[rocky10_0] History Rebuild to kernel-6.12.0-55.30.1.el10_0

Makefile.rhelver

@@ -12,7 +12,7 @@ RHEL_MINOR = 0
 #
 # Use this spot to avoid future merge conflicts.
 # Do not trim this comment.
-RHEL_RELEASE = 55.27.1
+RHEL_RELEASE = 55.30.1
 
 #
 # RHEL_REBASE_NUM

arch/powerpc/platforms/book3s/vas-api.c

@@ -464,7 +464,43 @@ static vm_fault_t vas_mmap_fault(struct vm_fault *vmf)
 	return VM_FAULT_SIGBUS;
 }
 
+/*
+ * During mmap() paste address, mapping VMA is saved in VAS window
+ * struct which is used to unmap during migration if the window is
+ * still open. But the user space can remove this mapping with
+ * munmap() before closing the window and the VMA address will
+ * be invalid. Set VAS window VMA to NULL in this function which
+ * is called before VMA free.
+ */
+static void vas_mmap_close(struct vm_area_struct *vma)
+{
+	struct file *fp = vma->vm_file;
+	struct coproc_instance *cp_inst = fp->private_data;
+	struct vas_window *txwin;
+
+	/* Should not happen */
+	if (!cp_inst || !cp_inst->txwin) {
+		pr_err("No attached VAS window for the paste address mmap\n");
+		return;
+	}
+
+	txwin = cp_inst->txwin;
+	/*
+	 * task_ref.vma is set in coproc_mmap() during mmap paste
+	 * address. So it has to be the same VMA that is getting freed.
+	 */
+	if (WARN_ON(txwin->task_ref.vma != vma)) {
+		pr_err("Invalid paste address mmaping\n");
+		return;
+	}
+
+	mutex_lock(&txwin->task_ref.mmap_mutex);
+	txwin->task_ref.vma = NULL;
+	mutex_unlock(&txwin->task_ref.mmap_mutex);
+}
+
 static const struct vm_operations_struct vas_vm_ops = {
+	.close = vas_mmap_close,
 	.fault = vas_mmap_fault,
 };
 

arch/s390/kernel/topology.c

@@ -555,6 +555,16 @@ static void __init alloc_masks(struct sysinfo_15_1_x *info,
 	}
 }
 
+static int __init detect_polarization(union topology_entry *tle)
+{
+	struct topology_core *tl_core;
+
+	while (tle->nl)
+		tle = next_tle(tle);
+	tl_core = (struct topology_core *)tle;
+	return tl_core->pp != POLARIZATION_HRZ;
+}
+
 void __init topology_init_early(void)
 {
 	struct sysinfo_15_1_x *info;
@@ -574,6 +584,7 @@ void __init topology_init_early(void)
 		      __func__, PAGE_SIZE, PAGE_SIZE);
 	info = tl_info;
 	store_topology(info);
+	cpu_management = detect_polarization(info->tle);
 	pr_info("The CPU configuration topology of the machine is: %d %d %d %d %d %d / %d\n",
 		info->mag[0], info->mag[1], info->mag[2], info->mag[3],
 		info->mag[4], info->mag[5], info->mnest);

arch/s390/pci/pci.c

@@ -44,6 +44,7 @@
 /* list of all detected zpci devices */
 static LIST_HEAD(zpci_list);
 static DEFINE_SPINLOCK(zpci_list_lock);
+static DEFINE_MUTEX(zpci_add_remove_lock);
 
 static DECLARE_BITMAP(zpci_domain, ZPCI_DOMAIN_BITMAP_SIZE);
 static DEFINE_SPINLOCK(zpci_domain_lock);
@@ -69,6 +70,15 @@ EXPORT_SYMBOL_GPL(zpci_aipb);
 struct airq_iv *zpci_aif_sbv;
 EXPORT_SYMBOL_GPL(zpci_aif_sbv);
 
+void zpci_zdev_put(struct zpci_dev *zdev)
+{
+	if (!zdev)
+		return;
+	mutex_lock(&zpci_add_remove_lock);
+	kref_put_lock(&zdev->kref, zpci_release_device, &zpci_list_lock);
+	mutex_unlock(&zpci_add_remove_lock);
+}
+
 struct zpci_dev *get_zdev_by_fid(u32 fid)
 {
 	struct zpci_dev *tmp, *zdev = NULL;
@@ -831,6 +841,7 @@ int zpci_add_device(struct zpci_dev *zdev)
 {
 	int rc;
 
