Exclude the core/runtime/v2/README.md symlink in the links workflow so
lychee does not fail resolving relative links from the symlink path.
docs/runtime-v2.md is already validated directly.
Update broken, moved, and rate-limited links across markdown documents:
- Replace moved Kubernetes cgroup driver and GCP registry docs.
- Replace retired RHEL 6 devmapper docs with kernel.org reference.
- Point NSSM, CRIU, and Ansible references to GitHub repositories.
- Fix relative paths in remote-snapshotter.md and root-relative paths.
Assisted-by: Antigravity
Signed-off-by: Samuel Karp <samuelkarp@google.com>
Restore support for -1 as an unknown-size sentinel for backward
compatibility. Normalize it to 0, the documented value for unknown
sizes in the Writer interface.
Other negative sizes remain invalid, and known-size copies retain
their existing bounds checks.
Updates 91f62bdd77
Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
The size comes from the local content store, but validate it as a
defense-in-depth before passing it to content.Copy.
Updates 91f62bdd77
Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
Push depended on content.Copy to invalidate negative sizes. Check for invalid
descriptor sizes in push itself, without relying on the validation in Copy.
Updates 91f62bdd77
Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
WriteBlob expects the size of the content to be known in advance, but
currently relies on Copy to validate the descriptor size.
Check for invalid descriptor sizes in WriteBlob itself, without relying
on the validation in Copy.
Updates 91f62bdd77
Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
Pin workflow runners to ubuntu-24.04 to avoid deprecation warnings and
breaking changes when ubuntu-latest migrates to Ubuntu 26.
Assisted-by: Antigravity
Signed-off-by: Samuel Karp <samuelkarp@google.com>
limactl start creates the default instance directory before downloading
the guest image. Delete the stopped instance before retrying so the
second limactl start invocation does not fail on an existing instance.
Assisted-by: Antigravity
Signed-off-by: Samuel Karp <samuelkarp@google.com>
Commit records a snapshot in the metadata store, which syncs its own
file, while layer.erofs and the snapshot directory may still be in the
page cache. A power loss shortly after a pull can then leave a committed
snapshot whose blob is corrupt or whose directory is missing. A later
pull does nothing because the chain ID exists, and running the image
fails to mount the layer.
With fsync_on_commit = true, Commit fsyncs everything in the snapshot
directory, the directory itself and snapshots/ before the snapshot is
recorded. Blobs served from a layer content cache are not synced. The
option is off by default and lives in the snapshotter because the
differ's SyncFs does not reach Commit, where the walking differ's blob
is created.
Assisted-by: Claude:claude-opus-5-5
Signed-off-by: Mykhaylo Sul <mykhaylo.sul@oss.qualcomm.com>
The pusher creates the upload request body as an io.Pipe. request.do
attaches the pipe reader to the request before calling Authorize. If
Authorize fails, do returns before client.Do, so the reader is never
handed to the HTTP client, which would have closed it, and nothing else
closes it.
A concurrent pushWriter.Write then blocks forever in pipe.Write. It only
reads the error reported by the request goroutine after the pipe is
closed, and it takes no context, so cancelling the caller does not help.
Callers such as content.Copy never return, and anything they hold, for
example a concurrency slot, is never released.
Close the body when authorization fails so the blocked Write returns the
authorization error. Transient errors keep going through the existing
retry and content.ErrReset handling.
Signed-off-by: Tonis Tiigi <tonistiigi@gmail.com>
Use the resolved platform for pulling and image-store updates. Fall back to
local pull when transfer cannot unpack the requested platform and snapshotter.
Signed-off-by: Chengyu Zhu <hudson@cyzhu.com>
Windows could only listen on and dial named pipes. Also accept AF_UNIX
sockets (Windows 10 1803+, Server 2019+) as unix://<path>, unix:<path>
or a bare path, next to \\.\pipe\<name>, //./pipe/<name> and npipe://.
A scheme is authoritative and named pipes stay the default.
The address is parsed in internal/localaddr, shared by the listener,
the dialer and the debug server's local/TCP check. os.Chmod and
os.Chown do nothing on Windows, so the parent directory is created with
MkdirAllWithACL and the socket gets a protected DACL for Builtin
Administrators and Local System.
