* [RFC PATCH 1/4] KVM: arm/arm64: Introduce kvm_arch_vcpu_run_pid_change
2018-03-21 17:47 [RFC PATCH 0/4] KVM: arm64: Optimise FPSIMD context switching Dave Martin
@ 2018-03-21 17:47 ` Dave Martin
2018-03-21 17:47 ` [RFC PATCH 2/4] KVM: arm64: Convert lazy FPSIMD context switch trap to C Dave Martin
` (2 subsequent siblings)
3 siblings, 0 replies; 5+ messages in thread
From: Dave Martin @ 2018-03-21 17:47 UTC (permalink / raw)
To: linux-arm-kernel
From: Christoffer Dall <christoffer.dall@linaro.org>
KVM/ARM differs from other architectures in having to maintain an
additional virtual address space from that of the host and the guest,
because we split the execution of KVM across both EL1 and EL2.
This results in a need to explicitly map data structures into EL2 (hyp)
which are accessed from the hyp code. As we are about to be more clever
with our FPSIMD handling, which stores data on the task struct and uses
thread_info flags, we have to map the currently executing task struct
into the EL2 virtual address space.
However, we don't want to do this on every KVM_RUN, because it is a
fairly expensive operation to walk the page tables, and the common
execution mode is to map a single thread to a VCPU. By introducing a
hook that architectures can select with HAVE_KVM_VCPU_RUN_PID_CHANGE, we
do not introduce overhead for other architectures, but have a simple way
to only map the data we need when required for arm64.
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
Signed-off-by: Dave Martin <Dave.Martin@arm.com>
---
The use of hyp_current to pass the value of "current" to hyp is backed
out compared with Chritoffer's original patch.
Instead, I do most of the grubbing around in task_struct to the host
side to reduce the amount of stuff that needs to be duplicated in hyp.
---
arch/arm64/kvm/Kconfig | 1 +
include/linux/kvm_host.h | 9 +++++++++
virt/kvm/Kconfig | 3 +++
virt/kvm/arm/arm.c | 10 ++++++++++
virt/kvm/kvm_main.c | 7 ++++++-
5 files changed, 29 insertions(+), 1 deletion(-)
diff --git a/arch/arm64/kvm/Kconfig b/arch/arm64/kvm/Kconfig
index 2257dfc..5b2c8d8 100644
--- a/arch/arm64/kvm/Kconfig
+++ b/arch/arm64/kvm/Kconfig
@@ -39,6 +39,7 @@ config KVM
select HAVE_KVM_IRQ_ROUTING
select IRQ_BYPASS_MANAGER
select HAVE_KVM_IRQ_BYPASS
+ select HAVE_KVM_VCPU_RUN_PID_CHANGE
---help---
Support hosting virtualized guest machines.
We don't support KVM with 16K page tables yet, due to the multiple
diff --git a/include/linux/kvm_host.h b/include/linux/kvm_host.h
index ac0062b..10a37b1 100644
--- a/include/linux/kvm_host.h
+++ b/include/linux/kvm_host.h
@@ -1272,4 +1272,13 @@ static inline long kvm_arch_vcpu_async_ioctl(struct file *filp,
}
#endif /* CONFIG_HAVE_KVM_VCPU_ASYNC_IOCTL */
+#ifdef CONFIG_HAVE_KVM_VCPU_RUN_PID_CHANGE
+int kvm_arch_vcpu_run_pid_change(struct kvm_vcpu *vcpu);
+#else
+static inline int kvm_arch_vcpu_run_pid_change(struct kvm_vcpu *vcpu)
+{
+ return 0;
+}
+#endif /* CONFIG_HAVE_KVM_VCPU_RUN_PID_CHANGE */
+
#endif
diff --git a/virt/kvm/Kconfig b/virt/kvm/Kconfig
index cca7e06..72143cf 100644
--- a/virt/kvm/Kconfig
+++ b/virt/kvm/Kconfig
@@ -54,3 +54,6 @@ config HAVE_KVM_IRQ_BYPASS
config HAVE_KVM_VCPU_ASYNC_IOCTL
bool
+
+config HAVE_KVM_VCPU_RUN_PID_CHANGE
+ bool
diff --git a/virt/kvm/arm/arm.c b/virt/kvm/arm/arm.c
index 86941f6..02a153a 100644
--- a/virt/kvm/arm/arm.c
+++ b/virt/kvm/arm/arm.c
@@ -822,6 +822,16 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run)
return ret;
}
+#ifdef CONFIG_ARM64
+int kvm_arch_vcpu_run_pid_change(struct kvm_vcpu *vcpu)
+{
+ struct task_struct *tsk = current;
+
+ /* Make sure the host task fpsimd state is visible to hyp: */
+ return create_hyp_mappings(tsk, tsk + 1, PAGE_HYP);
+}
+#endif
+
static int vcpu_interrupt_line(struct kvm_vcpu *vcpu, int number, bool level)
{
int bit_index;
diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index 4501e65..dbd35ab 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -2551,8 +2551,13 @@ static long kvm_vcpu_ioctl(struct file *filp,
oldpid = rcu_access_pointer(vcpu->pid);
if (unlikely(oldpid != current->pids[PIDTYPE_PID].pid)) {
/* The thread running this VCPU changed. */
- struct pid *newpid = get_task_pid(current, PIDTYPE_PID);
+ struct pid *newpid;
+ r = kvm_arch_vcpu_run_pid_change(vcpu);
+ if (r)
+ break;
+
+ newpid = get_task_pid(current, PIDTYPE_PID);
rcu_assign_pointer(vcpu->pid, newpid);
if (oldpid)
synchronize_rcu();
--
2.1.4
^ permalink raw reply related [flat|nested] 5+ messages in thread* [RFC PATCH 2/4] KVM: arm64: Convert lazy FPSIMD context switch trap to C
2018-03-21 17:47 [RFC PATCH 0/4] KVM: arm64: Optimise FPSIMD context switching Dave Martin
2018-03-21 17:47 ` [RFC PATCH 1/4] KVM: arm/arm64: Introduce kvm_arch_vcpu_run_pid_change Dave Martin
@ 2018-03-21 17:47 ` Dave Martin
2018-03-21 17:47 ` [RFC PATCH 3/4] arm64: fpsimd: Split cpu field out from struct fpsimd_state Dave Martin
2018-03-21 17:47 ` [RFC PATCH 4/4] KVM: arm64: Optimise FPSIMD context switching Dave Martin
3 siblings, 0 replies; 5+ messages in thread
From: Dave Martin @ 2018-03-21 17:47 UTC (permalink / raw)
To: linux-arm-kernel
To make the lazt FPSIMD context switch trap code easier to hack on,
this patch converts it to C.
