Merge branch irq/aic-v2 into irq/irqchip-next

* irq/aic-v2:
  : .
  : Add support for the interrupt controller found is the latest
  : incarnation of Apple M1 systems, courtesy of Hector Martin.
  : .
  irqchip/apple-aic: Add support for AICv2
  irqchip/apple-aic: Support multiple dies
  irqchip/apple-aic: Dynamically compute register offsets
  irqchip/apple-aic: Switch to irq_domain_create_tree and sparse hwirqs
  irqchip/apple-aic: Add Fast IPI support
  dt-bindings: interrupt-controller: apple,aic2: New binding for AICv2
  PCI: apple: Change MSI handling to handle 4-cell AIC fwspec form

Signed-off-by: Marc Zyngier <maz@kernel.org>
This commit is contained in:
Marc Zyngier 2022-03-11 09:10:12 +00:00
commit 89ea5be11a
4 changed files with 475 additions and 90 deletions

View File

@ -0,0 +1,98 @@
# SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause)
%YAML 1.2
---
$id: http://devicetree.org/schemas/interrupt-controller/apple,aic2.yaml#
$schema: http://devicetree.org/meta-schemas/core.yaml#
title: Apple Interrupt Controller 2
maintainers:
- Hector Martin <marcan@marcan.st>
description: |
The Apple Interrupt Controller 2 is a simple interrupt controller present on
Apple ARM SoC platforms starting with t600x (M1 Pro and Max).
It provides the following features:
- Level-triggered hardware IRQs wired to SoC blocks
- Single mask bit per IRQ
- Automatic masking on event delivery (auto-ack)
- Software triggering (ORed with hw line)
- Automatic prioritization (single event/ack register per CPU, lower IRQs =
higher priority)
- Automatic masking on ack
- Support for multiple dies
This device also represents the FIQ interrupt sources on platforms using AIC,
which do not go through a discrete interrupt controller. It also handles
FIQ-based Fast IPIs.
properties:
compatible:
items:
- const: apple,t6000-aic
- const: apple,aic2
interrupt-controller: true
'#interrupt-cells':
const: 4
description: |
The 1st cell contains the interrupt type:
- 0: Hardware IRQ
- 1: FIQ
The 2nd cell contains the die ID.
The next cell contains the interrupt number.
- HW IRQs: interrupt number
- FIQs:
- 0: physical HV timer
- 1: virtual HV timer
- 2: physical guest timer
- 3: virtual guest timer
The last cell contains the interrupt flags. This is normally
IRQ_TYPE_LEVEL_HIGH (4).
reg:
items:
- description: Address and size of the main AIC2 registers.
- description: Address and size of the AIC2 Event register.
reg-names:
items:
- const: core
- const: event
power-domains:
maxItems: 1
required:
- compatible
- '#interrupt-cells'
- interrupt-controller
- reg
- reg-names
additionalProperties: false
allOf:
- $ref: /schemas/interrupt-controller.yaml#
examples:
- |
soc {
#address-cells = <2>;
#size-cells = <2>;
aic: interrupt-controller@28e100000 {
compatible = "apple,t6000-aic", "apple,aic2";
#interrupt-cells = <4>;
interrupt-controller;
reg = <0x2 0x8e100000 0x0 0xc000>,
<0x2 0x8e10c000 0x0 0x4>;
reg-names = "core", "event";
};
};

View File

@ -1767,7 +1767,7 @@ T: git https://github.com/AsahiLinux/linux.git
F: Documentation/devicetree/bindings/arm/apple.yaml
F: Documentation/devicetree/bindings/arm/apple/*
F: Documentation/devicetree/bindings/i2c/apple,i2c.yaml
F: Documentation/devicetree/bindings/interrupt-controller/apple,aic.yaml
F: Documentation/devicetree/bindings/interrupt-controller/apple,*
F: Documentation/devicetree/bindings/mailbox/apple,mailbox.yaml
F: Documentation/devicetree/bindings/pci/apple,pcie.yaml
F: Documentation/devicetree/bindings/pinctrl/apple,pinctrl.yaml

View File

@ -24,7 +24,7 @@
* - Default "this CPU" register view and explicit per-CPU views
*
* In addition, this driver also handles FIQs, as these are routed to the same
* IRQ vector. These are used for Fast IPIs (TODO), the ARMv8 timer IRQs, and
* IRQ vector. These are used for Fast IPIs, the ARMv8 timer IRQs, and
* performance counters (TODO).
