Move the code that manages floating-point state to a separate package

This change is inspired by Adin's cl/355256448.

PiperOrigin-RevId: 364695931

32 files changed, 528 insertions, 421 deletions

diff --git a/pkg/ring0/BUILD b/pkg/ring0/BUILD
index d1b14efdb..885958456 100644
--- a/pkg/ring0/BUILD
+++ b/pkg/ring0/BUILD
@@ -80,6 +80,7 @@ go_library(
         "//pkg/ring0/pagetables",
         "//pkg/safecopy",
         "//pkg/sentry/arch",
+        "//pkg/sentry/arch/fpu",
         "//pkg/usermem",
     ],
 )
diff --git a/pkg/ring0/defs.go b/pkg/ring0/defs.go
index e2561e4c2..b6e2012e8 100644
--- a/pkg/ring0/defs.go
+++ b/pkg/ring0/defs.go
@@ -17,6 +17,7 @@ package ring0
 import (
 	"gvisor.dev/gvisor/pkg/ring0/pagetables"
 	"gvisor.dev/gvisor/pkg/sentry/arch"
+	"gvisor.dev/gvisor/pkg/sentry/arch/fpu"
 )
 
 // Kernel is a global kernel object.
@@ -96,7 +97,7 @@ type SwitchOpts struct {
 
 	// FloatingPointState is a byte pointer where floating point state is
 	// saved and restored.
-	FloatingPointState *byte
+	FloatingPointState *fpu.State
 
 	// PageTables are the application page tables.
 	PageTables *pagetables.PageTables
diff --git a/pkg/ring0/gen_offsets/BUILD b/pkg/ring0/gen_offsets/BUILD
index 15b93d61c..f421e1687 100644
--- a/pkg/ring0/gen_offsets/BUILD
+++ b/pkg/ring0/gen_offsets/BUILD
@@ -35,6 +35,7 @@ go_binary(
         "//pkg/cpuid",
         "//pkg/ring0/pagetables",
         "//pkg/sentry/arch",
+        "//pkg/sentry/arch/fpu",
         "//pkg/usermem",
     ],
 )
diff --git a/pkg/ring0/kernel_amd64.go b/pkg/ring0/kernel_amd64.go
index 36a60700e..33c259757 100644
--- a/pkg/ring0/kernel_amd64.go
+++ b/pkg/ring0/kernel_amd64.go
@@ -239,17 +239,17 @@ func (c *CPU) SwitchToUser(switchOpts SwitchOpts) (vector Vector) {
 	regs.Ss = uint64(Udata)   // Ditto.
 
 	// Perform the switch.
-	swapgs()                                         // GS will be swapped on return.
-	WriteFS(uintptr(regs.Fs_base))                   // escapes: no. Set application FS.
-	WriteGS(uintptr(regs.Gs_base))                   // escapes: no. Set application GS.
-	LoadFloatingPoint(switchOpts.FloatingPointState) // escapes: no. Copy in floating point.
+	swapgs()                                                       // GS will be swapped on return.
+	WriteFS(uintptr(regs.Fs_base))                                 // escapes: no. Set application FS.
+	WriteGS(uintptr(regs.Gs_base))                                 // escapes: no. Set application GS.
+	LoadFloatingPoint(switchOpts.FloatingPointState.BytePointer()) // escapes: no. Copy in floating point.
 	if switchOpts.FullRestore {
 		vector = iret(c, regs, uintptr(userCR3))
 	} else {
 		vector = sysret(c, regs, uintptr(userCR3))
 	}
-	SaveFloatingPoint(switchOpts.FloatingPointState) // escapes: no. Copy out floating point.
-	WriteFS(uintptr(c.registers.Fs_base))            // escapes: no. Restore kernel FS.
+	SaveFloatingPoint(switchOpts.FloatingPointState.BytePointer()) // escapes: no. Copy out floating point.
+	WriteFS(uintptr(c.registers.Fs_base))                          // escapes: no. Restore kernel FS.
 	return
 }
 
diff --git a/pkg/ring0/kernel_arm64.go b/pkg/ring0/kernel_arm64.go
index 41909b3a0..7975e5f92 100644
--- a/pkg/ring0/kernel_arm64.go
+++ b/pkg/ring0/kernel_arm64.go
@@ -62,7 +62,7 @@ func IsCanonical(addr uint64) bool {
 //go:nosplit
 func (c *CPU) SwitchToUser(switchOpts SwitchOpts) (vector Vector) {
 	storeAppASID(uintptr(switchOpts.UserASID))
-	storeEl0Fpstate(switchOpts.FloatingPointState)
+	storeEl0Fpstate(switchOpts.FloatingPointState.BytePointer())
 
 	if switchOpts.Flush {
 		FlushTlbByASID(uintptr(switchOpts.UserASID))
@@ -82,7 +82,7 @@ func (c *CPU) SwitchToUser(switchOpts SwitchOpts) (vector Vector) {
 
 	fpDisableTrap = CPACREL1()
 	if fpDisableTrap != 0 {
-		SaveFloatingPoint(switchOpts.FloatingPointState)
+		SaveFloatingPoint(switchOpts.FloatingPointState.BytePointer())
 	}
 
 	vector = c.vecCode
diff --git a/pkg/sentry/arch/BUILD b/pkg/sentry/arch/BUILD
index 85278b389..f660f1614 100644
--- a/pkg/sentry/arch/BUILD
+++ b/pkg/sentry/arch/BUILD
@@ -9,7 +9,6 @@ go_library(
         "arch.go",
         "arch_aarch64.go",
         "arch_amd64.go",
-        "arch_amd64.s",
         "arch_arm64.go",
         "arch_state_x86.go",
         "arch_x86.go",
@@ -36,8 +35,8 @@ go_library(
         "//pkg/log",
         "//pkg/marshal",
         "//pkg/marshal/primitive",
+        "//pkg/sentry/arch/fpu",
         "//pkg/sentry/limits",
-        "//pkg/sync",
         "//pkg/syserror",
         "//pkg/usermem",
         "@org_golang_x_sys//unix:go_default_library",
diff --git a/pkg/sentry/arch/arch.go b/pkg/sentry/arch/arch.go
index dd2effdf9..921151137 100644
--- a/pkg/sentry/arch/arch.go
+++ b/pkg/sentry/arch/arch.go
@@ -24,6 +24,7 @@ import (
 	"gvisor.dev/gvisor/pkg/cpuid"
 	"gvisor.dev/gvisor/pkg/log"
 	"gvisor.dev/gvisor/pkg/marshal"
+	"gvisor.dev/gvisor/pkg/sentry/arch/fpu"
 	"gvisor.dev/gvisor/pkg/sentry/limits"
 	"gvisor.dev/gvisor/pkg/usermem"
 )
@@ -50,12 +51,6 @@ func (a Arch) String() string {
 	}
 }
 
-// FloatingPointData is a generic type, and will always be passed as a pointer.
-// We rely on the individual arch implementations to meet all the necessary
-// requirements. For example, on x86 the region must be 16-byte aligned and 512
-// bytes in size.
-type FloatingPointData byte
-
 // Context provides architecture-dependent information for a specific thread.
 //
 // NOTE(b/34169503): Currently we use uintptr here to refer to a generic native
@@ -187,7 +182,7 @@ type Context interface {
 	ClearSingleStep()
 
 	// FloatingPointData will be passed to underlying save routines.
-	FloatingPointData() *FloatingPointData
+	FloatingPointData() *fpu.State
 
 	// NewMmapLayout returns a layout for a new MM, where MinAddr for the
 	// returned layout must be no lower than min, and MaxAddr for the returned
@@ -221,16 +216,6 @@ type Context interface {
 	// number of bytes read.
 	PtraceSetRegs(src io.Reader) (int, error)
 
-	// PtraceGetFPRegs implements ptrace(PTRACE_GETFPREGS) by writing the
-	// floating-point registers represented by this Context to addr in dst and
-	// returning the number of bytes written.
-	PtraceGetFPRegs(dst io.Writer) (int, error)
-
-	// PtraceSetFPRegs implements ptrace(PTRACE_SETFPREGS) by reading
-	// floating-point registers from src into this Context and returning the
-	// number of bytes read.
-	PtraceSetFPRegs(src io.Reader) (int, error)
-
 	// PtraceGetRegSet implements ptrace(PTRACE_GETREGSET) by writing the
 	// register set given by architecture-defined value regset from this
 	// Context to dst and returning the number of bytes written, which must be
@@ -365,18 +350,3 @@ func (a SyscallArgument) SizeT() uint {
 func (a SyscallArgument) ModeT() uint {
 	return uint(uint16(a.Value))
 }
-
-// ErrFloatingPoint indicates a failed restore due to unusable floating point
-// state.
-type ErrFloatingPoint struct {
-	// supported is the supported floating point state.
-	supported uint64
-
-	// saved is the saved floating point state.
-	saved uint64
-}
-
-// Error returns a sensible description of the restore error.
-func (e ErrFloatingPoint) Error() string {
-	return fmt.Sprintf("floating point state contains unsupported features; supported: %#x saved: %#x", e.supported, e.saved)
-}
diff --git a/pkg/sentry/arch/arch_aarch64.go b/pkg/sentry/arch/arch_aarch64.go
index fd73751e7..08789f517 100644
--- a/pkg/sentry/arch/arch_aarch64.go
+++ b/pkg/sentry/arch/arch_aarch64.go
@@ -23,6 +23,7 @@ import (
 	"gvisor.dev/gvisor/pkg/abi/linux"
 	"gvisor.dev/gvisor/pkg/cpuid"
 	"gvisor.dev/gvisor/pkg/log"
+	"gvisor.dev/gvisor/pkg/sentry/arch/fpu"
 	rpb "gvisor.dev/gvisor/pkg/sentry/arch/registers_go_proto"
 	"gvisor.dev/gvisor/pkg/syserror"
 )
@@ -40,65 +41,11 @@ type Registers struct {
 const (
 	// SyscallWidth is the width of insturctions.
 	SyscallWidth = 4
-
-	// fpsimdMagic is the magic number which is used in fpsimd_context.
-	fpsimdMagic = 0x46508001
-
-	// fpsimdContextSize is the size of fpsimd_context.
-	fpsimdContextSize = 0x210
 )
 