This is not amazingly efficient, but the trap should typically only
be taken once per host context switch.
Signed-off-by: Dave Martin <Dave.Martin@arm.com>
---
This patch adds the missing ! to ensure that the hyp stack pointer
is actually moved down when pushing regs in preparation for calling
__hyp_switch_fpsimd. Without this, bad things happen...
---
arch/arm64/kvm/hyp/entry.S | 57 +++++++++++++++++----------------------------
arch/arm64/kvm/hyp/switch.c | 24 +++++++++++++++++++
2 files changed, 46 insertions(+), 35 deletions(-)
diff --git a/arch/arm64/kvm/hyp/entry.S b/arch/arm64/kvm/hyp/entry.S
index fdd1068..47c6a78 100644
--- a/arch/arm64/kvm/hyp/entry.S
+++ b/arch/arm64/kvm/hyp/entry.S
@@ -176,41 +176,28 @@ ENTRY(__fpsimd_guest_restore)
// x1: vcpu
// x2-x29,lr: vcpu regs
// vcpu x0-x1 on the stack
- stp x2, x3, [sp, #-16]!
- stp x4, lr, [sp, #-16]!
-
-alternative_if_not ARM64_HAS_VIRT_HOST_EXTN
- mrs x2, cptr_el2
- bic x2, x2, #CPTR_EL2_TFP
- msr cptr_el2, x2
-alternative_else
- mrs x2, cpacr_el1
- orr x2, x2, #CPACR_EL1_FPEN
- msr cpacr_el1, x2
-alternative_endif
- isb
-
- mov x3, x1
-
- ldr x0, [x3, #VCPU_HOST_CONTEXT]
- kern_hyp_va x0
- add x0, x0, #CPU_GP_REG_OFFSET(CPU_FP_REGS)
- bl __fpsimd_save_state
-
- add x2, x3, #VCPU_CONTEXT
- add x0, x2, #CPU_GP_REG_OFFSET(CPU_FP_REGS)
- bl __fpsimd_restore_state
-
- // Skip restoring fpexc32 for AArch64 guests
- mrs x1, hcr_el2
- tbnz x1, #HCR_RW_SHIFT, 1f
- ldr x4, [x3, #VCPU_FPEXC32_EL2]
- msr fpexc32_el2, x4
-1:
- ldp x4, lr, [sp], #16
- ldp x2, x3, [sp], #16
- ldp x0, x1, [sp], #16
-
+ stp x2, x3, [sp, #-144]!
+ stp x4, x5, [sp, #16]
+ stp x6, x7, [sp, #32]
+ stp x8, x9, [sp, #48]
+ stp x10, x11, [sp, #64]
+ stp x12, x13, [sp, #80]
+ stp x14, x15, [sp, #96]
+ stp x16, x17, [sp, #112]
+ stp x18, lr, [sp, #128]
+
+ bl __hyp_switch_fpsimd
+
+ ldp x4, x5, [sp, #16]
+ ldp x6, x7, [sp, #32]
+ ldp x8, x9, [sp, #48]
+ ldp x10, x11, [sp, #64]
+ ldp x12, x13, [sp, #80]
+ ldp x14, x15, [sp, #96]
+ ldp x16, x17, [sp, #112]
+ ldp x18, lr, [sp, #128]
+ ldp x0, x1, [sp, #144]
+ ldp x2, x3, [sp], #160
eret
ENDPROC(__fpsimd_guest_restore)
diff --git a/arch/arm64/kvm/hyp/switch.c b/arch/arm64/kvm/hyp/switch.c
index 24f52fe..01e4254 100644
--- a/arch/arm64/kvm/hyp/switch.c
+++ b/arch/arm64/kvm/hyp/switch.c
@@ -436,6 +436,30 @@ int __hyp_text __kvm_vcpu_run(struct kvm_vcpu *vcpu)
return exit_code;
}
+void __hyp_text __hyp_switch_fpsimd(u64 esr __always_unused,
+ struct kvm_vcpu *vcpu)
+{
+ kvm_cpu_context_t *host_ctxt;
+
+ if (has_vhe())
+ write_sysreg(read_sysreg(cpacr_el1) | CPACR_EL1_FPEN,
+ cpacr_el1);
+ else
+ write_sysreg(read_sysreg(cptr_el2) & ~(u64)CPTR_EL2_TFP,
+ cptr_el2);
+
+ isb();
+
+ host_ctxt = kern_hyp_va(vcpu->arch.host_cpu_context);
+ __fpsimd_save_state(&host_ctxt->gp_regs.fp_regs);
+ __fpsimd_restore_state(&vcpu->arch.ctxt.gp_regs.fp_regs);
+
+ /* Skip restoring fpexc32 for AArch64 guests */
+ if (!(read_sysreg(hcr_el2) & HCR_RW))
+ write_sysreg(vcpu->arch.ctxt.sys_regs[FPEXC32_EL2],
+ fpexc32_el2);
+}
+
static const char __hyp_panic_string[] = "HYP panic:\nPS:%08llx PC:%016llx ESR:%08llx\nFAR:%016llx HPFAR:%016llx PAR:%016llx\nVCPU:%p\n";
static void __hyp_text __hyp_call_panic_nvhe(u64 spsr, u64 elr, u64 par,
--
2.1.4
^ permalink raw reply related [flat|nested] 5+ messages in thread* [RFC PATCH 3/4] arm64: fpsimd: Split cpu field out from struct fpsimd_state
2018-03-21 17:47 [RFC PATCH 0/4] KVM: arm64: Optimise FPSIMD context switching Dave Martin
2018-03-21 17:47 ` [RFC PATCH 1/4] KVM: arm/arm64: Introduce kvm_arch_vcpu_run_pid_change Dave Martin