*
* Implementation notes:
@ -52,10 +52,12 @@
#include <linux/irqchip.h>
#include <linux/irqchip/arm-vgic-info.h>
#include <linux/irqdomain.h>
#include <linux/jump_label.h>
#include <linux/limits.h>
#include <linux/of_address.h>
#include <linux/slab.h>
#include <asm/apple_m1_pmu.h>
#include <asm/cputype.h>
#include <asm/exception.h>
#include <asm/sysreg.h>
#include <asm/virt.h>
@ -63,20 +65,22 @@
#include <dt-bindings/interrupt-controller/apple-aic.h>
/*
* AIC registers (MMIO)
* AIC v1 registers (MMIO)
*/
#define AIC_INFO 0x0004
#define AIC_INFO_NR_HW GENMASK(15, 0)
#define AIC_INFO_NR_IRQ GENMASK(15, 0)
#define AIC_CONFIG 0x0010
#define AIC_WHOAMI 0x2000
#define AIC_EVENT 0x2004
#define AIC_EVENT_TYPE GENMASK(31, 16)
#define AIC_EVENT_DIE GENMASK(31, 24)
#define AIC_EVENT_TYPE GENMASK(23, 16)
#define AIC_EVENT_NUM GENMASK(15, 0)
#define AIC_EVENT_TYPE_HW 1
#define AIC_EVENT_TYPE_FIQ 0 /* Software use */
#define AIC_EVENT_TYPE_IRQ 1
#define AIC_EVENT_TYPE_IPI 4
#define AIC_EVENT_IPI_OTHER 1
#define AIC_EVENT_IPI_SELF 2
@ -92,22 +96,71 @@
#define AIC_IPI_SELF BIT(31)
#define AIC_TARGET_CPU 0x3000
#define AIC_SW_SET 0x4000
#define AIC_SW_CLR 0x4080
#define AIC_MASK_SET 0x4100
#define AIC_MASK_CLR 0x4180
#define AIC_CPU_IPI_SET(cpu) (0x5008 + ((cpu) << 7))
#define AIC_CPU_IPI_CLR(cpu) (0x500c + ((cpu) << 7))
#define AIC_CPU_IPI_MASK_SET(cpu) (0x5024 + ((cpu) << 7))
#define AIC_CPU_IPI_MASK_CLR(cpu) (0x5028 + ((cpu) << 7))
#define AIC_MAX_IRQ 0x400
/*
* AIC v2 registers (MMIO)
*/
#define AIC2_VERSION 0x0000
#define AIC2_VERSION_VER GENMASK(7, 0)
#define AIC2_INFO1 0x0004
#define AIC2_INFO1_NR_IRQ GENMASK(15, 0)
#define AIC2_INFO1_LAST_DIE GENMASK(27, 24)
#define AIC2_INFO2 0x0008
#define AIC2_INFO3 0x000c
#define AIC2_INFO3_MAX_IRQ GENMASK(15, 0)
#define AIC2_INFO3_MAX_DIE GENMASK(27, 24)
#define AIC2_RESET 0x0010
#define AIC2_RESET_RESET BIT(0)
#define AIC2_CONFIG 0x0014
#define AIC2_CONFIG_ENABLE BIT(0)
#define AIC2_CONFIG_PREFER_PCPU BIT(28)
#define AIC2_TIMEOUT 0x0028
#define AIC2_CLUSTER_PRIO 0x0030
#define AIC2_DELAY_GROUPS 0x0100
#define AIC2_IRQ_CFG 0x2000
/*
* AIC2 registers are laid out like this, starting at AIC2_IRQ_CFG:
*
* Repeat for each die:
* IRQ_CFG: u32 * MAX_IRQS
* SW_SET: u32 * (MAX_IRQS / 32)
* SW_CLR: u32 * (MAX_IRQS / 32)
* MASK_SET: u32 * (MAX_IRQS / 32)
* MASK_CLR: u32 * (MAX_IRQS / 32)
* HW_STATE: u32 * (MAX_IRQS / 32)
*
* This is followed by a set of event registers, each 16K page aligned.
* The first one is the AP event register we will use. Unfortunately,
* the actual implemented die count is not specified anywhere in the
* capability registers, so we have to explicitly specify the event
* register as a second reg entry in the device tree to remain
* forward-compatible.
*/
#define AIC2_IRQ_CFG_TARGET GENMASK(3, 0)
#define AIC2_IRQ_CFG_DELAY_IDX GENMASK(7, 5)
#define MASK_REG(x) (4 * ((x) >> 5))
#define MASK_BIT(x) BIT((x) & GENMASK(4, 0))
/*
* IMP-DEF sysregs that control FIQ sources
* Note: sysreg-based IPIs are not supported yet.
