First implementation of reading sectors from an SD card.
This is not the best performance for many reasons:
- No DMA
- Reads only one sector at a time
- Cannot read more than 512 bytes per syscall
Also there are major limitations:
- Cannot read less than 512 bytes. The hardware of course works in full
sectors. The mmc_disk driver should go through the io scheduler to
make sure requests have a reasonable size and offset, and nothing
tries to read just a few bytes in the middle of a sector.
- SD cards only (no SDHC, no MMC)
Architecture problems:
I think too much of the implementation is done in sdhci_pci and should
be moved to the upper layers. However it is difficult to say without
having implemented DMA (which indeed will be at the low level of the
sdhci controller). It doesn't help that the order of operations is a
bit different depending on wether there is DMA or not. In DMA mode you
first prepare the buffer, then run the command. In non-DMA mode you
first send the command, then read the data into the buffer. We need an
API at the mmc_bus level that doesn't care about that low-level detail.
There are other things that the MMC bus should be doing however, such
as switching to different clock speeds depending on which card is
activated and how fast it can go.
At least the following should be done:
- The read method for mmc_bus and sdhci_pci should use a scatter-gather
structure as a parameter instead of a single buffer
- See if can be integrated into ExecuteCommand at sdhci level (it's
essentially a command with an additional data phase)
Change-Id: I688b6c694561074535c9c0c2545f06dc04b06e7d
Reviewed-on: https://review.haiku-os.org/c/haiku/+/3466
Reviewed-by: Jérôme Duval <jerome.duval@gmail.com>
No read and write support for now. But we implement getting SD card
capacity. SDHC is not supported yet (it uses a different layout for the
CSD register which will be rejected by this version of the code)
Change-Id: Ife844a62f3846c0a780259e9a3a08195e2fd965e
Reviewed-on: https://review.haiku-os.org/c/haiku/+/1068
Reviewed-by: Jérôme Duval <jerome.duval@gmail.com>
C++ don't allow zero size class fields. If field with empty class field
is used, it's size will be 1 byte.
Create DeleteFunc instance as local variable at each use instead.
Fixes#16638.
Change-Id: Ifb76c45ea02e9fed014751542ee5f16f41e11d15
Reviewed-on: https://review.haiku-os.org/c/haiku/+/3458
Reviewed-by: Adrien Destugues <pulkomandy@gmail.com>
The same as CObjectDeleter.
Change-Id: I85c4cb3635f01f13e529ca087324cc2fcb42cfc0
Reviewed-on: https://review.haiku-os.org/c/haiku/+/3456
Reviewed-by: Adrien Destugues <pulkomandy@gmail.com>
It allows to make typedef of pointer types and declaring pointers in headers.
Store of destructor function pointer in CObjectDeleter is no longer needed.
Change-Id: Ic629fd10b28b09f4190edf8ba6b911ca3108ab0e
Reviewed-on: https://review.haiku-os.org/c/haiku/+/3455
Reviewed-by: Adrien Destugues <pulkomandy@gmail.com>
Fix 'item' is used after delete.
Pointed out by Clang Static Analyzer.
Change-Id: I8eca3084c97b37015a2da1b96119a458d4eb9aa7
Reviewed-on: https://review.haiku-os.org/c/haiku/+/3392
Reviewed-by: Adrien Destugues <pulkomandy@gmail.com>
Reviewed-by: John Scipione <jscipione@gmail.com>
Only keep a fixed number of icons in memory at once.
Completes To #15370
Change-Id: I23e3a4fa7559894034f45afb3b536910ea037078
Reviewed-on: https://review.haiku-os.org/c/haiku/+/3367
Reviewed-by: Rene Gollent <rene@gollent.com>
this is queued for issue 8: https://www.austingroupbugs.net/view.php?id=415
this implementation calls the ioctl hook of the filesystem with BSD-like constants
FIOSEEKDATA and FIOSEEKHOLE. if the hook doesn't know the constants, we use the stat size
to return the last hole (as proposed in the draft spec).
Change-Id: I5d052eed87e29b70491a7ff534e244896469d03e
Reviewed-on: https://review.haiku-os.org/c/haiku/+/3385
Reviewed-by: Jérôme Duval <jerome.duval@gmail.com>
the preallocate syscall will call the preallocate filesystem hook, if available.
fix#6285
Change-Id: Ifff4595548610c8e009d4e5ffb64c37e0884e62d
Reviewed-on: https://review.haiku-os.org/c/haiku/+/3382
Reviewed-by: Jérôme Duval <jerome.duval@gmail.com>
According to https://tools.ietf.org/html/rfc3493:
3.2 IPv6 Address Structure
"The structure in6_addr above is usually implemented with an embedded
union with extra fields that force the desired alignment level in a
manner similar to BSD implementations of "struct in_addr"."
Change-Id: Ibe0ee685366398c143cdf9573dcb77566a12888b
Reviewed-on: https://review.haiku-os.org/c/haiku/+/3348
Reviewed-by: Rene Gollent <rene@gollent.com>
Currently used by fixup_next_boot_floppy.
