* [PATCH v3 0/5] mips/kvm: Fix ABI for compatibility with 64-bit guests.
@ 2013-05-20 21:01 David Daney
2013-05-20 21:01 ` [PATCH v3 1/5] mips/kvm: Fix ABI for use of FPU David Daney
` (4 more replies)
0 siblings, 5 replies; 12+ messages in thread
From: David Daney @ 2013-05-20 21:01 UTC (permalink / raw)
To: linux-mips, ralf, kvm, Sanjay Lal; +Cc: linux-kernel, David Daney
From: David Daney <david.daney@cavium.com>
The initial patch set implementing MIPS KVM does not handle 64-bit
guests or use of the FPU. This patch set corrects these ABI issues,
and does some very minor clean up.
Changes from v2: Split into five parts, no code change.
David Daney (5):
mips/kvm: Fix ABI for use of FPU.
mips/kvm: Fix ABI for use of 64-bit registers.
mips/kvm: Fix name of gpr field in struct kvm_regs.
mips/kvm: Use ARRAY_SIZE() instead of hardcoded constants in
kvm_arch_vcpu_ioctl_{s,g}et_regs
mips/kvm: Fix ABI by moving manipulation of CP0 registers to
KVM_{G,S}ET_MSRS
arch/mips/include/asm/kvm.h | 111 ++++++++++---
arch/mips/include/asm/kvm_host.h | 4 -
arch/mips/kvm/kvm_mips.c | 102 +-----------
arch/mips/kvm/kvm_trap_emul.c | 330 ++++++++++++++++++++++++++++++++++-----
4 files changed, 386 insertions(+), 161 deletions(-)
--
1.7.11.7
^ permalink raw reply [flat|nested] 12+ messages in thread
* [PATCH v3 1/5] mips/kvm: Fix ABI for use of FPU.
2013-05-20 21:01 [PATCH v3 0/5] mips/kvm: Fix ABI for compatibility with 64-bit guests David Daney
@ 2013-05-20 21:01 ` David Daney
2013-05-20 21:01 ` [PATCH v3 2/5] mips/kvm: Fix ABI for use of 64-bit registers David Daney
` (3 subsequent siblings)
4 siblings, 0 replies; 12+ messages in thread
From: David Daney @ 2013-05-20 21:01 UTC (permalink / raw)
To: linux-mips, ralf, kvm, Sanjay Lal; +Cc: linux-kernel, David Daney
From: David Daney <david.daney@cavium.com>
Define a non-empty struct kvm_fpu.
Signed-off-by: David Daney <david.daney@cavium.com>
---
arch/mips/include/asm/kvm.h | 29 +++++++++++++++++++++--------
1 file changed, 21 insertions(+), 8 deletions(-)
diff --git a/arch/mips/include/asm/kvm.h b/arch/mips/include/asm/kvm.h
index 85789ea..0e8f565 100644
--- a/arch/mips/include/asm/kvm.h
+++ b/arch/mips/include/asm/kvm.h
@@ -1,11 +1,12 @@
/*
-* This file is subject to the terms and conditions of the GNU General Public
-* License. See the file "COPYING" in the main directory of this archive
-* for more details.
-*
-* Copyright (C) 2012 MIPS Technologies, Inc. All rights reserved.
-* Authors: Sanjay Lal <sanjayl@kymasys.com>
-*/
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License. See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2012 MIPS Technologies, Inc. All rights reserved.
+ * Copyright (C) 2013 Cavium, Inc.
+ * Authors: Sanjay Lal <sanjayl@kymasys.com>
+ */
#ifndef __LINUX_KVM_MIPS_H
#define __LINUX_KVM_MIPS_H
@@ -31,8 +32,20 @@ struct kvm_regs {
struct kvm_sregs {
};
-/* for KVM_GET_FPU and KVM_SET_FPU */
+/*
+ * for KVM_GET_FPU and KVM_SET_FPU
+ *
+ * If Status[FR] is zero (32-bit FPU), the upper 32-bits of the FPRs
+ * are zero filled.
+ */
struct kvm_fpu {
+ __u64 fpr[32];
+ __u32 fir;
+ __u32 fccr;
+ __u32 fexr;
+ __u32 fenr;
+ __u32 fcsr;
+ __u32 pad;
};
struct kvm_debug_exit_arch {
--
1.7.11.7
^ permalink raw reply related [flat|nested] 12+ messages in thread
* [PATCH v3 2/5] mips/kvm: Fix ABI for use of 64-bit registers.
2013-05-20 21:01 [PATCH v3 0/5] mips/kvm: Fix ABI for compatibility with 64-bit guests David Daney
2013-05-20 21:01 ` [PATCH v3 1/5] mips/kvm: Fix ABI for use of FPU David Daney
@ 2013-05-20 21:01 ` David Daney
2013-05-20 21:01 ` [PATCH v3 3/5] mips/kvm: Fix name of gpr field in struct kvm_regs David Daney
` (2 subsequent siblings)
4 siblings, 0 replies; 12+ messages in thread
From: David Daney @ 2013-05-20 21:01 UTC (permalink / raw)
To: linux-mips, ralf, kvm, Sanjay Lal; +Cc: linux-kernel, David Daney
From: David Daney <david.daney@cavium.com>
All registers are 64-bits wide, 32-bit guests use the least
significant portion of the register storage fields.
Signed-off-by: David Daney <david.daney@cavium.com>
---
arch/mips/include/asm/kvm.h | 16 +++++++++++-----
1 file changed, 11 insertions(+), 5 deletions(-)
diff --git a/arch/mips/include/asm/kvm.h b/arch/mips/include/asm/kvm.h
index 0e8f565..86812fb 100644
--- a/arch/mips/include/asm/kvm.h
+++ b/arch/mips/include/asm/kvm.h
@@ -18,12 +18,18 @@
#define N_MIPS_COPROC_REGS 32
#define N_MIPS_COPROC_SEL 8
-/* for KVM_GET_REGS and KVM_SET_REGS */
+/*
+ * for KVM_GET_REGS and KVM_SET_REGS
+ *
+ * If Config[AT] is zero (32-bit CPU), the register contents are
+ * stored in the lower 32-bits of the struct kvm_regs fields and sign
+ * extended to 64-bits.
+ */
struct kvm_regs {
- __u32 gprs[32];
- __u32 hi;
- __u32 lo;
- __u32 pc;
+ __u64 gprs[32];
+ __u64 hi;
+ __u64 lo;
+ __u64 pc;
__u32 cp0reg[N_MIPS_COPROC_REGS][N_MIPS_COPROC_SEL];
};
--
1.7.11.7
^ permalink raw reply related [flat|nested] 12+ messages in thread
* [PATCH v3 3/5] mips/kvm: Fix name of gpr field in struct kvm_regs.
2013-05-20 21:01 [PATCH v3 0/5] mips/kvm: Fix ABI for compatibility with 64-bit guests David Daney
2013-05-20 21:01 ` [PATCH v3 1/5] mips/kvm: Fix ABI for use of FPU David Daney
2013-05-20 21:01 ` [PATCH v3 2/5] mips/kvm: Fix ABI for use of 64-bit registers David Daney
@ 2013-05-20 21:01 ` David Daney
2013-05-20 21:01 ` [PATCH v3 4/5] mips/kvm: Use ARRAY_SIZE() instead of hardcoded constants in kvm_arch_vcpu_ioctl_{s,g}et_regs David Daney
2013-05-20 21:01 ` [PATCH v3 5/5] mips/kvm: Fix ABI by moving manipulation of CP0 registers to KVM_{G,S}ET_MSRS David Daney
4 siblings, 0 replies; 12+ messages in thread
From: David Daney @ 2013-05-20 21:01 UTC (permalink / raw)
To: linux-mips, ralf, kvm, Sanjay Lal; +Cc: linux-kernel, David Daney
From: David Daney <david.daney@cavium.com>
Signed-off-by: David Daney <david.daney@cavium.com>
---
arch/mips/include/asm/kvm.h | 3 ++-
arch/mips/kvm/kvm_mips.c | 4 ++--
2 files changed, 4 insertions(+), 3 deletions(-)
diff --git a/arch/mips/include/asm/kvm.h b/arch/mips/include/asm/kvm.h
index 86812fb..d145ead 100644
--- a/arch/mips/include/asm/kvm.h
+++ b/arch/mips/include/asm/kvm.h
@@ -26,7 +26,8 @@
* extended to 64-bits.
*/
struct kvm_regs {
- __u64 gprs[32];
+ /* out (KVM_GET_REGS) / in (KVM_SET_REGS) */
+ __u64 gpr[32];
__u64 hi;
__u64 lo;
__u64 pc;
diff --git a/arch/mips/kvm/kvm_mips.c b/arch/mips/kvm/kvm_mips.c
index e0dad02..93da750 100644
--- a/arch/mips/kvm/kvm_mips.c
+++ b/arch/mips/kvm/kvm_mips.c
@@ -678,7 +678,7 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
int i;
for (i = 0; i < 32; i++)
- vcpu->arch.gprs[i] = regs->gprs[i];
+ vcpu->arch.gprs[i] = regs->gpr[i];
vcpu->arch.hi = regs->hi;
vcpu->arch.lo = regs->lo;
@@ -692,7 +692,7 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
int i;
for (i = 0; i < 32; i++)
- regs->gprs[i] = vcpu->arch.gprs[i];
+ regs->gpr[i] = vcpu->arch.gprs[i];
regs->hi = vcpu->arch.hi;
regs->lo = vcpu->arch.lo;
--
1.7.11.7
^ permalink raw reply related [flat|nested] 12+ messages in thread
* [PATCH v3 4/5] mips/kvm: Use ARRAY_SIZE() instead of hardcoded constants in kvm_arch_vcpu_ioctl_{s,g}et_regs
2013-05-20 21:01 [PATCH v3 0/5] mips/kvm: Fix ABI for compatibility with 64-bit guests David Daney
` (2 preceding siblings ...)