 // ARMTrapFlag is the mask for the trap flag.
 const ARMTrapFlag = uint64(1) << 21
 
-// aarch64FPState is aarch64 floating point state.
-type aarch64FPState []byte
-
-// initAarch64FPState sets up initial state.
-//
-// Related code in Linux kernel: fpsimd_flush_thread().
-// FPCR = FPCR_RM_RN (0x0 << 22).
-//
-// Currently, aarch64FPState is only a space of 0x210 length for fpstate.
-// The fp head is useless in sentry/ptrace/kvm.
-//
-func initAarch64FPState(data aarch64FPState) {
-}
-
-func newAarch64FPStateSlice() []byte {
-	return alignedBytes(4096, 16)[:fpsimdContextSize]
-}
-
-// newAarch64FPState returns an initialized floating point state.
-//
-// The returned state is large enough to store all floating point state
-// supported by host, even if the app won't use much of it due to a restricted
-// FeatureSet.
-func newAarch64FPState() aarch64FPState {
-	f := aarch64FPState(newAarch64FPStateSlice())
-	initAarch64FPState(f)
-	return f
-}
-
-// fork creates and returns an identical copy of the aarch64 floating point state.
-func (f aarch64FPState) fork() aarch64FPState {
-	n := aarch64FPState(newAarch64FPStateSlice())
-	copy(n, f)
-	return n
-}
-
-// FloatingPointData returns the raw data pointer.
-func (f aarch64FPState) FloatingPointData() *FloatingPointData {
-	return (*FloatingPointData)(&f[0])
-}
-
-// NewFloatingPointData returns a new floating point data blob.
-//
-// This is primarily for use in tests.
-func NewFloatingPointData() *FloatingPointData {
-	return (*FloatingPointData)(&(newAarch64FPState()[0]))
-}
-
 // State contains the common architecture bits for aarch64 (the build tag of this
 // file ensures it's only built on aarch64).
 //
@@ -108,7 +55,7 @@ type State struct {
 	Regs Registers
 
 	// Our floating point state.
-	aarch64FPState `state:"wait"`
+	fpState fpu.State `state:"wait"`
 
 	// FeatureSet is a pointer to the currently active feature set.
 	FeatureSet *cpuid.FeatureSet
@@ -162,10 +109,10 @@ func (s State) Proto() *rpb.Registers {
 // Fork creates and returns an identical copy of the state.
 func (s *State) Fork() State {
 	return State{
-		Regs:           s.Regs,
-		aarch64FPState: s.aarch64FPState.fork(),
-		FeatureSet:     s.FeatureSet,
-		OrigR0:         s.OrigR0,
+		Regs:       s.Regs,
+		fpState:    s.fpState.Fork(),
+		FeatureSet: s.FeatureSet,
+		OrigR0:     s.OrigR0,
 	}
 }
 
@@ -318,10 +265,10 @@ func New(arch Arch, fs *cpuid.FeatureSet) Context {
 	case ARM64:
 		return &context64{
 			State{
-				aarch64FPState: newAarch64FPState(),
-				FeatureSet:     fs,
+				fpState:    fpu.NewState(),
+				FeatureSet: fs,
 			},
-			[]aarch64FPState(nil),
+			[]fpu.State(nil),
 		}
 	}
 	panic(fmt.Sprintf("unknown architecture %v", arch))
diff --git a/pkg/sentry/arch/arch_amd64.go b/pkg/sentry/arch/arch_amd64.go
index 15d8ddb40..2571be60f 100644
--- a/pkg/sentry/arch/arch_amd64.go
+++ b/pkg/sentry/arch/arch_amd64.go
@@ -25,6 +25,7 @@ import (
 	"gvisor.dev/gvisor/pkg/cpuid"
 	"gvisor.dev/gvisor/pkg/marshal"
 	"gvisor.dev/gvisor/pkg/marshal/primitive"
+	"gvisor.dev/gvisor/pkg/sentry/arch/fpu"
 	"gvisor.dev/gvisor/pkg/sentry/limits"
 	"gvisor.dev/gvisor/pkg/usermem"
 )
@@ -105,7 +106,7 @@ const (
 // +stateify savable
 type context64 struct {
 	State
-	sigFPState []x86FPState // fpstate to be restored on sigreturn.
+	sigFPState []fpu.State // fpstate to be restored on sigreturn.
 }
 
 // Arch implements Context.Arch.
@@ -113,14 +114,18 @@ func (c *context64) Arch() Arch {
 	return AMD64
 }
 
-func (c *context64) copySigFPState() []x86FPState {
-	var sigfps []x86FPState
+func (c *context64) copySigFPState() []fpu.State {
+	var sigfps []fpu.State
 	for _, s := range c.sigFPState {
-		sigfps = append(sigfps, s.fork())
+		sigfps = append(sigfps, s.Fork())
 	}
 	return sigfps
 }
 
+func (c *context64) FloatingPointData() *fpu.State {
+	return &c.State.fpState
+}
+
 // Fork returns an exact copy of this context.
 func (c *context64) Fork() Context {
 	return &context64{
diff --git a/pkg/sentry/arch/arch_arm64.go b/pkg/sentry/arch/arch_arm64.go
index 0c61a3ff7..14ad9483b 100644
--- a/pkg/sentry/arch/arch_arm64.go
+++ b/pkg/sentry/arch/arch_arm64.go
@@ -24,6 +24,7 @@ import (
 	"gvisor.dev/gvisor/pkg/cpuid"
 	"gvisor.dev/gvisor/pkg/marshal"
 	"gvisor.dev/gvisor/pkg/marshal/primitive"
+	"gvisor.dev/gvisor/pkg/sentry/arch/fpu"
 	"gvisor.dev/gvisor/pkg/sentry/limits"
 	"gvisor.dev/gvisor/pkg/usermem"
 )
@@ -79,7 +80,7 @@ const (
 // +stateify savable
 type context64 struct {
 	State
-	sigFPState []aarch64FPState // fpstate to be restored on sigreturn.
+	sigFPState []fpu.State // fpstate to be restored on sigreturn.
 }
 
 // Arch implements Context.Arch.
@@ -87,10 +88,10 @@ func (c *context64) Arch() Arch {
 	return ARM64
 }
 
-func (c *context64) copySigFPState() []aarch64FPState {
-	var sigfps []aarch64FPState
+func (c *context64) copySigFPState() []fpu.State {
+	var sigfps []fpu.State
 	for _, s := range c.sigFPState {
-		sigfps = append(sigfps, s.fork())
+		sigfps = append(sigfps, s.Fork())
 	}
 	return sigfps
 }
@@ -286,3 +287,7 @@ func (c *context64) PtracePokeUser(addr, data uintptr) error {
 	// TODO(gvisor.dev/issue/1239): Full ptrace supporting for Arm64.
 	return nil
 }
+
+func (c *context64) FloatingPointData() *fpu.State {
+	return &c.State.fpState
+}
diff --git a/pkg/sentry/arch/arch_state_x86.go b/pkg/sentry/arch/arch_state_x86.go
index 840e53d33..b2b94c304 100644
--- a/pkg/sentry/arch/arch_state_x86.go
+++ b/pkg/sentry/arch/arch_state_x86.go
@@ -16,59 +16,7 @@
 
 package arch
 
-import (
-	"gvisor.dev/gvisor/pkg/cpuid"
-	"gvisor.dev/gvisor/pkg/usermem"
-)
-
-// XSTATE_BV does not exist if FXSAVE is used, but FXSAVE implicitly saves x87
-// and SSE state, so this is the equivalent XSTATE_BV value.
-const fxsaveBV uint64 = cpuid.XSAVEFeatureX87 | cpuid.XSAVEFeatureSSE
-
 // afterLoadFPState is invoked by afterLoad.
 func (s *State) afterLoadFPState() {
-	old := s.x86FPState
-
-	// Recreate the slice. This is done to ensure that it is aligned
-	// appropriately in memory, and large enough to accommodate any new
-	// state that may be saved by the new CPU. Even if extraneous new state
-	// is saved, the state we care about is guaranteed to be a subset of
-	// new state. Later optimizations can use less space when using a
-	// smaller state component bitmap. Intel SDM Volume 1 Chapter 13 has
-	// more info.
-	s.x86FPState = newX86FPState()
-
-	// x86FPState always contains all the FP state supported by the host.
-	// We may have come from a newer machine that supports additional state
-	// which we cannot restore.
-	//
-	// The x86 FP state areas are backwards compatible, so we can simply
-	// truncate the additional floating point state.
-	//
-	// Applications should not depend on the truncated state because it
-	// should relate only to features that were not exposed in the app
-	// FeatureSet. However, because we do not *prevent* them from using
-	// this state, we must verify here that there is no in-use state
-	// (according to XSTATE_BV) which we do not support.
-	if len(s.x86FPState) < len(old) {
-		// What do we support?
-		supportedBV := fxsaveBV
-		if fs := cpuid.HostFeatureSet(); fs.UseXsave() {
-			supportedBV = fs.ValidXCR0Mask()
-		}
-
-		// What was in use?
-		savedBV := fxsaveBV
-		if len(old) >= xstateBVOffset+8 {
-			savedBV = usermem.ByteOrder.Uint64(old[xstateBVOffset:])
-		}
-
-		// Supported features must be a superset of saved features.
-		if savedBV&^supportedBV != 0 {
-			panic(ErrFloatingPoint{supported: supportedBV, saved: savedBV})
-		}
-	}
-
-	// Copy to the new, aligned location.
-	copy(s.x86FPState, old)
+	s.fpState.AfterLoad()
 }
diff --git a/pkg/sentry/arch/arch_x86.go b/pkg/sentry/arch/arch_x86.go
index 641ada92f..e8e52d3a8 100644
--- a/pkg/sentry/arch/arch_x86.go
+++ b/pkg/sentry/arch/arch_x86.go
@@ -24,10 +24,9 @@ import (
 	"gvisor.dev/gvisor/pkg/abi/linux"
 	"gvisor.dev/gvisor/pkg/cpuid"
 	"gvisor.dev/gvisor/pkg/log"
+	"gvisor.dev/gvisor/pkg/sentry/arch/fpu"
 	rpb "gvisor.dev/gvisor/pkg/sentry/arch/registers_go_proto"
-	"gvisor.dev/gvisor/pkg/sync"
 	"gvisor.dev/gvisor/pkg/syserror"
-	"gvisor.dev/gvisor/pkg/usermem"
 )
 