2018-03-21 17:47 ` [RFC PATCH 2/4] KVM: arm64: Convert lazy FPSIMD context switch trap to C Dave Martin
@ 2018-03-21 17:47 ` Dave Martin
2018-03-21 17:47 ` [RFC PATCH 4/4] KVM: arm64: Optimise FPSIMD context switching Dave Martin
3 siblings, 0 replies; 5+ messages in thread
From: Dave Martin @ 2018-03-21 17:47 UTC (permalink / raw)
To: linux-arm-kernel
In preparation for using a common representation of the FPSIMD
state for tasks and KVM vcpus, this patch separates out the "cpu"
field that is used to track the cpu on which the state was most
recently loaded.
This will allow common code to operate on task and vcpu contexts
without requiring the cpu field to be stored at the same offset
from the FPSIMD register data in both cases. This should avoid the
need for messing with the definition of those parts of struct
vcpu_arch that are exposed in the KVM user ABI.
Signed-off-by: Dave Martin <Dave.Martin@arm.com>
---
arch/arm64/include/asm/fpsimd.h | 29 ++------------------------
arch/arm64/include/asm/processor.h | 4 ++--
arch/arm64/kernel/fpsimd.c | 42 +++++++++++++++++++++-----------------
arch/arm64/kernel/ptrace.c | 10 ++++-----
arch/arm64/kernel/signal.c | 3 +--
arch/arm64/kernel/signal32.c | 3 +--
6 files changed, 34 insertions(+), 57 deletions(-)
diff --git a/arch/arm64/include/asm/fpsimd.h b/arch/arm64/include/asm/fpsimd.h
index 8857a0f..1bfc920 100644
--- a/arch/arm64/include/asm/fpsimd.h
+++ b/arch/arm64/include/asm/fpsimd.h
@@ -24,31 +24,6 @@
#include <linux/cache.h>
#include <linux/stddef.h>
-/*
- * FP/SIMD storage area has:
- * - FPSR and FPCR
- * - 32 128-bit data registers
- *
- * Note that user_fpsimd forms a prefix of this structure, which is
- * relied upon in the ptrace FP/SIMD accessors.
- */
-struct fpsimd_state {
- union {
- struct user_fpsimd_state user_fpsimd;
- struct {
- __uint128_t vregs[32];
- u32 fpsr;
- u32 fpcr;
- /*
- * For ptrace compatibility, pad to next 128-bit
- * boundary here if extending this struct.
- */
- };
- };
- /* the id of the last cpu to have restored this state */
- unsigned int cpu;
-};
-
#if defined(__KERNEL__) && defined(CONFIG_COMPAT)
/* Masks for extracting the FPSR and FPCR from the FPSCR */
#define VFP_FPSCR_STAT_MASK 0xf800009f
@@ -62,8 +37,8 @@ struct fpsimd_state {
struct task_struct;
-extern void fpsimd_save_state(struct fpsimd_state *state);
-extern void fpsimd_load_state(struct fpsimd_state *state);
+extern void fpsimd_save_state(struct user_fpsimd_state *state);
+extern void fpsimd_load_state(struct user_fpsimd_state *state);
extern void fpsimd_thread_switch(struct task_struct *next);
extern void fpsimd_flush_thread(void);
diff --git a/arch/arm64/include/asm/processor.h b/arch/arm64/include/asm/processor.h
index fce604e..4a04535 100644
--- a/arch/arm64/include/asm/processor.h
+++ b/arch/arm64/include/asm/processor.h
@@ -37,7 +37,6 @@
#include <linux/string.h>
#include <asm/alternative.h>
-#include <asm/fpsimd.h>
#include <asm/hw_breakpoint.h>
#include <asm/lse.h>
#include <asm/pgtable-hwdef.h>
@@ -107,7 +106,8 @@ struct thread_struct {
#ifdef CONFIG_COMPAT
unsigned long tp2_value;
#endif
- struct fpsimd_state fpsimd_state;
+ struct user_fpsimd_state fpsimd_state;
+ unsigned int fpsimd_cpu;
void *sve_state; /* SVE registers, if any */
unsigned int sve_vl; /* SVE vector length */
unsigned int sve_vl_onexec; /* SVE vl after next exec */
diff --git a/arch/arm64/kernel/fpsimd.c b/arch/arm64/kernel/fpsimd.c
index e7226c4..c4be311 100644
--- a/arch/arm64/kernel/fpsimd.c
+++ b/arch/arm64/kernel/fpsimd.c
@@ -64,7 +64,7 @@
* been loaded into its FPSIMD registers most recently, or whether it has
* been used to perform kernel mode NEON in the meantime.