2013-05-20 21:01 ` [PATCH v3 3/5] mips/kvm: Fix name of gpr field in struct kvm_regs David Daney
@ 2013-05-20 21:01 ` David Daney
2013-05-20 21:01 ` [PATCH v3 5/5] mips/kvm: Fix ABI by moving manipulation of CP0 registers to KVM_{G,S}ET_MSRS David Daney
4 siblings, 0 replies; 12+ messages in thread
From: David Daney @ 2013-05-20 21:01 UTC (permalink / raw)
To: linux-mips, ralf, kvm, Sanjay Lal; +Cc: linux-kernel, David Daney
From: David Daney <david.daney@cavium.com>
Also we cannot set special zero register, so force it to zero.
Signed-off-by: David Daney <david.daney@cavium.com>
---
arch/mips/kvm/kvm_mips.c | 6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)
diff --git a/arch/mips/kvm/kvm_mips.c b/arch/mips/kvm/kvm_mips.c
index 93da750..71a1fc1 100644
--- a/arch/mips/kvm/kvm_mips.c
+++ b/arch/mips/kvm/kvm_mips.c
@@ -677,9 +677,9 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
{
int i;
- for (i = 0; i < 32; i++)
+ for (i = 1; i < ARRAY_SIZE(vcpu->arch.gprs); i++)
vcpu->arch.gprs[i] = regs->gpr[i];
-
+ vcpu->arch.gprs[0] = 0; /* zero is special, and cannot be set. */
vcpu->arch.hi = regs->hi;
vcpu->arch.lo = regs->lo;
vcpu->arch.pc = regs->pc;
@@ -691,7 +691,7 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
{
int i;
- for (i = 0; i < 32; i++)
+ for (i = 0; i < ARRAY_SIZE(vcpu->arch.gprs); i++)
regs->gpr[i] = vcpu->arch.gprs[i];
regs->hi = vcpu->arch.hi;
--
1.7.11.7
^ permalink raw reply related [flat|nested] 12+ messages in thread
* [PATCH v3 5/5] mips/kvm: Fix ABI by moving manipulation of CP0 registers to KVM_{G,S}ET_MSRS
2013-05-20 21:01 [PATCH v3 0/5] mips/kvm: Fix ABI for compatibility with 64-bit guests David Daney
` (3 preceding siblings ...)
2013-05-20 21:01 ` [PATCH v3 4/5] mips/kvm: Use ARRAY_SIZE() instead of hardcoded constants in kvm_arch_vcpu_ioctl_{s,g}et_regs David Daney
@ 2013-05-20 21:01 ` David Daney
2013-05-21 15:37 ` Gleb Natapov
4 siblings, 1 reply; 12+ messages in thread
From: David Daney @ 2013-05-20 21:01 UTC (permalink / raw)
To: linux-mips, ralf, kvm, Sanjay Lal; +Cc: linux-kernel, David Daney
From: David Daney <david.daney@cavium.com>
Because not all 256 CP0 registers are ever implemented, we need a
different method of manipulating them. Use the
KVM_GET_MSRS/KVM_SET_MSRS mechanism as x86 does for its MSRs.
Code related to implementing KVM_GET_MSRS/KVM_SET_MSRS is consolidated
in to kvm_trap_emul.c, now unused code and definitions are removed.
Signed-off-by: David Daney <david.daney@cavium.com>
---
arch/mips/include/asm/kvm.h | 69 ++++++--
arch/mips/include/asm/kvm_host.h | 4 -
arch/mips/kvm/kvm_mips.c | 92 +----------
arch/mips/kvm/kvm_trap_emul.c | 330 ++++++++++++++++++++++++++++++++++-----
4 files changed, 350 insertions(+), 145 deletions(-)
diff --git a/arch/mips/include/asm/kvm.h b/arch/mips/include/asm/kvm.h
index d145ead..cae080c 100644
--- a/arch/mips/include/asm/kvm.h
+++ b/arch/mips/include/asm/kvm.h
@@ -13,10 +13,11 @@
#include <linux/types.h>
-#define __KVM_MIPS
-
-#define N_MIPS_COPROC_REGS 32
-#define N_MIPS_COPROC_SEL 8
+/*
+ * KVM MIPS specific structures and definitions.
+ *
+ * Some parts derived from the x86 version of this file.
+ */
/*
* for KVM_GET_REGS and KVM_SET_REGS
@@ -31,12 +32,6 @@ struct kvm_regs {
__u64 hi;
__u64 lo;
__u64 pc;
-
- __u32 cp0reg[N_MIPS_COPROC_REGS][N_MIPS_COPROC_SEL];
-};
-
-/* for KVM_GET_SREGS and KVM_SET_SREGS */
-struct kvm_sregs {
};
/*
@@ -55,21 +50,67 @@ struct kvm_fpu {
__u32 pad;
};
+
+/*
+ * For MIPS, we use the same APIs as x86, where 'msr' corresponds to a
+ * CP0 register. The index field is broken down as follows:
+ *
+ * bits[2..0] - Register 'sel' index.
+ * bits[7..3] - Register 'rd' index.
+ * bits[15..8] - Must be zero.
+ * bits[31..16] - 0 -> CP0 registers.
+ *
+ * Other sets registers may be added in the future. Each set would
+ * have its own identifier in bits[31..16].
+ *
+ * For MSRs that are narrower than 64-bits, the value is stored in the
+ * low order bits of the data field, and sign extended to 64-bits.
+ */
+#define KVM_MIPS_MSR_CP0 0
+struct kvm_msr_entry {
+ __u32 index;
+ __u32 reserved;
+ __u64 data;
+};
+
+/* for KVM_GET_MSRS and KVM_SET_MSRS */
+struct kvm_msrs {
+ __u32 nmsrs; /* number of msrs in entries */
+ __u32 pad;
+
+ struct kvm_msr_entry entries[0];
+};
+
+/* for KVM_GET_MSR_INDEX_LIST */
+struct kvm_msr_list {
+ __u32 nmsrs; /* number of msrs in entries */
+ __u32 indices[0];
+};
+
+/*
+ * KVM MIPS specific structures and definitions
+ *
+ */
struct kvm_debug_exit_arch {
+ __u64 epc;
};
/* for KVM_SET_GUEST_DEBUG */
struct kvm_guest_debug_arch {
};
+/* definition of registers in kvm_run */
+struct kvm_sync_regs {
+};
+
+/* dummy definition */
+struct kvm_sregs {
+};
+
struct kvm_mips_interrupt {
/* in */
__u32 cpu;
__u32 irq;
};
-/* definition of registers in kvm_run */
-struct kvm_sync_regs {
-};
-
#endif /* __LINUX_KVM_MIPS_H */
diff --git a/arch/mips/include/asm/kvm_host.h b/arch/mips/include/asm/kvm_host.h
index 143875c..4d6fa0b 100644
--- a/arch/mips/include/asm/kvm_host.h
+++ b/arch/mips/include/asm/kvm_host.h
@@ -496,10 +496,6 @@ struct kvm_mips_callbacks {
uint32_t cause);
int (*irq_clear) (struct kvm_vcpu *vcpu, unsigned int priority,
uint32_t cause);
- int (*vcpu_ioctl_get_regs) (struct kvm_vcpu *vcpu,
- struct kvm_regs *regs);
- int (*vcpu_ioctl_set_regs) (struct kvm_vcpu *vcpu,
- struct kvm_regs *regs);
};
extern struct kvm_mips_callbacks *kvm_mips_callbacks;
int kvm_mips_emulation_init(struct kvm_mips_callbacks **install_callbacks);
diff --git a/arch/mips/kvm/kvm_mips.c b/arch/mips/kvm/kvm_mips.c
index 71a1fc1..26a40e3 100644
--- a/arch/mips/kvm/kvm_mips.c
+++ b/arch/mips/kvm/kvm_mips.c
@@ -51,16 +51,6 @@ struct kvm_stats_debugfs_item debugfs_entries[] = {
{NULL}
};
-static int kvm_mips_reset_vcpu(struct kvm_vcpu *vcpu)
-{
- int i;
- for_each_possible_cpu(i) {
- vcpu->arch.guest_kernel_asid[i] = 0;
- vcpu->arch.guest_user_asid[i] = 0;
- }
- return 0;
-}
-
gfn_t unalias_gfn(struct kvm *kvm, gfn_t gfn)
{
return gfn;
@@ -192,12 +182,6 @@ void kvm_arch_destroy_vm(struct kvm *kvm)
}
}
-long
-kvm_arch_dev_ioctl(struct file *filp, unsigned int ioctl, unsigned long arg)
-{
- return -EINVAL;
-}
-
void kvm_arch_free_memslot(struct kvm_memory_slot *free,
struct kvm_memory_slot *dont)
{
@@ -435,42 +419,6 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run)
return r;
}
-
-int
-kvm_vcpu_ioctl_interrupt(struct kvm_vcpu *vcpu, struct kvm_mips_interrupt *irq)
-{
- int intr = (int)irq->irq;
- struct kvm_vcpu *dvcpu = NULL;
-
- if (intr == 3 || intr == -3 || intr == 4 || intr == -4)
- kvm_debug("%s: CPU: %d, INTR: %d\n", __func__, irq->cpu,
- (int)intr);
-
- if (irq->cpu == -1)
- dvcpu = vcpu;
- else
- dvcpu = vcpu->kvm->vcpus[irq->cpu];
-
- if (intr == 2 || intr == 3 || intr == 4) {
- kvm_mips_callbacks->queue_io_int(dvcpu, irq);
-
- } else if (intr == -2 || intr == -3 || intr == -4) {
- kvm_mips_callbacks->dequeue_io_int(dvcpu, irq);
- } else {
- kvm_err("%s: invalid interrupt ioctl (%d:%d)\n", __func__,
- irq->cpu, irq->irq);
- return -EINVAL;
- }
-
- dvcpu->arch.wait = 0;
-
- if (waitqueue_active(&dvcpu->wq)) {
- wake_up_interruptible(&dvcpu->wq);
- }
-
- return 0;
-}
-
int
kvm_arch_vcpu_ioctl_get_mpstate(struct kvm_vcpu *vcpu,
struct kvm_mp_state *mp_state)
@@ -485,42 +433,6 @@ kvm_arch_vcpu_ioctl_set_mpstate(struct kvm_vcpu *vcpu,
return -EINVAL;
}
-long
-kvm_arch_vcpu_ioctl(struct file *filp, unsigned int ioctl, unsigned long arg)
-{
- struct kvm_vcpu *vcpu = filp->private_data;
- void __user *argp = (void __user *)arg;
- long r;
- int intr;
-
- switch (ioctl) {
- case KVM_NMI:
- /* Treat the NMI as a CPU reset */
- r = kvm_mips_reset_vcpu(vcpu);
- break;
- case KVM_INTERRUPT:
- {
- struct kvm_mips_interrupt irq;
- r = -EFAULT;
- if (copy_from_user(&irq, argp, sizeof(irq)))
- goto out;
-
- intr = (int)irq.irq;
-
- kvm_debug("[%d] %s: irq: %d\n", vcpu->vcpu_id, __func__,
- irq.irq);
-
- r = kvm_vcpu_ioctl_interrupt(vcpu, &irq);
- break;
- }
- default:
- r = -EINVAL;
- }
-
-out:
- return r;
-}
-
/*
* Get (and clear) the dirty memory log for a memory slot.
