Hi, you probably didn't see my complains about previous attempts to create such a patch.
I didn't see any version which supports VMX and SVM natively i.e. will walk for you extended page tables as well and provide true physical address.
Or at least I am Okay to have two functions, one dump guest page tables virtual->guest physical and another dumps virtual->host physical.
Current code works virtual->host physical but only for first 4GB.
Wouldn't it be easier to have START and END virtual address param to the function (and to the debugger command) so you will choose which virtual space you want to translate and dump ?
Thanks,
Stanislav
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
No, I didn't see previous complaints.
Certainly, I think that a VMX function seperate from the host function is a good idea. Being unfamiliar with VMX and SVM, I did not consider these advanced CPU features.
As a 64 bit OS developer, I have found the 4Gb limit a major problem with debugging, as my kernel is at 0xFFFFC00000000000. Certainly, specifying a start and end would seem unnecessarily time consuming.
So would a seperate function for SVM/VMX be suitable under "info tab-vm"?
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Hi, you probably didn't see my complains about previous attempts to create such a patch.
I didn't see any version which supports VMX and SVM natively i.e. will walk for you extended page tables as well and provide true physical address.
Or at least I am Okay to have two functions, one dump guest page tables virtual->guest physical and another dumps virtual->host physical.
Current code works virtual->host physical but only for first 4GB.
Wouldn't it be easier to have START and END virtual address param to the function (and to the debugger command) so you will choose which virtual space you want to translate and dump ?
Thanks,
Stanislav
No, I didn't see previous complaints.
Certainly, I think that a VMX function seperate from the host function is a good idea. Being unfamiliar with VMX and SVM, I did not consider these advanced CPU features.
As a 64 bit OS developer, I have found the 4Gb limit a major problem with debugging, as my kernel is at 0xFFFFC00000000000. Certainly, specifying a start and end would seem unnecessarily time consuming.
So would a seperate function for SVM/VMX be suitable under "info tab-vm"?