Adding upstream version 6.1.22-dfsg.upstream/6.1.22-dfsgupstream

Signed-off-by: Daniel Baumann <daniel.baumann@progress-linux.org>

1 files changed, 1308 insertions, 0 deletions

diff --git a/src/VBox/Additions/common/VBoxGuest/VBoxGuest-linux.c b/src/VBox/Additions/common/VBoxGuest/VBoxGuest-linux.c
new file mode 100644
index 00000000..b46f7d65
--- /dev/null
+++ b/src/VBox/Additions/common/VBoxGuest/VBoxGuest-linux.c
@@ -0,0 +1,1308 @@
+/* $Id: VBoxGuest-linux.c $ */
+/** @file
+ * VBoxGuest - Linux specifics.
+ *
+ * Note. Unfortunately, the difference between this and SUPDrv-linux.c is
+ *       a little bit too big to be helpful.
+ */
+
+/*
+ * Copyright (C) 2006-2020 Oracle Corporation
+ *
+ * This file is part of VirtualBox Open Source Edition (OSE), as
+ * available from http://www.virtualbox.org. This file is free software;
+ * you can redistribute it and/or modify it under the terms of the GNU
+ * General Public License (GPL) as published by the Free Software
+ * Foundation, in version 2 as it comes in the "COPYING" file of the
+ * VirtualBox OSE distribution. VirtualBox OSE is distributed in the
+ * hope that it will be useful, but WITHOUT ANY WARRANTY of any kind.
+ *
+ * The contents of this file may alternatively be used under the terms
+ * of the Common Development and Distribution License Version 1.0
+ * (CDDL) only, as it comes in the "COPYING.CDDL" file of the
+ * VirtualBox OSE distribution, in which case the provisions of the
+ * CDDL are applicable instead of those of the GPL.
+ *
+ * You may elect to license modified versions of this file under the
+ * terms and conditions of either the GPL or the CDDL or both.
+ */
+
+
+/*********************************************************************************************************************************
+*   Header Files                                                                                                                 *
+*********************************************************************************************************************************/
+#define LOG_GROUP LOG_GROUP_SUP_DRV
+
+#include "the-linux-kernel.h"
+
+#if RTLNX_VER_MIN(2,6,15)
+# define VBOXGUEST_WITH_INPUT_DRIVER
+#endif
+
+#if RTLNX_VER_MIN(4,15,0)
+# define CONST_4_15 const
+#else
+# define CONST_4_15
+#endif
+
+#include "VBoxGuestInternal.h"
+#ifdef VBOXGUEST_WITH_INPUT_DRIVER
+# include <linux/input.h>
+#endif
+#include <linux/miscdevice.h>
+#include <linux/poll.h>
+#if RTLNX_VER_MIN(2,6,28)
+# include <linux/tty.h>
+#endif
+#include <VBox/version.h>
+#include "revision-generated.h"
+
+#include <iprt/assert.h>
+#include <iprt/asm.h>
+#include <iprt/ctype.h>
+#include <iprt/initterm.h>
+#include <iprt/mem.h>
+#include <iprt/mp.h>
+#include <iprt/process.h>
+#include <iprt/spinlock.h>
+#include <iprt/semaphore.h>
+#include <iprt/string.h>
+#include <VBox/err.h>
+#include <VBox/log.h>
+
+
+/*********************************************************************************************************************************
+*   Defined Constants And Macros                                                                                                 *
+*********************************************************************************************************************************/
+/** The device name. */
+#define DEVICE_NAME             "vboxguest"
+/** The device name for the device node open to everyone. */
+#define DEVICE_NAME_USER        "vboxuser"
+/** The name of the PCI driver */
+#define DRIVER_NAME             DEVICE_NAME
+
+
+/* 2.4.x compatibility macros that may or may not be defined. */
+#ifndef IRQ_RETVAL
+# define irqreturn_t            void
+# define IRQ_RETVAL(n)
+#endif
+
+/* uidgid.h was introduced in 3.5.0. */
+#if RTLNX_VER_MAX(3,5,0)
+# define kgid_t gid_t
+# define kuid_t uid_t
+#endif
+
+
+/*********************************************************************************************************************************
+*   Internal Functions                                                                                                           *
+*********************************************************************************************************************************/
+static void vgdrvLinuxTermPci(struct pci_dev *pPciDev);
+static int  vgdrvLinuxProbePci(struct pci_dev *pPciDev, const struct pci_device_id *id);
+static int  __init vgdrvLinuxModInit(void);
+static void __exit vgdrvLinuxModExit(void);
+static int  vgdrvLinuxOpen(struct inode *pInode, struct file *pFilp);
+static int  vgdrvLinuxRelease(struct inode *pInode, struct file *pFilp);
+#ifdef HAVE_UNLOCKED_IOCTL
+static long vgdrvLinuxIOCtl(struct file *pFilp, unsigned int uCmd, unsigned long ulArg);
+#else
+static int  vgdrvLinuxIOCtl(struct inode *pInode, struct file *pFilp, unsigned int uCmd, unsigned long ulArg);
+#endif
+static int  vgdrvLinuxIOCtlSlow(struct file *pFilp, unsigned int uCmd, unsigned long ulArg, PVBOXGUESTSESSION pSession);
+static int  vgdrvLinuxFAsync(int fd, struct file *pFile, int fOn);
+static unsigned int vgdrvLinuxPoll(struct file *pFile, poll_table *pPt);
+static ssize_t vgdrvLinuxRead(struct file *pFile, char *pbBuf, size_t cbRead, loff_t *poff);
+
+
+/*********************************************************************************************************************************
+*   Global Variables                                                                                                             *
+*********************************************************************************************************************************/
+/**
+ * Device extention & session data association structure.