*/
/* IPI request registers */
@ -146,6 +199,17 @@
#define SYS_IMP_APL_UPMSR_EL1 sys_reg(3, 7, 15, 6, 4)
#define UPMSR_IACT BIT(0)
/* MPIDR fields */
#define MPIDR_CPU(x) MPIDR_AFFINITY_LEVEL(x, 0)
#define MPIDR_CLUSTER(x) MPIDR_AFFINITY_LEVEL(x, 1)
#define AIC_IRQ_HWIRQ(die, irq) (FIELD_PREP(AIC_EVENT_DIE, die) | \
FIELD_PREP(AIC_EVENT_TYPE, AIC_EVENT_TYPE_IRQ) | \
FIELD_PREP(AIC_EVENT_NUM, irq))
#define AIC_FIQ_HWIRQ(x) (FIELD_PREP(AIC_EVENT_TYPE, AIC_EVENT_TYPE_FIQ) | \
FIELD_PREP(AIC_EVENT_NUM, x))
#define AIC_HWIRQ_IRQ(x) FIELD_GET(AIC_EVENT_NUM, x)
#define AIC_HWIRQ_DIE(x) FIELD_GET(AIC_EVENT_DIE, x)
#define AIC_NR_FIQ 6
#define AIC_NR_SWIPI 32
@ -164,14 +228,81 @@
#define AIC_TMR_EL02_PHYS AIC_TMR_GUEST_PHYS
#define AIC_TMR_EL02_VIRT AIC_TMR_GUEST_VIRT
DEFINE_STATIC_KEY_TRUE(use_fast_ipi);
struct aic_info {
int version;
/* Register offsets */
u32 event;
u32 target_cpu;
u32 irq_cfg;
u32 sw_set;
u32 sw_clr;
u32 mask_set;
u32 mask_clr;
u32 die_stride;
/* Features */
bool fast_ipi;
};
static const struct aic_info aic1_info = {
.version = 1,
.event = AIC_EVENT,
.target_cpu = AIC_TARGET_CPU,
};
static const struct aic_info aic1_fipi_info = {
.version = 1,
.event = AIC_EVENT,
.target_cpu = AIC_TARGET_CPU,
.fast_ipi = true,
};
static const struct aic_info aic2_info = {
.version = 2,
.irq_cfg = AIC2_IRQ_CFG,
.fast_ipi = true,
};
static const struct of_device_id aic_info_match[] = {
{
.compatible = "apple,t8103-aic",
.data = &aic1_fipi_info,
},
{
.compatible = "apple,aic",
.data = &aic1_info,
},
{
.compatible = "apple,aic2",
.data = &aic2_info,
},
{}
};
struct aic_irq_chip {
void __iomem *base;
void __iomem *event;
struct irq_domain *hw_domain;
struct irq_domain *ipi_domain;
struct {
cpumask_t aff;
} *fiq_aff[AIC_NR_FIQ];
int nr_hw;
int nr_irq;
int max_irq;
int nr_die;
int max_die;
struct aic_info info;
};
static DEFINE_PER_CPU(uint32_t, aic_fiq_unmasked);
@ -199,18 +330,24 @@ static void aic_ic_write(struct aic_irq_chip *ic, u32 reg, u32 val)
static void aic_irq_mask(struct irq_data *d)
{
irq_hw_number_t hwirq = irqd_to_hwirq(d);
struct aic_irq_chip *ic = irq_data_get_irq_chip_data(d);
aic_ic_write(ic, AIC_MASK_SET + MASK_REG(irqd_to_hwirq(d)),
MASK_BIT(irqd_to_hwirq(d)));
u32 off = AIC_HWIRQ_DIE(hwirq) * ic->info.die_stride;
u32 irq = AIC_HWIRQ_IRQ(hwirq);
aic_ic_write(ic, ic->info.mask_set + off + MASK_REG(irq), MASK_BIT(irq));
}
static void aic_irq_unmask(struct irq_data *d)
{
irq_hw_number_t hwirq = irqd_to_hwirq(d);
struct aic_irq_chip *ic = irq_data_get_irq_chip_data(d);
aic_ic_write(ic, AIC_MASK_CLR + MASK_REG(d->hwirq),
MASK_BIT(irqd_to_hwirq(d)));
u32 off = AIC_HWIRQ_DIE(hwirq) * ic->info.die_stride;
u32 irq = AIC_HWIRQ_IRQ(hwirq);
aic_ic_write(ic, ic->info.mask_clr + off + MASK_REG(irq), MASK_BIT(irq));
}
static void aic_irq_eoi(struct irq_data *d)
@ -233,12 +370,12 @@ static void __exception_irq_entry aic_handle_irq(struct pt_regs *regs)
* We cannot use a relaxed read here, as reads from DMA buffers
* need to be ordered after the IRQ fires.
