KVM: MIPS: Add Loongson-3 Virtual IPI interrupt support

This patch add Loongson-3 Virtual IPI interrupt support in the kernel.
The current implementation of IPI emulation in QEMU is based on GIC for
MIPS, but Loongson-3 doesn't use GIC. Furthermore, IPI emulation in QEMU
is too expensive for performance (because of too many context switches
between Host and Guest). With current solution, the IPI delay may even
cause RCU stall warnings in a multi-core Guest. So, we design a faster
solution that emulate IPI interrupt in kernel (only used by Loongson-3
now).

Reviewed-by: Aleksandar Markovic <aleksandar.qemu.devel@gmail.com>
Signed-off-by: Huacai Chen <chenhc@lemote.com>
Co-developed-by: Jiaxun Yang <jiaxun.yang@flygoat.com>
Message-Id: <1590220602-3547-11-git-send-email-chenhc@lemote.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

4 files changed, 245 insertions, 1 deletions

diff --git a/arch/mips/kvm/Makefile b/arch/mips/kvm/Makefile
index 0a3cef6153fd..506c4ac0ba1c 100644
--- a/arch/mips/kvm/Makefile
+++ b/arch/mips/kvm/Makefile
@@ -13,6 +13,9 @@ kvm-objs := $(common-objs-y) mips.o emulate.o entry.o \
 	    fpu.o
 kvm-objs += hypcall.o
 kvm-objs += mmu.o
+ifdef CONFIG_CPU_LOONGSON64
+kvm-objs += loongson_ipi.o
+endif
 
 ifdef CONFIG_KVM_MIPS_VZ
 kvm-objs		+= vz.o
diff --git a/arch/mips/kvm/emulate.c b/arch/mips/kvm/emulate.c
index 7ccf9b096783..4fa93d6a2ec5 100644
--- a/arch/mips/kvm/emulate.c
+++ b/arch/mips/kvm/emulate.c
@@ -1600,6 +1600,7 @@ enum emulation_result kvm_mips_emulate_store(union mips_instruction inst,
 					     struct kvm_run *run,
 					     struct kvm_vcpu *vcpu)
 {
+	int r;
 	enum emulation_result er;
 	u32 rt;
 	void *data = run->mmio.data;
@@ -1666,9 +1667,18 @@ enum emulation_result kvm_mips_emulate_store(union mips_instruction inst,
 		goto out_fail;
 	}
 
-	run->mmio.is_write = 1;
 	vcpu->mmio_needed = 1;
+	run->mmio.is_write = 1;
 	vcpu->mmio_is_write = 1;
+
+	r = kvm_io_bus_write(vcpu, KVM_MMIO_BUS,
+			run->mmio.phys_addr, run->mmio.len, data);
+
+	if (!r) {
+		vcpu->mmio_needed = 0;
+		return EMULATE_DONE;
+	}
+
 	return EMULATE_DO_MMIO;
 
 out_fail:
@@ -1681,6 +1691,7 @@ enum emulation_result kvm_mips_emulate_load(union mips_instruction inst,
 					    u32 cause, struct kvm_run *run,
 					    struct kvm_vcpu *vcpu)
 {
+	int r;
 	enum emulation_result er;
 	unsigned long curr_pc;
 	u32 op, rt;
@@ -1745,6 +1756,16 @@ enum emulation_result kvm_mips_emulate_load(union mips_instruction inst,
 
 	run->mmio.is_write = 0;
 	vcpu->mmio_is_write = 0;
+
+	r = kvm_io_bus_read(vcpu, KVM_MMIO_BUS,
+			run->mmio.phys_addr, run->mmio.len, run->mmio.data);
+
+	if (!r) {
+		kvm_mips_complete_mmio_load(vcpu, run);
+		vcpu->mmio_needed = 0;
+		return EMULATE_DONE;
+	}
+
 	return EMULATE_DO_MMIO;
 }
 
