docs: dev-tools: Add documentation for the device focused kselftests

Add documentation for the kselftests focused on testing devices and
point to it from the kselftest documentation. There are multiple tests
in this category so the aim of this page is to make it clear when to run
each test.

Signed-off-by: Nícolas F. R. A. Prado <nfraprado@collabora.com>
Reviewed-by: Shuah Khan <skhan@linuxfoundation.org>
Signed-off-by: Shuah Khan <skhan@linuxfoundation.org>

Documentation/dev-tools/kselftest.rst

@@ -31,6 +31,15 @@ kselftest runs as a userspace process.  Tests that can be written/run in
 userspace may wish to use the `Test Harness`_.  Tests that need to be
 run in kernel space may wish to use a `Test Module`_.
 
+Documentation on the tests
+==========================
+
+For documentation on the kselftests themselves, see:
+
+.. toctree::
+
+   testing-devices
+
 Running the selftests (hotplug tests are run in limited mode)
 =============================================================
 

Documentation/dev-tools/testing-devices.rst

@@ -0,0 +1,47 @@
+.. SPDX-License-Identifier: GPL-2.0
+.. Copyright (c) 2024 Collabora Ltd
+
+=============================
+Device testing with kselftest
+=============================
+
+
+There are a few different kselftests available for testing devices generically,
+with some overlap in coverage and different requirements. This document aims to
+give an overview of each one.
+
+Note: Paths in this document are relative to the kselftest folder
+(``tools/testing/selftests``).
+
+Device oriented kselftests:
+
+* Devicetree (``dt``)
+
+  * **Coverage**: Probe status for devices described in Devicetree
+  * **Requirements**: None
+
+* Error logs (``devices/error_logs``)
+
+  * **Coverage**: Error (or more critical) log messages presence coming from any
+    device
+  * **Requirements**: None
+
+* Discoverable bus (``devices/probe``)
+
+  * **Coverage**: Presence and probe status of USB or PCI devices that have been
+    described in the reference file
+  * **Requirements**: Manually describe the devices that should be tested in a
+    YAML reference file (see ``devices/probe/boards/google,spherion.yaml`` for
+    an example)
+
+* Exist (``devices/exist``)
+
+  * **Coverage**: Presence of all devices
+  * **Requirements**: Generate the reference (see ``devices/exist/README.rst``
+    for details) on a known-good kernel
+
+Therefore, the suggestion is to enable the error log and devicetree tests on all
+(DT-based) platforms, since they don't have any requirements. Then to greatly
+improve coverage, generate the reference for each platform and enable the exist
+test. The discoverable bus test can be used to verify the probe status of
+specific USB or PCI devices, but is probably not worth it for most cases.

Documentation/devicetree/bindings/mailbox/mtk,adsp-mbox.yaml

@@ -17,9 +17,15 @@ description: |
 
 properties:
   compatible:
-    enum:
-      - mediatek,mt8195-adsp-mbox
+    oneOf:
+      - enum:
           - mediatek,mt8186-adsp-mbox
+          - mediatek,mt8195-adsp-mbox
+      - items:
+          - enum:
+              - mediatek,mt8188-adsp-mbox
+          - const: mediatek,mt8186-adsp-mbox
+
 
   "#mbox-cells":
     const: 0

Documentation/devicetree/bindings/mailbox/qcom-ipcc.yaml

@@ -24,7 +24,9 @@ properties:
   compatible:
     items:
       - enum:
+          - qcom,qcs8300-ipcc
           - qcom,qdu1000-ipcc
+          - qcom,sa8255p-ipcc
           - qcom,sa8775p-ipcc
           - qcom,sc7280-ipcc
           - qcom,sc8280xp-ipcc

Makefile

@@ -1,8 +1,8 @@
 # SPDX-License-Identifier: GPL-2.0
 VERSION = 6
-PATCHLEVEL = 11
+PATCHLEVEL = 12
 SUBLEVEL = 0
-EXTRAVERSION =
+EXTRAVERSION = -rc1
 NAME = Baby Opossum Posse
 
 # *DOCUMENTATION*

arch/x86/include/asm/reboot.h

@@ -31,6 +31,8 @@ void cpu_emergency_register_virt_callback(cpu_emergency_virt_cb *callback);
 void cpu_emergency_unregister_virt_callback(cpu_emergency_virt_cb *callback);
 void cpu_emergency_disable_virtualization(void);
 #else
+static inline void cpu_emergency_register_virt_callback(cpu_emergency_virt_cb *callback) {}
+static inline void cpu_emergency_unregister_virt_callback(cpu_emergency_virt_cb *callback) {}
 static inline void cpu_emergency_disable_virtualization(void) {}
 #endif /* CONFIG_KVM_INTEL || CONFIG_KVM_AMD */
 

drivers/i2c/busses/Kconfig

@@ -782,6 +782,7 @@ config I2C_JZ4780
 config I2C_KEBA
 	tristate "KEBA I2C controller support"
 	depends on HAS_IOMEM
+	depends on KEBA_CP500 || COMPILE_TEST
 	select AUXILIARY_BUS
 	help
 	  This driver supports the I2C controller found in KEBA system FPGA

drivers/i2c/busses/i2c-designware-common.c

@@ -523,6 +523,7 @@ int i2c_dw_set_sda_hold(struct dw_i2c_dev *dev)
 
 void __i2c_dw_disable(struct dw_i2c_dev *dev)
 {
+	struct i2c_timings *t = &dev->timings;
 	unsigned int raw_intr_stats;
 	unsigned int enable;
 	int timeout = 100;
@@ -535,6 +536,19 @@ void __i2c_dw_disable(struct dw_i2c_dev *dev)
 
 	abort_needed = raw_intr_stats & DW_IC_INTR_MST_ON_HOLD;
 	if (abort_needed) {
+		if (!(enable & DW_IC_ENABLE_ENABLE)) {
+			regmap_write(dev->map, DW_IC_ENABLE, DW_IC_ENABLE_ENABLE);
+			/*
+			 * Wait 10 times the signaling period of the highest I2C
+			 * transfer supported by the driver (for 400KHz this is
+			 * 25us) to ensure the I2C ENABLE bit is already set
+			 * as described in the DesignWare I2C databook.
+			 */
+			fsleep(DIV_ROUND_CLOSEST_ULL(10 * MICRO, t->bus_freq_hz));
+			/* Set ENABLE bit before setting ABORT */
+			enable |= DW_IC_ENABLE_ENABLE;
+		}
+
 		regmap_write(dev->map, DW_IC_ENABLE, enable | DW_IC_ENABLE_ABORT);
 		ret = regmap_read_poll_timeout(dev->map, DW_IC_ENABLE, enable,
 					       !(enable & DW_IC_ENABLE_ABORT), 10,

drivers/i2c/busses/i2c-designware-core.h

@@ -108,6 +108,7 @@
 						 DW_IC_INTR_RX_UNDER | \
 						 DW_IC_INTR_RD_REQ)
 
+#define DW_IC_ENABLE_ENABLE			BIT(0)
 #define DW_IC_ENABLE_ABORT			BIT(1)
 
 #define DW_IC_STATUS_ACTIVITY			BIT(0)

drivers/i2c/busses/i2c-designware-master.c

@@ -271,6 +271,34 @@ static void i2c_dw_xfer_init(struct dw_i2c_dev *dev)
 	__i2c_dw_write_intr_mask(dev, DW_IC_INTR_MASTER_MASK);
 }
 
+/*
+ * This function waits for the controller to be idle before disabling I2C
+ * When the controller is not in the IDLE state, the MST_ACTIVITY bit
+ * (IC_STATUS[5]) is set.
+ *
+ * Values:
+ * 0x1 (ACTIVE): Controller not idle
+ * 0x0 (IDLE): Controller is idle
+ *
+ * The function is called after completing the current transfer.
+ *
+ * Returns:
+ * False when the controller is in the IDLE state.
+ * True when the controller is in the ACTIVE state.
+ */
+static bool i2c_dw_is_controller_active(struct dw_i2c_dev *dev)
+{
+	u32 status;
+
+	regmap_read(dev->map, DW_IC_STATUS, &status);
+	if (!(status & DW_IC_STATUS_MASTER_ACTIVITY))
+		return false;
+
+	return regmap_read_poll_timeout(dev->map, DW_IC_STATUS, status,
+				       !(status & DW_IC_STATUS_MASTER_ACTIVITY),
+				       1100, 20000) != 0;
+}
+
 static int i2c_dw_check_stopbit(struct dw_i2c_dev *dev)
 {
 	u32 val;
@@ -806,6 +834,16 @@ i2c_dw_xfer(struct i2c_adapter *adap, struct i2c_msg msgs[], int num)
 		goto done;
 	}
 
+	/*
+	 * This happens rarely (~1:500) and is hard to reproduce. Debug trace
+	 * showed that IC_STATUS had value of 0x23 when STOP_DET occurred,
+	 * if disable IC_ENABLE.ENABLE immediately that can result in
+	 * IC_RAW_INTR_STAT.MASTER_ON_HOLD holding SCL low. Check if
+	 * controller is still ACTIVE before disabling I2C.
+	 */
+	if (i2c_dw_is_controller_active(dev))
+		dev_err(dev->dev, "controller active\n");
+
 	/*
 	 * We must disable the adapter before returning and signaling the end
 	 * of the current transfer. Otherwise the hardware might continue

drivers/i2c/busses/i2c-synquacer.c

@@ -550,11 +550,12 @@ static int synquacer_i2c_probe(struct platform_device *pdev)
 	device_property_read_u32(&pdev->dev, "socionext,pclk-rate",
 				 &i2c->pclkrate);
 
-	pclk = devm_clk_get_enabled(&pdev->dev, "pclk");
+	pclk = devm_clk_get_optional_enabled(&pdev->dev, "pclk");
 	if (IS_ERR(pclk))
 		return dev_err_probe(&pdev->dev, PTR_ERR(pclk),
 				     "failed to get and enable clock\n");
 
+	if (pclk)
 		i2c->pclkrate = clk_get_rate(pclk);
 
 	if (i2c->pclkrate < SYNQUACER_I2C_MIN_CLK_RATE ||

drivers/i2c/busses/i2c-xiic.c

@@ -1337,8 +1337,8 @@ static int xiic_i2c_probe(struct platform_device *pdev)
 	return 0;
 
 err_pm_disable:
-	pm_runtime_set_suspended(&pdev->dev);
 	pm_runtime_disable(&pdev->dev);
+	pm_runtime_set_suspended(&pdev->dev);
 
 	return ret;
 }

drivers/mailbox/Kconfig

@@ -25,6 +25,7 @@ config ARM_MHU_V2
 
 config ARM_MHU_V3
 	tristate "ARM MHUv3 Mailbox"
+	depends on ARM64 || COMPILE_TEST
 	depends on HAS_IOMEM || COMPILE_TEST
 	depends on OF
 	help
@@ -73,7 +74,7 @@ config ARMADA_37XX_RWTM_MBOX
 
 config OMAP2PLUS_MBOX
 	tristate "OMAP2+ Mailbox framework support"
-	depends on ARCH_OMAP2PLUS || ARCH_K3
+	depends on ARCH_OMAP2PLUS || ARCH_K3 || COMPILE_TEST
 	help
 	  Mailbox implementation for OMAP family chips with hardware for
 	  interprocessor communication involving DSP, IVA1.0 and IVA2 in

drivers/mailbox/bcm2835-mailbox.c

@@ -145,7 +145,8 @@ static int bcm2835_mbox_probe(struct platform_device *pdev)
 	spin_lock_init(&mbox->lock);
 
 	ret = devm_request_irq(dev, irq_of_parse_and_map(dev->of_node, 0),
-			       bcm2835_mbox_irq, 0, dev_name(dev), mbox);
+			       bcm2835_mbox_irq, IRQF_NO_SUSPEND, dev_name(dev),
+			       mbox);
 	if (ret) {
 		dev_err(dev, "Failed to register a mailbox IRQ handler: %d\n",
 			ret);

drivers/mailbox/imx-mailbox.c

@@ -30,7 +30,7 @@
 #define IMX_MU_SCU_CHANS	6
 /* TX0/RX0 */
 #define IMX_MU_S4_CHANS		2
-#define IMX_MU_CHAN_NAME_SIZE	20
+#define IMX_MU_CHAN_NAME_SIZE	32
 
 #define IMX_MU_V2_PAR_OFF	0x4
 #define IMX_MU_V2_TR_MASK	GENMASK(7, 0)
@@ -782,7 +782,7 @@ static int imx_mu_init_generic(struct imx_mu_priv *priv)
 		cp->chan = &priv->mbox_chans[i];
 		priv->mbox_chans[i].con_priv = cp;
 		snprintf(cp->irq_desc, sizeof(cp->irq_desc),
-			 "imx_mu_chan[%i-%i]", cp->type, cp->idx);
+			 "%s[%i-%i]", dev_name(priv->dev), cp->type, cp->idx);
 	}
 
 	priv->mbox.num_chans = IMX_MU_CHANS;
@@ -819,7 +819,7 @@ static int imx_mu_init_specific(struct imx_mu_priv *priv)
 		cp->chan = &priv->mbox_chans[i];
 		priv->mbox_chans[i].con_priv = cp;
 		snprintf(cp->irq_desc, sizeof(cp->irq_desc),
-			 "imx_mu_chan[%i-%i]", cp->type, cp->idx);
+			 "%s[%i-%i]", dev_name(priv->dev), cp->type, cp->idx);
 	}
 
 	priv->mbox.num_chans = num_chans;

drivers/mailbox/mailbox.c

@@ -450,31 +450,21 @@ struct mbox_chan *mbox_request_channel_byname(struct mbox_client *cl,
 					      const char *name)
 {
 	struct device_node *np = cl->dev->of_node;
-	struct property *prop;
-	const char *mbox_name;
-	int index = 0;
+	int index;
 
 	if (!np) {
 		dev_err(cl->dev, "%s() currently only supports DT\n", __func__);
 		return ERR_PTR(-EINVAL);
 	}
 
-	if (!of_get_property(np, "mbox-names", NULL)) {
-		dev_err(cl->dev,
-			"%s() requires an \"mbox-names\" property\n", __func__);
-		return ERR_PTR(-EINVAL);
-	}
-
-	of_property_for_each_string(np, "mbox-names", prop, mbox_name) {
-		if (!strncmp(name, mbox_name, strlen(name)))
-			return mbox_request_channel(cl, index);
-		index++;
-	}
-
+	index = of_property_match_string(np, "mbox-names", name);
+	if (index < 0) {
 		dev_err(cl->dev, "%s() could not locate channel named \"%s\"\n",
 			__func__, name);
 		return ERR_PTR(-EINVAL);
 	}
+	return mbox_request_channel(cl, index);
+}
 EXPORT_SYMBOL_GPL(mbox_request_channel_byname);
 
 /**

drivers/mailbox/omap-mailbox.c

@@ -603,7 +603,7 @@ static struct platform_driver omap_mbox_driver = {
 	.driver	= {
 		.name = "omap-mailbox",
 		.pm = &omap_mbox_pm_ops,
-		.of_match_table = of_match_ptr(omap_mailbox_of_match),
+		.of_match_table = omap_mailbox_of_match,
 	},
 };
 module_platform_driver(omap_mbox_driver);

drivers/mailbox/rockchip-mailbox.c

@@ -159,7 +159,7 @@ static const struct of_device_id rockchip_mbox_of_match[] = {
 	{ .compatible = "rockchip,rk3368-mailbox", .data = &rk3368_drv_data},
 	{ },
 };
-MODULE_DEVICE_TABLE(of, rockchp_mbox_of_match);
+MODULE_DEVICE_TABLE(of, rockchip_mbox_of_match);
 
 static int rockchip_mbox_probe(struct platform_device *pdev)
 {

drivers/mailbox/sprd-mailbox.c

@@ -62,7 +62,6 @@ struct sprd_mbox_priv {
 	void __iomem		*outbox_base;
 	/*  Base register address for supplementary outbox */
 	void __iomem		*supp_base;
-	struct clk		*clk;
 	u32			outbox_fifo_depth;
 
 	struct mutex		lock;
@@ -291,19 +290,13 @@ static const struct mbox_chan_ops sprd_mbox_ops = {
 	.shutdown	= sprd_mbox_shutdown,
 };
 
-static void sprd_mbox_disable(void *data)
-{
-	struct sprd_mbox_priv *priv = data;
-
-	clk_disable_unprepare(priv->clk);
-}
-
 static int sprd_mbox_probe(struct platform_device *pdev)
 {
 	struct device *dev = &pdev->dev;
 	struct sprd_mbox_priv *priv;
 	int ret, inbox_irq, outbox_irq, supp_irq;
 	unsigned long id, supp;
+	struct clk *clk;
 
 	priv = devm_kzalloc(dev, sizeof(*priv), GFP_KERNEL);
 	if (!priv)
@@ -331,20 +324,10 @@ static int sprd_mbox_probe(struct platform_device *pdev)
 	if (IS_ERR(priv->outbox_base))
 		return PTR_ERR(priv->outbox_base);
 
-	priv->clk = devm_clk_get(dev, "enable");
-	if (IS_ERR(priv->clk)) {
+	clk = devm_clk_get_enabled(dev, "enable");
+	if (IS_ERR(clk)) {
 		dev_err(dev, "failed to get mailbox clock\n");
-		return PTR_ERR(priv->clk);
-	}
-
-	ret = clk_prepare_enable(priv->clk);
-	if (ret)
-		return ret;
-
-	ret = devm_add_action_or_reset(dev, sprd_mbox_disable, priv);
-	if (ret) {
-		dev_err(dev, "failed to add mailbox disable action\n");
-		return ret;
+		return PTR_ERR(clk);
 	}
 
 	inbox_irq = platform_get_irq_byname(pdev, "inbox");

drivers/message/fusion/mptctl.c

@@ -1609,7 +1609,7 @@ mptctl_eventreport (MPT_ADAPTER *ioc, unsigned long arg)
 	maxEvents = numBytes/sizeof(MPT_IOCTL_EVENTS);
 
 
-	max = MPTCTL_EVENT_LOG_SIZE < maxEvents ? MPTCTL_EVENT_LOG_SIZE : maxEvents;
+	max = min(maxEvents, MPTCTL_EVENT_LOG_SIZE);
 
 	/* If fewer than 1 event is requested, there must have
 	 * been some type of error.

drivers/remoteproc/Kconfig

@@ -330,8 +330,7 @@ config STM32_RPROC
 config TI_K3_DSP_REMOTEPROC
 	tristate "TI K3 DSP remoteproc support"
 	depends on ARCH_K3
-	select MAILBOX
-	select OMAP2PLUS_MBOX
+	depends on OMAP2PLUS_MBOX
 	help
 	  Say m here to support TI's C66x and C71x DSP remote processor
 	  subsystems on various TI K3 family of SoCs through the remote
@@ -356,8 +355,7 @@ config TI_K3_M4_REMOTEPROC
 config TI_K3_R5_REMOTEPROC
 	tristate "TI K3 R5 remoteproc support"
 	depends on ARCH_K3
-	select MAILBOX
-	select OMAP2PLUS_MBOX
+	depends on OMAP2PLUS_MBOX
 	help
 	  Say m here to support TI's R5F remote processor subsystems
 	  on various TI K3 family of SoCs through the remote processor

drivers/scsi/cxgbi/libcxgbi.h

@@ -485,7 +485,6 @@ struct cxgbi_device {
 	unsigned char nmtus;
 	unsigned char nports;
 	struct pci_dev *pdev;
-	struct dentry *debugfs_root;
 	struct iscsi_transport *itp;
 	struct module *owner;
 
@@ -499,7 +498,6 @@ struct cxgbi_device {
 	unsigned int rxq_idx_cntr;
 	struct cxgbi_ports_map pmap;
 