*/
@@ -684,7 +596,7 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
vcpu->arch.lo = regs->lo;
vcpu->arch.pc = regs->pc;
- return kvm_mips_callbacks->vcpu_ioctl_set_regs(vcpu, regs);
+ return 0;
}
int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
@@ -698,7 +610,7 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
regs->lo = vcpu->arch.lo;
regs->pc = vcpu->arch.pc;
- return kvm_mips_callbacks->vcpu_ioctl_get_regs(vcpu, regs);
+ return 0;
}
void kvm_mips_comparecount_func(unsigned long data)
diff --git a/arch/mips/kvm/kvm_trap_emul.c b/arch/mips/kvm/kvm_trap_emul.c
index 466aeef..df98dcb 100644
--- a/arch/mips/kvm/kvm_trap_emul.c
+++ b/arch/mips/kvm/kvm_trap_emul.c
@@ -13,7 +13,7 @@
#include <linux/err.h>
#include <linux/module.h>
#include <linux/vmalloc.h>
-
+#include <linux/fs.h>
#include <linux/kvm_host.h>
#include "kvm_mips_opcode.h"
@@ -345,54 +345,312 @@ static int kvm_trap_emul_handle_break(struct kvm_vcpu *vcpu)
return ret;
}
-static int
-kvm_trap_emul_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
+#define MSR_MIPS_CP0_INDEX (8 * 0 + 0)
+#define MSR_MIPS_CP0_ENTRYLO0 (8 * 2 + 0)
+#define MSR_MIPS_CP0_ENTRYLO1 (8 * 3 + 0)
+#define MSR_MIPS_CP0_CONTEXT (8 * 4 + 0)
+#define MSR_MIPS_CP0_USERLOCAL (8 * 4 + 2)
+#define MSR_MIPS_CP0_PAGEMASK (8 * 5 + 0)
+#define MSR_MIPS_CP0_PAGEGRAIN (8 * 5 + 1)
+#define MSR_MIPS_CP0_WIRED (8 * 6 + 0)
+#define MSR_MIPS_CP0_HWRENA (8 * 7 + 0)
+#define MSR_MIPS_CP0_BADVADDR (8 * 8 + 0)
+#define MSR_MIPS_CP0_COUNT (8 * 9 + 0)
+#define MSR_MIPS_CP0_ENTRYHI (8 * 10 + 0)
+#define MSR_MIPS_CP0_COMPARE (8 * 11 + 0)
+#define MSR_MIPS_CP0_STATUS (8 * 12 + 0)
+#define MSR_MIPS_CP0_CAUSE (8 * 13 + 0)
+#define MSR_MIPS_CP0_EBASE (8 * 15 + 1)
+#define MSR_MIPS_CP0_CONFIG (8 * 16 + 0)
+#define MSR_MIPS_CP0_CONFIG1 (8 * 16 + 1)
+#define MSR_MIPS_CP0_CONFIG2 (8 * 16 + 2)
+#define MSR_MIPS_CP0_CONFIG3 (8 * 16 + 3)
+#define MSR_MIPS_CP0_CONFIG7 (8 * 16 + 7)
+#define MSR_MIPS_CP0_XCONTEXT (8 * 20 + 0)
+#define MSR_MIPS_CP0_ERROREPC (8 * 30 + 0)
+
+static u32 msrs_to_save[] = {
+ MSR_MIPS_CP0_INDEX,
+ MSR_MIPS_CP0_CONTEXT,
+ MSR_MIPS_CP0_PAGEMASK,
+ MSR_MIPS_CP0_WIRED,
+ MSR_MIPS_CP0_BADVADDR,
+ MSR_MIPS_CP0_ENTRYHI,
+ MSR_MIPS_CP0_STATUS,
+ MSR_MIPS_CP0_CAUSE,
+ /* EPC set via kvm_regs, et al. */
+ MSR_MIPS_CP0_CONFIG,
+ MSR_MIPS_CP0_CONFIG1,
+ MSR_MIPS_CP0_CONFIG2,
+ MSR_MIPS_CP0_CONFIG3,
+ MSR_MIPS_CP0_CONFIG7,
+ MSR_MIPS_CP0_XCONTEXT,
+ MSR_MIPS_CP0_ERROREPC
+};
+
+#define MAX_IO_MSRS 128
+
+static int mipsvz_get_msr(struct kvm_vcpu *vcpu, unsigned index, u64 *data)
{
struct mips_coproc *cop0 = vcpu->arch.cop0;
+ s64 v;
- kvm_write_c0_guest_index(cop0, regs->cp0reg[MIPS_CP0_TLB_INDEX][0]);
- kvm_write_c0_guest_context(cop0, regs->cp0reg[MIPS_CP0_TLB_CONTEXT][0]);
- kvm_write_c0_guest_badvaddr(cop0, regs->cp0reg[MIPS_CP0_BAD_VADDR][0]);
- kvm_write_c0_guest_entryhi(cop0, regs->cp0reg[MIPS_CP0_TLB_HI][0]);
- kvm_write_c0_guest_epc(cop0, regs->cp0reg[MIPS_CP0_EXC_PC][0]);
+ switch (index) {
+ case MSR_MIPS_CP0_INDEX:
+ v = (long)kvm_read_c0_guest_index(cop0);
+ break;
+ case MSR_MIPS_CP0_CONTEXT:
+ v = (long)kvm_read_c0_guest_context(cop0);
+ break;
+ case MSR_MIPS_CP0_PAGEMASK:
+ v = (long)kvm_read_c0_guest_pagemask(cop0);
+ break;
+ case MSR_MIPS_CP0_WIRED:
+ v = (long)kvm_read_c0_guest_wired(cop0);
+ break;
+ case MSR_MIPS_CP0_BADVADDR:
+ v = (long)kvm_read_c0_guest_badvaddr(cop0);
+ break;
+ case MSR_MIPS_CP0_ENTRYHI:
+ v = (long)kvm_read_c0_guest_entryhi(cop0);
+ break;
+ case MSR_MIPS_CP0_STATUS:
+ v = (long)kvm_read_c0_guest_status(cop0);
+ break;
+ case MSR_MIPS_CP0_CAUSE:
+ v = (long)kvm_read_c0_guest_cause(cop0);
+ break;
+ case MSR_MIPS_CP0_ERROREPC:
+ v = (long)kvm_read_c0_guest_errorepc(cop0);
+ break;
+ case MSR_MIPS_CP0_CONFIG:
+ v = (long)kvm_read_c0_guest_config(cop0);
+ break;
+ case MSR_MIPS_CP0_CONFIG1:
+ v = (long)kvm_read_c0_guest_config1(cop0);
+ break;
+ case MSR_MIPS_CP0_CONFIG2:
+ v = (long)kvm_read_c0_guest_config2(cop0);
+ break;
+ case MSR_MIPS_CP0_CONFIG3:
+ v = (long)kvm_read_c0_guest_config3(cop0);
+ break;
+ case MSR_MIPS_CP0_CONFIG7:
+ v = (long)kvm_read_c0_guest_config7(cop0);
+ break;
+ default:
+ return -EINVAL;
+ }
+ *data = v;
+ return 0;
+}
- kvm_write_c0_guest_status(cop0, regs->cp0reg[MIPS_CP0_STATUS][0]);
- kvm_write_c0_guest_cause(cop0, regs->cp0reg[MIPS_CP0_CAUSE][0]);
- kvm_write_c0_guest_pagemask(cop0,
- regs->cp0reg[MIPS_CP0_TLB_PG_MASK][0]);
- kvm_write_c0_guest_wired(cop0, regs->cp0reg[MIPS_CP0_TLB_WIRED][0]);
- kvm_write_c0_guest_errorepc(cop0, regs->cp0reg[MIPS_CP0_ERROR_PC][0]);
+static int mipsvz_set_msr(struct kvm_vcpu *vcpu, unsigned index, u64 *data)
+{
+ struct mips_coproc *cop0 = vcpu->arch.cop0;
+ u64 v = *data;
+ switch (index) {
+ case MSR_MIPS_CP0_INDEX:
+ kvm_write_c0_guest_index(cop0, v);
+ break;
+ case MSR_MIPS_CP0_CONTEXT:
+ kvm_write_c0_guest_context(cop0, v);
+ break;
+ case MSR_MIPS_CP0_PAGEMASK:
+ kvm_write_c0_guest_pagemask(cop0, v);
+ break;
+ case MSR_MIPS_CP0_WIRED:
+ kvm_write_c0_guest_wired(cop0, v);
+ break;
+ case MSR_MIPS_CP0_BADVADDR:
+ kvm_write_c0_guest_badvaddr(cop0, v);
+ break;
+ case MSR_MIPS_CP0_ENTRYHI:
+ kvm_write_c0_guest_entryhi(cop0, v);
+ break;
+ case MSR_MIPS_CP0_STATUS:
+ kvm_write_c0_guest_status(cop0, v);
+ break;
+ case MSR_MIPS_CP0_CAUSE:
+ kvm_write_c0_guest_cause(cop0, v);
+ break;
+ case MSR_MIPS_CP0_ERROREPC:
+ kvm_write_c0_guest_errorepc(cop0, v);
+ break;
+ default:
+ return -EINVAL;
+ }
return 0;
}
-static int
-kvm_trap_emul_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
+/*
+ * Read or write a bunch of msrs. Parameters are user addresses.