*
- * For (a), we add a 'cpu' field to struct fpsimd_state, which gets updated to
+ * For (a), we add a fpsimd_cpu field to thread_struct, which gets updated to
* the id of the current CPU every time the state is loaded onto a CPU. For (b),
* we add the per-cpu variable 'fpsimd_last_state' (below), which contains the
* address of the userland FPSIMD state of the task that was loaded onto the CPU
@@ -73,7 +73,7 @@
* With this in place, we no longer have to restore the next FPSIMD state right
* when switching between tasks. Instead, we can defer this check to userland
* resume, at which time we verify whether the CPU's fpsimd_last_state and the
- * task's fpsimd_state.cpu are still mutually in sync. If this is the case, we
+ * task's fpsimd_cpu are still mutually in sync. If this is the case, we
* can omit the FPSIMD restore.
*
* As an optimization, we use the thread_info flag TIF_FOREIGN_FPSTATE to
@@ -90,14 +90,14 @@
* flag with local_bh_disable() unless softirqs are already masked.
*
* For a certain task, the sequence may look something like this:
- * - the task gets scheduled in; if both the task's fpsimd_state.cpu field
+ * - the task gets scheduled in; if both the task's fpsimd_cpu field
* contains the id of the current CPU, and the CPU's fpsimd_last_state per-cpu
* variable points to the task's fpsimd_state, the TIF_FOREIGN_FPSTATE flag is
* cleared, otherwise it is set;
*
* - the task returns to userland; if TIF_FOREIGN_FPSTATE is set, the task's
* userland FPSIMD state is copied from memory to the registers, the task's
- * fpsimd_state.cpu field is set to the id of the current CPU, the current
+ * fpsimd_cpu field is set to the id of the current CPU, the current
* CPU's fpsimd_last_state pointer is set to this task's fpsimd_state and the
* TIF_FOREIGN_FPSTATE flag is cleared;
*
@@ -115,7 +115,7 @@
* whatever is in the FPSIMD registers is not saved to memory, but discarded.
*/
struct fpsimd_last_state_struct {
- struct fpsimd_state *st;
+ struct user_fpsimd_state *st;
bool sve_in_use;
};
@@ -417,7 +417,7 @@ static void fpsimd_to_sve(struct task_struct *task)
{
unsigned int vq;
void *sst = task->thread.sve_state;
- struct fpsimd_state const *fst = &task->thread.fpsimd_state;
+ struct user_fpsimd_state const *fst = &task->thread.fpsimd_state;
unsigned int i;
if (!system_supports_sve())
@@ -443,7 +443,7 @@ static void sve_to_fpsimd(struct task_struct *task)
{
unsigned int vq;
void const *sst = task->thread.sve_state;
- struct fpsimd_state *fst = &task->thread.fpsimd_state;
+ struct user_fpsimd_state *fst = &task->thread.fpsimd_state;
unsigned int i;
if (!system_supports_sve())
@@ -539,7 +539,7 @@ void sve_sync_from_fpsimd_zeropad(struct task_struct *task)
{
unsigned int vq;
void *sst = task->thread.sve_state;
- struct fpsimd_state const *fst = &task->thread.fpsimd_state;
+ struct user_fpsimd_state const *fst = &task->thread.fpsimd_state;
unsigned int i;
if (!test_tsk_thread_flag(task, TIF_SVE))
@@ -908,10 +908,9 @@ void fpsimd_thread_switch(struct task_struct *next)
* the TIF_FOREIGN_FPSTATE flag so the state will be loaded
* upon the next return to userland.
*/
- struct fpsimd_state *st = &next->thread.fpsimd_state;
-
- if (__this_cpu_read(fpsimd_last_state.st) == st
- && st->cpu == smp_processor_id())
+ if (__this_cpu_read(fpsimd_last_state.st) ==
+ &next->thread.fpsimd_state
+ && next->thread.fpsimd_cpu == smp_processor_id())
clear_tsk_thread_flag(next, TIF_FOREIGN_FPSTATE);
else
set_tsk_thread_flag(next, TIF_FOREIGN_FPSTATE);
@@ -927,7 +926,8 @@ void fpsimd_flush_thread(void)
local_bh_disable();
- memset(¤t->thread.fpsimd_state, 0, sizeof(struct fpsimd_state));
+ memset(¤t->thread.fpsimd_state, 0,
+ sizeof current->thread.fpsimd_state);
fpsimd_flush_task_state(current);
if (system_supports_sve()) {
@@ -1004,11 +1004,10 @@ static void fpsimd_bind_to_cpu(void)
{
struct fpsimd_last_state_struct *last =
this_cpu_ptr(&fpsimd_last_state);
- struct fpsimd_state *st = ¤t->thread.fpsimd_state;
- last->st = st;
+ last->st = ¤t->thread.fpsimd_state;
last->sve_in_use = test_thread_flag(TIF_SVE);
- st->cpu = smp_processor_id();
+ current->thread.fpsimd_cpu = smp_processor_id();
}
/*
@@ -1043,7 +1042,7 @@ void fpsimd_update_current_state(struct user_fpsimd_state const *state)
local_bh_disable();
- current->thread.fpsimd_state.user_fpsimd = *state;
+ current->thread.fpsimd_state = *state;
if (system_supports_sve() && test_thread_flag(TIF_SVE))
