For backward compatibility, node uses X509_NAME_oneline to format
X509_NAME entries in PrintGeneralName. However, the format produced by
this function is non-standard and its use is discouraged. It also does
not handle Unicode names correctly.
This change switches to X509_NAME_print_ex with flags that produce an
RFC2253-compatible format. Non-ASCII strings are converted to UTF-8 and
preserved in the output. Control characters are not escaped by OpenSSL
when producing the RFC2253 format because they will be escaped by node
in a JSON-compatible manner afterwards.
PR-URL: https://github.com/nodejs/node/pull/42002
Refs: https://github.com/nodejs/node/pull/42001
Reviewed-By: Rich Trott <rtrott@gmail.com>
Reviewed-By: Matteo Collina <matteo.collina@gmail.com>
Rename X509_NAME_FLAGS to kX509NameFlagsMultiline to better align with
the naming conventions we use for constexpr values, to distinguish it
from OpenSSL's built-in X509 constants, and to clarify what specific
X509 name flags the constant represents.
PR-URL: https://github.com/nodejs/node/pull/42001
Reviewed-By: Anna Henningsen <anna@addaleax.net>
Reviewed-By: Antoine du Hamel <duhamelantoine1995@gmail.com>
This addresses a `TODO` comment, and makes use of the opportunity
to also clean up our `MaybeLocal` handling in this area and
start accepting `std::string_view` where we accept `std::string`.
PR-URL: https://github.com/nodejs/node/pull/41757
Reviewed-By: Darshan Sen <raisinten@gmail.com>
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: Tobias Nießen <tniessen@tnie.de>
The usage of `CHECK_LE` to verify that the index is within bounds
of a vector's size allows for reading one item past the vector's end,
which is in invalid memory read. This commit fixes the off-by-one error
by changing the bounds check to use `CHECK_LT`.
PR-URL: https://github.com/nodejs/node/pull/41452
Reviewed-By: Anna Henningsen <anna@addaleax.net>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: Tobias Nießen <tniessen@tnie.de>
Reviewed-By: Rich Trott <rtrott@gmail.com>
Reviewed-By: Luigi Pinca <luigipinca@gmail.com>
Reviewed-By: Richard Lau <rlau@redhat.com>
When subject and verifier are represented as strings, escape special
characters (such as '+') to guarantee unambiguity. Previously, different
distinguished names could result in the same string when encoded. In
particular, inserting a '+' in a single-value Relative Distinguished
Name (e.g., L or OU) would produce a string that is indistinguishable
from a multi-value Relative Distinguished Name. Third-party code that
correctly interprets the generated string representation as a
multi-value Relative Distinguished Name could then be vulnerable to an
injection attack, e.g., when an attacker includes a single-value RDN
with type OU and value 'HR + CN=example.com', the string representation
produced by unpatched versions of Node.js would be
'OU=HR + CN=example.com', which represents a multi-value RDN.
Node.js itself is not vulnerable to this attack because the current
implementation that parses such strings into objects does not handle '+'
at all. This oversight leads to incorrect results, but at the same time
appears to prevent injection attacks (as described above).
With this change, the JavaScript objects representing the subject and
issuer Relative Distinguished Names are constructed in C++ directly,
instead of (incorrectly) encoding them as strings and then (incorrectly)
decoding the strings in JavaScript.
This addresses CVE-2021-44533.
CVE-ID: CVE-2021-44533
PR-URL: https://github.com/nodejs-private/node-private/pull/300
Reviewed-By: Michael Dawson <midawson@redhat.com>
Reviewed-By: Rich Trott <rtrott@gmail.com>
This change introduces JSON-compatible escaping rules for strings that
include X.509 GeneralName components (see RFC 5280). This non-standard
format avoids ambiguities and prevents injection attacks that could
previously lead to X.509 certificates being accepted even though they
were not valid for the target hostname.
