| Age | Commit message (Collapse) | Author |
|---|
|
Run vs. exec, VFS1 vs. VFS2 were executable lookup were
slightly different from each other. Combine them all
into the same logic.
PiperOrigin-RevId: 315426443
|
|
As in VFS1, the mode, uid, and gid options are supported.
Updates #1197
PiperOrigin-RevId: 315340510
|
|
Netstack has traditionally parsed headers on-demand as a packet moves up the
stack. This is conceptually simple and convenient, but incompatible with
iptables, where headers can be inspected and mangled before even a routing
decision is made.
This changes header parsing to happen early in the incoming packet path, as soon
as the NIC gets the packet from a link endpoint. Even if an invalid packet is
found (e.g. a TCP header of insufficient length), the packet is passed up the
stack for proper stats bookkeeping.
PiperOrigin-RevId: 315179302
|
|
This is mostly syscall plumbing, VFS2 already implements the internals of
mounts. In addition to the syscall defintions, the following mount-related
mechanisms are updated:
- Implement MS_NOATIME for VFS2, but only for tmpfs and goferfs. The other VFS2
filesystems don't implement node-level timestamps yet.
- Implement the 'mode', 'uid' and 'gid' mount options for VFS2's tmpfs.
- Plumb mount namespace ownership, which is necessary for checking appropriate
capabilities during mount(2).
Updates #1035
PiperOrigin-RevId: 315035352
|
|
This analysis also catches a potential bug, which is a split on mapPhysical.
This would have led to potential guest-exit during Mapping (although this
would have been handled by the now-unecessary retryInGuest loop).
PiperOrigin-RevId: 315025106
|
|
This change has multiple small components.
First, the chunk size is bumped to 1GB in order to avoid creating excessive
VMAs in the Sentry, which can lead to VMA exhaustion (and hitting limits).
Second, gap-tracking is added to the usage set in order to efficiently scan
for available regions.
Third, reclaim is moved to a simple segment set. This is done to allow the
order of reclaim to align with the Allocate order (which becomes much more
complex when trying to track a "max page" as opposed to "min page", so we
just track explicit segments instead, which should make reclaim scanning
faster anyways).
Finally, the findAvailable function attempts to scan from the top-down, in
order to maximize opportunities for VMA merging in applications (hopefully
preventing the same VMA exhaustion that can affect the Sentry).
PiperOrigin-RevId: 315009249
|
|
The current task can share its fdtable with a few other tasks,
but after exec, this should be a completely separate process.
PiperOrigin-RevId: 314999565
|
|
For TCP sockets gVisor incorrectly returns EAGAIN when no ephemeral ports are
available to bind during a connect. Linux returns EADDRNOTAVAIL. This change
fixes gVisor to return the correct code and adds a test for the same.
This change also fixes a minor bug for ping sockets where connect() would fail
with EINVAL unless the socket was bound first.
Also added tests for testing UDP Port exhaustion and Ping socket port
exhaustion.
PiperOrigin-RevId: 314988525
|
|
IPTables.connections contains a sync.RWMutex. Copying it will trigger copylocks
analysis. Tested by manually enabling nogo tests.
sync.RWMutex is added to IPTables for the additional race condition discovered.
PiperOrigin-RevId: 314817019
|
|
Historically we've been passing PacketBuffer by shallow copying through out
the stack. Right now, this is only correct as the caller would not use
PacketBuffer after passing into the next layer in netstack.
With new buffer management effort in gVisor/netstack, PacketBuffer will
own a Buffer (to be added). Internally, both PacketBuffer and Buffer may
have pointers and shallow copying shouldn't be used.
Updates #2404.
PiperOrigin-RevId: 314610879
|
|
We must hold f.mu to write f.offset.
PiperOrigin-RevId: 314582968
|
|
PiperOrigin-RevId: 314570894
|
|
Updates #179
PiperOrigin-RevId: 314563830
|
|
PiperOrigin-RevId: 314450191
|
|
PiperOrigin-RevId: 314449030
|
|
PiperOrigin-RevId: 314415253
|
|
This change adds more information about what needs to be done
to implement `/dev/fuse`
|
|
Splice, setxattr and removexattr should generate events. Note that VFS2 already
generates events for extended attributes.
