Commit Graph

25416 Commits

Author SHA1 Message Date
Stephan Aßmus
3a0814fdae patch by Andrej Spielmann (GSOC):
* Prepared the AGG base_renderer code with two more functions needed to
  handle subpixel scanline coverage values.


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26356 a95241bf-73f2-0310-859d-f6bbb57e9c96
2008-07-10 08:11:04 +00:00
Stephan Aßmus
c816464287 patch by Andrej Spielmann (GSOC):
Added complementary functions to the set of functions that implement a
drawing_mode, these new functions interpret the coverage values passed
from the AGG rasterizer or another scanline storage as subpixel triplets.


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26355 a95241bf-73f2-0310-859d-f6bbb57e9c96
2008-07-10 08:08:59 +00:00
Stephan Aßmus
60d98a49ed patch by Andrej Spielmann (GSOC):
* Added marco definitions for subpixel based blending of two 32 bit pixels.


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26354 a95241bf-73f2-0310-859d-f6bbb57e9c96
2008-07-10 08:06:59 +00:00
Stephan Aßmus
674c8bc8bb Added TODO about a problem that the late mouse moved message dropping code
may have.


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26353 a95241bf-73f2-0310-859d-f6bbb57e9c96
2008-07-10 07:35:47 +00:00
Stephan Aßmus
dde9faab86 * Made the mouse moved message dropping code slightly nicer to read.
* Don't drop messages which carry an important transit value. (Thanks, Axel!)


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26352 a95241bf-73f2-0310-859d-f6bbb57e9c96
2008-07-10 07:21:16 +00:00
Jérôme Duval
6d3c6e761a reduced buffer size: with 4096, no buffer is received by the multiaudio addon
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26349 a95241bf-73f2-0310-859d-f6bbb57e9c96
2008-07-09 23:14:57 +00:00
Ingo Weinhold
f937627dc8 Blue screen pagination:
* Changed prompt and keys: Now it's 'q' to quit (abort) the debugger
  command and 's' to skip the following output.
* Don't offer or try to quit a command in the boot debug output.
* Cleanup the prompt and go to the next line also when aborting the
  command.


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26348 a95241bf-73f2-0310-859d-f6bbb57e9c96
2008-07-09 22:59:45 +00:00
Axel Dörfler
8d79c7db81 * Made Dump() const.
* Removed superfluous whitespace.


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26347 a95241bf-73f2-0310-859d-f6bbb57e9c96
2008-07-09 21:32:37 +00:00
Jérôme Duval
14f3368d36 fix warning
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26346 a95241bf-73f2-0310-859d-f6bbb57e9c96
2008-07-09 18:26:16 +00:00
Oliver Ruiz Dorantes
d04eb939af -Move the launching of the Local Device to an IOCTL from the open hook to be more compatible with the net_stack
-Implement Launch method in the accesors regarding previous comment
-Fix KDL at removing device



git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26345 a95241bf-73f2-0310-859d-f6bbb57e9c96
2008-07-09 17:01:30 +00:00
Ingo Weinhold
383a9ac435 * Added an explicit boot_splash_uninit(). Not really needed, but makes
it clear when boot_splash_set_stage() must no longer be used.
* Free the memory associated with the kernel args before starting the
  init process. Unlike the original TODO stated there are quite a few
  more users of the kernel args (including the boot splash screen),
  hence we can't really do that earlier, unless we decide to copy the
  data over to the kernel heap. 


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26344 a95241bf-73f2-0310-859d-f6bbb57e9c96
2008-07-09 15:25:01 +00:00
Ingo Weinhold
fa0a5cd0e8 insert_preloaded_image() now copies the debug symbols and string table
over to the kernel heap. This allows us to simply free them in
unload_elf_image(). Now we no longer leak the memory for non-preloaded
images. On my T61 this amounted to more than 10 MB of kernel heap after
booting.


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26343 a95241bf-73f2-0310-859d-f6bbb57e9c96
2008-07-09 15:17:39 +00:00
Axel Dörfler
daa861bc29 Work-in-progress of the module notification implementation:
* Most things are working, it just doesn't notify the listeners on changes yet.


