Re: [PATCH 0/4] kvm: x86: introduce CONFIG_KVM_DEBUG

2019-07-04 Thread Paolo Bonzini
On 04/07/19 04:11, wang.y...@zte.com.cn wrote: >> For MMU_DEBUG, the way to go is to add more tracepoints, but then >> converting all pgprintk occurrences to tracepoints would be wrong. You >> can only find the "right" tracepoints when debugging MMU code. I do >> have a couple patches in this are

Re: [PATCH 0/4] kvm: x86: introduce CONFIG_KVM_DEBUG

2019-07-04 Thread Paolo Bonzini
On 04/07/19 04:11, wang.y...@zte.com.cn wrote: >> For MMU_DEBUG, the way to go is to add more tracepoints, but then >> converting all pgprintk occurrences to tracepoints would be wrong. You >> can only find the "right" tracepoints when debugging MMU code. I do >> have a couple patches in this are

Re: [PATCH 0/4] kvm: x86: introduce CONFIG_KVM_DEBUG

2019-07-03 Thread Paolo Bonzini
On 01/07/19 08:21, Yi Wang wrote: > This series introduce CONFIG_KVM_DEBUG, using which we can make > the invoking *_debug in KVM simly and uniform. > > FYI: the former discussion can been found in: > https://www.spinics.net/lists/kvm/msg187026.html Basically everything except MMU_DEBUG can just

[PATCH 0/4] kvm: x86: introduce CONFIG_KVM_DEBUG

2019-06-30 Thread Yi Wang
This series introduce CONFIG_KVM_DEBUG, using which we can make the invoking *_debug in KVM simly and uniform. FYI: the former discussion can been found in: https://www.spinics.net/lists/kvm/msg187026.html Yi Wang (4): kvm: x86: Add CONFIG_KVM_DEBUG kvm: x86: allow set apic and ioapic debug d