Updates #1479.
PiperOrigin-RevId: 314244261
|
|
gofer.filesystem.createAndOpenChildLocked() doesn't need to take a reference on
the new dentry since vfs.FileDescription.Init() will do so.
PiperOrigin-RevId: 314242127
|
|
Using tee instead of read to detect when a O_RDONLY|O_NONBLOCK pipe FD has a
writer circumvents the problem of what to do with the byte read from the pipe,
avoiding much of the complexity of the fdpipe package.
PiperOrigin-RevId: 314216146
|
|
PiperOrigin-RevId: 314186752
|
|
PiperOrigin-RevId: 313871804
|
|
None of the dependencies have changed in 1.15. It may be possible to simplify
some of the wrappers in rawfile following 1.13, but that can come in a later
change.
PiperOrigin-RevId: 313863264
|
|
PiperOrigin-RevId: 313842690
|
|
Limited to tmpfs. Inotify support in other filesystem implementations to
follow.
Updates #1479
PiperOrigin-RevId: 313828648
|
|
PiperOrigin-RevId: 313817646
|
|
Support in other filesystem impls is still needed. Unlike in Linux and vfs1, we
need to plumb inotify down to each filesystem implementation in order to keep
track of links/inode structures properly.
IN_EXCL_UNLINK still needs to be implemented, as well as a few inotify hooks
that are not present in either vfs1 or vfs2. Those will be addressed in
subsequent changes.
Updates #1479.
PiperOrigin-RevId: 313781995
|
|
This makes it straightforward to create bind mounts internally in VFS2: Given a
bind mount root represented by vfs.VirtualDentry vd:
- Create a new mount with VFS.NewDisconnectedMount(vd.Mount().Filesystem(),
vd.Dentry()).
- Connect the resulting mount in the appropriate namespace with
VFS.ConnectMountAt().
Note that the resulting bind mount is non-recursive; recursive bind mounting
requires explicitly duplicating all children of the original mount, which is
best handled internally by VFS.
Updates #179
PiperOrigin-RevId: 313703963
|
|
|
|
PiperOrigin-RevId: 313600051
|
|
Inotify sends events when a watch target is reaches a link count of 0 (see
include/linux/fsnotify.h:fsnotify_inoderemove). Currently, we do not account
for both dir/ and dir/.. in unlink, causing
syscalls/linux/inotify.cc:WatchTargetDeletionGeneratesEvent to fail because
the expected inotify events are not generated.
Furthermore, we should DecRef() once the inode reaches zero links; otherwise,
we will leak a reference.
PiperOrigin-RevId: 313502091
|
|
|
|
PiperOrigin-RevId: 313332542
|
|
Updates #138
PiperOrigin-RevId: 313326354
|
|
PiperOrigin-RevId: 313300882
|
|
|
|
This change was derived from a change by:
Reapor-Yurnero <reapor.yurnero@gmail.com>
And has been modified by:
Adin Scannell <ascannell@google.com>
(The original change author is preserved for the commit.)
This change implements gap tracking in the segment set by adding additional
information in each node, and using that information to speed up gap finding
from a linear scan to a O(log(n)) walk of the tree.
This gap tracking is optional, and will default to off except for segment
instances that set gapTracking equal to 1 in their const lists.
PiperOrigin-RevId: 312621607
|
|
PiperOrigin-RevId: 312559861
|
|
In VFS1, both fs/host and fs/gofer used the same utils for host file mappings.
Refactor parts of fsimpl/gofer to create similar utils to share with
fsimpl/host (memory accounting code moved to fsutil, page rounding arithmetic
moved to usermem).
Updates #1476.
PiperOrigin-RevId: 312345090
|
|
On native Linux, calling recv/read right after send/write sometimes returns
EWOULDBLOCK, if the data has not made it to the receiving socket (even though
the endpoints are on the same host). Poll before reading to avoid this.
Making this change also uncovered a hostinet bug (gvisor.dev/issue/2726),
which is noted in this CL.
PiperOrigin-RevId: 312320587
|
|
PiperOrigin-RevId: 312299234
|
|
As new functionality is added to VFS2, corresponding files in VFS1
don't need to be changed.