+ *
+ * @return number of msrs set successfully.
+ */
+static int msr_io(struct kvm_vcpu *vcpu, struct kvm_msrs __user *user_msrs,
+ int (*do_msr)(struct kvm_vcpu *vcpu,
+ unsigned index, u64 *data),
+ bool writeback)
{
- struct mips_coproc *cop0 = vcpu->arch.cop0;
+ struct kvm_msrs msrs;
+ struct kvm_msr_entry *entries;
+ int r, n;
+ unsigned size;
+
+ r = -EFAULT;
+ if (copy_from_user(&msrs, user_msrs, sizeof(msrs)))
+ goto out;
+
+ r = -E2BIG;
+ if (msrs.nmsrs >= MAX_IO_MSRS)
+ goto out;
+
+ size = sizeof(struct kvm_msr_entry) * msrs.nmsrs;
+ entries = memdup_user(user_msrs->entries, size);
+ if (IS_ERR(entries)) {
+ r = PTR_ERR(entries);
+ goto out;
+ }
- regs->cp0reg[MIPS_CP0_TLB_INDEX][0] = kvm_read_c0_guest_index(cop0);
- regs->cp0reg[MIPS_CP0_TLB_CONTEXT][0] = kvm_read_c0_guest_context(cop0);
- regs->cp0reg[MIPS_CP0_BAD_VADDR][0] = kvm_read_c0_guest_badvaddr(cop0);
- regs->cp0reg[MIPS_CP0_TLB_HI][0] = kvm_read_c0_guest_entryhi(cop0);
- regs->cp0reg[MIPS_CP0_EXC_PC][0] = kvm_read_c0_guest_epc(cop0);
-
- regs->cp0reg[MIPS_CP0_STATUS][0] = kvm_read_c0_guest_status(cop0);
- regs->cp0reg[MIPS_CP0_CAUSE][0] = kvm_read_c0_guest_cause(cop0);
- regs->cp0reg[MIPS_CP0_TLB_PG_MASK][0] =
- kvm_read_c0_guest_pagemask(cop0);
- regs->cp0reg[MIPS_CP0_TLB_WIRED][0] = kvm_read_c0_guest_wired(cop0);
- regs->cp0reg[MIPS_CP0_ERROR_PC][0] = kvm_read_c0_guest_errorepc(cop0);
-
- regs->cp0reg[MIPS_CP0_CONFIG][0] = kvm_read_c0_guest_config(cop0);
- regs->cp0reg[MIPS_CP0_CONFIG][1] = kvm_read_c0_guest_config1(cop0);
- regs->cp0reg[MIPS_CP0_CONFIG][2] = kvm_read_c0_guest_config2(cop0);
- regs->cp0reg[MIPS_CP0_CONFIG][3] = kvm_read_c0_guest_config3(cop0);
- regs->cp0reg[MIPS_CP0_CONFIG][7] = kvm_read_c0_guest_config7(cop0);
+ for (n = 0; n < msrs.nmsrs; ++n)
+ if (do_msr(vcpu, entries[n].index, &entries[n].data))
+ break;
+
+ r = -EFAULT;
+ if (writeback && copy_to_user(user_msrs->entries, entries, size))
+ goto out_free;
+
+ r = n;
+
+out_free:
+ kfree(entries);
+out:
+ return r;
+}
+
+static int kvm_mips_reset_vcpu(struct kvm_vcpu *vcpu)
+{
+ int i;
+ for_each_possible_cpu(i) {
+ vcpu->arch.guest_kernel_asid[i] = 0;
+ vcpu->arch.guest_user_asid[i] = 0;
+ }
+ return 0;
+}
+
+int
+kvm_vcpu_ioctl_interrupt(struct kvm_vcpu *vcpu, struct kvm_mips_interrupt *irq)
+{
+ int intr = (int)irq->irq;
+ struct kvm_vcpu *dvcpu = NULL;
+
+ if (intr == 3 || intr == -3 || intr == 4 || intr == -4)
+ kvm_debug("%s: CPU: %d, INTR: %d\n", __func__, irq->cpu,
+ (int)intr);
+
+ if (irq->cpu == -1)
+ dvcpu = vcpu;
+ else
+ dvcpu = vcpu->kvm->vcpus[irq->cpu];
+
+ if (intr == 2 || intr == 3 || intr == 4) {
+ kvm_mips_callbacks->queue_io_int(dvcpu, irq);
+
+ } else if (intr == -2 || intr == -3 || intr == -4) {
+ kvm_mips_callbacks->dequeue_io_int(dvcpu, irq);
+ } else {
+ kvm_err("%s: invalid interrupt ioctl (%d:%d)\n", __func__,
+ irq->cpu, irq->irq);
+ return -EINVAL;
+ }
+
+ dvcpu->arch.wait = 0;
+
+ if (waitqueue_active(&dvcpu->wq))
+ wake_up_interruptible(&dvcpu->wq);
return 0;
}
+long
+kvm_arch_vcpu_ioctl(struct file *filp, unsigned int ioctl, unsigned long arg)
+{
+ struct kvm_vcpu *vcpu = filp->private_data;
+ void __user *argp = (void __user *)arg;
+ long r;
+
+ switch (ioctl) {
+ case KVM_GET_MSRS:
+ r = msr_io(vcpu, argp, mipsvz_get_msr, true);
+ break;
+ case KVM_SET_MSRS:
+ r = msr_io(vcpu, argp, mipsvz_set_msr, false);
+ break;
+ case KVM_NMI:
+ /* Treat the NMI as a CPU reset */
+ r = kvm_mips_reset_vcpu(vcpu);
+ break;
+ case KVM_INTERRUPT:
+ {
+ struct kvm_mips_interrupt irq;
+ r = -EFAULT;
+ if (copy_from_user(&irq, argp, sizeof(irq)))
+ goto out;
+
+ kvm_debug("[%d] %s: irq: %d\n", vcpu->vcpu_id, __func__,
+ irq.irq);
+
+ r = kvm_vcpu_ioctl_interrupt(vcpu, &irq);
+ break;
+ }
+ default:
+ r = -ENOIOCTLCMD;
+ }
+
+out:
+ return r;
+}
+
+long kvm_arch_dev_ioctl(struct file *filp,
+ unsigned int ioctl, unsigned long arg)
+{
+ long r;
+ void __user *argp = (void __user *)arg;
+
+ switch (ioctl) {
+ case KVM_GET_MSR_INDEX_LIST: {
+ struct kvm_msr_list __user *user_msr_list = argp;
+ struct kvm_msr_list msr_list;
+ unsigned n;
+ unsigned num_msrs_to_save;
+
+ r = -EFAULT;
+ if (copy_from_user(&msr_list, user_msr_list, sizeof(msr_list)))
+ goto out;
+ n = msr_list.nmsrs;
+ msr_list.nmsrs = ARRAY_SIZE(msrs_to_save);
+ num_msrs_to_save = min(n, msr_list.nmsrs);
+ if (copy_to_user(user_msr_list, &msr_list, sizeof(msr_list)))
+ goto out;
+ r = -E2BIG;
+ if (n < msr_list.nmsrs)
+ goto out;
+ r = -EFAULT;
+ if (copy_to_user(user_msr_list->indices, &msrs_to_save,
+ num_msrs_to_save * sizeof(u32)))
+ goto out;
+ r = 0;
+ break;
+ }
+ default:
+ r = -ENOIOCTLCMD;
+ }
+out:
+ return r;
+}
+
static int kvm_trap_emul_vm_init(struct kvm *kvm)
{
return 0;
@@ -471,8 +729,6 @@ static struct kvm_mips_callbacks kvm_trap_emul_callbacks = {
.dequeue_io_int = kvm_mips_dequeue_io_int_cb,
.irq_deliver = kvm_mips_irq_deliver_cb,
.irq_clear = kvm_mips_irq_clear_cb,
- .vcpu_ioctl_get_regs = kvm_trap_emul_ioctl_get_regs,
- .vcpu_ioctl_set_regs = kvm_trap_emul_ioctl_set_regs,
};
int kvm_mips_emulation_init(struct kvm_mips_callbacks **install_callbacks)
--
1.7.11.7
^ permalink raw reply related [flat|nested] 12+ messages in thread
* Re: [PATCH v3 5/5] mips/kvm: Fix ABI by moving manipulation of CP0 registers to KVM_{G,S}ET_MSRS
2013-05-20 21:01 ` [PATCH v3 5/5] mips/kvm: Fix ABI by moving manipulation of CP0 registers to KVM_{G,S}ET_MSRS David Daney
@ 2013-05-21 15:37 ` Gleb Natapov
2013-05-21 16:21 ` David Daney
0 siblings, 1 reply; 12+ messages in thread
From: Gleb Natapov @ 2013-05-21 15:37 UTC (permalink / raw)
To: David Daney; +Cc: linux-mips, ralf, kvm, Sanjay Lal, linux-kernel, David Daney
On Mon, May 20, 2013 at 02:01:26PM -0700, David Daney wrote:
> From: David Daney <david.daney@cavium.com>
>
> Because not all 256 CP0 registers are ever implemented, we need a
> different method of manipulating them. Use the
> KVM_GET_MSRS/KVM_SET_MSRS mechanism as x86 does for its MSRs.
