MFC 264353,264509,264768,264770,264825,264846,264988,265114,265165,265365,

265941,265951,266390,266550,266910:
Various bhyve fixes:
- Don't save host's return address in 'struct vmxctx'.
- Permit non-32-bit accesses to local APIC registers.
- Factor out common ioport handler code.
- Use calloc() in favor of malloc + memset.
- Change the vlapic timer frequency to be in the ballpark of contemporary
  hardware.
- Allow the guest to read the TSC via MSR 0x10.
- A VMCS is always inactive when it exits the vmx_run() loop.  Remove
  redundant code and the misleading comment that suggest otherwise.
- Ignore writes to microcode update MSR.  This MSR is accessed by RHEL7
  guest.
  Add KTR tracepoints to annotate wrmsr and rdmsr VM exits.
- Provide an alias for the userboot console and name it 'comconsole'.
- Use EV_ADD to create an mevent and EV_ENABLE to enable it.
- abort(3) the process in response to a VMEXIT_ABORT.
- Don't include the guest memory segments in the bhyve(8) process core dump.
- Make the vmx asm code dtrace-fbt-friendly.
- Allow vmx_getdesc() and vmx_setdesc() to be called for a vcpu that is in
  the VCPU_RUNNING state.
- Enable VMX in the IA32_FEATURE_CONTROL MSR if it not enabled and the MSR
  isn't locked.

13 files changed, 53 insertions, 64 deletions

diff --git a/usr.sbin/bhyve/bhyve.8 b/usr.sbin/bhyve/bhyve.8
index 2d2d396..aad1aef 100644
--- a/usr.sbin/bhyve/bhyve.8
+++ b/usr.sbin/bhyve/bhyve.8
@@ -32,7 +32,7 @@
 .Nd "run a guest operating system inside a virtual machine"
 .Sh SYNOPSIS
 .Nm
-.Op Fl aehwxAHPW
+.Op Fl aehwxACHPW
 .Op Fl c Ar numcpus
 .Op Fl g Ar gdbport
 .Op Fl p Ar vcpu:hostcpu
@@ -69,6 +69,8 @@ guests.
 .It Fl c Ar numcpus
 Number of guest virtual CPUs.
 The default is 1 and the maximum is 16.
+.It Fl C
+Include guest memory in core file.
 .It Fl H
 Yield the virtual CPU thread when a HLT instruction is detected.
 If this option is not specified, virtual CPUs will use 100% of a host CPU.
diff --git a/usr.sbin/bhyve/bhyverun.c b/usr.sbin/bhyve/bhyverun.c
index e662ca3..b2354c9 100644
--- a/usr.sbin/bhyve/bhyverun.c
+++ b/usr.sbin/bhyve/bhyverun.c
@@ -68,7 +68,6 @@ __FBSDID("$FreeBSD$");
 
 #define GUEST_NIO_PORT		0x488	/* guest upcalls via i/o port */
 
-#define	VMEXIT_SWITCH		0	/* force vcpu switch in mux mode */
 #define	VMEXIT_CONTINUE		1	/* continue from next instruction */
 #define	VMEXIT_RESTART		2	/* restart current instruction */
 #define	VMEXIT_ABORT		3	/* abort the vm run loop */
@@ -135,6 +134,7 @@ usage(int code)
 		"       -A: create an ACPI table\n"
 		"       -g: gdb port\n"
 		"       -c: # cpus (default 1)\n"
+		"       -C: include guest memory in core file\n"
 		"       -p: pin 'vcpu' to 'hostcpu'\n"
 		"       -H: vmexit from the guest on hlt\n"
 		"       -P: vmexit from the guest on pause\n"
@@ -272,12 +272,6 @@ fbsdrun_deletecpu(struct vmctx *ctx, int vcpu)
 }
 
 static int
-vmexit_catch_inout(void)
-{
-	return (VMEXIT_ABORT);
-}
-
-static int
 vmexit_handle_notify(struct vmctx *ctx, struct vm_exit *vme, int *pvcpu,
 		     uint32_t eax)
 {
@@ -330,7 +324,7 @@ vmexit_inout(struct vmctx *ctx, struct vm_exit *vme, int *pvcpu)
 		fprintf(stderr, "Unhandled %s%c 0x%04x\n",
 			in ? "in" : "out",
 			bytes == 1 ? 'b' : (bytes == 2 ? 'w' : 'l'), port);
-		return (vmexit_catch_inout());
+		return (VMEXIT_ABORT);
 	}
 }
 