Change-Id: I47c10657b5280f00e470a3171ad11744859ce76c
Reviewed-on: https://review.haiku-os.org/c/haiku/+/3310
Reviewed-by: Adrien Destugues <pulkomandy@gmail.com>
This API change forces all creation of BUrlRequest to be done via
BUrlProtocolRoster::MakeRequest(). This allows the structure of protocol
addons to be altered without breaking ABI for client applications.
Change-Id: I1785c9136c50d19eaa9e57cb9d259ed8d88a5b56
Reviewed-on: https://review.haiku-os.org/c/haiku/+/3080
Reviewed-by: waddlesplash <waddlesplash@gmail.com>
For now it just forwards the command to the SDHCI controller.
The bus will gain more features and functions as work advances (tracking
which card is active, arbitration of DMA transfers, etc).
Change-Id: I094eb84f27e7789387a3f8fb65fba1e5fcfa3e8a
Reviewed-on: https://review.haiku-os.org/c/haiku/+/3094
Reviewed-by: waddlesplash <waddlesplash@gmail.com>
Update BTab::DrawTab() to pass the current index, the index of the
selected tab, and the index of the first and last tabs into
BControlLook::DrawActiveTab() and BControlLook::DrawInactiveTab().
This allows you to draw tabs differently in your BTab or BControlLook
subclass in many different circumstances.
Modify BControlLook API to add indexes to DrawActiveTab() and
DrawInactiveTab() like so:
void DrawActiveTab(..., int32 index = 0, int32 selected = -1,
int32 first = 0, int32 last = 0);
void DrawInactiveTab(..., int32 index = 0, int32 selected = -1,
int32 first = 0, int32 last = 0);
These extra indexes are not used by HaikuControlLook which relies only
on if the tab is active or inactive to draw.
Add IndexOf(BTab* tab) method to BTabView and document it to get the
index of the current tab in BTab::DrawTab(). Also add a warning in the
BTabView::DrawTab() method not to use the position and full parameters
anymore, use BTabView::IndexOf(), BTabView::Selection(), and
BTabView::TabCount() to get the info you need.
Using a dynamic_cast to a BTabView in BeControlLook to determine if the
view is derived from a BTabView didn't work in the case of WebPositive.
Furthermore, WebPositive does custom tab drawing which needed to be
updated for alternative control look. These index parameters passed from
BTab to BeControlLook allow us to draw the tab like BeOS without relying
on a dynamic_cast to BTabView to get the info.
Reproduce the functionality described above for BTab in WebPositive's
custom tabs. Eliminate no longer needed code in favor of using indexes.
Update WebPositive custom tabs to use BControlLook::DrawTabFrame()
instead of BControlLook::DrawInactiveTab() matching the update made in
BTabView.
In BeControlLook::DrawTabFrame() fill rect with base color, WebPositive
doesn't draw any tab background, so it expects this work to be done for
it.
Eliminate hasFrames variable from WebPositive.
Rename TabSelected(index) to UpdateSelection(index) in WebPositive to
better reflect its purpose.
Adjusted HaikuControlLook::DrawInactiveTab() to draw the tab borders more
selectively. Only draw border if left border is set for top and bottom tabs
or top border is set for left and right tabs. Undo no longer needed frame
manipulation border drawing workaround in HaikuControlLook::DrawTabFrame().
Draw scroll bar triangle without using DrawArrowShape().
Unlike in HaikuControlLook, DrawArrowShape() is used to draw arrows in
BOutlineListView and menus distinctly from how it draws arrows in scroll
bars. Draw our distinct arrows in DrawSrollBarButtons() instead.
This fixes overflow of time edit up-down arrows in Clock prefs and the
collapse-expand arrow in Deskbar not being vertically centered.
In DrawBorders() only inset if we actually draw the border.
Fix alignment issues with DrawSliderThumb dots for example in
MediaPlayer volume knobs.
Draw using line arrays calling AddLine instead of StrokeLine in
several places.
DrawMenuBar() extends to draw final pixel which eliminates an extra
lines at the end of menu bars.
Truncate button labels better fixing a few issues for example keymap
keyboard layout button labels. Button insets has been updated a bit
to fix drawing issues with buttons missing a border.
Using a dynamic_cast to a BButton to determine if a view is a button
in BeControlLook didn't work in the case of the keymap label. Look for
B_FLAT, B_HOVER, or B_DEFAULT_BUTTON flag in BeControlLook::DrawLabel()
to draw the label inverted on click. Pass the B_FLAT flag from Keymap
keys when drawing using BControlLook so that the label is inverted.
Change-Id: I07631f4b006bdb9aeca2adc9cbdf2da54dae8e92
Reviewed-on: https://review.haiku-os.org/c/haiku/+/2866
Reviewed-by: Adrien Destugues <pulkomandy@gmail.com>
Just like post-install scripts which run shortly after a package is
installed, pre-uninstall scripts are run just before a package is
removed. Implements enhancement #13427
* Fix script exit code handling vs script launch errors.