diff --git a/arch/mips/kvm/loongson_ipi.c b/arch/mips/kvm/loongson_ipi.c
new file mode 100644
index 000000000000..3681fc8fba38
--- /dev/null
+++ b/arch/mips/kvm/loongson_ipi.c
@@ -0,0 +1,214 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ * Loongson-3 Virtual IPI interrupt support.
+ *
+ * Copyright (C) 2019  Loongson Technologies, Inc.  All rights reserved.
+ *
+ * Authors: Chen Zhu <zhuchen@loongson.cn>
+ * Authors: Huacai Chen <chenhc@lemote.com>
+ */
+
+#include <linux/kvm_host.h>
+
+#define IPI_BASE            0x3ff01000ULL
+
+#define CORE0_STATUS_OFF       0x000
+#define CORE0_EN_OFF           0x004
+#define CORE0_SET_OFF          0x008
+#define CORE0_CLEAR_OFF        0x00c
+#define CORE0_BUF_20           0x020
+#define CORE0_BUF_28           0x028
+#define CORE0_BUF_30           0x030
+#define CORE0_BUF_38           0x038
+
+#define CORE1_STATUS_OFF       0x100
+#define CORE1_EN_OFF           0x104
+#define CORE1_SET_OFF          0x108
+#define CORE1_CLEAR_OFF        0x10c
+#define CORE1_BUF_20           0x120
+#define CORE1_BUF_28           0x128
+#define CORE1_BUF_30           0x130
+#define CORE1_BUF_38           0x138
+
+#define CORE2_STATUS_OFF       0x200
+#define CORE2_EN_OFF           0x204
+#define CORE2_SET_OFF          0x208
+#define CORE2_CLEAR_OFF        0x20c
+#define CORE2_BUF_20           0x220
+#define CORE2_BUF_28           0x228
+#define CORE2_BUF_30           0x230
+#define CORE2_BUF_38           0x238
+
+#define CORE3_STATUS_OFF       0x300
+#define CORE3_EN_OFF           0x304
+#define CORE3_SET_OFF          0x308
+#define CORE3_CLEAR_OFF        0x30c
+#define CORE3_BUF_20           0x320
+#define CORE3_BUF_28           0x328
+#define CORE3_BUF_30           0x330
+#define CORE3_BUF_38           0x338
+
+static int loongson_vipi_read(struct loongson_kvm_ipi *ipi,
+				gpa_t addr, int len, void *val)
+{
+	uint32_t core = (addr >> 8) & 3;
+	uint32_t node = (addr >> 44) & 3;
+	uint32_t id = core + node * 4;
+	uint64_t offset = addr & 0xff;
+	void *pbuf;
+	struct ipi_state *s = &(ipi->ipistate[id]);
+
+	BUG_ON(offset & (len - 1));
+
+	switch (offset) {
+	case CORE0_STATUS_OFF:
+		*(uint64_t *)val = s->status;
+		break;
+
+	case CORE0_EN_OFF:
+		*(uint64_t *)val = s->en;
+		break;
+
+	case CORE0_SET_OFF:
+		*(uint64_t *)val = 0;
+		break;
+
+	case CORE0_CLEAR_OFF:
+		*(uint64_t *)val = 0;
+		break;
+
+	case CORE0_BUF_20 ... CORE0_BUF_38:
+		pbuf = (void *)s->buf + (offset - 0x20);
+		if (len == 8)
+			*(uint64_t *)val = *(uint64_t *)pbuf;
+		else /* Assume len == 4 */
+			*(uint32_t *)val = *(uint32_t *)pbuf;
+		break;
+
+	default:
+		pr_notice("%s with unknown addr %llx\n", __func__, addr);
+		break;
+	}
+
+	return 0;
+}
+
+static int loongson_vipi_write(struct loongson_kvm_ipi *ipi,
+				gpa_t addr, int len, const void *val)
+{
+	uint32_t core = (addr >> 8) & 3;
+	uint32_t node = (addr >> 44) & 3;
+	uint32_t id = core + node * 4;
+	uint64_t data, offset = addr & 0xff;
+	void *pbuf;
+	struct kvm *kvm = ipi->kvm;
+	struct kvm_mips_interrupt irq;
