summaryrefslogtreecommitdiffstats
path: root/Modules/socketmodule.c
Commit message (Collapse)AuthorAgeFilesLines
* Issue #24732, #23834: Fix sock_accept_impl() on WindowsVictor Stinner2015-07-271-0/+5
| | | | accept() returns INVALID_SOCKET on error, it's not necessary a negative number.
* Issue #22631: Added Linux-specific socket constant CAN_RAW_FD_FRAMES.Larry Hastings2015-04-131-0/+3
| | | | Patch courtesy of Joe Jevnik.
* Issue #23618: Fix internal_select() for negative timeout (blocking socket) whenVictor Stinner2015-04-091-4/+9
| | | | | | | poll() is not available. select() doesn't accept negative timeout, the timeout parameter must be NULL to block on select().
* Issue #23834: Fix the default socket timeoutVictor Stinner2015-04-091-1/+2
| | | | Use -1 second by default, not -1 nanosecond.
* Issue #22117: Fix sock_call_ex() for non-blocking socketVictor Stinner2015-04-061-1/+1
| | | | | Call internal_select() with a timeout of 0 second, not a timeout of -1 second (blocking)!
* Issue #23853: socket.socket.sendall() does no more reset the socket timeoutVictor Stinner2015-04-061-16/+41
| | | | | each time data is sent successfuly. The socket timeout is now the maximum total duration to send all data.
* Issue #23834: Fix initial value of the socket timeoutVictor Stinner2015-04-061-2/+2
| | | | | | Use _PyTime_FromSeconds() to initialize the default socket timeout to -1 second, instead of -1 nanosecond which causes rounding issues in internal_select().
* Issue #23834: Simplify timeout handlingVictor Stinner2015-04-031-27/+13
| | | | | | | | | | | * Use the new _PyTime_FromSeconds() function to set the timeout to -1 second for socket.settimeout(None). It avoids a special case in internal_select() because of a rounding issue: -1 nanosecond is rounded to 0 millisecond which means non-blocking, instead of blocking. * Check if the interval the negative in sock_call_ex() instead of doing the check in internal_select(). sock_call_ex() remembers if the socket has a timeout or not, which avoids a race condition if the timeout is modified in a different thread.
* Issue #23618: Ooops, remove abort() added for debug purposeVictor Stinner2015-04-021-1/+0
|
* Issue #23618: Fix sock_connect_impl(), set the socket error codeVictor Stinner2015-04-021-1/+7
| | | | | sock_call_ex() gets the socket error code when the socket function fails. sock_connect_impl() didn't set the error correctly.
* Issue #23618: socket.socket.connect() now waits until the connection completesVictor Stinner2015-04-021-124/+178
| | | | | | | | instead of raising InterruptedError if the connection is interrupted by signals, signal handlers don't raise an exception and the socket is blocking or has a timeout. socket.socket.connect() still raise InterruptedError for non-blocking sockets.
* Issue #23618: Don't declare recvmsg/sendmsg helper functions on WindowsVictor Stinner2015-04-021-7/+6
|
* Issue #23834: Fix sock_call(), set deadline_initialized to recompute the timeoutVictor Stinner2015-04-021-0/+1
|
* Issue #23834: Modify socket.sendall() to reuse sock_call() withVictor Stinner2015-04-011-29/+21
| | | | sock_send_impl()
* Issue #23834: Add sock_call() helper functionVictor Stinner2015-04-011-256/+323
| | | | | | | | | | The BEGIN_SELECT_LOOP and END_SELECT_LOOP macros of socketmodule.c don't handle EINTR. Functions using these macros use an inner loop to handle EINTR, but they don't recompute the timeout. This changes replaces the two macros with a new sock_call() function which takes a function as a parameter. sock_call() recomputes the timeout, handle false positive and handle EINTR.
* merge 3.4Benjamin Peterson2015-04-011-1/+2
|\
| * remove assignment in conditionalBenjamin Peterson2015-04-011-1/+2
| |
* | Issue #23618: Fix EINTR handling on WindowsVictor Stinner2015-04-011-2/+2
| | | | | | | | Windows uses WSAEINTR error code, not EINTR, for socket functions.
* | Issue #23618: Enhance EINTR handling in socket.connect()Victor Stinner2015-03-311-35/+37
| | | | | | | | | | | | | | Call PyErr_CheckSignals() immediatly if connect() or select() fails with EINTR in internal_connect(). Refactor also the code to limit indentaton and make it more readable.
* | Issue #23618: Fix EINTR handling in socket.connect()Victor Stinner2015-03-311-10/+8
| | | | | | | | | | Call PyErr_CheckSignals() if connect(), select() or getsockopt() failed with EINTR.
* | Issue #23618: Cleanup internal_connect() in socketmodule.cVictor Stinner2015-03-311-3/+1
| | | | | | | | | | On Windows, it looks like using the C type socklen_t for getsockopt() (instead of int) is fine, it was already used in socket.getsockopt().
* | Issue #23618: Refactor internal_connect()Victor Stinner2015-03-311-79/+30
| | | | | | | | | | On Windows, internal_connect() now reuses internal_connect_select() and always calls getsockopt().