*/
event = readl(ic->base + AIC_EVENT);
event = readl(ic->event + ic->info.event);
type = FIELD_GET(AIC_EVENT_TYPE, event);
irq = FIELD_GET(AIC_EVENT_NUM, event);
if (type == AIC_EVENT_TYPE_HW)
generic_handle_domain_irq(aic_irqc->hw_domain, irq);
if (type == AIC_EVENT_TYPE_IRQ)
generic_handle_domain_irq(aic_irqc->hw_domain, event);
else if (type == AIC_EVENT_TYPE_IPI && irq == 1)
aic_handle_ipi(regs);
else if (event != 0)
@ -265,12 +402,14 @@ static int aic_irq_set_affinity(struct irq_data *d,
struct aic_irq_chip *ic = irq_data_get_irq_chip_data(d);
int cpu;
BUG_ON(!ic->info.target_cpu);
if (force)
cpu = cpumask_first(mask_val);
else
cpu = cpumask_any_and(mask_val, cpu_online_mask);
aic_ic_write(ic, AIC_TARGET_CPU + hwirq * 4, BIT(cpu));
aic_ic_write(ic, ic->info.target_cpu + AIC_HWIRQ_IRQ(hwirq) * 4, BIT(cpu));
irq_data_update_effective_affinity(d, cpumask_of(cpu));
return IRQ_SET_MASK_OK;
@ -294,15 +433,21 @@ static struct irq_chip aic_chip = {
.irq_set_type = aic_irq_set_type,
};
static struct irq_chip aic2_chip = {
.name = "AIC2",
.irq_mask = aic_irq_mask,
.irq_unmask = aic_irq_unmask,
.irq_eoi = aic_irq_eoi,
.irq_set_type = aic_irq_set_type,
};
/*
* FIQ irqchip
*/
static unsigned long aic_fiq_get_idx(struct irq_data *d)
{
struct aic_irq_chip *ic = irq_data_get_irq_chip_data(d);
return irqd_to_hwirq(d) - ic->nr_hw;
return AIC_HWIRQ_IRQ(irqd_to_hwirq(d));
}
static void aic_fiq_set_mask(struct irq_data *d)
@ -380,17 +525,21 @@ static void __exception_irq_entry aic_handle_fiq(struct pt_regs *regs)
*/
if (read_sysreg_s(SYS_IMP_APL_IPI_SR_EL1) & IPI_SR_PENDING) {
pr_err_ratelimited("Fast IPI fired. Acking.\n");
write_sysreg_s(IPI_SR_PENDING, SYS_IMP_APL_IPI_SR_EL1);
if (static_branch_likely(&use_fast_ipi)) {
aic_handle_ipi(regs);
} else {
pr_err_ratelimited("Fast IPI fired. Acking.\n");
write_sysreg_s(IPI_SR_PENDING, SYS_IMP_APL_IPI_SR_EL1);
}
}
if (TIMER_FIRING(read_sysreg(cntp_ctl_el0)))
generic_handle_domain_irq(aic_irqc->hw_domain,
aic_irqc->nr_hw + AIC_TMR_EL0_PHYS);
AIC_FIQ_HWIRQ(AIC_TMR_EL0_PHYS));
if (TIMER_FIRING(read_sysreg(cntv_ctl_el0)))
generic_handle_domain_irq(aic_irqc->hw_domain,
aic_irqc->nr_hw + AIC_TMR_EL0_VIRT);
AIC_FIQ_HWIRQ(AIC_TMR_EL0_VIRT));
if (is_kernel_in_hyp_mode()) {
uint64_t enabled = read_sysreg_s(SYS_IMP_APL_VM_TMR_FIQ_ENA_EL2);
@ -398,12 +547,12 @@ static void __exception_irq_entry aic_handle_fiq(struct pt_regs *regs)