* Bump package and repo file version numbers due to new attribute,
unfortunately makes new .hpkg files not backwards compatible.
* Add pre-uninstall functionality, mostly cloning the post-install
except in a few places.
* Discover that _RunQueuedScripts() is never called, a future TODO:?
* Update package documentation for pre-uninstall scripts, and use of
the boot/post-install directory.
Change-Id: I45596255ce74bc102f6e5b606cbf83e4e4347a17
Reviewed-on: https://review.haiku-os.org/c/haiku/+/1504
Reviewed-by: Alexander G. M. Smith <agmsmith@ncf.ca>
Reviewed-by: Adrien Destugues <pulkomandy@gmail.com>
This is a version of the qsort function allowing you to pass a "cookie" to the function.
Change-Id: I60c645213b9c9590e38b112634fcac1d7969b6d9
Reviewed-on: https://review.haiku-os.org/c/haiku/+/2449
Reviewed-by: Adrien Destugues <pulkomandy@gmail.com>
when the cpufreq module is loaded, we let the scheduler update its policy.
Improve assert report
CoreEntry::GetLoad() could return more than kMaxLoad.
Change-Id: I127f9b3e8062b5996872aae30b4021b9904fa179
Reviewed-on: https://review.haiku-os.org/c/haiku/+/3216
Reviewed-by: Jérôme Duval <jerome.duval@gmail.com>
implement on x86 with APERFMPERF.
Change-Id: Ia484854c76dee76c5447983de15800a25d791d39
Reviewed-on: https://review.haiku-os.org/c/haiku/+/3213
Reviewed-by: Jérôme Duval <jerome.duval@gmail.com>
to call a function on the target cpu. Early mechanism not available.
Change-Id: I9d049e618c319c59729d1ab53fb313b748f82315
Reviewed-on: https://review.haiku-os.org/c/haiku/+/3212
Reviewed-by: Axel Dörfler <axeld@pinc-software.de>
This is probably incomplete. Is locking needed? Should we notify the
next writer (if any) that the port is writable when flushing the output?
Change-Id: I2566e2d036a61af4819894a44f57603179aa27df
Reviewed-on: https://review.haiku-os.org/c/haiku/+/2516
Reviewed-by: Adrien Destugues <pulkomandy@gmail.com>
This allows you to pass node_ref's around like you can entry_ref's.
Added node_ref_flatten(), node_ref_unflatten() and node_ref_swap() to
MessageUtils. These are close cousins to entry_ref_flatten(),
entry_ref_unflatten(), and entry_ref_swap() but for node_ref's.
Added B_NODE_REF_TYPE to TypeConstants.h in the Support Kit.
Added B_NODE_REF_TYPE to Debugger and ByteOrder in Support Kit,
B_NODE_REF_TYPE is treated the same as a B_REF_TYPE (entry_ref).
Add documentation for new NodeRef methods and B_NODE_REF_TYPE.
Change-Id: I32c6ed276bf1a7894a835b9fc9de5a882c35883c
Reviewed-on: https://review.haiku-os.org/c/haiku/+/3182
Reviewed-by: Adrien Destugues <pulkomandy@gmail.com>
Recalculate line breaks in FrameResized() if word-wrap is on, otherwise
only move the text rect into position. StyledEdit was recalculating line
breaks before on resize (we have to in this mode) and the frame offset
updates for non-wrapping text views are inexpensive. This makes resizing
text view's work like StyledEdit everywhere.
Scroll to cursor when word-wrap setting changes if text view is editable.
If you are viewing a long document changing word-wrap can move the cursor
quite far, so scroll back to it.
Fix _ActualTabWidth() pen location for right and center-aligned text views
so that tabs widths are calculated correctly.
Reset fTextRect horizontal limits to bounds minus insets in
_RecalculateLineBreaks(), then grow fTextRect based on alignment when
wrap is off.
Fixing insets also fixes right and center-aligned BTextViews.
Left-aligned text view's grow right, right-aligned ones grow left,
and center-aligned ones grow out.
Make extra scrolling space for all aligned text views go the other way
from how it did in hrev24130 (and on left-aligned text view's too) so
that half the text is visible when you edit past the end or before the
beginnning of a text view instead of none of it.
Fixes#1651#12608#13796#15189
Do not _RecalculateLineBreaks() if text view bounds are invalid.
In SetText() detect invalid text view bounds and resize the view to the
width and height of the first line. Then recalculate line breaks.
This fixes BAlert text view size issues.
Fixes#16481 (regression from hrev54496.)
Remove useless and heavy computation. There is no point in computing line
breaks for a 10px wide text view and it takes a long time because it needs
a lot of linebreaks. The view eventually gets laid out properly.
Fixes#5582 (which was not locale-related, after all.)
Only apply default insets if text rect is set to bounds. This ensures
that apps that manipulate the text rect can continue to do so without
the default insets interfering while apps that don't can benefit
from the defaults. If you want to set the text rect to bounds and
not use the default insets you must override the default by calling
SetInsets(). This prevent the default insets from being applied once
apps have changed the text rect fixing a bug in Icon-O-Matic where the
text rect insets were being applied incorrectly.
