[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [PATCH v3 4/4 gnumach] smp: Fix INIT/STARTUP IPI sequence
From: |
Samuel Thibault |
Subject: |
Re: [PATCH v3 4/4 gnumach] smp: Fix INIT/STARTUP IPI sequence |
Date: |
Fri, 9 Feb 2024 00:46:43 +0100 |
User-agent: |
NeoMutt/20170609 (1.8.3) |
Applied, thanks!
Damien Zammit, le mer. 07 févr. 2024 05:02:24 +0000, a ecrit:
> TODO: Don't hardcode 0x3000 as the starting eip.
>
> TESTED: works in qemu
> TESTED: works hardware with AMD cpu
> ---
> i386/i386/mp_desc.c | 15 +++--
> i386/i386/smp.c | 125 +++++++++++++++++++++++++++++-----------
> i386/i386/smp.h | 5 +-
> i386/i386at/cram.h | 5 ++
> i386/i386at/model_dep.c | 2 +-
> 5 files changed, 111 insertions(+), 41 deletions(-)
>
> diff --git a/i386/i386/mp_desc.c b/i386/i386/mp_desc.c
> index a5f6b8f6..61a7607b 100644
> --- a/i386/i386/mp_desc.c
> +++ b/i386/i386/mp_desc.c
> @@ -292,17 +292,22 @@ cpu_ap_main()
> kern_return_t
> cpu_start(int cpu)
> {
> + int err;
> +
> assert(machine_slot[cpu].running != TRUE);
>
> uint16_t apic_id = apic_get_cpu_apic_id(cpu);
>
> - printf("Trying to enable: %d\n", apic_id);
> -
> - smp_startup_cpu(apic_id, apboot_addr);
> + printf("Trying to enable: %d at 0x%lx\n", apic_id, apboot_addr);
>
> - printf("Started cpu %d (lapic id %04x)\n", cpu, apic_id);
> + err = smp_startup_cpu(apic_id, apboot_addr);
>
> - return KERN_SUCCESS;
> + if (!err) {
> + printf("Started cpu %d (lapic id %04x)\n", cpu, apic_id);
> + return KERN_SUCCESS;
> + }
> + printf("FATAL: Cannot init AP %d\n", cpu);
> + for (;;);
> }
>
> void
> diff --git a/i386/i386/smp.c b/i386/i386/smp.c
> index 87f59913..05e9de67 100644
> --- a/i386/i386/smp.c
> +++ b/i386/i386/smp.c
> @@ -18,10 +18,14 @@
> along with this program; if not, write to the Free Software
> Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111, USA. */
>
> +#include <string.h>
> #include <i386/apic.h>
> #include <i386/smp.h>
> #include <i386/cpu.h>
> +#include <i386/pio.h>
> +#include <i386/vm_param.h>
> #include <i386at/idt.h>
> +#include <i386at/cram.h>
> #include <i386at/acpi_parse_apic.h>
> #include <kern/printf.h>
> #include <mach/machine.h>
> @@ -75,59 +79,112 @@ void smp_pmap_update(unsigned apic_id)
> smp_send_ipi(apic_id, CALL_PMAP_UPDATE);
> }
>
> -/* See Intel IA32/64 Software Developer's Manual 3A Section 8.4.4.1 */
> -void smp_startup_cpu(unsigned apic_id, unsigned vector)
> +static void
> +wait_for_ipi(void)
> {
> - /* Clear APIC errors */
> - lapic->error_status.r = 0;
> + /* This could have a timeout, but if the IPI
> + * is never delivered, its a disaster anyway */
> + while (lapic->icr_low.delivery_status == SEND_PENDING) {
> + cpu_pause();
> + }
> +}
> +
> +static int
> +smp_send_ipi_init(int apic_id)
> +{
> + int err;
>
> - printf("Sending IPIs to APIC ID %u...", apic_id);
> + lapic->error_status.r = 0;
>
> - /* Assert INIT IPI */
> - apic_send_ipi(NO_SHORTHAND, INIT, PHYSICAL, ASSERT, LEVEL, 0, apic_id);
> + /* Assert INIT IPI:
> + *
> + * This is EDGE triggered to match the deassert
> + */
> + apic_send_ipi(NO_SHORTHAND, INIT, PHYSICAL, ASSERT, EDGE, 0, apic_id);
>
> /* Wait for delivery */
> - do {
> - cpu_pause();
> - } while(lapic->icr_low.delivery_status == SEND_PENDING);
> + wait_for_ipi();
> + hpet_mdelay(10);
>
> - /* Deassert INIT IPI */
> - apic_send_ipi(NO_SHORTHAND, INIT, PHYSICAL, DE_ASSERT, LEVEL, 0,
> apic_id);
> + /* Deassert INIT IPI:
> + *
> + * NB: This must be an EDGE triggered deassert signal.
> + * A LEVEL triggered deassert is only supported on very old hardware
> + * that does not support STARTUP IPIs at all, and instead jump
> + * via a warm reset vector.