>
Have you looked at KVM_(GET|SET)_ONE_REG interface (not used by x86, but is
used bu arm/ppc/s390). It looks like it is more suitable for your case.
Actually you can use it instead of KVM_(GET|SET)_REGS for all registers.
> Code related to implementing KVM_GET_MSRS/KVM_SET_MSRS is consolidated
> in to kvm_trap_emul.c, now unused code and definitions are removed.
>
> Signed-off-by: David Daney <david.daney@cavium.com>
> ---
> arch/mips/include/asm/kvm.h | 69 ++++++--
> arch/mips/include/asm/kvm_host.h | 4 -
> arch/mips/kvm/kvm_mips.c | 92 +----------
> arch/mips/kvm/kvm_trap_emul.c | 330 ++++++++++++++++++++++++++++++++++-----
> 4 files changed, 350 insertions(+), 145 deletions(-)
>
> diff --git a/arch/mips/include/asm/kvm.h b/arch/mips/include/asm/kvm.h
> index d145ead..cae080c 100644
> --- a/arch/mips/include/asm/kvm.h
> +++ b/arch/mips/include/asm/kvm.h
> @@ -13,10 +13,11 @@
>
> #include <linux/types.h>
>
> -#define __KVM_MIPS
> -
> -#define N_MIPS_COPROC_REGS 32
> -#define N_MIPS_COPROC_SEL 8
> +/*
> + * KVM MIPS specific structures and definitions.
> + *
> + * Some parts derived from the x86 version of this file.
> + */
>
> /*
> * for KVM_GET_REGS and KVM_SET_REGS
> @@ -31,12 +32,6 @@ struct kvm_regs {
> __u64 hi;
> __u64 lo;
> __u64 pc;
> -
> - __u32 cp0reg[N_MIPS_COPROC_REGS][N_MIPS_COPROC_SEL];
> -};
> -
> -/* for KVM_GET_SREGS and KVM_SET_SREGS */
> -struct kvm_sregs {
> };
>
> /*
> @@ -55,21 +50,67 @@ struct kvm_fpu {
> __u32 pad;
> };
>
> +
> +/*
> + * For MIPS, we use the same APIs as x86, where 'msr' corresponds to a
> + * CP0 register. The index field is broken down as follows:
> + *
> + * bits[2..0] - Register 'sel' index.
> + * bits[7..3] - Register 'rd' index.
> + * bits[15..8] - Must be zero.
> + * bits[31..16] - 0 -> CP0 registers.
> + *
> + * Other sets registers may be added in the future. Each set would
> + * have its own identifier in bits[31..16].
> + *
> + * For MSRs that are narrower than 64-bits, the value is stored in the
> + * low order bits of the data field, and sign extended to 64-bits.
> + */
> +#define KVM_MIPS_MSR_CP0 0
> +struct kvm_msr_entry {
> + __u32 index;
> + __u32 reserved;
> + __u64 data;
> +};
> +
> +/* for KVM_GET_MSRS and KVM_SET_MSRS */
> +struct kvm_msrs {
> + __u32 nmsrs; /* number of msrs in entries */
> + __u32 pad;
> +
> + struct kvm_msr_entry entries[0];
> +};
> +
> +/* for KVM_GET_MSR_INDEX_LIST */
> +struct kvm_msr_list {
> + __u32 nmsrs; /* number of msrs in entries */
> + __u32 indices[0];
> +};
> +
> +/*
> + * KVM MIPS specific structures and definitions
> + *
> + */
> struct kvm_debug_exit_arch {
> + __u64 epc;
> };
>
> /* for KVM_SET_GUEST_DEBUG */
> struct kvm_guest_debug_arch {
> };
>
> +/* definition of registers in kvm_run */
> +struct kvm_sync_regs {
> +};
> +
> +/* dummy definition */
> +struct kvm_sregs {
> +};
> +
> struct kvm_mips_interrupt {
> /* in */
> __u32 cpu;
> __u32 irq;
> };
>
> -/* definition of registers in kvm_run */
> -struct kvm_sync_regs {
> -};
> -
> #endif /* __LINUX_KVM_MIPS_H */
> diff --git a/arch/mips/include/asm/kvm_host.h b/arch/mips/include/asm/kvm_host.h
> index 143875c..4d6fa0b 100644
> --- a/arch/mips/include/asm/kvm_host.h
> +++ b/arch/mips/include/asm/kvm_host.h
> @@ -496,10 +496,6 @@ struct kvm_mips_callbacks {
> uint32_t cause);
> int (*irq_clear) (struct kvm_vcpu *vcpu, unsigned int priority,
> uint32_t cause);
> - int (*vcpu_ioctl_get_regs) (struct kvm_vcpu *vcpu,
> - struct kvm_regs *regs);
> - int (*vcpu_ioctl_set_regs) (struct kvm_vcpu *vcpu,
> - struct kvm_regs *regs);
> };
> extern struct kvm_mips_callbacks *kvm_mips_callbacks;
> int kvm_mips_emulation_init(struct kvm_mips_callbacks **install_callbacks);
> diff --git a/arch/mips/kvm/kvm_mips.c b/arch/mips/kvm/kvm_mips.c
> index 71a1fc1..26a40e3 100644
> --- a/arch/mips/kvm/kvm_mips.c
> +++ b/arch/mips/kvm/kvm_mips.c
> @@ -51,16 +51,6 @@ struct kvm_stats_debugfs_item debugfs_entries[] = {
> {NULL}
> };
>
> -static int kvm_mips_reset_vcpu(struct kvm_vcpu *vcpu)
> -{
> - int i;
> - for_each_possible_cpu(i) {
> - vcpu->arch.guest_kernel_asid[i] = 0;
> - vcpu->arch.guest_user_asid[i] = 0;
> - }
> - return 0;
> -}
> -
> gfn_t unalias_gfn(struct kvm *kvm, gfn_t gfn)
> {
> return gfn;
> @@ -192,12 +182,6 @@ void kvm_arch_destroy_vm(struct kvm *kvm)
> }
> }
>
> -long
> -kvm_arch_dev_ioctl(struct file *filp, unsigned int ioctl, unsigned long arg)
> -{
> - return -EINVAL;
> -}
> -
> void kvm_arch_free_memslot(struct kvm_memory_slot *free,
> struct kvm_memory_slot *dont)
> {
> @@ -435,42 +419,6 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run)
>
> return r;
> }
> -
> -int
> -kvm_vcpu_ioctl_interrupt(struct kvm_vcpu *vcpu, struct kvm_mips_interrupt *irq)
> -{
> - int intr = (int)irq->irq;
> - struct kvm_vcpu *dvcpu = NULL;
> -
> - if (intr == 3 || intr == -3 || intr == 4 || intr == -4)
> - kvm_debug("%s: CPU: %d, INTR: %d\n", __func__, irq->cpu,
> - (int)intr);
> -
> - if (irq->cpu == -1)
> - dvcpu = vcpu;
> - else
> - dvcpu = vcpu->kvm->vcpus[irq->cpu];
> -
> - if (intr == 2 || intr == 3 || intr == 4) {
> - kvm_mips_callbacks->queue_io_int(dvcpu, irq);
> -
> - } else if (intr == -2 || intr == -3 || intr == -4) {
> - kvm_mips_callbacks->dequeue_io_int(dvcpu, irq);
> - } else {
> - kvm_err("%s: invalid interrupt ioctl (%d:%d)\n", __func__,
> - irq->cpu, irq->irq);
> - return -EINVAL;
> - }
> -
> - dvcpu->arch.wait = 0;
> -
> - if (waitqueue_active(&dvcpu->wq)) {
> - wake_up_interruptible(&dvcpu->wq);
> - }
> -
> - return 0;
> -}
> -
> int
> kvm_arch_vcpu_ioctl_get_mpstate(struct kvm_vcpu *vcpu,
> struct kvm_mp_state *mp_state)
> @@ -485,42 +433,6 @@ kvm_arch_vcpu_ioctl_set_mpstate(struct kvm_vcpu *vcpu,
> return -EINVAL;
> }
>
> -long
> -kvm_arch_vcpu_ioctl(struct file *filp, unsigned int ioctl, unsigned long arg)
> -{
> - struct kvm_vcpu *vcpu = filp->private_data;
> - void __user *argp = (void __user *)arg;
> - long r;
> - int intr;
> -
> - switch (ioctl) {
> - case KVM_NMI:
> - /* Treat the NMI as a CPU reset */
> - r = kvm_mips_reset_vcpu(vcpu);
> - break;
> - case KVM_INTERRUPT:
> - {
> - struct kvm_mips_interrupt irq;
> - r = -EFAULT;
> - if (copy_from_user(&irq, argp, sizeof(irq)))
> - goto out;
> -
> - intr = (int)irq.irq;
> -
> - kvm_debug("[%d] %s: irq: %d\n", vcpu->vcpu_id, __func__,
> - irq.irq);
> -
> - r = kvm_vcpu_ioctl_interrupt(vcpu, &irq);
> - break;
> - }
> - default:
> - r = -EINVAL;
> - }
> -
> -out:
> - return r;
> -}
> -
> /*
> * Get (and clear) the dirty memory log for a memory slot.