Fixes#16488 (regression from hrev54496.)
Reduce left and right insets inside text views from full label spacing
to half label spacing. Unify padding between BTextControl and BTextView.
Move fLayoutData->UpdateInsets() to private BTextView::_UpdateInsets()
because we need access to BTextView member variables when deciding
whether or not to add the default padding or not.
_UpdateInsets() changes:
* Don't update insets if BTextView::SetInsets() was called.
* Don't add default insets unless fTextRect is set to view Bounds().
* Do not set the right and bottom insets to left and top if negative,
set them to 0 like we do to left and top -- DeskCalc bug otherwise.
Fixes#15688
Other BTextView fixes:
* Replace max_c and min_c with std::max and std::min respectively.
* Remove scrolling from one instance of BTextView::SetText() as it
produced undesired results while editing a scrolled text view.
* Add default insets in _UpdateInsets()
* Fix scrolling when entering and deleting text so that some part of
the text is always visible. Make visible scroll width depend on font
size.
* Allow scrolling to a negative offset in x but not y. This allows you
to scroll the entire contents of right and centered-aligned text views
whose content does not fit in the box.
* Change _Refresh() to take an offset instead of a bool so that you can
scroll to any offset.
* Replace TextLength() with fText->Length() in a couple of places.
TextControl changes:
* Set text rect in BTextControl::DoLayout().
* Remove AlignTextRect() from TextInput.
Fix the following problems in apps:
ScreenSaver: Set text rect in PreviewView::AddPreview().
Tracker: Set "Edit name" text view insets to 2. Tweek text rect position
to be on top of label in icon, mini-icon, and list mode. Add a TODO that
the text rect is a pixel off from the name on some files.
Mail: Remove _AlignTextRect() and FrameResized() from AddressTextControl.
Use default insets on the text view, defaults are fine here.
DeskCalc: Set insets based on font size in ExpressionTextView
SetTextRect() instead of manipulating the text rect.
Remove _CheckTextRect() and related methods from InputTextView.
Icon-O-Matic: Remove _CheckTextRect() and related methods from InputTextView.
WebPositive: Remove _AlignTextRect() and FrameResized() from URLTextView
and call SetInsets().
StyledEdit: Word-wrap and FrameResized() changes ported to BTextView.
Fixes#16476#16480#16488 (regressions from hrev54496.)
Change-Id: Ifeca6077f8815ccd86d5a3880f99556298aaf0fe
Reviewed-on: https://review.haiku-os.org/c/haiku/+/3152
Reviewed-by: John Scipione <jscipione@gmail.com>
When a thread is created, it is expected that some other thread (usually the
creating thread) will want to make sure it completes. This is done using the
pthread_join() or wait_for_thread() calls.
It is possible that threads end before another thread waits for its completion.
That's why there is a dead thread list for each team, which holds thread ids
and their exit status so that a call to pthread_join() or wait_for_thread() in
the future can complete succesfully.
The dead thread list was limited to 32 threads per team. If there would be
more, the oldest thread would be kicked off. This could cause issues in
situations where a team would create more than 32 threads, and would start
waiting for their result after they have finished. Some of the calls would fail
because the threads would no longer be in the dead list.
This specifically caused problems for cargo (the Rust package manager), which
could depending on the number of dependencies, could create more than 32
threads. See: https://github.com/nielx/rust/issues/3
This change removes the limit of dead threads within a team. Note that there is
a risk that a badly written program that does not detach or joins its threads
can make this an endless list, but the impact is relatively small (dead threads
only occupy a bit of kernel memory).
Change-Id: I0135dd54e10ee48a529f23228d21237d4f1a74e2
Reviewed-on: https://review.haiku-os.org/c/haiku/+/3178
Reviewed-by: Adrien Destugues <pulkomandy@gmail.com>
The catalogs are loaded from separate files, so there is no need to have
an app entry_ref to load them, just a MIME type is enough.
The implementation is a bit simplified: only the default catalog format
is allowed (unlike when loading from entry_ref, where extra catalog
formats can be added in add-ons).
Unrelated cleanup: remove unused code to load catalogs from attributes
of an application. We considered this when designing the locale kit, but
using resources or separate files works better.
Use this in Cortex, where some strings are in a static library, so they
don't have an associated executable or library or add-on to identify
them. The code in Cortex is not complete localization, several parts
should use StringForRate, BStringFormat, etc.
Change-Id: I09be22b1f50891250c4497c51e1db8dcee279140
Reviewed-on: https://review.haiku-os.org/c/haiku/+/3172
Reviewed-by: Kacper Kasper <kacperkasper@gmail.com>
When copying an area with vm_copy_area only the new protection would be
applied and any possibly existing page protections on the source area
were ignored.
For areas with stricter area protection than page protection, this lead
to faults when accessing the copy. In the opposite case it lead to too
relaxed protection. The currently only user of vm_copy_area is
fork_team which goes through all areas of the parent and copies them to
the new team. Hence page protections were ignored on all forked teams.