fpsimd_to_sve(current);
@@ -1055,12 +1054,17 @@ void fpsimd_update_current_state(struct user_fpsimd_state const *state)
local_bh_enable();
}
+void fpsimd_flush_state(unsigned int *cpu)
+{
+ *cpu = NR_CPUS;
+}
+
/*
* Invalidate live CPU copies of task t's FPSIMD state
*/
void fpsimd_flush_task_state(struct task_struct *t)
{
- t->thread.fpsimd_state.cpu = NR_CPUS;
+ fpsimd_flush_state(&t->thread.fpsimd_cpu);
}
static inline void fpsimd_flush_cpu_state(void)
@@ -1159,7 +1163,7 @@ EXPORT_SYMBOL(kernel_neon_end);
#ifdef CONFIG_EFI
-static DEFINE_PER_CPU(struct fpsimd_state, efi_fpsimd_state);
+static DEFINE_PER_CPU(struct user_fpsimd_state, efi_fpsimd_state);
static DEFINE_PER_CPU(bool, efi_fpsimd_state_used);
static DEFINE_PER_CPU(bool, efi_sve_state_used);
diff --git a/arch/arm64/kernel/ptrace.c b/arch/arm64/kernel/ptrace.c
index 9ae31f7..fdeaba0de 100644
--- a/arch/arm64/kernel/ptrace.c
+++ b/arch/arm64/kernel/ptrace.c
@@ -629,7 +629,7 @@ static int __fpr_get(struct task_struct *target,
sve_sync_to_fpsimd(target);
- uregs = &target->thread.fpsimd_state.user_fpsimd;
+ uregs = &target->thread.fpsimd_state;
return user_regset_copyout(&pos, &count, &kbuf, &ubuf, uregs,
start_pos, start_pos + sizeof(*uregs));
@@ -660,14 +660,14 @@ static int __fpr_set(struct task_struct *target,
*/
sve_sync_to_fpsimd(target);
- newstate = target->thread.fpsimd_state.user_fpsimd;
+ newstate = target->thread.fpsimd_state;
ret = user_regset_copyin(&pos, &count, &kbuf, &ubuf, &newstate,
start_pos, start_pos + sizeof(newstate));
if (ret)
return ret;
- target->thread.fpsimd_state.user_fpsimd = newstate;
+ target->thread.fpsimd_state = newstate;
return ret;
}
@@ -1169,7 +1169,7 @@ static int compat_vfp_get(struct task_struct *target,
compat_ulong_t fpscr;
int ret, vregs_end_pos;
- uregs = &target->thread.fpsimd_state.user_fpsimd;
+ uregs = &target->thread.fpsimd_state;
if (target == current)
fpsimd_preserve_current_state();
@@ -1202,7 +1202,7 @@ static int compat_vfp_set(struct task_struct *target,
compat_ulong_t fpscr;
int ret, vregs_end_pos;
- uregs = &target->thread.fpsimd_state.user_fpsimd;
+ uregs = &target->thread.fpsimd_state;
vregs_end_pos = VFP_STATE_SIZE - sizeof(compat_ulong_t);
ret = user_regset_copyin(&pos, &count, &kbuf, &ubuf, uregs, 0,
diff --git a/arch/arm64/kernel/signal.c b/arch/arm64/kernel/signal.c
index f60c052..d026615 100644
--- a/arch/arm64/kernel/signal.c
+++ b/arch/arm64/kernel/signal.c
@@ -178,8 +178,7 @@ static void __user *apply_user_offset(
static int preserve_fpsimd_context(struct fpsimd_context __user *ctx)
{
- struct user_fpsimd_state const *fpsimd =
- ¤t->thread.fpsimd_state.user_fpsimd;
+ struct user_fpsimd_state const *fpsimd = ¤t->thread.fpsimd_state;
int err;
/* copy the FP and status/control registers */
diff --git a/arch/arm64/kernel/signal32.c b/arch/arm64/kernel/signal32.c
index 79feb86..4ea38d3 100644
--- a/arch/arm64/kernel/signal32.c
+++ b/arch/arm64/kernel/signal32.c
@@ -148,8 +148,7 @@ union __fpsimd_vreg {
static int compat_preserve_vfp_context(struct compat_vfp_sigframe __user *frame)
{
- struct user_fpsimd_state const *fpsimd =
- ¤t->thread.fpsimd_state.user_fpsimd;
+ struct user_fpsimd_state const *fpsimd = ¤t->thread.fpsimd_state;
compat_ulong_t magic = VFP_MAGIC;
compat_ulong_t size = VFP_STORAGE_SIZE;
compat_ulong_t fpscr, fpexc;
--
2.1.4
^ permalink raw reply related [flat|nested] 5+ messages in thread* [RFC PATCH 4/4] KVM: arm64: Optimise FPSIMD context switching
2018-03-21 17:47 [RFC PATCH 0/4] KVM: arm64: Optimise FPSIMD context switching Dave Martin
` (2 preceding siblings ...)
2018-03-21 17:47 ` [RFC PATCH 3/4] arm64: fpsimd: Split cpu field out from struct fpsimd_state Dave Martin
@ 2018-03-21 17:47 ` Dave Martin
3 siblings, 0 replies; 5+ messages in thread
From: Dave Martin @ 2018-03-21 17:47 UTC (permalink / raw)
To: linux-arm-kernel
Currently, the optimisations to move most of the KVM save/restore
logic to vcpu_load()/vcpu_put() don't work for FPSIMD, so the host
and guest FPSIMD state must be saved/restored rather deep in the
guest entry code, in the local_irq_disable() region. This avoids
some nasty interactions with kernel-mode NEON, which might
otherwise corrupt the registers since it may use them any time that
softirqs are enabled: blocking softirqs and/or kernel-mode NEON use
throughout the KVM run loop is not considered acceptable as a
solution for this.
