Improved single stepping and added initial cpuid events

6 files changed, 355 insertions, 12 deletions

diff --git a/test/.gitignore b/test/.gitignore
index c2e8d30..f54caf0 100755
--- a/test/.gitignore
+++ b/test/.gitignore
@@ -5,3 +5,7 @@ sev
 sev-es
 sev-snp
 sevstep
+aes-detect_guest
+aes-detect_host
+access-detect_guest
+access-detect_host
diff --git a/test/access-detect_guest.c b/test/access-detect_guest.c
new file mode 100644
index 0000000..442c88c
--- /dev/null
+++ b/test/access-detect_guest.c
@@ -0,0 +1,26 @@
+#include "cachepc/uapi.h"
+
+#include <err.h>
+#include <unistd.h>
+#include <stdint.h>
+#include <string.h>
+#include <stdlib.h>
+
+int
+main(int argc, const char **argv)
+{
+	void *buf;
+
+	buf = NULL;
+	if (posix_memalign(&buf, L1_LINESIZE * L1_SETS, L1_LINESIZE * L1_SETS))
+		err(1, "memalign");
+	memset(buf, 0, L1_LINESIZE * L1_SETS);
+
+	while (1) {
+		CPC_CPUID_SIGNAL(CPC_CPUID_START_TRACK, 0);
+		
+		*(uint8_t *)(buf + L1_LINESIZE * 5) += 1;
+
+		CPC_CPUID_SIGNAL(CPC_CPUID_START_TRACK, 0);
+	}
+}
diff --git a/test/access-detect_host.c b/test/access-detect_host.c
new file mode 100644
index 0000000..9ede56f
--- /dev/null
+++ b/test/access-detect_host.c
@@ -0,0 +1,288 @@
+#define _GNU_SOURCE
+
+#include "cachepc/uapi.h"
+
+#include <linux/psp-sev.h>
+#include <linux/kvm.h>
+#include <sys/syscall.h>
+#include <sys/ioctl.h>
+#include <sys/user.h>
+#include <sys/wait.h>
+#include <sys/ioctl.h>
+#include <sys/mman.h>
+#include <sys/stat.h>
+#include <sys/types.h>
+#include <unistd.h>
+#include <signal.h>
+#include <dirent.h>
+#include <assert.h>
+#include <errno.h>
+#include <err.h>
+#include <fcntl.h>
+#include <sched.h>
+#include <string.h>
+#include <stdbool.h>
+#include <stdlib.h>
+#include <stdint.h>
+#include <stdio.h>
+#include <stdarg.h>
+
+#define ARRLEN(x) (sizeof(x) / sizeof((x)[0]))
+#define MIN(a,b) ((a) > (b) ? (b) : (a))
+
+#define TARGET_CORE 2
+#define SECONDARY_CORE 3
+
+#define TARGET_SET 15
+
+/* ioctl dev fds */
+static int kvm_dev;
+static int faultcnt;
+
+void
+hexdump(void *data, int len)
+{
+	int i;
+
+	for (i = 0; i < len; i++) {
+		if (i % 16 == 0 && i)
+			printf("\n");
+		printf("%02X ", *(uint8_t *)(data + i));
+	}
+	printf("\n");
+}
+
+bool
+pin_process(pid_t pid, int cpu, bool assert)
+{
+	cpu_set_t cpuset;
+	int ret;
+
+	CPU_ZERO(&cpuset);
+	CPU_SET(cpu, &cpuset);
+	ret = sched_setaffinity(pid, sizeof(cpu_set_t), &cpuset);
+	if (ret < 0) {
+		if (assert) err(1, "sched_setaffinity");
+		return false;
+	}
+
+	return true;
+}
+
+cpc_msrmt_t *
+read_counts()
+{
+	cpc_msrmt_t *counts;
+	int i, ret;
+
+	counts = malloc(L1_SETS * sizeof(cpc_msrmt_t));
+	if (!counts) err(1, "malloc");
+
+	ret = ioctl(kvm_dev, KVM_CPC_READ_COUNTS, counts);
+	if (ret == -1) err(1, "ioctl READ_COUNTS");
+
+	for (i = 0; i < L1_SETS; i++) {
+		if (counts[i] > 8)
+			errx(1, "Invalid counts set %i", i);
+	}	
+
+	return counts;
+}
+
+void
+print_counts(cpc_msrmt_t *counts)
+{
+	int i;
+
+	for (i = 0; i < 64; i++) {
+		if (i % 16 == 0 && i)
+			printf("\n");
+		if (counts[i] == 1)
+			printf("\x1b[38;5;88m");
+		else if (counts[i] > 1)
+			printf("\x1b[38;5;196m");
+		printf("%2i ", i);
+		if (counts[i] > 0)
+			printf("\x1b[0m");
+	}
+	printf("\n");
+}
+
+void
+print_counts_raw(cpc_msrmt_t *counts)
+{
+	int i;
+
+	for (i = 0; i < 64; i++) {
+		if (i % 16 == 0 && i)
+			printf("\n");
+		if (counts[i] == 1)
+			printf("\x1b[38;5;88m");
+		else if (counts[i] > 1)
+			printf("\x1b[38;5;196m");
+		printf("%02X ", (uint8_t) counts[i]);
+		if (counts[i] > 0)
+			printf("\x1b[0m");
+	}
+	printf("\n");
+}
+
+int
+monitor(bool baseline)
+{
+	struct cpc_event event;
+	cpc_msrmt_t counts[64];
+	uint64_t track_mode;
+	int ret, i;
+
+	/* Get page fault info */
+	ret = ioctl(kvm_dev, KVM_CPC_POLL_EVENT, &event);
+	if (!ret) {
+		if (event.type == CPC_EVENT_CPUID) {
+			printf("CPUID EVENT\n");
+			if (event.cpuid.type == CPC_CPUID_START_TRACK) {
+				track_mode = KVM_PAGE_TRACK_ACCESS;
+				ret = ioctl(kvm_dev, KVM_CPC_TRACK_ALL, &track_mode);