@@ -575,6 +569,8 @@ vm_loop(struct vmctx *ctx, int vcpu, uint64_t rip)
 			assert(error == 0 || errno == EALREADY);
                         rip = vmexit[vcpu].rip + vmexit[vcpu].inst_length;
 			break;
+		case VMEXIT_ABORT:
+			abort();
 		default:
 			exit(1);
 		}
@@ -647,19 +643,20 @@ int
 main(int argc, char *argv[])
 {
 	int c, error, gdb_port, err, bvmcons;
-	int max_vcpus, mptgen;
+	int dump_guest_memory, max_vcpus, mptgen;
 	struct vmctx *ctx;
 	uint64_t rip;
 	size_t memsize;
 
 	bvmcons = 0;
+	dump_guest_memory = 0;
 	progname = basename(argv[0]);
 	gdb_port = 0;
 	guest_ncpus = 1;
 	memsize = 256 * MB;
 	mptgen = 1;
 
-	while ((c = getopt(argc, argv, "abehwxAHIPWYp:g:c:s:m:l:U:")) != -1) {
+	while ((c = getopt(argc, argv, "abehwxACHIPWYp:g:c:s:m:l:U:")) != -1) {
 		switch (c) {
 		case 'a':
 			x2apic_mode = 0;
@@ -679,6 +676,9 @@ main(int argc, char *argv[])
                 case 'c':
 			guest_ncpus = atoi(optarg);
 			break;
+		case 'C':
+			dump_guest_memory = 1;
+			break;
 		case 'g':
 			gdb_port = atoi(optarg);
 			break;
@@ -760,6 +760,8 @@ main(int argc, char *argv[])
 
 	fbsdrun_set_capabilities(ctx, BSP);
 
+	if (dump_guest_memory)
+		vm_set_memflags(ctx, VM_MEM_F_INCORE);
 	err = vm_setup_memory(ctx, memsize, VM_MMAP_ALL);
 	if (err) {
 		fprintf(stderr, "Unable to setup memory (%d)\n", err);
diff --git a/usr.sbin/bhyve/block_if.c b/usr.sbin/bhyve/block_if.c
index ffa343c..b29bc78 100644
--- a/usr.sbin/bhyve/block_if.c
+++ b/usr.sbin/bhyve/block_if.c
@@ -270,13 +270,12 @@ blockif_open(const char *optstr, const char *ident)
 		assert(sectsz != 0);
 	}
 
-	bc = malloc(sizeof(struct blockif_ctxt));
+	bc = calloc(1, sizeof(struct blockif_ctxt));
 	if (bc == NULL) {
 		close(fd);
 		return (NULL);
 	}
 
-	memset(bc, 0, sizeof(*bc));
 	bc->bc_magic = BLOCKIF_SIG;
 	bc->bc_fd = fd;
 	bc->bc_size = size;
diff --git a/usr.sbin/bhyve/inout.c b/usr.sbin/bhyve/inout.c
index 9fec70a..5fbe99b 100644
--- a/usr.sbin/bhyve/inout.c
+++ b/usr.sbin/bhyve/inout.c
@@ -107,18 +107,19 @@ emulate_inout(struct vmctx *ctx, int vcpu, int in, int port, int bytes,
 	if (strict && handler == default_inout)
 		return (-1);
 
+	switch (bytes) {
+	case 1:
+		mask = 0xff;
+		break;
+	case 2:
+		mask = 0xffff;
+		break;
+	default:
+		mask = 0xffffffff;
+		break;
+	}
+
 	if (!in) {
-		switch (bytes) {
-		case 1:
-			mask = 0xff;
-			break;
-		case 2:
-			mask = 0xffff;
-			break;
-		default:
-			mask = 0xffffffff;
-			break;
-		}
 		val = *eax & mask;
 	}
 
@@ -131,17 +132,6 @@ emulate_inout(struct vmctx *ctx, int vcpu, int in, int port, int bytes,
 		error = -1;
 
 	if (!error && in) {
-		switch (bytes) {
-		case 1:
-			mask = 0xff;
-			break;
-		case 2:
-			mask = 0xffff;
-			break;
-		default:
-			mask = 0xffffffff;
-			break;
-		}
 		*eax &= ~mask;
 		*eax |= val & mask;
 	}
diff --git a/usr.sbin/bhyve/mevent.c b/usr.sbin/bhyve/mevent.c
index 82faab2..07d3baf 100644
--- a/usr.sbin/bhyve/mevent.c
+++ b/usr.sbin/bhyve/mevent.c
@@ -52,9 +52,10 @@ __FBSDID("$FreeBSD$");
 
 #define	MEVENT_MAX	64
 
-#define MEV_ENABLE	1
-#define MEV_DISABLE	2
-#define MEV_DEL_PENDING	3
+#define	MEV_ADD		1
+#define	MEV_ENABLE	2
+#define	MEV_DISABLE	3
+#define	MEV_DEL_PENDING	4
 
 extern char *vmname;
 
