On Wed, 2023-02-08 at 14:13 +0100, Thomas Gleixner wrote:
> On Wed, Feb 08 2023 at 00:09, Brian Gerst wrote:
> > On Tue, Feb 7, 2023 at 6:10 PM Usama Arif <usama.arif@xxxxxxxxxxxxx> wrote:
> > > Introduce a global variable 'smpboot_control' indicating to the AP how
> > > it should find its APIC ID. For a serialized bringup, the APIC ID is
> > > explicitly passed in the low bits of smpboot_control, while for parallel
> > > mode there are flags directing the AP to find its APIC ID in CPUID leaf
> > > 0x0b (for X2APIC mode) or CPUID leaf 0x01 where 8 bits are sufficient.
> >
> > For the serialized bringup case, it would be simpler to just put the
> > cpu number in the lower bits instead of the APIC id, skipping the
> > lookup.
>
> Yes and no. It can be done, but I rather prefer the consistency of the
> data and the mechanism. The "overhead" or "win" is not even measurable.
Yeah. I might do this much though, just to track CPU# simply in %ecx:
--- a/arch/x86/kernel/head_64.S
+++ b/arch/x86/kernel/head_64.S
@@ -285,17 +285,15 @@ SYM_INNER_LABEL(secondary_startup_64_no_verify,
SYM_L_GLOBAL)
leaq cpuid_to_apicid(%rip), %rbx
.Lfind_cpunr:
- cmpl (%rbx), %edx
+ cmpl (%rbx,%rcx,4), %edx
jz .Linit_cpu_data
- addq $4, %rbx
- addq $8, %rcx
+ inc %rcx
jmp .Lfind_cpunr
.Linit_cpu_data:
- /* Get the per cpu offset */
+ /* Get the per cpu offset for the given CPU# which is in ECX */
leaq __per_cpu_offset(%rip), %rbx
- addq %rcx, %rbx
- movq (%rbx), %rbx
+ movq (%rbx,%rcx,8), %rbx
/* Save it for GS BASE setup */
movq %rbx, initial_gs(%rip)
Attachment:
smime.p7s
Description: S/MIME cryptographic signature