+ */
+static VBOXGUESTDEVEXT          g_DevExt;
+/** The PCI device. */
+static struct pci_dev          *g_pPciDev = NULL;
+/** The base of the I/O port range. */
+static RTIOPORT                 g_IOPortBase;
+/** The base of the MMIO range. */
+static RTHCPHYS                 g_MMIOPhysAddr = NIL_RTHCPHYS;
+/** The size of the MMIO range as seen by PCI. */
+static uint32_t                 g_cbMMIO;
+/** The pointer to the mapping of the MMIO range. */
+static void                    *g_pvMMIOBase;
+/** Wait queue used by polling. */
+static wait_queue_head_t        g_PollEventQueue;
+/** Asynchronous notification stuff.  */
+static struct fasync_struct    *g_pFAsyncQueue;
+#ifdef VBOXGUEST_WITH_INPUT_DRIVER
+/** Pre-allocated mouse status VMMDev request for use in the IRQ
+ * handler. */
+static VMMDevReqMouseStatus    *g_pMouseStatusReq;
+#endif
+#if RTLNX_VER_MIN(2,6,0)
+/** Whether we've create the logger or not. */
+static volatile bool            g_fLoggerCreated;
+/** Release logger group settings. */
+static char                     g_szLogGrp[128];
+/** Release logger flags settings. */
+static char                     g_szLogFlags[128];
+/** Release logger destination settings. */
+static char                     g_szLogDst[128];
+# if 0
+/** Debug logger group settings. */
+static char                     g_szDbgLogGrp[128];
+/** Debug logger flags settings. */
+static char                     g_szDbgLogFlags[128];
+/** Debug logger destination settings. */
+static char                     g_szDbgLogDst[128];
+# endif
+#endif
+
+/** The input device handle */
+#ifdef VBOXGUEST_WITH_INPUT_DRIVER
+static struct input_dev        *g_pInputDevice = NULL;
+#endif
+
+/** The file_operations structure. */
+static struct file_operations   g_FileOps =
+{
+    owner:          THIS_MODULE,
+    open:           vgdrvLinuxOpen,
+    release:        vgdrvLinuxRelease,
+#ifdef HAVE_UNLOCKED_IOCTL
+    unlocked_ioctl: vgdrvLinuxIOCtl,
+#else
+    ioctl:          vgdrvLinuxIOCtl,
+#endif
+    fasync:         vgdrvLinuxFAsync,
+    read:           vgdrvLinuxRead,
+    poll:           vgdrvLinuxPoll,
+    llseek:         no_llseek,
+};
+
+/** The miscdevice structure. */
+static struct miscdevice        g_MiscDevice =
+{
+    minor:          MISC_DYNAMIC_MINOR,
+    name:           DEVICE_NAME,
+    fops:           &g_FileOps,
+};
+
+/** The file_operations structure for the user device.
+ * @remarks For the time being we'll be using the same implementation as
+ *          /dev/vboxguest here. */
+static struct file_operations   g_FileOpsUser =
+{
+    owner:          THIS_MODULE,
+    open:           vgdrvLinuxOpen,
+    release:        vgdrvLinuxRelease,
+#ifdef HAVE_UNLOCKED_IOCTL
+    unlocked_ioctl: vgdrvLinuxIOCtl,
+#else
+    ioctl:          vgdrvLinuxIOCtl,
+#endif
+};
+
+/** The miscdevice structure for the user device. */
+static struct miscdevice        g_MiscDeviceUser =
+{
+    minor:          MISC_DYNAMIC_MINOR,
+    name:           DEVICE_NAME_USER,
+    fops:           &g_FileOpsUser,
+};
+
+
+/** PCI hotplug structure. */
+static const struct pci_device_id g_VBoxGuestPciId[] =
+{
+    {
+        vendor:     VMMDEV_VENDORID,
+        device:     VMMDEV_DEVICEID
+    },
+    {
+        /* empty entry */
+    }
+};
+
+MODULE_DEVICE_TABLE(pci, g_VBoxGuestPciId);
+
+/** Structure for registering the PCI driver. */
+static struct pci_driver  g_PciDriver =
+{
+    name:           DRIVER_NAME,
+    id_table:       g_VBoxGuestPciId,
+    probe:          vgdrvLinuxProbePci,
+    remove:         vgdrvLinuxTermPci
+};
+
+#ifdef VBOXGUEST_WITH_INPUT_DRIVER
+/** Kernel IDC session to ourselves for use with the mouse events. */
+static PVBOXGUESTSESSION        g_pKernelSession = NULL;
+#endif
+
+
+
+/**
+ * Converts a VBox status code to a linux error code.
+ *
+ * @returns corresponding negative linux error code.
+ * @param   rc  supdrv error code (SUPDRV_ERR_* defines).
+ */
+static int vgdrvLinuxConvertToNegErrno(int rc)
+{
+    if (   rc > -1000
+        && rc < 1000)
+        return -RTErrConvertToErrno(rc);
+    switch (rc)
+    {
+        case VERR_HGCM_SERVICE_NOT_FOUND:      return -ESRCH;
+        case VINF_HGCM_CLIENT_REJECTED:        return 0;
+        case VERR_HGCM_INVALID_CMD_ADDRESS:    return -EFAULT;
+        case VINF_HGCM_ASYNC_EXECUTE:          return 0;
+        case VERR_HGCM_INTERNAL:               return -EPROTO;
+        case VERR_HGCM_INVALID_CLIENT_ID:      return -EINVAL;
+        case VINF_HGCM_SAVE_STATE:             return 0;
+        /* No reason to return this to a guest */
+        // case VERR_HGCM_SERVICE_EXISTS:         return -EEXIST;
+        default:
+            AssertMsgFailed(("Unhandled error code %Rrc\n", rc));
+            return -EPROTO;
+    }
+}
+
+
+/**
+ * Does the PCI detection and init of the device.
+ *
+ * @returns 0 on success, negated errno on failure.
+ */
+static int vgdrvLinuxProbePci(struct pci_dev *pPciDev, const struct pci_device_id *id)
+{
+    int rc;
+
+    NOREF(id);
+    AssertReturn(!g_pPciDev, -EINVAL);
+    rc = pci_enable_device(pPciDev);
+    if (rc >= 0)
+    {
+        /* I/O Ports are mandatory, the MMIO bit is not. */
+        g_IOPortBase = pci_resource_start(pPciDev, 0);
+        if (g_IOPortBase != 0)
+        {
+            /*
+             * Map the register address space.
+             */
+            g_MMIOPhysAddr = pci_resource_start(pPciDev, 1);
+            g_cbMMIO       = pci_resource_len(pPciDev, 1);
+            if (request_mem_region(g_MMIOPhysAddr, g_cbMMIO, DEVICE_NAME) != NULL)
+            {
+                g_pvMMIOBase = ioremap(g_MMIOPhysAddr, g_cbMMIO);
+                if (g_pvMMIOBase)
+                {
+                    /** @todo why aren't we requesting ownership of the I/O ports as well? */
+                    g_pPciDev = pPciDev;
+                    return 0;
+                }
+
+                /* failure cleanup path */
+                LogRel((DEVICE_NAME ": ioremap failed; MMIO Addr=%RHp cb=%#x\n", g_MMIOPhysAddr, g_cbMMIO));
+                rc = -ENOMEM;
+                release_mem_region(g_MMIOPhysAddr, g_cbMMIO);
+            }
+            else
+            {
+                LogRel((DEVICE_NAME ": failed to obtain adapter memory\n"));
+                rc = -EBUSY;
+            }
+            g_MMIOPhysAddr = NIL_RTHCPHYS;
+            g_cbMMIO       = 0;
+            g_IOPortBase   = 0;
+        }
+        else
+        {
+            LogRel((DEVICE_NAME ": did not find expected hardware resources\n"));
+            rc = -ENXIO;
+        }
+        pci_disable_device(pPciDev);
+    }
+    else
+        LogRel((DEVICE_NAME ": could not enable device: %d\n", rc));
+    return rc;
+}
+
+
+/**
+ * Clean up the usage of the PCI device.
