Date: Tue, 25 Sep 2012 16:25:21 -0400 From: Iordan Iordanov <iordan@cdf.toronto.edu> To: Bob Bishop <rb@gid.co.uk> Cc: freebsd-bugs@freebsd.org, CDF Admin <admin@cdf.toronto.edu> Subject: Re: kernel panic on boot with FreeBSD 8.1 and 9.0 Message-ID: <50621331.2020207@cdf.toronto.edu> In-Reply-To: <765938FA-B2D4-4553-AA76-A86E40294080@gid.co.uk> References: <505B7A32.4070206@cdf.toronto.edu> <DEC6B674-D2BE-472F-85DC-FE2E0B82F37F@gid.co.uk> <505CAD80.1070701@cdf.toronto.edu> <765938FA-B2D4-4553-AA76-A86E40294080@gid.co.uk>
next in thread | previous in thread | raw e-mail | index | archive | help
Hi Bob, On 09/21/12 14:25, Bob Bishop wrote: >> #5 0xffffffff806ab507 at uart_bus_attaeh+0x187 > ^^^^ > Hmm. Can you disable serial ports in the BIOS? Might be a workaround. Disabling the serial ports changed the backtrace, but it still crashed. Then, I decided to play around, and ALSO disabling the IDE controller on the motherboard (for which tere is no header, funnily) allowed FreeBSD to boot. The end result was that both the IDE controller AND serial ports had to be disabled for it to boot. Once again, FreeBSD boots up fine when the PCIe 4-port network adapter is removed from the pcie (8x in 4x) port with BIOS defaults loaded. Now start the ramblings of a person who does not understand how IRQs work. Is this basically an IRQ exhaustion issue where disabling serial ports and IDE controller frees up just enough IRQs for the OS to boot up? If so, what was done in Linux to allow "sharing" IRQs so that everyhing can be enabled in the BIOS and the for kernel to still manage to drive all devices attached to the system? Is there anything I can do to help debug this before we go production? Thanks! Iordan
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?50621331.2020207>