> */
> @@ -684,7 +596,7 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
> vcpu->arch.lo = regs->lo;
> vcpu->arch.pc = regs->pc;
>
> - return kvm_mips_callbacks->vcpu_ioctl_set_regs(vcpu, regs);
> + return 0;
> }
>
> int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
> @@ -698,7 +610,7 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
> regs->lo = vcpu->arch.lo;
> regs->pc = vcpu->arch.pc;
>
> - return kvm_mips_callbacks->vcpu_ioctl_get_regs(vcpu, regs);
> + return 0;
> }
>
> void kvm_mips_comparecount_func(unsigned long data)
> diff --git a/arch/mips/kvm/kvm_trap_emul.c b/arch/mips/kvm/kvm_trap_emul.c
> index 466aeef..df98dcb 100644
> --- a/arch/mips/kvm/kvm_trap_emul.c
> +++ b/arch/mips/kvm/kvm_trap_emul.c
> @@ -13,7 +13,7 @@
> #include <linux/err.h>
> #include <linux/module.h>
> #include <linux/vmalloc.h>
> -
> +#include <linux/fs.h>
> #include <linux/kvm_host.h>
>
> #include "kvm_mips_opcode.h"
> @@ -345,54 +345,312 @@ static int kvm_trap_emul_handle_break(struct kvm_vcpu *vcpu)
> return ret;
> }
>
> -static int
> -kvm_trap_emul_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
> +#define MSR_MIPS_CP0_INDEX (8 * 0 + 0)
> +#define MSR_MIPS_CP0_ENTRYLO0 (8 * 2 + 0)
> +#define MSR_MIPS_CP0_ENTRYLO1 (8 * 3 + 0)
> +#define MSR_MIPS_CP0_CONTEXT (8 * 4 + 0)
> +#define MSR_MIPS_CP0_USERLOCAL (8 * 4 + 2)
> +#define MSR_MIPS_CP0_PAGEMASK (8 * 5 + 0)
> +#define MSR_MIPS_CP0_PAGEGRAIN (8 * 5 + 1)
> +#define MSR_MIPS_CP0_WIRED (8 * 6 + 0)
> +#define MSR_MIPS_CP0_HWRENA (8 * 7 + 0)
> +#define MSR_MIPS_CP0_BADVADDR (8 * 8 + 0)
> +#define MSR_MIPS_CP0_COUNT (8 * 9 + 0)
> +#define MSR_MIPS_CP0_ENTRYHI (8 * 10 + 0)
> +#define MSR_MIPS_CP0_COMPARE (8 * 11 + 0)
> +#define MSR_MIPS_CP0_STATUS (8 * 12 + 0)
> +#define MSR_MIPS_CP0_CAUSE (8 * 13 + 0)
> +#define MSR_MIPS_CP0_EBASE (8 * 15 + 1)
> +#define MSR_MIPS_CP0_CONFIG (8 * 16 + 0)
> +#define MSR_MIPS_CP0_CONFIG1 (8 * 16 + 1)
> +#define MSR_MIPS_CP0_CONFIG2 (8 * 16 + 2)
> +#define MSR_MIPS_CP0_CONFIG3 (8 * 16 + 3)
> +#define MSR_MIPS_CP0_CONFIG7 (8 * 16 + 7)
> +#define MSR_MIPS_CP0_XCONTEXT (8 * 20 + 0)
> +#define MSR_MIPS_CP0_ERROREPC (8 * 30 + 0)
> +
> +static u32 msrs_to_save[] = {
> + MSR_MIPS_CP0_INDEX,
> + MSR_MIPS_CP0_CONTEXT,
> + MSR_MIPS_CP0_PAGEMASK,
> + MSR_MIPS_CP0_WIRED,
> + MSR_MIPS_CP0_BADVADDR,
> + MSR_MIPS_CP0_ENTRYHI,
> + MSR_MIPS_CP0_STATUS,
> + MSR_MIPS_CP0_CAUSE,
> + /* EPC set via kvm_regs, et al. */
> + MSR_MIPS_CP0_CONFIG,
> + MSR_MIPS_CP0_CONFIG1,
> + MSR_MIPS_CP0_CONFIG2,
> + MSR_MIPS_CP0_CONFIG3,
> + MSR_MIPS_CP0_CONFIG7,
> + MSR_MIPS_CP0_XCONTEXT,
> + MSR_MIPS_CP0_ERROREPC
> +};
> +
> +#define MAX_IO_MSRS 128
> +
> +static int mipsvz_get_msr(struct kvm_vcpu *vcpu, unsigned index, u64 *data)
> {
> struct mips_coproc *cop0 = vcpu->arch.cop0;
> + s64 v;
>
> - kvm_write_c0_guest_index(cop0, regs->cp0reg[MIPS_CP0_TLB_INDEX][0]);
> - kvm_write_c0_guest_context(cop0, regs->cp0reg[MIPS_CP0_TLB_CONTEXT][0]);
> - kvm_write_c0_guest_badvaddr(cop0, regs->cp0reg[MIPS_CP0_BAD_VADDR][0]);
> - kvm_write_c0_guest_entryhi(cop0, regs->cp0reg[MIPS_CP0_TLB_HI][0]);
> - kvm_write_c0_guest_epc(cop0, regs->cp0reg[MIPS_CP0_EXC_PC][0]);
> + switch (index) {
> + case MSR_MIPS_CP0_INDEX:
> + v = (long)kvm_read_c0_guest_index(cop0);
> + break;
> + case MSR_MIPS_CP0_CONTEXT:
> + v = (long)kvm_read_c0_guest_context(cop0);
> + break;
> + case MSR_MIPS_CP0_PAGEMASK:
> + v = (long)kvm_read_c0_guest_pagemask(cop0);
> + break;
> + case MSR_MIPS_CP0_WIRED:
> + v = (long)kvm_read_c0_guest_wired(cop0);
> + break;
> + case MSR_MIPS_CP0_BADVADDR:
> + v = (long)kvm_read_c0_guest_badvaddr(cop0);
> + break;
> + case MSR_MIPS_CP0_ENTRYHI:
> + v = (long)kvm_read_c0_guest_entryhi(cop0);
> + break;
> + case MSR_MIPS_CP0_STATUS:
> + v = (long)kvm_read_c0_guest_status(cop0);
> + break;
> + case MSR_MIPS_CP0_CAUSE:
> + v = (long)kvm_read_c0_guest_cause(cop0);
> + break;
> + case MSR_MIPS_CP0_ERROREPC:
> + v = (long)kvm_read_c0_guest_errorepc(cop0);
> + break;
> + case MSR_MIPS_CP0_CONFIG:
> + v = (long)kvm_read_c0_guest_config(cop0);
> + break;
> + case MSR_MIPS_CP0_CONFIG1:
> + v = (long)kvm_read_c0_guest_config1(cop0);
> + break;
> + case MSR_MIPS_CP0_CONFIG2:
> + v = (long)kvm_read_c0_guest_config2(cop0);
> + break;
> + case MSR_MIPS_CP0_CONFIG3:
> + v = (long)kvm_read_c0_guest_config3(cop0);
> + break;
> + case MSR_MIPS_CP0_CONFIG7:
> + v = (long)kvm_read_c0_guest_config7(cop0);
> + break;
> + default:
> + return -EINVAL;
> + }
> + *data = v;
> + return 0;
> +}
>
> - kvm_write_c0_guest_status(cop0, regs->cp0reg[MIPS_CP0_STATUS][0]);
> - kvm_write_c0_guest_cause(cop0, regs->cp0reg[MIPS_CP0_CAUSE][0]);
> - kvm_write_c0_guest_pagemask(cop0,
> - regs->cp0reg[MIPS_CP0_TLB_PG_MASK][0]);
> - kvm_write_c0_guest_wired(cop0, regs->cp0reg[MIPS_CP0_TLB_WIRED][0]);
> - kvm_write_c0_guest_errorepc(cop0, regs->cp0reg[MIPS_CP0_ERROR_PC][0]);
> +static int mipsvz_set_msr(struct kvm_vcpu *vcpu, unsigned index, u64 *data)
> +{
> + struct mips_coproc *cop0 = vcpu->arch.cop0;
> + u64 v = *data;
>
> + switch (index) {
> + case MSR_MIPS_CP0_INDEX:
> + kvm_write_c0_guest_index(cop0, v);
> + break;
> + case MSR_MIPS_CP0_CONTEXT:
> + kvm_write_c0_guest_context(cop0, v);
> + break;
> + case MSR_MIPS_CP0_PAGEMASK:
> + kvm_write_c0_guest_pagemask(cop0, v);
> + break;
> + case MSR_MIPS_CP0_WIRED:
> + kvm_write_c0_guest_wired(cop0, v);
> + break;
> + case MSR_MIPS_CP0_BADVADDR:
> + kvm_write_c0_guest_badvaddr(cop0, v);
> + break;
> + case MSR_MIPS_CP0_ENTRYHI:
> + kvm_write_c0_guest_entryhi(cop0, v);
> + break;
> + case MSR_MIPS_CP0_STATUS:
> + kvm_write_c0_guest_status(cop0, v);
> + break;
> + case MSR_MIPS_CP0_CAUSE:
> + kvm_write_c0_guest_cause(cop0, v);
> + break;
> + case MSR_MIPS_CP0_ERROREPC:
> + kvm_write_c0_guest_errorepc(cop0, v);
> + break;
> + default:
> + return -EINVAL;
> + }
> return 0;
> }
>
> -static int
> -kvm_trap_emul_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
> +/*
> + * Read or write a bunch of msrs. Parameters are user addresses.
> + *
> + * @return number of msrs set successfully.
