>>> On 21.09.15 at 16:09, wrote:
> On 21/09/15 15:00, Jan Beulich wrote:
> On 21.09.15 at 15:51, wrote:
>>> On 21/09/15 14:04, Jan Beulich wrote:
>>> On 17.09.15 at 13:40, wrote:
> +/* Mask out features not currently understood by Xen. */
> +eax &= (cpufeat_mask(X86_FEATU
On 21/09/15 15:00, Jan Beulich wrote:
On 21.09.15 at 15:51, wrote:
On 21/09/15 14:04, Jan Beulich wrote:
On 17.09.15 at 13:40, wrote:
Jan: I have opted for adding leaf 8 rather than reusing leaf 2, due to the
uncertainty with how this information is exposed in libxl. This patch
introduces
>>> On 21.09.15 at 15:51, wrote:
> On 21/09/15 14:04, Jan Beulich wrote:
> On 17.09.15 at 13:40, wrote:
>>> Jan: I have opted for adding leaf 8 rather than reusing leaf 2, due to the
>>> uncertainty with how this information is exposed in libxl. This patch
>>> introduces no change with how t
On 21/09/15 14:04, Jan Beulich wrote:
On 17.09.15 at 13:40, wrote:
Jan: I have opted for adding leaf 8 rather than reusing leaf 2, due to the
uncertainty with how this information is exposed in libxl. This patch
introduces no change with how the information is represented in userspace.
Mind e
>>> On 17.09.15 at 13:40, wrote:
> Jan: I have opted for adding leaf 8 rather than reusing leaf 2, due to the
> uncertainty with how this information is exposed in libxl. This patch
> introduces no change with how the information is represented in userspace.
Mind explaining this "uncertainty"? I
Convert existing cpu_has_x??? to being functions of boot_cpu_data (matching
the prevailing style), and mask out unsupported features.
Signed-off-by: Andrew Cooper
---
CC: Jan Beulich
CC: Shuai Ruan
This is another patch from my feature levelling series which is being posted
early because of in