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26342 a95241bf-73f2-0310-859d-f6bbb57e9c96
2008-07-09 14:23:29 +00:00
Stephan Aßmus
bb73c05fd7 * Introduced a new view event mask flag: B_FULL_POINTER_HISTORY which,
when set, prevents any old mouse moved message discarding.
* BWindow::DispatchMessage(B_MOUSE_MOVED) checks the event time of the
  message and discards too old events, but only if there is another event
  in the queue and the view does not specify B_FULL_POINTER_HISTORY.
* BView::GetMouse() ignores the checkHistory flag passed to the function
  in case the event mask specifies B_NO_POINTER_HISTORY.
  B_FULL_POINTER_HISTORY on the other hand prevents the dropping of old
  messages.


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26341 a95241bf-73f2-0310-859d-f6bbb57e9c96
2008-07-09 14:13:03 +00:00
Axel Dörfler
191a03fd11 * {add|remove}_node_listener() erroneously had "dev_t" as type of "node".
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26340 a95241bf-73f2-0310-859d-f6bbb57e9c96
2008-07-09 13:04:54 +00:00
Axel Dörfler
9c5c79ad86 * Made DIR definition private, and moved it into the new dirent_private.h
header.
* Added fields necessary to make it possible to retrieve more than one
  dirent from the underlying file system.
* Unless some app creates a DIR on its own to it to feed readdir(), this
  change should be binary compatible. If we find an application misbehaving,
  we can still make it a GCC4 only thing.
* fs_attr/fs_index/fs_query now all use readdir() directly (as that one
  contains the logic to iterate through a number of dirents in userspace).


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26339 a95241bf-73f2-0310-859d-f6bbb57e9c96
2008-07-09 11:16:00 +00:00
Axel Dörfler
6ad2f357dc * Both, rootfs, and devfs did not correctly report the number of entries
returned by fs_read_dir().
* Removed superfluous white space.


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26338 a95241bf-73f2-0310-859d-f6bbb57e9c96
2008-07-09 10:59:02 +00:00
Stephan Aßmus
fe3d5b3083 Implemented a simple fix for the message-pile-up problem of
BView::GetMouse( , ,useHistory = true) in case the application
calls GetMouse() in a loop with a longer delay then mouse
messages arrive at the queue. The "when" field of the messages
is used to discard old mouse moved messages. This also fixes
the possible problem of finding the same message over and over
in case it is not removed from the queue for other reasons.
This fix makes selecting text in Pe for example usable.


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26337 a95241bf-73f2-0310-859d-f6bbb57e9c96
2008-07-09 10:03:28 +00:00
Ingo Weinhold
74d5c1e499 * Fixed serious memory leak in team_create_thread_start(). The
destructor of the automatic KPath variable was never invoked, since
  normally arch_thread_enter_userspace() would not return and thus the
  function scope never be left. After a standard "jam @image" almost
  42 MB of kernel heap were lost this way.
* Added a few warning comments in functions that shouldn't use automatic
  variables with a destructor in the function scope either.


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26336 a95241bf-73f2-0310-859d-f6bbb57e9c96
2008-07-09 03:58:38 +00:00
Ingo Weinhold
d1d7044ed1 Added TODO regarding memory leak. Too tired ATM to solve it.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26335 a95241bf-73f2-0310-859d-f6bbb57e9c96
2008-07-09 03:04:39 +00:00
Ingo Weinhold
b64e66fbd0 Changes related to the "caches" command:
* The table used for sorting the caches could be too small for the
  number of caches in the system. We never checked that, which could
  cause a crash or triple faults when invoking the command.
* We now allocate the table as an area. Also increased the table size to
  102400 entries.
* Also print the committed memory for non-RAM caches, if they are the
  root cache of a cache tree that has committed memory.


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26333 a95241bf-73f2-0310-859d-f6bbb57e9c96
2008-07-09 01:56:58 +00:00
Marcus Overhagen
0cd4d5fa02 Report SCSI_STATUS_GOOD after successful ATAPI commands. This makes CDDA-fs work with AHCI.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26332 a95241bf-73f2-0310-859d-f6bbb57e9c96
2008-07-08 22:39:03 +00:00
Stephan Aßmus
f3312751b9 honor 80 char limit in comment
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26331 a95241bf-73f2-0310-859d-f6bbb57e9c96
2008-07-08 22:31:26 +00:00
Jérôme Duval
3787196a5a * some methods to access stream registers were defined with wrong types
* use HDAC_BIDIR_STREAM_OFFSET and HDAC_OUTPUT_STREAM_OFFSET when applicable
* use a PAGE_ALIGN macro


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26330 a95241bf-73f2-0310-859d-f6bbb57e9c96
2008-07-08 21:48:50 +00:00
Jérôme Duval
f67bbdb2a6 fixed warnings
use playback rate as preferred rate (instead of record)


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26329 a95241bf-73f2-0310-859d-f6bbb57e9c96
2008-07-08 21:31:14 +00:00
Stephan Aßmus
264df95726 All input device add-ons should already do this, but make sure that
mouse messages have a "when" field with the event system time.