 // Registers represents the CPU registers for this architecture.
@@ -111,57 +110,6 @@ var (
 	X86TrapFlag uint64 = (1 << 8)
 )
 
-// x86FPState is x86 floating point state.
-type x86FPState []byte
-
-// initX86FPState (defined in asm files) sets up initial state.
-func initX86FPState(data *FloatingPointData, useXsave bool)
-
-func newX86FPStateSlice() []byte {
-	size, align := cpuid.HostFeatureSet().ExtendedStateSize()
-	capacity := size
-	// Always use at least 4096 bytes.
-	//
-	// For the KVM platform, this state is a fixed 4096 bytes, so make sure
-	// that the underlying array is at _least_ that size otherwise we will
-	// corrupt random memory. This is not a pleasant thing to debug.
-	if capacity < 4096 {
-		capacity = 4096
-	}
-	return alignedBytes(capacity, align)[:size]
-}
-
-// newX86FPState returns an initialized floating point state.
-//
-// The returned state is large enough to store all floating point state
-// supported by host, even if the app won't use much of it due to a restricted
-// FeatureSet. Since they may still be able to see state not advertised by
-// CPUID we must ensure it does not contain any sentry state.
-func newX86FPState() x86FPState {
-	f := x86FPState(newX86FPStateSlice())
-	initX86FPState(f.FloatingPointData(), cpuid.HostFeatureSet().UseXsave())
-	return f
-}
-
-// fork creates and returns an identical copy of the x86 floating point state.
-func (f x86FPState) fork() x86FPState {
-	n := x86FPState(newX86FPStateSlice())
-	copy(n, f)
-	return n
-}
-
-// FloatingPointData returns the raw data pointer.
-func (f x86FPState) FloatingPointData() *FloatingPointData {
-	return (*FloatingPointData)(&f[0])
-}
-
-// NewFloatingPointData returns a new floating point data blob.
-//
-// This is primarily for use in tests.
-func NewFloatingPointData() *FloatingPointData {
-	return (*FloatingPointData)(&(newX86FPState()[0]))
-}
-
 // Proto returns a protobuf representation of the system registers in State.
 func (s State) Proto() *rpb.Registers {
 	regs := &rpb.AMD64Registers{
@@ -200,7 +148,7 @@ func (s State) Proto() *rpb.Registers {
 func (s *State) Fork() State {
 	return State{
 		Regs:       s.Regs,
-		x86FPState: s.x86FPState.fork(),
+		fpState:    s.fpState.Fork(),
 		FeatureSet: s.FeatureSet,
 	}
 }
@@ -393,149 +341,6 @@ func isValidSegmentBase(reg uint64) bool {
 	return reg < uint64(maxAddr64)
 }
 
-// ptraceFPRegsSize is the size in bytes of Linux's user_i387_struct, the type
-// manipulated by PTRACE_GETFPREGS and PTRACE_SETFPREGS on x86. Equivalently,
-// ptraceFPRegsSize is the size in bytes of the x86 FXSAVE area.
-const ptraceFPRegsSize = 512
-
-// PtraceGetFPRegs implements Context.PtraceGetFPRegs.
-func (s *State) PtraceGetFPRegs(dst io.Writer) (int, error) {
-	return dst.Write(s.x86FPState[:ptraceFPRegsSize])
-}
-
-// PtraceSetFPRegs implements Context.PtraceSetFPRegs.
-func (s *State) PtraceSetFPRegs(src io.Reader) (int, error) {
-	var f [ptraceFPRegsSize]byte
-	n, err := io.ReadFull(src, f[:])
-	if err != nil {
-		return 0, err
-	}
-	// Force reserved bits in MXCSR to 0. This is consistent with Linux.
-	sanitizeMXCSR(x86FPState(f[:]))
-	// N.B. this only copies the beginning of the FP state, which
-	// corresponds to the FXSAVE area.
-	copy(s.x86FPState, f[:])
-	return n, nil
-}
-
-const (
-	// mxcsrOffset is the offset in bytes of the MXCSR field from the start of
-	// the FXSAVE area. (Intel SDM Vol. 1, Table 10-2 "Format of an FXSAVE
-	// Area")
-	mxcsrOffset = 24
-
-	// mxcsrMaskOffset is the offset in bytes of the MXCSR_MASK field from the
-	// start of the FXSAVE area.
-	mxcsrMaskOffset = 28
-)
-
-var (
-	mxcsrMask     uint32
-	initMXCSRMask sync.Once
-)
-
-// sanitizeMXCSR coerces reserved bits in the MXCSR field of f to 0. ("FXRSTOR
-// generates a general-protection fault (#GP) in response to an attempt to set
-// any of the reserved bits of the MXCSR register." - Intel SDM Vol. 1, Section
-// 10.5.1.2 "SSE State")
-func sanitizeMXCSR(f x86FPState) {
-	mxcsr := usermem.ByteOrder.Uint32(f[mxcsrOffset:])
-	initMXCSRMask.Do(func() {
-		temp := x86FPState(alignedBytes(uint(ptraceFPRegsSize), 16))
-		initX86FPState(temp.FloatingPointData(), false /* useXsave */)
-		mxcsrMask = usermem.ByteOrder.Uint32(temp[mxcsrMaskOffset:])
-		if mxcsrMask == 0 {
-			// "If the value of the MXCSR_MASK field is 00000000H, then the
-			// MXCSR_MASK value is the default value of 0000FFBFH." - Intel SDM
-			// Vol. 1, Section 11.6.6 "Guidelines for Writing to the MXCSR
-			// Register"
-			mxcsrMask = 0xffbf
-		}
-	})
-	mxcsr &= mxcsrMask
-	usermem.ByteOrder.PutUint32(f[mxcsrOffset:], mxcsr)
-}
-
-const (
-	// minXstateBytes is the minimum size in bytes of an x86 XSAVE area, equal
-	// to the size of the XSAVE legacy area (512 bytes) plus the size of the
-	// XSAVE header (64 bytes). Equivalently, minXstateBytes is GDB's
-	// X86_XSTATE_SSE_SIZE.
-	minXstateBytes = 512 + 64
-
-	// userXstateXCR0Offset is the offset in bytes of the USER_XSTATE_XCR0_WORD
-	// field in Linux's struct user_xstateregs, which is the type manipulated
-	// by ptrace(PTRACE_GET/SETREGSET, NT_X86_XSTATE). Equivalently,
-	// userXstateXCR0Offset is GDB's I386_LINUX_XSAVE_XCR0_OFFSET.
-	userXstateXCR0Offset = 464
-
-	// xstateBVOffset is the offset in bytes of the XSTATE_BV field in an x86
-	// XSAVE area.
-	xstateBVOffset = 512
-
-	// xsaveHeaderZeroedOffset and xsaveHeaderZeroedBytes indicate parts of the
-	// XSAVE header that we coerce to zero: "Bytes 15:8 of the XSAVE header is
-	// a state-component bitmap called XCOMP_BV. ... Bytes 63:16 of the XSAVE
-	// header are reserved." - Intel SDM Vol. 1, Section 13.4.2 "XSAVE Header".
-	// Linux ignores XCOMP_BV, but it's able to recover from XRSTOR #GP
-	// exceptions resulting from invalid values; we aren't. Linux also never
-	// uses the compacted format when doing XSAVE and doesn't even define the
-	// compaction extensions to XSAVE as a CPU feature, so for simplicity we
-	// assume no one is using them.
-	xsaveHeaderZeroedOffset = 512 + 8
-	xsaveHeaderZeroedBytes  = 64 - 8
-)
-
-func (s *State) ptraceGetXstateRegs(dst io.Writer, maxlen int) (int, error) {
-	// N.B. s.x86FPState may contain more state than the application
-	// expects. We only copy the subset that would be in their XSAVE area.
-	ess, _ := s.FeatureSet.ExtendedStateSize()
-	f := make([]byte, ess)
-	copy(f, s.x86FPState)
-	// "The XSAVE feature set does not use bytes 511:416; bytes 463:416 are
-	// reserved." - Intel SDM Vol 1., Section 13.4.1 "Legacy Region of an XSAVE
-	// Area". Linux uses the first 8 bytes of this area to store the OS XSTATE
-	// mask. GDB relies on this: see
-	// gdb/x86-linux-nat.c:x86_linux_read_description().
-	usermem.ByteOrder.PutUint64(f[userXstateXCR0Offset:], s.FeatureSet.ValidXCR0Mask())
-	if len(f) > maxlen {
-		f = f[:maxlen]
-	}
-	return dst.Write(f)
-}
-
-func (s *State) ptraceSetXstateRegs(src io.Reader, maxlen int) (int, error) {
-	// Allow users to pass an xstate register set smaller than ours (they can
-	// mask bits out of XSTATE_BV), as long as it's at least minXstateBytes.
-	// Also allow users to pass a register set larger than ours; anything after
-	// their ExtendedStateSize will be ignored. (I think Linux technically
-	// permits setting a register set smaller than minXstateBytes, but it has
-	// the same silent truncation behavior in kernel/ptrace.c:ptrace_regset().)
-	if maxlen < minXstateBytes {
-		return 0, unix.EFAULT
-	}
-	ess, _ := s.FeatureSet.ExtendedStateSize()
-	if maxlen > int(ess) {
-		maxlen = int(ess)
-	}
-	f := make([]byte, maxlen)
-	if _, err := io.ReadFull(src, f); err != nil {
-		return 0, err
-	}
-	// Force reserved bits in MXCSR to 0. This is consistent with Linux.
-	sanitizeMXCSR(x86FPState(f))
-	// Users can't enable *more* XCR0 bits than what we, and the CPU, support.
-	xstateBV := usermem.ByteOrder.Uint64(f[xstateBVOffset:])
-	xstateBV &= s.FeatureSet.ValidXCR0Mask()
-	usermem.ByteOrder.PutUint64(f[xstateBVOffset:], xstateBV)
-	// Force XCOMP_BV and reserved bytes in the XSAVE header to 0.
-	reserved := f[xsaveHeaderZeroedOffset : xsaveHeaderZeroedOffset+xsaveHeaderZeroedBytes]
-	for i := range reserved {
-		reserved[i] = 0
-	}
-	return copy(s.x86FPState, f), nil
-}
-
 // Register sets defined in include/uapi/linux/elf.h.
 const (
 	_NT_PRSTATUS   = 1
@@ -552,12 +357,9 @@ func (s *State) PtraceGetRegSet(regset uintptr, dst io.Writer, maxlen int) (int,
 		}
 		return s.PtraceGetRegs(dst)
 	case _NT_PRFPREG:
-		if maxlen < ptraceFPRegsSize {
-			return 0, syserror.EFAULT
-		}
-		return s.PtraceGetFPRegs(dst)
+		return s.fpState.PtraceGetFPRegs(dst, maxlen)
 	case _NT_X86_XSTATE:
-		return s.ptraceGetXstateRegs(dst, maxlen)
+		return s.fpState.PtraceGetXstateRegs(dst, maxlen, s.FeatureSet)
 	default:
 		return 0, syserror.EINVAL
 	}
@@ -572,12 +374,9 @@ func (s *State) PtraceSetRegSet(regset uintptr, src io.Reader, maxlen int) (int,
 		}
 		return s.PtraceSetRegs(src)
 	case _NT_PRFPREG:
-		if maxlen < ptraceFPRegsSize {
-			return 0, syserror.EFAULT
-		}
-		return s.PtraceSetFPRegs(src)
+		return s.fpState.PtraceSetFPRegs(src, maxlen)
 	case _NT_X86_XSTATE:
-		return s.ptraceSetXstateRegs(src, maxlen)
+		return s.fpState.PtraceSetXstateRegs(src, maxlen, s.FeatureSet)
 	default:
 		return 0, syserror.EINVAL
 	}
@@ -609,10 +408,10 @@ func New(arch Arch, fs *cpuid.FeatureSet) Context {
 	case AMD64:
 		return &context64{
 			State{
-				x86FPState: newX86FPState(),
+				fpState:    fpu.NewState(),
 				FeatureSet: fs,
 			},
-			[]x86FPState(nil),
+			[]fpu.State(nil),
 		}
 	}
 	panic(fmt.Sprintf("unknown architecture %v", arch))
diff --git a/pkg/sentry/arch/arch_x86_impl.go b/pkg/sentry/arch/arch_x86_impl.go
index 0c73fcbfb..5d7b99bd9 100644
--- a/pkg/sentry/arch/arch_x86_impl.go
+++ b/pkg/sentry/arch/arch_x86_impl.go
@@ -18,6 +18,7 @@ package arch
 
