summaryrefslogtreecommitdiffstats
path: root/generic/tclIO.h
Commit message (Collapse)AuthorAgeFilesLines
* * generic/tclIO.h (CHANNEL_INCLOSE): New flag. Set inandreas_kupries2004-07-151-1/+10
| | | | | | | | | | | | | | | | | | | | | | * generic/tclIO.c (Tcl_UnregisterChannel): 'Tcl_Close' while the * generic/tclIO.c (Tcl_Close): close callbacks are run. Checked in 'Tcl_Close' and 'Tcl_Unregister' to prevent recursive call of 'close' in the close-callbacks. This is a possible error made by implementors of virtual filesystems based on 'tclvfs', thinking that they have to close the channel in the close handler for the filesystem. * generic/tclIO.c: * generic/tclIO.h: * Not reverting, but #ifdef'ing the changes from May 19, 2004 out of the core. This removes the ***POTENTIAL INCOMPATIBILITY*** for channel drivers it introduced. This has become possible due to Expect gaining a BlockModeProc and now handling blockingg and non-blocking modes correctly. Thus [SF Tcl Bug 943274] is still fixed if a recent enough version of Expect is used. * doc/CrtChannel.3: Added warning about usage of a channel without a BlockModeProc.
* * tclIO.c: Fixed [SF Tcl Bug 943274]. This is the same problem asandreas_kupries2004-05-191-1/+17
| | | | | | | | | | | | | * tclIO.h: [SF Tcl Bug 462317], see ChangeLog entry 2001-09-26. The fix done at that time is incomplete. It is possible to get around it if the actual read operation is defered and not executed in the event handler itself. Instead of tracking if we are in an read caused by a synthesized fileevent we now track if the OS has delivered a true event = actual data and bypass the driver if a read finds that there is no actual data waiting. The flag is cleared by a short or full read.
* * Updated APIs in the file generic/tclIO.c according to the guidelinesdgp2002-01-151-4/+4
| | | | | | | | | | | of TIP 27. Several minor documentation corrections as well. * Updated channel driver interface according to the guidelines of TIP 27. See also [Bug 500348]. * Moved Tcl_EolTranslation enum declaration from generic/tcl.h to generic/tclInt.h (renamed to TclEolTranslation). It is not used anywhere in Tcl's public interface.
* * The changes below fix [Bug #462317] where Expect tried to readandreas_kupries2001-09-271-1/+12
| | | | | | | | | | | | | | | | | | | | | | | more than was in the buffers and then blocked in the OS call as its pty channel driver provides no blockmodeproc through which the OS could be notified of blocking-behaviour. Because of this the general I/O core has to take more care than usual to preserve the semantics of non-blocking channels. * generic/tclIO.c (Tcl_ReadRaw): Do not read from the driver if the channel is non-blocking and the fileevent causing the read was generated by a timer. We do not know if there is data available from the OS. Instead of going to the OS for more and potentially blocking we simply signal EWOULDBLOCK to the higher levels to cause the system to wait for true fileevents. (GetInput): Same as before. (ChannelTimerProc): Added set and clear of CHANNEL_TIMER_FEV. * generic/tclIO.h (CHANNEL_TIMER_FEV): New flag for channels. Is set if a fileevent was generated by a timer, the channel is not blocking and the driver did not provide a blockmodeproc. In that case the I/O core has to be especially careful about going to the driver for more data.
* 2001-03-30 Andreas Kupries <andreas_kupries@users.sourceforge.net>andreas_kupries2001-03-301-1/+3
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | * All of the changes below belong to TIP #10 [Tcl I/O Enhancement: Thread-Aware Channels]. See also [Patch #403358] at SF. * generic/tclIO.h (struct ChannelState, line 236f): Extended the structure with a new field of type 'Tcl_ThreadId' to hold the id of the thread currently managing all channels with this state. Note: This structure is shared by all channels in a stack of transformations. * generic/tclIO.c (Tcl_CreateChannel, lines 1058-1065): Modified to store the Id of the current thread in the 'ChannelState' of the new channel. * generic/tclIO.c (Tcl_SpliceChannel, lines 2265-2270): Modified in the same manner as 'Tcl_CreateChannel' as the channel will be managed by the current thread afterward. * generic/tclIO.c (Tcl_GetChannelThread, lines 1478-1503): * generic/tcl.decls (Tcl_GetChannelThread, lines 1504-1506): New API function to retrieve the Id of the managing thread from a channel. Implementation and declaration. * generic/tclTest.c (TestChannelCmd, lines 4520-4532): Added subcommand 'mthread' to query a channel about its managing thread.
* up-port of the stacked channel implementation rewrite in 8.3.2 tohobbs2000-09-281-46/+77
| | | | 8.4a2 code base, merged in with some existing new 8.4a2 features.
* * generic/tclTest.c:hobbs2000-05-191-0/+348
* generic/tclIO.c: moved channel test commands from tclIO.c to tclTest.c. * generic/tclIO.h: new file, split out from tclIO.c to allow test commands to be moved to tclTest.c. * generic/tclStubInit.c: * generic/tclIntDecls.h: * generic/tclInt.decls: removed TclTestChannel*Cmd from internal stubs table and added TclChannelEventScriptInvoker to the internal stubs table so it can be used from the test code.