if ((enabled & VM_TMR_FIQ_ENABLE_P) &&
TIMER_FIRING(read_sysreg_s(SYS_CNTP_CTL_EL02)))
generic_handle_domain_irq(aic_irqc->hw_domain,
aic_irqc->nr_hw + AIC_TMR_EL02_PHYS);
AIC_FIQ_HWIRQ(AIC_TMR_EL02_PHYS));
if ((enabled & VM_TMR_FIQ_ENABLE_V) &&
TIMER_FIRING(read_sysreg_s(SYS_CNTV_CTL_EL02)))
generic_handle_domain_irq(aic_irqc->hw_domain,
aic_irqc->nr_hw + AIC_TMR_EL02_VIRT);
AIC_FIQ_HWIRQ(AIC_TMR_EL02_VIRT));
}
if (read_sysreg_s(SYS_IMP_APL_PMCR0_EL1) & PMCR0_IACT) {
@ -414,7 +563,7 @@ static void __exception_irq_entry aic_handle_fiq(struct pt_regs *regs)
else
irq = AIC_CPU_PMU_E;
generic_handle_domain_irq(aic_irqc->hw_domain,
aic_irqc->nr_hw + irq);
AIC_FIQ_HWIRQ(irq));
}
if (FIELD_GET(UPMCR0_IMODE, read_sysreg_s(SYS_IMP_APL_UPMCR0_EL1)) == UPMCR0_IMODE_FIQ &&
@ -448,13 +597,18 @@ static int aic_irq_domain_map(struct irq_domain *id, unsigned int irq,
irq_hw_number_t hw)
{
struct aic_irq_chip *ic = id->host_data;
u32 type = FIELD_GET(AIC_EVENT_TYPE, hw);
struct irq_chip *chip = &aic_chip;
if (hw < ic->nr_hw) {
irq_domain_set_info(id, irq, hw, &aic_chip, id->host_data,
if (ic->info.version == 2)
chip = &aic2_chip;
if (type == AIC_EVENT_TYPE_IRQ) {
irq_domain_set_info(id, irq, hw, chip, id->host_data,
handle_fasteoi_irq, NULL, NULL);
irqd_set_single_target(irq_desc_get_irq_data(irq_to_desc(irq)));
} else {
int fiq = hw - ic->nr_hw;
int fiq = FIELD_GET(AIC_EVENT_NUM, hw);
switch (fiq) {
case AIC_CPU_PMU_P:
@ -479,32 +633,46 @@ static int aic_irq_domain_translate(struct irq_domain *id,
unsigned int *type)
{
struct aic_irq_chip *ic = id->host_data;
u32 *args;
u32 die = 0;
if (fwspec->param_count != 3 || !is_of_node(fwspec->fwnode))
if (fwspec->param_count < 3 || fwspec->param_count > 4 ||
!is_of_node(fwspec->fwnode))
return -EINVAL;
args = &fwspec->param[1];
if (fwspec->param_count == 4) {
die = args[0];
args++;
}
switch (fwspec->param[0]) {
case AIC_IRQ:
if (fwspec->param[1] >= ic->nr_hw)
if (die >= ic->nr_die)
return -EINVAL;
*hwirq = fwspec->param[1];
if (args[0] >= ic->nr_irq)
return -EINVAL;
*hwirq = AIC_IRQ_HWIRQ(die, args[0]);
break;
case AIC_FIQ:
if (fwspec->param[1] >= AIC_NR_FIQ)
if (die != 0)
return -EINVAL;
*hwirq = ic->nr_hw + fwspec->param[1];
if (args[0] >= AIC_NR_FIQ)
return -EINVAL;
*hwirq = AIC_FIQ_HWIRQ(args[0]);
/*
* In EL1 the non-redirected registers are the guest's,
* not EL2's, so remap the hwirqs to match.