 import (
 	"gvisor.dev/gvisor/pkg/cpuid"
+	"gvisor.dev/gvisor/pkg/sentry/arch/fpu"
 )
 
 // State contains the common architecture bits for X86 (the build tag of this
@@ -29,7 +30,7 @@ type State struct {
 	Regs Registers
 
 	// Our floating point state.
-	x86FPState `state:"wait"`
+	fpState fpu.State `state:"wait"`
 
 	// FeatureSet is a pointer to the currently active feature set.
 	FeatureSet *cpuid.FeatureSet
diff --git a/pkg/sentry/arch/fpu/BUILD b/pkg/sentry/arch/fpu/BUILD
new file mode 100644
index 000000000..0a5395267
--- /dev/null
+++ b/pkg/sentry/arch/fpu/BUILD
@@ -0,0 +1,21 @@
+load("//tools:defs.bzl", "go_library")
+
+package(licenses = ["notice"])
+
+go_library(
+    name = "fpu",
+    srcs = [
+        "fpu.go",
+        "fpu_amd64.go",
+        "fpu_amd64.s",
+        "fpu_arm64.go",
+    ],
+    visibility = ["//:sandbox"],
+    deps = [
+        "//pkg/cpuid",
+        "//pkg/sync",
+        "//pkg/syserror",
+        "//pkg/usermem",
+        "@org_golang_x_sys//unix:go_default_library",
+    ],
+)
diff --git a/pkg/sentry/arch/fpu/fpu.go b/pkg/sentry/arch/fpu/fpu.go
new file mode 100644
index 000000000..867d309a3
--- /dev/null
+++ b/pkg/sentry/arch/fpu/fpu.go
@@ -0,0 +1,54 @@
+// Copyright 2021 The gVisor Authors.
+//
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+//     http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+// Package fpu provides basic floating point helpers.
+package fpu
+
+import (
+	"fmt"
+	"reflect"
+)
+
+// State represents floating point state.
+//
+// This is a simple byte slice, but may have architecture-specific methods
+// attached to it.
+type State []byte
+
+// ErrLoadingState indicates a failed restore due to unusable floating point
+// state.
+type ErrLoadingState struct {
+	// supported is the supported floating point state.
+	supportedFeatures uint64
+
+	// saved is the saved floating point state.
+	savedFeatures uint64
+}
+
+// Error returns a sensible description of the restore error.
+func (e ErrLoadingState) Error() string {
+	return fmt.Sprintf("floating point state contains unsupported features; supported: %#x saved: %#x", e.supportedFeatures, e.savedFeatures)
+}
+
+// alignedBytes returns a slice of size bytes, aligned in memory to the given
+// alignment. This is used because we require certain structures to be aligned
+// in a specific way (for example, the X86 floating point data).
+func alignedBytes(size, alignment uint) []byte {
+	data := make([]byte, size+alignment-1)
+	offset := uint(reflect.ValueOf(data).Index(0).Addr().Pointer() % uintptr(alignment))
+	if offset == 0 {
+		return data[:size:size]
+	}
+	return data[alignment-offset:][:size:size]
+}
diff --git a/pkg/sentry/arch/fpu/fpu_amd64.go b/pkg/sentry/arch/fpu/fpu_amd64.go
new file mode 100644
index 000000000..3a62f51be
--- /dev/null
+++ b/pkg/sentry/arch/fpu/fpu_amd64.go
@@ -0,0 +1,280 @@
+// Copyright 2021 The gVisor Authors.
+//
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+//     http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+// +build amd64 i386
+
+package fpu
+
+import (
+	"io"
+
+	"golang.org/x/sys/unix"
+	"gvisor.dev/gvisor/pkg/cpuid"
+	"gvisor.dev/gvisor/pkg/sync"
+	"gvisor.dev/gvisor/pkg/syserror"
+	"gvisor.dev/gvisor/pkg/usermem"
+)
+
+// initX86FPState (defined in asm files) sets up initial state.
+func initX86FPState(data *byte, useXsave bool)
+
+func newX86FPStateSlice() State {
+	size, align := cpuid.HostFeatureSet().ExtendedStateSize()
+	capacity := size
+	// Always use at least 4096 bytes.
+	//
+	// For the KVM platform, this state is a fixed 4096 bytes, so make sure
+	// that the underlying array is at _least_ that size otherwise we will
+	// corrupt random memory. This is not a pleasant thing to debug.
+	if capacity < 4096 {
+		capacity = 4096
+	}
+	return alignedBytes(capacity, align)[:size]
+}
+
+// NewState returns an initialized floating point state.
+//
+// The returned state is large enough to store all floating point state
+// supported by host, even if the app won't use much of it due to a restricted
+// FeatureSet. Since they may still be able to see state not advertised by
+// CPUID we must ensure it does not contain any sentry state.
+func NewState() State {
+	f := newX86FPStateSlice()
+	initX86FPState(&f[0], cpuid.HostFeatureSet().UseXsave())
+	return f
+}
+
+// Fork creates and returns an identical copy of the x86 floating point state.
+func (s *State) Fork() State {
+	n := newX86FPStateSlice()
+	copy(n, *s)
+	return n
+}
+
+// ptraceFPRegsSize is the size in bytes of Linux's user_i387_struct, the type
+// manipulated by PTRACE_GETFPREGS and PTRACE_SETFPREGS on x86. Equivalently,
+// ptraceFPRegsSize is the size in bytes of the x86 FXSAVE area.
+const ptraceFPRegsSize = 512
+
+// PtraceGetFPRegs implements Context.PtraceGetFPRegs.
+func (s *State) PtraceGetFPRegs(dst io.Writer, maxlen int) (int, error) {
+	if maxlen < ptraceFPRegsSize {
+		return 0, syserror.EFAULT
+	}
+
+	return dst.Write((*s)[:ptraceFPRegsSize])
+}
+
+// PtraceSetFPRegs implements Context.PtraceSetFPRegs.
+func (s *State) PtraceSetFPRegs(src io.Reader, maxlen int) (int, error) {
+	if maxlen < ptraceFPRegsSize {
+		return 0, syserror.EFAULT
+	}
+
+	var f [ptraceFPRegsSize]byte
+	n, err := io.ReadFull(src, f[:])
+	if err != nil {
+		return 0, err
+	}
+	// Force reserved bits in MXCSR to 0. This is consistent with Linux.
+	sanitizeMXCSR(State(f[:]))
+	// N.B. this only copies the beginning of the FP state, which
+	// corresponds to the FXSAVE area.
+	copy(*s, f[:])
+	return n, nil
+}
+
+const (
+	// mxcsrOffset is the offset in bytes of the MXCSR field from the start of
+	// the FXSAVE area. (Intel SDM Vol. 1, Table 10-2 "Format of an FXSAVE
+	// Area")
+	mxcsrOffset = 24
+
+	// mxcsrMaskOffset is the offset in bytes of the MXCSR_MASK field from the
+	// start of the FXSAVE area.
+	mxcsrMaskOffset = 28
+)
+
+var (
+	mxcsrMask     uint32
+	initMXCSRMask sync.Once
+)
+
+const (
+	// minXstateBytes is the minimum size in bytes of an x86 XSAVE area, equal
+	// to the size of the XSAVE legacy area (512 bytes) plus the size of the
+	// XSAVE header (64 bytes). Equivalently, minXstateBytes is GDB's
+	// X86_XSTATE_SSE_SIZE.
+	minXstateBytes = 512 + 64
+
+	// userXstateXCR0Offset is the offset in bytes of the USER_XSTATE_XCR0_WORD
+	// field in Linux's struct user_xstateregs, which is the type manipulated
+	// by ptrace(PTRACE_GET/SETREGSET, NT_X86_XSTATE). Equivalently,
+	// userXstateXCR0Offset is GDB's I386_LINUX_XSAVE_XCR0_OFFSET.
+	userXstateXCR0Offset = 464
+
+	// xstateBVOffset is the offset in bytes of the XSTATE_BV field in an x86
+	// XSAVE area.
+	xstateBVOffset = 512
+
+	// xsaveHeaderZeroedOffset and xsaveHeaderZeroedBytes indicate parts of the
+	// XSAVE header that we coerce to zero: "Bytes 15:8 of the XSAVE header is
+	// a state-component bitmap called XCOMP_BV. ... Bytes 63:16 of the XSAVE
+	// header are reserved." - Intel SDM Vol. 1, Section 13.4.2 "XSAVE Header".
+	// Linux ignores XCOMP_BV, but it's able to recover from XRSTOR #GP
+	// exceptions resulting from invalid values; we aren't. Linux also never
+	// uses the compacted format when doing XSAVE and doesn't even define the
+	// compaction extensions to XSAVE as a CPU feature, so for simplicity we
+	// assume no one is using them.
+	xsaveHeaderZeroedOffset = 512 + 8
+	xsaveHeaderZeroedBytes  = 64 - 8
+)
+
+// sanitizeMXCSR coerces reserved bits in the MXCSR field of f to 0. ("FXRSTOR
+// generates a general-protection fault (#GP) in response to an attempt to set
+// any of the reserved bits of the MXCSR register." - Intel SDM Vol. 1, Section
+// 10.5.1.2 "SSE State")
+func sanitizeMXCSR(f State) {