+ */
+static void vgdrvLinuxTermPci(struct pci_dev *pPciDev)
+{
+    g_pPciDev = NULL;
+    if (pPciDev)
+    {
+        iounmap(g_pvMMIOBase);
+        g_pvMMIOBase = NULL;
+
+        release_mem_region(g_MMIOPhysAddr, g_cbMMIO);
+        g_MMIOPhysAddr = NIL_RTHCPHYS;
+        g_cbMMIO = 0;
+
+        pci_disable_device(pPciDev);
+    }
+}
+
+
+/**
+ * Interrupt service routine.
+ *
+ * @returns In 2.4 it returns void.
+ *          In 2.6 we indicate whether we've handled the IRQ or not.
+ *
+ * @param   iIrq            The IRQ number.
+ * @param   pvDevId         The device ID, a pointer to g_DevExt.
+ * @param   pRegs           Register set. Removed in 2.6.19.
+ */
+#if RTLNX_VER_MIN(2,6,19) && !defined(DOXYGEN_RUNNING)
+static irqreturn_t vgdrvLinuxISR(int iIrq, void *pvDevId)
+#else
+static irqreturn_t vgdrvLinuxISR(int iIrq, void *pvDevId, struct pt_regs *pRegs)
+#endif
+{
+    bool fTaken = VGDrvCommonISR(&g_DevExt);
+    return IRQ_RETVAL(fTaken);
+}
+
+
+/**
+ * Registers the ISR and initializes the poll wait queue.
+ */
+static int __init vgdrvLinuxInitISR(void)
+{
+    int rc;
+
+    init_waitqueue_head(&g_PollEventQueue);
+    rc = request_irq(g_pPciDev->irq,
+                     vgdrvLinuxISR,
+#if RTLNX_VER_MIN(2,6,20)
+                     IRQF_SHARED,
+#else
+                     SA_SHIRQ,
+#endif
+                     DEVICE_NAME,
+                     &g_DevExt);
+    if (rc)
+    {
+        LogRel((DEVICE_NAME ": could not request IRQ %d: err=%d\n", g_pPciDev->irq, rc));
+        return rc;
+    }
+    return 0;
+}
+
+
+/**
+ * Deregisters the ISR.
+ */
+static void vgdrvLinuxTermISR(void)
+{
+    free_irq(g_pPciDev->irq, &g_DevExt);
+}
+
+
+#ifdef VBOXGUEST_WITH_INPUT_DRIVER
+
+/**
+ * Reports the mouse integration status to the host.
+ *
+ * Calls the kernel IOCtl to report mouse status to the host on behalf of
+ * our kernel session.
+ *
+ * @param   fStatus     The mouse status to report.
+ */
+static int vgdrvLinuxSetMouseStatus(uint32_t fStatus)
+{
+    int rc;
+    VBGLIOCSETMOUSESTATUS Req;
+    VBGLREQHDR_INIT(&Req.Hdr, SET_MOUSE_STATUS);
+    Req.u.In.fStatus = fStatus;
+    rc = VGDrvCommonIoCtl(VBGL_IOCTL_SET_MOUSE_STATUS, &g_DevExt, g_pKernelSession, &Req.Hdr, sizeof(Req));
+    if (RT_SUCCESS(rc))
+        rc = Req.Hdr.rc;
+    return rc;
+}
+
+
+/**
+ * Called when the input device is first opened.
+ *
+ * Sets up absolute mouse reporting.
+ */
+static int vboxguestOpenInputDevice(struct input_dev *pDev)
+{
+    int rc = vgdrvLinuxSetMouseStatus(VMMDEV_MOUSE_GUEST_CAN_ABSOLUTE | VMMDEV_MOUSE_NEW_PROTOCOL);
+    if (RT_FAILURE(rc))
+        return ENODEV;
+    NOREF(pDev);
+    return 0;
+}
+
+
+/**
+ * Called if all open handles to the input device are closed.
+ *
+ * Disables absolute reporting.
+ */
+static void vboxguestCloseInputDevice(struct input_dev *pDev)
+{
+    NOREF(pDev);
+    vgdrvLinuxSetMouseStatus(0);
+}
+
+
+/**
+ * Creates the kernel input device.
+ */
+static int __init vgdrvLinuxCreateInputDevice(void)
+{
+    int rc = VbglR0GRAlloc((VMMDevRequestHeader **)&g_pMouseStatusReq, sizeof(*g_pMouseStatusReq), VMMDevReq_GetMouseStatus);
+    if (RT_SUCCESS(rc))
+    {
+        g_pInputDevice = input_allocate_device();
+        if (g_pInputDevice)
+        {
+            g_pInputDevice->id.bustype = BUS_PCI;
+            g_pInputDevice->id.vendor  = VMMDEV_VENDORID;
+            g_pInputDevice->id.product = VMMDEV_DEVICEID;
+            g_pInputDevice->id.version = VBOX_SHORT_VERSION;
+            g_pInputDevice->open       = vboxguestOpenInputDevice;
+            g_pInputDevice->close      = vboxguestCloseInputDevice;
+# if RTLNX_VER_MAX(2,6,22)
+            g_pInputDevice->cdev.dev   = &g_pPciDev->dev;
+# else
+            g_pInputDevice->dev.parent = &g_pPciDev->dev;
+# endif
+            rc = input_register_device(g_pInputDevice);
+            if (rc == 0)
+            {
+                /* Do what one of our competitors apparently does as that works. */
+                ASMBitSet(g_pInputDevice->evbit, EV_ABS);
+                ASMBitSet(g_pInputDevice->evbit, EV_KEY);
+# ifdef EV_SYN
+                ASMBitSet(g_pInputDevice->evbit, EV_SYN);
+# endif
+                input_set_abs_params(g_pInputDevice, ABS_X, VMMDEV_MOUSE_RANGE_MIN, VMMDEV_MOUSE_RANGE_MAX, 0, 0);
+                input_set_abs_params(g_pInputDevice, ABS_Y, VMMDEV_MOUSE_RANGE_MIN, VMMDEV_MOUSE_RANGE_MAX, 0, 0);
+                ASMBitSet(g_pInputDevice->keybit, BTN_MOUSE);
+                /** @todo this string should be in a header file somewhere. */
+                g_pInputDevice->name = "VirtualBox mouse integration";
+                return 0;
+            }
+
+            input_free_device(g_pInputDevice);
+        }
+        else
+            rc = -ENOMEM;
+        VbglR0GRFree(&g_pMouseStatusReq->header);
+        g_pMouseStatusReq = NULL;
+    }
+    else
+        rc = -ENOMEM;
+    return rc;
+}
+
+
+/**
+ * Terminates the kernel input device.