Remove the protection argument and instead always carry over the source
area protection and duplicate the page protections when present.
Also make sure to take the page protections into account for deciding
whether or not the copy is writable and therefore needs to have copy on
write semantics.
Change-Id: I52f295f2aaa66e31b4900b754343b3be9a19ba30
Reviewed-on: https://review.haiku-os.org/c/haiku/+/3166
Reviewed-by: waddlesplash <waddlesplash@gmail.com>
Change-Id: Id050fad59ede444f2eab7eca681c6ec44612aaf9
Reviewed-on: https://review.haiku-os.org/c/haiku/+/3160
Reviewed-by: Alex von Gluck IV <kallisti5@unixzen.com>
Reviewed-by: François Revol <revol@free.fr>
Somehow the first review merged only the commit log.
FreeBSD doesn't have m68k anyway, so use fenv from musl with as less
changes as possible.
Change-Id: I6372af6679e6773fbb6bf4c8b5b30512971a97a6
Reviewed-on: https://review.haiku-os.org/c/haiku/+/3161
Reviewed-by: Alex von Gluck IV <kallisti5@unixzen.com>
This enables generation of exceptions that are due to uncorrected
hardware errors. The exception handlers were already in place and will
now actually trigger kernel panics.
Note that this is the simplest form of MCE "handling" and does not add
anything of the broader machine check architecture (MCA) that also allow
reporting of corrected errors. As MCEs are generally hard to decode due
to their hardware specifity, this merely makes such problems more
obvious.
Might help to discern hardware issues in cases that would otherwise just
triple fault and cause a reboot.
Change-Id: I9e3a2640458f7c562066478d0ca90e3a46c3a325
Reviewed-on: https://review.haiku-os.org/c/haiku/+/3155
Reviewed-by: waddlesplash <waddlesplash@gmail.com>
Reviewed-by: Axel Dörfler <axeld@pinc-software.de>
The BGradient class is a bit strange as it can store any gradient on its
own, butonly the subclasses allow to set some of the fields.
In the asignment operator, the non-base data (which is in an union) was
not copied over.
More importantly, the missing copy constructor led to the default
implementation being used, and BList (used for the color stops) was
being copied using its default copy constructor, resulting in the two
BGradient (original and copy) poinitng to the same stops data. Heap
corruption resulted whenever one of them was deleted.
Having a working copy ocnstructor fixes this. The alternative is making
the copy constructor private or protected to make sure gradients are not
copied, since normally you'd copy only the subclasses, preserving the
C++ type. However there is nothing enforcing that, and manipulating a
BGradient copied from a subclass works just fine.
Change-Id: I28e733eb8a2970b76ae623eabb75ef8435f508af
Reviewed-on: https://review.haiku-os.org/c/haiku/+/3144
Reviewed-by: waddlesplash <waddlesplash@gmail.com>
Preserve passed in text rect in fTextRext (unless in layout)
and create an internal version fAlignedTextRect which is used
in place of fTextRect. fAlignedTextRext is aligned to fit the
text rect bounds and grows to fit. fAlignedTextRect always grows
vertically but only grows horizontally if wrap is off.
Left-aligned text view's grow right, right-aligned ones grow left,
and center center aligned ones grow out.
Set fTextRect to bounds in _DoLayout().
Reduce left and right padding inside text views from full label
spacing to half label spacing. Unify padding between BTextControl
and BTextView.
Fixing padding also fixes right and center-aligned BTextViews.
Undo extra scrolling for non-left text views from hrev24130 fixing
a scrolling left and right with mouse bug when it shouldn't.
Replace max_c and min_c with std::max and std::min respectively.
Remove scrolling from one instance of BTextView::SetText as it
produced undesired results while editing a scrolled text view.
Set text rect in BTextControl::DoLayout() and ScreenSaver
PreviewView::AddPreview().
Don't add padding if BTextView::SetInsets() is called. Set insets
to 0 in Tracker "Edit name" setting which prevents default padding
from being added. This is so that when you rename a file in Tracker
the TextView appears on top of the file name text with no padding.
80 char limit fixes.
Fixes#1651#12608#13796#15189#15688
Change-Id: I8c6106effc612f49aff374f29742471628b5df86
Reviewed-on: https://review.haiku-os.org/c/haiku/+/3054
Reviewed-by: Adrien Destugues <pulkomandy@gmail.com>
- libicule and libiculx do not exist anymore in newer ICU versions
(harfbuzz replaces them), but we didn't actually use them, so remove
them from the build feature and from the package dependencies
- Add namespace usage marcos since the newer ICU packages put ICU things
in a namespace, making it easier to have multiple versions of ICU used
side by side.
No functional change intended, but this makes it possible to build the
code with either ICU 57 (for gcc2) or 66 (for other architectures).
It allows an application to signal that it no longer needs the data in
the given address range and the underlying pages can be discarded and
reused elsewhere. This is finer grained than working with full areas
or mappings at a time and enables unmapping sections of partially used
mappings without giving up its address space.