+	mxcsr := usermem.ByteOrder.Uint32(f[mxcsrOffset:])
+	initMXCSRMask.Do(func() {
+		temp := State(alignedBytes(uint(ptraceFPRegsSize), 16))
+		initX86FPState(&temp[0], false /* useXsave */)
+		mxcsrMask = usermem.ByteOrder.Uint32(temp[mxcsrMaskOffset:])
+		if mxcsrMask == 0 {
+			// "If the value of the MXCSR_MASK field is 00000000H, then the
+			// MXCSR_MASK value is the default value of 0000FFBFH." - Intel SDM
+			// Vol. 1, Section 11.6.6 "Guidelines for Writing to the MXCSR
+			// Register"
+			mxcsrMask = 0xffbf
+		}
+	})
+	mxcsr &= mxcsrMask
+	usermem.ByteOrder.PutUint32(f[mxcsrOffset:], mxcsr)
+}
+
+// PtraceGetXstateRegs implements ptrace(PTRACE_GETREGS, NT_X86_XSTATE) by
+// writing the floating point registers from this state to dst and returning the
+// number of bytes written, which must be less than or equal to maxlen.
+func (s *State) PtraceGetXstateRegs(dst io.Writer, maxlen int, featureSet *cpuid.FeatureSet) (int, error) {
+	// N.B. s.x86FPState may contain more state than the application
+	// expects. We only copy the subset that would be in their XSAVE area.
+	ess, _ := featureSet.ExtendedStateSize()
+	f := make([]byte, ess)
+	copy(f, *s)
+	// "The XSAVE feature set does not use bytes 511:416; bytes 463:416 are
+	// reserved." - Intel SDM Vol 1., Section 13.4.1 "Legacy Region of an XSAVE
+	// Area". Linux uses the first 8 bytes of this area to store the OS XSTATE
+	// mask. GDB relies on this: see
+	// gdb/x86-linux-nat.c:x86_linux_read_description().
+	usermem.ByteOrder.PutUint64(f[userXstateXCR0Offset:], featureSet.ValidXCR0Mask())
+	if len(f) > maxlen {
+		f = f[:maxlen]
+	}
+	return dst.Write(f)
+}
+
+// PtraceSetXstateRegs implements ptrace(PTRACE_SETREGS, NT_X86_XSTATE) by
+// reading floating point registers from src and returning the number of bytes
+// read, which must be less than or equal to maxlen.
+func (s *State) PtraceSetXstateRegs(src io.Reader, maxlen int, featureSet *cpuid.FeatureSet) (int, error) {
+	// Allow users to pass an xstate register set smaller than ours (they can
+	// mask bits out of XSTATE_BV), as long as it's at least minXstateBytes.
+	// Also allow users to pass a register set larger than ours; anything after
+	// their ExtendedStateSize will be ignored. (I think Linux technically
+	// permits setting a register set smaller than minXstateBytes, but it has
+	// the same silent truncation behavior in kernel/ptrace.c:ptrace_regset().)
+	if maxlen < minXstateBytes {
+		return 0, unix.EFAULT
+	}
+	ess, _ := featureSet.ExtendedStateSize()
+	if maxlen > int(ess) {
+		maxlen = int(ess)
+	}
+	f := make([]byte, maxlen)
+	if _, err := io.ReadFull(src, f); err != nil {
+		return 0, err
+	}
+	// Force reserved bits in MXCSR to 0. This is consistent with Linux.
+	sanitizeMXCSR(State(f))
+	// Users can't enable *more* XCR0 bits than what we, and the CPU, support.
+	xstateBV := usermem.ByteOrder.Uint64(f[xstateBVOffset:])
+	xstateBV &= featureSet.ValidXCR0Mask()
+	usermem.ByteOrder.PutUint64(f[xstateBVOffset:], xstateBV)
+	// Force XCOMP_BV and reserved bytes in the XSAVE header to 0.
+	reserved := f[xsaveHeaderZeroedOffset : xsaveHeaderZeroedOffset+xsaveHeaderZeroedBytes]
+	for i := range reserved {
+		reserved[i] = 0
+	}
+	return copy(*s, f), nil
+}
+
+// BytePointer returns a pointer to the first byte of the state.
+//
+//go:nosplit
+func (s *State) BytePointer() *byte {
+	return &(*s)[0]
+}
+
+// XSTATE_BV does not exist if FXSAVE is used, but FXSAVE implicitly saves x87
+// and SSE state, so this is the equivalent XSTATE_BV value.
+const fxsaveBV uint64 = cpuid.XSAVEFeatureX87 | cpuid.XSAVEFeatureSSE
+
+// AfterLoad converts the loaded state to the format that compatible with the
+// current processor.
+func (s *State) AfterLoad() {
+	old := *s
+
+	// Recreate the slice. This is done to ensure that it is aligned
+	// appropriately in memory, and large enough to accommodate any new
+	// state that may be saved by the new CPU. Even if extraneous new state
+	// is saved, the state we care about is guaranteed to be a subset of
+	// new state. Later optimizations can use less space when using a
+	// smaller state component bitmap. Intel SDM Volume 1 Chapter 13 has
+	// more info.
+	*s = NewState()
+
+	// x86FPState always contains all the FP state supported by the host.
+	// We may have come from a newer machine that supports additional state
+	// which we cannot restore.
+	//
+	// The x86 FP state areas are backwards compatible, so we can simply
+	// truncate the additional floating point state.
+	//
+	// Applications should not depend on the truncated state because it
+	// should relate only to features that were not exposed in the app
+	// FeatureSet. However, because we do not *prevent* them from using
+	// this state, we must verify here that there is no in-use state
+	// (according to XSTATE_BV) which we do not support.
+	if len(*s) < len(old) {
+		// What do we support?
+		supportedBV := fxsaveBV
+		if fs := cpuid.HostFeatureSet(); fs.UseXsave() {
+			supportedBV = fs.ValidXCR0Mask()
+		}
+
+		// What was in use?
+		savedBV := fxsaveBV
+		if len(old) >= xstateBVOffset+8 {
+			savedBV = usermem.ByteOrder.Uint64(old[xstateBVOffset:])
+		}
+
+		// Supported features must be a superset of saved features.
+		if savedBV&^supportedBV != 0 {
+			panic(ErrLoadingState{supportedFeatures: supportedBV, savedFeatures: savedBV})
+		}
+	}
+
+	// Copy to the new, aligned location.
+	copy(*s, old)
+}
diff --git a/pkg/sentry/arch/arch_amd64.s b/pkg/sentry/arch/fpu/fpu_amd64.s
index 6c10336e7..6c10336e7 100644
--- a/pkg/sentry/arch/arch_amd64.s
+++ b/pkg/sentry/arch/fpu/fpu_amd64.s
diff --git a/pkg/sentry/arch/fpu/fpu_arm64.go b/pkg/sentry/arch/fpu/fpu_arm64.go
new file mode 100644
index 000000000..d2f62631d
--- /dev/null
+++ b/pkg/sentry/arch/fpu/fpu_arm64.go
@@ -0,0 +1,63 @@
+// Copyright 2020 The gVisor Authors.
+//
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+//     http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+// +build arm64
+
+package fpu
+
+const (
+	// fpsimdMagic is the magic number which is used in fpsimd_context.
+	fpsimdMagic = 0x46508001
+
+	// fpsimdContextSize is the size of fpsimd_context.
+	fpsimdContextSize = 0x210
+)
+
+// initAarch64FPState sets up initial state.
+//
+// Related code in Linux kernel: fpsimd_flush_thread().
+// FPCR = FPCR_RM_RN (0x0 << 22).
+//
+// Currently, aarch64FPState is only a space of 0x210 length for fpstate.
+// The fp head is useless in sentry/ptrace/kvm.
+//
+func initAarch64FPState(data *State) {
+}
+
+func newAarch64FPStateSlice() []byte {
+	return alignedBytes(4096, 16)[:fpsimdContextSize]
+}
+
+// NewState returns an initialized floating point state.
+//
+// The returned state is large enough to store all floating point state
+// supported by host, even if the app won't use much of it due to a restricted
+// FeatureSet.
+func NewState() State {
+	f := State(newAarch64FPStateSlice())
+	initAarch64FPState(&f)
+	return f
+}
+
+// Fork creates and returns an identical copy of the aarch64 floating point state.
+func (s *State) Fork() State {
+	n := State(newAarch64FPStateSlice())
+	copy(n, *s)
+	return n
+}
+
+// BytePointer returns a pointer to the first byte of the state.
+func (s *State) BytePointer() *byte {
+	return &(*s)[0]
+}
diff --git a/pkg/sentry/arch/signal_amd64.go b/pkg/sentry/arch/signal_amd64.go
index e6557cab6..ee3743483 100644
--- a/pkg/sentry/arch/signal_amd64.go
+++ b/pkg/sentry/arch/signal_amd64.go
@@ -23,6 +23,7 @@ import (
 	"gvisor.dev/gvisor/pkg/abi/linux"
 	"gvisor.dev/gvisor/pkg/log"
 	"gvisor.dev/gvisor/pkg/marshal/primitive"
+	"gvisor.dev/gvisor/pkg/sentry/arch/fpu"
 	"gvisor.dev/gvisor/pkg/usermem"
 )
 