+ */
+static void vgdrvLinuxTermInputDevice(void)
+{
+    VbglR0GRFree(&g_pMouseStatusReq->header);
+    g_pMouseStatusReq = NULL;
+
+    /* See documentation of input_register_device(): input_free_device()
+     * should not be called after a device has been registered. */
+    input_unregister_device(g_pInputDevice);
+}
+
+#endif /* VBOXGUEST_WITH_INPUT_DRIVER */
+
+/**
+ * Creates the device nodes.
+ *
+ * @returns 0 on success, negated errno on failure.
+ */
+static int __init vgdrvLinuxInitDeviceNodes(void)
+{
+    /*
+     * The full feature device node.
+     */
+    int rc = misc_register(&g_MiscDevice);
+    if (!rc)
+    {
+        /*
+         * The device node intended to be accessible by all users.
+         */
+        rc = misc_register(&g_MiscDeviceUser);
+        if (!rc)
+            return 0;
+        LogRel((DEVICE_NAME ": misc_register failed for %s (rc=%d)\n", DEVICE_NAME_USER, rc));
+        misc_deregister(&g_MiscDevice);
+    }
+    else
+        LogRel((DEVICE_NAME ": misc_register failed for %s (rc=%d)\n", DEVICE_NAME, rc));
+    return rc;
+}
+
+
+/**
+ * Deregisters the device nodes.
+ */
+static void vgdrvLinuxTermDeviceNodes(void)
+{
+    misc_deregister(&g_MiscDevice);
+    misc_deregister(&g_MiscDeviceUser);
+}
+
+
+/**
+ * Initialize module.
+ *
+ * @returns appropriate status code.
+ */
+static int __init vgdrvLinuxModInit(void)
+{
+    static const char * const   s_apszGroups[] = VBOX_LOGGROUP_NAMES;
+    PRTLOGGER                   pRelLogger;
+    int                         rc;
+
+    /*
+     * Initialize IPRT first.
+     */
+    rc = RTR0Init(0);
+    if (RT_FAILURE(rc))
+    {
+        printk(KERN_ERR DEVICE_NAME ": RTR0Init failed, rc=%d.\n", rc);
+        return -EINVAL;
+    }
+
+    /*
+     * Create the release log.
+     * (We do that here instead of common code because we want to log
+     * early failures using the LogRel macro.)
+     */
+    rc = RTLogCreate(&pRelLogger, 0 /* fFlags */, "all",
+                     "VBOX_RELEASE_LOG", RT_ELEMENTS(s_apszGroups), s_apszGroups,
+                     RTLOGDEST_STDOUT | RTLOGDEST_DEBUGGER | RTLOGDEST_USER, NULL);
+    if (RT_SUCCESS(rc))
+    {
+#if RTLNX_VER_MIN(2,6,0)
+        RTLogGroupSettings(pRelLogger, g_szLogGrp);
+        RTLogFlags(pRelLogger, g_szLogFlags);
+        RTLogDestinations(pRelLogger, g_szLogDst);
+#endif
+        RTLogRelSetDefaultInstance(pRelLogger);
+    }
+#if RTLNX_VER_MIN(2,6,0)
+    g_fLoggerCreated = true;
+#endif
+
+    /*
+     * Locate and initialize the PCI device.
+     */
+    rc = pci_register_driver(&g_PciDriver);
+    if (rc >= 0 && g_pPciDev)
+    {
+        /*
+         * Call the common device extension initializer.
+         */
+#if RTLNX_VER_MIN(2,6,0) && defined(RT_ARCH_X86)
+        VBOXOSTYPE enmOSType = VBOXOSTYPE_Linux26;
+#elif RTLNX_VER_MIN(2,6,0) && defined(RT_ARCH_AMD64)
+        VBOXOSTYPE enmOSType = VBOXOSTYPE_Linux26_x64;
+#elif RTLNX_VER_MIN(2,4,0) && defined(RT_ARCH_X86)
+        VBOXOSTYPE enmOSType = VBOXOSTYPE_Linux24;
+#elif RTLNX_VER_MIN(2,4,0) && defined(RT_ARCH_AMD64)
+        VBOXOSTYPE enmOSType = VBOXOSTYPE_Linux24_x64;
+#else
+# warning "huh? which arch + version is this?"
+        VBOXOSTYPE enmOsType = VBOXOSTYPE_Linux;
+#endif
+        rc = VGDrvCommonInitDevExt(&g_DevExt,
+                                   g_IOPortBase,
+                                   g_pvMMIOBase,
+                                   g_cbMMIO,
+                                   enmOSType,
+                                   VMMDEV_EVENT_MOUSE_POSITION_CHANGED);
+        if (RT_SUCCESS(rc))
+        {
+            /*
+             * Register the interrupt service routine for it now that g_DevExt can handle IRQs.
+             */
+            rc = vgdrvLinuxInitISR();
+            if (rc >= 0) /** @todo r=bird: status check differs from that inside vgdrvLinuxInitISR. */
+            {
+#ifdef VBOXGUEST_WITH_INPUT_DRIVER
+                /*
+                 * Create the kernel session for this driver.
+                 */
+                rc = VGDrvCommonCreateKernelSession(&g_DevExt, &g_pKernelSession);
+                if (RT_SUCCESS(rc))
+                {
+                    /*
+                     * Create the kernel input device.
+                     */
+                    rc = vgdrvLinuxCreateInputDevice();
+                    if (rc >= 0)
+                    {
+#endif
+                        /*
+                         * Read host configuration.
+                         */
+                        VGDrvCommonProcessOptionsFromHost(&g_DevExt);
+
+                        /*
+                         * Finally, create the device nodes.