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26328 a95241bf-73f2-0310-859d-f6bbb57e9c96
2008-07-08 20:29:18 +00:00
Stephan Aßmus
d36ae78cc6 Added two tests:
* statusbar was used to test and improve some BStatusBar behavior
* lagging_get_mouse demonstrates what the problem is with older
  applications using synchronous GetMouse() calls without specifying
  that they actually don't care about mouse history.


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26327 a95241bf-73f2-0310-859d-f6bbb57e9c96
2008-07-08 20:01:31 +00:00
Oliver Ruiz Dorantes
48cca06a81 Add new virtual method in the delegates to launch the device, Instantiation meant only to open/register the localdevice
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26326 a95241bf-73f2-0310-859d-f6bbb57e9c96
2008-07-08 18:39:41 +00:00
Stephan Aßmus
6be60b2d19 Applied patch by Jan Kloetzke:
The PUSHF/POPF instructions were not emulated correctly.


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26325 a95241bf-73f2-0310-859d-f6bbb57e9c96
2008-07-08 18:25:45 +00:00
Stephan Aßmus
fc8e77f863 When a BBitmap is drawn with scale, but also only partly, the inverse
image scan generator matrix was calculated wrongly. The part of the offset
that lies within the bitmap bounds needs to have the scale applied as well.
Maybe this code can be simplified. Appearantly there is not a lot of code
that uses BBitmap drawing this way, perhaps because the R5 version has had
issues with rounding. But MediaPlayer uses this feature for drawing the peak
levels and this is fixed by this change.


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26324 a95241bf-73f2-0310-859d-f6bbb57e9c96
2008-07-08 17:48:30 +00:00
Stephan Aßmus
93016537eb Check if the Painter even has a clipping region set before trying to take the
short cut when no text needs to be rendered. I've seen a crash yesterday in
the app_server test environment when the decorater was drawing something,
although it may also have been because I had a screwed up objects folder
where some objects were not recompiled because I am switching back and forth
between two app_server code folders.


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26323 a95241bf-73f2-0310-859d-f6bbb57e9c96
2008-07-08 17:41:15 +00:00
Stephan Aßmus
bafcc63bcc This is either a fix or a workaround for a problem in Pe, where the
application modifies the scrollbars one by one and the changes would
have some weird cyclic effect where the constrains of one scrollbar
affected the scrolling already while they would be lifted afterwards.
It sounds weird, but maybe it is simply a resulting behavior of the
BeOS implementation which does not check the scrollbar restrictions
in the BView code. 


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26322 a95241bf-73f2-0310-859d-f6bbb57e9c96
2008-07-08 17:21:39 +00:00
Jérôme Duval
81a7dedd6b init buffer_cycle
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26321 a95241bf-73f2-0310-859d-f6bbb57e9c96
2008-07-08 16:41:52 +00:00
Ingo Weinhold
65748dc8d9 Remove() was moving an additional array element. Mostly harmless, but
would cause a crash when the array end was page-aligned and the next
page was not accessible. Happened to me during a "rm -rf ...".


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26320 a95241bf-73f2-0310-859d-f6bbb57e9c96
2008-07-08 16:15:41 +00:00
Axel Dörfler
1bde8b03e7 * Replaced BFS's ReadWriteLock implementation with the kernel's rw_lock.
* The BlockAllocator now uses the new mutex_transfer_lock() function instead
  of clobbering the mutex directly.
* Removed Lock.h - it's no longer needed.
* Minor white space cleanup.


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26317 a95241bf-73f2-0310-859d-f6bbb57e9c96
2008-07-08 08:04:37 +00:00
Axel Dörfler
78b32dd026 * The rw_lock is now able to nest read/write locks when you have a write lock.
It follows the semantics of the BFS R/W lock, though, that is, if you unlock
  your write lock before the read locks, the read locks effectively become
  write locks, too.
* Added a mutex_transfer_lock() function that will allow you to unlock a mutex
  in a different thread than the one which locked it (only matters if KDEBUG
  is enabled, though).