-	void (*dev_ddp_cleanup)(struct cxgbi_device *);
 	struct cxgbi_ppm* (*cdev2ppm)(struct cxgbi_device *);
 	int (*csk_ddp_set_map)(struct cxgbi_ppm *, struct cxgbi_sock *,
 			       struct cxgbi_task_tag_info *);
@@ -512,7 +510,6 @@ struct cxgbi_device {
 				   unsigned int, int);
 
 	void (*csk_release_offload_resources)(struct cxgbi_sock *);
-	int (*csk_rx_pdu_ready)(struct cxgbi_sock *, struct sk_buff *);
 	u32 (*csk_send_rx_credits)(struct cxgbi_sock *, u32);
 	int (*csk_push_tx_frames)(struct cxgbi_sock *, int);
 	void (*csk_send_abort_req)(struct cxgbi_sock *);

drivers/scsi/ibmvscsi/ibmvfc.c

@@ -37,6 +37,7 @@ static unsigned int default_timeout = IBMVFC_DEFAULT_TIMEOUT;
 static u64 max_lun = IBMVFC_MAX_LUN;
 static unsigned int max_targets = IBMVFC_MAX_TARGETS;
 static unsigned int max_requests = IBMVFC_MAX_REQUESTS_DEFAULT;
+static u16 max_sectors = IBMVFC_MAX_SECTORS;
 static u16 scsi_qdepth = IBMVFC_SCSI_QDEPTH;
 static unsigned int disc_threads = IBMVFC_MAX_DISC_THREADS;
 static unsigned int ibmvfc_debug = IBMVFC_DEBUG;
@@ -83,6 +84,9 @@ MODULE_PARM_DESC(default_timeout,
 module_param_named(max_requests, max_requests, uint, S_IRUGO);
 MODULE_PARM_DESC(max_requests, "Maximum requests for this adapter. "
 		 "[Default=" __stringify(IBMVFC_MAX_REQUESTS_DEFAULT) "]");
+module_param_named(max_sectors, max_sectors, ushort, S_IRUGO);
+MODULE_PARM_DESC(max_sectors, "Maximum sectors for this adapter. "
+		 "[Default=" __stringify(IBMVFC_MAX_SECTORS) "]");
 module_param_named(scsi_qdepth, scsi_qdepth, ushort, S_IRUGO);
 MODULE_PARM_DESC(scsi_qdepth, "Maximum scsi command depth per adapter queue. "
 		 "[Default=" __stringify(IBMVFC_SCSI_QDEPTH) "]");
@@ -1494,7 +1498,7 @@ static void ibmvfc_set_login_info(struct ibmvfc_host *vhost)
 	memset(login_info, 0, sizeof(*login_info));
 
 	login_info->ostype = cpu_to_be32(IBMVFC_OS_LINUX);
-	login_info->max_dma_len = cpu_to_be64(IBMVFC_MAX_SECTORS << 9);
+	login_info->max_dma_len = cpu_to_be64(max_sectors << 9);
 	login_info->max_payload = cpu_to_be32(sizeof(struct ibmvfc_fcp_cmd_iu));
 	login_info->max_response = cpu_to_be32(sizeof(struct ibmvfc_fcp_rsp));
 	login_info->partition_num = cpu_to_be32(vhost->partition_number);
@@ -5230,7 +5234,7 @@ static void ibmvfc_npiv_login_done(struct ibmvfc_event *evt)
 	}
 
 	vhost->logged_in = 1;
-	npiv_max_sectors = min((uint)(be64_to_cpu(rsp->max_dma_len) >> 9), IBMVFC_MAX_SECTORS);
+	npiv_max_sectors = min((uint)(be64_to_cpu(rsp->max_dma_len) >> 9), max_sectors);
 	dev_info(vhost->dev, "Host partition: %s, device: %s %s %s max sectors %u\n",
 		 rsp->partition_name, rsp->device_name, rsp->port_loc_code,
 		 rsp->drc_name, npiv_max_sectors);
@@ -6329,7 +6333,7 @@ static int ibmvfc_probe(struct vio_dev *vdev, const struct vio_device_id *id)
 	shost->can_queue = scsi_qdepth;
 	shost->max_lun = max_lun;
 	shost->max_id = max_targets;
-	shost->max_sectors = IBMVFC_MAX_SECTORS;
+	shost->max_sectors = max_sectors;
 	shost->max_cmd_len = IBMVFC_MAX_CDB_LEN;
 	shost->unique_id = shost->host_no;
 	shost->nr_hw_queues = mq_enabled ? min(max_scsi_queues, nr_scsi_hw_queues) : 1;
@@ -6556,6 +6560,7 @@ static struct fc_function_template ibmvfc_transport_functions = {
  **/
 static int __init ibmvfc_module_init(void)
 {
+	int min_max_sectors = PAGE_SIZE >> 9;
 	int rc;
 
 	if (!firmware_has_feature(FW_FEATURE_VIO))
@@ -6564,6 +6569,16 @@ static int __init ibmvfc_module_init(void)
 	printk(KERN_INFO IBMVFC_NAME": IBM Virtual Fibre Channel Driver version: %s %s\n",
 	       IBMVFC_DRIVER_VERSION, IBMVFC_DRIVER_DATE);
 
+	/*
+	 * Range check the max_sectors module parameter. The upper bounds is
+	 * implicity checked since the parameter is a ushort.
+	 */
+	if (max_sectors < min_max_sectors) {
+		printk(KERN_ERR IBMVFC_NAME ": max_sectors must be at least %d.\n",
+			min_max_sectors);
+		max_sectors = min_max_sectors;
+	}
+
 	ibmvfc_transport_template = fc_attach_transport(&ibmvfc_transport_functions);
 	if (!ibmvfc_transport_template)
 		return -ENOMEM;

drivers/scsi/ibmvscsi/ibmvfc.h

@@ -32,7 +32,7 @@
 #define IBMVFC_DEBUG			0
 #define IBMVFC_MAX_TARGETS		1024
 #define IBMVFC_MAX_LUN			0xffffffff
-#define IBMVFC_MAX_SECTORS		0xffffu
+#define IBMVFC_MAX_SECTORS		2048
 #define IBMVFC_MAX_DISC_THREADS	4
 #define IBMVFC_TGT_MEMPOOL_SZ		64
 #define IBMVFC_MAX_CMDS_PER_LUN	64

drivers/scsi/lpfc/lpfc_bsg.c

@@ -3208,6 +3208,9 @@ lpfc_bsg_diag_loopback_run(struct bsg_job *job)
 	cmdiocbq->num_bdes = num_bde;
 	cmdiocbq->cmd_flag |= LPFC_IO_LIBDFC;
 	cmdiocbq->cmd_flag |= LPFC_IO_LOOPBACK;
+	if (phba->cfg_vmid_app_header)
+		cmdiocbq->cmd_flag |= LPFC_IO_VMID;
+
 	cmdiocbq->vport = phba->pport;
 	cmdiocbq->cmd_cmpl = NULL;
 	cmdiocbq->bpl_dmabuf = txbmp;

drivers/scsi/lpfc/lpfc_ct.c

@@ -1572,8 +1572,8 @@ lpfc_cmpl_ct_cmd_gft_id(struct lpfc_hba *phba, struct lpfc_iocbq *cmdiocb,
 			}
 		}
 	} else
-		lpfc_printf_vlog(vport, KERN_ERR, LOG_TRACE_EVENT,
-				 "3065 GFT_ID failed x%08x\n", ulp_status);
+		lpfc_vlog_msg(vport, KERN_WARNING, LOG_DISCOVERY,
+			      "3065 GFT_ID status x%08x\n", ulp_status);
 
 out:
 	lpfc_ct_free_iocb(phba, cmdiocb);
@@ -1647,6 +1647,18 @@ lpfc_cmpl_ct(struct lpfc_hba *phba, struct lpfc_iocbq *cmdiocb,
 	}
 
 out:
+	/* If the caller wanted a synchronous DA_ID completion, signal the
+	 * wait obj and clear flag to reset the vport.
+	 */
+	if (ndlp->save_flags & NLP_WAIT_FOR_DA_ID) {
+		if (ndlp->da_id_waitq)
+			wake_up(ndlp->da_id_waitq);
+	}
+
+	spin_lock_irq(&ndlp->lock);
+	ndlp->save_flags &= ~NLP_WAIT_FOR_DA_ID;
+	spin_unlock_irq(&ndlp->lock);
+
 	lpfc_ct_free_iocb(phba, cmdiocb);
 	lpfc_nlp_put(ndlp);
 	return;
@@ -2246,7 +2258,7 @@ lpfc_cmpl_ct_disc_fdmi(struct lpfc_hba *phba, struct lpfc_iocbq *cmdiocb,
 		}
 
 		lpfc_printf_vlog(vport, KERN_INFO, LOG_DISCOVERY,
-				 "0229 FDMI cmd %04x failed, latt = %d "
+				 "0229 FDMI cmd %04x latt = %d "
 				 "ulp_status: x%x, rid x%x\n",
 				 be16_to_cpu(fdmi_cmd), latt, ulp_status,
 				 ulp_word4);
@@ -2263,9 +2275,9 @@ lpfc_cmpl_ct_disc_fdmi(struct lpfc_hba *phba, struct lpfc_iocbq *cmdiocb,
 	/* Check for a CT LS_RJT response */
 	cmd =  be16_to_cpu(fdmi_cmd);
 	if (be16_to_cpu(fdmi_rsp) == SLI_CT_RESPONSE_FS_RJT) {
-		/* FDMI rsp failed */
+		/* Log FDMI reject */
 		lpfc_printf_vlog(vport, KERN_INFO, LOG_DISCOVERY | LOG_ELS,
-				 "0220 FDMI cmd failed FS_RJT Data: x%x", cmd);
+				 "0220 FDMI cmd FS_RJT Data: x%x", cmd);
 
 		/* Should we fallback to FDMI-2 / FDMI-1 ? */
 		switch (cmd) {

drivers/scsi/lpfc/lpfc_disc.h

@@ -90,6 +90,8 @@ enum lpfc_nlp_save_flags {
 	NLP_IN_RECOV_POST_DEV_LOSS	= 0x1,
 	/* wait for outstanding LOGO to cmpl */
 	NLP_WAIT_FOR_LOGO		= 0x2,
+	/* wait for outstanding DA_ID to finish */
+	NLP_WAIT_FOR_DA_ID              = 0x4
 };
 
 struct lpfc_nodelist {
@@ -159,7 +161,12 @@ struct lpfc_nodelist {
 	uint32_t nvme_fb_size; /* NVME target's supported byte cnt */
 #define NVME_FB_BIT_SHIFT 9    /* PRLI Rsp first burst in 512B units. */
 	uint32_t nlp_defer_did;
+
+	/* These wait objects are NPIV specific.  These IOs must complete
+	 * synchronously.
+	 */
 	wait_queue_head_t *logo_waitq;
+	wait_queue_head_t *da_id_waitq;
 };
 
 struct lpfc_node_rrq {

drivers/scsi/lpfc/lpfc_els.c

@@ -979,7 +979,7 @@ lpfc_cmpl_els_flogi(struct lpfc_hba *phba, struct lpfc_iocbq *cmdiocb,
 				phba->fcoe_cvl_eventtag_attn =
 					phba->fcoe_cvl_eventtag;
 			lpfc_printf_log(phba, KERN_WARNING, LOG_FIP | LOG_ELS,
-					"2611 FLOGI failed on FCF (x%x), "
+					"2611 FLOGI FCF (x%x), "
 					"status:x%x/x%x, tmo:x%x, perform "
 					"roundrobin FCF failover\n",
 					phba->fcf.current_rec.fcf_indx,
@@ -997,8 +997,8 @@ lpfc_cmpl_els_flogi(struct lpfc_hba *phba, struct lpfc_iocbq *cmdiocb,
 		if (!(ulp_status == IOSTAT_LOCAL_REJECT &&
 		      ((ulp_word4 & IOERR_PARAM_MASK) ==
 					IOERR_LOOP_OPEN_FAILURE)))
-			lpfc_printf_vlog(vport, KERN_ERR, LOG_TRACE_EVENT,
-					 "2858 FLOGI failure Status:x%x/x%x TMO"
+			lpfc_vlog_msg(vport, KERN_WARNING, LOG_ELS,
+				      "2858 FLOGI Status:x%x/x%x TMO"
 				      ":x%x Data x%lx x%x\n",
 				      ulp_status, ulp_word4, tmo,
 				      phba->hba_flag, phba->fcf.fcf_flag);
@@ -1023,7 +1023,7 @@ lpfc_cmpl_els_flogi(struct lpfc_hba *phba, struct lpfc_iocbq *cmdiocb,
 			lpfc_nlp_put(ndlp);
 
 		lpfc_printf_vlog(vport, KERN_WARNING, LOG_ELS,
-				 "0150 FLOGI failure Status:x%x/x%x "
+				 "0150 FLOGI Status:x%x/x%x "
 				 "xri x%x TMO:x%x refcnt %d\n",
 				 ulp_status, ulp_word4, cmdiocb->sli4_xritag,
 				 tmo, kref_read(&ndlp->kref));
@@ -1032,9 +1032,9 @@ lpfc_cmpl_els_flogi(struct lpfc_hba *phba, struct lpfc_iocbq *cmdiocb,
 		if (!(ulp_status == IOSTAT_LOCAL_REJECT &&
 		      ((ulp_word4 & IOERR_PARAM_MASK) ==
 					IOERR_LOOP_OPEN_FAILURE))) {
-			/* FLOGI failure */
-			lpfc_printf_vlog(vport, KERN_ERR, LOG_TRACE_EVENT,
-					 "0100 FLOGI failure Status:x%x/x%x "
+			/* Warn FLOGI status */
+			lpfc_vlog_msg(vport, KERN_WARNING, LOG_ELS,
+				      "0100 FLOGI Status:x%x/x%x "
 				      "TMO:x%x\n",
 				      ulp_status, ulp_word4, tmo);
 			goto flogifail;
@@ -1964,13 +1964,13 @@ lpfc_cmpl_els_rrq(struct lpfc_hba *phba, struct lpfc_iocbq *cmdiocb,
 
 	if (ulp_status) {
 		/* Check for retry */
-		/* RRQ failed Don't print the vport to vport rjts */
+		/* Warn RRQ status Don't print the vport to vport rjts */
 		if (ulp_status != IOSTAT_LS_RJT ||
 		    (((ulp_word4) >> 16 != LSRJT_INVALID_CMD) &&
 		     ((ulp_word4) >> 16 != LSRJT_UNABLE_TPC)) ||
 		    (phba)->pport->cfg_log_verbose & LOG_ELS)
-			lpfc_printf_vlog(vport, KERN_ERR, LOG_TRACE_EVENT,
-					 "2881 RRQ failure DID:%06X Status:"
+			lpfc_vlog_msg(vport, KERN_WARNING, LOG_ELS,
+				      "2881 RRQ DID:%06X Status:"
 				      "x%x/x%x\n",
 				      ndlp->nlp_DID, ulp_status,
 				      ulp_word4);
@@ -2077,13 +2077,13 @@ lpfc_cmpl_els_plogi(struct lpfc_hba *phba, struct lpfc_iocbq *cmdiocb,
 			}
 			goto out;
 		}
-		/* PLOGI failed Don't print the vport to vport rjts */
+		/* Warn PLOGI status Don't print the vport to vport rjts */
 		if (ulp_status != IOSTAT_LS_RJT ||
 		    (((ulp_word4) >> 16 != LSRJT_INVALID_CMD) &&
 		     ((ulp_word4) >> 16 != LSRJT_UNABLE_TPC)) ||
 		    (phba)->pport->cfg_log_verbose & LOG_ELS)
-			lpfc_printf_vlog(vport, KERN_ERR, LOG_TRACE_EVENT,
-					 "2753 PLOGI failure DID:%06X "
+			lpfc_vlog_msg(vport, KERN_WARNING, LOG_ELS,
+				      "2753 PLOGI DID:%06X "
 				      "Status:x%x/x%x\n",
 				      ndlp->nlp_DID, ulp_status,
 				      ulp_word4);
@@ -2323,7 +2323,6 @@ lpfc_cmpl_els_prli(struct lpfc_hba *phba, struct lpfc_iocbq *cmdiocb,
 	struct lpfc_vport *vport = cmdiocb->vport;
 	struct lpfc_nodelist *ndlp;
 	char *mode;
-	u32 loglevel;
 	u32 ulp_status;
 	u32 ulp_word4;
 	bool release_node = false;
@@ -2372,17 +2371,14 @@ lpfc_cmpl_els_prli(struct lpfc_hba *phba, struct lpfc_iocbq *cmdiocb,
 		 * could be expected.
 		 */
 		if (test_bit(FC_FABRIC, &vport->fc_flag) ||
-		    vport->cfg_enable_fc4_type != LPFC_ENABLE_BOTH) {
-			mode = KERN_ERR;
-			loglevel =  LOG_TRACE_EVENT;
-		} else {
+		    vport->cfg_enable_fc4_type != LPFC_ENABLE_BOTH)
+			mode = KERN_WARNING;
+		else
 			mode = KERN_INFO;
-			loglevel =  LOG_ELS;
-		}
 
-		/* PRLI failed */
-		lpfc_printf_vlog(vport, mode, loglevel,
-				 "2754 PRLI failure DID:%06X Status:x%x/x%x, "
+		/* Warn PRLI status */
+		lpfc_printf_vlog(vport, mode, LOG_ELS,
+				 "2754 PRLI DID:%06X Status:x%x/x%x, "
 				 "data: x%x x%x x%x\n",
 				 ndlp->nlp_DID, ulp_status,
 				 ulp_word4, ndlp->nlp_state,
@@ -2854,9 +2850,9 @@ lpfc_cmpl_els_adisc(struct lpfc_hba *phba, struct lpfc_iocbq *cmdiocb,
 			}
 			goto out;
 		}
-		/* ADISC failed */
-		lpfc_printf_vlog(vport, KERN_ERR, LOG_TRACE_EVENT,
-				 "2755 ADISC failure DID:%06X Status:x%x/x%x\n",
+		/* Warn ADISC status */
+		lpfc_vlog_msg(vport, KERN_WARNING, LOG_ELS,
+			      "2755 ADISC DID:%06X Status:x%x/x%x\n",
 			      ndlp->nlp_DID, ulp_status,
 			      ulp_word4);
 		lpfc_disc_state_machine(vport, ndlp, cmdiocb,
@@ -3045,9 +3041,9 @@ lpfc_cmpl_els_logo(struct lpfc_hba *phba, struct lpfc_iocbq *cmdiocb,
 	 * discovery.  The PLOGI will retry.
 	 */
 	if (ulp_status) {
-		/* LOGO failed */
-		lpfc_printf_vlog(vport, KERN_ERR, LOG_TRACE_EVENT,
-				 "2756 LOGO failure, No Retry DID:%06X "
+		/* Warn LOGO status */
+		lpfc_vlog_msg(vport, KERN_WARNING, LOG_ELS,
+			      "2756 LOGO, No Retry DID:%06X "
 			      "Status:x%x/x%x\n",
 			      ndlp->nlp_DID, ulp_status,
 			      ulp_word4);
@@ -4837,9 +4833,8 @@ lpfc_els_retry(struct lpfc_hba *phba, struct lpfc_iocbq *cmdiocb,
 			if ((phba->sli3_options & LPFC_SLI3_NPIV_ENABLED) &&
 			  (cmd == ELS_CMD_FDISC) &&
 			  (stat.un.b.lsRjtRsnCodeExp == LSEXP_OUT_OF_RESOURCE)){
-				lpfc_printf_vlog(vport, KERN_ERR,
-						 LOG_TRACE_EVENT,
-						 "0125 FDISC Failed (x%x). "
+				lpfc_vlog_msg(vport, KERN_WARNING, LOG_ELS,
+					      "0125 FDISC (x%x). "
 					      "Fabric out of resources\n",
 					      stat.un.lsRjtError);
 				lpfc_vport_set_state(vport,
@@ -4877,9 +4872,8 @@ lpfc_els_retry(struct lpfc_hba *phba, struct lpfc_iocbq *cmdiocb,
 						LSEXP_NOTHING_MORE) {
 				vport->fc_sparam.cmn.bbRcvSizeMsb &= 0xf;
 				retry = 1;
-				lpfc_printf_vlog(vport, KERN_ERR,
-						 LOG_TRACE_EVENT,
-						 "0820 FLOGI Failed (x%x). "
+				lpfc_vlog_msg(vport, KERN_WARNING, LOG_ELS,
+					      "0820 FLOGI (x%x). "
 					      "BBCredit Not Supported\n",
 					      stat.un.lsRjtError);
 			}
@@ -4891,9 +4885,8 @@ lpfc_els_retry(struct lpfc_hba *phba, struct lpfc_iocbq *cmdiocb,
 			  ((stat.un.b.lsRjtRsnCodeExp == LSEXP_INVALID_PNAME) ||
 			  (stat.un.b.lsRjtRsnCodeExp == LSEXP_INVALID_NPORT_ID))
 			  ) {
-				lpfc_printf_vlog(vport, KERN_ERR,
-						 LOG_TRACE_EVENT,
-						 "0122 FDISC Failed (x%x). "
+				lpfc_vlog_msg(vport, KERN_WARNING, LOG_ELS,
+					      "0122 FDISC (x%x). "
 					      "Fabric Detected Bad WWN\n",
 					      stat.un.lsRjtError);
 				lpfc_vport_set_state(vport,
@@ -5355,8 +5348,8 @@ lpfc_cmpl_els_rsp(struct lpfc_hba *phba, struct lpfc_iocbq *cmdiocb,
 	u32 ulp_status, ulp_word4, tmo, did, iotag;
 
 	if (!vport) {
-		lpfc_printf_log(phba, KERN_ERR, LOG_TRACE_EVENT,
-				"3177 ELS response failed\n");
+		lpfc_printf_log(phba, KERN_WARNING, LOG_ELS,
+				"3177 null vport in ELS rsp\n");
 		goto out;
 	}
 	if (cmdiocb->context_un.mbox)
@@ -9658,11 +9651,12 @@ lpfc_els_flush_cmd(struct lpfc_vport *vport)
 		if (piocb->cmd_flag & LPFC_DRIVER_ABORTED && !mbx_tmo_err)
 			continue;
 
-		/* On the ELS ring we can have ELS_REQUESTs or
-		 * GEN_REQUESTs waiting for a response.
+		/* On the ELS ring we can have ELS_REQUESTs, ELS_RSPs,
+		 * or GEN_REQUESTs waiting for a CQE response.
 		 */
 		ulp_command = get_job_cmnd(phba, piocb);
-		if (ulp_command == CMD_ELS_REQUEST64_CR) {
+		if (ulp_command == CMD_ELS_REQUEST64_WQE ||
+		    ulp_command == CMD_XMIT_ELS_RSP64_WQE) {
 			list_add_tail(&piocb->dlist, &abort_list);
 
 			/* If the link is down when flushing ELS commands
@@ -11327,9 +11321,9 @@ lpfc_cmpl_els_fdisc(struct lpfc_hba *phba, struct lpfc_iocbq *cmdiocb,
 		/* Check for retry */
 		if (lpfc_els_retry(phba, cmdiocb, rspiocb))
 			goto out;
-		/* FDISC failed */
-		lpfc_printf_vlog(vport, KERN_ERR, LOG_TRACE_EVENT,
-				 "0126 FDISC failed. (x%x/x%x)\n",
+		/* Warn FDISC status */
+		lpfc_vlog_msg(vport, KERN_WARNING, LOG_ELS,
+			      "0126 FDISC cmpl status: x%x/x%x)\n",
 			      ulp_status, ulp_word4);
 		goto fdisc_failed;
 	}

drivers/scsi/lpfc/lpfc_hbadisc.c

@@ -527,6 +527,9 @@ lpfc_dev_loss_tmo_handler(struct lpfc_nodelist *ndlp)
 		 * the following lpfc_nlp_put is necessary after fabric node is
 		 * recovered.
 		 */
+		spin_lock_irqsave(&ndlp->lock, iflags);
+		ndlp->nlp_flag &= ~NLP_IN_DEV_LOSS;
+		spin_unlock_irqrestore(&ndlp->lock, iflags);
 		if (recovering) {
 			lpfc_printf_vlog(vport, KERN_INFO,
 					 LOG_DISCOVERY | LOG_NODE,
@@ -539,6 +542,7 @@ lpfc_dev_loss_tmo_handler(struct lpfc_nodelist *ndlp)
 			spin_lock_irqsave(&ndlp->lock, iflags);
 			ndlp->save_flags |= NLP_IN_RECOV_POST_DEV_LOSS;
 			spin_unlock_irqrestore(&ndlp->lock, iflags);
+			return fcf_inuse;
 		} else if (ndlp->nlp_state == NLP_STE_UNMAPPED_NODE) {
 			/* Fabric node fully recovered before this dev_loss_tmo
 			 * queue work is processed.  Thus, ignore the
@@ -552,15 +556,9 @@ lpfc_dev_loss_tmo_handler(struct lpfc_nodelist *ndlp)
 					 ndlp->nlp_DID, kref_read(&ndlp->kref),
 					 ndlp, ndlp->nlp_flag,
 					 vport->port_state);
-			spin_lock_irqsave(&ndlp->lock, iflags);
-			ndlp->nlp_flag &= ~NLP_IN_DEV_LOSS;
-			spin_unlock_irqrestore(&ndlp->lock, iflags);
 			return fcf_inuse;
 		}
 
