Re: [kvm-unit-tests PATCH 3/3] x86: Support environments without test-devices

[Nadav Amit <nadav.amit@gmail.com>]

> On Jul 2, 2019, at 10:24 AM, Paolo Bonzini <pbonzini@redhat.com> wrote:
> 
> On 02/07/19 18:43, Nadav Amit wrote:
>>>> Remember that the output goes to the serial port.
>>> 
>>> RAM size can use the multiboot info (see lib/x86/setup.c).
>> 
>> The multiboot info, as provided by the boot-loader is not good enough as far
>> as I remember. The info just defines where to kernel can be loaded, but does
>> not say how big the memory is. For that, e820 decoding is needed, which I
>> was too lazy to do.
> 
> The multiboot info has both e801 memory size and e820 memory map info.
> e801 is basically the first contiguous chunk of memory below 4GB, it
> should be enough for kvm-unit-tests.

It is not enough for some of the access/ept_access tests, and 1GB page
allocation can fail on certain configurations. But anyhow I didn’t solve
this problem (IIRC the allocator only allocates from addresses that are
below 4GB). So, I can do that (using e801).

>>> For the # of CPUs I'm not sure what you're supposed to do on bare metal
>>> though. :)
>> 
>> I know you are not “serious”, but I’ll use this opportunity for a small
>> clarification. You do need to provide the real number of CPUs as otherwise
>> things will fail. I do not use cpuid, as my machine, for example has two
>> sockets. Decoding the ACPI tables is the right way, but I was too lazy to
>> implement it.
> 
> What about the mptables, too?

If you mean to reuse mptable.c from [1] or [2] - I can give it a shot. I am
not about to write my own parser.

[1] https://github.com/kernelslacker/x86info/blob/master/mptable.c
[2] https://people.freebsd.org/~fsmp/SMP/mptable/mptable.c