> + */
> +static int msr_io(struct kvm_vcpu *vcpu, struct kvm_msrs __user *user_msrs,
> + int (*do_msr)(struct kvm_vcpu *vcpu,
> + unsigned index, u64 *data),
> + bool writeback)
> {
> - struct mips_coproc *cop0 = vcpu->arch.cop0;
> + struct kvm_msrs msrs;
> + struct kvm_msr_entry *entries;
> + int r, n;
> + unsigned size;
> +
> + r = -EFAULT;
> + if (copy_from_user(&msrs, user_msrs, sizeof(msrs)))
> + goto out;
> +
> + r = -E2BIG;
> + if (msrs.nmsrs >= MAX_IO_MSRS)
> + goto out;
> +
> + size = sizeof(struct kvm_msr_entry) * msrs.nmsrs;
> + entries = memdup_user(user_msrs->entries, size);
> + if (IS_ERR(entries)) {
> + r = PTR_ERR(entries);
> + goto out;
> + }
>
> - regs->cp0reg[MIPS_CP0_TLB_INDEX][0] = kvm_read_c0_guest_index(cop0);
> - regs->cp0reg[MIPS_CP0_TLB_CONTEXT][0] = kvm_read_c0_guest_context(cop0);
> - regs->cp0reg[MIPS_CP0_BAD_VADDR][0] = kvm_read_c0_guest_badvaddr(cop0);
> - regs->cp0reg[MIPS_CP0_TLB_HI][0] = kvm_read_c0_guest_entryhi(cop0);
> - regs->cp0reg[MIPS_CP0_EXC_PC][0] = kvm_read_c0_guest_epc(cop0);
> -
> - regs->cp0reg[MIPS_CP0_STATUS][0] = kvm_read_c0_guest_status(cop0);
> - regs->cp0reg[MIPS_CP0_CAUSE][0] = kvm_read_c0_guest_cause(cop0);
> - regs->cp0reg[MIPS_CP0_TLB_PG_MASK][0] =
> - kvm_read_c0_guest_pagemask(cop0);
> - regs->cp0reg[MIPS_CP0_TLB_WIRED][0] = kvm_read_c0_guest_wired(cop0);
> - regs->cp0reg[MIPS_CP0_ERROR_PC][0] = kvm_read_c0_guest_errorepc(cop0);
> -
> - regs->cp0reg[MIPS_CP0_CONFIG][0] = kvm_read_c0_guest_config(cop0);
> - regs->cp0reg[MIPS_CP0_CONFIG][1] = kvm_read_c0_guest_config1(cop0);
> - regs->cp0reg[MIPS_CP0_CONFIG][2] = kvm_read_c0_guest_config2(cop0);
> - regs->cp0reg[MIPS_CP0_CONFIG][3] = kvm_read_c0_guest_config3(cop0);
> - regs->cp0reg[MIPS_CP0_CONFIG][7] = kvm_read_c0_guest_config7(cop0);
> + for (n = 0; n < msrs.nmsrs; ++n)
> + if (do_msr(vcpu, entries[n].index, &entries[n].data))
> + break;
> +
> + r = -EFAULT;
> + if (writeback && copy_to_user(user_msrs->entries, entries, size))
> + goto out_free;
> +
> + r = n;
> +
> +out_free:
> + kfree(entries);
> +out:
> + return r;
> +}
> +
> +static int kvm_mips_reset_vcpu(struct kvm_vcpu *vcpu)
> +{
> + int i;
> + for_each_possible_cpu(i) {
> + vcpu->arch.guest_kernel_asid[i] = 0;
> + vcpu->arch.guest_user_asid[i] = 0;
> + }
> + return 0;
> +}
> +
> +int
> +kvm_vcpu_ioctl_interrupt(struct kvm_vcpu *vcpu, struct kvm_mips_interrupt *irq)
> +{
> + int intr = (int)irq->irq;
> + struct kvm_vcpu *dvcpu = NULL;
> +
> + if (intr == 3 || intr == -3 || intr == 4 || intr == -4)
> + kvm_debug("%s: CPU: %d, INTR: %d\n", __func__, irq->cpu,
> + (int)intr);
> +
> + if (irq->cpu == -1)
> + dvcpu = vcpu;
> + else
> + dvcpu = vcpu->kvm->vcpus[irq->cpu];
> +
> + if (intr == 2 || intr == 3 || intr == 4) {
> + kvm_mips_callbacks->queue_io_int(dvcpu, irq);
> +
> + } else if (intr == -2 || intr == -3 || intr == -4) {
> + kvm_mips_callbacks->dequeue_io_int(dvcpu, irq);
> + } else {
> + kvm_err("%s: invalid interrupt ioctl (%d:%d)\n", __func__,
> + irq->cpu, irq->irq);
> + return -EINVAL;
> + }
> +
> + dvcpu->arch.wait = 0;
> +
> + if (waitqueue_active(&dvcpu->wq))
> + wake_up_interruptible(&dvcpu->wq);
>
> return 0;
> }
>
> +long
> +kvm_arch_vcpu_ioctl(struct file *filp, unsigned int ioctl, unsigned long arg)
> +{
> + struct kvm_vcpu *vcpu = filp->private_data;
> + void __user *argp = (void __user *)arg;
> + long r;
> +
> + switch (ioctl) {
> + case KVM_GET_MSRS:
> + r = msr_io(vcpu, argp, mipsvz_get_msr, true);
> + break;
> + case KVM_SET_MSRS:
> + r = msr_io(vcpu, argp, mipsvz_set_msr, false);
> + break;
> + case KVM_NMI:
> + /* Treat the NMI as a CPU reset */
> + r = kvm_mips_reset_vcpu(vcpu);
> + break;
> + case KVM_INTERRUPT:
> + {
> + struct kvm_mips_interrupt irq;
> + r = -EFAULT;
> + if (copy_from_user(&irq, argp, sizeof(irq)))
> + goto out;
> +
> + kvm_debug("[%d] %s: irq: %d\n", vcpu->vcpu_id, __func__,
> + irq.irq);
> +
> + r = kvm_vcpu_ioctl_interrupt(vcpu, &irq);
> + break;
> + }
> + default:
> + r = -ENOIOCTLCMD;
> + }
> +
> +out:
> + return r;
> +}
> +
> +long kvm_arch_dev_ioctl(struct file *filp,
> + unsigned int ioctl, unsigned long arg)
> +{
> + long r;
> + void __user *argp = (void __user *)arg;
> +
> + switch (ioctl) {
> + case KVM_GET_MSR_INDEX_LIST: {
> + struct kvm_msr_list __user *user_msr_list = argp;
> + struct kvm_msr_list msr_list;
> + unsigned n;
> + unsigned num_msrs_to_save;
> +
> + r = -EFAULT;
> + if (copy_from_user(&msr_list, user_msr_list, sizeof(msr_list)))
> + goto out;
> + n = msr_list.nmsrs;
> + msr_list.nmsrs = ARRAY_SIZE(msrs_to_save);
> + num_msrs_to_save = min(n, msr_list.nmsrs);
> + if (copy_to_user(user_msr_list, &msr_list, sizeof(msr_list)))
> + goto out;
> + r = -E2BIG;
> + if (n < msr_list.nmsrs)
> + goto out;
> + r = -EFAULT;
> + if (copy_to_user(user_msr_list->indices, &msrs_to_save,
> + num_msrs_to_save * sizeof(u32)))
> + goto out;
> + r = 0;
> + break;
> + }
> + default:
> + r = -ENOIOCTLCMD;
> + }
> +out:
> + return r;
> +}
> +
> static int kvm_trap_emul_vm_init(struct kvm *kvm)
> {
> return 0;
> @@ -471,8 +729,6 @@ static struct kvm_mips_callbacks kvm_trap_emul_callbacks = {
> .dequeue_io_int = kvm_mips_dequeue_io_int_cb,
> .irq_deliver = kvm_mips_irq_deliver_cb,
> .irq_clear = kvm_mips_irq_clear_cb,
> - .vcpu_ioctl_get_regs = kvm_trap_emul_ioctl_get_regs,
> - .vcpu_ioctl_set_regs = kvm_trap_emul_ioctl_set_regs,
> };
>
> int kvm_mips_emulation_init(struct kvm_mips_callbacks **install_callbacks)
> --
> 1.7.11.7
>
> --
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at http://www.tux.org/lkml/
--
Gleb.
^ permalink raw reply [flat|nested] 12+ messages in thread
* Re: [PATCH v3 5/5] mips/kvm: Fix ABI by moving manipulation of CP0 registers to KVM_{G,S}ET_MSRS
@ 2013-05-21 16:21 ` David Daney
0 siblings, 0 replies; 12+ messages in thread
From: David Daney @ 2013-05-21 16:21 UTC (permalink / raw)
To: Gleb Natapov
Cc: David Daney, linux-mips, ralf, kvm, Sanjay Lal, linux-kernel,
David Daney
On 05/21/2013 08:37 AM, Gleb Natapov wrote:
> On Mon, May 20, 2013 at 02:01:26PM -0700, David Daney wrote:
>> From: David Daney <david.daney@cavium.com>
>>
>> Because not all 256 CP0 registers are ever implemented, we need a
>> different method of manipulating them. Use the
>> KVM_GET_MSRS/KVM_SET_MSRS mechanism as x86 does for its MSRs.
>>
> Have you looked at KVM_(GET|SET)_ONE_REG interface (not used by x86, but is
> used bu arm/ppc/s390). It looks like it is more suitable for your case.
> Actually you can use it instead of KVM_(GET|SET)_REGS for all registers.
Yes, I suppose it could be used. One problem it has is that there is no
way to query the set of supported registers. Also you have to make
multiple calls to set multiple registers, which involves vcpu_{load,put}
for each register.