-		spin_lock_irqsave(&ndlp->lock, iflags);
-		ndlp->nlp_flag &= ~NLP_IN_DEV_LOSS;
-		spin_unlock_irqrestore(&ndlp->lock, iflags);
 		lpfc_nlp_put(ndlp);
 		return fcf_inuse;
 	}

drivers/scsi/lpfc/lpfc_hw.h

@@ -561,6 +561,27 @@ struct fc_vft_header {
 
 #include <uapi/scsi/fc/fc_els.h>
 
+/*
+ * Application Header
+ */
+struct fc_app_header {
+	uint32_t dst_app_id;
+	uint32_t src_app_id;
+#define LOOPBACK_SRC_APPID	0x4321
+	uint32_t word2;
+	uint32_t word3;
+};
+
+/*
+ * dfctl optional header definition
+ */
+enum lpfc_fc_dfctl {
+	LPFC_FC_NO_DEVICE_HEADER,
+	LPFC_FC_16B_DEVICE_HEADER,
+	LPFC_FC_32B_DEVICE_HEADER,
+	LPFC_FC_64B_DEVICE_HEADER,
+};
+
 /*
  *  Extended Link Service LS_COMMAND codes (Payload Word 0)
  */

drivers/scsi/lpfc/lpfc_hw4.h

@@ -4847,6 +4847,7 @@ struct fcp_iwrite64_wqe {
 #define	cmd_buff_len_SHIFT  16
 #define	cmd_buff_len_MASK  0x00000ffff
 #define	cmd_buff_len_WORD  word3
+/* Note: payload_offset_len field depends on ASIC support */
 #define payload_offset_len_SHIFT 0
 #define payload_offset_len_MASK 0x0000ffff
 #define payload_offset_len_WORD word3
@@ -4863,6 +4864,7 @@ struct fcp_iread64_wqe {
 #define	cmd_buff_len_SHIFT  16
 #define	cmd_buff_len_MASK  0x00000ffff
 #define	cmd_buff_len_WORD  word3
+/* Note: payload_offset_len field depends on ASIC support */
 #define payload_offset_len_SHIFT 0
 #define payload_offset_len_MASK 0x0000ffff
 #define payload_offset_len_WORD word3
@@ -4879,6 +4881,7 @@ struct fcp_icmnd64_wqe {
 #define	cmd_buff_len_SHIFT  16
 #define	cmd_buff_len_MASK  0x00000ffff
 #define	cmd_buff_len_WORD  word3
+/* Note: payload_offset_len field depends on ASIC support */
 #define payload_offset_len_SHIFT 0
 #define payload_offset_len_MASK 0x0000ffff
 #define payload_offset_len_WORD word3

drivers/scsi/lpfc/lpfc_init.c

@@ -4699,6 +4699,7 @@ lpfc_create_port(struct lpfc_hba *phba, int instance, struct device *dev)
 	uint64_t wwn;
 	bool use_no_reset_hba = false;
 	int rc;
+	u8 if_type;
 
 	if (lpfc_no_hba_reset_cnt) {
 		if (phba->sli_rev < LPFC_SLI_REV4 &&
@@ -4773,10 +4774,24 @@ lpfc_create_port(struct lpfc_hba *phba, int instance, struct device *dev)
 	shost->max_id = LPFC_MAX_TARGET;
 	shost->max_lun = vport->cfg_max_luns;
 	shost->this_id = -1;
-	if (phba->sli_rev == LPFC_SLI_REV4)
+
+	/* Set max_cmd_len applicable to ASIC support */
+	if (phba->sli_rev == LPFC_SLI_REV4) {
+		if_type = bf_get(lpfc_sli_intf_if_type,
+				 &phba->sli4_hba.sli_intf);
+		switch (if_type) {
+		case LPFC_SLI_INTF_IF_TYPE_2:
+			fallthrough;
+		case LPFC_SLI_INTF_IF_TYPE_6:
 			shost->max_cmd_len = LPFC_FCP_CDB_LEN_32;
-	else
+			break;
+		default:
 			shost->max_cmd_len = LPFC_FCP_CDB_LEN;
+			break;
+		}
+	} else {
+		shost->max_cmd_len = LPFC_FCP_CDB_LEN;
+	}
 
 	if (phba->sli_rev == LPFC_SLI_REV4) {
 		if (!phba->cfg_fcp_mq_threshold ||
@@ -10436,6 +10451,7 @@ lpfc_sli4_queue_create(struct lpfc_hba *phba)
 	struct lpfc_vector_map_info *cpup;
 	struct lpfc_vector_map_info *eqcpup;
 	struct lpfc_eq_intr_info *eqi;
+	u32 wqesize;
 
 	/*
 	 * Create HBA Record arrays.
@@ -10655,9 +10671,15 @@ lpfc_sli4_queue_create(struct lpfc_hba *phba)
 	 * Create ELS Work Queues
 	 */
 
-	/* Create slow-path ELS Work Queue */
+	/*
+	 * Create slow-path ELS Work Queue.
+	 * Increase the ELS WQ size when WQEs contain an embedded cdb
+	 */
+	wqesize = (phba->fcp_embed_io) ?
+			LPFC_WQE128_SIZE : phba->sli4_hba.wq_esize;
+
 	qdesc = lpfc_sli4_queue_alloc(phba, LPFC_DEFAULT_PAGE_SIZE,
-				      phba->sli4_hba.wq_esize,
+				      wqesize,
 				      phba->sli4_hba.wq_ecount, cpu);
 	if (!qdesc) {
 		lpfc_printf_log(phba, KERN_ERR, LOG_TRACE_EVENT,

drivers/scsi/lpfc/lpfc_scsi.c

@@ -4760,7 +4760,7 @@ static int lpfc_scsi_prep_cmnd_buf_s4(struct lpfc_vport *vport,
 
 	 /* Word 3 */
 	bf_set(payload_offset_len, &wqe->fcp_icmd,
-	       sizeof(struct fcp_cmnd32) + sizeof(struct fcp_rsp));
+	       sizeof(struct fcp_cmnd) + sizeof(struct fcp_rsp));
 
 	/* Word 6 */
 	bf_set(wqe_ctxt_tag, &wqe->generic.wqe_com,

drivers/scsi/lpfc/lpfc_sli.c

@@ -1940,12 +1940,15 @@ lpfc_issue_cmf_sync_wqe(struct lpfc_hba *phba, u32 ms, u64 total)
 	atot = atomic_xchg(&phba->cgn_sync_alarm_cnt, 0);
 	wtot = atomic_xchg(&phba->cgn_sync_warn_cnt, 0);
 
+	spin_lock_irqsave(&phba->hbalock, iflags);
+
 	/* ONLY Managed mode will send the CMF_SYNC_WQE to the HBA */
 	if (phba->cmf_active_mode != LPFC_CFG_MANAGED ||
-	    phba->link_state == LPFC_LINK_DOWN)
-		return 0;
+	    phba->link_state < LPFC_LINK_UP) {
+		ret_val = 0;
+		goto out_unlock;
+	}
 
-	spin_lock_irqsave(&phba->hbalock, iflags);
 	sync_buf = __lpfc_sli_get_iocbq(phba);
 	if (!sync_buf) {
 		lpfc_printf_log(phba, KERN_ERR, LOG_CGN_MGMT,
@@ -11090,9 +11093,17 @@ __lpfc_sli_prep_xmit_seq64_s4(struct lpfc_iocbq *cmdiocbq,
 	/* Word 9 */
 	bf_set(wqe_rcvoxid, &wqe->xmit_sequence.wqe_com, ox_id);
 
+	if (cmdiocbq->cmd_flag & (LPFC_IO_LIBDFC | LPFC_IO_LOOPBACK)) {
+		/* Word 10 */
+		if (cmdiocbq->cmd_flag & LPFC_IO_VMID) {
+			bf_set(wqe_appid, &wqe->xmit_sequence.wqe_com, 1);
+			bf_set(wqe_wqes, &wqe->xmit_sequence.wqe_com, 1);
+			wqe->words[31] = LOOPBACK_SRC_APPID;
+		}
+
 		/* Word 12 */
-	if (cmdiocbq->cmd_flag & (LPFC_IO_LIBDFC | LPFC_IO_LOOPBACK))
 		wqe->xmit_sequence.xmit_len = full_size;
+	}
 	else
 		wqe->xmit_sequence.xmit_len =
 			wqe->xmit_sequence.bde.tus.f.bdeSize;
@@ -18431,6 +18442,7 @@ lpfc_fc_frame_check(struct lpfc_hba *phba, struct fc_frame_header *fc_hdr)
 {
 	/*  make rctl_names static to save stack space */
 	struct fc_vft_header *fc_vft_hdr;
+	struct fc_app_header *fc_app_hdr;
 	uint32_t *header = (uint32_t *) fc_hdr;
 
 #define FC_RCTL_MDS_DIAGS	0xF4
@@ -18486,6 +18498,32 @@ lpfc_fc_frame_check(struct lpfc_hba *phba, struct fc_frame_header *fc_hdr)
 		goto drop;
 	}
 
+	if (unlikely(phba->link_flag == LS_LOOPBACK_MODE &&
+				phba->cfg_vmid_app_header)) {
+		/* Application header is 16B device header */
+		if (fc_hdr->fh_df_ctl & LPFC_FC_16B_DEVICE_HEADER) {
+			fc_app_hdr = (struct fc_app_header *) (fc_hdr + 1);
+			if (be32_to_cpu(fc_app_hdr->src_app_id) !=
+					LOOPBACK_SRC_APPID) {
+				lpfc_printf_log(phba, KERN_WARNING,
+						LOG_ELS | LOG_LIBDFC,
+						"1932 Loopback src app id "
+						"not matched, app_id:x%x\n",
+						be32_to_cpu(fc_app_hdr->src_app_id));
+
+				goto drop;
+			}
+		} else {
+			lpfc_printf_log(phba, KERN_WARNING,
+					LOG_ELS | LOG_LIBDFC,
+					"1933 Loopback df_ctl bit not set, "
+					"df_ctl:x%x\n",
+					fc_hdr->fh_df_ctl);
+
+			goto drop;
+		}
+	}
+
 	lpfc_printf_log(phba, KERN_INFO, LOG_ELS,
 			"2538 Received frame rctl:x%x, type:x%x, "
 			"frame Data:%08x %08x %08x %08x %08x %08x %08x\n",

drivers/scsi/lpfc/lpfc_version.h

@@ -20,7 +20,7 @@
  * included with this package.                                     *
  *******************************************************************/
 
-#define LPFC_DRIVER_VERSION "14.4.0.4"
+#define LPFC_DRIVER_VERSION "14.4.0.5"
 #define LPFC_DRIVER_NAME		"lpfc"
 
 /* Used for SLI 2/3 */

drivers/scsi/lpfc/lpfc_vport.c

@@ -626,6 +626,7 @@ lpfc_vport_delete(struct fc_vport *fc_vport)
 	struct Scsi_Host *shost = lpfc_shost_from_vport(vport);
 	struct lpfc_hba  *phba = vport->phba;
 	int rc;
+	DECLARE_WAIT_QUEUE_HEAD_ONSTACK(waitq);
 
 	if (vport->port_type == LPFC_PHYSICAL_PORT) {
 		lpfc_printf_vlog(vport, KERN_ERR, LOG_TRACE_EVENT,
@@ -679,21 +680,49 @@ lpfc_vport_delete(struct fc_vport *fc_vport)
 	if (!ndlp)
 		goto skip_logo;
 
+	/* Send the DA_ID and Fabric LOGO to cleanup the NPIV fabric entries. */
 	if (ndlp && ndlp->nlp_state == NLP_STE_UNMAPPED_NODE &&
 	    phba->link_state >= LPFC_LINK_UP &&
 	    phba->fc_topology != LPFC_TOPOLOGY_LOOP) {
 		if (vport->cfg_enable_da_id) {
-			/* Send DA_ID and wait for a completion. */
+			/* Send DA_ID and wait for a completion.  This is best
+			 * effort.  If the DA_ID fails, likely the fabric will
+			 * "leak" NportIDs but at least the driver issued the
+			 * command.
+			 */
+			ndlp = lpfc_findnode_did(vport, NameServer_DID);
+			if (!ndlp)
+				goto issue_logo;
+
+			spin_lock_irq(&ndlp->lock);
+			ndlp->da_id_waitq = &waitq;
+			ndlp->save_flags |= NLP_WAIT_FOR_DA_ID;
+			spin_unlock_irq(&ndlp->lock);
+
 			rc = lpfc_ns_cmd(vport, SLI_CTNS_DA_ID, 0, 0);
-			if (rc) {
-				lpfc_printf_log(vport->phba, KERN_WARNING,
-						LOG_VPORT,
-						"1829 CT command failed to "
-						"delete objects on fabric, "
-						"rc %d\n", rc);
-			}
+			if (!rc) {
+				wait_event_timeout(waitq,
+				   !(ndlp->save_flags & NLP_WAIT_FOR_DA_ID),
+				   msecs_to_jiffies(phba->fc_ratov * 2000));
 			}
 
+			lpfc_printf_vlog(vport, KERN_INFO, LOG_VPORT | LOG_ELS,
+					 "1829 DA_ID issue status %d. "
+					 "SFlag x%x NState x%x, NFlag x%x "
+					 "Rpi x%x\n",
+					 rc, ndlp->save_flags, ndlp->nlp_state,
+					 ndlp->nlp_flag, ndlp->nlp_rpi);
+
+			/* Remove the waitq and save_flags.  It no
+			 * longer matters if the wake happened.
+			 */
+			spin_lock_irq(&ndlp->lock);
+			ndlp->da_id_waitq = NULL;
+			ndlp->save_flags &= ~NLP_WAIT_FOR_DA_ID;
+			spin_unlock_irq(&ndlp->lock);
+		}
+
+issue_logo:
 		/*
 		 * If the vpi is not registered, then a valid FDISC doesn't
 		 * exist and there is no need for a ELS LOGO.  Just cleanup

drivers/scsi/mpi3mr/mpi/mpi30_cnfg.h

@@ -67,6 +67,7 @@
 #define MPI3_SECURITY_PGAD_SLOT_GROUP_MASK              (0x0000ff00)
 #define MPI3_SECURITY_PGAD_SLOT_GROUP_SHIFT		(8)
 #define MPI3_SECURITY_PGAD_SLOT_MASK                    (0x000000ff)
+#define MPI3_INSTANCE_PGAD_INSTANCE_MASK                (0x0000ffff)
 struct mpi3_config_request {
 	__le16             host_tag;
 	u8                 ioc_use_only02;
@@ -75,7 +76,8 @@ struct mpi3_config_request {
 	u8                 ioc_use_only06;
 	u8                 msg_flags;
 	__le16             change_count;
-	__le16             reserved0a;
+	u8                 proxy_ioc_number;
+	u8                 reserved0b;
 	u8                 page_version;
 	u8                 page_number;
 	u8                 page_type;
@@ -206,6 +208,9 @@ struct mpi3_config_page_header {
 #define MPI3_MFGPAGE_DEVID_SAS5116_MPI_MGMT		(0x00b5)
 #define MPI3_MFGPAGE_DEVID_SAS5116_NVME_MGMT		(0x00b6)
 #define MPI3_MFGPAGE_DEVID_SAS5116_PCIE_SWITCH		(0x00b8)
+#define MPI3_MFGPAGE_DEVID_SAS5248_MPI			(0x00f0)
+#define MPI3_MFGPAGE_DEVID_SAS5248_MPI_NS		(0x00f1)
+#define MPI3_MFGPAGE_DEVID_SAS5248_PCIE_SWITCH		(0x00f2)
 struct mpi3_man_page0 {
 	struct mpi3_config_page_header         header;
 	u8                                 chip_revision[8];
@@ -1074,6 +1079,8 @@ struct mpi3_io_unit_page8 {
 #define MPI3_IOUNIT8_SBSTATE_SVN_UPDATE_PENDING   (0x04)
 #define MPI3_IOUNIT8_SBSTATE_KEY_UPDATE_PENDING   (0x02)
 #define MPI3_IOUNIT8_SBSTATE_SECURE_BOOT_ENABLED  (0x01)
+#define MPI3_IOUNIT8_SBMODE_CURRENT_KEY_IOUNIT17	(0x10)
+#define MPI3_IOUNIT8_SBMODE_HARD_SECURE_RECERTIFIED	(0x08)
 struct mpi3_io_unit_page9 {
 	struct mpi3_config_page_header         header;
 	__le32                             flags;
@@ -1089,6 +1096,8 @@ struct mpi3_io_unit_page9 {
 #define MPI3_IOUNIT9_FLAGS_UBM_ENCLOSURE_ORDER_BACKPLANE_TYPE     (0x00000004)
 #define MPI3_IOUNIT9_FLAGS_VDFIRST_ENABLED                        (0x00000001)
 #define MPI3_IOUNIT9_FIRSTDEVICE_UNKNOWN                          (0xffff)
+#define MPI3_IOUNIT9_FIRSTDEVICE_IN_DRIVER_PAGE_0                 (0xfffe)
+
 struct mpi3_io_unit_page10 {
 	struct mpi3_config_page_header         header;
 	u8                                 flags;
@@ -1224,6 +1233,19 @@ struct mpi3_io_unit_page15 {
 #define MPI3_IOUNIT15_FLAGS_EPRSUPPORT_WITHOUT_POWER_BRAKE_GPIO     (0x01)
 #define MPI3_IOUNIT15_FLAGS_EPRSUPPORT_WITH_POWER_BRAKE_GPIO        (0x02)
 #define MPI3_IOUNIT15_NUMPOWERBUDGETDATA_POWER_BUDGETING_DISABLED   (0x00)
+
+struct mpi3_io_unit_page17 {
+	struct mpi3_config_page_header         header;
+	u8                                 num_instances;
+	u8                                 instance;
+	__le16                             reserved0a;
+	__le32                             reserved0c[4];
+	__le16                             key_length;
+	u8                                 encryption_algorithm;
+	u8                                 reserved1f;
+	__le32                             current_key[];
+};
+#define MPI3_IOUNIT17_PAGEVERSION		(0x00)
 struct mpi3_ioc_page0 {
 	struct mpi3_config_page_header         header;
 	__le32                             reserved08;
@@ -1311,7 +1333,7 @@ struct mpi3_driver_page0 {
 	u8                                 tur_interval;
 	u8                                 reserved10;
 	u8                                 security_key_timeout;
-	__le16                             reserved12;
+	__le16                             first_device;
 	__le32                             reserved14;
 	__le32                             reserved18;
 };
@@ -1324,10 +1346,13 @@ struct mpi3_driver_page0 {
 #define MPI3_DRIVER0_BSDOPTS_REGISTRATION_IOC_AND_DEVS      (0x00000000)
 #define MPI3_DRIVER0_BSDOPTS_REGISTRATION_IOC_ONLY          (0x00000001)
 #define MPI3_DRIVER0_BSDOPTS_REGISTRATION_IOC_AND_INTERNAL_DEVS		(0x00000002)
+#define MPI3_DRIVER0_FIRSTDEVICE_IGNORE1                            (0x0000)
+#define MPI3_DRIVER0_FIRSTDEVICE_IGNORE2                            (0xffff)
 struct mpi3_driver_page1 {
 	struct mpi3_config_page_header         header;
 	__le32                             flags;
-	__le32                             reserved0c;
+	u8                                 time_stamp_update;
+	u8                                 reserved0d[3];
 	__le16                             host_diag_trace_max_size;
 	__le16                             host_diag_trace_min_size;
 	__le16                             host_diag_trace_decrement_size;
@@ -2347,6 +2372,10 @@ struct mpi3_device0_vd_format {
 #define MPI3_DEVICE0_VD_DEVICE_INFO_SAS                     (0x0001)
 #define MPI3_DEVICE0_VD_FLAGS_IO_THROTTLE_GROUP_QD_MASK     (0xf000)
 #define MPI3_DEVICE0_VD_FLAGS_IO_THROTTLE_GROUP_QD_SHIFT    (12)
+#define MPI3_DEVICE0_VD_FLAGS_OSEXPOSURE_MASK               (0x0003)
+#define MPI3_DEVICE0_VD_FLAGS_OSEXPOSURE_HDD                (0x0000)
+#define MPI3_DEVICE0_VD_FLAGS_OSEXPOSURE_SSD                (0x0001)
+#define MPI3_DEVICE0_VD_FLAGS_OSEXPOSURE_NO_GUIDANCE        (0x0002)
 union mpi3_device0_dev_spec_format {
 	struct mpi3_device0_sas_sata_format        sas_sata_format;
 	struct mpi3_device0_pcie_format            pcie_format;

drivers/scsi/mpi3mr/mpi/mpi30_image.h

@@ -205,13 +205,14 @@ struct mpi3_encrypted_hash_entry {
 	u8                         hash_image_type;
 	u8                         hash_algorithm;
 	u8                         encryption_algorithm;
-	u8                         reserved03;
+	u8                         flags;
 	__le16                     public_key_size;
 	__le16                     signature_size;
 	__le32                     public_key[MPI3_PUBLIC_KEY_MAX];
 };
-
-#define MPI3_HASH_IMAGE_TYPE_KEY_WITH_SIGNATURE      (0x03)
+#define MPI3_HASH_IMAGE_TYPE_KEY_WITH_HASH           (0x03)
+#define MPI3_HASH_IMAGE_TYPE_KEY_WITH_HASH_1_OF_2    (0x04)
+#define MPI3_HASH_IMAGE_TYPE_KEY_WITH_HASH_2_OF_2    (0x05)
 #define MPI3_HASH_ALGORITHM_VERSION_MASK             (0xe0)
 #define MPI3_HASH_ALGORITHM_VERSION_NONE             (0x00)
 #define MPI3_HASH_ALGORITHM_VERSION_SHA1             (0x20)
@@ -230,6 +231,12 @@ struct mpi3_encrypted_hash_entry {
 #define MPI3_ENCRYPTION_ALGORITHM_RSA4096            (0x05)
 #define MPI3_ENCRYPTION_ALGORITHM_RSA3072            (0x06)
 
