aboutsummaryrefslogtreecommitdiff
path: root/libports/manage-multithread.c
Commit message (Collapse)AuthorAgeFilesLines
* 2008-12-29 Samuel Thibault <samuel.thibault@ens-lyon.org>Samuel Thibault2008-12-291-2/+0
| | | | | * manage-multithread.c (ports_manage_port_operations_multithread): Drop reset of thread_timeout and global_timeout to 0.
* *** empty log message ***Thomas Bushnell1997-08-201-4/+0
|
* Wed Aug 20 13:56:40 1997 Thomas Bushnell, n/BSG <thomas@gnu.ai.mit.edu>Thomas Bushnell1997-08-201-7/+4
| | | | | | | | | * manage-multithread.c (ports_manage_port_operations_multithread): Drop parms `wire_cthreads' and `wire_threads'. Add parm `hook'. (thread_function): Call HOOK instead of cthread_wire and/or thread_wire. * ports.h (ports_manage_port_operations_multithread): Adjust declaration.
* (ports_manage_port_operations_multithread): Fill in meaningfulMichael I. Bushnell1996-03-281-4/+32
| | | | statuses and errors rather than just zero.
* (ports_manage_port_operations_multithread):Miles Bader1996-01-271-8/+15
| | | | Pass INP->msgh_id to ports_begin_rpc, and only call DEMUXER if it returns 0.
* Don't wire timeouts to zero.Roland McGrath1995-11-221-11/+9
|
* (ports_manage_port_operations_multithread): Actually, just useMichael I. Bushnell1995-11-211-20/+3
| | | | hurd_thread_cancel and then begin the RPC normally.
* (ports_manage_port_operations_multithread): Lock _ports_lock aroundMichael I. Bushnell1995-11-211-1/+6
| | | | new critical section.
* (ports_manage_port_operations_multithread/internal_demuxer): If theMichael I. Bushnell1995-11-211-1/+17
| | | | | | incoming RPC has already been cancelled, then call cancel_rpc instead of the user's demuxer. (cancel_rpc): New function.
* (ports_manage_port_operations_multithread/internal_demuxer): Don'tMichael I. Bushnell1995-10-241-4/+9
| | | | attempt RPC if we can't get a valid port struct.
* (ports_manage_port_operations_multithread): Hammer WIRE_CTHREADS onMichael I. Bushnell1995-08-291-0/+4
| | | | for now.
* (ports_manage_port_operations_multithread): For now, wireMichael I. Bushnell1995-07-211-0/+2
| | | | GLOBAL_TIMEOUT and THREAD_TIMEOUT to zero.
* (ports_manage_port_operations_multithread): typoMichael I. Bushnell1995-07-181-0/+1
|
* (ports_manage_port_operations_multithread) [thread_function]: ReturnMichael I. Bushnell1995-07-181-4/+2
| | | | int. Don't cthread_exit child threads; just let them return normally.
* (ports_manage_port_operations_multithread) [internal_demuxer]:Michael I. Bushnell1995-07-121-8/+11
| | | | | | Increment NREQTHREADS and TOTALTHREADS *before* forking newthread. [thread_function]: Don't increment NREQTHREADS and TOTALTHREADS here. Initialize NREQTHREADS and TOTALTHREADS in main body of function.
* (ports_manage_port_operations_multithread): Declare NREQTHREADS asdMichael I. Bushnell1995-06-261-2/+2
| | | | TOTALTHREADS volatile.
* (ports_manage_port_operations_multithread): Declare ERR insideMichael I. Bushnell1995-06-261-1/+1
| | | | THREAD_FUNCTION so it doesn't get shared by multiple threads.
* (ports_manage_port_operations_multithread): When master goes back toMichael I. Bushnell1995-06-261-1/+4
| | | | wait for messages, it should unlock LOCK.
* (ports_manage_port_operations_multithread/thread_function): Don't holdMichael I. Bushnell1995-06-261-1/+2
| | | | lock while setting TIMEOUT.
* Initial revisionMichael I. Bushnell1995-06-201-0/+123