On Tue, 2023-02-21 at 09:05 +0100, Oleksandr Natalenko wrote: > > Please see here: http://ix.io/4oLm Was that just for one CPU? Can we have them all please? The interesting part is the line starting 00000001. We're looking at the top 8 bits of EBX: Leaf Subleaf EAX EBX ECX EDX 00000000 00000000: 00000010 .... 68747541 Auth 444d4163 cAMD 69746e65 enti 00000001 00000000: 00a20f12 .... 00200800 .. . 7ef8320b .2.~ 178bfbff .... ↑↑ So the first CPU is CPU0. Could have told you that... what about the others? :) If anyone can reproduce this with a serial port, can you try this? From 98ad11d0fb88f081f49f7b1496420dbfbeff8833 Mon Sep 17 00:00:00 2001 From: David Woodhouse <dwmw@xxxxxxxxxxxx> Date: Sat, 4 Feb 2023 15:20:24 +0000 Subject: [PATCH] parallel debug --- arch/x86/kernel/head_64.S | 17 +++++++++++++++++ 1 file changed, 17 insertions(+) diff --git a/arch/x86/kernel/head_64.S b/arch/x86/kernel/head_64.S index 0e4e53d231db..da7f4d2d9951 100644 --- a/arch/x86/kernel/head_64.S +++ b/arch/x86/kernel/head_64.S @@ -281,6 +281,15 @@ SYM_INNER_LABEL(secondary_startup_64_no_verify, SYM_L_GLOBAL) .Lsetup_AP: /* EDX contains the APIC ID of the current CPU */ +#if 1 + /* Test hack: Print APIC ID and then CPU# when we find it. */ + mov %edx, %ecx + mov %edx, %eax + addb $'A', %al + mov $0x3f8, %dx + outb %al, %dx + mov %ecx, %edx +#endif xorq %rcx, %rcx leaq cpuid_to_apicid(%rip), %rbx @@ -302,6 +311,14 @@ SYM_INNER_LABEL(secondary_startup_64_no_verify, SYM_L_GLOBAL) .Linit_cpu_data: /* Get the per cpu offset for the given CPU# which is in ECX */ +#if 1 + mov %rcx, %rax + shr $3, %rax + addb $'a', %al + + mov $0x3f8, %dx + outb %al, %dx +#endif leaq __per_cpu_offset(%rip), %rbx movq (%rbx,%rcx,8), %rbx /* Save it for GS BASE setup */ -- 2.39.0
Attachment:
smime.p7s
Description: S/MIME cryptographic signature