+	mutex_lock(&zpci_add_remove_lock);
 	zpci_dbg(1, "add fid:%x, fh:%x, c:%d\n", zdev->fid, zdev->fh, zdev->state);
 	rc = zpci_init_iommu(zdev);
 	if (rc)
@@ -844,12 +855,14 @@ int zpci_add_device(struct zpci_dev *zdev)
 	spin_lock(&zpci_list_lock);
 	list_add_tail(&zdev->entry, &zpci_list);
 	spin_unlock(&zpci_list_lock);
+	mutex_unlock(&zpci_add_remove_lock);
 	return 0;
 
 error_destroy_iommu:
 	zpci_destroy_iommu(zdev);
 error:
 	zpci_dbg(0, "add fid:%x, rc:%d\n", zdev->fid, rc);
+	mutex_unlock(&zpci_add_remove_lock);
 	return rc;
 }
 
@@ -919,60 +932,44 @@ int zpci_deconfigure_device(struct zpci_dev *zdev)
  * @zdev: the zpci_dev that was reserved
  *
  * Handle the case that a given zPCI function was reserved by another system.
- * After a call to this function the zpci_dev can not be found via
- * get_zdev_by_fid() anymore but may still be accessible via existing
- * references though it will not be functional anymore.
  */
 void zpci_device_reserved(struct zpci_dev *zdev)
 {
-	/*
-	 * Remove device from zpci_list as it is going away. This also
-	 * makes sure we ignore subsequent zPCI events for this device.
-	 */
-	spin_lock(&zpci_list_lock);
-	list_del(&zdev->entry);
-	spin_unlock(&zpci_list_lock);
+	lockdep_assert_held(&zdev->state_lock);
+	/* We may declare the device reserved multiple times */
+	if (zdev->state == ZPCI_FN_STATE_RESERVED)
+		return;
 	zdev->state = ZPCI_FN_STATE_RESERVED;
 	zpci_dbg(3, "rsv fid:%x\n", zdev->fid);
+	/*
+	 * The underlying device is gone. Allow the zdev to be freed
+	 * as soon as all other references are gone by accounting for
+	 * the removal as a dropped reference.
+	 */
 	zpci_zdev_put(zdev);
 }
 
 void zpci_release_device(struct kref *kref)
 {
 	struct zpci_dev *zdev = container_of(kref, struct zpci_dev, kref);
-	int ret;
+
+	lockdep_assert_held(&zpci_add_remove_lock);
+	WARN_ON(zdev->state != ZPCI_FN_STATE_RESERVED);
+	/*
+	 * We already hold zpci_list_lock thanks to kref_put_lock().
+	 * This makes sure no new reference can be taken from the list.
+	 */
+	list_del(&zdev->entry);
+	spin_unlock(&zpci_list_lock);
 
 	if (zdev->has_hp_slot)
 		zpci_exit_slot(zdev);
 
-	if (zdev->zbus->bus)
-		zpci_bus_remove_device(zdev, false);
-
-	if (zdev_enabled(zdev))
-		zpci_disable_device(zdev);
+	if (zdev->has_resources)
+		zpci_cleanup_bus_resources(zdev);
 
-	switch (zdev->state) {
-	case ZPCI_FN_STATE_CONFIGURED:
-		ret = sclp_pci_deconfigure(zdev->fid);
-		zpci_dbg(3, "deconf fid:%x, rc:%d\n", zdev->fid, ret);
-		fallthrough;
-	case ZPCI_FN_STATE_STANDBY:
-		if (zdev->has_hp_slot)
-			zpci_exit_slot(zdev);
-		spin_lock(&zpci_list_lock);
-		list_del(&zdev->entry);
-		spin_unlock(&zpci_list_lock);
-		zpci_dbg(3, "rsv fid:%x\n", zdev->fid);
-		fallthrough;
-	case ZPCI_FN_STATE_RESERVED:
-		if (zdev->has_resources)
-			zpci_cleanup_bus_resources(zdev);
-		zpci_bus_device_unregister(zdev);
-		zpci_destroy_iommu(zdev);
-		fallthrough;
-	default:
-		break;
-	}
+	zpci_bus_device_unregister(zdev);
+	zpci_destroy_iommu(zdev);
 	zpci_dbg(3, "rem fid:%x\n", zdev->fid);
 	kfree_rcu(zdev, rcu);
 }

arch/s390/pci/pci_bus.c

@@ -171,7 +171,6 @@ void zpci_bus_scan_busses(void)
 static bool zpci_bus_is_multifunction_root(struct zpci_dev *zdev)
 {
 	return !s390_pci_no_rid && zdev->rid_available &&
-		zpci_is_device_configured(zdev) &&
 		!zdev->vfn;
 }
 