Compared with punching holes into a mapping by "mapping over" with
PROT_NONE and MAP_NORESERVE, this has the obvious advantage of not
producing a lot of unused extra areas and saves the corresponding
resources. It is also a lot "lighter" of an operation than cutting
existing areas.
This introduces madvise() alongside the existing posix_madvise() to
allow for OS specific extensions. The constants for both functions are
aliased, the POSIX_MADV_* being a subset of the MADV_* ones without the
non-POSIX extensions. Internally posix_madvise() simply calls madvise().
MADV_FREE is commonly supported in other OSes with various subtle
semantic differences as to when pages are actually freed/cleared and how
or whether the pages are counted against the memory use of a process.
In the variant implemented here, pages are always immediately discarded
and memory counting is not altered. This behaviour should be considered
an implementation detail and may be altered later. The actual unmap and
discard could for example be delayed until pages are needed elsewhere to
reduce overhead in case of repeated discarding and remapping.
Note that MADV_FREE doesn't really align with the rest of the madvise()
API as it works like a command (i.e. discard these pages) and does not
add an attribute to the pages in the given range (i.e. mark these pages
for quick access from now on). As such, an MADV_FREE does not need to be
undone by setting a different advice later on, unlike how the other
flags work. This discrepancy may be the reason why it is not part of
POSIX.
Change-Id: Icc093379125a43e465dc4409d8f5ae0f64e107e0
Reviewed-on: https://review.haiku-os.org/c/haiku/+/2844
Reviewed-by: waddlesplash <waddlesplash@gmail.com>
Pages in the given range are unmapped and freed without getting written
back anywhere. It can be used whenever a caller does not care about the
data in the given range anymore and wants to reduce page pressure.
Change-Id: I8bcce68fab278efef710d3714677e1d463504a56
Reviewed-on: https://review.haiku-os.org/c/haiku/+/2843
Reviewed-by: waddlesplash <waddlesplash@gmail.com>
The limit was set when creating limits.h back in hrev88. It seems to be
used only in glob(). The limit is quite low by today standards and
prevents building icu 67.
I guess the liit was put at 32K because that's what BeOS did, but I see
no reason to keep it that way.
Change-Id: I74f95d9b56891dd90c79b7ced35ca8d1ec81d6ab
Reviewed-on: https://review.haiku-os.org/c/haiku/+/3117
Reviewed-by: waddlesplash <waddlesplash@gmail.com>
Change-Id: Ifbd82ef7bfc2c39b2aeb5c25be177421cd22d246
Reviewed-on: https://review.haiku-os.org/c/haiku/+/2920
Reviewed-by: Alex von Gluck IV <kallisti5@unixzen.com>
Reviewed-by: Adrien Destugues <pulkomandy@gmail.com>
This replaces the use of the hard-coded scroll bar size constants
and instead asks the scrollbars for their preferred sizes directly.
Right now, this is a giant no-op since BScrollBar just returns
the same hard-coded size when asked. The next commit will, however,
change that.
It was always enabled, and disabling it would break ABI.
If we want to make it disable-able, it needs to be a setting,
but I don't really see a reason for that.
* Migrate some platform agnostic architecture code into
boot/arch from efi/arch. This helps to avoid conflicts
between kernel and boot sources as well.
* Conflicts between arch_cpu in efi and kernel code means
bootcode really should *never* directly use kernel arch
headers. (other platforms don't, which is why they don't
have this same issue)
* We carefully thread any needed kernel headers (namely
assembly helper macros) into the bootloader headers without
mixing in the whole conflicting kernel/arch headers.
* ARM now properly get its cpu init code called, and we
progress further into the EFI bootloader.
Change-Id: If67ec9758b5ce68563ebd9eb45d5196401911c67
Reviewed-on: https://review.haiku-os.org/c/haiku/+/2975
Reviewed-by: waddlesplash <waddlesplash@gmail.com>
None of these were used; they were all imported with the original
root Haiku commit, and they are totally unrelated to PulkoMandy's
new SPARC work. Plus, they were also under a BSD Advertising Clause
license.
The BSD implementation was under the Advertising Clause,
so we might as well take the opportunity to replace the
implementation entirely with musl's.
Header also rewritten to be a Haiku one; the constants
are left unchanged of course.
Taken from FreeBSD; some minor cleanup elsewhere.
udp.h rewritten entirely as it contained no copyrightable
material and bears little resemblance to BSD's.
* Remove functions not even FreeBSD defines.
* Remove dependency on unnecessary headers.
* Update copyright headers to match FreeBSD's; includes
removal of the advertising clause.
* Move some private structs to netresolv port_before.
Unfortunately, it seems musl implements very little of this file,
so we may wind up sticking with netresolv for its implementation.
* Remove all functions and a number of constants that neither
glibc nor musl define or support (and even FreeBSD does not
declare a good number of these anymore.)
* Redeclare the primary flags in terms of (1 << X) instead
of raw 0x... for readability (the constants at the end
do NOT match up to their definitions in glibc, musl, and BSDs!)
* Remove usage of unneeded headers, and __BEGIN/END_DECLS.
