Fixed the application signature in the .rdef file.
Removed some "how to write an .rdef file" comments.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@12124 a95241bf-73f2-0310-859d-f6bbb57e9c96
if $TARGET_PLATFORM requires this.
Note, "jam app_server" for R5 did not work for me before and does not work now
although the number of errors are reduced - whoever knows the requirements for
the app_server under R5 better than me, please fix this.
Added the app_server as well as some more drivers to the Haiku HD image. Since
RUN_WITHOUT_APP_SERVER is still defined, it won't work correctly, though.
For the time being, you need to manually alter this and the Bootscript in order
to run it.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@12121 a95241bf-73f2-0310-859d-f6bbb57e9c96
- all libraries needed for the app_server are now on the image
- the fonts will be installed (currently requires to have built the app_server before)
- includes MiniTerminal
- includes the radeon driver and accelerant (others will follow)
- creates beos/apps and beos/prefs directories
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@12117 a95241bf-73f2-0310-859d-f6bbb57e9c96
process ID), it also searched for a group with a negative ID.
Some minor cleanup.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@12111 a95241bf-73f2-0310-859d-f6bbb57e9c96
any chance to come back (without being signaled), it now will.
For example, this is used by the sleep command (which now actually works
as it should :)).
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@12110 a95241bf-73f2-0310-859d-f6bbb57e9c96
- SupportDefs.h only includes stdbool.h if included from C
- stdbool.h for C++ now includes a macro for "bool" as defined by that header
- stdbool.h does nothing if __bool_true_false_are_defined is already defined
- stdbool.h no longer defines a _Bool enum, but defines _Bool as unsigned char, as
previously done by SupportDefs.h (the previous version changed the size).
- The gensyscalls Jamfile now preprocesses its headers in C++ mode so that "bool"
stays "bool", and doesn't become _Bool.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@12042 a95241bf-73f2-0310-859d-f6bbb57e9c96
Implemented them and moved them to unistd/terminal.c - not yet tested, though,
but should work. As a side effect, the TTY should now send signals.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@12034 a95241bf-73f2-0310-859d-f6bbb57e9c96
instead of hoardGetThreadID()).
Removed unused arch functions.
__init_heap() now calls hoardHeap::initNumProcs(); it's no longer initialized
through a global constructor - that ensures that the heap can really be used
by libroot's constructors (although it did work before as well, somehow).
Moved the heap to 0x30000000 for now (limiting its size, but also ensuring free space
after it) - until the VM can avoid the free space after the heap.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@11959 a95241bf-73f2-0310-859d-f6bbb57e9c96
The new heap backend will probably be taken from the boot loader's heap
implementation (the free chunk stuff).
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@11958 a95241bf-73f2-0310-859d-f6bbb57e9c96
that we no longer need an initialize_after().
Put the architecture dependent hoard functions into the BPrivate namespace
as well. Minor cleanup.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@11956 a95241bf-73f2-0310-859d-f6bbb57e9c96
(and added support for that function in the emulated kernel).
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@11951 a95241bf-73f2-0310-859d-f6bbb57e9c96
properties; it should pass on to the next handler (and only if there is none
reply with a B_MESSAGE_NOT_UNDERSTOOD).
Those B_MESSAGE_NOT_UNDERSTOOD are no longer the answer to a B_MESSAGE_NOT_UNDERSTOOD
message anymore, as this lead to an endless loop (for example when starting the
media_server).
Added some ToDo comments - someone please check what has to be done and fill it
out some more.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@11950 a95241bf-73f2-0310-859d-f6bbb57e9c96
-S will dump over the serial line (kernel debug output).
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@11948 a95241bf-73f2-0310-859d-f6bbb57e9c96
Renamed RLD_STARTUP() runtime_loader() - the new entry point of the loader.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@11943 a95241bf-73f2-0310-859d-f6bbb57e9c96
from kernel/Jamfile. Actually kernel/apps is still there as Subversion
doesn't let me commit it -- I'll remove that later.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@11940 a95241bf-73f2-0310-859d-f6bbb57e9c96
Moved the rld.so creation from kernel/Jamfile to kernel/runtime_loader/Jamfile.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@11936 a95241bf-73f2-0310-859d-f6bbb57e9c96
The default option of "m" is now honored.