+                         */
+                        rc = vgdrvLinuxInitDeviceNodes();
+                        if (rc >= 0)
+                        {
+                            /* some useful information for the user but don't show this on the console */
+                            LogRel((DEVICE_NAME ": Successfully loaded version " VBOX_VERSION_STRING " r" __stringify(VBOX_SVN_REV) "\n"));
+                            LogRel((DEVICE_NAME ": misc device minor %d, IRQ %d, I/O port %RTiop, MMIO at %RHp (size 0x%x)\n",
+                                    g_MiscDevice.minor, g_pPciDev->irq, g_IOPortBase, g_MMIOPhysAddr, g_cbMMIO));
+                            printk(KERN_DEBUG DEVICE_NAME ": Successfully loaded version "
+                                   VBOX_VERSION_STRING " r" __stringify(VBOX_SVN_REV) " (interface " RT_XSTR(VMMDEV_VERSION) ")\n");
+                            return rc;
+                        }
+
+                        /* bail out */
+#ifdef VBOXGUEST_WITH_INPUT_DRIVER
+                        vgdrvLinuxTermInputDevice();
+                    }
+                    else
+                    {
+                        LogRel((DEVICE_NAME ": vboxguestCreateInputDevice failed with rc=%Rrc\n", rc));
+                        rc = RTErrConvertFromErrno(rc);
+                    }
+                    VGDrvCommonCloseSession(&g_DevExt, g_pKernelSession);
+                }
+#endif
+                vgdrvLinuxTermISR();
+            }
+            VGDrvCommonDeleteDevExt(&g_DevExt);
+        }
+        else
+        {
+            LogRel((DEVICE_NAME ": VGDrvCommonInitDevExt failed with rc=%Rrc\n", rc));
+            rc = RTErrConvertFromErrno(rc);
+        }
+    }
+    else
+    {
+        LogRel((DEVICE_NAME ": PCI device not found, probably running on physical hardware.\n"));
+        rc = -ENODEV;
+    }
+    pci_unregister_driver(&g_PciDriver);
+    RTLogDestroy(RTLogRelSetDefaultInstance(NULL));
+    RTLogDestroy(RTLogSetDefaultInstance(NULL));
+    RTR0Term();
+    return rc;
+}
+
+
+/**
+ * Unload the module.
+ */
+static void __exit vgdrvLinuxModExit(void)
+{
+    /*
+     * Inverse order of init.
+     */
+    vgdrvLinuxTermDeviceNodes();
+#ifdef VBOXGUEST_WITH_INPUT_DRIVER
+    vgdrvLinuxTermInputDevice();
+    VGDrvCommonCloseSession(&g_DevExt, g_pKernelSession);
+#endif
+    vgdrvLinuxTermISR();
+    VGDrvCommonDeleteDevExt(&g_DevExt);
+    pci_unregister_driver(&g_PciDriver);
+    RTLogDestroy(RTLogRelSetDefaultInstance(NULL));
+    RTLogDestroy(RTLogSetDefaultInstance(NULL));
+    RTR0Term();
+}
+
+
+/**
+ * Get the process user ID.
+ *
+ * @returns UID.
+ */
+DECLINLINE(RTUID) vgdrvLinuxGetUid(void)
+{
+#if RTLNX_VER_MIN(2,6,29)
+# if RTLNX_VER_MIN(3,5,0)
+    return from_kuid(current_user_ns(), current->cred->uid);
+# else
+    return current->cred->uid;
+# endif
+#else
+    return current->uid;
+#endif
+}
+
+
+/**
+ * Checks if the given group number is zero or not.
+ *
+ * @returns true / false.
+ * @param   gid                 The group to check for.
+ */
+DECLINLINE(bool) vgdrvLinuxIsGroupZero(kgid_t gid)
+{
+#if RTLNX_VER_MIN(3,5,0)
+    return from_kgid(current_user_ns(), gid);
+#else
+    return gid == 0;
+#endif
+}
+
+
+/**
+ * Searches the effective group and supplementary groups for @a gid.
+ *
+ * @returns true if member, false if not.
+ * @param   gid                 The group to check for.
+ */
+DECLINLINE(RTGID) vgdrvLinuxIsInGroupEff(kgid_t gid)
+{
+    return in_egroup_p(gid) != 0;
+}
+
+
+/**
+ * Check if we can positively or negatively determine that the process is
+ * running under a login on the physical machine console.
+ *
+ * Havne't found a good way to figure this out for graphical sessions, so this
+ * is mostly pointless.  But let us try do what we can do.
+ *
+ * @returns VMMDEV_REQUESTOR_CON_XXX.
+ */
+static uint32_t vgdrvLinuxRequestorOnConsole(void)
+{
+    uint32_t           fRet = VMMDEV_REQUESTOR_CON_DONT_KNOW;
+
+#if RTLNX_VER_MIN(2,6,28) /* First with tty_kref_put(). */
+    /*
+     * Check for tty0..63, ASSUMING that these are only used for the physical console.
+     */
+    struct tty_struct *pTty = get_current_tty();
+    if (pTty)
+    {
+# if RTLNX_VER_MIN(4,2,0)
+        const char *pszName = tty_name(pTty);
+# else
+        char szBuf[64];
+        const char *pszName = tty_name(pTty, szBuf);
+# endif
+        if (   pszName
+            && pszName[0] == 't'
+            && pszName[1] == 't'
+            && pszName[2] == 'y'
+            && RT_C_IS_DIGIT(pszName[3])
+            && (   pszName[4] == '\0'
+                || (   RT_C_IS_DIGIT(pszName[4])
+                    && pszName[5] == '\0'
+                    && (pszName[3] - '0') * 10 + (pszName[4] - '0') <= 63)) )
+               fRet = VMMDEV_REQUESTOR_CON_YES;
+        tty_kref_put(pTty);
+    }
+#endif
+
+    return fRet;
+}
+
+
+/**
+ * Device open. Called on open /dev/vboxdrv
+ *
+ * @param   pInode      Pointer to inode info structure.
+ * @param   pFilp       Associated file pointer.
+ */
+static int vgdrvLinuxOpen(struct inode *pInode, struct file *pFilp)
+{
+    int                 rc;
+    PVBOXGUESTSESSION   pSession;
+    uint32_t            fRequestor;
+    Log((DEVICE_NAME ": pFilp=%p pid=%d/%d %s\n", pFilp, RTProcSelf(), current->pid, current->comm));
+
+    /*
+     * Figure out the requestor flags.
+     * ASSUMES that the gid of /dev/vboxuser is what we should consider the special vbox group.
+     */
+    fRequestor = VMMDEV_REQUESTOR_USERMODE | VMMDEV_REQUESTOR_TRUST_NOT_GIVEN;
+    if (vgdrvLinuxGetUid() == 0)
+        fRequestor |= VMMDEV_REQUESTOR_USR_ROOT;
+    else
+        fRequestor |= VMMDEV_REQUESTOR_USR_USER;
+    if (MINOR(pInode->i_rdev) == g_MiscDeviceUser.minor)
+    {
+        fRequestor |= VMMDEV_REQUESTOR_USER_DEVICE;
+        if (!vgdrvLinuxIsGroupZero(pInode->i_gid) && vgdrvLinuxIsInGroupEff(pInode->i_gid))
+            fRequestor |= VMMDEV_REQUESTOR_GRP_VBOX;
+    }
+    fRequestor |= vgdrvLinuxRequestorOnConsole();
+
+    /*
+     * Call common code to create the user session. Associate it with
+     * the file so we can access it in the other methods.