> + */
> + apic_send_ipi(NO_SHORTHAND, INIT, PHYSICAL, DE_ASSERT, EDGE, 0, apic_id);
>
> /* Wait for delivery */
> - do {
> - cpu_pause();
> - } while(lapic->icr_low.delivery_status == SEND_PENDING);
> + wait_for_ipi();
>
> - /* Wait 10 msec */
> - hpet_mdelay(10);
> + err = lapic->error_status.r;
> + if (err) {
> + printf("ESR error upon INIT 0x%x\n", err);
> + }
> + return 0;
> +}
>
> - /* Clear APIC errors */
> - lapic->error_status.r = 0;
> +static int
> +smp_send_ipi_startup(int apic_id, int vector)
> +{
> + int err;
>
> - /* First StartUp IPI */
> - apic_send_ipi(NO_SHORTHAND, STARTUP, PHYSICAL, ASSERT, LEVEL, vector >>
> 12, apic_id);
> + lapic->error_status.r = 0;
>
> - /* Wait 200 usec */
> - hpet_udelay(200);
> + /* StartUp IPI:
> + *
> + * Have not seen any documentation for trigger mode for this IPI
> + * but it seems to work with EDGE. (AMD BKDG FAM16h document specifies
> dont care)
> + */
> + apic_send_ipi(NO_SHORTHAND, STARTUP, PHYSICAL, ASSERT, EDGE, vector,
> apic_id);
>
> /* Wait for delivery */
> - do {
> - cpu_pause();
> - } while(lapic->icr_low.delivery_status == SEND_PENDING);
> + wait_for_ipi();
>
> - /* Second StartUp IPI */
> - apic_send_ipi(NO_SHORTHAND, STARTUP, PHYSICAL, ASSERT, LEVEL, vector >>
> 12, apic_id);
> + err = lapic->error_status.r;
> + if (err) {
> + printf("ESR error upon STARTUP 0x%x\n", err);
> + }
> + return 0;
> +}
>
> - /* Wait 200 usec */
> +/* See Intel IA32/64 Software Developer's Manual 3A Section 8.4.4.1 */
> +int smp_startup_cpu(unsigned apic_id, phys_addr_t start_eip)
> +{
> +#if 0
> + /* This block goes with a legacy method of INIT that only works with
> + * old hardware that does not support SIPIs.
> + * Must use INIT DEASSERT LEVEL triggered IPI to use this block.
> + * (At least one AMD FCH does not support this IPI mode,
> + * See AMD BKDG FAM16h document # 48751 page 461).
> + */
> +
> + /* Tell CMOS to warm reset through through 40:67 */
> + outb(CMOS_ADDR, CMOS_SHUTDOWN);
> + outb(CMOS_DATA, CM_JMP_467);
> +
> + /* Set warm reset vector to point to AP startup code */
> + uint16_t dword[2];
> + dword[0] = 0;
> + dword[1] = start_eip >> 4;
> + memcpy((uint8_t *)phystokv(0x467), dword, 4);
> +#endif
> +
> + /* Local cache flush */
> + asm("wbinvd":::"memory");
> +
> + printf("Sending IPIs to APIC ID %u...\n", apic_id);
> +
> + smp_send_ipi_init(apic_id);
> + hpet_mdelay(10);
> + smp_send_ipi_startup(apic_id, start_eip >> STARTUP_VECTOR_SHIFT);
> + hpet_udelay(200);
> + smp_send_ipi_startup(apic_id, start_eip >> STARTUP_VECTOR_SHIFT);
> hpet_udelay(200);
> -
> - /* Wait for delivery */
> - do {
> - cpu_pause();
> - } while(lapic->icr_low.delivery_status == SEND_PENDING);
>
> printf("done\n");
> + return 0;
> }
>
> /*
> diff --git a/i386/i386/smp.h b/i386/i386/smp.h
> index 784936ea..73d273ef 100644
> --- a/i386/i386/smp.h
> +++ b/i386/i386/smp.h
> @@ -21,11 +21,14 @@
> #ifndef _SMP_H_
> #define _SMP_H_
>
> +#include <mach/machine/vm_types.h>
> +
> int smp_init(void);
> void smp_remote_ast(unsigned apic_id);
> void smp_pmap_update(unsigned apic_id);
> -void smp_startup_cpu(unsigned apic_id, unsigned vector);
> +int smp_startup_cpu(unsigned apic_id, phys_addr_t start_eip);
>
> #define cpu_pause() asm volatile ("pause" : : : "memory")
> +#define STARTUP_VECTOR_SHIFT (20 - 8)
>
> #endif
> diff --git a/i386/i386at/cram.h b/i386/i386at/cram.h
> index 8a3a6ec9..ac40cf13 100644
> --- a/i386/i386at/cram.h
> +++ b/i386/i386at/cram.h
> @@ -71,6 +71,11 @@ WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
>
> /* Addresses, related masks, and potential results */
>
> +#define CMOS_SHUTDOWN 0xf
> +#define CM_NORM_RST 0x0
> +#define CM_LOAD_SYS 0x4
> +#define CM_JMP_467 0xa
> +
> #define CMOS_EB 0x14 /* read Equipment Byte */
> #define CM_SCRMSK 0x30 /* mask for EB query to get screen */
> #define CM_EGA_VGA 0x00 /* "not CGA or MONO" */
> diff --git a/i386/i386at/model_dep.c b/i386/i386at/model_dep.c
> index b5f56c7d..05892791 100644
> --- a/i386/i386at/model_dep.c
> +++ b/i386/i386at/model_dep.c
> @@ -220,7 +220,7 @@ void machine_init(void)
> * Grab an early page for AP boot code
> */
> /* FIXME: this may not allocate from below 1MB, if within first 16MB */
> - apboot_addr = vm_page_to_pa(vm_page_grab_contig(PAGE_SIZE,
> VM_PAGE_SEL_DMA));
> + apboot_addr = 0x3000; //vm_page_to_pa(vm_page_grab_contig(PAGE_SIZE,
> VM_PAGE_SEL_DMA));
> assert (apboot_addr < 0x100000);
>
> /*
> --
> 2.43.0
>
>
>
--
Samuel
---
Pour une évaluation indépendante, transparente et rigoureuse !
Je soutiens la Commission d'Évaluation de l'Inria.