On Thursday 11 September 2008 03:59:54 pm Lombard, David N wrote: > On Thu, Sep 11, 2008 at 01:23:58PM -0700, Bjorn Helgaas wrote: > > Has kexec been tested on x86 with EFI firmware? > > > > I'm testing it on ia64 with Tiano-based EFI firmware, and I > > tripped over an issue with SetVirtualAddressMap(). I would > > expect a similar problem to happen on x86, but I don't see > > any code to deal with it. > > Hmmm. I run on 32- and 64-bit kernels on a 32-bit IA EFI, and > haven't seen this problem, including chains of kexec invocations. > > I'll take a closer look at what's going on on my systems. Thanks. In case it wasn't clear, here's my confusion about x86: I don't see anything in kexec-tools that touches SetVirtualAddressMap(). And I don't see anything in Linux that prevents x86 from calling SetVirtualAddressMap() in the new kernel. If we do call SetVirtualAddressMap() in the new kernel, I would expect it to return EFI_UNSUPPORTED, which should cause Linux to panic. Bjorn