+     */
+    rc = VGDrvCommonCreateUserSession(&g_DevExt, fRequestor, &pSession);
+    if (RT_SUCCESS(rc))
+        pFilp->private_data = pSession;
+
+    Log(("vgdrvLinuxOpen: g_DevExt=%p pSession=%p rc=%d/%d (pid=%d/%d %s)\n",
+         &g_DevExt, pSession, rc, vgdrvLinuxConvertToNegErrno(rc), RTProcSelf(), current->pid, current->comm));
+    return vgdrvLinuxConvertToNegErrno(rc);
+}
+
+
+/**
+ * Close device.
+ *
+ * @param   pInode      Pointer to inode info structure.
+ * @param   pFilp       Associated file pointer.
+ */
+static int vgdrvLinuxRelease(struct inode *pInode, struct file *pFilp)
+{
+    Log(("vgdrvLinuxRelease: pFilp=%p pSession=%p pid=%d/%d %s\n",
+         pFilp, pFilp->private_data, RTProcSelf(), current->pid, current->comm));
+
+#if RTLNX_VER_MAX(2,6,28)
+    /* This housekeeping was needed in older kernel versions to ensure that
+     * the file pointer didn't get left on the polling queue. */
+    vgdrvLinuxFAsync(-1, pFilp, 0);
+#endif
+    VGDrvCommonCloseSession(&g_DevExt, (PVBOXGUESTSESSION)pFilp->private_data);
+    pFilp->private_data = NULL;
+    return 0;
+}
+
+
+/**
+ * Device I/O Control entry point.
+ *
+ * @param   pFilp       Associated file pointer.
+ * @param   uCmd        The function specified to ioctl().
+ * @param   ulArg       The argument specified to ioctl().
+ */
+#if defined(HAVE_UNLOCKED_IOCTL) || defined(DOXYGEN_RUNNING)
+static long vgdrvLinuxIOCtl(struct file *pFilp, unsigned int uCmd, unsigned long ulArg)
+#else
+static int vgdrvLinuxIOCtl(struct inode *pInode, struct file *pFilp, unsigned int uCmd, unsigned long ulArg)
+#endif
+{
+    PVBOXGUESTSESSION pSession = (PVBOXGUESTSESSION)pFilp->private_data;
+    int rc;
+#ifndef HAVE_UNLOCKED_IOCTL
+    unlock_kernel();
+#endif
+
+#if 0 /* no fast I/O controls defined atm. */
+    if (RT_LIKELY(   (   uCmd == SUP_IOCTL_FAST_DO_RAW_RUN
+                      || uCmd == SUP_IOCTL_FAST_DO_HM_RUN
+                      || uCmd == SUP_IOCTL_FAST_DO_NOP)
+                  && pSession->fUnrestricted == true))
+        rc = VGDrvCommonIoCtlFast(uCmd, ulArg, &g_DevExt, pSession);
+    else
+#endif
+        rc = vgdrvLinuxIOCtlSlow(pFilp, uCmd, ulArg, pSession);
+
+#ifndef HAVE_UNLOCKED_IOCTL
+    lock_kernel();
+#endif
+    return rc;
+}
+
+
+/**
+ * Device I/O Control entry point, slow variant.
+ *
+ * @param   pFilp       Associated file pointer.
+ * @param   uCmd        The function specified to ioctl().
+ * @param   ulArg       The argument specified to ioctl().
+ * @param   pSession    The session instance.
+ */
+static int vgdrvLinuxIOCtlSlow(struct file *pFilp, unsigned int uCmd, unsigned long ulArg, PVBOXGUESTSESSION pSession)
+{
+    int                 rc;
+    VBGLREQHDR          Hdr;
+    PVBGLREQHDR         pHdr;
+    uint32_t            cbBuf;
+
+    Log6(("vgdrvLinuxIOCtlSlow: pFilp=%p uCmd=%#x ulArg=%p pid=%d/%d\n", pFilp, uCmd, (void *)ulArg, RTProcSelf(), current->pid));
+
+    /*
+     * Read the header.
+     */
+    if (RT_FAILURE(RTR0MemUserCopyFrom(&Hdr, ulArg, sizeof(Hdr))))
+    {
+        Log(("vgdrvLinuxIOCtlSlow: copy_from_user(,%#lx,) failed; uCmd=%#x\n", ulArg, uCmd));
+        return -EFAULT;
+    }
+    if (RT_UNLIKELY(Hdr.uVersion != VBGLREQHDR_VERSION))
+    {
+        Log(("vgdrvLinuxIOCtlSlow: bad header version %#x; uCmd=%#x\n", Hdr.uVersion, uCmd));
+        return -EINVAL;
+    }
+
+    /*
+     * Buffer the request.
+     * Note! The header is revalidated by the common code.
+     */
+    cbBuf = RT_MAX(Hdr.cbIn, Hdr.cbOut);
+    if (RT_UNLIKELY(cbBuf > _1M*16))
+    {
+        Log(("vgdrvLinuxIOCtlSlow: too big cbBuf=%#x; uCmd=%#x\n", cbBuf, uCmd));
+        return -E2BIG;
+    }
+    if (RT_UNLIKELY(   Hdr.cbIn < sizeof(Hdr)
+                    || (cbBuf != _IOC_SIZE(uCmd) && _IOC_SIZE(uCmd) != 0)))
+    {
+        Log(("vgdrvLinuxIOCtlSlow: bad ioctl cbBuf=%#x _IOC_SIZE=%#x; uCmd=%#x\n", cbBuf, _IOC_SIZE(uCmd), uCmd));
+        return -EINVAL;
+    }
+    pHdr = RTMemAlloc(cbBuf);
+    if (RT_UNLIKELY(!pHdr))
+    {
+        LogRel(("vgdrvLinuxIOCtlSlow: failed to allocate buffer of %d bytes for uCmd=%#x\n", cbBuf, uCmd));
+        return -ENOMEM;
+    }
+    if (RT_FAILURE(RTR0MemUserCopyFrom(pHdr, ulArg, Hdr.cbIn)))
+    {
+        Log(("vgdrvLinuxIOCtlSlow: copy_from_user(,%#lx, %#x) failed; uCmd=%#x\n", ulArg, Hdr.cbIn, uCmd));
+        RTMemFree(pHdr);
+        return -EFAULT;
+    }
+    if (Hdr.cbIn < cbBuf)
+        RT_BZERO((uint8_t *)pHdr + Hdr.cbIn, cbBuf - Hdr.cbIn);
+
+    /*
+     * Process the IOCtl.
+     */
+    rc = VGDrvCommonIoCtl(uCmd, &g_DevExt, pSession, pHdr, cbBuf);
+
+    /*
+     * Copy ioctl data and output buffer back to user space.