*/
if (!is_kernel_in_hyp_mode()) {
switch (fwspec->param[1]) {
switch (args[0]) {
case AIC_TMR_GUEST_PHYS:
*hwirq = ic->nr_hw + AIC_TMR_EL0_PHYS;
*hwirq = AIC_FIQ_HWIRQ(AIC_TMR_EL0_PHYS);
break;
case AIC_TMR_GUEST_VIRT:
*hwirq = ic->nr_hw + AIC_TMR_EL0_VIRT;
*hwirq = AIC_FIQ_HWIRQ(AIC_TMR_EL0_VIRT);
break;
case AIC_TMR_HV_PHYS:
case AIC_TMR_HV_VIRT:
@ -518,7 +686,7 @@ static int aic_irq_domain_translate(struct irq_domain *id,
return -EINVAL;
}
*type = fwspec->param[2] & IRQ_TYPE_SENSE_MASK;
*type = args[1] & IRQ_TYPE_SENSE_MASK;
return 0;
}
@ -567,6 +735,22 @@ static const struct irq_domain_ops aic_irq_domain_ops = {
* IPI irqchip
*/
static void aic_ipi_send_fast(int cpu)
{
u64 mpidr = cpu_logical_map(cpu);
u64 my_mpidr = read_cpuid_mpidr();
u64 cluster = MPIDR_CLUSTER(mpidr);
u64 idx = MPIDR_CPU(mpidr);
if (MPIDR_CLUSTER(my_mpidr) == cluster)
write_sysreg_s(FIELD_PREP(IPI_RR_CPU, idx),
SYS_IMP_APL_IPI_RR_LOCAL_EL1);
else
write_sysreg_s(FIELD_PREP(IPI_RR_CPU, idx) | FIELD_PREP(IPI_RR_CLUSTER, cluster),
SYS_IMP_APL_IPI_RR_GLOBAL_EL1);
isb();
}
static void aic_ipi_mask(struct irq_data *d)
{
u32 irq_bit = BIT(irqd_to_hwirq(d));
@ -592,8 +776,12 @@ static void aic_ipi_unmask(struct irq_data *d)
* If a pending vIPI was unmasked, raise a HW IPI to ourselves.
* No barriers needed here since this is a self-IPI.
*/
if (atomic_read(this_cpu_ptr(&aic_vipi_flag)) & irq_bit)
aic_ic_write(ic, AIC_IPI_SEND, AIC_IPI_SEND_CPU(smp_processor_id()));
if (atomic_read(this_cpu_ptr(&aic_vipi_flag)) & irq_bit) {
if (static_branch_likely(&use_fast_ipi))
aic_ipi_send_fast(smp_processor_id());
else
aic_ic_write(ic, AIC_IPI_SEND, AIC_IPI_SEND_CPU(smp_processor_id()));
}
}
static void aic_ipi_send_mask(struct irq_data *d, const struct cpumask *mask)
@ -621,8 +809,12 @@ static void aic_ipi_send_mask(struct irq_data *d, const struct cpumask *mask)
smp_mb__after_atomic();
if (!(pending & irq_bit) &&
(atomic_read(per_cpu_ptr(&aic_vipi_enable, cpu)) & irq_bit))
send |= AIC_IPI_SEND_CPU(cpu);
(atomic_read(per_cpu_ptr(&aic_vipi_enable, cpu)) & irq_bit)) {
if (static_branch_likely(&use_fast_ipi))
aic_ipi_send_fast(cpu);
else
send |= AIC_IPI_SEND_CPU(cpu);
}
}
/*
@ -654,8 +846,16 @@ static void aic_handle_ipi(struct pt_regs *regs)
/*
* Ack the IPI. We need to order this after the AIC event read, but
* that is enforced by normal MMIO ordering guarantees.
*
* For the Fast IPI case, this needs to be ordered before the vIPI
* handling below, so we need to isb();
*/
aic_ic_write(aic_irqc, AIC_IPI_ACK, AIC_IPI_OTHER);
if (static_branch_likely(&use_fast_ipi)) {
write_sysreg_s(IPI_SR_PENDING, SYS_IMP_APL_IPI_SR_EL1);
isb();
} else {
aic_ic_write(aic_irqc, AIC_IPI_ACK, AIC_IPI_OTHER);
}
/*
* The mask read does not need to be ordered. Only we can change
@ -683,7 +883,8 @@ static void aic_handle_ipi(struct pt_regs *regs)
* No ordering needed here; at worst this just changes the timing of
* when the next IPI will be delivered.
*/
aic_ic_write(aic_irqc, AIC_IPI_MASK_CLR, AIC_IPI_OTHER);
if (!static_branch_likely(&use_fast_ipi))
aic_ic_write(aic_irqc, AIC_IPI_MASK_CLR, AIC_IPI_OTHER);
}
static int aic_ipi_alloc(struct irq_domain *d, unsigned int virq,
@ -770,20 +971,27 @@ static int aic_init_cpu(unsigned int cpu)
/* Commit all of the above */
isb();
/*
* Make sure the kernel's idea of logical CPU order is the same as AIC's
* If we ever end up with a mismatch here, we will have to introduce
* a mapping table similar to what other irqchip drivers do.
*/
WARN_ON(aic_ic_read(aic_irqc, AIC_WHOAMI) != smp_processor_id());
if (aic_irqc->info.version == 1) {
/*
* Make sure the kernel's idea of logical CPU order is the same as AIC's
* If we ever end up with a mismatch here, we will have to introduce
* a mapping table similar to what other irqchip drivers do.
