> On two new PC's? Showing identical symptoms? I find that hard to believe. > But on the other hand... If they are the same board then its likely they will have the same problems - software or hardware. > > newish so later kernel SHOULD have better support. Maybe try a development > > kernel on one of the machines e.g. 2.6.23.----- > > This is what I am wondering...if it *is* the kernel, udev, or something > like that. This thing has 2 gb/sec throughput...it shouldn't be doing > this. Should make no difference. Assuming the BIOS has vaguely recent microcode and the like it should just work. > Not yet. But I can tell you "top" gives the full 4 GB it says I have. Of > course, that doesn't mean much. Again, I find it very difficult to believe > that two machines will have this problem. That said, I'm not ruling out > anything. Its worth checking as we've seen reports before of problems with >2GB memory that turned out to be BIOS bugs, and you can get entire batches of dud RAM together. It probably isn't the problem but it is easy to eliminate .... Also its important to get the kernel logs that go with the journal abort - the disk error sequence in particular. The fact nothing is getting displayed (if you are out of X does anything get logged to the console ?) before the abort is peculiar indeed