@@ -147,10 +148,11 @@ mevent_kq_flags(struct mevent *mevp)
 	int ret;
 
 	switch (mevp->me_state) {
+	case MEV_ADD:
+		ret = EV_ADD;		/* implicitly enabled */
+		break;
 	case MEV_ENABLE:
-		ret = EV_ADD;
-		if (mevp->me_type == EVF_TIMER)
-			ret |= EV_ENABLE;
+		ret = EV_ENABLE;
 		break;
 	case MEV_DISABLE:
 		ret = EV_DISABLE;
@@ -158,6 +160,9 @@ mevent_kq_flags(struct mevent *mevp)
 	case MEV_DEL_PENDING:
 		ret = EV_DELETE;
 		break;
+	default:
+		assert(0);
+		break;
 	}
 
 	return (ret);
@@ -268,12 +273,11 @@ mevent_add(int tfd, enum ev_type type,
 	/*
 	 * Allocate an entry, populate it, and add it to the change list.
 	 */
-	mevp = malloc(sizeof(struct mevent));
+	mevp = calloc(1, sizeof(struct mevent));
 	if (mevp == NULL) {
 		goto exit;
 	}
 
-	memset(mevp, 0, sizeof(struct mevent));
 	if (type == EVF_TIMER) {
 		mevp->me_msecs = tfd;
 		mevp->me_timid = mevent_timid++;
@@ -285,7 +289,7 @@ mevent_add(int tfd, enum ev_type type,
 
 	LIST_INSERT_HEAD(&change_head, mevp, me_list);
 	mevp->me_cq = 1;
-	mevp->me_state = MEV_ENABLE;
+	mevp->me_state = MEV_ADD;
 	mevent_notify();
 
 exit:
diff --git a/usr.sbin/bhyve/pci_ahci.c b/usr.sbin/bhyve/pci_ahci.c
index dd9bfb9..9f61107 100644
--- a/usr.sbin/bhyve/pci_ahci.c
+++ b/usr.sbin/bhyve/pci_ahci.c
@@ -1786,8 +1786,7 @@ pci_ahci_init(struct vmctx *ctx, struct pci_devinst *pi, char *opts, int atapi)
 	dbg = fopen("/tmp/log", "w+");
 #endif
 
-       	sc = malloc(sizeof(struct pci_ahci_softc));
-	memset(sc, 0, sizeof(struct pci_ahci_softc));
+	sc = calloc(1, sizeof(struct pci_ahci_softc));
 	pi->pi_arg = sc;
 	sc->asc_pi = pi;
 	sc->ports = MAX_PORTS;
diff --git a/usr.sbin/bhyve/pci_emul.c b/usr.sbin/bhyve/pci_emul.c
index 8fd19ac..5455f40 100644
--- a/usr.sbin/bhyve/pci_emul.c
+++ b/usr.sbin/bhyve/pci_emul.c
@@ -705,8 +705,7 @@ pci_emul_init(struct vmctx *ctx, struct pci_devemu *pde, int bus, int slot,
 	struct pci_devinst *pdi;
 	int err;
 
-	pdi = malloc(sizeof(struct pci_devinst));
-	bzero(pdi, sizeof(*pdi));
+	pdi = calloc(1, sizeof(struct pci_devinst));
 
 	pdi->pi_vmctx = ctx;
 	pdi->pi_bus = bus;
@@ -798,8 +797,7 @@ pci_msix_table_init(struct pci_devinst *pi, int table_entries)
 	assert(table_entries <= MAX_MSIX_TABLE_ENTRIES);
 
 	table_size = table_entries * MSIX_TABLE_ENTRY_SIZE;
-	pi->pi_msix.table = malloc(table_size);
-	bzero(pi->pi_msix.table, table_size);
+	pi->pi_msix.table = calloc(1, table_size);
 
 	/* set mask bit of vector control register */
 	for (i = 0; i < table_entries; i++)
@@ -1809,8 +1807,7 @@ pci_emul_dinit(struct vmctx *ctx, struct pci_devinst *pi, char *opts)
 	int error;
 	struct pci_emul_dsoftc *sc;
 
-	sc = malloc(sizeof(struct pci_emul_dsoftc));
-	memset(sc, 0, sizeof(struct pci_emul_dsoftc));
+	sc = calloc(1, sizeof(struct pci_emul_dsoftc));
 
 	pi->pi_arg = sc;
 
diff --git a/usr.sbin/bhyve/pci_passthru.c b/usr.sbin/bhyve/pci_passthru.c
index 562d532..04d68c4 100644
--- a/usr.sbin/bhyve/pci_passthru.c
+++ b/usr.sbin/bhyve/pci_passthru.c
@@ -232,8 +232,7 @@ cfginitmsi(struct passthru_softc *sc)
 
