Linux: release only the volume's own loop and device-mapper devices

Unmounting released the loop device and device-mapper devices recorded
at mount time, by name. After a dismount that failed partway, for
example on a busy auxiliary FUSE mount, these names could be reused by
unrelated objects, which the listing then showed as the volume's and a
retry, emergency cleanup, filesystem check or auto-dismount released.
Without kernel cryptography, the loop device, which carries the
decrypted volume, was also released while the filesystem was mounted
elsewhere or the device was held open. A failed mount could remove a
pre-existing device-mapper device of the same name, and a failed
cascade mount could leave its layers and loop device behind.

Each destructive step now acts only on a device verified to belong to
the volume. A loop device must be bound to the container or volume
image (LOOP_GET_STATUS64). Device-mapper devices get the UUID
VERACRYPT-<serial>-<name> and are removed by UUID; those of earlier
versions, which have none, are verified through their sysfs slaves,
and the dismount fails, saying why, if that is not possible. Emergency
cleanup is not offered then, as it would hide the volume but leave its
devices. While the filesystem is mounted elsewhere, the dismount fails
and keeps the volume's devices. With kernel cryptography, the loop
device is marked for autoclear as soon as it is attached and verified,
so the kernel detaches it once device-mapper releases it; a process
holding it open sees only ciphertext and does not fail the dismount.
Without kernel cryptography, the dismount also fails while the loop
device is held open. Dismount, emergency cleanup, filesystem check and
the listing re-read the volume's devices from its auxiliary mount, and
the English emergency warning mentions these causes.

Add Tests/test_linux_loop_identity.py, covering these cases with and
without kernel cryptography, and with --legacy-binary, volumes mounted
by an earlier version. All 89 checks pass, as do the 11 added by
--legacy-binary 1.26.29; 1.26.29 itself fails 21 of the 89.

Author: idrassi

Tests/test_linux_loop_identity.py

@@ -1666,7 +1666,7 @@
     <entry lang="en" key="LINUX_PREF_MOUNT_NTFS_WITH_KERNEL_DRIVER">Mount NTFS volumes with an in-kernel Linux driver</entry>
     <entry lang="en" key="LINUX_PREF_MOUNT_NTFS_WITH_KERNEL_DRIVER_HELP">Linux only. When enabled and no explicit filesystem type was supplied, VeraCrypt probes the decrypted virtual device with blkid -p and mounts detected NTFS filesystems with an available in-kernel NTFS driver, bypassing mount helpers such as ntfs-3g. VeraCrypt uses ntfs when it is positively identified as a modern read/write driver or expected on Linux 7.1 or later, and otherwise uses ntfs3. If NTFS detection fails, VeraCrypt uses the normal automatic filesystem selection. If no supported in-kernel NTFS driver is available or loadable, mounting fails. This opt-in option can avoid suspend or hibernate hangs caused by frozen user-space FUSE filesystems.</entry>
     <entry lang="en" key="LINUX_KERNEL_NTFS_DRIVER_UNAVAILABLE">No supported in-kernel NTFS driver is available or loadable. To use the system default NTFS backend, disable the NTFS kernel-driver preference or do not request kernel NTFS explicitly.</entry>
-    <entry lang="en" key="LINUX_EMERGENCY_UNMOUNT_WARNING">Normal unmount of volume {0} failed. This can happen when applications still have files or directories open on the volume, or when the backing device was disconnected and the mount became stale.\n\nIf the device is still connected, choose No, close applications using the volume, and try unmounting again.\n\nIf the device was disconnected or the mount is stale, VeraCrypt can attempt emergency cleanup by lazy-detaching the filesystem and removing or scheduling removal of VeraCrypt kernel objects. Pending writes may have failed, data may be lost, and cleanup may remain pending until applications close open files. Check the filesystem with fsck or the appropriate repair tool before using it again.\n\nContinue?</entry>
+    <entry lang="en" key="LINUX_EMERGENCY_UNMOUNT_WARNING">Normal unmount of volume {0} failed. This can happen when applications still have files or directories open on the volume, when the volume's filesystem is also mounted elsewhere (for example, a bind mount or a mount in another mount namespace), when another process has the volume's device open, or when the backing device was disconnected and the mount became stale.\n\nIf the device is still connected, choose No, close applications using the volume, unmount any other mounts of its filesystem, and try unmounting again.\n\nIf the device was disconnected or the mount is stale, VeraCrypt can attempt emergency cleanup by lazy-detaching the filesystem and removing or scheduling removal of VeraCrypt kernel objects. Pending writes may have failed, data may be lost, and cleanup may remain pending until applications close open files. Check the filesystem with fsck or the appropriate repair tool before using it again.\n\nContinue?</entry>
     <entry lang="en" key="LINUX_EMERGENCY_UNMOUNTED">Emergency cleanup for volume {0} has been initiated. If the volume was disconnected, the mount was stale, or there were pending writes, check the filesystem with fsck or the appropriate repair tool before using it again.</entry>
     <entry lang="en" key="FORMAT_STAGE_WRITING_DATA">Creating volume data. Please wait.</entry>
     <entry lang="en" key="FORMAT_STAGE_WRITING_BACKUP_HEADER">Finalizing volume creation: writing backup header.</entry>

