrote: > Barney Cordoba wrote: > > S=F8ren Schmidt wrote: > >> This server has been running -current all the time > >> with about biweekly=20 > >> updates, so it has made the transition all the way > >> from about 6.1 to=20 > >> todays current without problems. > >> =20 > >>> Something happened between 6.2 and 7.0, as we were > >>> able to run 6.2 on these systems. > >>> > >>> =20 > >>> =20 > >> Looks like a pattern, something outside of ATA also > >> caused HW issues on=20 > >> some Promise controllers to appear in this period. > >> Just checked and nothing in ATA changed in regards > >> to either the HT1000=20 > >> or the Promise HW, its the same support code thats > >> been there all the time. > >> It could be a sideeffect of now having more fine > >> grained locking etc,=20 > >> that makes these problems show up, your guess is as > >> good as mine. > >> > >> Again, the sticky point is that if I cant reproduce > >> a problem here its=20 > >> very hard to work on it, if not just plain > >> impossible... > >> =20 > > > > A sticky point on this end is that it will cost more > > to ship you a system than the cost of the product; you > > being in Europe and the US dollar sucking in general. > > I'm also not going to be able to get authorization to > > ship overseas.=20 > > > > =20 > Maybe I can get the HW over here, paypal can be used to transfer the=20 > funds in that case... > > Could you send me the config file you're using on your > > 7.0 system with the H8SSL MB so I can give it a try > > and see if its a config issue, or some device > > conflict? You have the MB we're having trouble with, > > so maybe we can learn enough to give you some ideas.I > > have a system with an H8SSL-i that just plain doesn't > > work with either SATA or IDE. > > =20 > Sure find my config attached currently used with plain releng_7. > It might be worth while to compare HW revisions etc as well as there=20 > might be differences. > One thing you could try would be to install a 6.2 system and if that=20 > still works just grap ATA from releng_6 as that is now the same as in=20 > -current and should fit almost cleanly into 6.2 as is, then give that a=20 > run. Depending on the outcome we'll know more about where to look for=20 > the problem. >=20 > For references sake could we please get a list of the actual HW that=20 > people have problems with listed here, as detailed as possible please.=20 =46YI, I've seen weird in-memory corruption with machines with the HT1000_S= 1=20 atapci device. In all the cases I've seen so far, a single page is corrupt= ed=20 with garbage and the page happens to be used by UMA to hold credentials=20 including proc0's credentials. I've seen this corruption (trashed creds fo= r=20 proc0 and other creds in that page) on many of the same boxes (Dell 1435's= =20 IIRC) running on 6.2. I've tried switching the HT1000_S1 to use SWKSMIO=20 rather SWKS100 as I mentioned to you in an earlier e-mail (the Linux driver= =20 uses equivalent of SWKSMIO FWIW) but don't have any conclusive tests on tha= t. =2D-=20 John Baldwin
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200711271639.09601.jhb>