This results in some unnecessary work, and extension to interrupt
blackout times around guest entry/exit. These issues will tend to
worsen when full SVE support is added to KVM, due to the
corresponding increase in the amount of register state that may
need to be switched.
This patch attempts to improve the situation by giving hyp and the
host FPSIMD context handling code some visibility of each other.
TIF_FOREIGN_FPSTATE, TIF_SVE and fpsimd_last_state_struct now
equally describe the host or vcpu context, so that kernel-mode NEON
can transparently save state in the correct place
(__this_cpu_read(fpsimd_last_state.st)) and signal that the
registers have been corrupted (via setting TIF_FOREIGN_FPSTATE).
Hyp is also taught how to use the same bookkeeping mechanism to
save the host context in the correct place when preparing to load
the vcpu FPSIMD state during an FPSIMD access trap, and the
required host task data (thread flags and user_fpsimd_state data)
is mapped to hyp as appropriate in order to enable this to work.
An additional vcpu flag vcpu->arch.fp_enabled is added to indicate
that the "current" FPSIMD context (as described by the above
bookkeeping data) is the vcpu's state, as opposed to the host's.
On entry via vcpu_load(), fp_enabled is initially false, indicating
that the hosts's FPSIMD state (or possibly garbage) remains in the
FPSIMD regs. On taking an FPSIMD trap to hyp, the host's state is
saved if required, the vcpu's state is loaded and fp_enabled is
set. On return to the KVM run loop, this flag is checked to update
the bookkeeping data appropriately before exposing it to the host
kernel by enabling interrupts.
fp_enabled and TIF_FOREIGN_FPSTATE need to be rechecked on each
guest entry, in order to determine whether to trap FPSIMD while in
the guest. This is how reload of the guest's state is guaranteed
after the registers are corrupted by kernel-mode NEON in a softirq.
Currently SVE is not supported for guests, so TIF_SVE is simply
cleared when loading the guest's state. A flag
vcpu->arch.host_sve_in_use is added to back up the state of this
flag while the guest's state is loaded.
In vcpu_put() it is necessary to tidy up so that there is no stray
guest state in the registers before the host context switch code
takes over: if fp_enabled was set, then the vcpu's FPSIMD state is
saved back to the vcpu struct and the register content is
invalidated. TIF_SVE is restored from the backed-up value in
host_sve_in_use.
Signed-off-by: Dave Martin <Dave.Martin@arm.com>
---
arch/arm64/include/asm/fpsimd.h | 5 ++++
arch/arm64/include/asm/kvm_host.h | 7 +++++
arch/arm64/kernel/fpsimd.c | 24 ++++++++++++-----
arch/arm64/kvm/hyp/switch.c | 44 +++++++++++++++++++------------
virt/kvm/arm/arm.c | 54 ++++++++++++++++++++++++++++++++++++---
5 files changed, 108 insertions(+), 26 deletions(-)
diff --git a/arch/arm64/include/asm/fpsimd.h b/arch/arm64/include/asm/fpsimd.h
index 1bfc920..dbe7340 100644
--- a/arch/arm64/include/asm/fpsimd.h
+++ b/arch/arm64/include/asm/fpsimd.h
@@ -40,6 +40,8 @@ struct task_struct;
extern void fpsimd_save_state(struct user_fpsimd_state *state);
extern void fpsimd_load_state(struct user_fpsimd_state *state);
+extern void task_fpsimd_save(void);
+
extern void fpsimd_thread_switch(struct task_struct *next);
extern void fpsimd_flush_thread(void);
@@ -48,7 +50,10 @@ extern void fpsimd_preserve_current_state(void);
extern void fpsimd_restore_current_state(void);
extern void fpsimd_update_current_state(struct user_fpsimd_state const *state);
+extern void fpsimd_bind_state_to_cpu(struct user_fpsimd_state *state);
+
extern void fpsimd_flush_task_state(struct task_struct *target);
+extern void fpsimd_flush_cpu_state(void);
extern void sve_flush_cpu_state(void);
/* Maximum VL that SVE VL-agnostic software can transparently support */
diff --git a/arch/arm64/include/asm/kvm_host.h b/arch/arm64/include/asm/kvm_host.h
index 596f8e4..56c60ff 100644
--- a/arch/arm64/include/asm/kvm_host.h
+++ b/arch/arm64/include/asm/kvm_host.h
@@ -30,6 +30,7 @@
#include <asm/kvm.h>
#include <asm/kvm_asm.h>
#include <asm/kvm_mmio.h>
+#include <asm/thread_info.h>
#define __KVM_HAVE_ARCH_INTC_INITIALIZED
@@ -235,6 +236,12 @@ struct kvm_vcpu_arch {
/* Pointer to host CPU context */
kvm_cpu_context_t *host_cpu_context;
+
+ struct thread_info *host_thread_info; /* hyp VA */
+ struct user_fpsimd_state *host_fpsimd_state; /* hyp VA */
+ bool host_sve_in_use; /* backup for host TIF_SVE while in guest */
+ bool fp_enabled;
+
struct {
/* {Break,watch}point registers */
struct kvm_guest_debug_arch regs;
diff --git a/arch/arm64/kernel/fpsimd.c b/arch/arm64/kernel/fpsimd.c
index c4be311..8e7dffc 100644
--- a/arch/arm64/kernel/fpsimd.c
+++ b/arch/arm64/kernel/fpsimd.c
@@ -268,13 +268,15 @@ static void task_fpsimd_load(void)
}
/*
- * Ensure current's FPSIMD/SVE storage in thread_struct is up to date
+ * Ensure current's FPSIMD/SVE storage in memory is up to date
* with respect to the CPU registers.