+	struct ipi_state *s = &(ipi->ipistate[id]);
+
+	data = *(uint64_t *)val;
+	BUG_ON(offset & (len - 1));
+
+	switch (offset) {
+	case CORE0_STATUS_OFF:
+		break;
+
+	case CORE0_EN_OFF:
+		s->en = data;
+		break;
+
+	case CORE0_SET_OFF:
+		s->status |= data;
+		irq.cpu = id;
+		irq.irq = 6;
+		kvm_vcpu_ioctl_interrupt(kvm->vcpus[id], &irq);
+		break;
+
+	case CORE0_CLEAR_OFF:
+		s->status &= ~data;
+		if (!s->status) {
+			irq.cpu = id;
+			irq.irq = -6;
+			kvm_vcpu_ioctl_interrupt(kvm->vcpus[id], &irq);
+		}
+		break;
+
+	case CORE0_BUF_20 ... CORE0_BUF_38:
+		pbuf = (void *)s->buf + (offset - 0x20);
+		if (len == 8)
+			*(uint64_t *)pbuf = (uint64_t)data;
+		else /* Assume len == 4 */
+			*(uint32_t *)pbuf = (uint32_t)data;
+		break;
+
+	default:
+		pr_notice("%s with unknown addr %llx\n", __func__, addr);
+		break;
+	}
+
+	return 0;
+}
+
+static int kvm_ipi_read(struct kvm_vcpu *vcpu, struct kvm_io_device *dev,
+			gpa_t addr, int len, void *val)
+{
+	unsigned long flags;
+	struct loongson_kvm_ipi *ipi;
+	struct ipi_io_device *ipi_device;
+
+	ipi_device = container_of(dev, struct ipi_io_device, device);
+	ipi = ipi_device->ipi;
+
+	spin_lock_irqsave(&ipi->lock, flags);
+	loongson_vipi_read(ipi, addr, len, val);
+	spin_unlock_irqrestore(&ipi->lock, flags);
+
+	return 0;
+}
+
+static int kvm_ipi_write(struct kvm_vcpu *vcpu, struct kvm_io_device *dev,
+			gpa_t addr, int len, const void *val)
+{
+	unsigned long flags;
+	struct loongson_kvm_ipi *ipi;
+	struct ipi_io_device *ipi_device;
+
+	ipi_device = container_of(dev, struct ipi_io_device, device);
+	ipi = ipi_device->ipi;
+
+	spin_lock_irqsave(&ipi->lock, flags);
+	loongson_vipi_write(ipi, addr, len, val);
+	spin_unlock_irqrestore(&ipi->lock, flags);
+
+	return 0;
+}
+
+static const struct kvm_io_device_ops kvm_ipi_ops = {
+	.read     = kvm_ipi_read,
+	.write    = kvm_ipi_write,
+};
+
+void kvm_init_loongson_ipi(struct kvm *kvm)
+{
+	int i;
+	unsigned long addr;
+	struct loongson_kvm_ipi *s;
+	struct kvm_io_device *device;
+
+	s = &kvm->arch.ipi;
+	s->kvm = kvm;
+	spin_lock_init(&s->lock);
+
+	/*
+	 * Initialize IPI device
+	 */
+	for (i = 0; i < 4; i++) {
+		device = &s->dev_ipi[i].device;
+		kvm_iodevice_init(device, &kvm_ipi_ops);
+		addr = (((unsigned long)i) << 44) + IPI_BASE;
+		mutex_lock(&kvm->slots_lock);
+		kvm_io_bus_register_dev(kvm, KVM_MMIO_BUS, addr, 0x400, device);
+		mutex_unlock(&kvm->slots_lock);
+		s->dev_ipi[i].ipi = s;
+		s->dev_ipi[i].node_id = i;
+	}
+}
diff --git a/arch/mips/kvm/mips.c b/arch/mips/kvm/mips.c
index 6b435c6e5018..e4d42e03ec37 100644
--- a/arch/mips/kvm/mips.c
+++ b/arch/mips/kvm/mips.c
@@ -129,6 +129,8 @@ int kvm_arch_check_processor_compat(void *opaque)
 	return 0;
 }
 
+extern void kvm_init_loongson_ipi(struct kvm *kvm);
+
 int kvm_arch_init_vm(struct kvm *kvm, unsigned long type)
 {
 	switch (type) {
@@ -148,6 +150,10 @@ int kvm_arch_init_vm(struct kvm *kvm, unsigned long type)
 	if (!kvm->arch.gpa_mm.pgd)
 		return -ENOMEM;
 
+#ifdef CONFIG_CPU_LOONGSON64
+	kvm_init_loongson_ipi(kvm);
+#endif
+
 	return 0;
 }