summaryrefslogtreecommitdiffhomepage
path: root/pkg/sentry/fsimpl
AgeCommit message (Collapse)Author
2020-10-23Check for verity file/Merkle file when reopenChong Cai
Even if the child dentry has been cached, we should still check whether the file and the corresponding Merkle tree file exist as expected. This ensures that we can detect deletion/renaming of files that have been previous enabled and opened. Also make all verification failures to return EIO. This helps to test verify failures. PiperOrigin-RevId: 338709055
2020-10-23Rewrite reference leak checker without finalizers.Dean Deng
Our current reference leak checker uses finalizers to verify whether an object has reached zero references before it is garbage collected. There are multiple problems with this mechanism, so a rewrite is in order. With finalizers, there is no way to guarantee that a finalizer will run before the program exits. When an unreachable object with a finalizer is garbage collected, its finalizer will be added to a queue and run asynchronously. The best we can do is run garbage collection upon sandbox exit to make sure that all finalizers are enqueued. Furthermore, if there is a chain of finalized objects, e.g. A points to B points to C, garbage collection needs to run multiple times before all of the finalizers are enqueued. The first GC run will register the finalizer for A but not free it. It takes another GC run to free A, at which point B's finalizer can be registered. As a result, we need to run GC as many times as the length of the longest such chain to have a somewhat reliable leak checker. Finally, a cyclical chain of structs pointing to one another will never be garbage collected if a finalizer is set. This is a well-known issue with Go finalizers (https://github.com/golang/go/issues/7358). Using leak checking on filesystem objects that produce cycles will not work and even result in memory leaks. The new leak checker stores reference counted objects in a global map when leak check is enabled and removes them once they are destroyed. At sandbox exit, any remaining objects in the map are considered as leaked. This provides a deterministic way of detecting leaks without relying on the complexities of finalizers and garbage collection. This approach has several benefits over the former, including: - Always detects leaks of objects that should be destroyed very close to sandbox exit. The old checker very rarely detected these leaks, because it relied on garbage collection to be run in a short window of time. - Panics if we forgot to enable leak check on a ref-counted object (we will try to remove it from the map when it is destroyed, but it will never have been added). - Can store extra logging information in the map values without adding to the size of the ref count struct itself. With the size of just an int64, the ref count object remains compact, meaning frequent operations like IncRef/DecRef are more cache-efficient. - Can aggregate leak results in a single report after the sandbox exits. Instead of having warnings littered in the log, which were non-deterministically triggered by garbage collection, we can print all warning messages at once. Note that this could also be a limitation--the sandbox must exit properly for leaks to be detected. Some basic benchmarking indicates that this change does not significantly affect performance when leak checking is enabled, which is understandable since registering/unregistering is only done once for each filesystem object. Updates #1486. PiperOrigin-RevId: 338685972
2020-10-19Fix reference counting on kcov mappings.Dean Deng
Reported-by: syzbot+078580ce5dd6d607fcd8@syzkaller.appspotmail.com Reported-by: syzbot+2096681f6891e7bf8aed@syzkaller.appspotmail.com PiperOrigin-RevId: 337973519
2020-10-19Fix runsc tests on VFS2 overlay.Jamie Liu