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26316 a95241bf-73f2-0310-859d-f6bbb57e9c96
2008-07-08 07:56:49 +00:00
Michael Lotz
f90a5e346d Adding bus_barrier macro to compatibility layer.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26312 a95241bf-73f2-0310-859d-f6bbb57e9c96
2008-07-07 21:52:40 +00:00
Oliver Tappe
a042907733 * removed debuggin leftover
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26311 a95241bf-73f2-0310-859d-f6bbb57e9c96
2008-07-07 21:06:04 +00:00
Jérôme Duval
b5f75a7ec8 forgot to close the stream before deletion
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26310 a95241bf-73f2-0310-859d-f6bbb57e9c96
2008-07-07 20:45:54 +00:00
Karsten Heimrich
685483eada * set the what field to okok on success, as the epson returns 'Good' for some reason
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26309 a95241bf-73f2-0310-859d-f6bbb57e9c96
2008-07-07 20:04:03 +00:00
Karsten Heimrich
c5ea3d591d * rearranged header a bit
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26308 a95241bf-73f2-0310-859d-f6bbb57e9c96
2008-07-07 20:01:46 +00:00
Karsten Heimrich
a836d47d56 * cleanup, no functional change
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26307 a95241bf-73f2-0310-859d-f6bbb57e9c96
2008-07-07 19:39:10 +00:00
Karsten Heimrich
30cf4e6a54 * revert the last change as we can get stop etc. also
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26306 a95241bf-73f2-0310-859d-f6bbb57e9c96
2008-07-07 19:23:40 +00:00
Karsten Heimrich
7630cacab6 * check against baad as e.g. the epson addon does not return okok on success
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26305 a95241bf-73f2-0310-859d-f6bbb57e9c96
2008-07-07 19:14:08 +00:00
Karsten Heimrich
f56301c9e8 * maintain our own rect while printing, this makes it possible to print all pages e.g. syslog from StyledEdit
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26304 a95241bf-73f2-0310-859d-f6bbb57e9c96
2008-07-07 19:10:39 +00:00
Oliver Tappe
3ad86c7ff5 Fixed R5 compatibility "layer" in libnetwork:
* instead of always converting from the expected r5_sockaddr_in to haiku's own,
  we now explicitly check whether or not the given sockaddr is an r5_sockaddr_in
  or not, naturally doing the conversion only if it is. This is necessary since
  even R5 applications may not always pass in r5_sockaddr_in structs (as for
  instance gethostbyname() will return a native [haiku-]sockaddr_in)
* cleaned up the confusion between the name r5addr and it's actual meaning
  (holding a haiku sockaddr_in) - renaming it to haikuAddr instead
* undid the part of Ingo's r25489 described as: "Extended R5 compatibility 
  check to also consider calls from libbnetapi" - as I fail to see why this
  would be desirable and in fact it stops at least Beam from working.
  Ingo: if you can remember, please enlighten me what was the reason behind 
  this change.
This finally makes Beam "work" (well: connect to servers and download mails ;-)
Vision, NetworkTime and NetPositive are still working, too, so hopefully there
are no regressions.

git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26303 a95241bf-73f2-0310-859d-f6bbb57e9c96
2008-07-07 18:53:01 +00:00
Axel Dörfler
06d5e1b3b6 * Added ReadLocker/WriteLocker classes to auto lock an rw_lock.
* Added *_init_etc() functions to the fs_shell.


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26302 a95241bf-73f2-0310-859d-f6bbb57e9c96
2008-07-07 18:12:33 +00:00
Axel Dörfler
e2dfe00484 * Replaced most of the error enums with #defines.
* fssh_auto_locker.h should probably better use an FS-Shell header instead of
  stddef.h.


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26301 a95241bf-73f2-0310-859d-f6bbb57e9c96
2008-07-07 18:10:21 +00:00
Axel Dörfler
846f4fa2ab * Work-around for the KDEBUG "holder" mechanism; we actually want to lock the
mutex in another thread here than we will unlock it.


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26300 a95241bf-73f2-0310-859d-f6bbb57e9c96
2008-07-07 16:34:38 +00:00
Axel Dörfler
e0fc5e7380 * The fssh_auto_locker.h header was not self-contained.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26299 a95241bf-73f2-0310-859d-f6bbb57e9c96
2008-07-07 16:26:57 +00:00