We will definitely implement it for all the FP and General Purpose
registers.
>
>> Code related to implementing KVM_GET_MSRS/KVM_SET_MSRS is consolidated
^ permalink raw reply [flat|nested] 12+ messages in thread
* Re: [PATCH v3 5/5] mips/kvm: Fix ABI by moving manipulation of CP0 registers to KVM_{G,S}ET_MSRS
@ 2013-05-21 16:21 ` David Daney
0 siblings, 0 replies; 12+ messages in thread
From: David Daney @ 2013-05-21 16:21 UTC (permalink / raw)
To: Gleb Natapov
Cc: David Daney, linux-mips, ralf, kvm, Sanjay Lal, linux-kernel,
David Daney
On 05/21/2013 08:37 AM, Gleb Natapov wrote:
> On Mon, May 20, 2013 at 02:01:26PM -0700, David Daney wrote:
>> From: David Daney <david.daney@cavium.com>
>>
>> Because not all 256 CP0 registers are ever implemented, we need a
>> different method of manipulating them. Use the
>> KVM_GET_MSRS/KVM_SET_MSRS mechanism as x86 does for its MSRs.
>>
> Have you looked at KVM_(GET|SET)_ONE_REG interface (not used by x86, but is
> used bu arm/ppc/s390). It looks like it is more suitable for your case.
> Actually you can use it instead of KVM_(GET|SET)_REGS for all registers.
Yes, I suppose it could be used. One problem it has is that there is no
way to query the set of supported registers. Also you have to make
multiple calls to set multiple registers, which involves vcpu_{load,put}
for each register.
We will definitely implement it for all the FP and General Purpose
registers.
>
>> Code related to implementing KVM_GET_MSRS/KVM_SET_MSRS is consolidated
^ permalink raw reply [flat|nested] 12+ messages in thread
* Re: [PATCH v3 5/5] mips/kvm: Fix ABI by moving manipulation of CP0 registers to KVM_{G,S}ET_MSRS
2013-05-21 16:21 ` David Daney
(?)
@ 2013-05-21 16:28 ` Gleb Natapov
2013-05-21 16:34 ` David Daney
-1 siblings, 1 reply; 12+ messages in thread
From: Gleb Natapov @ 2013-05-21 16:28 UTC (permalink / raw)
To: David Daney
Cc: David Daney, linux-mips, ralf, kvm, Sanjay Lal, linux-kernel,
David Daney
On Tue, May 21, 2013 at 09:21:06AM -0700, David Daney wrote:
> On 05/21/2013 08:37 AM, Gleb Natapov wrote:
> >On Mon, May 20, 2013 at 02:01:26PM -0700, David Daney wrote:
> >>From: David Daney <david.daney@cavium.com>
> >>
> >>Because not all 256 CP0 registers are ever implemented, we need a
> >>different method of manipulating them. Use the
> >>KVM_GET_MSRS/KVM_SET_MSRS mechanism as x86 does for its MSRs.
> >>
> >Have you looked at KVM_(GET|SET)_ONE_REG interface (not used by x86, but is
> >used bu arm/ppc/s390). It looks like it is more suitable for your case.
> >Actually you can use it instead of KVM_(GET|SET)_REGS for all registers.
>
> Yes, I suppose it could be used. One problem it has is that there
> is no way to query the set of supported registers.
KVM_GET_REG_LIST
> Also you have to
> make multiple calls to set multiple registers, which involves
> vcpu_{load,put} for each register.
>
How often this happens on the fast path on mips? On x86 this never
happens on the fast path so it uses KVM_(GET|SET)_REGS mostly for
historical reasons.
> We will definitely implement it for all the FP and General Purpose
> registers.
>
> >
> >>Code related to implementing KVM_GET_MSRS/KVM_SET_MSRS is consolidated
>
--
Gleb.
^ permalink raw reply [flat|nested] 12+ messages in thread
* Re: [PATCH v3 5/5] mips/kvm: Fix ABI by moving manipulation of CP0 registers to KVM_{G,S}ET_MSRS
@ 2013-05-21 16:34 ` David Daney
0 siblings, 0 replies; 12+ messages in thread
From: David Daney @ 2013-05-21 16:34 UTC (permalink / raw)
To: Gleb Natapov
Cc: David Daney, linux-mips, ralf, kvm, Sanjay Lal, linux-kernel,
David Daney
On 05/21/2013 09:28 AM, Gleb Natapov wrote:
> On Tue, May 21, 2013 at 09:21:06AM -0700, David Daney wrote:
>> On 05/21/2013 08:37 AM, Gleb Natapov wrote:
>>> On Mon, May 20, 2013 at 02:01:26PM -0700, David Daney wrote:
>>>> From: David Daney <david.daney@cavium.com>
>>>>
>>>> Because not all 256 CP0 registers are ever implemented, we need a
>>>> different method of manipulating them. Use the
>>>> KVM_GET_MSRS/KVM_SET_MSRS mechanism as x86 does for its MSRs.
>>>>
>>> Have you looked at KVM_(GET|SET)_ONE_REG interface (not used by x86, but is
>>> used bu arm/ppc/s390). It looks like it is more suitable for your case.
>>> Actually you can use it instead of KVM_(GET|SET)_REGS for all registers.
>>
>> Yes, I suppose it could be used. One problem it has is that there
>> is no way to query the set of supported registers.
> KVM_GET_REG_LIST
OK, I guess we can switch to KVM_GET_REG_LIST/KVM_(GET|SET)_ONE_REG.
I will revise this patch.
Thanks.
>
>> Also you have to
>> make multiple calls to set multiple registers, which involves
>> vcpu_{load,put} for each register.
>>
> How often this happens on the fast path on mips?
Rarely, if ever once, the VCPU is running.
Only when initializing the VCPU do we need to set a bunch of registers.
> On x86 this never
> happens on the fast path so it uses KVM_(GET|SET)_REGS mostly for
> historical reasons.
>
>> We will definitely implement it for all the FP and General Purpose
>> registers.
>>
>>>
>>>> Code related to implementing KVM_GET_MSRS/KVM_SET_MSRS is consolidated
>>
>
> --
> Gleb.
>
^ permalink raw reply [flat|nested] 12+ messages in thread
* Re: [PATCH v3 5/5] mips/kvm: Fix ABI by moving manipulation of CP0 registers to KVM_{G,S}ET_MSRS
@ 2013-05-21 16:34 ` David Daney
0 siblings, 0 replies; 12+ messages in thread
From: David Daney @ 2013-05-21 16:34 UTC (permalink / raw)
To: Gleb Natapov
Cc: David Daney, linux-mips, ralf, kvm, Sanjay Lal, linux-kernel,
David Daney
On 05/21/2013 09:28 AM, Gleb Natapov wrote:
> On Tue, May 21, 2013 at 09:21:06AM -0700, David Daney wrote:
>> On 05/21/2013 08:37 AM, Gleb Natapov wrote:
>>> On Mon, May 20, 2013 at 02:01:26PM -0700, David Daney wrote:
>>>> From: David Daney <david.daney@cavium.com>
>>>>
>>>> Because not all 256 CP0 registers are ever implemented, we need a
>>>> different method of manipulating them. Use the
>>>> KVM_GET_MSRS/KVM_SET_MSRS mechanism as x86 does for its MSRs.
>>>>
>>> Have you looked at KVM_(GET|SET)_ONE_REG interface (not used by x86, but is
>>> used bu arm/ppc/s390). It looks like it is more suitable for your case.
>>> Actually you can use it instead of KVM_(GET|SET)_REGS for all registers.
>>
>> Yes, I suppose it could be used. One problem it has is that there
>> is no way to query the set of supported registers.
> KVM_GET_REG_LIST
OK, I guess we can switch to KVM_GET_REG_LIST/KVM_(GET|SET)_ONE_REG.
I will revise this patch.
Thanks.
>
>> Also you have to
>> make multiple calls to set multiple registers, which involves
>> vcpu_{load,put} for each register.
>>
> How often this happens on the fast path on mips?
Rarely, if ever once, the VCPU is running.
Only when initializing the VCPU do we need to set a bunch of registers.
> On x86 this never
> happens on the fast path so it uses KVM_(GET|SET)_REGS mostly for
> historical reasons.
>
>> We will definitely implement it for all the FP and General Purpose
>> registers.
>>
>>>
>>>> Code related to implementing KVM_GET_MSRS/KVM_SET_MSRS is consolidated
>>
>
> --
> Gleb.
>
^ permalink raw reply [flat|nested] 12+ messages in thread
end of thread, other threads:[~2013-05-21 16:34 UTC | newest]
Thread overview: 12+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2013-05-20 21:01 [PATCH v3 0/5] mips/kvm: Fix ABI for compatibility with 64-bit guests David Daney
2013-05-20 21:01 ` [PATCH v3 1/5] mips/kvm: Fix ABI for use of FPU David Daney
2013-05-20 21:01 ` [PATCH v3 2/5] mips/kvm: Fix ABI for use of 64-bit registers David Daney
2013-05-20 21:01 ` [PATCH v3 3/5] mips/kvm: Fix name of gpr field in struct kvm_regs David Daney
2013-05-20 21:01 ` [PATCH v3 4/5] mips/kvm: Use ARRAY_SIZE() instead of hardcoded constants in kvm_arch_vcpu_ioctl_{s,g}et_regs David Daney
2013-05-20 21:01 ` [PATCH v3 5/5] mips/kvm: Fix ABI by moving manipulation of CP0 registers to KVM_{G,S}ET_MSRS David Daney
2013-05-21 15:37 ` Gleb Natapov
2013-05-21 16:21 ` David Daney
2013-05-21 16:21 ` David Daney
2013-05-21 16:28 ` Gleb Natapov
2013-05-21 16:34 ` David Daney
2013-05-21 16:34 ` David Daney
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.