source: trunk/source/lisp-kernel/thread_manager.c

Revision Log Mode:


Legend:

Added
Modified
Copied or renamed
Diff Rev Age Author Log Message
(edit) @14233   9 years gb Don't use the (Linux-specific ?) MAP_GROWSDOWN mmap option when …
(edit) @14199   9 years gb Free the (separate) signal stack on thread exit on ARM Linux.
(edit) @14197   9 years rme Rename Threads.h to threads.h (with no capital letter).
(edit) @14119   9 years gb Changes from ARM branch. Need testing …
(edit) @13869   9 years gb For Windows: as suggested by 'maetbag' on #ccl: suspend_tcr() and …
(edit) @13455   10 years rme In count_cpus(): we can use sysconf(_SC_NPROCESSORS_ONLN) on Mac OS X …
(edit) @13434   10 years gb Add a cast to the Unix version of create_system_thread, to avoid a warning.
(edit) @13431   10 years gb Try to address ticket:649 by clearly distinguishing between …
(edit) @13279   10 years gb Lots of changes from "purify" branch, mostly involving: - new memory …
(edit) @13102   10 years gb In the windows version of raise_thread_interrupt(), don't try to use …
(edit) @13067   10 years rme Update copyright notices.
(edit) @13066   10 years rme Change "OpenMCL" to "Clozure CL" in comments and docstrings.
(edit) @12798   10 years gb Don't copy a thread's termination semaphore to the TCR (so don't …
(edit) @12476   10 years gb Whoops; no tcr->interrupt_context. Move along, nothing to see here.
(edit) @12475   10 years gb In the Windows version of suspend_tcr(), if the suspended thread is in …
(edit) @12379   10 years gb destroy_semaphore(): if lisp_global(IN_GC) is true in the …
(edit) @11761   11 years gb In lisp_thread_entry(): declare start_vsp unconditionally.
(edit) @11750   11 years gb In lisp_thread_entry() and cooperative_thread_startup(), ensure that …
(edit) @11742   11 years gb In suspend_resume_handler: if we don't have a TCR, create one (with …
(edit) @11712   11 years gb The Windows interrupt code seems safe (it just sets the deferred …
(edit) @11710   11 years gb Try to handle the case of suspending a Windows thread that's in the …
(edit) @11623   11 years gb Write debugging/diagnostic output to the FILE* "dbgout", which is …
(edit) @11609   11 years gb Back out of recent change. FreeBSD amd64 kernels don't change %fs on …
(edit) @11608   11 years gb FreeBSD 7.1 on i386 doesn't seem able to ensure that 'current_tcr' is …
(edit) @11581   11 years gb Hold TCR_AREA_LOCK when calling pthread_create(), to avoid Darwin lossage.
(edit) @11574   11 years gb free_tcr_extra_segment() for win32.
(edit) @11548   11 years gb Not really the bug I've been looking for, but ... allocate_tcr() …
(edit) @11501   11 years gb Try to use non-conflicting, platform-dependent signals instead of …
(edit) @11500   11 years gb Since SIQQUIT can be sent to the process, using a handler for it to …
(edit) @11375   11 years gb Declare rcontext_readonly.
(edit) @11370   11 years gb Using i386_set_ldt() to point %fs at the tcr on 32-bit FreeBSD doesn't …
(edit) @11345   11 years gb In 32-bit x86 FreeBSD setup_tcr_extra_segment(), exit (don't just …
(edit) @11328   11 years gb In setup_tcr_extra_segment() for X8632 FreeBSD: helps to set …
(edit) @11245   11 years gb In Solaris x8632 version of setup_tcr_extra_segment(), loop until we …
(edit) @11241   11 years gb 32-bit Solaris x86 changes.
(edit) @11231   11 years gb In windows suspend_tcr(): writing to a (new, uncommitted) page on …
(edit) @11184   11 years gb #include <sys/syscall.h> on Linux. use gettid() (which is only …
(edit) @11170   11 years gb Set tcr->shutdown_count to 1 in new_tcr() on Windows. …
(edit) @11149   11 years gb Need a cast to pthread_t in kill_tcr().
(edit) @11144   11 years gb TerminateThread?() needs another arg.
(edit) @11143   11 years gb Start to bootstrap a change in which (rude) thread termination …
(edit) @11115   11 years gb Always queue an APC when interrupting a thread that's in foreign code: …
(edit) @11095   11 years gb Create a thread-specific Windows event handle (a sort of one-shot …
(edit) @11082   11 years gb raise_thread_interrupt(): don't assume that the TCR has a non-null …
(edit) @11015   11 years gb FreeBSD x8632 changes.
(edit) @10953   11 years gb In Windows raise_thread_interrupt, setup call to interrupt_handler for …
(edit) @10933   11 years gb Work around win32 malloc alignment limitations when allocating tcr
(edit) @10916   11 years gb setup_tcr_extra_segment() for win32. (Note that it looked like we …
(edit) @10889   11 years gb Try to use new format-string constants. CALLBACK on thread startup …
(edit) @10831   11 years gb When interrupting a thread running foreign code, "queue an …
(edit) @10770   11 years gb If we HAVE_TLS, try to ensure that the TCR is allocated on a 16-byte …
(edit) @10765   11 years gb In linux x8632 version of setup_tcr_extra_segment(), segment limit is …
(edit) @10727   11 years gb Provide LDT setup/free stuff for ia32 Linux.
(edit) @10700   11 years gb Not sure if this is right, but resuming the thread before cancelling …
(edit) @10597   11 years gb Use WIN_64 (not WIN64) to conditionalize.
(edit) @10594   11 years gb Reference 'restore_windows_context*', not win64-specific. Stub out …
(edit) @10565   11 years gb Merge changes from branches/win64. As well as the expected low-level …
(edit) @10464   11 years gb Call pthread_attr_destroy() on pthread_attr objects when done with …
(edit) @10462   11 years gb in shutdown_thread_tcr(), zero the tlb_limit field before deallocating …
(edit) @10386   11 years gb Always use SIGRETURN macro to return from suspend_resume_handler().
(edit) @10385   11 years gb Remove some old/experimental conditionalization, making some of this …
(edit) @10183   11 years gb Need an os_get_stack_bounds() implementation for Solaris.
(edit) @10177   11 years gb lisp_suspend_other_threads: lock tcr_area_lock before. …
(edit) @10102   11 years rme Additions for Darwin/x8632.
(edit) @10024   11 years gb Use Chris Curtis' idea (_lwp_private()) for setup_tcr_extra_segment() …
(edit) @10010   11 years gb Mostly cosmetic changes (to avoid some gcc warnings at -Wall). …
(edit) @9997   11 years gb Before calling pthread_kill(), ensure that tcr->osid is non-zero …
(edit) @9901   11 years gb Remove unused variables. (May need to compile with -Wall to find more …
(edit) @9785   11 years gb Try to ensure that the stack size(s) passed around are "natural" …
(edit) @9303   11 years gb ==, not =, when comparing err to EINTR in rwlock_wlock. (Not sure if …
(edit) @8547   12 years andreas Big #ifdef WINDOWS hack&slash: stub out every function that references …
(edit) @8408   12 years gb Win64 changes; see if they break other platforms.
(copy) @8177   12 years gb move to trunk
copied from branches/1.2/devel/source/lisp-kernel/thread_manager.c:
(copy) @8123   12 years gb relocate
Note: See TracRevisionLog for help on using the revision log.