@@ -332,6 +331,20 @@ static int zpci_bus_add_device(struct zpci_bus *zbus, struct zpci_dev *zdev)
 	return rc;
 }
 
+static bool zpci_bus_is_isolated_vf(struct zpci_bus *zbus, struct zpci_dev *zdev)
+{
+	struct pci_dev *pdev;
+
+	if (!zdev->vfn)
+		return false;
+
+	pdev = zpci_iov_find_parent_pf(zbus, zdev);
+	if (!pdev)
+		return true;
+	pci_dev_put(pdev);
+	return false;
+}
+
 int zpci_bus_device_register(struct zpci_dev *zdev, struct pci_ops *ops)
 {
 	bool topo_is_tid = zdev->tid_avail;
@@ -346,6 +359,15 @@ int zpci_bus_device_register(struct zpci_dev *zdev, struct pci_ops *ops)
 
 	topo = topo_is_tid ? zdev->tid : zdev->pchid;
 	zbus = zpci_bus_get(topo, topo_is_tid);
+	/*
+	 * An isolated VF gets its own domain/bus even if there exists
+	 * a matching domain/bus already
+	 */
+	if (zbus && zpci_bus_is_isolated_vf(zbus, zdev)) {
+		zpci_bus_put(zbus);
+		zbus = NULL;
+	}
+
 	if (!zbus) {
 		zbus = zpci_bus_alloc(topo, topo_is_tid);
 		if (!zbus)

arch/s390/pci/pci_bus.h

@@ -17,11 +17,8 @@ int zpci_bus_scan_device(struct zpci_dev *zdev);
 void zpci_bus_remove_device(struct zpci_dev *zdev, bool set_error);
 
 void zpci_release_device(struct kref *kref);
-static inline void zpci_zdev_put(struct zpci_dev *zdev)
-{
-	if (zdev)
-		kref_put(&zdev->kref, zpci_release_device);
-}
+
+void zpci_zdev_put(struct zpci_dev *zdev);
 
 static inline void zpci_zdev_get(struct zpci_dev *zdev)
 {

arch/s390/pci/pci_clp.c

@@ -423,6 +423,8 @@ static void __clp_add(struct clp_fh_list_entry *entry, void *data)
 		return;
 	}
 	zdev = zpci_create_device(entry->fid, entry->fh, entry->config_state);
+	if (IS_ERR(zdev))
+		return;
 	list_add_tail(&zdev->entry, scan_list);
 }
 

arch/s390/pci/pci_event.c

@@ -335,6 +335,22 @@ static void zpci_event_hard_deconfigured(struct zpci_dev *zdev, u32 fh)
 	zdev->state = ZPCI_FN_STATE_STANDBY;
 }
 
+static void zpci_event_reappear(struct zpci_dev *zdev)
+{
+	lockdep_assert_held(&zdev->state_lock);
+	/*
+	 * The zdev is in the reserved state. This means that it was presumed to
+	 * go away but there are still undropped references. Now, the platform
+	 * announced its availability again. Bring back the lingering zdev
+	 * to standby. This is safe because we hold a temporary reference
+	 * now so that it won't go away. Account for the re-appearance of the
+	 * underlying device by incrementing the reference count.
+	 */
+	zdev->state = ZPCI_FN_STATE_STANDBY;
+	zpci_zdev_get(zdev);
+	zpci_dbg(1, "rea fid:%x, fh:%x\n", zdev->fid, zdev->fh);
+}
+
 static void __zpci_event_availability(struct zpci_ccdf_avail *ccdf)
 {
 	struct zpci_dev *zdev = get_zdev_by_fid(ccdf->fid);
@@ -358,8 +374,10 @@ static void __zpci_event_availability(struct zpci_ccdf_avail *ccdf)
 				break;
 			}
 		} else {
+			if (zdev->state == ZPCI_FN_STATE_RESERVED)
+				zpci_event_reappear(zdev);
 			/* the configuration request may be stale */
-			if (zdev->state != ZPCI_FN_STATE_STANDBY)
+			else if (zdev->state != ZPCI_FN_STATE_STANDBY)
 				break;
 			zdev->state = ZPCI_FN_STATE_CONFIGURED;
 		}
@@ -375,6 +393,8 @@ static void __zpci_event_availability(struct zpci_ccdf_avail *ccdf)
 				break;
 			}
 		} else {
+			if (zdev->state == ZPCI_FN_STATE_RESERVED)
+				zpci_event_reappear(zdev);
 			zpci_update_fh(zdev, ccdf->fh);
 		}
 		break;

