Throw, don't abort. `new Buffer(0x3fffffff + 1)` used to bring down the process
with the following error message:
FATAL ERROR: v8::Object::SetIndexedPropertiesToExternalArrayData() length
exceeds max acceptable value
Fixes#2280.
* Upgrade V8 to 3.6.6.24
* dtrace ustack helper improvements (Dave Pacheco)
* API Documentation refactor (isaacs)
* #2827 net: fix race write() before and after connect() (koichik)
* #2554#2567 throw if fs args for 'start' or 'end' are strings (AJ ONeal)
* punycode: Update to v1.0.0 (Mathias Bynens)
* Make a fat binary for the OS X pkg (isaacs)
* Fix hang on accessing process.stdin (isaacs)
* repl: make tab completion work on non-objects (Nathan Rajlich)
* Fix fs.watch on OS X (Ben Noordhuis)
* Fix#2515 nested setTimeouts cause premature process exit (Ben Noordhuis)
* windows: fix time conversion in stat (Igor Zinkovsky)
* windows: fs: handle EOF in read (Brandon Philips)
* windows: avoid IOCP short-circuit on non-ifs lsps (Igor Zinkovsky)
* Upgrade npm to 1.1.4 (isaacs)
- windows fixes
- Bundle nested bundleDependencies properly
- install: support --save with url install targets
- shrinkwrap: behave properly with url-installed modules
- support installing uncompressed tars or single file modules from urls etc.
- don't run make clean on rebuild
- support HTTPS-over-HTTP proxy tunneling
This reverts commit b0c1541227.
Reverted commit introduced a regression causing `process.argv[0]` to be
invalid in node processes spawned from `PATH` (without explicit path to
executable file - for example when using global node installation).
Instead of finding a correct path to the executable, `process.cwd()`
would be prepended to `process.argv[0]`.
* http: allow multiple WebSocket RFC6455 headers (Einar Otto Stangvik)
* http: allow multiple WWW-Authenticate headers (Ben Noordhuis)
* windows: support unicode argv and environment variables (Bert Belder)
* tls: mitigate session renegotiation attacks (Ben Noordhuis)
* tcp, pipe: don't assert on uv_accept() errors (Ben Noordhuis)
* tls: Allow establishing secure connection on the existing socket (koichik)
* dgram: handle close of dgram socket before DNS lookup completes (Seth Fitzsimmons)
* windows: Support half-duplex pipes (Igor Zinkovsky)
* build: disable omit-frame-pointer on solaris systems (Dave Pacheco)
* debugger: fix --debug-brk (Ben Noordhuis)
* net: fix large file downloads failing (koichik)
* fs: fix ReadStream failure to read from existing fd (Christopher Jeffrey)
* net: destroy socket on DNS error (Stefan Rusu)
* dtrace: add missing translator (Dave Pacheco)
* unix: don't flush tty on switch to raw mode (Ben Noordhuis)
* windows: reset brightness when reverting to default text color (Bert Belder)
* npm: update to 1.1.1
- Update which, fstream, mkdirp, request, and rimraf
- Fix#2123 Set path properly for lifecycle scripts on windows
- Mark the root as seen, so we don't recurse into it. Fixes#1838. (Martin Cooper)
The TLS protocol allows (and sometimes requires) clients to renegotiate the
session. However, renegotiation requires a disproportional amount of server-side
resources, particularly CPU time, which makes it a potential vector for
denial-of-service attacks.
To mitigate this issue, we keep track of and limit the number of renegotiation
requests over time, emitting an error if the threshold is exceeded.
These methods were removed a long time ago. Keeping these notices here
makes no sense anymore.
Also, removing this part of code slightly speeds up the startup.
It's possible for a new connection to be closed in the window between the
accept() syscall and the call to uv_accept(). Deal with it and move on, don't
assert.
`process.debug_port` is useful for changing debugger port in runtime,
before starting it (via SIGUSR1).
Using `--port=` argument for debugger repl, tests will run debugger
server on a `common.PORT` (as it usually does for any other servers).
`process._debugEnd()` stops debugger and its server.
* debugger: implemented process._debugEnd(), `node debug --port=5858 app.js`
* test: start debugger repl on common.PORT
* fixes#2613
* fixes#2614
11d1eca9 added a v8 locker to ease development of 3rd party threading
extensions but it created a condition which would cause node to exit
uncleanly while in debug mode; it was reverted in 7543c38d.
The problem here is that the Locker was being disposed after V8 was torn
down. Adding some scoping fixes that.
* Upgrade V8 to 3.9.2
* Revert support for isolates. (Ben Noordhuis)
* cluster: Cleanup docs, event handling, and process.disconnect (Andreas Madsen)
* gyp_addon: link with node.lib on Windows (Nathan Rajlich)
* http: fix case where http-parser is freed twice (koichik)
* Windows: disable RTTI and exceptions (Bert Belder)
It was decided that the performance benefits that isolates offer (faster spin-up
times for worker processes, faster inter-worker communication, possibly a lower
memory footprint) are not actual bottlenecks for most people and do not outweigh
the potential stability issues and intrusive changes to the code base that
first-class support for isolates requires.
Hence, this commit backs out all isolates-related changes.
Good bye, isolates. We hardly knew ye.
This reverts commit 11d1eca9f3.
It sporadically (but reproducibly) triggers an assert inside V8:
Fatal error in /path/to/node/deps/v8/src/isolate.cc, line 1857
CHECK(CurrentPerIsolateThreadData()->isolate_ == this) failed
Needs further investigation.
v8 requires a lock of each thread using the vm, but if none is
explicitly is created it will implicitly create one for you. This
creates issues when trying to build modules which use v8's
multi-threading features because there's no lock to unlock.
Ben Noordhuis
Bert Belder
Nathan Rajlich
Shigeki Ohtsu
isaacs
Dave Pacheco
Maciej Małecki
Tomasz Buchert
Andreas Madsen
Fedor Indutny
Igor Zinkovsky
Marcel Laverdet