@@ -98,7 +99,7 @@ func (c *context64) NewSignalStack() NativeSignalStack {
 const _FP_XSTATE_MAGIC2_SIZE = 4
 
 func (c *context64) fpuFrameSize() (size int, useXsave bool) {
-	size = len(c.x86FPState)
+	size = len(c.fpState)
 	if size > 512 {
 		// Make room for the magic cookie at the end of the xsave frame.
 		size += _FP_XSTATE_MAGIC2_SIZE
@@ -226,10 +227,10 @@ func (c *context64) SignalSetup(st *Stack, act *SignalAct, info *SignalInfo, alt
 	c.Regs.Ss = userDS
 
 	// Save the thread's floating point state.
-	c.sigFPState = append(c.sigFPState, c.x86FPState)
+	c.sigFPState = append(c.sigFPState, c.fpState)
 
 	// Signal handler gets a clean floating point state.
-	c.x86FPState = newX86FPState()
+	c.fpState = fpu.NewState()
 
 	return nil
 }
@@ -273,7 +274,7 @@ func (c *context64) SignalRestore(st *Stack, rt bool) (linux.SignalSet, SignalSt
 	// Restore floating point state.
 	l := len(c.sigFPState)
 	if l > 0 {
-		c.x86FPState = c.sigFPState[l-1]
+		c.fpState = c.sigFPState[l-1]
 		// NOTE(cl/133042258): State save requires that any slice
 		// elements from '[len:cap]' to be zero value.
 		c.sigFPState[l-1] = nil
diff --git a/pkg/sentry/arch/signal_arm64.go b/pkg/sentry/arch/signal_arm64.go
index 4491008c2..53281dcba 100644
--- a/pkg/sentry/arch/signal_arm64.go
+++ b/pkg/sentry/arch/signal_arm64.go
@@ -20,6 +20,7 @@ import (
 	"golang.org/x/sys/unix"
 	"gvisor.dev/gvisor/pkg/abi/linux"
 	"gvisor.dev/gvisor/pkg/log"
+	"gvisor.dev/gvisor/pkg/sentry/arch/fpu"
 	"gvisor.dev/gvisor/pkg/usermem"
 )
 
@@ -139,9 +140,9 @@ func (c *context64) SignalSetup(st *Stack, act *SignalAct, info *SignalInfo, alt
 	c.Regs.Regs[30] = uint64(act.Restorer)
 
 	// Save the thread's floating point state.
-	c.sigFPState = append(c.sigFPState, c.aarch64FPState)
+	c.sigFPState = append(c.sigFPState, c.fpState)
 	// Signal handler gets a clean floating point state.
-	c.aarch64FPState = newAarch64FPState()
+	c.fpState = fpu.NewState()
 	return nil
 }
 
@@ -166,7 +167,7 @@ func (c *context64) SignalRestore(st *Stack, rt bool) (linux.SignalSet, SignalSt
 	// Restore floating point state.
 	l := len(c.sigFPState)
 	if l > 0 {
-		c.aarch64FPState = c.sigFPState[l-1]
+		c.fpState = c.sigFPState[l-1]
 		// NOTE(cl/133042258): State save requires that any slice
 		// elements from '[len:cap]' to be zero value.
 		c.sigFPState[l-1] = nil
diff --git a/pkg/sentry/kernel/ptrace_amd64.go b/pkg/sentry/kernel/ptrace_amd64.go
index 609ad3941..7aea3dcd8 100644
--- a/pkg/sentry/kernel/ptrace_amd64.go
+++ b/pkg/sentry/kernel/ptrace_amd64.go
@@ -51,14 +51,15 @@ func (t *Task) ptraceArch(target *Task, req int64, addr, data usermem.Addr) erro
 		return err
 
 	case linux.PTRACE_GETFPREGS:
-		_, err := target.Arch().PtraceGetFPRegs(&usermem.IOReadWriter{
+		s := target.Arch().FloatingPointData()
+		_, err := target.Arch().FloatingPointData().PtraceGetFPRegs(&usermem.IOReadWriter{
 			Ctx:  t,
 			IO:   t.MemoryManager(),
 			Addr: data,
 			Opts: usermem.IOOpts{
 				AddressSpaceActive: true,
 			},
-		})
+		}, len(*s))
 		return err
 
 	case linux.PTRACE_SETREGS:
@@ -73,14 +74,15 @@ func (t *Task) ptraceArch(target *Task, req int64, addr, data usermem.Addr) erro
 		return err
 
 	case linux.PTRACE_SETFPREGS:
-		_, err := target.Arch().PtraceSetFPRegs(&usermem.IOReadWriter{
+		s := target.Arch().FloatingPointData()
+		_, err := s.PtraceSetFPRegs(&usermem.IOReadWriter{
 			Ctx:  t,
 			IO:   t.MemoryManager(),
 			Addr: data,
 			Opts: usermem.IOOpts{
 				AddressSpaceActive: true,
 			},
-		})
+		}, len(*s))
 		return err
 
 	default:
diff --git a/pkg/sentry/platform/kvm/BUILD b/pkg/sentry/platform/kvm/BUILD
index 4f9e781af..03a76eb9b 100644
--- a/pkg/sentry/platform/kvm/BUILD
+++ b/pkg/sentry/platform/kvm/BUILD
@@ -50,6 +50,7 @@ go_library(
         "//pkg/safecopy",
         "//pkg/seccomp",
         "//pkg/sentry/arch",
+        "//pkg/sentry/arch/fpu",
         "//pkg/sentry/memmap",
         "//pkg/sentry/platform",
         "//pkg/sentry/platform/interrupt",
@@ -78,6 +79,7 @@ go_test(
         "//pkg/ring0",
         "//pkg/ring0/pagetables",
         "//pkg/sentry/arch",
+        "//pkg/sentry/arch/fpu",
         "//pkg/sentry/platform",
         "//pkg/sentry/platform/kvm/testutil",
         "//pkg/sentry/time",
diff --git a/pkg/sentry/platform/kvm/bluepill_amd64.go b/pkg/sentry/platform/kvm/bluepill_amd64.go
index f4b9a5321..d761bbdee 100644
--- a/pkg/sentry/platform/kvm/bluepill_amd64.go
+++ b/pkg/sentry/platform/kvm/bluepill_amd64.go
@@ -73,7 +73,7 @@ func (c *vCPU) KernelSyscall() {
 	// We only trigger a bluepill entry in the bluepill function, and can
 	// therefore be guaranteed that there is no floating point state to be
 	// loaded on resuming from halt. We only worry about saving on exit.
-	ring0.SaveFloatingPoint((*byte)(c.floatingPointState)) // escapes: no.
+	ring0.SaveFloatingPoint(c.floatingPointState.BytePointer()) // escapes: no.
 	ring0.Halt()
 	ring0.WriteFS(uintptr(regs.Fs_base)) // escapes: no, reload host segment.
 }
@@ -92,7 +92,7 @@ func (c *vCPU) KernelException(vector ring0.Vector) {
 		regs.Rip = 0
 	}
 	// See above.
-	ring0.SaveFloatingPoint((*byte)(c.floatingPointState)) // escapes: no.
+	ring0.SaveFloatingPoint(c.floatingPointState.BytePointer()) // escapes: no.
 	ring0.Halt()
 	ring0.WriteFS(uintptr(regs.Fs_base)) // escapes: no; reload host segment.
 }
@@ -124,5 +124,5 @@ func bluepillArchExit(c *vCPU, context *arch.SignalContext64) {
 	// Set the context pointer to the saved floating point state. This is
 	// where the guest data has been serialized, the kernel will restore
 	// from this new pointer value.
-	context.Fpstate = uint64(uintptrValue((*byte)(c.floatingPointState)))
+	context.Fpstate = uint64(uintptrValue(c.floatingPointState.BytePointer()))
 }
diff --git a/pkg/sentry/platform/kvm/bluepill_arm64.go b/pkg/sentry/platform/kvm/bluepill_arm64.go
index e26b7da8d..578852c3f 100644
--- a/pkg/sentry/platform/kvm/bluepill_arm64.go
+++ b/pkg/sentry/platform/kvm/bluepill_arm64.go
@@ -92,7 +92,7 @@ func bluepillArchExit(c *vCPU, context *arch.SignalContext64) {
 