arch/s390/pci/pci_iov.c

@@ -60,18 +60,35 @@ static int zpci_iov_link_virtfn(struct pci_dev *pdev, struct pci_dev *virtfn, in
 	return 0;
 }
 
-int zpci_iov_setup_virtfn(struct zpci_bus *zbus, struct pci_dev *virtfn, int vfn)
+/**
+ * zpci_iov_find_parent_pf - Find the parent PF, if any, of the given function
+ * @zbus:	The bus that the PCI function is on, or would be added on
+ * @zdev:	The PCI function
+ *
+ * Finds the parent PF, if it exists and is configured, of the given PCI function
+ * and increments its refcount. Th PF is searched for on the provided bus so the
+ * caller has to ensure that this is the correct bus to search. This function may
+ * be used before adding the PCI function to a zbus.
+ *
+ * Return: Pointer to the struct pci_dev of the parent PF or NULL if it not
+ * found. If the function is not a VF or has no RequesterID information,
+ * NULL is returned as well.
+ */
+struct pci_dev *zpci_iov_find_parent_pf(struct zpci_bus *zbus, struct zpci_dev *zdev)
 {
-	int i, cand_devfn;
-	struct zpci_dev *zdev;
+	int i, vfid, devfn, cand_devfn;
 	struct pci_dev *pdev;
-	int vfid = vfn - 1; /* Linux' vfid's start at 0 vfn at 1*/
-	int rc = 0;
 
 	if (!zbus->multifunction)
-		return 0;
-
-	/* If the parent PF for the given VF is also configured in the
+		return NULL;
+	/* Non-VFs and VFs without RID available don't have a parent */
+	if (!zdev->vfn || !zdev->rid_available)
+		return NULL;
+	/* Linux vfid starts at 0 vfn at 1 */
+	vfid = zdev->vfn - 1;
+	devfn = zdev->rid & ZPCI_RID_MASK_DEVFN;
+	/*
+	 * If the parent PF for the given VF is also configured in the
 	 * instance, it must be on the same zbus.
 	 * We can then identify the parent PF by checking what
 	 * devfn the VF would have if it belonged to that PF using the PF's
@@ -85,15 +102,26 @@ int zpci_iov_setup_virtfn(struct zpci_bus *zbus, struct pci_dev *virtfn, int vfn
 			if (!pdev)
 				continue;
 			cand_devfn = pci_iov_virtfn_devfn(pdev, vfid);
-			if (cand_devfn == virtfn->devfn) {
-				rc = zpci_iov_link_virtfn(pdev, virtfn, vfid);
-				/* balance pci_get_slot() */
-				pci_dev_put(pdev);
-				break;
-			}
+			if (cand_devfn == devfn)
+				return pdev;
 			/* balance pci_get_slot() */
 			pci_dev_put(pdev);
 		}
 	}
+	return NULL;
+}
+
+int zpci_iov_setup_virtfn(struct zpci_bus *zbus, struct pci_dev *virtfn, int vfn)
+{
+	struct zpci_dev *zdev = to_zpci(virtfn);
+	struct pci_dev *pdev_pf;
+	int rc = 0;
+
+	pdev_pf = zpci_iov_find_parent_pf(zbus, zdev);
+	if (pdev_pf) {
+		/* Linux' vfids start at 0 while zdev->vfn starts at 1 */
+		rc = zpci_iov_link_virtfn(pdev_pf, virtfn, zdev->vfn - 1);
+		pci_dev_put(pdev_pf);
+	}
 	return rc;
 }