+/* hierarchical signature system (hss) */
+#define MPI3_ENCRYPTION_ALGORITHM_ML_DSA_87	    (0x0b)
+#define MPI3_ENCRYPTION_ALGORITHM_ML_DSA_65	    (0x0c)
+#define MPI3_ENCRYPTION_ALGORITHM_ML_DSA_44	    (0x0d)
+#define MPI3_ENCRYPTED_HASH_ENTRY_FLAGS_PAIRED_KEY_MASK		(0x0f)
+
 #ifndef MPI3_ENCRYPTED_HASH_ENTRY_MAX
 #define MPI3_ENCRYPTED_HASH_ENTRY_MAX               (1)
 #endif

drivers/scsi/mpi3mr/mpi/mpi30_ioc.h

@@ -39,6 +39,12 @@ struct mpi3_ioc_init_request {
 #define MPI3_WHOINIT_HOST_DRIVER                         (0x03)
 #define MPI3_WHOINIT_MANUFACTURER                        (0x04)
 
+#define MPI3_IOCINIT_DRIVERCAP_OSEXPOSURE_MASK              (0x00000003)
+#define MPI3_IOCINIT_DRIVERCAP_OSEXPOSURE_NO_GUIDANCE       (0x00000000)
+#define MPI3_IOCINIT_DRIVERCAP_OSEXPOSURE_NO_SPECIAL        (0x00000001)
+#define MPI3_IOCINIT_DRIVERCAP_OSEXPOSURE_REPORT_AS_HDD     (0x00000002)
+#define MPI3_IOCINIT_DRIVERCAP_OSEXPOSURE_REPORT_AS_SSD     (0x00000003)
+
 struct mpi3_ioc_facts_request {
 	__le16                 host_tag;
 	u8                     ioc_use_only02;
@@ -140,6 +146,8 @@ struct mpi3_ioc_facts_data {
 #define MPI3_IOCFACTS_EXCEPT_MANUFACT_CHECKSUM_FAIL           (0x0020)
 #define MPI3_IOCFACTS_EXCEPT_FW_CHECKSUM_FAIL                 (0x0010)
 #define MPI3_IOCFACTS_EXCEPT_CONFIG_CHECKSUM_FAIL             (0x0008)
+#define MPI3_IOCFACTS_EXCEPT_BLOCKING_BOOT_EVENT              (0x0004)
+#define MPI3_IOCFACTS_EXCEPT_SECURITY_SELFTEST_FAILURE        (0x0002)
 #define MPI3_IOCFACTS_EXCEPT_BOOTSTAT_MASK                    (0x0001)
 #define MPI3_IOCFACTS_EXCEPT_BOOTSTAT_PRIMARY                 (0x0000)
 #define MPI3_IOCFACTS_EXCEPT_BOOTSTAT_SECONDARY               (0x0001)

drivers/scsi/mpi3mr/mpi/mpi30_transport.h

@@ -18,7 +18,7 @@ union mpi3_version_union {
 
 #define MPI3_VERSION_MAJOR                                              (3)
 #define MPI3_VERSION_MINOR                                              (0)
-#define MPI3_VERSION_UNIT                                               (31)
+#define MPI3_VERSION_UNIT                                               (34)
 #define MPI3_VERSION_DEV                                                (0)
 #define MPI3_DEVHANDLE_INVALID                                          (0xffff)
 struct mpi3_sysif_oper_queue_indexes {
@@ -158,6 +158,7 @@ struct mpi3_sysif_registers {
 #define MPI3_SYSIF_FAULT_CODE_SOFT_RESET_NEEDED                         (0x0000f004)
 #define MPI3_SYSIF_FAULT_CODE_POWER_CYCLE_REQUIRED                      (0x0000f005)
 #define MPI3_SYSIF_FAULT_CODE_TEMP_THRESHOLD_EXCEEDED                   (0x0000f006)
+#define MPI3_SYSIF_FAULT_CODE_INSUFFICIENT_PCI_SLOT_POWER               (0x0000f007)
 #define MPI3_SYSIF_FAULT_INFO0_OFFSET                                   (0x00001c14)
 #define MPI3_SYSIF_FAULT_INFO1_OFFSET                                   (0x00001c18)
 #define MPI3_SYSIF_FAULT_INFO2_OFFSET                                   (0x00001c1c)
@@ -410,6 +411,7 @@ struct mpi3_default_reply {
 #define MPI3_IOCSTATUS_INSUFFICIENT_RESOURCES       (0x0006)
 #define MPI3_IOCSTATUS_INVALID_FIELD                (0x0007)
 #define MPI3_IOCSTATUS_INVALID_STATE                (0x0008)
+#define MPI3_IOCSTATUS_SHUTDOWN_ACTIVE              (0x0009)
 #define MPI3_IOCSTATUS_INSUFFICIENT_POWER           (0x000a)
 #define MPI3_IOCSTATUS_INVALID_CHANGE_COUNT         (0x000b)
 #define MPI3_IOCSTATUS_ALLOWED_CMD_BLOCK            (0x000c)

drivers/scsi/mpi3mr/mpi3mr.h

@@ -57,8 +57,8 @@ extern struct list_head mrioc_list;
 extern int prot_mask;
 extern atomic64_t event_counter;
 
-#define MPI3MR_DRIVER_VERSION	"8.10.0.5.50"
-#define MPI3MR_DRIVER_RELDATE	"08-Aug-2024"
+#define MPI3MR_DRIVER_VERSION	"8.12.0.0.50"
+#define MPI3MR_DRIVER_RELDATE	"05-Sept-2024"
 
 #define MPI3MR_DRIVER_NAME	"mpi3mr"
 #define MPI3MR_DRIVER_LICENSE	"GPL"
@@ -178,7 +178,7 @@ extern atomic64_t event_counter;
 #define MPI3MR_DEFAULT_SDEV_QD	32
 
 /* Definitions for Threaded IRQ poll*/
-#define MPI3MR_IRQ_POLL_SLEEP			2
+#define MPI3MR_IRQ_POLL_SLEEP			20
 #define MPI3MR_IRQ_POLL_TRIGGER_IOCOUNT		8
 
 /* Definitions for the controller security status*/
@@ -1090,6 +1090,7 @@ struct scmd_priv {
  * @evtack_cmds_bitmap: Event Ack bitmap
  * @delayed_evtack_cmds_list: Delayed event acknowledgment list
  * @ts_update_counter: Timestamp update counter
+ * @ts_update_interval: Timestamp update interval
  * @reset_in_progress: Reset in progress flag
  * @unrecoverable: Controller unrecoverable flag
  * @prev_reset_result: Result of previous reset
@@ -1277,7 +1278,8 @@ struct mpi3mr_ioc {
 	unsigned long *evtack_cmds_bitmap;
 	struct list_head delayed_evtack_cmds_list;
 
-	u32 ts_update_counter;
+	u16 ts_update_counter;
+	u16 ts_update_interval;
 	u8 reset_in_progress;
 	u8 unrecoverable;
 	int prev_reset_result;

drivers/scsi/mpi3mr/mpi3mr_fw.c

@@ -728,7 +728,7 @@ static irqreturn_t mpi3mr_isr_poll(int irq, void *privdata)
 			    mpi3mr_process_op_reply_q(mrioc,
 				intr_info->op_reply_q);
 
-		usleep_range(MPI3MR_IRQ_POLL_SLEEP, 10 * MPI3MR_IRQ_POLL_SLEEP);
+		usleep_range(MPI3MR_IRQ_POLL_SLEEP, MPI3MR_IRQ_POLL_SLEEP + 1);
 
 	} while (atomic_read(&intr_info->op_reply_q->pend_ios) &&
 	    (num_op_reply < mrioc->max_host_ios));
@@ -1362,6 +1362,10 @@ static int mpi3mr_bring_ioc_ready(struct mpi3mr_ioc *mrioc)
 	int retval = 0;
 	enum mpi3mr_iocstate ioc_state;
 	u64 base_info;
+	u8 retry = 0;
+	u64 start_time, elapsed_time_sec;
+
+retry_bring_ioc_ready:
 
 	ioc_status = readl(&mrioc->sysif_regs->ioc_status);
 	ioc_config = readl(&mrioc->sysif_regs->ioc_configuration);
@@ -1380,26 +1384,23 @@ static int mpi3mr_bring_ioc_ready(struct mpi3mr_ioc *mrioc)
 	ioc_info(mrioc, "controller is in %s state during detection\n",
 	    mpi3mr_iocstate_name(ioc_state));
 
-	if (ioc_state == MRIOC_STATE_BECOMING_READY ||
-	    ioc_state == MRIOC_STATE_RESET_REQUESTED) {
 	timeout = mrioc->ready_timeout * 10;
+
 	do {
-			msleep(100);
-		} while (--timeout);
+		ioc_state = mpi3mr_get_iocstate(mrioc);
+
+		if (ioc_state != MRIOC_STATE_BECOMING_READY &&
+		    ioc_state != MRIOC_STATE_RESET_REQUESTED)
+			break;
 
 		if (!pci_device_is_present(mrioc->pdev)) {
 			mrioc->unrecoverable = 1;
-			ioc_err(mrioc,
-			    "controller is not present while waiting to reset\n");
-			retval = -1;
+			ioc_err(mrioc, "controller is not present while waiting to reset\n");
 			goto out_device_not_present;
 		}
 
-		ioc_state = mpi3mr_get_iocstate(mrioc);
-		ioc_info(mrioc,
-		    "controller is in %s state after waiting to reset\n",
-		    mpi3mr_iocstate_name(ioc_state));
-	}
+		msleep(100);
+	} while (--timeout);
 
 	if (ioc_state == MRIOC_STATE_READY) {
 		ioc_info(mrioc, "issuing message unit reset (MUR) to bring to reset state\n");
@@ -1460,6 +1461,9 @@ static int mpi3mr_bring_ioc_ready(struct mpi3mr_ioc *mrioc)
 	ioc_config |= MPI3_SYSIF_IOC_CONFIG_ENABLE_IOC;
 	writel(ioc_config, &mrioc->sysif_regs->ioc_configuration);
 
+	if (retry == 0)
+		start_time = jiffies;
+
 	timeout = mrioc->ready_timeout * 10;
 	do {
 		ioc_state = mpi3mr_get_iocstate(mrioc);
@@ -1469,6 +1473,12 @@ static int mpi3mr_bring_ioc_ready(struct mpi3mr_ioc *mrioc)
 			    mpi3mr_iocstate_name(ioc_state));
 			return 0;
 		}
+		ioc_status = readl(&mrioc->sysif_regs->ioc_status);
+		if ((ioc_status & MPI3_SYSIF_IOC_STATUS_RESET_HISTORY) ||
+		    (ioc_status & MPI3_SYSIF_IOC_STATUS_FAULT)) {
+			mpi3mr_print_fault_info(mrioc);
+			goto out_failed;
+		}
 		if (!pci_device_is_present(mrioc->pdev)) {
 			mrioc->unrecoverable = 1;
 			ioc_err(mrioc,
@@ -1477,9 +1487,19 @@ static int mpi3mr_bring_ioc_ready(struct mpi3mr_ioc *mrioc)
 			goto out_device_not_present;
 		}
 		msleep(100);
-	} while (--timeout);
+		elapsed_time_sec = jiffies_to_msecs(jiffies - start_time)/1000;
+	} while (elapsed_time_sec < mrioc->ready_timeout);
 
 out_failed:
+	elapsed_time_sec = jiffies_to_msecs(jiffies - start_time)/1000;
+	if ((retry < 2) && (elapsed_time_sec < (mrioc->ready_timeout - 60))) {
+		retry++;
+
+		ioc_warn(mrioc, "retrying to bring IOC ready, retry_count:%d\n"
+				" elapsed time =%llu\n", retry, elapsed_time_sec);
+
+		goto retry_bring_ioc_ready;
+	}
 	ioc_state = mpi3mr_get_iocstate(mrioc);
 	ioc_err(mrioc,
 	    "failed to bring to ready state,  current state: %s\n",
@@ -2671,7 +2691,7 @@ static void mpi3mr_watchdog_work(struct work_struct *work)
 		return;
 	}
 
-	if (mrioc->ts_update_counter++ >= MPI3MR_TSUPDATE_INTERVAL) {
+	if (mrioc->ts_update_counter++ >= mrioc->ts_update_interval) {
 		mrioc->ts_update_counter = 0;
 		mpi3mr_sync_timestamp(mrioc);
 	}
@@ -3844,6 +3864,29 @@ static int mpi3mr_repost_diag_bufs(struct mpi3mr_ioc *mrioc)
 	return retval;
 }
 
+/**
+ * mpi3mr_read_tsu_interval - Update time stamp interval
+ * @mrioc: Adapter instance reference
+ *
+ * Update time stamp interval if its defined in driver page 1,
+ * otherwise use default value.
+ *
+ * Return: Nothing
+ */
+static void
+mpi3mr_read_tsu_interval(struct mpi3mr_ioc *mrioc)
+{
+	struct mpi3_driver_page1 driver_pg1;
+	u16 pg_sz = sizeof(driver_pg1);
+	int retval = 0;
+
+	mrioc->ts_update_interval = MPI3MR_TSUPDATE_INTERVAL;
+
+	retval = mpi3mr_cfg_get_driver_pg1(mrioc, &driver_pg1, pg_sz);
+	if (!retval && driver_pg1.time_stamp_update)
+		mrioc->ts_update_interval = (driver_pg1.time_stamp_update * 60);
+}
+
 /**
  * mpi3mr_print_ioc_info - Display controller information
  * @mrioc: Adapter instance reference
@@ -4140,6 +4183,7 @@ int mpi3mr_init_ioc(struct mpi3mr_ioc *mrioc)
 		goto out_failed_noretry;
 	}
 
+	mpi3mr_read_tsu_interval(mrioc);
 	mpi3mr_print_ioc_info(mrioc);
 
 	if (!mrioc->cfg_page) {
@@ -4321,6 +4365,7 @@ int mpi3mr_reinit_ioc(struct mpi3mr_ioc *mrioc, u8 is_resume)
 		goto out_failed_noretry;
 	}
 
+	mpi3mr_read_tsu_interval(mrioc);
 	mpi3mr_print_ioc_info(mrioc);
 
 	if (is_resume) {

drivers/scsi/mpt3sas/mpt3sas_base.c

@@ -8898,9 +8898,8 @@ _base_check_ioc_facts_changes(struct MPT3SAS_ADAPTER *ioc)
 		    ioc->device_remove_in_progress, pd_handles_sz, GFP_KERNEL);
 		if (!device_remove_in_progress) {
 			ioc_info(ioc,
-			    "Unable to allocate the memory for "
-			    "device_remove_in_progress of sz: %d\n "
-			    , pd_handles_sz);
+			    "Unable to allocate the memory for device_remove_in_progress of sz: %d\n",
+			    pd_handles_sz);
 			return -ENOMEM;
 		}
 		memset(device_remove_in_progress +

drivers/scsi/pm8001/pm8001_init.c

@@ -100,10 +100,12 @@ static void pm8001_map_queues(struct Scsi_Host *shost)
 	struct pm8001_hba_info *pm8001_ha = sha->lldd_ha;
 	struct blk_mq_queue_map *qmap = &shost->tag_set.map[HCTX_TYPE_DEFAULT];
 
-	if (pm8001_ha->number_of_intr > 1)
+	if (pm8001_ha->number_of_intr > 1) {
 		blk_mq_pci_map_queues(qmap, pm8001_ha->pdev, 1);
+		return;
+	}
 
-	return blk_mq_map_queues(qmap);
+	blk_mq_map_queues(qmap);
 }
 
 /*

drivers/scsi/pmcraid.c

@@ -1946,7 +1946,7 @@ static void pmcraid_soft_reset(struct pmcraid_cmd *cmd)
 	}
 
 	iowrite32(doorbell, pinstance->int_regs.host_ioa_interrupt_reg);
-	ioread32(pinstance->int_regs.host_ioa_interrupt_reg),
+	ioread32(pinstance->int_regs.host_ioa_interrupt_reg);
 	int_reg = ioread32(pinstance->int_regs.ioa_host_interrupt_reg);
 
 	pmcraid_info("Waiting for IOA to become operational %x:%x\n",

drivers/scsi/scsi_debug.c

@@ -2760,7 +2760,6 @@ static int resp_mode_sense(struct scsi_cmnd *scp,
 	else
 		bd_len = 0;
 	alloc_len = msense_6 ? cmd[4] : get_unaligned_be16(cmd + 7);
-	memset(arr, 0, SDEBUG_MAX_MSENSE_SZ);
 	if (0x3 == pcontrol) {  /* Saving values not supported */
 		mk_sense_buffer(scp, ILLEGAL_REQUEST, SAVING_PARAMS_UNSUP, 0);
 		return check_condition_result;

drivers/scsi/sd.c

@@ -38,7 +38,6 @@
 #include <linux/fs.h>
 #include <linux/kernel.h>
 #include <linux/mm.h>
-#include <linux/bio-integrity.h>
 #include <linux/hdreg.h>
 #include <linux/errno.h>
 #include <linux/idr.h>
@@ -3404,7 +3403,7 @@ static void sd_read_block_characteristics(struct scsi_disk *sdkp,
 	rcu_read_lock();
 	vpd = rcu_dereference(sdkp->device->vpd_pgb1);
 
-	if (!vpd || vpd->len < 8) {
+	if (!vpd || vpd->len <= 8) {
 		rcu_read_unlock();
 	        return;
 	}
@@ -4093,9 +4092,38 @@ static int sd_start_stop_device(struct scsi_disk *sdkp, int start)
 {
 	unsigned char cmd[6] = { START_STOP };	/* START_VALID */
 	struct scsi_sense_hdr sshdr;
+	struct scsi_failure failure_defs[] = {
+		{
+			/* Power on, reset, or bus device reset occurred */
+			.sense = UNIT_ATTENTION,
+			.asc = 0x29,
+			.ascq = 0,
+			.result = SAM_STAT_CHECK_CONDITION,
+		},
+		{
+			/* Power on occurred */
+			.sense = UNIT_ATTENTION,
+			.asc = 0x29,
+			.ascq = 1,
+			.result = SAM_STAT_CHECK_CONDITION,
+		},
+		{
+			/* SCSI bus reset */
+			.sense = UNIT_ATTENTION,
+			.asc = 0x29,
+			.ascq = 2,
+			.result = SAM_STAT_CHECK_CONDITION,
+		},
+		{}
+	};
+	struct scsi_failures failures = {
+		.total_allowed = 3,
+		.failure_definitions = failure_defs,
+	};
 	const struct scsi_exec_args exec_args = {
 		.sshdr = &sshdr,
 		.req_flags = BLK_MQ_REQ_PM,
+		.failures = &failures,
 	};
 	struct scsi_device *sdp = sdkp->device;
 	int res;

drivers/scsi/st.c

@@ -834,6 +834,9 @@ static int flush_buffer(struct scsi_tape *STp, int seek_next)
 	int backspace, result;
 	struct st_partstat *STps;
 
+	if (STp->ready != ST_READY)
+		return 0;
+
 	/*
 	 * If there was a bus reset, block further access
 	 * to this device.
@@ -841,8 +844,6 @@ static int flush_buffer(struct scsi_tape *STp, int seek_next)
 	if (STp->pos_unknown)
 		return (-EIO);
 
-	if (STp->ready != ST_READY)
-		return 0;
 	STps = &(STp->ps[STp->partition]);
 	if (STps->rw == ST_WRITING)	/* Writing */
 		return st_flush_write_buffer(STp);

drivers/ufs/host/ufs-qcom.c

@@ -93,7 +93,7 @@ static const struct __ufs_qcom_bw_table {
 	[MODE_HS_RB][UFS_HS_G3][UFS_LANE_2] = { 1492582,	204800 },
 	[MODE_HS_RB][UFS_HS_G4][UFS_LANE_2] = { 2915200,	409600 },
 	[MODE_HS_RB][UFS_HS_G5][UFS_LANE_2] = { 5836800,	819200 },
-	[MODE_MAX][0][0]		    = { 7643136,	307200 },
+	[MODE_MAX][0][0]		    = { 7643136,	819200 },
 };
 
 static void ufs_qcom_get_default_testbus_cfg(struct ufs_qcom_host *host);

fs/bcachefs/backpointers.c

@@ -501,7 +501,7 @@ static int check_extent_checksum(struct btree_trans *trans,
 	prt_printf(&buf, "\n  %s ", bch2_btree_id_str(o_btree));
 	bch2_bkey_val_to_text(&buf, c, extent2);
 
-	struct nonce nonce = extent_nonce(extent.k->version, p.crc);
+	struct nonce nonce = extent_nonce(extent.k->bversion, p.crc);
 	struct bch_csum csum = bch2_checksum(c, p.crc.csum_type, nonce, data_buf, bytes);
 	if (fsck_err_on(bch2_crc_cmp(csum, p.crc.csum),
 			trans, dup_backpointer_to_bad_csum_extent,

fs/bcachefs/bcachefs.h

@@ -594,6 +594,7 @@ struct bch_dev {
 #define BCH_FS_FLAGS()			\
 	x(new_fs)			\
 	x(started)			\
+	x(clean_recovery)		\
 	x(btree_running)		\
 	x(accounting_replay_done)	\
 	x(may_go_rw)			\
@@ -776,7 +777,7 @@ struct bch_fs {
 		unsigned	nsec_per_time_unit;
 		u64		features;
 		u64		compat;
-		unsigned long	errors_silent[BITS_TO_LONGS(BCH_SB_ERR_MAX)];
+		unsigned long	errors_silent[BITS_TO_LONGS(BCH_FSCK_ERR_MAX)];
 		u64		btrees_lost_data;
 	}			sb;
 

fs/bcachefs/bcachefs_format.h

@@ -217,7 +217,7 @@ struct bkey {
 #if __BYTE_ORDER__ == __ORDER_LITTLE_ENDIAN__
 	__u8		pad[1];
 