* Replace non-Haiku license headers with the ones from FreeBSD,
which notably contain a removal of the advertising clause.
ABIs remain unchanged, but a small set of applications that
use these esoteric APIs may not compile anymore (are there
any remaining?)
These are not in the standard and are not declared by glibc at all.
The symbols remain for any applications that are still using them,
for now.\
Change-Id: Ie6b4a6b5ec3231c304e05ce9cb38c67d9ee51ad7
Reviewed-on: https://review.haiku-os.org/c/haiku/+/2942
Reviewed-by: waddlesplash <waddlesplash@gmail.com>
Utilize user_memcpy and IS_USER_ADDRESS when necessary to prevent SMAP violations.
Also add a "wacom_device_header" struct to more easily share data between the wacom
kernel driver and input_server addon.
Should fix#14589
Change-Id: Ie2784020b21523f82fd450a2db2de60ccf9d6620
Reviewed-on: https://review.haiku-os.org/c/haiku/+/2783
Reviewed-by: Adrien Destugues <pulkomandy@gmail.com>
* This allows file systems to retrieve the actual error code on a
failure, and report it to the user.
* All affected file systems have been adjusted to the API change.
This is a binary incompatible change.
Change-Id: Id73392aaf9c6cb7d643ff9adcb8bf80f3037874c
Reviewed-on: https://review.haiku-os.org/c/haiku/+/2913
Reviewed-by: Axel Dörfler <axeld@pinc-software.de>
Reviewed-by: Adrien Destugues <pulkomandy@gmail.com>
xsave or xsavec are supported.
breaks vregs compatibility.
change the thread structure object cache alignment to 64
the xsave fpu_state size isn't defined, it is for instance 832 here, thus I picked 1024.
Change-Id: I4a0cab0bc42c1d37f24dcafb8259f8ff24a330d2
Reviewed-on: https://review.haiku-os.org/c/haiku/+/2849
Reviewed-by: Adrien Destugues <pulkomandy@gmail.com>
Otherwise it clashes with the implementation in OpenSSL which uses the
same names but now has a different ABI.
Change-Id: I5cb3ff97d7b28de978cdcbd8a06f25f65fb53784
Reviewed-on: https://review.haiku-os.org/c/haiku/+/2854
Reviewed-by: Kyle Ambroff-Kao <kyle@ambroffkao.com>
Reviewed-by: Jérôme Duval <jerome.duval@gmail.com>
The code to parse the resource table reads one entry at a time because
the table size isn't known. This resulted in a lot of read syscalls,
each reading just 12 bytes. Use a BBufferIO to buffer these and reduce
the number of syscalls. This helps especially when there are lot of
resources, for example in libbe with all the country flags.
It also removes some spam from strace output for all these read calls.
Change-Id: Ib165a0eacc2bc5f3d319c22c2fac4f439efbdef2
Reviewed-on: https://review.haiku-os.org/c/haiku/+/2858
Reviewed-by: Rene Gollent <rene@gollent.com>
The goal here is to avoid potentially expensive fork()ing.
The time for a fork() is (for a process with no real heap usage
and thus few areas) 300-400us on my system. load_image() takes
3000us (3ms) or so, but this of course includes exec() time.
Overall, for compiling HaikuDepot (with a tweaked jam to use
posix_spawn on Haiku, not just on Linux) there is a slight
decrease in time:
before:
real 1m21.727s
user 1m2.131s
sys 0m43.029s
after:
real 1m19.472s
user 1m1.752s
sys 0m41.740s
Which is probably within the realm of "noise", so more benchmarks
are needed. Likely if we tweak our jam usage to not need as many
shells when running commands, this would be a much more noticeable
change.
Change-Id: I217f2476b1ed9aa18322b3c2bc8986571d89549a
It iterates over all areas intersecting a given address range and
removes the need for manually skipping uninteresting initial areas. It
uses VMAddressSpace::FindClosestArea() to efficiently find the starting
area.
This speeds up the two iterations in unmap_address_range and one in
wait_if_address_range_is_wired and resolves a TODO in the latter hinting
at such a solution.
Change-Id: Iba1d39942db4e4b27e17706be194496f9d4279ed
Reviewed-on: https://review.haiku-os.org/c/haiku/+/2841
Reviewed-by: waddlesplash <waddlesplash@gmail.com>
This introduces VMAddressSpace::FindClosestArea() that can be used to
find the closest area to a given address in either direction. This is
now trivial and efficient since both kernel and user address spaces use
a binary search tree.
Using FindClosestArea() getting multiple area infos is sped up
dramatically as it removes the need for a linear search from the first
area to the one given in the cookie on each successive invocation.
Change-Id: I227da87d915f6f3d3ef88bfeb6be5d4c97c3baaa
Reviewed-on: https://review.haiku-os.org/c/haiku/+/2840
Reviewed-by: waddlesplash <waddlesplash@gmail.com>
They return the left and right most nodes of the entire tree, i.e.
starting from the root node.