Windows has the 108 byte sun_path limit of Linux, but there is hope,
see https://lwn.net/Articles/987098/
Continues https://github.com/containerd/containerd/pull/12921
Co-authored-by: David Scott <dave@recoil.org>
Co-authored-by: Derek McGowan <derek@mcg.dev>
Assisted-by: OpenCode (claude-sonnet-5-5)
Signed-off-by: Eric Curtin <eric.curtin@docker.com>
These releases include 15 security fixes following the security policy:
- net/http: HTTP/2 server crash due to HPACK encoder race
HTTP/2 servers could end up crashing due to inadvertently
modifying its HPACK encoder concurrently. This happens because the
server modifies the HPACK encoder from two goroutines without
synchronization: one uses the encoder to encode a HEADERS frame as part
of a response sent to a client and the other modifies the encoder's
table size when handling a SETTINGS frame containing
SETTINGS_HEADER_TABLE_SIZE that a client sends. A malicious client can
repeatedly send a request while changing the header table size to crash
the server.
Fix this issue by not applying SETTINGS_HEADER_TABLE_SIZE immediately.
Instead, buffer any SETTINGS_HEADER_TABLE_SIZE received, and only apply
the new value prior to the next time the server writes a frame.
Thanks to RyotaK (https://ryotak.net) of GMO Flatt Security Inc. for reporting this issue.
This is CVE-2026-97032 and Go issue https://go.dev/issue/81867.
- net/http: HTTP/2 server memory exhaustion due to Trailer headers
When "Trailer" headers are sent by a client, the HTTP server internally
uses the header values to populate the Request.Trailer map passed to the
server handler. Because Request.Trailer is a map, each entry incurs
memory overhead. For HTTP/2 servers, a malicious client can exploit this
by sending a "Trailer" header that declares a large number of fields,
causing the server to allocate a disproportionate amount of memory while
bypassing Server.MaxHeaderValueCount and Server.MaxHeaderBytes limits.
This exploit is not applicable for HTTP/1 servers, which do not support
multiplexing a large number of requests over one TCP connection, and
whose Server.MaxHeaderBytes are calculated differently.
Server.MaxHeaderValueCount and Server.MaxHeaderBytes limits are now
applied towards the trailer fields declared in "Trailer" headers.
Thanks to RyotaK (https://ryotak.net) of GMO Flatt Security Inc. for reporting this issue.
This is CVE-2026-78659 and Go issue https://go.dev/issue/81857.
- crypto/tls: reject malformed ECH outer extension references
Multiple ECH outer extension references are not
permitted under RFC 9849; previously, a client
could send a well-crafted packet that could
trigger memory exhaustion in the server process
by specifying multiple references.
We now reject these as malformed and curb the
memory amplification vector as a result.
This is CVE-2026-97031 and Go issue https://go.dev/issue/81855.
- cmd/go: checksum bypass for golang.org/fips140
Previously, a user operating inside of a malicious
Go project that defines a bogus golang.org/fips140
and operates a malicious GOMODPROXY the user chooses
to connect to can serve an arbitrary module in its
place.
We now unpack the trusted ziphash for the bundled
golang.org/fips140 module and construct its entry
in the GOMODCACHE such that it can be verified by
the toolchain.
This is CVE-2026-94444 and Go issue https://go.dev/issue/81833.
- cmd/go: checksum database bypass for golang.org/toolchain
Previously, a user operating inside of a malicious
Go project that defines a bogus golang.org/toolchain
go.sum entry and operates a malicious GOMODPROXY the
user chooses to use can bypass the intended checksum.
We now ensure that golang.org/toolchain always goes
to the network for the canonical checksum.
This is CVE-2026-94447 and Go issue https://go.dev/issue/81834.
- html/template: reset context tracking on consecutive template expressions
When a JavaScript template literal contains
consecutive expressions, the context tracking
state was not properly reset upon entering a
new expression.
We now ensure that template-literal expression
entries correctly reset context variables so all
subsequent regular expression literals are
accurately recognized and escaped.
This is CVE-2026-94448 and Go issue https://go.dev/issue/81821.
- html/template: recognize yield as regexp preceder keyword
A trusted template author may have previously
written a valid template wherein the use of
the yield keyword would not be correctly
escaped.
We now ensure that valid keyword uses are
escaped and non-keyword uses are not escaped.
This is CVE-2026-97030 and Go issue https://go.dev/issue/81823.
- net/textproto, mime/multipart: memory limit bypass when parsing MIME headers
Parsing a multipart form could bypass memory limits and read an
arbitrarily long line into memory when the remaining limit at the
start of a part was less than 400 bytes.
Multipart form memory limits are now properly enforced in this situation.
Thanks to Jakub Ciolek (https://ciolek.dev) for reporting this issue.
This is CVE-2026-94440 and Go issue https://go.dev/issue/81741.