-	struct bversion	version;
+	struct bversion	bversion;
 	__u32		size;		/* extent size, in sectors */
 	struct bpos	p;
 #elif __BYTE_ORDER__ == __ORDER_BIG_ENDIAN__
@@ -328,8 +328,8 @@ enum bch_bkey_fields {
 		bkey_format_field(OFFSET,	p.offset),		\
 		bkey_format_field(SNAPSHOT,	p.snapshot),		\
 		bkey_format_field(SIZE,		size),			\
-		bkey_format_field(VERSION_HI,	version.hi),		\
-		bkey_format_field(VERSION_LO,	version.lo),		\
+		bkey_format_field(VERSION_HI,	bversion.hi),		\
+		bkey_format_field(VERSION_LO,	bversion.lo),		\
 	},								\
 })
 

fs/bcachefs/bkey.h

@@ -214,9 +214,9 @@ static __always_inline int bversion_cmp(struct bversion l, struct bversion r)
 #define ZERO_VERSION	((struct bversion) { .hi = 0, .lo = 0 })
 #define MAX_VERSION	((struct bversion) { .hi = ~0, .lo = ~0ULL })
 
-static __always_inline int bversion_zero(struct bversion v)
+static __always_inline bool bversion_zero(struct bversion v)
 {
-	return !bversion_cmp(v, ZERO_VERSION);
+	return bversion_cmp(v, ZERO_VERSION) == 0;
 }
 
 #ifdef CONFIG_BCACHEFS_DEBUG
@@ -554,8 +554,8 @@ static inline void bch2_bkey_pack_test(void) {}
 	x(BKEY_FIELD_OFFSET,		p.offset)			\
 	x(BKEY_FIELD_SNAPSHOT,		p.snapshot)			\
 	x(BKEY_FIELD_SIZE,		size)				\
-	x(BKEY_FIELD_VERSION_HI,	version.hi)			\
-	x(BKEY_FIELD_VERSION_LO,	version.lo)
+	x(BKEY_FIELD_VERSION_HI,	bversion.hi)			\
+	x(BKEY_FIELD_VERSION_LO,	bversion.lo)
 
 struct bkey_format_state {
 	u64 field_min[BKEY_NR_FIELDS];

fs/bcachefs/bkey_methods.c

@@ -289,7 +289,7 @@ void bch2_bkey_to_text(struct printbuf *out, const struct bkey *k)
 
 		bch2_bpos_to_text(out, k->p);
 
-		prt_printf(out, " len %u ver %llu", k->size, k->version.lo);
+		prt_printf(out, " len %u ver %llu", k->size, k->bversion.lo);
 	} else {
 		prt_printf(out, "(null)");
 	}

fs/bcachefs/bkey_methods.h

@@ -70,7 +70,7 @@ bool bch2_bkey_normalize(struct bch_fs *, struct bkey_s);
 static inline bool bch2_bkey_maybe_mergable(const struct bkey *l, const struct bkey *r)
 {
 	return l->type == r->type &&
-		!bversion_cmp(l->version, r->version) &&
+		!bversion_cmp(l->bversion, r->bversion) &&
 		bpos_eq(l->p, bkey_start_pos(r));
 }
 

fs/bcachefs/btree_gc.c

@@ -513,6 +513,8 @@ int bch2_check_topology(struct bch_fs *c)
 	struct bpos pulled_from_scan = POS_MIN;
 	int ret = 0;
 
+	bch2_trans_srcu_unlock(trans);
+
 	for (unsigned i = 0; i < btree_id_nr_alive(c) && !ret; i++) {
 		struct btree_root *r = bch2_btree_id_root(c, i);
 		bool reconstructed_root = false;
@@ -599,15 +601,15 @@ static int bch2_gc_mark_key(struct btree_trans *trans, enum btree_id btree_id,
 
 	if (initial) {
 		BUG_ON(bch2_journal_seq_verify &&
-		       k.k->version.lo > atomic64_read(&c->journal.seq));
+		       k.k->bversion.lo > atomic64_read(&c->journal.seq));
 
 		if (fsck_err_on(btree_id != BTREE_ID_accounting &&
-				k.k->version.lo > atomic64_read(&c->key_version),
+				k.k->bversion.lo > atomic64_read(&c->key_version),
 				trans, bkey_version_in_future,
 				"key version number higher than recorded %llu\n  %s",
 				atomic64_read(&c->key_version),
 				(bch2_bkey_val_to_text(&buf, c, k), buf.buf)))
-			atomic64_set(&c->key_version, k.k->version.lo);
+			atomic64_set(&c->key_version, k.k->bversion.lo);
 	}
 
 	if (mustfix_fsck_err_on(level && !bch2_dev_btree_bitmap_marked(c, k),

fs/bcachefs/btree_io.c

@@ -1195,6 +1195,10 @@ int bch2_btree_node_read_done(struct bch_fs *c, struct bch_dev *ca,
 	set_btree_bset(b, b->set, &b->data->keys);
 
 	b->nr = bch2_key_sort_fix_overlapping(c, &sorted->keys, iter);
+	memset((uint8_t *)(sorted + 1) + b->nr.live_u64s * sizeof(u64), 0,
+			btree_buf_bytes(b) -
+			sizeof(struct btree_node) -
+			b->nr.live_u64s * sizeof(u64));
 
 	u64s = le16_to_cpu(sorted->keys.u64s);
 	*sorted = *b->data;
@@ -1219,7 +1223,7 @@ int bch2_btree_node_read_done(struct bch_fs *c, struct bch_dev *ca,
 		ret = bch2_bkey_val_validate(c, u.s_c, READ);
 		if (ret == -BCH_ERR_fsck_delete_bkey ||
 		    (bch2_inject_invalid_keys &&
-		     !bversion_cmp(u.k->version, MAX_VERSION))) {
+		     !bversion_cmp(u.k->bversion, MAX_VERSION))) {
 			btree_keys_account_key_drop(&b->nr, 0, k);
 
 			i->u64s = cpu_to_le16(le16_to_cpu(i->u64s) - k->u64s);

fs/bcachefs/btree_node_scan.c

@@ -275,7 +275,7 @@ static int read_btree_nodes(struct find_btree_nodes *f)
 		w->ca		= ca;
 
 		t = kthread_run(read_btree_nodes_worker, w, "read_btree_nodes/%s", ca->name);
-		ret = IS_ERR_OR_NULL(t);
+		ret = PTR_ERR_OR_ZERO(t);
 		if (ret) {
 			percpu_ref_put(&ca->io_ref);
 			closure_put(&cl);

fs/bcachefs/btree_trans_commit.c

@@ -684,10 +684,10 @@ bch2_trans_commit_write_locked(struct btree_trans *trans, unsigned flags,
 	    !(flags & BCH_TRANS_COMMIT_no_journal_res)) {
 		if (bch2_journal_seq_verify)
 			trans_for_each_update(trans, i)
-				i->k->k.version.lo = trans->journal_res.seq;
+				i->k->k.bversion.lo = trans->journal_res.seq;
 		else if (bch2_inject_invalid_keys)
 			trans_for_each_update(trans, i)
-				i->k->k.version = MAX_VERSION;
+				i->k->k.bversion = MAX_VERSION;
 	}
 
 	h = trans->hooks;
@@ -700,27 +700,31 @@ bch2_trans_commit_write_locked(struct btree_trans *trans, unsigned flags,
 
 	struct jset_entry *entry = trans->journal_entries;
 
-	if (likely(!(flags & BCH_TRANS_COMMIT_skip_accounting_apply))) {
 	percpu_down_read(&c->mark_lock);
 
 	for (entry = trans->journal_entries;
 	     entry != (void *) ((u64 *) trans->journal_entries + trans->journal_entries_u64s);
 	     entry = vstruct_next(entry))
-			if (jset_entry_is_key(entry) && entry->start->k.type == KEY_TYPE_accounting) {
+		if (entry->type == BCH_JSET_ENTRY_write_buffer_keys &&
+		    entry->start->k.type == KEY_TYPE_accounting) {
+			BUG_ON(!trans->journal_res.ref);
+
 			struct bkey_i_accounting *a = bkey_i_to_accounting(entry->start);
 
-				a->k.version = journal_pos_to_bversion(&trans->journal_res,
+			a->k.bversion = journal_pos_to_bversion(&trans->journal_res,
 							(u64 *) entry - (u64 *) trans->journal_entries);
-				BUG_ON(bversion_zero(a->k.version));
-				ret = bch2_accounting_mem_mod_locked(trans, accounting_i_to_s_c(a), false, false);
+			BUG_ON(bversion_zero(a->k.bversion));
+
+			if (likely(!(flags & BCH_TRANS_COMMIT_skip_accounting_apply))) {
+				ret = bch2_accounting_mem_mod_locked(trans, accounting_i_to_s_c(a), BCH_ACCOUNTING_normal);
 				if (ret)
 					goto revert_fs_usage;
 			}
+		}
 	percpu_up_read(&c->mark_lock);
 
 	/* XXX: we only want to run this if deltas are nonzero */
 	bch2_trans_account_disk_usage_change(trans);
-	}
 
 	trans_for_each_update(trans, i)
 		if (btree_node_type_has_atomic_triggers(i->bkey_type)) {
@@ -735,6 +739,40 @@ bch2_trans_commit_write_locked(struct btree_trans *trans, unsigned flags,
 			goto fatal_err;
 	}
 
+	trans_for_each_update(trans, i) {
+		enum bch_validate_flags invalid_flags = 0;
+
+		if (!(flags & BCH_TRANS_COMMIT_no_journal_res))
+			invalid_flags |= BCH_VALIDATE_write|BCH_VALIDATE_commit;
+
+		ret = bch2_bkey_validate(c, bkey_i_to_s_c(i->k),
+					 i->bkey_type, invalid_flags);
+		if (unlikely(ret)){
+			bch2_trans_inconsistent(trans, "invalid bkey on insert from %s -> %ps\n",
+						trans->fn, (void *) i->ip_allocated);
+			goto fatal_err;
+		}
+		btree_insert_entry_checks(trans, i);
+	}
+
+	for (struct jset_entry *i = trans->journal_entries;
+	     i != (void *) ((u64 *) trans->journal_entries + trans->journal_entries_u64s);
+	     i = vstruct_next(i)) {
+		enum bch_validate_flags invalid_flags = 0;
+
+		if (!(flags & BCH_TRANS_COMMIT_no_journal_res))
+			invalid_flags |= BCH_VALIDATE_write|BCH_VALIDATE_commit;
+
+		ret = bch2_journal_entry_validate(c, NULL, i,
+						  bcachefs_metadata_version_current,
+						  CPU_BIG_ENDIAN, invalid_flags);
+		if (unlikely(ret)) {
+			bch2_trans_inconsistent(trans, "invalid journal entry on insert from %s\n",
+						trans->fn);
+			goto fatal_err;
+		}
+	}
+
 	if (likely(!(flags & BCH_TRANS_COMMIT_no_journal_res))) {
 		struct journal *j = &c->journal;
 		struct jset_entry *entry;
@@ -798,7 +836,7 @@ bch2_trans_commit_write_locked(struct btree_trans *trans, unsigned flags,
 			struct bkey_s_accounting a = bkey_i_to_s_accounting(entry2->start);
 
 			bch2_accounting_neg(a);
-			bch2_accounting_mem_mod_locked(trans, a.c, false, false);
+			bch2_accounting_mem_mod_locked(trans, a.c, BCH_ACCOUNTING_normal);
 			bch2_accounting_neg(a);
 		}
 	percpu_up_read(&c->mark_lock);
@@ -1019,40 +1057,6 @@ int __bch2_trans_commit(struct btree_trans *trans, unsigned flags)
 	if (ret)
 		goto out_reset;
 
-	trans_for_each_update(trans, i) {
-		enum bch_validate_flags invalid_flags = 0;
-
-		if (!(flags & BCH_TRANS_COMMIT_no_journal_res))
-			invalid_flags |= BCH_VALIDATE_write|BCH_VALIDATE_commit;
-
-		ret = bch2_bkey_validate(c, bkey_i_to_s_c(i->k),
-					 i->bkey_type, invalid_flags);
-		if (unlikely(ret)){
-			bch2_trans_inconsistent(trans, "invalid bkey on insert from %s -> %ps\n",
-						trans->fn, (void *) i->ip_allocated);
-			return ret;
-		}
-		btree_insert_entry_checks(trans, i);
-	}
-
-	for (struct jset_entry *i = trans->journal_entries;
-	     i != (void *) ((u64 *) trans->journal_entries + trans->journal_entries_u64s);
-	     i = vstruct_next(i)) {
-		enum bch_validate_flags invalid_flags = 0;
-
-		if (!(flags & BCH_TRANS_COMMIT_no_journal_res))
-			invalid_flags |= BCH_VALIDATE_write|BCH_VALIDATE_commit;
-
-		ret = bch2_journal_entry_validate(c, NULL, i,
-						  bcachefs_metadata_version_current,
-						  CPU_BIG_ENDIAN, invalid_flags);
-		if (unlikely(ret)) {
-			bch2_trans_inconsistent(trans, "invalid journal entry on insert from %s\n",
-						trans->fn);
-			return ret;
-		}
-	}
-
 	if (unlikely(!test_bit(BCH_FS_may_go_rw, &c->flags))) {
 		ret = do_bch2_trans_commit_to_journal_replay(trans);
 		goto out_reset;

fs/bcachefs/btree_update.h

@@ -220,7 +220,8 @@ static inline struct bkey_i *__bch2_bkey_make_mut_noupdate(struct btree_trans *t
 	if (type && k.k->type != type)
 		return ERR_PTR(-ENOENT);
 
-	mut = bch2_trans_kmalloc_nomemzero(trans, bytes);
+	/* extra padding for varint_decode_fast... */
+	mut = bch2_trans_kmalloc_nomemzero(trans, bytes + 8);
 	if (!IS_ERR(mut)) {
 		bkey_reassemble(mut, k);
 

fs/bcachefs/data_update.c

@@ -639,7 +639,7 @@ int bch2_data_update_init(struct btree_trans *trans,
 
 	bch2_write_op_init(&m->op, c, io_opts);
 	m->op.pos	= bkey_start_pos(k.k);
-	m->op.version	= k.k->version;
+	m->op.version	= k.k->bversion;
 	m->op.target	= data_opts.target;
 	m->op.write_point = wp;
 	m->op.nr_replicas = 0;

fs/bcachefs/disk_accounting.c

@@ -134,6 +134,10 @@ int bch2_accounting_validate(struct bch_fs *c, struct bkey_s_c k,
 	void *end = &acc_k + 1;
 	int ret = 0;
 
+	bkey_fsck_err_on(bversion_zero(k.k->bversion),
+			 c, accounting_key_version_0,
+			 "accounting key with version=0");
+
 	switch (acc_k.type) {
 	case BCH_DISK_ACCOUNTING_nr_inodes:
 		end = field_end(acc_k, nr_inodes);
@@ -291,7 +295,7 @@ static int __bch2_accounting_mem_insert(struct bch_fs *c, struct bkey_s_c_accoun
 
 	struct accounting_mem_entry n = {
 		.pos		= a.k->p,
-		.version	= a.k->version,
+		.bversion	= a.k->bversion,
 		.nr_counters	= bch2_accounting_counters(a.k),
 		.v[0]		= __alloc_percpu_gfp(n.nr_counters * sizeof(u64),
 						     sizeof(u64), GFP_KERNEL),
@@ -319,11 +323,13 @@ static int __bch2_accounting_mem_insert(struct bch_fs *c, struct bkey_s_c_accoun
 	return -BCH_ERR_ENOMEM_disk_accounting;
 }
 
-int bch2_accounting_mem_insert(struct bch_fs *c, struct bkey_s_c_accounting a, bool gc)
+int bch2_accounting_mem_insert(struct bch_fs *c, struct bkey_s_c_accounting a,
+			       enum bch_accounting_mode mode)
 {
 	struct bch_replicas_padded r;
 
-	if (accounting_to_replicas(&r.e, a.k->p) &&
+	if (mode != BCH_ACCOUNTING_read &&
+	    accounting_to_replicas(&r.e, a.k->p) &&
 	    !bch2_replicas_marked_locked(c, &r.e))
 		return -BCH_ERR_btree_insert_need_mark_replicas;
 
@@ -566,7 +572,9 @@ int bch2_gc_accounting_done(struct bch_fs *c)
 					struct { __BKEY_PADDED(k, BCH_ACCOUNTING_MAX_COUNTERS); } k_i;
 
 					accounting_key_init(&k_i.k, &acc_k, src_v, nr);
-					bch2_accounting_mem_mod_locked(trans, bkey_i_to_s_c_accounting(&k_i.k), false, false);
+					bch2_accounting_mem_mod_locked(trans,
+								bkey_i_to_s_c_accounting(&k_i.k),
+								BCH_ACCOUNTING_normal);
 
 					preempt_disable();
 					struct bch_fs_usage_base *dst = this_cpu_ptr(c->usage);
@@ -589,30 +597,14 @@ int bch2_gc_accounting_done(struct bch_fs *c)
 static int accounting_read_key(struct btree_trans *trans, struct bkey_s_c k)
 {
 	struct bch_fs *c = trans->c;
-	struct printbuf buf = PRINTBUF;
 
 	if (k.k->type != KEY_TYPE_accounting)
 		return 0;
 
 	percpu_down_read(&c->mark_lock);
-	int ret = bch2_accounting_mem_mod_locked(trans, bkey_s_c_to_accounting(k), false, true);
+	int ret = bch2_accounting_mem_mod_locked(trans, bkey_s_c_to_accounting(k),
+						 BCH_ACCOUNTING_read);
 	percpu_up_read(&c->mark_lock);
-
-	if (bch2_accounting_key_is_zero(bkey_s_c_to_accounting(k)) &&
-	    ret == -BCH_ERR_btree_insert_need_mark_replicas)
-		ret = 0;
-
-	struct disk_accounting_pos acc;
-	bpos_to_disk_accounting_pos(&acc, k.k->p);
-
-	if (fsck_err_on(ret == -BCH_ERR_btree_insert_need_mark_replicas,
-			trans, accounting_replicas_not_marked,
-			"accounting not marked in superblock replicas\n  %s",
-			(bch2_accounting_key_to_text(&buf, &acc),
-			 buf.buf)))
-		ret = bch2_accounting_update_sb_one(c, k.k->p);
-fsck_err:
-	printbuf_exit(&buf);
 	return ret;
 }
 
@@ -624,6 +616,7 @@ int bch2_accounting_read(struct bch_fs *c)
 {
 	struct bch_accounting_mem *acc = &c->accounting;
 	struct btree_trans *trans = bch2_trans_get(c);
+	struct printbuf buf = PRINTBUF;
 
 	int ret = for_each_btree_key(trans, iter,
 				BTREE_ID_accounting, POS_MIN,
@@ -647,7 +640,7 @@ int bch2_accounting_read(struct bch_fs *c)
 						accounting_pos_cmp, &k.k->p);
 
 			bool applied = idx < acc->k.nr &&
-				bversion_cmp(acc->k.data[idx].version, k.k->version) >= 0;
+				bversion_cmp(acc->k.data[idx].bversion, k.k->bversion) >= 0;
 
 			if (applied)
 				continue;
@@ -655,7 +648,7 @@ int bch2_accounting_read(struct bch_fs *c)
 			if (i + 1 < &darray_top(*keys) &&
 			    i[1].k->k.type == KEY_TYPE_accounting &&
 			    !journal_key_cmp(i, i + 1)) {
-				BUG_ON(bversion_cmp(i[0].k->k.version, i[1].k->k.version) >= 0);
+				WARN_ON(bversion_cmp(i[0].k->k.bversion, i[1].k->k.bversion) >= 0);
 
 				i[1].journal_seq = i[0].journal_seq;
 
@@ -674,6 +667,45 @@ int bch2_accounting_read(struct bch_fs *c)
 	keys->gap = keys->nr = dst - keys->data;
 
 	percpu_down_read(&c->mark_lock);
+	for (unsigned i = 0; i < acc->k.nr; i++) {
+		u64 v[BCH_ACCOUNTING_MAX_COUNTERS];
+		bch2_accounting_mem_read_counters(acc, i, v, ARRAY_SIZE(v), false);
+
+		if (bch2_is_zero(v, sizeof(v[0]) * acc->k.data[i].nr_counters))
+			continue;
+
+		struct bch_replicas_padded r;
+		if (!accounting_to_replicas(&r.e, acc->k.data[i].pos))
+			continue;
+
+		/*
+		 * If the replicas entry is invalid it'll get cleaned up by
+		 * check_allocations:
+		 */
+		if (bch2_replicas_entry_validate(&r.e, c, &buf))
+			continue;
+
+		struct disk_accounting_pos k;
+		bpos_to_disk_accounting_pos(&k, acc->k.data[i].pos);
+
+		if (fsck_err_on(!bch2_replicas_marked_locked(c, &r.e),
+				trans, accounting_replicas_not_marked,
+				"accounting not marked in superblock replicas\n  %s",
+				(printbuf_reset(&buf),
+				 bch2_accounting_key_to_text(&buf, &k),
+				 buf.buf))) {
+			/*
+			 * We're not RW yet and still single threaded, dropping
+			 * and retaking lock is ok:
+			 */
+			percpu_up_read(&c->mark_lock);
+			ret = bch2_mark_replicas(c, &r.e);
+			if (ret)
+				goto fsck_err;
+			percpu_down_read(&c->mark_lock);
+		}
+	}
+
 	preempt_disable();
 	struct bch_fs_usage_base *usage = this_cpu_ptr(c->usage);
 