*
* Softirqs (and preemption) must be disabled.
*/
-static void task_fpsimd_save(void)
+void task_fpsimd_save(void)
{
+ struct user_fpsimd_state *st = __this_cpu_read(fpsimd_last_state.st);
+
WARN_ON(!in_softirq() && !irqs_disabled());
if (!test_thread_flag(TIF_FOREIGN_FPSTATE)) {
@@ -290,10 +292,9 @@ static void task_fpsimd_save(void)
return;
}
- sve_save_state(sve_pffr(current),
- ¤t->thread.fpsimd_state.fpsr);
+ sve_save_state(sve_pffr(current), &st->fpsr);
} else
- fpsimd_save_state(¤t->thread.fpsimd_state);
+ fpsimd_save_state(st);
}
}
@@ -1010,6 +1011,17 @@ static void fpsimd_bind_to_cpu(void)
current->thread.fpsimd_cpu = smp_processor_id();
}
+void fpsimd_bind_state_to_cpu(struct user_fpsimd_state *st)
+{
+ struct fpsimd_last_state_struct *last =
+ this_cpu_ptr(&fpsimd_last_state);
+
+ WARN_ON(!in_softirq() && !irqs_disabled());
+
+ last->st = st;
+ last->sve_in_use = false;
+}
+
/*
* Load the userland FPSIMD state of 'current' from memory, but only if the
* FPSIMD state already held in the registers is /not/ the most recent FPSIMD
@@ -1067,7 +1079,7 @@ void fpsimd_flush_task_state(struct task_struct *t)
fpsimd_flush_state(&t->thread.fpsimd_cpu);
}
-static inline void fpsimd_flush_cpu_state(void)
+void fpsimd_flush_cpu_state(void)
{
__this_cpu_write(fpsimd_last_state.st, NULL);
}
diff --git a/arch/arm64/kvm/hyp/switch.c b/arch/arm64/kvm/hyp/switch.c
index 01e4254..c177534 100644
--- a/arch/arm64/kvm/hyp/switch.c
+++ b/arch/arm64/kvm/hyp/switch.c
@@ -27,6 +27,7 @@
#include <asm/kvm_mmu.h>
#include <asm/fpsimd.h>
#include <asm/debug-monitors.h>
+#include <asm/thread_info.h>
static bool __hyp_text __fpsimd_enabled_nvhe(void)
{
@@ -47,24 +48,38 @@ bool __hyp_text __fpsimd_enabled(void)
return __fpsimd_is_enabled()();
}
-static void __hyp_text __activate_traps_vhe(void)
+static bool update_fp_enabled(struct kvm_vcpu *vcpu)
+{
+ if (vcpu->arch.host_thread_info->flags & _TIF_FOREIGN_FPSTATE)
+ vcpu->arch.fp_enabled = false;
+
+ return vcpu->arch.fp_enabled;
+}
+
+static void __hyp_text __activate_traps_vhe(struct kvm_vcpu *vcpu)
{
u64 val;
val = read_sysreg(cpacr_el1);
val |= CPACR_EL1_TTA;
- val &= ~(CPACR_EL1_FPEN | CPACR_EL1_ZEN);
+ val &= ~CPACR_EL1_ZEN;
+ if (!update_fp_enabled(vcpu))
+ val &= ~CPACR_EL1_FPEN;
+
write_sysreg(val, cpacr_el1);
write_sysreg(kvm_get_hyp_vector(), vbar_el1);
}
-static void __hyp_text __activate_traps_nvhe(void)
+static void __hyp_text __activate_traps_nvhe(struct kvm_vcpu *vcpu)
{
u64 val;
val = CPTR_EL2_DEFAULT;
- val |= CPTR_EL2_TTA | CPTR_EL2_TFP | CPTR_EL2_TZ;
+ val |= CPTR_EL2_TTA | CPTR_EL2_TZ;
+ if (!update_fp_enabled(vcpu))
+ val |= CPTR_EL2_TFP;
+
write_sysreg(val, cptr_el2);
}
@@ -107,7 +122,7 @@ static void __hyp_text __activate_traps(struct kvm_vcpu *vcpu)
write_sysreg(0, pmselr_el0);
write_sysreg(ARMV8_PMU_USERENR_MASK, pmuserenr_el0);
write_sysreg(vcpu->arch.mdcr_el2, mdcr_el2);
- __activate_traps_arch()();
+ __activate_traps_arch()(vcpu);
}
static void __hyp_text __deactivate_traps_vhe(void)
@@ -305,7 +320,6 @@ int __hyp_text __kvm_vcpu_run(struct kvm_vcpu *vcpu)
{
struct kvm_cpu_context *host_ctxt;
struct kvm_cpu_context *guest_ctxt;
- bool fp_enabled;
u64 exit_code;
vcpu = kern_hyp_va(vcpu);
@@ -409,8 +423,6 @@ int __hyp_text __kvm_vcpu_run(struct kvm_vcpu *vcpu)
}
}
- fp_enabled = __fpsimd_enabled();
-
__sysreg_save_guest_state(guest_ctxt);
__sysreg32_save_state(vcpu);
__timer_disable_traps(vcpu);
@@ -421,11 +433,6 @@ int __hyp_text __kvm_vcpu_run(struct kvm_vcpu *vcpu)
__sysreg_restore_host_state(host_ctxt);
- if (fp_enabled) {
- __fpsimd_save_state(&guest_ctxt->gp_regs.fp_regs);
- __fpsimd_restore_state(&host_ctxt->gp_regs.fp_regs);