* | Issue #23618: Refactor internal_connect()Victor Stinner2015-03-311-35/+57
| | | | | | | | | | | | | | The function now returns the error code instead of using the global errno (POSIX) or WSAGetLastError() (Windows). internal_connect() now returns errno if getsockopt() fails.
* | Issue #22117: Fix integer overflow check in socket_parse_timeout() on WindowsVictor Stinner2015-03-311-3/+6
| |
* | Issue #23618: Fix internal_connect_select()Victor Stinner2015-03-311-1/+1
| |
* | Issue #23618: internal_connect_select() now waits also for error eventsVictor Stinner2015-03-311-6/+15
| |
* | Issue #23618: Refactor internal_select() to prepare socket.connect() for EINTRVictor Stinner2015-03-311-13/+23
| |
* | Issue #23618: Refactor the _socket moduleVictor Stinner2015-03-311-17/+11
| | | | | | | | | | * Inline internal_select() function * Rename internal_select_ex() internal_select()
* | Issue #22117: Replace usage of _PyTime_ROUND_UP with _PyTime_ROUND_CEILINGVictor Stinner2015-03-301-6/+7
| | | | | | | | | | All these functions only accept positive timeouts, so this change has no effect in practice.
* | Issue #22117: Fix usage of _PyTime_AsTimeval()Victor Stinner2015-03-301-6/+2
| | | | | | | | | | Add _PyTime_AsTimeval_noraise() function. Call it when it's not possible (or not useful) to raise a Python exception on overflow.
* | Issue #22117: The socket module uses _PyTime_t timestamp for timeoutsVictor Stinner2015-03-281-90/+123
| |
* | Issue #23618, #22117: refactor socketmodule.cVictor Stinner2015-03-271-83/+113
| | | | | | | | | | Move Py_BEGIN_ALLOW_THREADS/Py_END_ALLOW_THREADS inside internal_select_ex() to prepare a switch to the _PyTime_t type and retry syscall on EINTR.
* | Fix compiler warnings: comparison between signed and unsigned numbersVictor Stinner2015-03-201-1/+1
| |
* | Issue #23001: Few functions in modules mmap, ossaudiodev, socket, ssl, andSerhiy Storchaka2015-03-201-47/+60
| | | | | | | | | | codecs, that accepted only read-only bytes-like object now accept writable bytes-like object too.
* | Issue #23646: Enhance precision of time.sleep() and socket timeout whenVictor Stinner2015-03-201-1/+1
| | | | | | | | | | | | | | | | interrupted by a signal Add a new _PyTime_AddDouble() function and remove _PyTime_ADD_SECONDS() macro. The _PyTime_ADD_SECONDS only supported an integer number of seconds, the _PyTime_AddDouble() has subsecond resolution.
* | Fixed GCC version testing.Serhiy Storchaka2015-03-051-2/+2
| |
* | Fix "GCC diagnostic" in socketmodule.cVictor Stinner2015-03-051-2/+2
| | | | | | | | | | Fix regression of changeset 7c6e3358221a on GCC < 4.4. The _socket module cannot be compiled on "x86 FreeBSD 7.2 3.x" buildbot anymore.
* | Silenced minor GCC warnings.Serhiy Storchaka2015-02-261-0/+7
| |
* | Issue #23446: Use PyMem_New instead of PyMem_Malloc to avoid possible integerSerhiy Storchaka2015-02-161-2/+4
|\ \ | |/ | | | | overflows. Added few missed PyErr_NoMemory().
| * Issue #23446: Use PyMem_New instead of PyMem_Malloc to avoid possible integerSerhiy Storchaka2015-02-161-2/+4
| | | | | | | | overflows. Added few missed PyErr_NoMemory().
* | Issue #23285: PEP 475 -- Retry system calls failing with EINTR.Charles-François Natali2015-02-071-112/+127
| |
* | Issue #22919: Windows build updated to support VC 14.0 (Visual Studio 2015), ↵Steve Dower2014-11-221-1/+10
| | | | | | | | which will be used for the official 3.5 release.
* | Issue #22581: Use more "bytes-like object" throughout the docs and comments.Serhiy Storchaka2014-12-051-3/+3
|\ \ | |/
| * Issue #22581: Use more "bytes-like object" throughout the docs and comments.Serhiy Storchaka2014-12-051-3/+3
| |
* | Issue #22378: socket module: add SO_MARK.Charles-François Natali2014-09-181-0/+3
| |
* | Issue #22215: Now ValueError is raised instead of TypeError when str or bytesSerhiy Storchaka2014-09-061-1/+1
| | | | | | | | argument contains not permitted null character or byte.
* | Issue #22043: time.monotonic() is now always availableVictor Stinner2014-09-021-2/+2
| | | | | | | | | | threading.Lock.acquire(), threading.RLock.acquire() and socket operations now use a monotonic clock, instead of the system clock, when a timeout is used.
* | Issue #22218: Fix "comparison between signed and unsigned integers" warnings inVictor Stinner2014-08-171-5/+6
| | | | | | | | socketmodule.c.
* | Issue #22127: fix typo.Martin v. Löwis2014-08-051-1/+1
| |
* | Issue #22127: Bypass IDNA for pure-ASCII host names (in particular for ↵Martin v. Löwis2014-08-051-10/+75
| | | | | | | | numeric IPs).
generated by cgit v0.12 at 2025-01-06 08:46:58 (GMT)