- Check the sticky bit in overlay.filesystem.UnlinkAt(). Fixes StickyTest.StickyBitPermDenied. - When configuring a VFS2 overlay in runsc, copy the lower layer's root owner/group/mode to the upper layer's root (as in the VFS1 equivalent, boot.addOverlay()). This makes the overlay root owned by UID/GID 65534 with mode 0755 rather than owned by UID/GID 0 with mode 01777. Fixes CreateTest.CreateFailsOnUnpermittedDir, which assumes that the test cannot create files in /. - MknodTest.UnimplementedTypesReturnError assumes that the creation of device special files is not supported. However, while the VFS2 gofer client still doesn't support device special files, VFS2 tmpfs does, and in the overlay test dimension mknod() targets a tmpfs upper layer. The test initially has all capabilities, including CAP_MKNOD, so its creation of these files succeeds. Constrain these tests to VFS1. - Rename overlay.nonDirectoryFD to overlay.regularFileFD and only use it for regular files, using the original FD for pipes and device special files. This is more consistent with Linux (which gets the original inode_operations, and therefore file_operations, for these file types from ovl_fill_inode() => init_special_inode()) and fixes remaining mknod and pipe tests. - Read/write 1KB at a time in PipeTest.Streaming, rather than 4 bytes. This isn't strictly necessary, but it makes the test less obnoxiously slow on ptrace. Fixes #4407 PiperOrigin-RevId: 337971042
2020-10-19[vfs2] Fix fork reference leaks.Dean Deng
PiperOrigin-RevId: 337919424
2020-10-15Change verity isEnable to be a member of dentryChong Cai
PiperOrigin-RevId: 337384146
2020-10-13Don't read beyond EOF when inserting into sentry page cache.Jamie Liu
The sentry page cache stores file contents at page granularity; this is necessary for memory mappings. Thus file offset ranges passed to fsutil.FileRangeSet.Fill() must be page-aligned. If the read callback passed to Fill() returns (partial read, nil error) when reading up to EOF (which is the case for p9.ClientFile.ReadAt() since 9P's Rread cannot convey both a partial read and EOF), Fill() will re-invoke the read callback to try to read from EOF to the end of the containing page, which is harmless but needlessly expensive. Fix this by handling file size explicitly in fsutil.FileRangeSet.Fill(). PiperOrigin-RevId: 336934075
2020-10-13[vfs2] Don't take reference in Task.MountNamespaceVFS2 and MountNamespace.Root.Dean Deng
This fixes reference leaks related to accidentally forgetting to DecRef() after calling one or the other. PiperOrigin-RevId: 336918922
2020-10-13[vfs2] Destroy all tmpfs files when the filesystem is released.Dean Deng
In addition to fixing reference leaks, this change also releases memory used by regular tmpfs files once the containing filesystem is released. PiperOrigin-RevId: 336833111
2020-10-13[vfs2] Add FilesystemType.Release to avoid reference leaks.Dean Deng
Singleton filesystem like devpts and devtmpfs have a single filesystem shared among all mounts, so they acquire a "self-reference" when initialized that must be released when the entire virtual filesystem is released at sandbox exit. PiperOrigin-RevId: 336828852
2020-10-12Change verity mu to be per file systemChong Cai
verity Mu should be per file system instead of global, so that enabling and verifying in different file systems won't block each other. Also Lock verity Mu in PRead. PiperOrigin-RevId: 336779356
2020-10-12Change Merkle tree library to use ReaderAtChong Cai
Merkle tree library was originally using Read/Seek to access data and tree, since the parameters are io.ReadSeeker. This could cause race conditions if multiple threads accesses the same fd to read. Here we change to use ReaderAt, and implement it with PRead to make it thread safe. PiperOrigin-RevId: 336779260
2020-10-12[vfs] kernfs: Fix inode memory leak issue.Ayush Ranjan