PiperOrigin-RevId: 312153799
|
|
* Aggregate architecture Overview in "What is gVisor?" as it makes more sense
in one place.
* Drop "user-space kernel" and use "application kernel". The term "user-space
kernel" is confusing when some platform implementation do not run in
user-space (instead running in guest ring zero).
* Clear up the relationship between the Platform page in the user guide and the
Platform page in the architecture guide, and ensure they are cross-linked.
* Restore the call-to-action quick start link in the main page, and drop the
GitHub link (which also appears in the top-right).
* Improve image formatting by centering all doc and blog images, and move the
image captions to the alt text.
PiperOrigin-RevId: 311845158
|
|
PiperOrigin-RevId: 311808460
|
|
PiperOrigin-RevId: 311657502
|
|
Closes #2612.
PiperOrigin-RevId: 311548074
|
|
This change adds support for TCP_SYNCNT and TCP_WINDOW_CLAMP options
in GetSockOpt/SetSockOpt. This change does not really change any
behaviour in Netstack and only stores/returns the stored value.
Actual honoring of these options will be added as required.
Fixes #2626, #2625
PiperOrigin-RevId: 311453777
|
|
Closes #1197
PiperOrigin-RevId: 311438223
|
|
Linux 4.18 and later make reads and writes coherent between pre-copy-up and
post-copy-up FDs representing the same file on an overlay filesystem. However,
memory mappings remain incoherent:
- Documentation/filesystems/overlayfs.rst, "Non-standard behavior": "If a file
residing on a lower layer is opened for read-only and then memory mapped with
MAP_SHARED, then subsequent changes to the file are not reflected in the
memory mapping."
- fs/overlay/file.c:ovl_mmap() passes through to the underlying FD without any
management of coherence in the overlay.
- Experimentally on Linux 5.2:
```
$ cat mmap_cat_page.c
#include <err.h>
#include <fcntl.h>
#include <stdio.h>
#include <string.h>
#include <sys/mman.h>
#include <unistd.h>
int main(int argc, char **argv) {
if (argc < 2) {
errx(1, "syntax: %s [FILE]", argv[0]);
}
const int fd = open(argv[1], O_RDONLY);
if (fd < 0) {
err(1, "open(%s)", argv[1]);
}
const size_t page_size = sysconf(_SC_PAGE_SIZE);
void* page = mmap(NULL, page_size, PROT_READ, MAP_SHARED, fd, 0);
if (page == MAP_FAILED) {
err(1, "mmap");
}
for (;;) {
write(1, page, strnlen(page, page_size));
if (getc(stdin) == EOF) {
break;
}
}
return 0;
}
$ gcc -O2 -o mmap_cat_page mmap_cat_page.c
$ mkdir lowerdir upperdir workdir overlaydir
$ echo old > lowerdir/file
$ sudo mount -t overlay -o "lowerdir=lowerdir,upperdir=upperdir,workdir=workdir" none overlaydir
$ ./mmap_cat_page overlaydir/file
old
^Z
[1]+ Stopped ./mmap_cat_page overlaydir/file
$ echo new > overlaydir/file
$ cat overlaydir/file
new
$ fg
./mmap_cat_page overlaydir/file
old
```
Therefore, while the VFS1 gofer client's behavior of reopening read FDs is only
necessary pre-4.18, replacing existing memory mappings (in both sentry and
application address spaces) with mappings of the new FD is required regardless
of kernel version, and this latter behavior is common to both VFS1 and VFS2.
Re-document accordingly, and change the runsc flag to enabled by default.
New test:
- Before this CL: https://source.cloud.google.com/results/invocations/5b222d2c-e918-4bae-afc4-407f5bac509b
- After this CL: https://source.cloud.google.com/results/invocations/f28c747e-d89c-4d8c-a461-602b33e71aab
PiperOrigin-RevId: 311361267
|
|
It's a workaround to treat PROT_NONE as RDONLY temporarily.
TODO(gvisor.dev/issue/2686): PROT_NONE should be specially treated.
Signed-off-by: Bin Lu <bin.lu@arm.com>
|