+				if (ret) err(1, "ioctl TRACK_ALL");
+			} else if (event.cpuid.type == CPC_CPUID_STOP_TRACK) {
+				track_mode = KVM_PAGE_TRACK_ACCESS;
+				ret = ioctl(kvm_dev, KVM_CPC_UNTRACK_ALL, &track_mode);
+				if (ret) err(1, "ioctl UNTRACK_ALL");
+			}
+			return 0;
+		} else if (event.type != CPC_EVENT_TRACK) {
+			return 0;
+		}
+
+		printf("EVENT\n");
+
+		ret = ioctl(kvm_dev, KVM_CPC_READ_COUNTS, counts);
+		if (ret == -1) err(1, "ioctl READ_COUNTS");
+
+		if (!baseline) {
+			printf("Event: inst:%llu data:%llu retired:%llu\n",
+				event.track.inst_fault_gfn,
+				event.track.data_fault_gfn,
+				event.track.retinst);
+			print_counts(counts);
+			printf("\n");
+		}
+
+		for (i = 0; i < 64; i++) {
+			if (counts[i] > 8) {
+				errx(1, "Invalid count for set %i (%llu)",
+					i, counts[i]);
+			}
+		}
+
+		ret = ioctl(kvm_dev, KVM_CPC_ACK_EVENT, &event.track.id);
+		if (ret == -1) err(1, "ioctl ACK_EVENT");
+
+		faultcnt++;
+	} else if (errno != EAGAIN) {
+		perror("ioctl POLL_EVENT");
+		return 1;
+	}
+
+	return 0;
+}
+
+int
+main(int argc, const char **argv)
+{
+	pid_t pid;
+	uint32_t arg;
+	struct cpc_event event;
+	cpc_msrmt_t baseline[64];
+	int ret, i;
+
+	if (argc <= 1 || !atoi(argv[1])) {
+		printf("Specify qemu process to pin\n");
+		return 0;
+	}
+	
+	kvm_dev = open("/dev/kvm", O_RDWR);
+	if (!kvm_dev) err(1, "open /dev/kvm");
+
+	setvbuf(stdout, NULL, _IONBF, 0);
+
+	pid = atoi(argv[1]);
+	pin_process(pid, TARGET_CORE, true);
+
+	pin_process(0, TARGET_CORE, true);
+
+	/* Setup needed performance counters */
+	ret = ioctl(kvm_dev, KVM_CPC_SETUP_PMC, NULL);
+	if (ret < 0) err(1, "ioctl SETUP_PMC");
+
+	/* Reset previous tracking */
+	ret = ioctl(kvm_dev, KVM_CPC_RESET_TRACKING, NULL);
+	if (ret == -1) err(1, "ioctl RESET_TRACKING");
+
+	/* Do data access stepping */
+	arg = CPC_TRACK_DATA_ACCESS;
+	ret = ioctl(kvm_dev, KVM_CPC_TRACK_MODE, &arg);
+	if (ret == -1) err(1, "ioctl TRACK_MODE");
+
+	pin_process(0, SECONDARY_CORE, true);
+	printf("PINNED\n");
+
+	arg = false;
+	ret = ioctl(kvm_dev, KVM_CPC_SUB_BASELINE, &arg);
+	if (ret == -1) err(1, "ioctl SUB_BASELINE");
+
+	arg = true;
+	ret = ioctl(kvm_dev, KVM_CPC_MEASURE_BASELINE, &arg);
+	if (ret == -1) err(1, "ioctl MEASURE_BASELINE");
+
+	arg = KVM_PAGE_TRACK_ACCESS;
+	ret = ioctl(kvm_dev, KVM_CPC_TRACK_ALL, &arg);
+	if (ret) err(1, "ioctl TRACK_ALL");
+
+	faultcnt = 0;
+	while (faultcnt < 100) {
+		if (monitor(true)) break;
+	}
+
+	arg = KVM_PAGE_TRACK_ACCESS;
+	ret = ioctl(kvm_dev, KVM_CPC_UNTRACK_ALL, &arg);
+	if (ret) err(1, "ioctl TRACK_ALL");
+
+	do {
+		ret = ioctl(kvm_dev, KVM_CPC_POLL_EVENT, &event);
+		if (ret == -1 && errno != EAGAIN)
+			err(1, "ioctl POLL_EVENT");
+	} while (ret == -1 && errno == EAGAIN);
+
+	arg = false;
+	ret = ioctl(kvm_dev, KVM_CPC_MEASURE_BASELINE, &arg);
+	if (ret == -1) err(1, "ioctl MEASURE_BASELINE");
+
+	ret = ioctl(kvm_dev, KVM_CPC_READ_BASELINE, baseline);
+	if (ret == -1) err(1, "ioctl READ_BASELINE");
+
+	printf("\n>>> BASELINE:\n");
+	print_counts(baseline);
+	printf("\n");
+	print_counts_raw(baseline);
+	printf("\n");
+
+	/* Check baseline for saturated sets */
+	for (i = 0; i < 64; i++) {
+		if (baseline[i] >= 8)
+			errx(1, "!!! Baseline set %i full\n", i);
+	}
+
+	arg = true;
+	ret = ioctl(kvm_dev, KVM_CPC_SUB_BASELINE, &arg);
+	if (ret == -1) err(1, "ioctl SUB_BASELINE");
+
+	ret = ioctl(kvm_dev, KVM_CPC_ACK_EVENT, &event.track.id);
+	if (ret == -1) err(1, "ioctl ACK_EVENT");
+
+	faultcnt = 0;
+	while (faultcnt < 10) {
+		if (monitor(false)) break;
+	}
+}
+
diff --git a/test/aes-detect_guest b/test/aes-detect_guest
index de1ec2f..d482fd2 100755
--- a/test/aes-detect_guest
+++ b/test/aes-detect_guest
diff --git a/test/aes-detect_guest.c b/test/aes-detect_guest.c
index 4012b74..80a6a64 100644
--- a/test/aes-detect_guest.c
+++ b/test/aes-detect_guest.c
@@ -30,8 +30,24 @@ main(int argc, const char **argv)
 {
 	struct kcapi_handle *kcapi;
 	uint8_t block[128];
+	uint8_t *buf;
 	size_t n;
 
+	buf = NULL;
+	if (posix_memalign((void *)&buf, L1_LINESIZE * L1_SETS, L1_LINESIZE * L1_SETS))
+		err(1, "memalign");
+	memset(buf, 0, L1_LINESIZE * L1_SETS);
+
+	while (1) {
+		CPC_CPUID_SIGNAL(CPC_CPUID_START_TRACK, 0);
+		
+		buf[L1_LINESIZE * 5] += 1;
+
+		CPC_CPUID_SIGNAL(CPC_CPUID_START_TRACK, 0);
+	}
+
+	return 0;
+
 	kcapi = NULL;
 	if (kcapi_cipher_init(&kcapi, "ecb(aes)", 0))
 		err(1, "kcapi init");
diff --git a/test/sevstep.c b/test/sevstep.c
index 3307245..40130c5 100644
--- a/test/sevstep.c
+++ b/test/sevstep.c
@@ -481,7 +481,7 @@ int
 monitor(struct kvm *kvm, bool baseline)
 {
 	static uint64_t rip_prev = 1;
-	struct cpc_track_event event;
+	struct cpc_event event;
 	cpc_msrmt_t counts[64];
 	uint64_t rip;
 	int ret, i;
@@ -489,17 +489,24 @@ monitor(struct kvm *kvm, bool baseline)
 	/* Get page fault info */
 	ret = ioctl(kvm_dev, KVM_CPC_POLL_EVENT, &event);
 	if (!ret) {
+		if (event.type != CPC_EVENT_TRACK)
+			return 0;
+
 		ret = ioctl(kvm_dev, KVM_CPC_READ_COUNTS, counts);
 		if (ret == -1) err(1, "ioctl READ_COUNTS");
 
