Index: sys/arm64/arm64/elf32_machdep.c
===================================================================
--- sys/arm64/arm64/elf32_machdep.c
+++ sys/arm64/arm64/elf32_machdep.c
@@ -51,6 +51,7 @@
 #include <sys/vnode.h>
 
 #include <machine/elf.h>
+#include <machine/machdep.h>
 
 #include <compat/freebsd32/freebsd32_util.h>
 
@@ -251,6 +252,8 @@
 	tf->tf_x[14] = imgp->entry_addr;
 	tf->tf_elr = imgp->entry_addr;
 	tf->tf_spsr = PSR_M_32;
+
+	exec_reset_vfp_state(td, td->td_pcb);
 }
 
 void
Index: sys/arm64/arm64/machdep.c
===================================================================
--- sys/arm64/arm64/machdep.c
+++ sys/arm64/arm64/machdep.c
@@ -548,10 +548,28 @@
 	return (0);
 }
 
+/*
+ * Reset the FP state to avoid leaking state from the parent process across
+ * execve() (and to ensure that we get a consitent floating point environment
+ * in every new process).
+ */
+void
+exec_reset_vfp_state(struct thread *td, struct pcb *pcb)
+{
+	bzero(&pcb->pcb_fpustate.vfp_regs, sizeof(pcb->pcb_fpustate.vfp_regs));
+	KASSERT(pcb->pcb_fpusaved == &pcb->pcb_fpustate,
+	    ("pcb_fpusaved should point to pcb_fpustate."));
+	pcb->pcb_fpustate.vfp_fpcr = initial_fpcr;
+	pcb->pcb_fpustate.vfp_fpsr = 0;
+	pcb->pcb_vfpcpu = UINT_MAX;
+	pcb->pcb_fpflags = 0;
+}
+
 void
 exec_setregs(struct thread *td, struct image_params *imgp, uintptr_t stack)
 {
 	struct trapframe *tf = td->td_frame;
+	struct pcb *pcb = td->td_pcb;
 
 	memset(tf, 0, sizeof(struct trapframe));
 
@@ -559,6 +577,11 @@
 	tf->tf_sp = STACKALIGN(stack);
 	tf->tf_lr = imgp->entry_addr;
 	tf->tf_elr = imgp->entry_addr;
+
+	exec_reset_vfp_state(td, pcb);
+
+	/* FIXME: Shouldn't we also reset pcb_dbg_regs? */
+	/* XXX: should exec clear single-step in pcb_flags? */
 }
 
 /* Sanity check these are the same size, they will be memcpy'd to and fro */
Index: sys/arm64/arm64/vm_machdep.c
===================================================================
--- sys/arm64/arm64/vm_machdep.c
+++ sys/arm64/arm64/vm_machdep.c
@@ -108,7 +108,6 @@
 	td2->td_pcb->pcb_sp = (uintptr_t)td2->td_frame;
 	td2->td_pcb->pcb_fpusaved = &td2->td_pcb->pcb_fpustate;
 	td2->td_pcb->pcb_vfpcpu = UINT_MAX;
-	td2->td_pcb->pcb_fpusaved->vfp_fpcr = initial_fpcr;
 
 	/* Setup to release spin count in fork_exit(). */
 	td2->td_md.md_spinlock_count = 1;
Index: sys/arm64/include/machdep.h
===================================================================
--- sys/arm64/include/machdep.h
+++ sys/arm64/include/machdep.h
@@ -59,6 +59,7 @@
 #endif
 int memory_mapping_mode(vm_paddr_t pa);
 extern void (*pagezero)(void *);
+void exec_reset_vfp_state(struct thread *, struct pcb *);
 
 #ifdef SOCDEV_PA
 /*
Index: sys/arm64/linux/linux_sysvec.c
===================================================================
--- sys/arm64/linux/linux_sysvec.c
+++ sys/arm64/linux/linux_sysvec.c
@@ -57,6 +57,8 @@
 #include <compat/linux/linux_util.h>
 #include <compat/linux/linux_vdso.h>
 
+#include <machine/machdep.h>
+
 MODULE_VERSION(linux64elf, 1);
 
 const char *linux_kplatform;
@@ -360,6 +362,8 @@
         regs->tf_lr = 0xffffffffffffffff;
 #endif
         regs->tf_elr = imgp->entry_addr;
+
+	exec_reset_vfp_state(td, td->td_pcb);
 }
 
 int