 	lazyVfp := c.GetLazyVFP()
 	if lazyVfp != 0 {
-		fpsimd := fpsimdPtr((*byte)(c.floatingPointState))
+		fpsimd := fpsimdPtr(c.floatingPointState.BytePointer())
 		context.Fpsimd64.Fpsr = fpsimd.Fpsr
 		context.Fpsimd64.Fpcr = fpsimd.Fpcr
 		context.Fpsimd64.Vregs = fpsimd.Vregs
@@ -112,12 +112,12 @@ func (c *vCPU) KernelSyscall() {
 
 	fpDisableTrap := ring0.CPACREL1()
 	if fpDisableTrap != 0 {
-		fpsimd := fpsimdPtr((*byte)(c.floatingPointState))
+		fpsimd := fpsimdPtr(c.floatingPointState.BytePointer())
 		fpcr := ring0.GetFPCR()
 		fpsr := ring0.GetFPSR()
 		fpsimd.Fpcr = uint32(fpcr)
 		fpsimd.Fpsr = uint32(fpsr)
-		ring0.SaveVRegs((*byte)(c.floatingPointState))
+		ring0.SaveVRegs(c.floatingPointState.BytePointer())
 	}
 
 	ring0.Halt()
@@ -136,12 +136,12 @@ func (c *vCPU) KernelException(vector ring0.Vector) {
 
 	fpDisableTrap := ring0.CPACREL1()
 	if fpDisableTrap != 0 {
-		fpsimd := fpsimdPtr((*byte)(c.floatingPointState))
+		fpsimd := fpsimdPtr(c.floatingPointState.BytePointer())
 		fpcr := ring0.GetFPCR()
 		fpsr := ring0.GetFPSR()
 		fpsimd.Fpcr = uint32(fpcr)
 		fpsimd.Fpsr = uint32(fpsr)
-		ring0.SaveVRegs((*byte)(c.floatingPointState))
+		ring0.SaveVRegs(c.floatingPointState.BytePointer())
 	}
 
 	ring0.Halt()
diff --git a/pkg/sentry/platform/kvm/context.go b/pkg/sentry/platform/kvm/context.go
index aeae01dbd..706fa53dc 100644
--- a/pkg/sentry/platform/kvm/context.go
+++ b/pkg/sentry/platform/kvm/context.go
@@ -65,7 +65,7 @@ func (c *context) Switch(ctx pkgcontext.Context, mm platform.MemoryManager, ac a
 	// Prepare switch options.
 	switchOpts := ring0.SwitchOpts{
 		Registers:          &ac.StateData().Regs,
-		FloatingPointState: (*byte)(ac.FloatingPointData()),
+		FloatingPointState: ac.FloatingPointData(),
 		PageTables:         localAS.pageTables,
 		Flush:              localAS.Touch(cpu),
 		FullRestore:        ac.FullRestore(),
diff --git a/pkg/sentry/platform/kvm/kvm_amd64_test.go b/pkg/sentry/platform/kvm/kvm_amd64_test.go
index 76fc594a0..e44e995a0 100644
--- a/pkg/sentry/platform/kvm/kvm_amd64_test.go
+++ b/pkg/sentry/platform/kvm/kvm_amd64_test.go
@@ -33,7 +33,7 @@ func TestSegments(t *testing.T) {
 			var si arch.SignalInfo
 			if _, err := c.SwitchToUser(ring0.SwitchOpts{
 				Registers:          regs,
-				FloatingPointState: dummyFPState,
+				FloatingPointState: &dummyFPState,
 				PageTables:         pt,
 				FullRestore:        true,
 			}, &si); err == platform.ErrContextInterrupt {
diff --git a/pkg/sentry/platform/kvm/kvm_test.go b/pkg/sentry/platform/kvm/kvm_test.go
index 6243b9a04..5bce16dde 100644
--- a/pkg/sentry/platform/kvm/kvm_test.go
+++ b/pkg/sentry/platform/kvm/kvm_test.go
@@ -25,13 +25,14 @@ import (
 	"gvisor.dev/gvisor/pkg/ring0"
 	"gvisor.dev/gvisor/pkg/ring0/pagetables"
 	"gvisor.dev/gvisor/pkg/sentry/arch"
+	"gvisor.dev/gvisor/pkg/sentry/arch/fpu"
 	"gvisor.dev/gvisor/pkg/sentry/platform"
 	"gvisor.dev/gvisor/pkg/sentry/platform/kvm/testutil"
 	ktime "gvisor.dev/gvisor/pkg/sentry/time"
 	"gvisor.dev/gvisor/pkg/usermem"
 )
 
-var dummyFPState = (*byte)(arch.NewFloatingPointData())
+var dummyFPState = fpu.NewState()
 
