On September 9, linux@horizon.com wrote: > A while ago (2.5.46 or so), I tried using the 2.6 kernel and > discovered that if you have large stripe sizes, it will send > a single bio for more than 256 sectors to the IDE driver, > which complains and doesn't do the I/O. > > File system corruption follows shortly thereafter. Ouch. > > The obvious fix was just to propagate the minimum of the underlying > devices' maximum bio size to the md driver, but the general solution had > to cope with hot adds, which might include a new device with a smaller > bio size limit, and was difficult. > > Has anyone managed a fix yet? I've been a bit scared to try, although > read-only should be safe enough. Yes, I think this is fixed now. Success reports are always welcome. NeilBrown - To unsubscribe from this list: send the line "unsubscribe linux-raid" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
