On Thu, 20 Feb 2020 at 18:33, James Morse [off-list ref] wrote:

Hi Ard,

On 20/02/2020 17:04, Ard Biesheuvel wrote:

quoted

On Thu, 20 Feb 2020 at 17:58, James Morse [off-list ref] wrote:

quoted

It turns out KVM relies on the inline hint being honoured by the compiler
in quite a few more places than expected. Something about the Shadow Call
Stack support[0] causes the compiler to avoid inline-ing and to place
these functions outside the __hyp_text. This ruins KVM's day.

Add the simon-says __always_inline annotation to all the static
inlines that KVM calls from HYP code.

quoted

This isn't quite as yuck as I expected, fortunately, but it does beg
the question whether we shouldn't simply map the entire kernel at EL2
instead?

If the kernel is big enough to need internal veneers (the 128M range?), these would
certainly go horribly wrong because its running somewhere other than the relocation-time
address. We would need a way of telling the linker to keep the bits of KVM close together...

Ah, of course, there is that as well ...

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel