hyperv-fixes for v6.13-rc4

-----BEGIN PGP SIGNATURE-----
 
 iQFHBAABCgAxFiEEIbPD0id6easf0xsudhRwX5BBoF4FAmdhxFsTHHdlaS5saXVA
 a2VybmVsLm9yZwAKCRB2FHBfkEGgXmGuB/0ZsXKBh9eCZnmVSZwdVkcSeE/HEHHc
 K5qeIxjqirAzajvLbAtgnTQpi5w0kLYYOT/X/8b6L1nnefcfk1cz5cbyQKxXx7J6
 Zuob5KpjX7FtJbMc7QQtzrLyApw2k2OYe1QPCRxuWkYzQjaus/6kM27ivjjYFqDs
 Qv6IKCcebomAbJTN8IwF38KsFQ2JS3moWr+kcVyna7+Kg1ymEto2QlnI01Z+M1n7
 BUYvLgYUD/5HzbNjqC0HysUuX5PsVQ45OsmTgjkvs/XzLcAGGEHJGrJHJkEccx7H
 AsiNzHoB6wUrFhCAX7IPCSE0g9vjtwI21ozqYllzOTa/Q5KnUE+q9fhJ
 =Fw3U
 -----END PGP SIGNATURE-----

Merge tag 'hyperv-fixes-signed-20241217' of git://git.kernel.org/pub/scm/linux/kernel/git/hyperv/linux

Pull hyperv fixes from Wei Liu:

 - Various fixes to Hyper-V tools in the kernel tree (Dexuan Cui, Olaf
   Hering, Vitaly Kuznetsov)

 - Fix a bug in the Hyper-V TSC page based sched_clock() (Naman Jain)

 - Two bug fixes in the Hyper-V utility functions (Michael Kelley)

 - Convert open-coded timeouts to secs_to_jiffies() in Hyper-V drivers
   (Easwar Hariharan)

* tag 'hyperv-fixes-signed-20241217' of git://git.kernel.org/pub/scm/linux/kernel/git/hyperv/linux:
  tools/hv: reduce resource usage in hv_kvp_daemon
  tools/hv: add a .gitignore file
  tools/hv: reduce resouce usage in hv_get_dns_info helper
  hv/hv_kvp_daemon: Pass NIC name to hv_get_dns_info as well
  Drivers: hv: util: Avoid accessing a ringbuffer not initialized yet
  Drivers: hv: util: Don't force error code to ENODEV in util_probe()
  tools/hv: terminate fcopy daemon if read from uio fails
  drivers: hv: Convert open-coded timeouts to secs_to_jiffies()
  tools: hv: change permissions of NetworkManager configuration file
  x86/hyperv: Fix hv tsc page based sched_clock for hibernation
  tools: hv: Fix a complier warning in the fcopy uio daemon

arch/x86/kernel/cpu/mshyperv.c

@@ -223,6 +223,63 @@ static void hv_machine_crash_shutdown(struct pt_regs *regs)
 	hyperv_cleanup();
 }
 #endif /* CONFIG_CRASH_DUMP */
+
+static u64 hv_ref_counter_at_suspend;
+static void (*old_save_sched_clock_state)(void);
+static void (*old_restore_sched_clock_state)(void);
+
+/*
+ * Hyper-V clock counter resets during hibernation. Save and restore clock
+ * offset during suspend/resume, while also considering the time passed
+ * before suspend. This is to make sure that sched_clock using hv tsc page
+ * based clocksource, proceeds from where it left off during suspend and
+ * it shows correct time for the timestamps of kernel messages after resume.
+ */
+static void save_hv_clock_tsc_state(void)
+{
+	hv_ref_counter_at_suspend = hv_read_reference_counter();
+}
+
+static void restore_hv_clock_tsc_state(void)
+{
+	/*
+	 * Adjust the offsets used by hv tsc clocksource to
+	 * account for the time spent before hibernation.
+	 * adjusted value = reference counter (time) at suspend
+	 *                - reference counter (time) now.
+	 */
+	hv_adj_sched_clock_offset(hv_ref_counter_at_suspend - hv_read_reference_counter());
+}
+
+/*
+ * Functions to override save_sched_clock_state and restore_sched_clock_state
+ * functions of x86_platform. The Hyper-V clock counter is reset during
+ * suspend-resume and the offset used to measure time needs to be
+ * corrected, post resume.
+ */
+static void hv_save_sched_clock_state(void)
+{
+	old_save_sched_clock_state();
+	save_hv_clock_tsc_state();
+}
+
+static void hv_restore_sched_clock_state(void)
+{
+	restore_hv_clock_tsc_state();
+	old_restore_sched_clock_state();
+}
+
+static void __init x86_setup_ops_for_tsc_pg_clock(void)
+{
+	if (!(ms_hyperv.features & HV_MSR_REFERENCE_TSC_AVAILABLE))
+		return;
+
+	old_save_sched_clock_state = x86_platform.save_sched_clock_state;
+	x86_platform.save_sched_clock_state = hv_save_sched_clock_state;
+
+	old_restore_sched_clock_state = x86_platform.restore_sched_clock_state;
+	x86_platform.restore_sched_clock_state = hv_restore_sched_clock_state;
+}
 #endif /* CONFIG_HYPERV */
 
 static uint32_t  __init ms_hyperv_platform(void)
@@ -579,6 +636,7 @@ static void __init ms_hyperv_init_platform(void)
 
 	/* Register Hyper-V specific clocksource */
 	hv_init_clocksource();
+	x86_setup_ops_for_tsc_pg_clock();
 	hv_vtl_init_platform();
 #endif
 	/*

drivers/clocksource/hyperv_timer.c

@@ -27,7 +27,8 @@
 #include <asm/mshyperv.h>
 
 static struct clock_event_device __percpu *hv_clock_event;
-static u64 hv_sched_clock_offset __ro_after_init;
+/* Note: offset can hold negative values after hibernation. */
+static u64 hv_sched_clock_offset __read_mostly;
 
 /*
  * If false, we're using the old mechanism for stimer0 interrupts
@@ -470,6 +471,17 @@ static void resume_hv_clock_tsc(struct clocksource *arg)
 	hv_set_msr(HV_MSR_REFERENCE_TSC, tsc_msr.as_uint64);
 }
 
+/*
+ * Called during resume from hibernation, from overridden
+ * x86_platform.restore_sched_clock_state routine. This is to adjust offsets
+ * used to calculate time for hv tsc page based sched_clock, to account for
+ * time spent before hibernation.
+ */
+void hv_adj_sched_clock_offset(u64 offset)
+{
+	hv_sched_clock_offset -= offset;
+}
+
 #ifdef HAVE_VDSO_CLOCKMODE_HVCLOCK
 static int hv_cs_enable(struct clocksource *cs)
 {

drivers/hv/hv_balloon.c

@@ -756,7 +756,7 @@ static void hv_mem_hot_add(unsigned long start, unsigned long size,
 		 * adding succeeded, it is ok to proceed even if the memory was
 		 * not onlined in time.
 		 */
-		wait_for_completion_timeout(&dm_device.ol_waitevent, 5 * HZ);
+		wait_for_completion_timeout(&dm_device.ol_waitevent, secs_to_jiffies(5));
 		post_status(&dm_device);
 	}
 }
@@ -1373,7 +1373,8 @@ static int dm_thread_func(void *dm_dev)
 	struct hv_dynmem_device *dm = dm_dev;
 
 	while (!kthread_should_stop()) {
-		wait_for_completion_interruptible_timeout(&dm_device.config_event, 1 * HZ);
+		wait_for_completion_interruptible_timeout(&dm_device.config_event,
+								secs_to_jiffies(1));
 		/*
 		 * The host expects us to post information on the memory
 		 * pressure every second.
@@ -1748,7 +1749,7 @@ static int balloon_connect_vsp(struct hv_device *dev)
 	if (ret)
 		goto out;
 
-	t = wait_for_completion_timeout(&dm_device.host_event, 5 * HZ);
+	t = wait_for_completion_timeout(&dm_device.host_event, secs_to_jiffies(5));
 	if (t == 0) {
 		ret = -ETIMEDOUT;
 		goto out;
@@ -1806,7 +1807,7 @@ static int balloon_connect_vsp(struct hv_device *dev)
 	if (ret)
 		goto out;
 
-	t = wait_for_completion_timeout(&dm_device.host_event, 5 * HZ);
+	t = wait_for_completion_timeout(&dm_device.host_event, secs_to_jiffies(5));
 	if (t == 0) {
 		ret = -ETIMEDOUT;
 		goto out;

drivers/hv/hv_kvp.c

@@ -655,7 +655,7 @@ void hv_kvp_onchannelcallback(void *context)
 		if (host_negotiatied == NEGO_NOT_STARTED) {
 			host_negotiatied = NEGO_IN_PROGRESS;
 			schedule_delayed_work(&kvp_host_handshake_work,
-				      HV_UTIL_NEGO_TIMEOUT * HZ);
+						secs_to_jiffies(HV_UTIL_NEGO_TIMEOUT));
 		}
 		return;
 	}
@@ -724,7 +724,7 @@ void hv_kvp_onchannelcallback(void *context)
 		 */
 		schedule_work(&kvp_sendkey_work);
 		schedule_delayed_work(&kvp_timeout_work,
-					HV_UTIL_TIMEOUT * HZ);
+				      secs_to_jiffies(HV_UTIL_TIMEOUT));
 
 		return;
 
@@ -767,6 +767,12 @@ hv_kvp_init(struct hv_util_service *srv)
 	 */
 	kvp_transaction.state = HVUTIL_DEVICE_INIT;
 
+	return 0;
+}
+
+int
+hv_kvp_init_transport(void)
+{
 	hvt = hvutil_transport_init(kvp_devname, CN_KVP_IDX, CN_KVP_VAL,
 				    kvp_on_msg, kvp_on_reset);
 	if (!hvt)

drivers/hv/hv_snapshot.c

@@ -193,7 +193,8 @@ static void vss_send_op(void)
 	vss_transaction.state = HVUTIL_USERSPACE_REQ;
 
 	schedule_delayed_work(&vss_timeout_work, op == VSS_OP_FREEZE ?
-			VSS_FREEZE_TIMEOUT * HZ : HV_UTIL_TIMEOUT * HZ);
+				secs_to_jiffies(VSS_FREEZE_TIMEOUT) :
+				secs_to_jiffies(HV_UTIL_TIMEOUT));
 
 	rc = hvutil_transport_send(hvt, vss_msg, sizeof(*vss_msg), NULL);
 	if (rc) {
@@ -388,6 +389,12 @@ hv_vss_init(struct hv_util_service *srv)
 	 */
 	vss_transaction.state = HVUTIL_DEVICE_INIT;
 
+	return 0;
+}
+
+int
+hv_vss_init_transport(void)
+{
 	hvt = hvutil_transport_init(vss_devname, CN_VSS_IDX, CN_VSS_VAL,
 				    vss_on_msg, vss_on_reset);
 	if (!hvt) {

drivers/hv/hv_util.c

@@ -141,6 +141,7 @@ static struct hv_util_service util_heartbeat = {
 static struct hv_util_service util_kvp = {
 	.util_cb = hv_kvp_onchannelcallback,
 	.util_init = hv_kvp_init,
+	.util_init_transport = hv_kvp_init_transport,
 	.util_pre_suspend = hv_kvp_pre_suspend,
 	.util_pre_resume = hv_kvp_pre_resume,
 	.util_deinit = hv_kvp_deinit,
@@ -149,6 +150,7 @@ static struct hv_util_service util_kvp = {
 static struct hv_util_service util_vss = {
 	.util_cb = hv_vss_onchannelcallback,
 	.util_init = hv_vss_init,
+	.util_init_transport = hv_vss_init_transport,
 	.util_pre_suspend = hv_vss_pre_suspend,
 	.util_pre_resume = hv_vss_pre_resume,
 	.util_deinit = hv_vss_deinit,
@@ -590,11 +592,9 @@ static int util_probe(struct hv_device *dev,
 	srv->channel = dev->channel;
 	if (srv->util_init) {
 		ret = srv->util_init(srv);
-		if (ret) {
-			ret = -ENODEV;
+		if (ret)
 			goto error1;
 	}
-	}
 
 	/*
 	 * The set of services managed by the util driver are not performance
@@ -613,6 +613,13 @@ static int util_probe(struct hv_device *dev,
 	if (ret)
 		goto error;
 
+	if (srv->util_init_transport) {
+		ret = srv->util_init_transport();
+		if (ret) {
+			vmbus_close(dev->channel);
+			goto error;
+		}
+	}
 	return 0;
 
 error:

drivers/hv/hyperv_vmbus.h

@@ -370,12 +370,14 @@ void vmbus_on_event(unsigned long data);
 void vmbus_on_msg_dpc(unsigned long data);
 
 int hv_kvp_init(struct hv_util_service *srv);
+int hv_kvp_init_transport(void);
 void hv_kvp_deinit(void);
 int hv_kvp_pre_suspend(void);
 int hv_kvp_pre_resume(void);
 void hv_kvp_onchannelcallback(void *context);
 
 int hv_vss_init(struct hv_util_service *srv);
+int hv_vss_init_transport(void);
 void hv_vss_deinit(void);
 int hv_vss_pre_suspend(void);
 int hv_vss_pre_resume(void);

drivers/hv/vmbus_drv.c

@@ -2507,7 +2507,7 @@ static int vmbus_bus_resume(struct device *dev)
 	vmbus_request_offers();
 
 	if (wait_for_completion_timeout(
-		&vmbus_connection.ready_for_resume_event, 10 * HZ) == 0)
+		&vmbus_connection.ready_for_resume_event, secs_to_jiffies(10)) == 0)
 		pr_err("Some vmbus device is missing after suspending?\n");
 
 	/* Reset the event for the next suspend. */

include/clocksource/hyperv_timer.h

@@ -38,6 +38,8 @@ extern void hv_remap_tsc_clocksource(void);
 extern unsigned long hv_get_tsc_pfn(void);
 extern struct ms_hyperv_tsc_page *hv_get_tsc_page(void);
 
+extern void hv_adj_sched_clock_offset(u64 offset);
+
 static __always_inline bool
 hv_read_tsc_page_tsc(const struct ms_hyperv_tsc_page *tsc_pg,
 		     u64 *cur_tsc, u64 *time)

include/linux/hyperv.h

@@ -1559,6 +1559,7 @@ struct hv_util_service {
 	void *channel;
 	void (*util_cb)(void *);
 	int (*util_init)(struct hv_util_service *);
+	int (*util_init_transport)(void);
 	void (*util_deinit)(void);
 	int (*util_pre_suspend)(void);
 	int (*util_pre_resume)(void);

tools/hv/.gitignore

@@ -0,0 +1,3 @@
+hv_fcopy_uio_daemon
+hv_kvp_daemon
+hv_vss_daemon

tools/hv/hv_fcopy_uio_daemon.c

@@ -35,8 +35,6 @@
 #define WIN8_SRV_MINOR		1
 #define WIN8_SRV_VERSION	(WIN8_SRV_MAJOR << 16 | WIN8_SRV_MINOR)
 
-#define MAX_FOLDER_NAME		15
-#define MAX_PATH_LEN		15
 #define FCOPY_UIO		"/sys/bus/vmbus/devices/eb765408-105f-49b6-b4aa-c123b64d17d4/uio"
 
 #define FCOPY_VER_COUNT		1
@@ -51,7 +49,7 @@ static const int fw_versions[] = {
 
 #define HV_RING_SIZE		0x4000 /* 16KB ring buffer size */
 
-unsigned char desc[HV_RING_SIZE];
+static unsigned char desc[HV_RING_SIZE];
 
 static int target_fd;
 static char target_fname[PATH_MAX];
@@ -409,8 +407,8 @@ int main(int argc, char *argv[])
 	struct vmbus_br txbr, rxbr;
 	void *ring;
 	uint32_t len = HV_RING_SIZE;
-	char uio_name[MAX_FOLDER_NAME] = {0};
-	char uio_dev_path[MAX_PATH_LEN] = {0};
+	char uio_name[NAME_MAX] = {0};
+	char uio_dev_path[PATH_MAX] = {0};
 
 	static struct option long_options[] = {
 		{"help",	no_argument,	   0,  'h' },
@@ -468,8 +466,10 @@ int main(int argc, char *argv[])
 		 */
 		ret = pread(fcopy_fd, &tmp, sizeof(int), 0);
 		if (ret < 0) {
-			syslog(LOG_ERR, "pread failed: %s", strerror(errno));
+			if (errno == EINTR || errno == EAGAIN)
 				continue;
+			syslog(LOG_ERR, "pread failed: %s", strerror(errno));
+			goto close;
 		}
 
 		len = HV_RING_SIZE;

tools/hv/hv_get_dns_info.sh

@@ -1,4 +1,4 @@
-#!/bin/bash
+#!/bin/sh
 
 # This example script parses /etc/resolv.conf to retrive DNS information.
 # In the interest of keeping the KVP daemon code free of distro specific
@@ -10,4 +10,4 @@
 # this script can be based on the Network Manager APIs for retrieving DNS
 # entries.
 
-cat /etc/resolv.conf 2>/dev/null | awk '/^nameserver/ { print $2 }'
+exec awk '/^nameserver/ { print $2 }' /etc/resolv.conf 2>/dev/null

tools/hv/hv_kvp_daemon.c

@@ -725,7 +725,7 @@ static void kvp_get_ipconfig_info(char *if_name,
 	 * .
 	 */
 
-	sprintf(cmd, KVP_SCRIPTS_PATH "%s",  "hv_get_dns_info");
+	sprintf(cmd, "exec %s %s", KVP_SCRIPTS_PATH "hv_get_dns_info", if_name);
 
 	/*
 	 * Execute the command to gather DNS info.
@@ -742,7 +742,7 @@ static void kvp_get_ipconfig_info(char *if_name,
 	 * Enabled: DHCP enabled.
 	 */
 
-	sprintf(cmd, KVP_SCRIPTS_PATH "%s %s", "hv_get_dhcp_info", if_name);
+	sprintf(cmd, "exec %s %s", KVP_SCRIPTS_PATH "hv_get_dhcp_info", if_name);
 
 	file = popen(cmd, "r");
 	if (file == NULL)
@@ -1606,8 +1606,9 @@ static int kvp_set_ip_info(char *if_name, struct hv_kvp_ipaddr_value *new_val)
 	 * invoke the external script to do its magic.
 	 */
 
-	str_len = snprintf(cmd, sizeof(cmd), KVP_SCRIPTS_PATH "%s %s %s",
-			   "hv_set_ifconfig", if_filename, nm_filename);
+	str_len = snprintf(cmd, sizeof(cmd), "exec %s %s %s",
+			   KVP_SCRIPTS_PATH "hv_set_ifconfig",
+			   if_filename, nm_filename);
 	/*
 	 * This is a little overcautious, but it's necessary to suppress some
 	 * false warnings from gcc 8.0.1.

tools/hv/hv_set_ifconfig.sh

@@ -81,7 +81,7 @@ echo "ONBOOT=yes" >> $1
 
 cp $1 /etc/sysconfig/network-scripts/
 
-chmod 600 $2
+umask 0177
 interface=$(echo $2 | awk -F - '{ print $2 }')
 filename="${2##*/}"