lguest: use KVM hypercalls

Impact: cleanup

This patch allow us to use KVM hypercalls

Signed-off-by: Matias Zabaljauregui <zabaljauregui at gmail.com>
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>

3 files changed, 67 insertions, 6 deletions

diff --git a/drivers/lguest/interrupts_and_traps.c b/drivers/lguest/interrupts_and_traps.c
index 415fab0..504091d 100644
--- a/drivers/lguest/interrupts_and_traps.c
+++ b/drivers/lguest/interrupts_and_traps.c
@@ -288,9 +288,10 @@ static int direct_trap(unsigned int num)
 
 	/* The Host needs to see page faults (for shadow paging and to save the
 	 * fault address), general protection faults (in/out emulation) and
-	 * device not available (TS handling), and of course, the hypercall
-	 * trap. */
-	return num != 14 && num != 13 && num != 7 && num != LGUEST_TRAP_ENTRY;
+	 * device not available (TS handling), invalid opcode fault (kvm hcall),
+	 * and of course, the hypercall trap. */
+	return num != 14 && num != 13 && num != 7 &&
+			num != 6 && num != LGUEST_TRAP_ENTRY;
 }
 /*:*/
 
diff --git a/drivers/lguest/lguest_device.c b/drivers/lguest/lguest_device.c
index 8132533..df44d96 100644
--- a/drivers/lguest/lguest_device.c
+++ b/drivers/lguest/lguest_device.c
@@ -161,7 +161,7 @@ static void set_status(struct virtio_device *vdev, u8 status)
 
 	/* We set the status. */
 	to_lgdev(vdev)->desc->status = status;
-	hcall(LHCALL_NOTIFY, (max_pfn<<PAGE_SHIFT) + offset, 0, 0);
+	kvm_hypercall1(LHCALL_NOTIFY, (max_pfn << PAGE_SHIFT) + offset);
 }
 
 static void lg_set_status(struct virtio_device *vdev, u8 status)
@@ -209,7 +209,7 @@ static void lg_notify(struct virtqueue *vq)
 	 * virtqueue structure. */
 	struct lguest_vq_info *lvq = vq->priv;
 
-	hcall(LHCALL_NOTIFY, lvq->config.pfn << PAGE_SHIFT, 0, 0);
+	kvm_hypercall1(LHCALL_NOTIFY, lvq->config.pfn << PAGE_SHIFT);
 }
 
 /* An extern declaration inside a C file is bad form.  Don't do it. */
diff --git a/drivers/lguest/x86/core.c b/drivers/lguest/x86/core.c
index bf79423..a6b7176 100644
--- a/drivers/lguest/x86/core.c
+++ b/drivers/lguest/x86/core.c
@@ -290,6 +290,57 @@ static int emulate_insn(struct lg_cpu *cpu)
 	return 1;
 }
 
+/* Our hypercalls mechanism used to be based on direct software interrupts.
+ * After Anthony's "Refactor hypercall infrastructure" kvm patch, we decided to
+ * change over to using kvm hypercalls.
+ *
+ * KVM_HYPERCALL is actually a "vmcall" instruction, which generates an invalid
+ * opcode fault (fault 6) on non-VT cpus, so the easiest solution seemed to be
+ * an *emulation approach*: if the fault was really produced by an hypercall
+ * (is_hypercall() does exactly this check), we can just call the corresponding
+ * hypercall host implementation function.
+ *
+ * But these invalid opcode faults are notably slower than software interrupts.
+ * So we implemented the *patching (or rewriting) approach*: every time we hit
+ * the KVM_HYPERCALL opcode in Guest code, we patch it to the old "int 0x1f"
+ * opcode, so next time the Guest calls this hypercall it will use the
+ * faster trap mechanism.
+ *
+ * Matias even benchmarked it to convince you: this shows the average cycle
+ * cost of a hypercall.  For each alternative solution mentioned above we've
+ * made 5 runs of the benchmark:
+ *
+ * 1) direct software interrupt: 2915, 2789, 2764, 2721, 2898
+ * 2) emulation technique: 3410, 3681, 3466, 3392, 3780
+ * 3) patching (rewrite) technique: 2977, 2975, 2891, 2637, 2884
+ *
+ * One two-line function is worth a 20% hypercall speed boost!
+ */
+static void rewrite_hypercall(struct lg_cpu *cpu)
+{
+	/* This are the opcodes we use to patch the Guest.  The opcode for "int
+	 * $0x1f" is "0xcd 0x1f" but vmcall instruction is 3 bytes long, so we
+	 * complete the sequence with a NOP (0x90). */
+	u8 insn[3] = {0xcd, 0x1f, 0x90};
+
+	__lgwrite(cpu, guest_pa(cpu, cpu->regs->eip), insn, sizeof(insn));
+}
+
+static bool is_hypercall(struct lg_cpu *cpu)
+{
+	u8 insn[3];
+
+	/* This must be the Guest kernel trying to do something.
+	 * The bottom two bits of the CS segment register are the privilege
+	 * level. */
+	if ((cpu->regs->cs & 3) != GUEST_PL)
+		return false;
+
+	/* Is it a vmcall? */
+	__lgread(cpu, insn, guest_pa(cpu, cpu->regs->eip), sizeof(insn));
+	return insn[0] == 0x0f && insn[1] == 0x01 && insn[2] == 0xc1;
+}
+
 /*H:050 Once we've re-enabled interrupts, we look at why the Guest exited. */
 void lguest_arch_handle_trap(struct lg_cpu *cpu)
 {
@@ -337,7 +388,7 @@ void lguest_arch_handle_trap(struct lg_cpu *cpu)
 		break;
 	case 32 ... 255:
 		/* These values mean a real interrupt occurred, in which case
-		 * the Host handler has already been run.  We just do a
+		 * the Host handler has already been run. We just do a
 		 * friendly check if another process should now be run, then
 		 * return to run the Guest again */
 		cond_resched();
@@ -347,6 +398,15 @@ void lguest_arch_handle_trap(struct lg_cpu *cpu)
 		 * up the pointer now to indicate a hypercall is pending. */
 		cpu->hcall = (struct hcall_args *)cpu->regs;
 		return;
+	case 6:
+		/* kvm hypercalls trigger an invalid opcode fault (6).
+		 * We need to check if ring == GUEST_PL and
+		 * faulting instruction == vmcall. */
+		if (is_hypercall(cpu)) {
+			rewrite_hypercall(cpu);
+			return;
+		}
+		break;
 	}
 
 	/* We didn't handle the trap, so it needs to go to the Guest. */