 type testHarness interface {
 	Errorf(format string, args ...interface{})
@@ -159,7 +160,7 @@ func TestApplicationSyscall(t *testing.T) {
 		var si arch.SignalInfo
 		if _, err := c.SwitchToUser(ring0.SwitchOpts{
 			Registers:          regs,
-			FloatingPointState: dummyFPState,
+			FloatingPointState: &dummyFPState,
 			PageTables:         pt,
 			FullRestore:        true,
 		}, &si); err == platform.ErrContextInterrupt {
@@ -173,7 +174,7 @@ func TestApplicationSyscall(t *testing.T) {
 		var si arch.SignalInfo
 		if _, err := c.SwitchToUser(ring0.SwitchOpts{
 			Registers:          regs,
-			FloatingPointState: dummyFPState,
+			FloatingPointState: &dummyFPState,
 			PageTables:         pt,
 		}, &si); err == platform.ErrContextInterrupt {
 			return true // Retry.
@@ -190,7 +191,7 @@ func TestApplicationFault(t *testing.T) {
 		var si arch.SignalInfo
 		if _, err := c.SwitchToUser(ring0.SwitchOpts{
 			Registers:          regs,
-			FloatingPointState: dummyFPState,
+			FloatingPointState: &dummyFPState,
 			PageTables:         pt,
 			FullRestore:        true,
 		}, &si); err == platform.ErrContextInterrupt {
@@ -205,7 +206,7 @@ func TestApplicationFault(t *testing.T) {
 		var si arch.SignalInfo
 		if _, err := c.SwitchToUser(ring0.SwitchOpts{
 			Registers:          regs,
-			FloatingPointState: dummyFPState,
+			FloatingPointState: &dummyFPState,
 			PageTables:         pt,
 		}, &si); err == platform.ErrContextInterrupt {
 			return true // Retry.
@@ -223,7 +224,7 @@ func TestRegistersSyscall(t *testing.T) {
 			var si arch.SignalInfo
 			if _, err := c.SwitchToUser(ring0.SwitchOpts{
 				Registers:          regs,
-				FloatingPointState: dummyFPState,
+				FloatingPointState: &dummyFPState,
 				PageTables:         pt,
 			}, &si); err == platform.ErrContextInterrupt {
 				continue // Retry.
@@ -246,7 +247,7 @@ func TestRegistersFault(t *testing.T) {
 			var si arch.SignalInfo
 			if _, err := c.SwitchToUser(ring0.SwitchOpts{
 				Registers:          regs,
-				FloatingPointState: dummyFPState,
+				FloatingPointState: &dummyFPState,
 				PageTables:         pt,
 				FullRestore:        true,
 			}, &si); err == platform.ErrContextInterrupt {
@@ -272,7 +273,7 @@ func TestBounce(t *testing.T) {
 		var si arch.SignalInfo
 		if _, err := c.SwitchToUser(ring0.SwitchOpts{
 			Registers:          regs,
-			FloatingPointState: dummyFPState,
+			FloatingPointState: &dummyFPState,
 			PageTables:         pt,
 		}, &si); err != platform.ErrContextInterrupt {
 			t.Errorf("application partial restore: got %v, wanted %v", err, platform.ErrContextInterrupt)
@@ -287,7 +288,7 @@ func TestBounce(t *testing.T) {
 		var si arch.SignalInfo
 		if _, err := c.SwitchToUser(ring0.SwitchOpts{
 			Registers:          regs,
-			FloatingPointState: dummyFPState,
+			FloatingPointState: &dummyFPState,
 			PageTables:         pt,
 			FullRestore:        true,
 		}, &si); err != platform.ErrContextInterrupt {
@@ -319,7 +320,7 @@ func TestBounceStress(t *testing.T) {
 			var si arch.SignalInfo
 			if _, err := c.SwitchToUser(ring0.SwitchOpts{
 				Registers:          regs,
-				FloatingPointState: dummyFPState,
+				FloatingPointState: &dummyFPState,
 				PageTables:         pt,
 			}, &si); err != platform.ErrContextInterrupt {
 				t.Errorf("application partial restore: got %v, wanted %v", err, platform.ErrContextInterrupt)
@@ -340,7 +341,7 @@ func TestInvalidate(t *testing.T) {
 			var si arch.SignalInfo
 			if _, err := c.SwitchToUser(ring0.SwitchOpts{
 				Registers:          regs,
-				FloatingPointState: dummyFPState,
+				FloatingPointState: &dummyFPState,
 				PageTables:         pt,
 			}, &si); err == platform.ErrContextInterrupt {
 				continue // Retry.
@@ -355,7 +356,7 @@ func TestInvalidate(t *testing.T) {
 			var si arch.SignalInfo
 			if _, err := c.SwitchToUser(ring0.SwitchOpts{
 				Registers:          regs,
-				FloatingPointState: dummyFPState,
+				FloatingPointState: &dummyFPState,
 				PageTables:         pt,
 				Flush:              true,
 			}, &si); err == platform.ErrContextInterrupt {
@@ -379,7 +380,7 @@ func TestEmptyAddressSpace(t *testing.T) {
 		var si arch.SignalInfo
 		if _, err := c.SwitchToUser(ring0.SwitchOpts{
 			Registers:          regs,
-			FloatingPointState: dummyFPState,
+			FloatingPointState: &dummyFPState,
 			PageTables:         pt,
 		}, &si); err == platform.ErrContextInterrupt {
 			return true // Retry.
@@ -393,7 +394,7 @@ func TestEmptyAddressSpace(t *testing.T) {
 		var si arch.SignalInfo
 		if _, err := c.SwitchToUser(ring0.SwitchOpts{
 			Registers:          regs,
-			FloatingPointState: dummyFPState,
+			FloatingPointState: &dummyFPState,
 			PageTables:         pt,
 			FullRestore:        true,
 		}, &si); err == platform.ErrContextInterrupt {
@@ -469,7 +470,7 @@ func BenchmarkApplicationSyscall(b *testing.B) {
 		var si arch.SignalInfo
 		if _, err := c.SwitchToUser(ring0.SwitchOpts{
 			Registers:          regs,
-			FloatingPointState: dummyFPState,
+			FloatingPointState: &dummyFPState,
 			PageTables:         pt,
 		}, &si); err == platform.ErrContextInterrupt {
 			a++
@@ -506,7 +507,7 @@ func BenchmarkWorldSwitchToUserRoundtrip(b *testing.B) {
 		var si arch.SignalInfo
 		if _, err := c.SwitchToUser(ring0.SwitchOpts{
 			Registers:          regs,
-			FloatingPointState: dummyFPState,
+			FloatingPointState: &dummyFPState,
 			PageTables:         pt,
 		}, &si); err == platform.ErrContextInterrupt {
 			a++
diff --git a/pkg/sentry/platform/kvm/machine_amd64.go b/pkg/sentry/platform/kvm/machine_amd64.go
index 6e583baa3..8f2c82e73 100644
--- a/pkg/sentry/platform/kvm/machine_amd64.go
+++ b/pkg/sentry/platform/kvm/machine_amd64.go
@@ -27,6 +27,7 @@ import (
 	"gvisor.dev/gvisor/pkg/ring0"
 	"gvisor.dev/gvisor/pkg/ring0/pagetables"
 	"gvisor.dev/gvisor/pkg/sentry/arch"
+	"gvisor.dev/gvisor/pkg/sentry/arch/fpu"
 	"gvisor.dev/gvisor/pkg/sentry/platform"
 	ktime "gvisor.dev/gvisor/pkg/sentry/time"
 	"gvisor.dev/gvisor/pkg/usermem"
@@ -70,7 +71,7 @@ type vCPUArchState struct {
 
 	// floatingPointState is the floating point state buffer used in guest
 	// to host transitions. See usage in bluepill_amd64.go.
-	floatingPointState *arch.FloatingPointData
+	floatingPointState fpu.State
 }
 
 const (
@@ -151,7 +152,7 @@ func (c *vCPU) initArchState() error {
 	// This will be saved prior to leaving the guest, and we restore from
 	// this always. We cannot use the pointer in the context alone because
 	// we don't know how large the area there is in reality.
-	c.floatingPointState = arch.NewFloatingPointData()
+	c.floatingPointState = fpu.NewState()
 
 	// Set the time offset to the host native time.
 	return c.setSystemTime()
diff --git a/pkg/sentry/platform/kvm/machine_arm64.go b/pkg/sentry/platform/kvm/machine_arm64.go
index 7d7857067..2edc9d1b2 100644
--- a/pkg/sentry/platform/kvm/machine_arm64.go
+++ b/pkg/sentry/platform/kvm/machine_arm64.go
@@ -20,6 +20,7 @@ import (
 	"gvisor.dev/gvisor/pkg/ring0"
 	"gvisor.dev/gvisor/pkg/ring0/pagetables"
 	"gvisor.dev/gvisor/pkg/sentry/arch"
+	"gvisor.dev/gvisor/pkg/sentry/arch/fpu"
 	"gvisor.dev/gvisor/pkg/sentry/platform"
 	"gvisor.dev/gvisor/pkg/usermem"
 )
@@ -32,7 +33,7 @@ type vCPUArchState struct {
 
 	// floatingPointState is the floating point state buffer used in guest
 	// to host transitions. See usage in bluepill_arm64.go.
-	floatingPointState *arch.FloatingPointData
+	floatingPointState fpu.State
 }
 
 const (
diff --git a/pkg/sentry/platform/kvm/machine_arm64_unsafe.go b/pkg/sentry/platform/kvm/machine_arm64_unsafe.go
index 059aa43d0..e7d5f3193 100644
--- a/pkg/sentry/platform/kvm/machine_arm64_unsafe.go
+++ b/pkg/sentry/platform/kvm/machine_arm64_unsafe.go
@@ -26,6 +26,7 @@ import (
 	"gvisor.dev/gvisor/pkg/ring0"
 	"gvisor.dev/gvisor/pkg/ring0/pagetables"
 	"gvisor.dev/gvisor/pkg/sentry/arch"
+	"gvisor.dev/gvisor/pkg/sentry/arch/fpu"
 	"gvisor.dev/gvisor/pkg/sentry/platform"
 	"gvisor.dev/gvisor/pkg/usermem"
 )
@@ -150,7 +151,7 @@ func (c *vCPU) initArchState() error {
 		c.PCIDs = pagetables.NewPCIDs(fixedKernelPCID+1, poolPCIDs)
 	}
 
-	c.floatingPointState = arch.NewFloatingPointData()
+	c.floatingPointState = fpu.NewState()
 
 	return c.setSystemTime()
 }
diff --git a/pkg/sentry/platform/ptrace/BUILD b/pkg/sentry/platform/ptrace/BUILD
index fc43cc3c0..47efde6a2 100644
--- a/pkg/sentry/platform/ptrace/BUILD
+++ b/pkg/sentry/platform/ptrace/BUILD
@@ -30,6 +30,7 @@ go_library(
         "//pkg/safecopy",
         "//pkg/seccomp",
         "//pkg/sentry/arch",
+        "//pkg/sentry/arch/fpu",
         "//pkg/sentry/memmap",
         "//pkg/sentry/platform",
         "//pkg/sentry/platform/interrupt",
diff --git a/pkg/sentry/platform/ptrace/ptrace_unsafe.go b/pkg/sentry/platform/ptrace/ptrace_unsafe.go
index 2c21f946e..01e73b019 100644
--- a/pkg/sentry/platform/ptrace/ptrace_unsafe.go
+++ b/pkg/sentry/platform/ptrace/ptrace_unsafe.go
@@ -20,6 +20,7 @@ import (
 	"golang.org/x/sys/unix"
 	"gvisor.dev/gvisor/pkg/abi/linux"
 	"gvisor.dev/gvisor/pkg/sentry/arch"
+	"gvisor.dev/gvisor/pkg/sentry/arch/fpu"
 	"gvisor.dev/gvisor/pkg/usermem"
 )
 
@@ -62,9 +63,9 @@ func (t *thread) setRegs(regs *arch.Registers) error {
 }
 
 // getFPRegs gets the floating-point data via the GETREGSET ptrace unix.
-func (t *thread) getFPRegs(fpState *arch.FloatingPointData, fpLen uint64, useXsave bool) error {
+func (t *thread) getFPRegs(fpState *fpu.State, fpLen uint64, useXsave bool) error {
 	iovec := unix.Iovec{
-		Base: (*byte)(fpState),
+		Base: fpState.BytePointer(),
 		Len:  fpLen,
 	}
 	_, _, errno := unix.RawSyscall6(
@@ -81,9 +82,9 @@ func (t *thread) getFPRegs(fpState *arch.FloatingPointData, fpLen uint64, useXsa
 }
 
 // setFPRegs sets the floating-point data via the SETREGSET ptrace unix.
-func (t *thread) setFPRegs(fpState *arch.FloatingPointData, fpLen uint64, useXsave bool) error {
+func (t *thread) setFPRegs(fpState *fpu.State, fpLen uint64, useXsave bool) error {
 	iovec := unix.Iovec{
-		Base: (*byte)(fpState),
+		Base: fpState.BytePointer(),
 		Len:  fpLen,
 	}
 	_, _, errno := unix.RawSyscall6(