- net/http: HTTP/1 client connection desynchronization after CONNECT rejection
When http.Transport sends an HTTP/1 CONNECT request with a non-empty
Request.Body, it writes the body directly to the connection without
framing after the request headers. If the server rejects the CONNECT
request with a non-2xx keep-alive response, Transport returns the
connection to the idle pool. Because CONNECT requests do not have a
request body, the server may interpret the trailing body bytes as a
subsequent pipelined HTTP/1.1 request on the connection, leaving the
pooled connection desynchronized and causing the next caller that reuses
it to read the response to the injected request. In reverse proxies
(including httputil.ReverseProxy) that forward CONNECT requests through
a shared Transport, this can lead to cross-user response poisoning.
The HTTP/1 transport now closes a connection after sending a CONNECT
request, regardless of the response status.
In addition, ReverseProxy now rejects incoming CONNECT requests
with a 405 Method Not Allowed response. ReverseProxy has never handled
CONNECT requests in a useful fashion (it does not convert the
connection into a bidirectional tunnel), so we do not expect this
change to negatively affect any current users.
Thanks to Xclow3n (Rajat Raghav) for reporting this issue.
This is CVE-2026-56866 and Go issue https://go.dev/issue/81740.
- net/http: HTTP/1 server connection desynchronization after 2xx CONNECT response
When an HTTP server handler sent a 2xx response to an HTTP/1 CONNECT request
and returned without hijacking the connection, the server improperly continued
to read and serve requests from the connection. Since a 2xx response to an
HTTP/1 CONNECT converts the connection into a tunnel, the server should not
treat the connection as continuing to contain HTTP.
The impact of this misbehavior is mostly limited to potential request smuggling,
where an intermediate proxy considers the data on the connection to be tunneled
and the server considers it to be HTTP.
The HTTP/1 server now always closes a connection after responding to a CONNECT
request, regardless of the response status.
Thanks to Jakub Ciolek (https://ciolek.dev) for reporting this issue.
This is CVE-2026-94439 and Go issue https://go.dev/issue/81744.
- net/http: excessive CPU consumption from repeated initial window changes
A malicious HTTP/2 peer could cause excessive CPU consumption in the
client or server by opening a large number of streams and then sending
many small SETTINGS frames containing SETTINGS_INITIAL_WINDOW_SIZE values.
The HTTP/2 client and server now efficiently handle changes to the
initial window size (O(1) rather than O(number of streams)).
Thanks to Jakub Ciolek (https://ciolek.dev) for reporting this issue.
This is CVE-2026-78669 and Go issue https://go.dev/issue/81742.
- net/http: HTTP/2 transport accepts malformed framing-related headers
Historically, we have been rather lax about malformed framing-related
headers in our HTTP/2 implementation, as they cannot interfere with
HTTP/2 framing. However, this makes it possible for our HTTP/2
implementation to forward responses containing such headers to an HTTP/1
client when acting as a reverse proxy. If the HTTP/1 client also does
not behave strictly enough, this can result in response smuggling.
We now delete malformed framing-related headers when received by our
HTTP/2 transport, so they will not be forwarded to a potentially
vulnerable HTTP/1 client.
Thanks to TJ Barton for reporting this issue.
This is CVE-2026-78660 and Go issue https://go.dev/issue/81115.
- os: Root.Mkdir(All) can follow junctions out of the root on Windows
On Windows, when the target of Root.Mkdir or Root.MkdirAll was a junction
pointing to an empty location, the operation would create a directory at
the junction target even when that target was located outside the root.
This only applies to operations where the last path component is a
junction (path/to/junction, but not path/junction/target).
Root.Mkdir and Root.MkdirAll now correctly avoid resolving junctions.
Thanks to Daniele Ballarini for reporting this issue.
This is CVE-2026-56857 and Go issue https://go.dev/issue/81739.
- net/http: lack of limit on size of parsed Range headers
When parsing a Range header containing a large number of small ranges,
FileServer(FS), ServeContent, and ServeFile(FS) could consume an excessive
amount of CPU.
These functions now ignore Range headers containing more than 200 ranges.
The limit is controlled by the new httpservecontentmaxranges=
GODEBUG setting. Setting GODEBUG=httpservecontentmaxranges=0
disables the limit.
Thanks to Jakub Ciolek (https://ciolek.dev) for reporting this issue.
This is CVE-2026-78667 and Go issue https://go.dev/issue/81858.