src/Common/Language.xml

@@ -306,15 +306,17 @@ namespace VeraCrypt
 #endif
 		catch (NotApplicable &) { }
 
-		// Earlier versions recorded the host device of a device-hosted volume mounted with Linux kernel cryptography as its loop device
-		if (!mountedVolume->LoopDevice.IsEmpty() && wstring (mountedVolume->LoopDevice) != wstring (mountedVolume->Path))
+		// On Linux, DismountNativeVolume detaches the loop device once it is verified to belong to the volume
+#ifndef TC_LINUX
+		if (!mountedVolume->LoopDevice.IsEmpty())
 		{
 			try
 			{
 				DetachLoopDevice (mountedVolume->LoopDevice);
 			}
 			catch (ExecutedProcessFailed&) { }
 		}
+#endif
 
 		if (syncVolumeInfo || mountedVolume->Protection == VolumeProtection::HiddenVolumeReadOnly)
 		{
@@ -411,25 +413,6 @@ namespace VeraCrypt
 				firstException.reset (e.CloneNew());
 		}
 
-		// Earlier versions recorded the host device of a device-hosted volume mounted with kernel cryptography as its loop device
-		if (!mountedVolume->LoopDevice.IsEmpty() && wstring (mountedVolume->LoopDevice) != wstring (mountedVolume->Path))
-		{
-			try
-			{
-				DetachLoopDevice (mountedVolume->LoopDevice);
-			}
-			catch (ExecutedProcessFailed &e)
-			{
-				if (IsLoopDeviceAttached (mountedVolume->LoopDevice) && !firstException.get())
-					firstException.reset (e.CloneNew());
-			}
-			catch (Exception &e)
-			{
-				if (!firstException.get())
-					firstException.reset (e.CloneNew());
-			}
-		}
-
 		if (!mountedVolume->AuxMountPoint.IsEmpty())
 		{
 			bool auxMountPointMounted = true;
@@ -725,16 +708,16 @@ namespace VeraCrypt
 			mountedVol->AuxMountPoint = mf.MountPoint;
 
 #ifndef TC_MACOSX
+			// Recorded devices may have been released and reused since
+			UpdateMountedVolumeInfo (mountedVol);
+
 			if (mountedVol->MountPoint.IsEmpty() && !mountedVol->VirtualDevice.IsEmpty())
 			{
 				MountedFilesystemList mpl = GetMountedFilesystems (mountedVol->VirtualDevice);
 
 				if (mpl.size() > 0)
 					mountedVol->MountPoint = mpl.front()->MountPoint;
 			}
-
-			if (mountedVol->MountPoint.IsEmpty() || mountedVol->VirtualDevice.IsEmpty())
-				UpdateMountedVolumeInfo (mountedVol);
 #endif
 
 			volumes.push_back (mountedVol);

src/Core/Unix/CoreUnix.cpp

@@ -63,7 +63,6 @@ namespace VeraCrypt
 #ifdef TC_LINUX
 		virtual void DismountFilesystemLazy (const DirectoryPath &mountPoint) const;
 		virtual void DismountNativeVolumeDeferred (shared_ptr <VolumeInfo> mountedVolume) const { DismountNativeVolume (mountedVolume); }
-		virtual bool IsLoopDeviceAttached (const DevicePath &devicePath) const { return devicePath.IsBlockDevice(); }
 #endif
 		virtual bool FilesystemSupportsUnixPermissions (const DevicePath &devicePath) const;
 		virtual string GetDefaultMountPointPrefix () const;