On 10/11/2016 15:10, gong lei wrote: > 4) The isssue can't be reporduced. I search the Intel VMX sepc about > reaseons > of vmptrld failure: > The instruction fails if its operand is not properly aligned, sets > unsupported physical-address bits, or is equal to the VMXON > pointer. In addition, the instruction fails if the 32 bits in memory > referenced by the operand do not match the VMCS > revision identifier supported by this processor. > > But I can't find any cues from the KVM source code. It seems each > error conditions is impossible in theory. :(
Yes, it should not happen. :( If it's not reproducible, it's really hard to say what it was, except a random memory corruption elsewhere or even a bit flip (!). Paolo