- }
-
__debug_save_state(vcpu, kern_hyp_va(vcpu->arch.debug_ptr), guest_ctxt);
/*
* This must come after restoring the host sysregs, since a non-VHE
@@ -439,8 +446,6 @@ int __hyp_text __kvm_vcpu_run(struct kvm_vcpu *vcpu)
void __hyp_text __hyp_switch_fpsimd(u64 esr __always_unused,
struct kvm_vcpu *vcpu)
{
- kvm_cpu_context_t *host_ctxt;
-
if (has_vhe())
write_sysreg(read_sysreg(cpacr_el1) | CPACR_EL1_FPEN,
cpacr_el1);
@@ -450,14 +455,19 @@ void __hyp_text __hyp_switch_fpsimd(u64 esr __always_unused,
isb();
- host_ctxt = kern_hyp_va(vcpu->arch.host_cpu_context);
- __fpsimd_save_state(&host_ctxt->gp_regs.fp_regs);
+ if (vcpu->arch.host_fpsimd_state) {
+ __fpsimd_save_state(vcpu->arch.host_fpsimd_state);
+ vcpu->arch.host_fpsimd_state = NULL;
+ }
+
__fpsimd_restore_state(&vcpu->arch.ctxt.gp_regs.fp_regs);
/* Skip restoring fpexc32 for AArch64 guests */
if (!(read_sysreg(hcr_el2) & HCR_RW))
write_sysreg(vcpu->arch.ctxt.sys_regs[FPEXC32_EL2],
fpexc32_el2);
+
+ vcpu->arch.fp_enabled = true;
}
static const char __hyp_panic_string[] = "HYP panic:\nPS:%08llx PC:%016llx ESR:%08llx\nFAR:%016llx HPFAR:%016llx PAR:%016llx\nVCPU:%p\n";
diff --git a/virt/kvm/arm/arm.c b/virt/kvm/arm/arm.c
index 02a153a..2c69def 100644
--- a/virt/kvm/arm/arm.c
+++ b/virt/kvm/arm/arm.c
@@ -48,6 +48,7 @@
#include <asm/kvm_emulate.h>
#include <asm/kvm_coproc.h>
#include <asm/sections.h>
+#include <asm/thread_info.h>
#ifdef REQUIRES_VIRT
__asm__(".arch_extension virt");
@@ -359,6 +360,14 @@ void kvm_arch_vcpu_load(struct kvm_vcpu *vcpu, int cpu)
vcpu->cpu = cpu;
vcpu->arch.host_cpu_context = this_cpu_ptr(&kvm_host_cpu_state);
+ BUG_ON(system_supports_sve());
+ BUG_ON(!current->mm);
+
+ vcpu->arch.fp_enabled = false;
+ vcpu->arch.host_fpsimd_state =
+ kern_hyp_va(¤t->thread.fpsimd_state);
+ vcpu->arch.host_sve_in_use = !!test_thread_flag(TIF_SVE);
+
kvm_arm_set_running_vcpu(vcpu);
kvm_vgic_load(vcpu);
kvm_timer_vcpu_load(vcpu);
@@ -371,6 +380,21 @@ void kvm_arch_vcpu_put(struct kvm_vcpu *vcpu)
vcpu->cpu = -1;
+ local_bh_disable();
+
+ if (vcpu->arch.fp_enabled) {
+ task_fpsimd_save();
+ fpsimd_flush_cpu_state();
+ set_thread_flag(TIF_FOREIGN_FPSTATE);
+ }
+
+ if (vcpu->arch.host_sve_in_use)
+ set_thread_flag(TIF_SVE);
+ else
+ clear_thread_flag(TIF_SVE);
+
+ local_bh_enable();
+
kvm_arm_set_running_vcpu(NULL);
}
@@ -778,6 +802,13 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run)
if (static_branch_unlikely(&userspace_irqchip_in_use))
kvm_timer_sync_hwstate(vcpu);
+ if (vcpu->arch.fp_enabled) {
+ fpsimd_bind_state_to_cpu(
+ &vcpu->arch.ctxt.gp_regs.fp_regs);
+ clear_thread_flag(TIF_FOREIGN_FPSTATE);
+ clear_thread_flag(TIF_SVE);
+ }
+
/*
* We may have taken a host interrupt in HYP mode (ie
* while executing the guest). This interrupt is still
@@ -825,10 +856,27 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run)
#ifdef CONFIG_ARM64
int kvm_arch_vcpu_run_pid_change(struct kvm_vcpu *vcpu)
{
- struct task_struct *tsk = current;
+ int ret;
+
+ struct thread_info *ti = ¤t->thread_info;
+ struct user_fpsimd_state *fpsimd = ¤t->thread.fpsimd_state;
- /* Make sure the host task fpsimd state is visible to hyp: */
- return create_hyp_mappings(tsk, tsk + 1, PAGE_HYP);
+ /*
+ * Make sure the host task thread flags and fpsimd state are
+ * visible to hyp:
+ */
+ ret = create_hyp_mappings(ti, ti + 1, PAGE_HYP);
+ if (ret)
+ goto error;
+
+ ret = create_hyp_mappings(fpsimd, fpsimd + 1, PAGE_HYP);
+ if (ret)
+ goto error;
+
+ vcpu->arch.host_thread_info = kern_hyp_va(ti);
+ vcpu->arch.host_fpsimd_state = kern_hyp_va(fpsimd);
+error:
+ return ret;
}
#endif
--
2.1.4
^ permalink raw reply related [flat|nested] 5+ messages in thread