@@ -709,8 +741,10 @@ int bch2_accounting_read(struct bch_fs *c)
 		}
 	}
 	preempt_enable();
+fsck_err:
 	percpu_up_read(&c->mark_lock);
 err:
+	printbuf_exit(&buf);
 	bch2_trans_put(trans);
 	bch_err_fn(c, ret);
 	return ret;

fs/bcachefs/disk_accounting.h

@@ -36,8 +36,8 @@ static inline void bch2_accounting_accumulate(struct bkey_i_accounting *dst,
 
 	for (unsigned i = 0; i < bch2_accounting_counters(&dst->k); i++)
 		dst->v.d[i] += src.v->d[i];
-	if (bversion_cmp(dst->k.version, src.k->version) < 0)
-		dst->k.version = src.k->version;
+	if (bversion_cmp(dst->k.bversion, src.k->bversion) < 0)
+		dst->k.bversion = src.k->bversion;
 }
 
 static inline void fs_usage_data_type_to_base(struct bch_fs_usage_base *fs_usage,
@@ -103,23 +103,35 @@ static inline int accounting_pos_cmp(const void *_l, const void *_r)
 	return bpos_cmp(*l, *r);
 }
 
-int bch2_accounting_mem_insert(struct bch_fs *, struct bkey_s_c_accounting, bool);
+enum bch_accounting_mode {
+	BCH_ACCOUNTING_normal,
+	BCH_ACCOUNTING_gc,
+	BCH_ACCOUNTING_read,
+};
+
+int bch2_accounting_mem_insert(struct bch_fs *, struct bkey_s_c_accounting, enum bch_accounting_mode);
 void bch2_accounting_mem_gc(struct bch_fs *);
 
 /*
  * Update in memory counters so they match the btree update we're doing; called
  * from transaction commit path
  */
-static inline int bch2_accounting_mem_mod_locked(struct btree_trans *trans, struct bkey_s_c_accounting a, bool gc, bool read)
+static inline int bch2_accounting_mem_mod_locked(struct btree_trans *trans,
+						 struct bkey_s_c_accounting a,
+						 enum bch_accounting_mode mode)
 {
 	struct bch_fs *c = trans->c;
+	struct bch_accounting_mem *acc = &c->accounting;
 	struct disk_accounting_pos acc_k;
 	bpos_to_disk_accounting_pos(&acc_k, a.k->p);
+	bool gc = mode == BCH_ACCOUNTING_gc;
+
+	EBUG_ON(gc && !acc->gc_running);
 
 	if (acc_k.type == BCH_DISK_ACCOUNTING_inum)
 		return 0;
 
-	if (!gc && !read) {
+	if (mode == BCH_ACCOUNTING_normal) {
 		switch (acc_k.type) {
 		case BCH_DISK_ACCOUNTING_persistent_reserved:
 			trans->fs_usage_delta.reserved += acc_k.persistent_reserved.nr_replicas * a.v->d[0];
@@ -140,14 +152,11 @@ static inline int bch2_accounting_mem_mod_locked(struct btree_trans *trans, stru
 		}
 	}
 
-	struct bch_accounting_mem *acc = &c->accounting;
 	unsigned idx;
 
-	EBUG_ON(gc && !acc->gc_running);
-
 	while ((idx = eytzinger0_find(acc->k.data, acc->k.nr, sizeof(acc->k.data[0]),
 				      accounting_pos_cmp, &a.k->p)) >= acc->k.nr) {
-		int ret = bch2_accounting_mem_insert(c, a, gc);
+		int ret = bch2_accounting_mem_insert(c, a, mode);
 		if (ret)
 			return ret;
 	}
@@ -164,7 +173,7 @@ static inline int bch2_accounting_mem_mod_locked(struct btree_trans *trans, stru
 static inline int bch2_accounting_mem_add(struct btree_trans *trans, struct bkey_s_c_accounting a, bool gc)
 {
 	percpu_down_read(&trans->c->mark_lock);
-	int ret = bch2_accounting_mem_mod_locked(trans, a, gc, false);
+	int ret = bch2_accounting_mem_mod_locked(trans, a, gc ? BCH_ACCOUNTING_gc : BCH_ACCOUNTING_normal);
 	percpu_up_read(&trans->c->mark_lock);
 	return ret;
 }

fs/bcachefs/disk_accounting_types.h

@@ -6,7 +6,7 @@
 
 struct accounting_mem_entry {
 	struct bpos				pos;
-	struct bversion				version;
+	struct bversion				bversion;
 	unsigned				nr_counters;
 	u64 __percpu				*v[2];
 };

fs/bcachefs/error.c

@@ -239,7 +239,19 @@ int __bch2_fsck_err(struct bch_fs *c,
 	if (!c)
 		c = trans->c;
 
-	WARN_ON(!trans && bch2_current_has_btree_trans(c));
+	/*
+	 * Ugly: if there's a transaction in the current task it has to be
+	 * passed in to unlock if we prompt for user input.
+	 *
+	 * But, plumbing a transaction and transaction restarts into
+	 * bkey_validate() is problematic.
+	 *
+	 * So:
+	 * - make all bkey errors AUTOFIX, they're simple anyways (we just
+	 *   delete the key)
+	 * - and we don't need to warn if we're not prompting
+	 */
+	WARN_ON(!(flags & FSCK_AUTOFIX) && !trans && bch2_current_has_btree_trans(c));
 
 	if ((flags & FSCK_CAN_FIX) &&
 	    test_bit(err, c->sb.errors_silent))

fs/bcachefs/error.h

@@ -184,7 +184,7 @@ do {									\
 		ret = -BCH_ERR_fsck_delete_bkey;			\
 		goto fsck_err;						\
 	}								\
-	int _ret = __bch2_bkey_fsck_err(c, k, FSCK_CAN_FIX,		\
+	int _ret = __bch2_bkey_fsck_err(c, k, FSCK_CAN_FIX|FSCK_AUTOFIX,\
 				BCH_FSCK_ERR_##_err_type,		\
 				_err_msg, ##__VA_ARGS__);		\
 	if (_ret != -BCH_ERR_fsck_fix &&				\

fs/bcachefs/fsck.c

@@ -21,6 +21,49 @@
 #include <linux/bsearch.h>
 #include <linux/dcache.h> /* struct qstr */
 
+static bool inode_points_to_dirent(struct bch_inode_unpacked *inode,
+				   struct bkey_s_c_dirent d)
+{
+	return  inode->bi_dir		== d.k->p.inode &&
+		inode->bi_dir_offset	== d.k->p.offset;
+}
+
+static bool dirent_points_to_inode_nowarn(struct bkey_s_c_dirent d,
+				   struct bch_inode_unpacked *inode)
+{
+	if (d.v->d_type == DT_SUBVOL
+	    ? le32_to_cpu(d.v->d_child_subvol)	== inode->bi_subvol
+	    : le64_to_cpu(d.v->d_inum)		== inode->bi_inum)
+		return 0;
+	return -BCH_ERR_ENOENT_dirent_doesnt_match_inode;
+}
+
+static void dirent_inode_mismatch_msg(struct printbuf *out,
+				      struct bch_fs *c,
+				      struct bkey_s_c_dirent dirent,
+				      struct bch_inode_unpacked *inode)
+{
+	prt_str(out, "inode points to dirent that does not point back:");
+	prt_newline(out);
+	bch2_bkey_val_to_text(out, c, dirent.s_c);
+	prt_newline(out);
+	bch2_inode_unpacked_to_text(out, inode);
+}
+
+static int dirent_points_to_inode(struct bch_fs *c,
+				  struct bkey_s_c_dirent dirent,
+				  struct bch_inode_unpacked *inode)
+{
+	int ret = dirent_points_to_inode_nowarn(dirent, inode);
+	if (ret) {
+		struct printbuf buf = PRINTBUF;
+		dirent_inode_mismatch_msg(&buf, c, dirent, inode);
+		bch_warn(c, "%s", buf.buf);
+		printbuf_exit(&buf);
+	}
+	return ret;
+}
+
 /*
  * XXX: this is handling transaction restarts without returning
  * -BCH_ERR_transaction_restart_nested, this is not how we do things anymore:
@@ -346,14 +389,17 @@ static int reattach_inode(struct btree_trans *trans,
 static int remove_backpointer(struct btree_trans *trans,
 			      struct bch_inode_unpacked *inode)
 {
-	struct btree_iter iter;
-	struct bkey_s_c_dirent d;
-	int ret;
+	if (!inode->bi_dir)
+		return 0;
 
-	d = bch2_bkey_get_iter_typed(trans, &iter, BTREE_ID_dirents,
-				     POS(inode->bi_dir, inode->bi_dir_offset), 0,
+	struct bch_fs *c = trans->c;
+	struct btree_iter iter;
+	struct bkey_s_c_dirent d =
+		bch2_bkey_get_iter_typed(trans, &iter, BTREE_ID_dirents,
+				     SPOS(inode->bi_dir, inode->bi_dir_offset, inode->bi_snapshot), 0,
 				     dirent);
-	ret =   bkey_err(d) ?:
+	int ret =   bkey_err(d) ?:
+		dirent_points_to_inode(c, d, inode) ?:
 		__remove_dirent(trans, d.k->p);
 	bch2_trans_iter_exit(trans, &iter);
 	return ret;
@@ -371,6 +417,7 @@ static int reattach_subvol(struct btree_trans *trans, struct bkey_s_c_subvolume
 		return ret;
 
 	ret = remove_backpointer(trans, &inode);
+	if (!bch2_err_matches(ret, ENOENT))
 		bch_err_msg(c, ret, "removing dirent");
 	if (ret)
 		return ret;
@@ -626,12 +673,12 @@ static int ref_visible2(struct bch_fs *c,
 struct inode_walker_entry {
 	struct bch_inode_unpacked inode;
 	u32			snapshot;
-	bool			seen_this_pos;
 	u64			count;
 };
 
 struct inode_walker {
 	bool				first_this_inode;
+	bool				have_inodes;
 	bool				recalculate_sums;
 	struct bpos			last_pos;
 
@@ -669,6 +716,12 @@ static int get_inodes_all_snapshots(struct btree_trans *trans,
 	struct bkey_s_c k;
 	int ret;
 
+	/*
+	 * We no longer have inodes for w->last_pos; clear this to avoid
+	 * screwing up check_i_sectors/check_subdir_count if we take a
+	 * transaction restart here:
+	 */
+	w->have_inodes = false;
 	w->recalculate_sums = false;
 	w->inodes.nr = 0;
 
@@ -686,6 +739,7 @@ static int get_inodes_all_snapshots(struct btree_trans *trans,
 		return ret;
 
 	w->first_this_inode = true;
+	w->have_inodes = true;
 	return 0;
 }
 
@@ -740,9 +794,6 @@ static struct inode_walker_entry *walk_inode(struct btree_trans *trans,
 		int ret = get_inodes_all_snapshots(trans, w, k.k->p.inode);
 		if (ret)
 			return ERR_PTR(ret);
-	} else if (bkey_cmp(w->last_pos, k.k->p)) {
-		darray_for_each(w->inodes, i)
-			i->seen_this_pos = false;
 	}
 
 	w->last_pos = k.k->p;
@@ -896,21 +947,6 @@ static struct bkey_s_c_dirent inode_get_dirent(struct btree_trans *trans,
 	return dirent_get_by_pos(trans, iter, SPOS(inode->bi_dir, inode->bi_dir_offset, *snapshot));
 }
 
-static bool inode_points_to_dirent(struct bch_inode_unpacked *inode,
-				   struct bkey_s_c_dirent d)
-{
-	return  inode->bi_dir		== d.k->p.inode &&
-		inode->bi_dir_offset	== d.k->p.offset;
-}
-
-static bool dirent_points_to_inode(struct bkey_s_c_dirent d,
-				   struct bch_inode_unpacked *inode)
-{
-	return d.v->d_type == DT_SUBVOL
-		? le32_to_cpu(d.v->d_child_subvol)	== inode->bi_subvol
-		: le64_to_cpu(d.v->d_inum)		== inode->bi_inum;
-}
-
 static int check_inode_deleted_list(struct btree_trans *trans, struct bpos p)
 {
 	struct btree_iter iter;
@@ -920,13 +956,14 @@ static int check_inode_deleted_list(struct btree_trans *trans, struct bpos p)
 	return ret;
 }
 
-static int check_inode_dirent_inode(struct btree_trans *trans, struct bkey_s_c inode_k,
+static int check_inode_dirent_inode(struct btree_trans *trans,
 				    struct bch_inode_unpacked *inode,
-				    u32 inode_snapshot, bool *write_inode)
+				    bool *write_inode)
 {
 	struct bch_fs *c = trans->c;
 	struct printbuf buf = PRINTBUF;
 
+	u32 inode_snapshot = inode->bi_snapshot;
 	struct btree_iter dirent_iter = {};
 	struct bkey_s_c_dirent d = inode_get_dirent(trans, &dirent_iter, inode, &inode_snapshot);
 	int ret = bkey_err(d);
@@ -936,13 +973,13 @@ static int check_inode_dirent_inode(struct btree_trans *trans, struct bkey_s_c i
 	if (fsck_err_on(ret,
 			trans, inode_points_to_missing_dirent,
 			"inode points to missing dirent\n%s",
-			(bch2_bkey_val_to_text(&buf, c, inode_k), buf.buf)) ||
-	    fsck_err_on(!ret && !dirent_points_to_inode(d, inode),
+			(bch2_inode_unpacked_to_text(&buf, inode), buf.buf)) ||
+	    fsck_err_on(!ret && dirent_points_to_inode_nowarn(d, inode),
 			trans, inode_points_to_wrong_dirent,
-			"inode points to dirent that does not point back:\n%s",
-			(bch2_bkey_val_to_text(&buf, c, inode_k),
-			 prt_newline(&buf),
-			 bch2_bkey_val_to_text(&buf, c, d.s_c), buf.buf))) {
+			"%s",
+			(printbuf_reset(&buf),
+			 dirent_inode_mismatch_msg(&buf, c, d, inode),
+			 buf.buf))) {
 		/*
 		 * We just clear the backpointer fields for now. If we find a
 		 * dirent that points to this inode in check_dirents(), we'll
@@ -963,7 +1000,7 @@ static int check_inode_dirent_inode(struct btree_trans *trans, struct bkey_s_c i
 	return ret;
 }
 
-static bool bch2_inode_open(struct bch_fs *c, struct bpos p)
+static bool bch2_inode_is_open(struct bch_fs *c, struct bpos p)
 {
 	subvol_inum inum = {
 		.subvol = snapshot_t(c, p.snapshot)->subvol,
@@ -972,7 +1009,7 @@ static bool bch2_inode_open(struct bch_fs *c, struct bpos p)
 
 	/* snapshot tree corruption, can't safely delete */
 	if (!inum.subvol) {
-		bch_err_ratelimited(c, "%s(): snapshot %u has no subvol", __func__, p.snapshot);
+		bch_warn_ratelimited(c, "%s(): snapshot %u has no subvol, unlinked but can't safely delete", __func__, p.snapshot);
 		return true;
 	}
 
@@ -1045,6 +1082,17 @@ static int check_inode(struct btree_trans *trans,
 	}
 
 	if (u.bi_flags & BCH_INODE_unlinked) {
+		if (!test_bit(BCH_FS_started, &c->flags)) {
+			/*
+			 * If we're not in online fsck, don't delete unlinked
+			 * inodes, just make sure they're on the deleted list.
+			 *
+			 * They might be referred to by a logged operation -
+			 * i.e. we might have crashed in the middle of a
+			 * truncate on an unlinked but open file - so we want to
+			 * let the delete_dead_inodes kill it after resuming
+			 * logged ops.
+			 */
 			ret = check_inode_deleted_list(trans, k.k->p);
 			if (ret < 0)
 				return ret;
@@ -1053,22 +1101,25 @@ static int check_inode(struct btree_trans *trans,
 				    trans, unlinked_inode_not_on_deleted_list,
 				    "inode %llu:%u unlinked, but not on deleted list",
 				    u.bi_inum, k.k->p.snapshot);
-		ret = 0;
-	}
 
-	if (u.bi_flags & BCH_INODE_unlinked &&
-	    !bch2_inode_open(c, k.k->p) &&
-	    (!c->sb.clean ||
-	     fsck_err(trans, inode_unlinked_but_clean,
-		      "filesystem marked clean, but inode %llu unlinked",
-		      u.bi_inum))) {
+			ret = bch2_btree_bit_mod_buffered(trans, BTREE_ID_deleted_inodes, k.k->p, 1);
+			if (ret)
+				goto err;
+		} else {
+			if (fsck_err_on(bch2_inode_is_open(c, k.k->p),
+					trans, inode_unlinked_and_not_open,
+				      "inode %llu%u unlinked and not open",
+				      u.bi_inum, u.bi_snapshot)) {
 				ret = bch2_inode_rm_snapshot(trans, u.bi_inum, iter->pos.snapshot);
 				bch_err_msg(c, ret, "in fsck deleting inode");
 				return ret;
 			}
+		}
+	}
 
+	/* i_size_dirty is vestigal, since we now have logged ops for truncate * */
 	if (u.bi_flags & BCH_INODE_i_size_dirty &&
-	    (!c->sb.clean ||
+	    (!test_bit(BCH_FS_clean_recovery, &c->flags) ||
 	     fsck_err(trans, inode_i_size_dirty_but_clean,
 		      "filesystem marked clean, but inode %llu has i_size dirty",
 		      u.bi_inum))) {
@@ -1097,8 +1148,9 @@ static int check_inode(struct btree_trans *trans,
 		do_update = true;
 	}
 
+	/* i_sectors_dirty is vestigal, i_sectors is always updated transactionally */
 	if (u.bi_flags & BCH_INODE_i_sectors_dirty &&
-	    (!c->sb.clean ||
+	    (!test_bit(BCH_FS_clean_recovery, &c->flags) ||
 	     fsck_err(trans, inode_i_sectors_dirty_but_clean,
 		      "filesystem marked clean, but inode %llu has i_sectors dirty",
 		      u.bi_inum))) {
@@ -1126,7 +1178,7 @@ static int check_inode(struct btree_trans *trans,
 	}
 
 	if (u.bi_dir || u.bi_dir_offset) {
-		ret = check_inode_dirent_inode(trans, k, &u, k.k->p.snapshot, &do_update);
+		ret = check_inode_dirent_inode(trans, &u, &do_update);
 		if (ret)
 			goto err;
 	}
@@ -1555,10 +1607,10 @@ static int check_extent(struct btree_trans *trans, struct btree_iter *iter,
 			struct bkey_s_c k,
 			struct inode_walker *inode,
 			struct snapshots_seen *s,
-			struct extent_ends *extent_ends)
+			struct extent_ends *extent_ends,
+			struct disk_reservation *res)
 {
 	struct bch_fs *c = trans->c;
-	struct inode_walker_entry *i;
 	struct printbuf buf = PRINTBUF;
 	int ret = 0;
 
@@ -1568,7 +1620,7 @@ static int check_extent(struct btree_trans *trans, struct btree_iter *iter,
 		goto out;
 	}
 
-	if (inode->last_pos.inode != k.k->p.inode) {
+	if (inode->last_pos.inode != k.k->p.inode && inode->have_inodes) {
 		ret = check_i_sectors(trans, inode);
 		if (ret)
 			goto err;
@@ -1578,12 +1630,12 @@ static int check_extent(struct btree_trans *trans, struct btree_iter *iter,
 	if (ret)
 		goto err;
 
-	i = walk_inode(trans, inode, k);
-	ret = PTR_ERR_OR_ZERO(i);
+	struct inode_walker_entry *extent_i = walk_inode(trans, inode, k);
+	ret = PTR_ERR_OR_ZERO(extent_i);
 	if (ret)
 		goto err;
 
-	ret = check_key_has_inode(trans, iter, inode, i, k);
+	ret = check_key_has_inode(trans, iter, inode, extent_i, k);
 	if (ret)
 		goto err;
 
@@ -1592,24 +1644,19 @@ static int check_extent(struct btree_trans *trans, struct btree_iter *iter,
 						&inode->recalculate_sums);
 		if (ret)
 			goto err;
-	}
 
 		/*
-	 * Check inodes in reverse order, from oldest snapshots to newest,
-	 * starting from the inode that matches this extent's snapshot. If we
-	 * didn't have one, iterate over all inodes:
+		 * Check inodes in reverse order, from oldest snapshots to
+		 * newest, starting from the inode that matches this extent's
+		 * snapshot. If we didn't have one, iterate over all inodes:
 		 */
-	if (!i)
-		i = &darray_last(inode->inodes);
-
-	for (;
+		for (struct inode_walker_entry *i = extent_i ?: &darray_last(inode->inodes);
 		     inode->inodes.data && i >= inode->inodes.data;
 		     --i) {
 			if (i->snapshot > k.k->p.snapshot ||
 			    !key_visible_in_snapshot(c, s, i->snapshot, k.k->p.snapshot))
 				continue;
 
-		if (k.k->type != KEY_TYPE_whiteout) {
 			if (fsck_err_on(!(i->inode.bi_flags & BCH_INODE_i_size_dirty) &&
 					k.k->p.offset > round_up(i->inode.bi_size, block_bytes(c)) >> 9 &&
 					!bkey_extent_is_reservation(k),
@@ -1629,13 +1676,25 @@ static int check_extent(struct btree_trans *trans, struct btree_iter *iter,
 					goto err;
 
 				iter->k.type = KEY_TYPE_whiteout;
+				break;
+			}
+		}
 	}
 
-			if (bkey_extent_is_allocation(k.k))
+	ret = bch2_trans_commit(trans, res, NULL, BCH_TRANS_COMMIT_no_enospc);
+	if (ret)
+		goto err;
+
+	if (bkey_extent_is_allocation(k.k)) {
+		for (struct inode_walker_entry *i = extent_i ?: &darray_last(inode->inodes);
+		     inode->inodes.data && i >= inode->inodes.data;
+		     --i) {
+			if (i->snapshot > k.k->p.snapshot ||
+			    !key_visible_in_snapshot(c, s, i->snapshot, k.k->p.snapshot))
+				continue;
+
 			i->count += k.k->size;
 		}
-
-		i->seen_this_pos = true;
 	}
 
 	if (k.k->type != KEY_TYPE_whiteout) {
@@ -1666,13 +1725,11 @@ int bch2_check_extents(struct bch_fs *c)
 	extent_ends_init(&extent_ends);
 
 	int ret = bch2_trans_run(c,
-		for_each_btree_key_commit(trans, iter, BTREE_ID_extents,
+		for_each_btree_key(trans, iter, BTREE_ID_extents,
 				POS(BCACHEFS_ROOT_INO, 0),
-				BTREE_ITER_prefetch|BTREE_ITER_all_snapshots, k,
-				&res, NULL,
-				BCH_TRANS_COMMIT_no_enospc, ({
+				BTREE_ITER_prefetch|BTREE_ITER_all_snapshots, k, ({
 			bch2_disk_reservation_put(c, &res);
-			check_extent(trans, &iter, k, &w, &s, &extent_ends) ?:
+			check_extent(trans, &iter, k, &w, &s, &extent_ends, &res) ?:
 			check_extent_overbig(trans, &iter, k);
 		})) ?:
 		check_i_sectors_notnested(trans, &w));
@@ -1758,6 +1815,7 @@ static int check_dirent_inode_dirent(struct btree_trans *trans,
 {
 	struct bch_fs *c = trans->c;
 	struct printbuf buf = PRINTBUF;
+	struct btree_iter bp_iter = { NULL };
 	int ret = 0;
 
 	if (inode_points_to_dirent(target, d))
@@ -1770,7 +1828,7 @@ static int check_dirent_inode_dirent(struct btree_trans *trans,
 		       prt_printf(&buf, "\n  "),
 		       bch2_inode_unpacked_to_text(&buf, target),
 		       buf.buf)))
-		goto out_noiter;
+		goto err;
 
 	if (!target->bi_dir &&
 	    !target->bi_dir_offset) {
@@ -1779,7 +1837,6 @@ static int check_dirent_inode_dirent(struct btree_trans *trans,
 		return __bch2_fsck_write_inode(trans, target, target_snapshot);
 	}
 