*/
WARN_ON(aic_ic_read(aic_irqc, AIC_WHOAMI) != smp_processor_id());
/*
* Always keep IPIs unmasked at the hardware level (except auto-masking
* by AIC during processing). We manage masks at the vIPI level.
*/
aic_ic_write(aic_irqc, AIC_IPI_ACK, AIC_IPI_SELF | AIC_IPI_OTHER);
aic_ic_write(aic_irqc, AIC_IPI_MASK_SET, AIC_IPI_SELF);
aic_ic_write(aic_irqc, AIC_IPI_MASK_CLR, AIC_IPI_OTHER);
/*
* Always keep IPIs unmasked at the hardware level (except auto-masking
* by AIC during processing). We manage masks at the vIPI level.
* These registers only exist on AICv1, AICv2 always uses fast IPIs.
*/
aic_ic_write(aic_irqc, AIC_IPI_ACK, AIC_IPI_SELF | AIC_IPI_OTHER);
if (static_branch_likely(&use_fast_ipi)) {
aic_ic_write(aic_irqc, AIC_IPI_MASK_SET, AIC_IPI_SELF | AIC_IPI_OTHER);
} else {
aic_ic_write(aic_irqc, AIC_IPI_MASK_SET, AIC_IPI_SELF);
aic_ic_write(aic_irqc, AIC_IPI_MASK_CLR, AIC_IPI_OTHER);
}
}
/* Initialize the local mask state */
__this_cpu_write(aic_fiq_unmasked, 0);
@ -836,43 +1044,97 @@ static void build_fiq_affinity(struct aic_irq_chip *ic, struct device_node *aff)
static int __init aic_of_ic_init(struct device_node *node, struct device_node *parent)
{
int i;
int i, die;
u32 off, start_off;
void __iomem *regs;
u32 info;
struct aic_irq_chip *irqc;
struct device_node *affs;
const struct of_device_id *match;
regs = of_iomap(node, 0);
if (WARN_ON(!regs))
return -EIO;
irqc = kzalloc(sizeof(*irqc), GFP_KERNEL);
if (!irqc)
if (!irqc) {
iounmap(regs);
return -ENOMEM;
}
aic_irqc = irqc;
irqc->base = regs;
info = aic_ic_read(irqc, AIC_INFO);
irqc->nr_hw = FIELD_GET(AIC_INFO_NR_HW, info);
match = of_match_node(aic_info_match, node);
if (!match)
goto err_unmap;
irqc->hw_domain = irq_domain_create_linear(of_node_to_fwnode(node),
irqc->nr_hw + AIC_NR_FIQ,
&aic_irq_domain_ops, irqc);
if (WARN_ON(!irqc->hw_domain)) {
iounmap(irqc->base);
kfree(irqc);
return -ENODEV;
irqc->info = *(struct aic_info *)match->data;
aic_irqc = irqc;
switch (irqc->info.version) {
case 1: {
u32 info;
info = aic_ic_read(irqc, AIC_INFO);
irqc->nr_irq = FIELD_GET(AIC_INFO_NR_IRQ, info);
irqc->max_irq = AIC_MAX_IRQ;
irqc->nr_die = irqc->max_die = 1;
off = start_off = irqc->info.target_cpu;
off += sizeof(u32) * irqc->max_irq; /* TARGET_CPU */
irqc->event = irqc->base;
break;
}
case 2: {
u32 info1, info3;
info1 = aic_ic_read(irqc, AIC2_INFO1);
info3 = aic_ic_read(irqc, AIC2_INFO3);
irqc->nr_irq = FIELD_GET(AIC2_INFO1_NR_IRQ, info1);
irqc->max_irq = FIELD_GET(AIC2_INFO3_MAX_IRQ, info3);
irqc->nr_die = FIELD_GET(AIC2_INFO1_LAST_DIE, info1) + 1;
irqc->max_die = FIELD_GET(AIC2_INFO3_MAX_DIE, info3);
off = start_off = irqc->info.irq_cfg;
off += sizeof(u32) * irqc->max_irq; /* IRQ_CFG */
irqc->event = of_iomap(node, 1);
if (WARN_ON(!irqc->event))
goto err_unmap;
break;
}
}
irqc->info.sw_set = off;
off += sizeof(u32) * (irqc->max_irq >> 5); /* SW_SET */
irqc->info.sw_clr = off;
off += sizeof(u32) * (irqc->max_irq >> 5); /* SW_CLR */
irqc->info.mask_set = off;
