This change makes it simpler to write mount tests that examine the optional
field of mountinfo. It also cleans up the mounts in existing tests so they
don't pollute mountinfo after exit.
PiperOrigin-RevId: 561439839
This change implements only the basic functions of mount namespaces.
All features that depends on user namespaces will be implemented separately.
PiperOrigin-RevId: 552673896
Before this fix, the vd provided to connectMountAt wouldn't be consumed if
we ran into a dead mount, the caller was responsible. This dead mount
could be rooted at a different vd than the one provided, which meant that we
were DecRef'ing the wrong vd if the method returned an error. The method now
always consumes the vd to simplify its use.
Reported-by: [email protected]
PiperOrigin-RevId: 532636615
This is the first in a series of changes that will enabled shared mount
subtrees. This will emulate the Linux kernel implementation described here:
https://www.kernel.org/doc/Documentation/filesystems/sharedsubtree.txt
Mounts can now a be part of a mount groups. These groups are
identified by ids assigned by the VFS. Group IDs can be reused after all
mounts that are a member of that group are destroyed.
Newly created mounts inherit the propagation type of their parent mount. A
mount created on a shared mount will also be a shared mount, though it will
be part of a different group. A shared mount 'A' bound to a shared mount 'B'
will replicate that bind to all of 'B's peers. These new mounts will all be
part of a new shared peer group.
Unmounting a mount 'A' that is a direct child of a shared mount 'B' mounted
at dentry 'b' will propagate that event to all its peers. So peers B1, B2,
B3, etc will all unmount the mounts (A1, A2...) located at 'b'. However, if
any of peers of A have children, they are skipped. If the original mount A
has children, the mount is failed entirely.
The initial root mount has propagation type MS_PRIVATE.
This change only implements a basic version of mount groups. Notably it does
not implement MS_SLAVE, MS_UNBINDABLE, or MS_REC.
PiperOrigin-RevId: 483792757
Implementing memory allocation taking into consideration the fact
that the page allocation does not changes whenever the size of a
file changes. For instance, ftruncate may increase the size of a
file without actually allocating that size. So it changes when the
pages are actually allocated.
Also making amends to allow for partial allocations. For instance,
a write, mmap can allocate pages as long as the size option
is not exhausted.
Fixes#7827
PiperOrigin-RevId: 470426017
Adding test case to check empty size parsing in Linux.
If a value is not passed with size the mount(2) should return EINVAL.
Handling empty size parsing in gVisor.
PiperOrigin-RevId: 448132149
Fixing partial write memory allocation.
Accounting for memory allocation in hard links.
Adding test cases to check for hard link memory allocation for the same.
PiperOrigin-RevId: 443742013
When tmpfs is mounted with `size` option it imposes a limit for
the tmpfs mount. The total size of all files in that mount must not exceed this
limit. See tmpfs(5) for more details. When this option was not supported and
specifying it caused mount(2) to return EINVAL instead of ENOSPC.
In Linux, tmpfs charges symlinks, regular files and directories against this
`size` limit. All accounting is done on the granularity of page size. The `size`
limit and all individual file sizes are rounded up to number of pages while
performing calculations. gVisor has aimed to replicate the same behavior.
One difference from Linux is that if the `size` option is not specified, then
Linux uses 50% of physical RAM as the `size`. We don't replicate this behavior
in gVisor as the actual size of host physical RAM should not be exposed to
application. In gVisor, if `size` option is not specified, then no limit is
imposed. This is consistent with the behavior before this change.
Added syscall tests for regular files and symlink.
PiperOrigin-RevId: 442686814
Linux exempts special files (sockets, named pipes, and character/block device
special files) from mount writability checks in both open
(fs/open.c:do_dentry_open()) and access (fs/open.c:do_faccessat()) paths. We
don't currently do so in open (in VFS1 or VFS2) or access (in VFS1), so we
don't do so in VFS2 access either.
PiperOrigin-RevId: 439398002
Inode number consistency checks are now skipped in save/restore tests for
reasons described in greatest detail in StatTest.StateDoesntChangeAfterRename.
They pass in VFS1 due to the bug described in new test case
SimpleStatTest.DifferentFilesHaveDifferentDeviceInodeNumberPairs.
Fixes#1663
PiperOrigin-RevId: 338776148
Overlayfs does not persist a directory's inode number even while it is mounted.
See fs/overlayfs/inode.c:ovl_map_dev_ino(). VFS2 generates a new inode number
for directories everytime in lookup.
PiperOrigin-RevId: 331045037
Before kernel version 4.16-rc6, fuse mount is protected by
capable(CAP_SYS_ADMIN). After this version, it uses
ns_capable(CAP_SYS_ADMIN) to protect. Before the 4.16 kernel,
it was not allowed to mount fuse file systems without the
global CAP_SYS_ADMIN.
Fixes#3360
Inode ids are only stable across Save/Restore if we have an open FD on the
inode. All tests that compare inode ids must therefor hold an FD open.
PiperOrigin-RevId: 254086603
Based on the guidelines at
https://opensource.google.com/docs/releasing/authors/.
1. $ rg -l "Google LLC" | xargs sed -i 's/Google LLC.*/The gVisor Authors./'
2. Manual fixup of "Google Inc" references.
3. Add AUTHORS file. Authors may request to be added to this file.
4. Point netstack AUTHORS to gVisor AUTHORS. Drop CONTRIBUTORS.
Fixes#209
PiperOrigin-RevId: 245823212
Change-Id: I64530b24ad021a7d683137459cafc510f5ee1de9
