from the USER_STACK_REGION (until B_BASE_ADDRESS is available), made it much better
readable. It now grows from the lower USER_STACK_REGION up to the start of the main
thread's stack location.
Calls arch_thread_tls_init() and accounts for the TLS area in the stack region.
Added some comments, some cleanups.
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@2382 a95241bf-73f2-0310-859d-f6bbb57e9c96
area. It now also calls arch_thread_init_tls(). Cleaned it up to make it more
readable, added some comments.
Some other cleanups.
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@2381 a95241bf-73f2-0310-859d-f6bbb57e9c96
and TSS_BASE_SEGMENT in i386_set_tss_and_kstack() (thanks Daniel for pointing this
out)).
Added a arch_thread_init_tls() function which fills the reserved slots - the find_thread()
inline asm in OS.h is finally working.
Added some more comments, resolved the ToDo items from last time :-)
arch_thread_context_switch() now calls set_tls_context() at the correct location.
arch_thread_enter_uspace() now calls set_tls_context() as there is no context
switch in this case for the current thread after TLS has been set up, and it
now gets a pointer to the thread structure directly, and calculates the user
stack top itself, rather than the callee.
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@2380 a95241bf-73f2-0310-859d-f6bbb57e9c96
at all, and assumes a fixed and wrong number of preallocated GDT entries.
Implemented TLS: there is one GDT entry per CPU. If a context switch happens,
the FS register of the new thread is set to the matching GDT, and the GDT is
changed so that it points to the current thread's TLS storage area.
This area currently resides unsafely at the bottom of the user stack - for
some reason I could not figure out, it doesn't even work correctly most of
the time (it segfaults when accessing a slot via FS). I've added a ToDo item
explaining the situation - hopefully I have more ideas when I slept a bit more...
The GDT is now no longer static in arch_cpu.c and has been renamed from gdt to
gGDT. It's now also referenced in arch_thread.c, and it would make sense for
arch_selector.c to use it as well (instead of another local copy).
arch_cpu_init2() now uses the set_tss_descriptor() inline function and the
TSS_BASE_SEGMENT macro to set up the TSS section. It now also sets up the
TLS segment descriptors (as TSS, one entry per CPU).
Since I removed desc_table from the headers (the GDT is now a (segment_descriptor *)),
I added it locally to arch_int.c.
i386_enter_uspace() now don't set the FS register to 0x23 anymore, since it's
now already set correctly at the end of arch_thread_context_switch().
Some clean-ups.
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@2366 a95241bf-73f2-0310-859d-f6bbb57e9c96
I was too lazy to revert that little change back, so that I would be able
to write this short paragraph explaining what I've done.
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@2364 a95241bf-73f2-0310-859d-f6bbb57e9c96
some structure definitions (to outline some basic ideas), and the symbols
exported to userland.
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@2214 a95241bf-73f2-0310-859d-f6bbb57e9c96
Some int -> cpu_state cleanup on the way, and implement the
B_KDEBUG_CONT and B_KDEBUG_QUIT features in our kernel debugger.
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@2173 a95241bf-73f2-0310-859d-f6bbb57e9c96
code and made it much simpler (it now just uses a stack of path names
instead of this complex doubly-linked module_iterator_dir list).
Now works together with the rest of the code without making any problems.
Added a module_test() function which iterates over all existing modules.
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@2149 a95241bf-73f2-0310-859d-f6bbb57e9c96
Some cleanups, moved user_*() functions to the end of the file.
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@2144 a95241bf-73f2-0310-859d-f6bbb57e9c96
field to the thread structure to keep track of the current state.
forbids/permits can be nested.
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@2141 a95241bf-73f2-0310-859d-f6bbb57e9c96
disabled. If a page fault happend while a interrupt handler is executed,
or while a section of code is running with interrupts disabled (with the
exception of the initial kernel startup), the page fault handler will
now panic, to give you the chance to fix the bug.
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@2124 a95241bf-73f2-0310-859d-f6bbb57e9c96
thread_exit() for now prints a warning if it was called with interrupts
disabled.
Fixed a bug in thread_exit2() that restored the interrupts too early,
instead of simply enabling them (and restoring them later) [found by
Marcus O. great debug additions].
Fixes for the khash changes.
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@2119 a95241bf-73f2-0310-859d-f6bbb57e9c96
Now the d_reclen field of struct dirent is correctly calculated in all
file systems (read_dir() function).
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@2114 a95241bf-73f2-0310-859d-f6bbb57e9c96
leak fixed, some other bugs fixed, removed the global queues as there weren't
needed and used at all, moved to the other hash table implementation
(because the new_hash_table stuff was buggy). Should really work okay now :-)
Added the last missing function get_next_loaded_module_name() - I am not
sure the Be version works this way, but we could easily change that after
actually having found out how that one is really working.
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@2113 a95241bf-73f2-0310-859d-f6bbb57e9c96
and to have better performance on Hyper Threading systems. Should be
backward compatible with all x86 systems, according to Intel documentation.
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@2068 a95241bf-73f2-0310-859d-f6bbb57e9c96
some minor rearrangement (sys_xxx() calls now placed together)
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@1997 a95241bf-73f2-0310-859d-f6bbb57e9c96
