On 3/21/06, Declan Moriarty <[EMAIL PROTECTED]> wrote: > Recently, Somebody Somewhere wrote these words > > Hi JPS, > > > > >When I last had this problem (kernel panic), everything that I > > >thought might be useful scrolled past the console too quickly > > >to reliably see, and when I wrote in what was left, I was told > > >it was insufficient. > > > Hear, hear. > > > >Running on the "real box", I see more/textual data from the > > >panic, but I'm not sure if that is sufficiently useful, either > > >- and I don't want to waste anyone's time (and be more annoying > > >than I've so far managed) with more useless data. Or is there > > >some way to capture the data, if the panic happens before > > >logging to disk works, short of redirecting > > the > > >console to a serial port and capturing it from another box? > > > I'd like to know that as well - otherwise, it seems I could > > simply cut'n'paste from your earlier mail ;-). Some registers > > may be different, but the error is essentially the same. > > > > Next question: > > > > How do I find the grsecurity patch for kernel 2.6.14.3? I have a > > strong suspicion that 2.6.14.6 might be the real problem. > > > I have both patches on disk. Mail me off list if you want them. I > never tried 2.6.14.6. > > What I really wanted to see was > > Kernel Panic!! <Error> or > > Error > Kernel Panic!! > > > If you don't get that, the kernel is borked and you don't have to > think. I presume init=/bin/bash has been tried, so it doesn't even > get that far? That's pretty fundemental. > > As for capturing them, surely you could do it on either box by > booting in with another system and making /dev/stdout or /dev/tty1 > a file, or a serial port or some such?
For this to work, the root drive has to be writeable at boot, otherwise, it might just work, I have no idea and am not going to be near a computer i have enough control of to test it until tonight (USEastern time) > There is/was a 'pause' feature used in Dos at these moments to > sort what was borked. The pause button remains on the keyboard, > needing Ctrl as well in some implementations. Hm, Pause key works for for me while loading through my BIOS but not sure about after that, with a working system, I think Scroll-Lock works for me once the kernel decompresses, but I'll have to test that at home, but getting the next *page* requires being very quick or being on a very slow machine, as it scrolls by faster than you can start/stop it. > Then > dd if=/dev/ttySx of=big_file bs=1 count=10000 grabs the first > 10k :-D. > > This is an idea, not a tested trick btw. That seems it would only work if you could reliable run commands there, but since it's dying out before that, it wouldn't be possible on a real box, but I have no experience with VMware, so I haven't a guess what it can do. > With best Regards, > > > Declan Moriarty. > -- > http://linuxfromscratch.org/mailman/listinfo/hlfs-dev > FAQ: http://www.linuxfromscratch.org/faq/ > Unsubscribe: See the above information page > Good Luck. -- Poison [BLX] Joshua M. Murphy -- http://linuxfromscratch.org/mailman/listinfo/hlfs-dev FAQ: http://www.linuxfromscratch.org/faq/ Unsubscribe: See the above information page