These changes affect the format of subject alternative names and the
format of authority information access. The checkServerIdentity function
has been modified to safely handle the new format, eliminating the
possibility of injecting subject alternative names into the verification
logic.
Because each subject alternative name is only encoded as a JSON string
literal if necessary for security purposes, this change will only be
visible in rare cases.
This addresses CVE-2021-44532.
CVE-ID: CVE-2021-44532
PR-URL: https://github.com/nodejs-private/node-private/pull/300
Reviewed-By: Michael Dawson <midawson@redhat.com>
Reviewed-By: Rich Trott <rtrott@gmail.com>
* The createHistogram(options) options weren't actually implemented
* Add a new count property that tracks the number of samples
* Adds BigInt options for relevant properties
* Adds add(other) method for RecordableHistogram
* Cleans up and expands tests
* Eliminates unnecessary ELDHistogram native class
* Improve/Simplify histogram transfer impl
Signed-off-by: James M Snell <jasnell@gmail.com>
perf_hooks: simplify Histogram constructor options
Signed-off-by: James M Snell <jasnell@gmail.com>
PR-URL: https://github.com/nodejs/node/pull/41153
Reviewed-By: Matteo Collina <matteo.collina@gmail.com>
This is supposed to be a public alternative of the private APIs,
`process._getActiveResources()` and `process._getActiveHandles()`. When
called, it returns an array of strings containing the types of the
active resources that are currently keeping the event loop alive.
Signed-off-by: Darshan Sen <darshan.sen@postman.com>
PR-URL: https://github.com/nodejs/node/pull/40813
Reviewed-By: Stephen Belanger <admin@stephenbelanger.com>
Reviewed-By: Vladimir de Turckheim <vlad2t@hotmail.com>
Reviewed-By: Matteo Collina <matteo.collina@gmail.com>
Coverity reported that the use of sizeof along with pointer
arithmetic was likely an error as the pointer arithmetic
would already be accounting for the size of what the
pointer points to.
Looking at the code that looked right but removing the
extra sizeOf caused tests to fail.
Looking more closely it seems like we were not allocating
a big enough buffer but the extra sizeof was allowing
us to convert even though it might have been corrupting
memory.
Signed-off-by: Michael Dawson <mdawson@devrus.com>
PR-URL: https://github.com/nodejs/node/pull/41026
Reviewed-By: Antoine du Hamel <duhamelantoine1995@gmail.com>
RFC 5246 section-7.2.1 requires that the implementation must immediately
stop reading from the stream, as it is no longer TLS-encrypted. The
underlying stream is permitted to still pump events (and errors) to
other users, but those are now unencrypted, so we should not process
them here. But therefore, we do not want to stop the underlying stream,
as there could be another user of it, but we do need to remove ourselves
as a listener.
Per TLS v1.2, we should have also destroy the TLS state entirely here
(including the writing side), but this was revised in TLS v1.3 to permit
the stream to continue to flush output.
There appears to be some inconsistencies in the way nodejs handles
ownership of the underlying stream, with `TLS.close()` on the write side
also calling shutdown on the underlying stream (thus assuming other
users of the underlying stream are not permitted), while receiving EOF
on the read side leaves the underlying channel open. These
inconsistencies are left for a later person to resolve, if the extra
functionality is needed (as described in #35904). The current goal here
is to the fix the occasional CI exceptions depending on the timing of
these kernel messages through the TCP stack.
PR-URL: https://github.com/nodejs/node/pull/36111
Fixes: https://github.com/nodejs/node/pull/35946
Refs: https://github.com/libuv/libuv/pull/3036
Refs: https://github.com/nodejs/node/issues/35904
Co-authored-by: Momtchil Momtchev <momtchil@momtchev.com>
Reviewed-By: Robert Nagy <ronagy@icloud.com>
Reviewed-By: Matteo Collina <matteo.collina@gmail.com>
Reviewed-By: Luigi Pinca <luigipinca@gmail.com>