On 05/17/2018 11:02 AM, Jan Beulich wrote:
>>>> On 17.05.18 at 16:47, <boris.ostrov...@oracle.com> wrote:
>> @@ -64,6 +67,9 @@ ENTRY(pvh_start_xen)
>> mov %eax,%es
>> mov %eax,%ss
>> + mov $PVH_CANARY_SEL,%eax
>> + mov %eax,%gs
> I doubt this is needed for 64-bit (you could equally well load zero or leave
> in place what's there in that case),
I don't understand this.
> and loading the selector before setting
> the base address in the descriptor won't have the intended effect.
I wasn't sure about this either but then I noticed that
secondary_startup_64() does it in the same order (although not using the
>> @@ -150,9 +170,12 @@ gdt_start:
>> .quad GDT_ENTRY(0xc09a, 0, 0xfffff) /* __KERNEL_CS */
>> .quad GDT_ENTRY(0xc092, 0, 0xfffff) /* __KERNEL_DS */
>> + .quad GDT_ENTRY(0x4090, 0, 0x18) /* PVH_CANARY_SEL */
>> - .balign 4
>> + .balign 16
>> + .fill 24, 1, 0
> This is too little space for 64-bit afaict (the canary lives at offset 40
> if I can trust asm/processor.h).
Yes, should be 48. I didn't realize the two modes use different offsets.
Xen-devel mailing list