+     */
+    if (RT_SUCCESS(rc))
+    {
+        uint32_t cbOut = pHdr->cbOut;
+        if (RT_UNLIKELY(cbOut > cbBuf))
+        {
+            LogRel(("vgdrvLinuxIOCtlSlow: too much output! %#x > %#x; uCmd=%#x!\n", cbOut, cbBuf, uCmd));
+            cbOut = cbBuf;
+        }
+        if (RT_FAILURE(RTR0MemUserCopyTo(ulArg, pHdr, cbOut)))
+        {
+            /* this is really bad! */
+            LogRel(("vgdrvLinuxIOCtlSlow: copy_to_user(%#lx,,%#x); uCmd=%#x!\n", ulArg, cbOut, uCmd));
+            rc = -EFAULT;
+        }
+    }
+    else
+    {
+        Log(("vgdrvLinuxIOCtlSlow: pFilp=%p uCmd=%#x ulArg=%p failed, rc=%d\n", pFilp, uCmd, (void *)ulArg, rc));
+        rc = -EINVAL;
+    }
+    RTMemFree(pHdr);
+
+    Log6(("vgdrvLinuxIOCtlSlow: returns %d (pid=%d/%d)\n", rc, RTProcSelf(), current->pid));
+    return rc;
+}
+
+
+/**
+ * @note This code is duplicated on other platforms with variations, so please
+ *       keep them all up to date when making changes!
+ */
+int VBOXCALL VBoxGuestIDC(void *pvSession, uintptr_t uReq, PVBGLREQHDR pReqHdr, size_t cbReq)
+{
+    /*
+     * Simple request validation (common code does the rest).
+     */
+    int rc;
+    if (   RT_VALID_PTR(pReqHdr)
+        && cbReq >= sizeof(*pReqHdr))
+    {
+        /*
+         * All requests except the connect one requires a valid session.
+         */
+        PVBOXGUESTSESSION pSession = (PVBOXGUESTSESSION)pvSession;
+        if (pSession)
+        {
+            if (   RT_VALID_PTR(pSession)
+                && pSession->pDevExt == &g_DevExt)
+                rc = VGDrvCommonIoCtl(uReq, &g_DevExt, pSession, pReqHdr, cbReq);
+            else
+                rc = VERR_INVALID_HANDLE;
+        }
+        else if (uReq == VBGL_IOCTL_IDC_CONNECT)
+        {
+            rc = VGDrvCommonCreateKernelSession(&g_DevExt, &pSession);
+            if (RT_SUCCESS(rc))
+            {
+                rc = VGDrvCommonIoCtl(uReq, &g_DevExt, pSession, pReqHdr, cbReq);
+                if (RT_FAILURE(rc))
+                    VGDrvCommonCloseSession(&g_DevExt, pSession);
+            }
+        }
+        else
+            rc = VERR_INVALID_HANDLE;
+    }
+    else
+        rc = VERR_INVALID_POINTER;
+    return rc;
+}
+EXPORT_SYMBOL(VBoxGuestIDC);
+
+
+/**
+ * Asynchronous notification activation method.
+ *
+ * @returns 0 on success, negative errno on failure.
+ *
+ * @param   fd          The file descriptor.
+ * @param   pFile       The file structure.
+ * @param   fOn         On/off indicator.
+ */
+static int vgdrvLinuxFAsync(int fd, struct file *pFile, int fOn)
+{
+    return fasync_helper(fd, pFile, fOn, &g_pFAsyncQueue);
+}
+
+
+/**
+ * Poll function.
+ *
+ * This returns ready to read if the mouse pointer mode or the pointer position
+ * has changed since last call to read.
+ *
+ * @returns 0 if no changes, POLLIN | POLLRDNORM if there are unseen changes.
+ *
+ * @param   pFile       The file structure.
+ * @param   pPt         The poll table.
+ *
+ * @remarks This is probably not really used, X11 is said to use the fasync
+ *          interface instead.
+ */
+static unsigned int vgdrvLinuxPoll(struct file *pFile, poll_table *pPt)
+{
+    PVBOXGUESTSESSION   pSession  = (PVBOXGUESTSESSION)pFile->private_data;
+    uint32_t            u32CurSeq = ASMAtomicUoReadU32(&g_DevExt.u32MousePosChangedSeq);
+    unsigned int        fMask     = pSession->u32MousePosChangedSeq != u32CurSeq
+                                  ? POLLIN | POLLRDNORM
+                                  : 0;
+    poll_wait(pFile, &g_PollEventQueue, pPt);
+    return fMask;
+}
+
+
+/**
+ * Read to go with our poll/fasync response.
+ *
+ * @returns 1 or -EINVAL.
+ *
+ * @param   pFile       The file structure.
+ * @param   pbBuf       The buffer to read into.
+ * @param   cbRead      The max number of bytes to read.
+ * @param   poff        The current file position.
+ *
+ * @remarks This is probably not really used as X11 lets the driver do its own
+ *          event reading. The poll condition is therefore also cleared when we
+ *          see VMMDevReq_GetMouseStatus in vgdrvIoCtl_VMMRequest.
+ */
+static ssize_t vgdrvLinuxRead(struct file *pFile, char *pbBuf, size_t cbRead, loff_t *poff)
+{
+    PVBOXGUESTSESSION   pSession  = (PVBOXGUESTSESSION)pFile->private_data;
+    uint32_t            u32CurSeq = ASMAtomicUoReadU32(&g_DevExt.u32MousePosChangedSeq);
+
+    if (*poff != 0)
+        return -EINVAL;
+
+    /*
+     * Fake a single byte read if we're not up to date with the current mouse position.
+     */
+    if (    pSession->u32MousePosChangedSeq != u32CurSeq
+        &&  cbRead > 0)
+    {
+        pSession->u32MousePosChangedSeq = u32CurSeq;
+        pbBuf[0] = 0;
+        return 1;
+    }
+    return 0;
+}
+
+
+void VGDrvNativeISRMousePollEvent(PVBOXGUESTDEVEXT pDevExt)
+{
+#ifdef VBOXGUEST_WITH_INPUT_DRIVER
+    int rc;
+#endif
+    NOREF(pDevExt);
+
+    /*
+     * Wake up everyone that's in a poll() and post anyone that has
+     * subscribed to async notifications.