off += sizeof(u32) * (irqc->max_irq >> 5); /* MASK_SET */
irqc->info.mask_clr = off;
off += sizeof(u32) * (irqc->max_irq >> 5); /* MASK_CLR */
off += sizeof(u32) * (irqc->max_irq >> 5); /* HW_STATE */
if (irqc->info.fast_ipi)
static_branch_enable(&use_fast_ipi);
else
static_branch_disable(&use_fast_ipi);
irqc->info.die_stride = off - start_off;
irqc->hw_domain = irq_domain_create_tree(of_node_to_fwnode(node),
&aic_irq_domain_ops, irqc);
if (WARN_ON(!irqc->hw_domain))
goto err_unmap;
irq_domain_update_bus_token(irqc->hw_domain, DOMAIN_BUS_WIRED);
if (aic_init_smp(irqc, node)) {
irq_domain_remove(irqc->hw_domain);
iounmap(irqc->base);
kfree(irqc);
return -ENODEV;
}
if (aic_init_smp(irqc, node))
goto err_remove_domain;
affs = of_get_child_by_name(node, "affinities");
if (affs) {
@ -885,26 +1147,51 @@ static int __init aic_of_ic_init(struct device_node *node, struct device_node *p
set_handle_irq(aic_handle_irq);
set_handle_fiq(aic_handle_fiq);
for (i = 0; i < BITS_TO_U32(irqc->nr_hw); i++)
aic_ic_write(irqc, AIC_MASK_SET + i * 4, U32_MAX);
for (i = 0; i < BITS_TO_U32(irqc->nr_hw); i++)
aic_ic_write(irqc, AIC_SW_CLR + i * 4, U32_MAX);
for (i = 0; i < irqc->nr_hw; i++)
aic_ic_write(irqc, AIC_TARGET_CPU + i * 4, 1);
off = 0;
for (die = 0; die < irqc->nr_die; die++) {
for (i = 0; i < BITS_TO_U32(irqc->nr_irq); i++)
aic_ic_write(irqc, irqc->info.mask_set + off + i * 4, U32_MAX);
for (i = 0; i < BITS_TO_U32(irqc->nr_irq); i++)
aic_ic_write(irqc, irqc->info.sw_clr + off + i * 4, U32_MAX);
if (irqc->info.target_cpu)
for (i = 0; i < irqc->nr_irq; i++)
aic_ic_write(irqc, irqc->info.target_cpu + off + i * 4, 1);
off += irqc->info.die_stride;
}
if (irqc->info.version == 2) {
u32 config = aic_ic_read(irqc, AIC2_CONFIG);
config |= AIC2_CONFIG_ENABLE;
aic_ic_write(irqc, AIC2_CONFIG, config);
}
if (!is_kernel_in_hyp_mode())
pr_info("Kernel running in EL1, mapping interrupts");
if (static_branch_likely(&use_fast_ipi))
pr_info("Using Fast IPIs");
cpuhp_setup_state(CPUHP_AP_IRQ_APPLE_AIC_STARTING,
"irqchip/apple-aic/ipi:starting",
aic_init_cpu, NULL);
vgic_set_kvm_info(&vgic_info);
pr_info("Initialized with %d IRQs, %d FIQs, %d vIPIs\n",
irqc->nr_hw, AIC_NR_FIQ, AIC_NR_SWIPI);
pr_info("Initialized with %d/%d IRQs * %d/%d die(s), %d FIQs, %d vIPIs",
irqc->nr_irq, irqc->max_irq, irqc->nr_die, irqc->max_die, AIC_NR_FIQ, AIC_NR_SWIPI);
return 0;
err_remove_domain:
irq_domain_remove(irqc->hw_domain);
err_unmap:
if (irqc->event && irqc->event != irqc->base)
iounmap(irqc->event);
iounmap(irqc->base);
kfree(irqc);
return -ENODEV;
}
IRQCHIP_DECLARE(apple_m1_aic, "apple,aic", aic_of_ic_init);
IRQCHIP_DECLARE(apple_aic, "apple,aic", aic_of_ic_init);
IRQCHIP_DECLARE(apple_aic2, "apple,aic2", aic_of_ic_init);

View File

@ -219,7 +219,7 @@ static int apple_msi_domain_alloc(struct irq_domain *domain, unsigned int virq,
if (hwirq < 0)
return -ENOSPC;
fwspec.param[1] += hwirq;
fwspec.param[fwspec.param_count - 2] += hwirq;
ret = irq_domain_alloc_irqs_parent(domain, virq, nr_irqs, &fwspec);
if (ret)