 		rip = sev_dbg_rip(kvm->vmfd);
 		if (!baseline && rip != rip_prev) {
 			printf("Event: inst:%llu data:%llu retired:%llu rip:%lu\n",
-				event.inst_fault_gfn, event.data_fault_gfn,
-				event.retinst, rip);
+				event.track.inst_fault_gfn,
+				event.track.data_fault_gfn,
+				event.track.retinst, rip);
 			print_counts(counts);
 			printf("\n");
 			rip_prev = rip;
+			faultcnt++;
+		} else if (baseline) {
+			faultcnt++;
 		}
 
 		for (i = 0; i < 64; i++) {
@@ -511,8 +518,6 @@ monitor(struct kvm *kvm, bool baseline)
 
 		ret = ioctl(kvm_dev, KVM_CPC_ACK_EVENT, &event.id);
 		if (ret == -1) err(1, "ioctl ACK_EVENT");
-
-		faultcnt++;
 	} else if (errno != EAGAIN) {
 		perror("ioctl POLL_EVENT");
 		return 1;
@@ -528,7 +533,7 @@ main(int argc, const char **argv)
 	uint64_t track_mode;
 	pid_t ppid, pid;
 	uint32_t arg;
-	struct cpc_track_event event;
+	struct cpc_event event;
 	cpc_msrmt_t baseline[64];
 	int ret, i;
 
@@ -571,6 +576,14 @@ main(int argc, const char **argv)
 	ret = ioctl(kvm_dev, KVM_CPC_TRACK_ALL, &track_mode);
 	if (ret == -1) err(1, "ioctl TRACK_ALL");
 
+	arg = false;
+	ret = ioctl(kvm_dev, KVM_CPC_SUB_BASELINE, &arg);
+	if (ret == -1) err(1, "ioctl SUB_BASELINE");
+
+	arg = true;
+	ret = ioctl(kvm_dev, KVM_CPC_MEASURE_BASELINE, &arg);
+	if (ret == -1) err(1, "ioctl MEASURE_BASELINE");
+
 	ppid = getpid();
 	if ((pid = fork())) {
 		if (pid < 0) err(1, "fork");
@@ -584,12 +597,8 @@ main(int argc, const char **argv)
 		pin_process(0, SECONDARY_CORE, true);
 		printf("PINNED\n");
 
-		arg = true;
-		ret = ioctl(kvm_dev, KVM_CPC_MEASURE_BASELINE, &arg);
-		if (ret == -1) err(1, "ioctl MEASURE_BASELINE");
-
 		faultcnt = 0;
-		while (faultcnt < 50) {
+		while (faultcnt < 300) {
 			if (monitor(&kvm_with_access, true)) break;
 		}
 
@@ -626,7 +635,7 @@ main(int argc, const char **argv)
 		if (ret == -1) err(1, "ioctl ACK_EVENT");
 
 		faultcnt = 0;
-		while (faultcnt < 50) {
+		while (faultcnt < 20) {
 			if (monitor(&kvm_with_access, false)) break;
 		}