-	struct btree_iter bp_iter = { NULL };
 	struct bkey_s_c_dirent bp_dirent = dirent_get_by_pos(trans, &bp_iter,
 			      SPOS(target->bi_dir, target->bi_dir_offset, target_snapshot));
 	ret = bkey_err(bp_dirent);
@@ -1840,7 +1897,6 @@ static int check_dirent_inode_dirent(struct btree_trans *trans,
 err:
 fsck_err:
 	bch2_trans_iter_exit(trans, &bp_iter);
-out_noiter:
 	printbuf_exit(&buf);
 	bch_err_fn(c, ret);
 	return ret;
@@ -2075,7 +2131,7 @@ static int check_dirent(struct btree_trans *trans, struct btree_iter *iter,
 	if (k.k->type == KEY_TYPE_whiteout)
 		goto out;
 
-	if (dir->last_pos.inode != k.k->p.inode) {
+	if (dir->last_pos.inode != k.k->p.inode && dir->have_inodes) {
 		ret = check_subdir_count(trans, dir);
 		if (ret)
 			goto err;
@@ -2137,11 +2193,15 @@ static int check_dirent(struct btree_trans *trans, struct btree_iter *iter,
 			if (ret)
 				goto err;
 		}
+	}
+
+	ret = bch2_trans_commit(trans, NULL, NULL, BCH_TRANS_COMMIT_no_enospc);
+	if (ret)
+		goto err;
 
 	if (d.v->d_type == DT_DIR)
 		for_each_visible_inode(c, s, dir, d.k->p.snapshot, i)
 			i->count++;
-	}
 out:
 err:
 fsck_err:
@@ -2164,12 +2224,9 @@ int bch2_check_dirents(struct bch_fs *c)
 	snapshots_seen_init(&s);
 
 	int ret = bch2_trans_run(c,
-		for_each_btree_key_commit(trans, iter, BTREE_ID_dirents,
+		for_each_btree_key(trans, iter, BTREE_ID_dirents,
 				POS(BCACHEFS_ROOT_INO, 0),
-				BTREE_ITER_prefetch|BTREE_ITER_all_snapshots,
-				k,
-				NULL, NULL,
-				BCH_TRANS_COMMIT_no_enospc,
+				BTREE_ITER_prefetch|BTREE_ITER_all_snapshots, k,
 			check_dirent(trans, &iter, k, &hash_info, &dir, &target, &s)) ?:
 		check_subdir_count_notnested(trans, &dir));
 
@@ -2314,22 +2371,6 @@ static bool darray_u32_has(darray_u32 *d, u32 v)
 	return false;
 }
 
-/*
- * We've checked that inode backpointers point to valid dirents; here, it's
- * sufficient to check that the subvolume root has a dirent:
- */
-static int subvol_has_dirent(struct btree_trans *trans, struct bkey_s_c_subvolume s)
-{
-	struct bch_inode_unpacked inode;
-	int ret = bch2_inode_find_by_inum_trans(trans,
-				(subvol_inum) { s.k->p.offset, le64_to_cpu(s.v->inode) },
-				&inode);
-	if (ret)
-		return ret;
-
-	return inode.bi_dir != 0;
-}
-
 static int check_subvol_path(struct btree_trans *trans, struct btree_iter *iter, struct bkey_s_c k)
 {
 	struct bch_fs *c = trans->c;
@@ -2348,14 +2389,24 @@ static int check_subvol_path(struct btree_trans *trans, struct btree_iter *iter,
 
 		struct bkey_s_c_subvolume s = bkey_s_c_to_subvolume(k);
 
-		ret = subvol_has_dirent(trans, s);
-		if (ret < 0)
+		struct bch_inode_unpacked subvol_root;
+		ret = bch2_inode_find_by_inum_trans(trans,
+					(subvol_inum) { s.k->p.offset, le64_to_cpu(s.v->inode) },
+					&subvol_root);
+		if (ret)
 			break;
 
-		if (fsck_err_on(!ret,
+		/*
+		 * We've checked that inode backpointers point to valid dirents;
+		 * here, it's sufficient to check that the subvolume root has a
+		 * dirent:
+		 */
+		if (fsck_err_on(!subvol_root.bi_dir,
 				trans, subvol_unreachable,
 				"unreachable subvolume %s",
 				(bch2_bkey_val_to_text(&buf, c, s.s_c),
+				 prt_newline(&buf),
+				 bch2_inode_unpacked_to_text(&buf, &subvol_root),
 				 buf.buf))) {
 			ret = reattach_subvol(trans, s);
 			break;
@@ -2450,10 +2501,8 @@ static int check_path(struct btree_trans *trans, pathbuf *p, struct bkey_s_c ino
 		if (ret && !bch2_err_matches(ret, ENOENT))
 			break;
 
-		if (!ret && !dirent_points_to_inode(d, &inode)) {
+		if (!ret && (ret = dirent_points_to_inode(c, d, &inode)))
 			bch2_trans_iter_exit(trans, &dirent_iter);
-			ret = -BCH_ERR_ENOENT_dirent_doesnt_match_inode;
-		}
 
 		if (bch2_err_matches(ret, ENOENT)) {
 			ret = 0;

fs/bcachefs/inode.c

@@ -320,9 +320,11 @@ static noinline int bch2_inode_unpack_slowpath(struct bkey_s_c k,
 int bch2_inode_unpack(struct bkey_s_c k,
 		      struct bch_inode_unpacked *unpacked)
 {
-	if (likely(k.k->type == KEY_TYPE_inode_v3))
-		return bch2_inode_unpack_v3(k, unpacked);
-	return bch2_inode_unpack_slowpath(k, unpacked);
+	unpacked->bi_snapshot = k.k->p.snapshot;
+
+	return likely(k.k->type == KEY_TYPE_inode_v3)
+		? bch2_inode_unpack_v3(k, unpacked)
+		: bch2_inode_unpack_slowpath(k, unpacked);
 }
 
 int bch2_inode_peek_nowarn(struct btree_trans *trans,
@@ -557,7 +559,7 @@ static void __bch2_inode_unpacked_to_text(struct printbuf *out,
 
 void bch2_inode_unpacked_to_text(struct printbuf *out, struct bch_inode_unpacked *inode)
 {
-	prt_printf(out, "inum: %llu ", inode->bi_inum);
+	prt_printf(out, "inum: %llu:%u ", inode->bi_inum, inode->bi_snapshot);
 	__bch2_inode_unpacked_to_text(out, inode);
 }
 
@@ -1111,7 +1113,7 @@ static int may_delete_deleted_inode(struct btree_trans *trans,
 			pos.offset, pos.snapshot))
 		goto delete;
 
-	if (c->sb.clean &&
+	if (test_bit(BCH_FS_clean_recovery, &c->flags) &&
 	    !fsck_err(trans, deleted_inode_but_clean,
 		      "filesystem marked as clean but have deleted inode %llu:%u",
 		      pos.offset, pos.snapshot)) {

fs/bcachefs/inode.h

@@ -69,6 +69,7 @@ typedef u64 u96;
 
 struct bch_inode_unpacked {
 	u64			bi_inum;
+	u32			bi_snapshot;
 	u64			bi_journal_seq;
 	__le64			bi_hash_seed;
 	u64			bi_size;

fs/bcachefs/io_read.c

@@ -517,7 +517,7 @@ static int __bch2_rbio_narrow_crcs(struct btree_trans *trans,
 	if ((ret = bkey_err(k)))
 		goto out;
 
-	if (bversion_cmp(k.k->version, rbio->version) ||
+	if (bversion_cmp(k.k->bversion, rbio->version) ||
 	    !bch2_bkey_matches_ptr(c, k, rbio->pick.ptr, data_offset))
 		goto out;
 
@@ -1031,7 +1031,7 @@ int __bch2_read_extent(struct btree_trans *trans, struct bch_read_bio *orig,
 	rbio->read_pos		= read_pos;
 	rbio->data_btree	= data_btree;
 	rbio->data_pos		= data_pos;
-	rbio->version		= k.k->version;
+	rbio->version		= k.k->bversion;
 	rbio->promote		= promote;
 	INIT_WORK(&rbio->work, NULL);
 

fs/bcachefs/io_write.c

@@ -697,7 +697,7 @@ static void init_append_extent(struct bch_write_op *op,
 	e = bkey_extent_init(op->insert_keys.top);
 	e->k.p		= op->pos;
 	e->k.size	= crc.uncompressed_size;
-	e->k.version	= version;
+	e->k.bversion	= version;
 
 	if (crc.csum_type ||
 	    crc.compression_type ||
@@ -1544,7 +1544,7 @@ static void bch2_write_data_inline(struct bch_write_op *op, unsigned data_len)
 
 	id = bkey_inline_data_init(op->insert_keys.top);
 	id->k.p		= op->pos;
-	id->k.version	= op->version;
+	id->k.bversion	= op->version;
 	id->k.size	= sectors;
 
 	iter = bio->bi_iter;

fs/bcachefs/journal_io.c

@@ -605,7 +605,7 @@ static int journal_entry_data_usage_validate(struct bch_fs *c,
 		goto out;
 	}
 
-	if (journal_entry_err_on(bch2_replicas_entry_validate(&u->r, c->disk_sb.sb, &err),
+	if (journal_entry_err_on(bch2_replicas_entry_validate(&u->r, c, &err),
 				 c, version, jset, entry,
 				 journal_entry_data_usage_bad_size,
 				 "invalid journal entry usage: %s", err.buf)) {

fs/bcachefs/logged_ops.c

@@ -37,6 +37,14 @@ static int resume_logged_op(struct btree_trans *trans, struct btree_iter *iter,
 	const struct bch_logged_op_fn *fn = logged_op_fn(k.k->type);
 	struct bkey_buf sk;
 	u32 restart_count = trans->restart_count;
+	struct printbuf buf = PRINTBUF;
+	int ret = 0;
+
+	fsck_err_on(test_bit(BCH_FS_clean_recovery, &c->flags),
+		    trans, logged_op_but_clean,
+		    "filesystem marked as clean but have logged op\n%s",
+		    (bch2_bkey_val_to_text(&buf, c, k),
+		     buf.buf));
 
 	if (!fn)
 		return 0;
@@ -47,8 +55,9 @@ static int resume_logged_op(struct btree_trans *trans, struct btree_iter *iter,
 	fn->resume(trans, sk.k);
 
 	bch2_bkey_buf_exit(&sk, c);
-
-	return trans_was_restarted(trans, restart_count);
+fsck_err:
+	printbuf_exit(&buf);
+	return ret ?: trans_was_restarted(trans, restart_count);
 }
 
 int bch2_resume_logged_ops(struct bch_fs *c)

fs/bcachefs/recovery.c

@@ -151,7 +151,7 @@ static int bch2_journal_replay_accounting_key(struct btree_trans *trans,
 	struct bkey_s_c old = bch2_btree_path_peek_slot(btree_iter_path(trans, &iter), &u);
 
 	/* Has this delta already been applied to the btree? */
-	if (bversion_cmp(old.k->version, k->k->k.version) >= 0) {
+	if (bversion_cmp(old.k->bversion, k->k->k.bversion) >= 0) {
 		ret = 0;
 		goto out;
 	}
@@ -717,6 +717,8 @@ int bch2_fs_recovery(struct bch_fs *c)
 
 	if (c->opts.fsck)
 		set_bit(BCH_FS_fsck_running, &c->flags);
+	if (c->sb.clean)
+		set_bit(BCH_FS_clean_recovery, &c->flags);
 
 	ret = bch2_blacklist_table_initialize(c);
 	if (ret) {
@@ -862,6 +864,9 @@ int bch2_fs_recovery(struct bch_fs *c)
 
 	clear_bit(BCH_FS_fsck_running, &c->flags);
 
+	/* in case we don't run journal replay, i.e. norecovery mode */
+	set_bit(BCH_FS_accounting_replay_done, &c->flags);
+
 	/* fsync if we fixed errors */
 	if (test_bit(BCH_FS_errors_fixed, &c->flags) &&
 	    bch2_write_ref_tryget(c, BCH_WRITE_REF_fsync)) {

fs/bcachefs/recovery_passes_types.h

@@ -50,7 +50,7 @@
 	x(check_directory_structure,		30, PASS_ONLINE|PASS_FSCK)	\
 	x(check_nlinks,				31, PASS_FSCK)			\
 	x(resume_logged_ops,			23, PASS_ALWAYS)		\
-	x(delete_dead_inodes,			32, PASS_FSCK|PASS_UNCLEAN)	\
+	x(delete_dead_inodes,			32, PASS_ALWAYS)		\
 	x(fix_reflink_p,			33, 0)				\
 	x(set_fs_needs_rebalance,		34, 0)				\
 

fs/bcachefs/reflink.c

@@ -367,7 +367,7 @@ static int bch2_make_extent_indirect(struct btree_trans *trans,
 	r_v->k.type	= bkey_type_to_indirect(&orig->k);
 	r_v->k.p	= reflink_iter.pos;
 	bch2_key_resize(&r_v->k, orig->k.size);
-	r_v->k.version	= orig->k.version;
+	r_v->k.bversion	= orig->k.bversion;
 
 	set_bkey_val_bytes(&r_v->k, sizeof(__le64) + bkey_val_bytes(&orig->k));
 

fs/bcachefs/replicas.c

@@ -66,7 +66,7 @@ void bch2_replicas_entry_to_text(struct printbuf *out,
 	prt_printf(out, "]");
 }
 
-int bch2_replicas_entry_validate(struct bch_replicas_entry_v1 *r,
+static int bch2_replicas_entry_validate_locked(struct bch_replicas_entry_v1 *r,
 					       struct bch_sb *sb,
 					       struct printbuf *err)
 {
@@ -94,6 +94,16 @@ int bch2_replicas_entry_validate(struct bch_replicas_entry_v1 *r,
 	return -BCH_ERR_invalid_replicas_entry;
 }
 
+int bch2_replicas_entry_validate(struct bch_replicas_entry_v1 *r,
+				 struct bch_fs *c,
+				 struct printbuf *err)
+{
+	mutex_lock(&c->sb_lock);
+	int ret = bch2_replicas_entry_validate_locked(r, c->disk_sb.sb, err);
+	mutex_unlock(&c->sb_lock);
+	return ret;
+}
+
 void bch2_cpu_replicas_to_text(struct printbuf *out,
 			       struct bch_replicas_cpu *r)
 {
@@ -676,7 +686,7 @@ static int bch2_cpu_replicas_validate(struct bch_replicas_cpu *cpu_r,
 		struct bch_replicas_entry_v1 *e =
 			cpu_replicas_entry(cpu_r, i);
 
-		int ret = bch2_replicas_entry_validate(e, sb, err);
+		int ret = bch2_replicas_entry_validate_locked(e, sb, err);
 		if (ret)
 			return ret;
 

fs/bcachefs/replicas.h

@@ -10,7 +10,7 @@ void bch2_replicas_entry_sort(struct bch_replicas_entry_v1 *);
 void bch2_replicas_entry_to_text(struct printbuf *,
 				 struct bch_replicas_entry_v1 *);
 int bch2_replicas_entry_validate(struct bch_replicas_entry_v1 *,
-				 struct bch_sb *, struct printbuf *);
+				 struct bch_fs *, struct printbuf *);
 void bch2_cpu_replicas_to_text(struct printbuf *, struct bch_replicas_cpu *);
 
 static inline struct bch_replicas_entry_v1 *

fs/bcachefs/sb-clean.c

@@ -167,6 +167,7 @@ struct bch_sb_field_clean *bch2_read_superblock_clean(struct bch_fs *c)
 
 	ret = bch2_sb_clean_validate_late(c, clean, READ);
 	if (ret) {
+		kfree(clean);
 		mutex_unlock(&c->sb_lock);
 		return ERR_PTR(ret);
 	}

fs/bcachefs/sb-downgrade.c

@@ -312,8 +312,7 @@ static void bch2_sb_downgrade_to_text(struct printbuf *out, struct bch_sb *sb,
 			if (!first)
 				prt_char(out, ',');
 			first = false;
-			unsigned e = le16_to_cpu(i->errors[j]);
-			prt_str(out, e < BCH_SB_ERR_MAX ? bch2_sb_error_strs[e] : "(unknown)");
+			bch2_sb_error_id_to_text(out, le16_to_cpu(i->errors[j]));
 		}
 		prt_newline(out);
 	}
@@ -353,7 +352,9 @@ int bch2_sb_downgrade_update(struct bch_fs *c)
 		for (unsigned i = 0; i < src->nr_errors; i++)
 			dst->errors[i] = cpu_to_le16(src->errors[i]);
 
-		downgrade_table_extra(c, &table);
+		ret = downgrade_table_extra(c, &table);
+		if (ret)
+			goto out;
 
 		if (!dst->recovery_passes[0] &&
 		    !dst->recovery_passes[1] &&
@@ -399,7 +400,7 @@ void bch2_sb_set_downgrade(struct bch_fs *c, unsigned new_minor, unsigned old_mi
 
 			for (unsigned j = 0; j < le16_to_cpu(i->nr_errors); j++) {
 				unsigned e = le16_to_cpu(i->errors[j]);
-				if (e < BCH_SB_ERR_MAX)
+				if (e < BCH_FSCK_ERR_MAX)
 					__set_bit(e, c->sb.errors_silent);
 				if (e < sizeof(ext->errors_silent) * 8)
 					__set_bit_le64(e, ext->errors_silent);

fs/bcachefs/sb-errors.c

@@ -7,12 +7,12 @@
 const char * const bch2_sb_error_strs[] = {
 #define x(t, n, ...) [n] = #t,
 	BCH_SB_ERRS()
-	NULL
+#undef x
 };
 
-static void bch2_sb_error_id_to_text(struct printbuf *out, enum bch_sb_error_id id)
+void bch2_sb_error_id_to_text(struct printbuf *out, enum bch_sb_error_id id)
 {
-	if (id < BCH_SB_ERR_MAX)
+	if (id < BCH_FSCK_ERR_MAX)
 		prt_str(out, bch2_sb_error_strs[id]);
 	else
 		prt_printf(out, "(unknown error %u)", id);

fs/bcachefs/sb-errors.h

@@ -6,6 +6,8 @@
 
 extern const char * const bch2_sb_error_strs[];
 