- net/http: double flow control refund on HTTP/2 server streams
The HTTP/2 server could refund connection-level flow control twice for the same data:
Once when a client resets a stream (refunding data for any sent-but-unread portion
of the stream), and again when a request handler reads the buffered data.
A malicious client could exploit this to bypass the configured connection-level
flow control limit (MaxReceiveBufferPerConnection). Total buffered data is still
limited by the concurrent stream limit and stream-level flow control.
The HTTP/2 server now waits to refund connection-level flow control for reset
streams until after the request handler is complete.
Thanks to Ali Sherif (https://www.linkedin.com/in/ali-sherif-13812b276/) for reporting this issue.
This is CVE-2026-78663 and Go issue https://go.dev/issue/81743.
release notes: https://go.dev/doc/devel/release#go1.27.2
Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
Resolve explicit, default, and annotation-based handlers to their configured
platform and snapshotter. Reject unknown handlers and preserve backend platform
defaults when no platform is configured.
Signed-off-by: Chengyu Zhu <hudson@cyzhu.com>
Stream chunks through a chunkReader backed by a channel of capacity
parallelism so at most parallelism pipes are allocated at any time.
Assisted-by: Antigravity
Signed-off-by: Samuel Karp <samuelkarp@google.com>
systemd-tmpfiles-clean.timer runs 15 minutes after boot and removes
empty directories under /tmp with mtimes older than 10 days. Some
integration tests place containerd's root in t.TempDir() and unpack
image layers with older directory timestamps under /tmp. If the timer
fires during extraction, systemd-tmpfiles removes newly created empty
directories before the final directory timestamp pass.
Assisted-by: Antigravity
Signed-off-by: Samuel Karp <samuelkarp@google.com>
Reject negative sizes in content.Copy, remotes.Fetch,
dockerFetcher.Fetch, and newHTTPReadSeeker.
Bound reads in content.Copy and httpReadSeeker.Read to the remaining
expected size from the current offset, returning
errdefs.ErrFailedPrecondition if the source yields extra bytes.
Assisted-by: Antigravity
Signed-off-by: Samuel Karp <samuelkarp@google.com>
A terminal exec or init process has no processIO. The shim keeps a writer
on the stdin FIFO (openStdin) so the console's stdin copier never sees a
premature EOF, and keeps the console master in `console`. delete closed
the closers only under `if io != nil`, which is never true on the
terminal path, and never closed the console. When a client exits without
calling CloseIO, the stdin copier therefore stays blocked on the FIFO and
the FIFO handles, the console master and the goroutine stay alive until a
later garbage collection finalizes the dropped process object, and for as
long as the client holds its own writer on the FIFO, past that too.
Release the I/O on every path and through one owner. delete shuts the
console down, drains, then closes closers, console and io regardless of
io; a process deleted from the created state never ran setExited, so this
is its only shutdown. CopyConsole returns a console that owns its
teardown: shutdown and deregistration from the epoller happen once, while
the fd is still open, and Close then closes the master, wakes a writer
parked on it, and closes the stdin FIFO reader, which ends a copier
blocked reading stdin. The copier, the setup-error path, ShutdownConsole
and both deletes go through it, so a bare Close can no longer leave a
copier asleep, and a late ShutdownConsole is a no-op instead of a
deregistration by a fd number that a newer console may own. CopyConsole
also closes the received master when registering it fails and the stdout
FIFO writer when opening its reader fails.
Six regression tests cover the delete ordering, the setup-error paths,
the copier wake-up and the fd reuse; each fails before its part of this
change.
Assisted-by: Claude Code
Signed-off-by: Jacob Gabrielson <jacob.gabrielson@docker.com>
With max_concurrent_downloads and concurrent_layer_fetch_buffer set,
containerd can use far more memory than intended. The download limiter
bounds how many requests run at once. A downloaded chunk stays in
memory until the reader consumes it. When the registry is faster than
the content store write, chunks accumulate up to the size of the layer.
Add a per-layer sliding window of max_workers chunks. Download a new
chunk only after the reader has consumed the oldest chunk in the
window. A layer then holds at most chunk_size * max_workers of chunk
buffers. When the fetch is cancelled, close the writers of chunks that
never reached a worker, as a reader would otherwise block on them
forever. Pooled buffers and other overhead add to total memory use.
Signed-off-by: Maksym Pavlenko <pavlenko.maksym@gmail.com>
CRI requires PullImage to reject an unknown runtime handler. Validate it
against the handlers configured in the CRI runtime service; an empty handler
is unchanged.
Part of #14252
Signed-off-by: Chengyu Zhu <hudson@cyzhu.com>
Maksym Pavlenko