Prints out <BEEP> to the debug output instead of printing it on-screen.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@11934 a95241bf-73f2-0310-859d-f6bbb57e9c96
now uses this one instead of using map() for every single page - this should
be faster in most cases, and also, more importantly, does not mess up the
map_count counter anymore.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@11916 a95241bf-73f2-0310-859d-f6bbb57e9c96
to work when jam was executed in the local directory (as I did for testing).
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@11914 a95241bf-73f2-0310-859d-f6bbb57e9c96
removed unused headers. Adapted sources to still compile with the new
header locations.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@11913 a95241bf-73f2-0310-859d-f6bbb57e9c96
so that mails I get from Mr. Wilber are shown correctly, too :)
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@11906 a95241bf-73f2-0310-859d-f6bbb57e9c96
Makes sure the test apps will also build with target=haiku.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@11894 a95241bf-73f2-0310-859d-f6bbb57e9c96
be private, as it serves no useful purpose outside of the kernel.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@11893 a95241bf-73f2-0310-859d-f6bbb57e9c96
team boundaries; if you didn't actually call close() from within the
application, the close-hook of the file system was never called.
Also, you could close files of other teams (ie. invoke close on a
shared file descriptor).
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@11892 a95241bf-73f2-0310-859d-f6bbb57e9c96
they are public now, they might become private later on.
It's almost the same as acquire_sem() but allows you to release another semaphore
atomically. This makes condvar implementations and the like very simple to do.
Added B_CHECK_PERMISSION flags to the user calls, although it's not honored yet.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@11887 a95241bf-73f2-0310-859d-f6bbb57e9c96
is perfectly okay on x86 anyway, but should be moved into the arch stuff).
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@11869 a95241bf-73f2-0310-859d-f6bbb57e9c96
Should work correctly, although there is a big race condition left when you're
using F_SETLKW. Will be fixed next.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@11868 a95241bf-73f2-0310-859d-f6bbb57e9c96
simple cases - it might not work properly though, since we don't know yet if
a cache has children; therefore it will print out a warning as a constant
reminder if used in that way. It will still not work if we would need to
insert a new cache inbetween (in case the cache has children or only read-only
areas (even if there is no way yet to specify how exactly the VM should behave
in this case yet).
Added a comment to vm_map_file() about what we still need to do there.
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@11751 a95241bf-73f2-0310-859d-f6bbb57e9c96
server accepts input (r5 method). This doesn't work yet, though, as our
input_server fires up the event loop too early.
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@11731 a95241bf-73f2-0310-859d-f6bbb57e9c96
just create an attribute directory.
Removed the now unnecessary UNSAFE_GET_VNODE lines.
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@11724 a95241bf-73f2-0310-859d-f6bbb57e9c96
bfs_inode::InitCheck() tested INODE_NOT_READY before making sure the block is
an inode at all, resulting in spurious "inode is not becoming unbusy" errors.
Since there is now publish_vnode() we don't even have to return B_BUSY at all
anymore, though - INODE_NOT_READY is no longer used at all.
bfs_inode::etc is no longer used either (and for the same reason).
Inode::Create() no longer locks the whole volume, but only the write lock
of the parent directory (if any).
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@11723 a95241bf-73f2-0310-859d-f6bbb57e9c96
can now just do what it's supposed to do, even the bfs_inode::etc pointer is
no longer used.
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@11722 a95241bf-73f2-0310-859d-f6bbb57e9c96
hold the lock and still call subsequent Lock() and LockWrite() without
the deadlock penalty.
Added another constructor to WriteLocked so that you can also pass NULL
to it (in this case, nothing happens, of course :)).
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@11720 a95241bf-73f2-0310-859d-f6bbb57e9c96
function publish_vnode(). When calling new_vnode(), the node is only
reserved, but not yet accessible for others. Only when you call
publish_vnode(), it will become available. That simplifies new node
handling/locking considerably.
Note, you don't have to call new_vnode() before publish_vnode(); ie.
publish_vnode() acts exactly like new_vnode() did before if used that
way.
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@11719 a95241bf-73f2-0310-859d-f6bbb57e9c96
Axel Dörfler