 		/* Allocate the emulated MSI-X table array */
 		table_size = pi->pi_msix.table_count * MSIX_TABLE_ENTRY_SIZE;
-		pi->pi_msix.table = malloc(table_size);
-		bzero(pi->pi_msix.table, table_size);
+		pi->pi_msix.table = calloc(1, table_size);
 
 		/* Mask all table entries */
 		for (i = 0; i < pi->pi_msix.table_count; i++) {
@@ -574,8 +573,7 @@ passthru_init(struct vmctx *ctx, struct pci_devinst *pi, char *opts)
 	if (vm_assign_pptdev(ctx, bus, slot, func) != 0)
 		goto done;
 
-	sc = malloc(sizeof(struct passthru_softc));
-	memset(sc, 0, sizeof(struct passthru_softc));
+	sc = calloc(1, sizeof(struct passthru_softc));
 
 	pi->pi_arg = sc;
 	sc->psc_pi = pi;
diff --git a/usr.sbin/bhyve/pci_virtio_block.c b/usr.sbin/bhyve/pci_virtio_block.c
index 2f37804..3474fd7 100644
--- a/usr.sbin/bhyve/pci_virtio_block.c
+++ b/usr.sbin/bhyve/pci_virtio_block.c
@@ -299,8 +299,7 @@ pci_vtblk_init(struct vmctx *ctx, struct pci_devinst *pi, char *opts)
 		assert(sectsz != 0);
 	}
 
-	sc = malloc(sizeof(struct pci_vtblk_softc));
-	memset(sc, 0, sizeof(struct pci_vtblk_softc));
+	sc = calloc(1, sizeof(struct pci_vtblk_softc));
 
 	/* record fd of storage device/file */
 	sc->vbsc_fd = fd;
diff --git a/usr.sbin/bhyve/pci_virtio_net.c b/usr.sbin/bhyve/pci_virtio_net.c
index e7b54bc..c3b8690 100644
--- a/usr.sbin/bhyve/pci_virtio_net.c
+++ b/usr.sbin/bhyve/pci_virtio_net.c
@@ -513,8 +513,7 @@ pci_vtnet_init(struct vmctx *ctx, struct pci_devinst *pi, char *opts)
 	char *vtopts;
 	int mac_provided;
 
-	sc = malloc(sizeof(struct pci_vtnet_softc));
-	memset(sc, 0, sizeof(struct pci_vtnet_softc));
+	sc = calloc(1, sizeof(struct pci_vtnet_softc));
 
 	pthread_mutex_init(&sc->vsc_mtx, NULL);
 
diff --git a/usr.sbin/bhyve/pci_virtio_rnd.c b/usr.sbin/bhyve/pci_virtio_rnd.c
index 38459d2..4d53183 100644
--- a/usr.sbin/bhyve/pci_virtio_rnd.c
+++ b/usr.sbin/bhyve/pci_virtio_rnd.c
@@ -155,8 +155,7 @@ pci_vtrnd_init(struct vmctx *ctx, struct pci_devinst *pi, char *opts)
 		return (1);
 	}
 
-	sc = malloc(sizeof(struct pci_vtrnd_softc));
-	memset(sc, 0, sizeof(struct pci_vtrnd_softc));
+	sc = calloc(1, sizeof(struct pci_vtrnd_softc));
 
 	vi_softc_linkup(&sc->vrsc_vs, &vtrnd_vi_consts, sc, pi, &sc->vrsc_vq);
 	sc->vrsc_vs.vs_mtx = &sc->vrsc_mtx;
diff --git a/usr.sbin/bhyve/uart_emul.c b/usr.sbin/bhyve/uart_emul.c
index ae4cab3..4242e5c 100644
--- a/usr.sbin/bhyve/uart_emul.c
+++ b/usr.sbin/bhyve/uart_emul.c
@@ -594,8 +594,7 @@ uart_init(uart_intr_func_t intr_assert, uart_intr_func_t intr_deassert,
 {
 	struct uart_softc *sc;
 
-	sc = malloc(sizeof(struct uart_softc));
-	bzero(sc, sizeof(struct uart_softc));
+	sc = calloc(1, sizeof(struct uart_softc));
 
 	sc->arg = arg;
 	sc->intr_assert = intr_assert;
diff --git a/usr.sbin/bhyve/xmsr.c b/usr.sbin/bhyve/xmsr.c
index ba94125..63522bf 100644
--- a/usr.sbin/bhyve/xmsr.c
+++ b/usr.sbin/bhyve/xmsr.c
@@ -47,6 +47,8 @@ emulate_wrmsr(struct vmctx *ctx, int vcpu, uint32_t code, uint64_t val)
 	case 0xd04:			/* Sandy Bridge uncore PMC MSRs */
 	case 0xc24:
 		return (0);
+	case 0x79:
+		return (0);		/* IA32_BIOS_UPDT_TRIG MSR */
 	default:
 		break;
 	}