+void bch2_sb_error_id_to_text(struct printbuf *, enum bch_sb_error_id);
+
 extern const struct bch_sb_field_ops bch_sb_field_ops_errors;
 
 void bch2_sb_error_count(struct bch_fs *, enum bch_sb_error_id);

fs/bcachefs/sb-errors_format.h

@@ -210,22 +210,23 @@ enum bch_fsck_flags {
 	x(inode_snapshot_mismatch,				196,	0)		\
 	x(inode_unlinked_but_clean,				197,	0)		\
 	x(inode_unlinked_but_nlink_nonzero,			198,	0)		\
+	x(inode_unlinked_and_not_open,				281,	0)		\
 	x(inode_checksum_type_invalid,				199,	0)		\
 	x(inode_compression_type_invalid,			200,	0)		\
 	x(inode_subvol_root_but_not_dir,			201,	0)		\
-	x(inode_i_size_dirty_but_clean,				202,	0)		\
-	x(inode_i_sectors_dirty_but_clean,			203,	0)		\
-	x(inode_i_sectors_wrong,				204,	0)		\
-	x(inode_dir_wrong_nlink,				205,	0)		\
-	x(inode_dir_multiple_links,				206,	0)		\
-	x(inode_multiple_links_but_nlink_0,			207,	0)		\
-	x(inode_wrong_backpointer,				208,	0)		\
-	x(inode_wrong_nlink,					209,	0)		\
-	x(inode_unreachable,					210,	0)		\
-	x(deleted_inode_but_clean,				211,	0)		\
-	x(deleted_inode_missing,				212,	0)		\
-	x(deleted_inode_is_dir,					213,	0)		\
-	x(deleted_inode_not_unlinked,				214,	0)		\
+	x(inode_i_size_dirty_but_clean,				202,	FSCK_AUTOFIX)	\
+	x(inode_i_sectors_dirty_but_clean,			203,	FSCK_AUTOFIX)	\
+	x(inode_i_sectors_wrong,				204,	FSCK_AUTOFIX)	\
+	x(inode_dir_wrong_nlink,				205,	FSCK_AUTOFIX)	\
+	x(inode_dir_multiple_links,				206,	FSCK_AUTOFIX)	\
+	x(inode_multiple_links_but_nlink_0,			207,	FSCK_AUTOFIX)	\
+	x(inode_wrong_backpointer,				208,	FSCK_AUTOFIX)	\
+	x(inode_wrong_nlink,					209,	FSCK_AUTOFIX)	\
+	x(inode_unreachable,					210,	FSCK_AUTOFIX)	\
+	x(deleted_inode_but_clean,				211,	FSCK_AUTOFIX)	\
+	x(deleted_inode_missing,				212,	FSCK_AUTOFIX)	\
+	x(deleted_inode_is_dir,					213,	FSCK_AUTOFIX)	\
+	x(deleted_inode_not_unlinked,				214,	FSCK_AUTOFIX)	\
 	x(extent_overlapping,					215,	0)		\
 	x(key_in_missing_inode,					216,	0)		\
 	x(key_in_wrong_inode_type,				217,	0)		\
@@ -255,7 +256,7 @@ enum bch_fsck_flags {
 	x(dir_loop,						241,	0)		\
 	x(hash_table_key_duplicate,				242,	0)		\
 	x(hash_table_key_wrong_offset,				243,	0)		\
-	x(unlinked_inode_not_on_deleted_list,			244,	0)		\
+	x(unlinked_inode_not_on_deleted_list,			244,	FSCK_AUTOFIX)	\
 	x(reflink_p_front_pad_bad,				245,	0)		\
 	x(journal_entry_dup_same_device,			246,	0)		\
 	x(inode_bi_subvol_missing,				247,	0)		\
@@ -270,7 +271,7 @@ enum bch_fsck_flags {
 	x(subvol_children_not_set,				256,	0)		\
 	x(subvol_children_bad,					257,	0)		\
 	x(subvol_loop,						258,	0)		\
-	x(subvol_unreachable,					259,	0)		\
+	x(subvol_unreachable,					259,	FSCK_AUTOFIX)	\
 	x(btree_node_bkey_bad_u64s,				260,	0)		\
 	x(btree_node_topology_empty_interior_node,		261,	0)		\
 	x(btree_ptr_v2_min_key_bad,				262,	0)		\
@@ -282,8 +283,8 @@ enum bch_fsck_flags {
 	x(btree_ptr_v2_written_0,				268,	0)		\
 	x(subvol_snapshot_bad,					269,	0)		\
 	x(subvol_inode_bad,					270,	0)		\
-	x(alloc_key_stripe_sectors_wrong,			271,	0)		\
-	x(accounting_mismatch,					272,	0)		\
+	x(alloc_key_stripe_sectors_wrong,			271,	FSCK_AUTOFIX)	\
+	x(accounting_mismatch,					272,	FSCK_AUTOFIX)	\
 	x(accounting_replicas_not_marked,			273,	0)		\
 	x(invalid_btree_id,					274,	0)		\
 	x(alloc_key_io_time_bad,				275,	0)		\
@@ -292,12 +293,14 @@ enum bch_fsck_flags {
 	x(accounting_key_replicas_nr_devs_0,			278,	FSCK_AUTOFIX)	\
 	x(accounting_key_replicas_nr_required_bad,		279,	FSCK_AUTOFIX)	\
 	x(accounting_key_replicas_devs_unsorted,		280,	FSCK_AUTOFIX)	\
+	x(accounting_key_version_0,				282,	FSCK_AUTOFIX)	\
+	x(logged_op_but_clean,					283,	FSCK_AUTOFIX)	\
+	x(MAX,							284,	0)
 
 enum bch_sb_error_id {
 #define x(t, n, ...) BCH_FSCK_ERR_##t = n,
 	BCH_SB_ERRS()
 #undef x
-	BCH_SB_ERR_MAX
 };
 
 struct bch_sb_field_errors {

fs/bcachefs/six.c

@@ -169,11 +169,17 @@ static int __do_six_trylock(struct six_lock *lock, enum six_lock_type type,
 				ret = -1 - SIX_LOCK_write;
 		}
 	} else if (type == SIX_LOCK_write && lock->readers) {
-		if (try) {
+		if (try)
 			atomic_add(SIX_LOCK_HELD_write, &lock->state);
-			smp_mb__after_atomic();
-		}
 
+		/*
+		 * Make sure atomic_add happens before pcpu_read_count and
+		 * six_set_bitmask in slow path happens before pcpu_read_count.
+		 *
+		 * Paired with the smp_mb() in read lock fast path (per-cpu mode)
+		 * and the one before atomic_read in read unlock path.
+		 */
+		smp_mb();
 		ret = !pcpu_read_count(lock);
 
 		if (try && !ret) {

fs/bcachefs/snapshot.c

@@ -469,6 +469,7 @@ static u32 bch2_snapshot_tree_oldest_subvol(struct bch_fs *c, u32 snapshot_root)
 	u32 id = snapshot_root;
 	u32 subvol = 0, s;
 
+	rcu_read_lock();
 	while (id) {
 		s = snapshot_t(c, id)->subvol;
 
@@ -477,6 +478,7 @@ static u32 bch2_snapshot_tree_oldest_subvol(struct bch_fs *c, u32 snapshot_root)
 
 		id = bch2_snapshot_tree_next(c, id);
 	}
+	rcu_read_unlock();
 
 	return subvol;
 }
@@ -1782,6 +1784,7 @@ static int bch2_propagate_key_to_snapshot_leaf(struct btree_trans *trans,
 	new->k.p.snapshot = leaf_id;
 	ret = bch2_trans_update(trans, &iter, new, 0);
 out:
+	bch2_set_btree_iter_dontneed(&iter);
 	bch2_trans_iter_exit(trans, &iter);
 	return ret;
 }

fs/bcachefs/subvolume.c

@@ -92,33 +92,31 @@ static int check_subvol(struct btree_trans *trans,
 	}
 
 	struct bch_inode_unpacked inode;
-	struct btree_iter inode_iter = {};
-	ret = bch2_inode_peek_nowarn(trans, &inode_iter, &inode,
+	ret = bch2_inode_find_by_inum_nowarn_trans(trans,
 				    (subvol_inum) { k.k->p.offset, le64_to_cpu(subvol.v->inode) },
-				    0);
-	bch2_trans_iter_exit(trans, &inode_iter);
-
-	if (ret && !bch2_err_matches(ret, ENOENT))
-		return ret;
-
-	if (fsck_err_on(ret,
-			trans, subvol_to_missing_root,
+				    &inode);
+	if (!ret) {
+		if (fsck_err_on(inode.bi_subvol != subvol.k->p.offset,
+				trans, subvol_root_wrong_bi_subvol,
+				"subvol root %llu:%u has wrong bi_subvol field: got %u, should be %llu",
+				inode.bi_inum, inode.bi_snapshot,
+				inode.bi_subvol, subvol.k->p.offset)) {
+			inode.bi_subvol = subvol.k->p.offset;
+			ret = __bch2_fsck_write_inode(trans, &inode, le32_to_cpu(subvol.v->snapshot));
+			if (ret)
+				goto err;
+		}
+	} else if (bch2_err_matches(ret, ENOENT)) {
+		if (fsck_err(trans, subvol_to_missing_root,
 			     "subvolume %llu points to missing subvolume root %llu:%u",
 			     k.k->p.offset, le64_to_cpu(subvol.v->inode),
 			     le32_to_cpu(subvol.v->snapshot))) {
 			ret = bch2_subvolume_delete(trans, iter->pos.offset);
 			bch_err_msg(c, ret, "deleting subvolume %llu", iter->pos.offset);
-		return ret ?: -BCH_ERR_transaction_restart_nested;
+			ret = ret ?: -BCH_ERR_transaction_restart_nested;
+			goto err;
 		}
-
-	if (fsck_err_on(inode.bi_subvol != subvol.k->p.offset,
-			trans, subvol_root_wrong_bi_subvol,
-			"subvol root %llu:%u has wrong bi_subvol field: got %u, should be %llu",
-			inode.bi_inum, inode_iter.k.p.snapshot,
-			inode.bi_subvol, subvol.k->p.offset)) {
-		inode.bi_subvol = subvol.k->p.offset;
-		ret = __bch2_fsck_write_inode(trans, &inode, le32_to_cpu(subvol.v->snapshot));
-		if (ret)
+	} else {
 		goto err;
 	}
 
@@ -137,7 +135,7 @@ static int check_subvol(struct btree_trans *trans,
 				"%s: snapshot tree %u not found", __func__, snapshot_tree);
 
 		if (ret)
-			return ret;
+			goto err;
 
 		if (fsck_err_on(le32_to_cpu(st.master_subvol) != subvol.k->p.offset,
 				trans, subvol_not_master_and_not_snapshot,
@@ -147,7 +145,7 @@ static int check_subvol(struct btree_trans *trans,
 				bch2_bkey_make_mut_typed(trans, iter, &subvol.s_c, 0, subvolume);
 			ret = PTR_ERR_OR_ZERO(s);
 			if (ret)
-				return ret;
+				goto err;
 
 			SET_BCH_SUBVOLUME_SNAP(&s->v, true);
 		}

fs/bcachefs/super-io.c

@@ -799,8 +799,10 @@ static int __bch2_read_super(const char *path, struct bch_opts *opts,
 	     i < layout.sb_offset + layout.nr_superblocks; i++) {
 		offset = le64_to_cpu(*i);
 
-		if (offset == opt_get(*opts, sb))
+		if (offset == opt_get(*opts, sb)) {
+			ret = -BCH_ERR_invalid;
 			continue;
+		}
 
 		ret = read_one_super(sb, offset, &err);
 		if (!ret)
@@ -1188,7 +1190,8 @@ static void bch2_sb_ext_to_text(struct printbuf *out, struct bch_sb *sb,
 		le_bitvector_to_cpu(errors_silent, (void *) e->errors_silent, sizeof(e->errors_silent) * 8);
 
 		prt_printf(out, "Errors to silently fix:\t");
-		prt_bitflags_vector(out, bch2_sb_error_strs, errors_silent, sizeof(e->errors_silent) * 8);
+		prt_bitflags_vector(out, bch2_sb_error_strs, errors_silent,
+				    min(BCH_FSCK_ERR_MAX, sizeof(e->errors_silent) * 8));
 		prt_newline(out);
 
 		kfree(errors_silent);

fs/bcachefs/tests.c

@@ -394,7 +394,7 @@ static int insert_test_extent(struct bch_fs *c,
 	k.k_i.k.p.offset = end;
 	k.k_i.k.p.snapshot = U32_MAX;
 	k.k_i.k.size = end - start;
-	k.k_i.k.version.lo = test_version++;
+	k.k_i.k.bversion.lo = test_version++;
 
 	ret = bch2_btree_insert(c, BTREE_ID_extents, &k.k_i, NULL, 0, 0);
 	bch_err_fn(c, ret);

include/trace/events/dma.h

@@ -176,9 +176,9 @@ TRACE_EVENT(dma_free,
 );
 
 TRACE_EVENT(dma_map_sg,
-	TP_PROTO(struct device *dev, struct scatterlist *sg, int nents,
+	TP_PROTO(struct device *dev, struct scatterlist *sgl, int nents,
 		 int ents, enum dma_data_direction dir, unsigned long attrs),
-	TP_ARGS(dev, sg, nents, ents, dir, attrs),
+	TP_ARGS(dev, sgl, nents, ents, dir, attrs),
 
 	TP_STRUCT__entry(
 		__string(device, dev_name(dev))
@@ -190,17 +190,17 @@ TRACE_EVENT(dma_map_sg,
 	),
 
 	TP_fast_assign(
+		struct scatterlist *sg;
 		int i;
 
 		__assign_str(device);
-		for (i = 0; i < nents; i++)
-			((u64 *)__get_dynamic_array(phys_addrs))[i] =
-				sg_phys(sg + i);
-		for (i = 0; i < ents; i++) {
+		for_each_sg(sgl, sg, nents, i)
+			((u64 *)__get_dynamic_array(phys_addrs))[i] = sg_phys(sg);
+		for_each_sg(sgl, sg, ents, i) {
 			((u64 *)__get_dynamic_array(dma_addrs))[i] =
-				sg_dma_address(sg + i);
+				sg_dma_address(sg);
 			((unsigned int *)__get_dynamic_array(lengths))[i] =
-				sg_dma_len(sg + i);
+				sg_dma_len(sg);
 		}
 		__entry->dir = dir;
 		__entry->attrs = attrs;
@@ -222,9 +222,9 @@ TRACE_EVENT(dma_map_sg,
 );
 
 TRACE_EVENT(dma_unmap_sg,
-	TP_PROTO(struct device *dev, struct scatterlist *sg, int nents,
+	TP_PROTO(struct device *dev, struct scatterlist *sgl, int nents,
 		 enum dma_data_direction dir, unsigned long attrs),
-	TP_ARGS(dev, sg, nents, dir, attrs),
+	TP_ARGS(dev, sgl, nents, dir, attrs),
 
 	TP_STRUCT__entry(
 		__string(device, dev_name(dev))
@@ -234,12 +234,12 @@ TRACE_EVENT(dma_unmap_sg,
 	),
 
 	TP_fast_assign(
+		struct scatterlist *sg;
 		int i;
 
 		__assign_str(device);
-		for (i = 0; i < nents; i++)
-			((u64 *)__get_dynamic_array(addrs))[i] =
-				sg_phys(sg + i);
+		for_each_sg(sgl, sg, nents, i)
+			((u64 *)__get_dynamic_array(addrs))[i] = sg_phys(sg);
 		__entry->dir = dir;
 		__entry->attrs = attrs;
 	),
@@ -290,9 +290,9 @@ DEFINE_EVENT(dma_sync_single, dma_sync_single_for_device,
 	TP_ARGS(dev, dma_addr, size, dir));
 
 DECLARE_EVENT_CLASS(dma_sync_sg,
-	TP_PROTO(struct device *dev, struct scatterlist *sg, int nents,
+	TP_PROTO(struct device *dev, struct scatterlist *sgl, int nents,
 		 enum dma_data_direction dir),
-	TP_ARGS(dev, sg, nents, dir),
+	TP_ARGS(dev, sgl, nents, dir),
 
 	TP_STRUCT__entry(
 		__string(device, dev_name(dev))
@@ -302,14 +302,15 @@ DECLARE_EVENT_CLASS(dma_sync_sg,
 	),
 
 	TP_fast_assign(
+		struct scatterlist *sg;
 		int i;
 
 		__assign_str(device);
-		for (i = 0; i < nents; i++) {
+		for_each_sg(sgl, sg, nents, i) {
 			((u64 *)__get_dynamic_array(dma_addrs))[i] =
-				sg_dma_address(sg + i);
+				sg_dma_address(sg);
 			((unsigned int *)__get_dynamic_array(lengths))[i] =
-				sg_dma_len(sg + i);
+				sg_dma_len(sg);
 		}
 		__entry->dir = dir;
 	),

tools/testing/selftests/timers/Makefile

@@ -1,5 +1,5 @@
 # SPDX-License-Identifier: GPL-2.0
-CFLAGS += -O3 -Wl,-no-as-needed -Wall
+CFLAGS += -O3 -Wl,-no-as-needed -Wall -I $(top_srcdir)
 LDLIBS += -lrt -lpthread -lm
 
 # these are all "safe" tests that don't modify

tools/testing/selftests/timers/adjtick.c

@@ -22,14 +22,12 @@
 #include <sys/time.h>
 #include <sys/timex.h>
 #include <time.h>
+#include <include/vdso/time64.h>
 
 #include "../kselftest.h"
 
 #define CLOCK_MONOTONIC_RAW	4
 
-#define NSEC_PER_SEC		1000000000LL
-#define USEC_PER_SEC		1000000
-
 #define MILLION			1000000
 
 long systick;

tools/testing/selftests/timers/alarmtimer-suspend.c

@@ -28,6 +28,8 @@
 #include <signal.h>
 #include <stdlib.h>
 #include <pthread.h>
+#include <include/vdso/time64.h>
+#include <errno.h>
 #include "../kselftest.h"
 
 #define CLOCK_REALTIME			0
@@ -45,7 +47,6 @@
 #define NR_CLOCKIDS			12
 
 
-#define NSEC_PER_SEC 1000000000ULL
 #define UNREASONABLE_LAT (NSEC_PER_SEC * 5) /* hopefully we resume in 5 secs */
 
 #define SUSPEND_SECS 15
@@ -142,8 +143,8 @@ int main(void)
 
 		alarmcount = 0;
 		if (timer_create(alarm_clock_id, &se, &tm1) == -1) {
-			printf("timer_create failed, %s unsupported?\n",
-					clockstring(alarm_clock_id));
+			printf("timer_create failed, %s unsupported?: %s\n",
+					clockstring(alarm_clock_id), strerror(errno));
 			break;
 		}
 

tools/testing/selftests/timers/inconsistency-check.c

@@ -28,10 +28,10 @@
 #include <sys/timex.h>
 #include <string.h>
 #include <signal.h>
+#include <include/vdso/time64.h>
 #include "../kselftest.h"
 
 #define CALLS_PER_LOOP 64
-#define NSEC_PER_SEC 1000000000ULL
 
 #define CLOCK_REALTIME			0
 #define CLOCK_MONOTONIC			1

tools/testing/selftests/timers/leap-a-day.c

@@ -48,9 +48,9 @@
 #include <string.h>
 #include <signal.h>
 #include <unistd.h>
+#include <include/vdso/time64.h>
 #include "../kselftest.h"
 
-#define NSEC_PER_SEC 1000000000ULL
 #define CLOCK_TAI 11
 
 time_t next_leap;

tools/testing/selftests/timers/mqueue-lat.c

@@ -29,9 +29,9 @@
 #include <signal.h>
 #include <errno.h>
 #include <mqueue.h>
+#include <include/vdso/time64.h>
 #include "../kselftest.h"
 
-#define NSEC_PER_SEC 1000000000ULL
 
 #define TARGET_TIMEOUT		100000000	/* 100ms in nanoseconds */
 #define UNRESONABLE_LATENCY	40000000	/* 40ms in nanosecs */

tools/testing/selftests/timers/nanosleep.c

@@ -27,10 +27,9 @@
 #include <sys/timex.h>
 #include <string.h>
 #include <signal.h>
+#include <include/vdso/time64.h>
 #include "../kselftest.h"
 
-#define NSEC_PER_SEC 1000000000ULL
-
 #define CLOCK_REALTIME			0
 #define CLOCK_MONOTONIC			1
 #define CLOCK_PROCESS_CPUTIME_ID	2

tools/testing/selftests/timers/nsleep-lat.c

@@ -24,10 +24,9 @@
 #include <sys/timex.h>
 #include <string.h>
 #include <signal.h>
+#include <include/vdso/time64.h>
 #include "../kselftest.h"
 
-#define NSEC_PER_SEC 1000000000ULL
-
 #define UNRESONABLE_LATENCY 40000000 /* 40ms in nanosecs */
 
 

tools/testing/selftests/timers/posix_timers.c

@@ -15,13 +15,12 @@
 #include <string.h>
 #include <unistd.h>
 #include <time.h>
+#include <include/vdso/time64.h>
 #include <pthread.h>
 
 #include "../kselftest.h"
 
 #define DELAY 2
-#define USECS_PER_SEC 1000000
-#define NSECS_PER_SEC 1000000000
 
 static void __fatal_error(const char *test, const char *name, const char *what)
 {
@@ -82,9 +81,9 @@ static int check_diff(struct timeval start, struct timeval end)
 	long long diff;
 
 	diff = end.tv_usec - start.tv_usec;
-	diff += (end.tv_sec - start.tv_sec) * USECS_PER_SEC;
+	diff += (end.tv_sec - start.tv_sec) * USEC_PER_SEC;
 
-	if (llabs(diff - DELAY * USECS_PER_SEC) > USECS_PER_SEC / 2) {
+	if (llabs(diff - DELAY * USEC_PER_SEC) > USEC_PER_SEC / 2) {
 		printf("Diff too high: %lld..", diff);
 		return -1;
 	}
@@ -444,7 +443,7 @@ static inline int64_t calcdiff_ns(struct timespec t1, struct timespec t2)
 {
 	int64_t diff;
 
-	diff = NSECS_PER_SEC * (int64_t)((int) t1.tv_sec - (int) t2.tv_sec);
+	diff = NSEC_PER_SEC * (int64_t)((int) t1.tv_sec - (int) t2.tv_sec);
 	diff += ((int) t1.tv_nsec - (int) t2.tv_nsec);
 	return diff;
 }
@@ -475,7 +474,7 @@ static void check_sigev_none(int which, const char *name)
 	do {
 		if (clock_gettime(which, &now))
 			fatal_error(name, "clock_gettime()");
-	} while (calcdiff_ns(now, start) < NSECS_PER_SEC);
+	} while (calcdiff_ns(now, start) < NSEC_PER_SEC);
 
 	if (timer_gettime(timerid, &its))
 		fatal_error(name, "timer_gettime()");
@@ -532,7 +531,7 @@ static void check_gettime(int which, const char *name)
 			wraps++;
 		prev = its;
 
-	} while (calcdiff_ns(now, start) < NSECS_PER_SEC);
+	} while (calcdiff_ns(now, start) < NSEC_PER_SEC);
 
 	if (timer_delete(timerid))
 		fatal_error(name, "timer_delete()");
@@ -583,7 +582,7 @@ static void check_overrun(int which, const char *name)
 	do {
 		if (clock_gettime(which, &now))
 			fatal_error(name, "clock_gettime()");
-	} while (calcdiff_ns(now, start) < NSECS_PER_SEC);
+	} while (calcdiff_ns(now, start) < NSEC_PER_SEC);
 
 	/* Unblock it, which should deliver a signal */
 	if (sigprocmask(SIG_UNBLOCK, &set, NULL))

tools/testing/selftests/timers/raw_skew.c

@@ -25,10 +25,10 @@
 #include <sys/time.h>
 #include <sys/timex.h>
 #include <time.h>
+#include <include/vdso/time64.h>
 #include "../kselftest.h"
 
 #define CLOCK_MONOTONIC_RAW		4
-#define NSEC_PER_SEC 1000000000LL
 
 #define shift_right(x, s) ({		\
 	__typeof__(x) __x = (x);	\

tools/testing/selftests/timers/set-2038.c

@@ -27,10 +27,9 @@
 #include <unistd.h>
 #include <time.h>
 #include <sys/time.h>
+#include <include/vdso/time64.h>
 #include "../kselftest.h"
 
-#define NSEC_PER_SEC 1000000000LL
-
 #define KTIME_MAX	((long long)~((unsigned long long)1 << 63))
 #define KTIME_SEC_MAX	(KTIME_MAX / NSEC_PER_SEC)