-
Notifications
You must be signed in to change notification settings - Fork 29.8k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
worker: fix nullptr deref after MessagePort deser failure #25076
Conversation
This would previously always have crashed when deserializing a `MessagePort` fails, because there was always at least one `nullptr` entry in the vector.
@@ -90,7 +90,8 @@ MaybeLocal<Value> Message::Deserialize(Environment* env, | |||
if (ports[i] == nullptr) { | |||
for (MessagePort* port : ports) { | |||
// This will eventually release the MessagePort object itself. | |||
port->Close(); | |||
if (port != nullptr) |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Does the enclosing for
loop need to loop through all ports
or just up to (but not including since ports[i] == nullptr
) i
?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
@richardlau Yes, I think that should work too … should we optimize here? This effectively only occurs during .terminate()
, and I don’t think it’s a typical case to pass more than one MessagePort
per Message anyway?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
¯\_(ツ)_/¯ I'll leave it to your judgement.
Landed in e1ab457. |
This would previously always have crashed when deserializing a `MessagePort` fails, because there was always at least one `nullptr` entry in the vector. PR-URL: #25076 Reviewed-By: Richard Lau <riclau@uk.ibm.com> Reviewed-By: James M Snell <jasnell@gmail.com> Reviewed-By: Gireesh Punathil <gpunathi@in.ibm.com>
This would previously always have crashed when deserializing a `MessagePort` fails, because there was always at least one `nullptr` entry in the vector. PR-URL: #25076 Reviewed-By: Richard Lau <riclau@uk.ibm.com> Reviewed-By: James M Snell <jasnell@gmail.com> Reviewed-By: Gireesh Punathil <gpunathi@in.ibm.com>
This would previously always have crashed when deserializing a `MessagePort` fails, because there was always at least one `nullptr` entry in the vector. PR-URL: nodejs#25076 Reviewed-By: Richard Lau <riclau@uk.ibm.com> Reviewed-By: James M Snell <jasnell@gmail.com> Reviewed-By: Gireesh Punathil <gpunathi@in.ibm.com>
This would previously always have crashed when deserializing a `MessagePort` fails, because there was always at least one `nullptr` entry in the vector. PR-URL: #25076 Reviewed-By: Richard Lau <riclau@uk.ibm.com> Reviewed-By: James M Snell <jasnell@gmail.com> Reviewed-By: Gireesh Punathil <gpunathi@in.ibm.com>
This would previously always have crashed when deserializing a `MessagePort` fails, because there was always at least one `nullptr` entry in the vector. PR-URL: #25076 Reviewed-By: Richard Lau <riclau@uk.ibm.com> Reviewed-By: James M Snell <jasnell@gmail.com> Reviewed-By: Gireesh Punathil <gpunathi@in.ibm.com>
This would previously always have crashed when deserializing a `MessagePort` fails, because there was always at least one `nullptr` entry in the vector. PR-URL: #25076 Reviewed-By: Richard Lau <riclau@uk.ibm.com> Reviewed-By: James M Snell <jasnell@gmail.com> Reviewed-By: Gireesh Punathil <gpunathi@in.ibm.com>
This would previously always have crashed when deserializing a `MessagePort` fails, because there was always at least one `nullptr` entry in the vector. PR-URL: #25076 Reviewed-By: Richard Lau <riclau@uk.ibm.com> Reviewed-By: James M Snell <jasnell@gmail.com> Reviewed-By: Gireesh Punathil <gpunathi@in.ibm.com>
Notable changes: - **deps**: upgrade openssl sources to 1.1.1c (Sam Roberts) [#28212](#28212) - **stream**: do not unconditionally call `\_read()` on `resume()` (Anna Henningsen) [#26965](#26965) - **worker**: fix nullptr deref after MessagePort deser failure (Anna Henningsen) [#25076](#25076) PR-URL: #28731
Notable changes: - **deps**: upgrade openssl sources to 1.1.1c (Sam Roberts) [#28212](#28212) - **stream**: do not unconditionally call `\_read()` on `resume()` (Anna Henningsen) [#26965](#26965) - **worker**: fix nullptr deref after MessagePort deser failure (Anna Henningsen) [#25076](#25076) PR-URL: #28731
Notable changes: - **deps**: upgrade openssl sources to 1.1.1c (Sam Roberts) [#28212](#28212) - **stream**: do not unconditionally call `\_read()` on `resume()` (Anna Henningsen) [#26965](#26965) - **worker**: fix nullptr deref after MessagePort deser failure (Anna Henningsen) [#25076](#25076) PR-URL: #28731
Notable changes: - **deps**: upgrade openssl sources to 1.1.1c (Sam Roberts) [#28212](#28212) - **stream**: do not unconditionally call `\_read()` on `resume()` (Anna Henningsen) [#26965](#26965) - **worker**: fix nullptr deref after MessagePort deser failure (Anna Henningsen) [#25076](#25076) PR-URL: #28731
This would previously always have crashed when deserializing
a
MessagePort
fails, because there was always at least onenullptr
entry in the vector.(Caught by @gireeshpunathil in #25061)
Checklist
make -j4 test
(UNIX), orvcbuild test
(Windows) passes