On Wed, Apr 09, 2014 at 02:04:50PM -0500, Romer, Benjamin M wrote: > This patch adds a function, is_spar_system(), to check that s-Par > firmware is present, and then uses this function at the beginning of > each module to verify that the modules are being run on an s-Par system > before beginning initialization. If the firmware is not detected, the > module will return a failure code. > > Checking for s-Par firmware uses the cpuid instruction to verify that > the processor is running with virtualization turned on, and then uses a > second cpuid instruction to check that the hypervisor ID is > "UnisysSpar64". Why not use the cpuid infrastructure to automatically load/bind your code and not rely on it being loaded "by hand"? Doesn't that work for CPU types already? > Additionally, some minor clean-up was done on copyright tags and > unnecessary messages were removed from the visorchipset_main() function. Patches need to do only one thing, so can you please split this up in to multiple patches, each one only doing one thing. > This patch was tested with KVM to ensure that the modules do not load > when s-Par firmware is not present, and tested with s-Par 4.0.12 to > verify that the modules load correctly when the firmware is present. > > Signed-off-by: Benjamin Romer <benjamin.romer@xxxxxxxxxx> You forgot to add a "Reported-by:" line here, and possibly a "Tested-by:" if someone tested it and reported that it solved the problem. Proper attribution is very important. thanks, greg k-h _______________________________________________ devel mailing list devel@xxxxxxxxxxxxxxxxxxxxxx http://driverdev.linuxdriverproject.org/mailman/listinfo/driverdev-devel