This change aims to fix the memory leak issue reported in #3933. Background: VFS2 kernfs kept accumulating invalid dentries if those dentries were not walked on. After substantial consideration of the problem by our team, we decided to have an LRU cache solution. This change is the first part to that solution, where we don't cache anything. The LRU cache can be added on top of this. What has changed: - Introduced the concept of an inode tree in kernfs.OrderedChildren. This is helpful is cases where the lifecycle of an inode is different from that of a dentry. - OrderedChildren now deals with initialized inodes instead of initialized dentries. It now implements Lookup() where it constructs a new dentry using the inode. - OrderedChildren holds a ref on all its children inodes. With this change, now an inode can "outlive" a dentry pointing to it. See comments in kernfs.OrderedChildren. - The kernfs dentry tree is solely maintained by kernfs only. Inode implementations can not modify the dentry tree. - Dentries that reach ref count 0 are removed from the dentry tree. - revalidateChildLocked now defer-DecRefs the newly created dentry from Inode.Lookup(), limiting its life to the current filesystem operation. If refs are picked on the dentry during the FS op (via an FD or something), then it will stick around and will be removed when the FD is closed. So there is essentially _no caching_ for Look()ed up dentries. - kernfs.DecRef does not have the precondition that fs.mu must be locked. Fixes #3933 PiperOrigin-RevId: 336768576
2020-10-09Include stat in Verity hashChong Cai
PiperOrigin-RevId: 336395445
2020-10-03Fix kcov enabling and disabling procedures.Dean Deng
- When the KCOV_ENABLE_TRACE ioctl is called with the trace kind KCOV_TRACE_PC, the kcov mode should be set to KCOV_*MODE*_TRACE_PC. - When the owning task of kcov exits, the memory mapping should not be cleared so it can be used by other tasks. - Add more tests (also tested on native Linux kcov). PiperOrigin-RevId: 335202585
2020-10-02Convert uses of the binary package in kernel to go-marshal.Rahat Mahmood
PiperOrigin-RevId: 335077195
2020-10-01Add a verity test for modified parent Merkle fileChong Cai
When a child's root hash or its Merkle path is modified in its parent's Merkle tree file, opening the file should fail, provided the directory is verity enabled. The test for this behavior is added. PiperOrigin-RevId: 334963690
2020-09-30Merge pull request #3824 from btw616:fix/issue-3823gVisor bot
PiperOrigin-RevId: 334721453
2020-09-30Implement ioctl with measure in verity fsChong Cai
PiperOrigin-RevId: 334682753
2020-09-30Internal change.Chong Cai
PiperOrigin-RevId: 334678513
2020-09-30[go-marshal] Port ext codebase to use go marshal.Ayush Ranjan
PiperOrigin-RevId: 334656292
2020-09-30Add verity fs testsChong Cai
The tests confirms that when a file is opened in verity, the corresponding Merkle trees are generated. Also a normal read succeeds on verity enabled files, but fails if either the verity file or the Merkle tree file is modified. PiperOrigin-RevId: 334640331
2020-09-29Add /proc/[pid]/cwdFabricio Voznika
PiperOrigin-RevId: 334478850
2020-09-28Support inotify in overlayfs.Dean Deng
Fixes #1479, #317. PiperOrigin-RevId: 334258052
2020-09-27Fix kernfs race condition.Dean Deng
Do not release dirMu between checking whether to create a child and actually inserting it. Also fixes a bug in fusefs which was causing it to deadlock under the new lock ordering. We do not need to call kernfs.Dentry.InsertChild from newEntry because it will always be called at the kernfs filesystem layer. Updates #1193. PiperOrigin-RevId: 334049264
2020-09-24[vfs] kernfs: Do not hold reference on the inode when opening FD.Ayush Ranjan
The FD should hold a reference on the dentry they were opened on which in turn holds a reference on the inode it points to. PiperOrigin-RevId: 333589223
2020-09-24[vfs] [2/2] kernfs: kernfs: Internally use kernfs.Dentry instead of vfs.Dentry.Ayush Ranjan
Update signatures for: - All methods in inodeDirectory - deferDecRef() and Filesystem.droppedDentries - newSyntheticDirectory() - `slot`s used in OrderedChildren and subsequent methods like replaceChildLocked() and checkExistingLocked() - stepExistingLocked(), walkParentDirLocked(), checkCreateLocked() Updates #1193 PiperOrigin-RevId: 333558866
2020-09-24Add basic stateify annotations.Adin Scannell