+     */
+    Log3(("VGDrvNativeISRMousePollEvent: wake_up_all\n"));
+    wake_up_all(&g_PollEventQueue);
+    Log3(("VGDrvNativeISRMousePollEvent: kill_fasync\n"));
+    kill_fasync(&g_pFAsyncQueue, SIGIO, POLL_IN);
+#ifdef VBOXGUEST_WITH_INPUT_DRIVER
+    /* Report events to the kernel input device */
+    g_pMouseStatusReq->mouseFeatures = 0;
+    g_pMouseStatusReq->pointerXPos = 0;
+    g_pMouseStatusReq->pointerYPos = 0;
+    rc = VbglR0GRPerform(&g_pMouseStatusReq->header);
+    if (RT_SUCCESS(rc))
+    {
+        input_report_abs(g_pInputDevice, ABS_X,
+                         g_pMouseStatusReq->pointerXPos);
+        input_report_abs(g_pInputDevice, ABS_Y,
+                         g_pMouseStatusReq->pointerYPos);
+# ifdef EV_SYN
+        input_sync(g_pInputDevice);
+# endif
+    }
+#endif
+    Log3(("VGDrvNativeISRMousePollEvent: done\n"));
+}
+
+
+bool VGDrvNativeProcessOption(PVBOXGUESTDEVEXT pDevExt, const char *pszName, const char *pszValue)
+{
+    RT_NOREF(pDevExt); RT_NOREF(pszName); RT_NOREF(pszValue);
+    return false;
+}
+
+
+#if RTLNX_VER_MIN(2,6,0)
+
+/** log and dbg_log parameter setter. */
+static int vgdrvLinuxParamLogGrpSet(const char *pszValue, CONST_4_15 struct kernel_param *pParam)
+{
+    if (g_fLoggerCreated)
+    {
+        PRTLOGGER pLogger = pParam->name[0] == 'd' ? RTLogDefaultInstance() : RTLogRelGetDefaultInstance();
+        if (pLogger)
+            RTLogGroupSettings(pLogger, pszValue);
+    }
+    else if (pParam->name[0] != 'd')
+        strlcpy(&g_szLogGrp[0], pszValue, sizeof(g_szLogGrp));
+
+    return 0;
+}
+
+/** log and dbg_log parameter getter. */
+static int vgdrvLinuxParamLogGrpGet(char *pszBuf, CONST_4_15 struct kernel_param *pParam)
+{
+    PRTLOGGER pLogger = pParam->name[0] == 'd' ? RTLogDefaultInstance() : RTLogRelGetDefaultInstance();
+    *pszBuf = '\0';
+    if (pLogger)
+        RTLogGetGroupSettings(pLogger, pszBuf, _4K);
+    return strlen(pszBuf);
+}
+
+
+/** log and dbg_log_flags parameter setter. */
+static int vgdrvLinuxParamLogFlagsSet(const char *pszValue, CONST_4_15 struct kernel_param *pParam)
+{
+    if (g_fLoggerCreated)
+    {
+        PRTLOGGER pLogger = pParam->name[0] == 'd' ? RTLogDefaultInstance() : RTLogRelGetDefaultInstance();
+        if (pLogger)
+            RTLogFlags(pLogger, pszValue);
+    }
+    else if (pParam->name[0] != 'd')
+        strlcpy(&g_szLogFlags[0], pszValue, sizeof(g_szLogFlags));
+    return 0;
+}
+
+/** log and dbg_log_flags parameter getter. */
+static int vgdrvLinuxParamLogFlagsGet(char *pszBuf, CONST_4_15 struct kernel_param *pParam)
+{
+    PRTLOGGER pLogger = pParam->name[0] == 'd' ? RTLogDefaultInstance() : RTLogRelGetDefaultInstance();
+    *pszBuf = '\0';
+    if (pLogger)
+        RTLogGetFlags(pLogger, pszBuf, _4K);
+    return strlen(pszBuf);
+}
+
+
+/** log and dbg_log_dest parameter setter. */
+static int vgdrvLinuxParamLogDstSet(const char *pszValue, CONST_4_15 struct kernel_param *pParam)
+{
+    if (g_fLoggerCreated)
+    {
+        PRTLOGGER pLogger = pParam->name[0] == 'd' ? RTLogDefaultInstance() : RTLogRelGetDefaultInstance();
+        if (pLogger)
+            RTLogDestinations(pLogger, pszValue);
+    }
+    else if (pParam->name[0] != 'd')
+        strlcpy(&g_szLogDst[0], pszValue, sizeof(g_szLogDst));
+    return 0;
+}
+
+/** log and dbg_log_dest parameter getter. */
+static int vgdrvLinuxParamLogDstGet(char *pszBuf, CONST_4_15 struct kernel_param *pParam)
+{
+    PRTLOGGER pLogger = pParam->name[0] == 'd' ? RTLogDefaultInstance() : RTLogRelGetDefaultInstance();
+    *pszBuf = '\0';
+    if (pLogger)
+        RTLogGetDestinations(pLogger, pszBuf, _4K);
+    return strlen(pszBuf);
+}
+
+
+/** r3_log_to_host parameter setter. */
+static int vgdrvLinuxParamR3LogToHostSet(const char *pszValue, CONST_4_15 struct kernel_param *pParam)
+{
+    g_DevExt.fLoggingEnabled = VBDrvCommonIsOptionValueTrue(pszValue);
+    return 0;
+}
+
+/** r3_log_to_host parameter getter. */
+static int vgdrvLinuxParamR3LogToHostGet(char *pszBuf, CONST_4_15 struct kernel_param *pParam)
+{
+    strcpy(pszBuf, g_DevExt.fLoggingEnabled ? "enabled" : "disabled");
+    return strlen(pszBuf);
+}
+
+
+/*
+ * Define module parameters.
+ */
+module_param_call(log,            vgdrvLinuxParamLogGrpSet,   vgdrvLinuxParamLogGrpGet,   NULL, 0664);
+module_param_call(log_flags,      vgdrvLinuxParamLogFlagsSet, vgdrvLinuxParamLogFlagsGet, NULL, 0664);
+module_param_call(log_dest,       vgdrvLinuxParamLogDstSet,   vgdrvLinuxParamLogDstGet,   NULL, 0664);
+# ifdef LOG_ENABLED
+module_param_call(dbg_log,        vgdrvLinuxParamLogGrpSet,   vgdrvLinuxParamLogGrpGet,   NULL, 0664);
+module_param_call(dbg_log_flags,  vgdrvLinuxParamLogFlagsSet, vgdrvLinuxParamLogFlagsGet, NULL, 0664);
+module_param_call(dbg_log_dest,   vgdrvLinuxParamLogDstSet,   vgdrvLinuxParamLogDstGet,   NULL, 0664);
+# endif
+module_param_call(r3_log_to_host, vgdrvLinuxParamR3LogToHostSet, vgdrvLinuxParamR3LogToHostGet, NULL, 0664);
+
+#endif /* 2.6.0 and later */
+
+
+module_init(vgdrvLinuxModInit);
+module_exit(vgdrvLinuxModExit);
+
+MODULE_AUTHOR(VBOX_VENDOR);
+MODULE_DESCRIPTION(VBOX_PRODUCT " Guest Additions for Linux Module");
+MODULE_LICENSE("GPL");
+#ifdef MODULE_VERSION
+MODULE_VERSION(VBOX_VERSION_STRING " r" RT_XSTR(VBOX_SVN_REV));
+#endif
+