Updates #1663 PiperOrigin-RevId: 333539293
2020-09-23fuse: don't call dentry.InsertChildAndrei Vagin
It is called from the kernfs code (OpenAt and revalidateChildLocked()). For RemoveChildLocked, it is opposed. We need to call it from fuse.RmDir and fuse.Unlink. PiperOrigin-RevId: 333453218
2020-09-23[vfs] kernfs: Enable leak checking consistently.Ayush Ranjan
There were some instances where we were not enabling leak checking. PiperOrigin-RevId: 333418571
2020-09-23Let underlying fs handle LockFD in verity fsChong Cai
PiperOrigin-RevId: 333412836
2020-09-23Set verity underlying fs mount as internalChong Cai
PiperOrigin-RevId: 333404727
2020-09-22[vfs] [1/2] kernfs: Internally use kernfs.Dentry instead of vfs.Dentry.Ayush Ranjan
Update signatures for: - walkExistingLocked - checkDeleteLocked - Inode.Open Updates #1193 PiperOrigin-RevId: 333163381
2020-09-21Allow partial writes for gofer.specialFileFD.Dean Deng
Originally, we avoided partial writes in case it caused us to write a partial packet to a socket-backed specialFileFD. However, this check causes splicing from a pipe to specialFileFD to fail if we hit EOF on the pipe. PiperOrigin-RevId: 333016216
2020-09-21Use kernfs.Dentry for kernfs.Lookup.Dean Deng
Updates #1193. PiperOrigin-RevId: 332939026
2020-09-20Merge pull request #3651 from ianlewis:ip-forwardinggVisor bot
PiperOrigin-RevId: 332760843
2020-09-18Implement fsimpl/overlay.filesystem.RenameAt.Jamie Liu
Updates #1199 PiperOrigin-RevId: 332539197
2020-09-18Use a tmpfs file for shared anonymous and /dev/zero mmap on VFS2.Jamie Liu
This is more consistent with Linux (see comment on MM.NewSharedAnonMappable()). We don't do the same thing on VFS1 for reasons documented by the updated comment. PiperOrigin-RevId: 332514849
2020-09-18Merge pull request #3972 from btw616:fix/commentsgVisor bot
PiperOrigin-RevId: 332486111
2020-09-17fuse.DeviceFD needs to hold a reference on the associated filesystem.Rahat Mahmood
This fixes a use-after-free in fuse.DeviceFD.Release. PiperOrigin-RevId: 332394146
2020-09-17Change sizeofInt32 to string sizeChong Cai
This constant is used to represent int32 stored in file xattrs. The integers are stored as strings there, so the real size should be the string size (number of digits) instead of an int size (4 bytes). PiperOrigin-RevId: 332353217
2020-09-17Set mode when creating Merkle tree fileChong Cai
PiperOrigin-RevId: 332340342
2020-09-17Complete vfs2 implementation of fallocate.Dean Deng
This change includes overlay, special regular gofer files, and hostfs. Fixes #3589. PiperOrigin-RevId: 332330860
2020-09-17Fix kernfs unlinkat and rmdirat incorrect resolved path nameJinmou Li
2020-09-17fsimpl: improve the "implements" commentsTiwei Bie
As noticed by @ayushr2, the "implements" comments are not consistent, e.g. // IterDirents implements kernfs.inodeDynamicLookup. // Generate implements vfs.DynamicBytesSource.Generate. This patch improves this by making the comments like this consistently include the package name (when the interface and struct are not in the same package) and method name. Signed-off-by: Tiwei Bie <tiwei.btw@antgroup.com>
2020-09-16Merge pull request #3934 from avagin:feature/fusegVisor bot
PiperOrigin-RevId: 332122081
2020-09-16Implement OpenAt() for verity fsChong Cai
OpenAt() for verity fs is implemented by opening both the target file or directory and the corresponding Merkle tree file in the underlying file system. Generally they are only open for read. In allowRuntimeEnable mode, the Merkle tree file is also open for write. PiperOrigin-RevId: 332116423
2020-09-16Implement PRead for verity fsChong Cai
PRead is implemented by read from the underlying file in blocks, and verify each block. The verified contents are saved into the output buffer. PiperOrigin-RevId: 332092267
2020-09-16fuse: fix data race in fusefs Release()Jinmou Li
fix #3956
2020-09-16fuse: fix FUSE_RELEASE reply handlingJinmou Li
fix #3963
generated by cgit v1.2.3 (git 2.39.1) at 2024-11-19 00:21:39 +0000