Change-Id: I651a9db6d12308aef4c2ed71484958428e58c9bc
Reviewed-on: https://review.haiku-os.org/c/haiku/+/2838
Reviewed-by: waddlesplash <waddlesplash@gmail.com>
This reverts parts of hrev52546 that removed the B_KERNEL_AREA
protection flag and replaced it with an address space comparison.
Checking for areas in the kernel address space inside a user address
space does not work, as areas can only ever belong to one address space.
This rendered these checks ineffective and allowed to unmap, delete or
resize kernel managed areas from their respective userland teams.
That protection was meant to be applied to the team user data area which
was introduced to reduce the kernel to userland overhead by directly
sharing some data between the two. It was intended to be set up in such
a manner that this is safe on the kernel side and the B_KERNEL_AREA flag
was introduced specifically for this purpose.
Incidentally the actual application of the B_KERNEL_AREA flag on the
team user data area was apparently forgotten in the original commit.
The absence of that protection allowed applications to induce KDLs by
modifying the user area and generating a signal for example.
This change restores the B_KERNEL_AREA flag and also applies it to the
team user data area.
Change-Id: I993bb1cf7c6ae10085100db7df7cc23fe66f4edd
Reviewed-on: https://review.haiku-os.org/c/haiku/+/2836
Reviewed-by: waddlesplash <waddlesplash@gmail.com>
The code in the Resize and Rebase methods was identical except for the
iterator.
Change-Id: I9f6b3c2c09af0c26778215bd627fed030c4d46f1
Reviewed-on: https://review.haiku-os.org/c/haiku/+/2835
Reviewed-by: waddlesplash <waddlesplash@gmail.com>
This allows switching from another recursive_lock, mutex or read-locked
rw_lock analogous to the switching possibilities already in mutex.
With this, recursive_locks can be used in more complex situations where
previously only mutexes would work.
Also add debugger command to dump a recursive_lock.
Change-Id: Ibeeae1b42c543d925dec61a3b257e1f3df7f8934
Reviewed-on: https://review.haiku-os.org/c/haiku/+/2834
Reviewed-by: waddlesplash <waddlesplash@gmail.com>
The memory map may be unordered and include overlapping ranges. To make
sure that nothing gets included as usable that should actually be
excluded, first scan for all usable ranges and add them, then remove
anything unusable from these ranges again.
To calculate the amount of unusable memory, count the total after the
first pass and then subtract the total after the second. This way, only
unusable ranges that actually overlap physical memory (and therefore
reduce the amount of usable memory) get excluded.
Note that the explicit ignore of the ACPI reclaim memory is subsumed by
the above. We still don't want to add this region to the usable memory
map, as that would allow the kernel to allocate pages into that region,
possibly corrupting ACPI tables before they were used. We also don't
want to add it as an allocated range, as it is not guaranteed that ACPI
is done with the tables before the unused bootloader ranges are freed in
the kernel.
Also add the missing unusable memory amount from ignoring the first MiB
of memory in the EFI loader.
May fix#16056 although it is not certain that graphics memory ranges
are actually included in the memory map.
Change-Id: Ie7991d2c4dcd988edac2995b3a7efc509fa0f4a3
Reviewed-on: https://review.haiku-os.org/c/haiku/+/2814
Reviewed-by: waddlesplash <waddlesplash@gmail.com>
APE reader was using a GPL licensed version of MD5. A similar
implementation in the public domain was available in libnetapi, which I
moved to libshared so the APE reader can use it (and made some fixes,
missing const mainly). It only needs a small wrapper to use it easily
from C++ in a way compatible with the previous implementation.
Part of #13814.
I forgot to change MUTEX_INITIALIZER following removal of the
unused field.
Change-Id: I011c023ae00bb4576c8bcecf83546892fef3a77e
Reviewed-on: https://review.haiku-os.org/c/haiku/+/2719
Reviewed-by: waddlesplash <waddlesplash@gmail.com>
As far as I can tell, there is no reason to ignore unlocks, ever;
if no threads are waiting, then mutex_unlock() will act appropriately.
So all we need to do is increment the lock's count here,
as we are relinquishing our request for locking.
On the other hand, if we did not find our structure in the lock,
that means we own the lock; so to return with an error from here
without changing the count would result in a deadlock, as the lock
would then be ours, despite our error code implying otherwise.
Additionally, take care of part of the case where we have woken up
by mutex_destroy(), by setting thread to NULL and checking for it
in that case. There is still a race here, however.
May fix#16044, as it appears there is a case where ACPICA
calls this with a timeout of 0 (we should make this be
a mutex_trylock, anyway.)
Change-Id: I98215df218514c70ac1922bc3a6f10e01087e44b
Reviewed-on: https://review.haiku-os.org/c/haiku/+/2716
Reviewed-by: waddlesplash <waddlesplash@gmail.com>
This is required for BMenuBar based toolbars.
Fixes#15785.
Change-Id: I8d108694b481e408e5c56e23a697c8e7829343dc
Reviewed-on: https://review.haiku-os.org/c/haiku/+/2316
Reviewed-by: Adrien Destugues <pulkomandy@gmail.com>
It is used by several of the filesystems, so it seems a good idea to
move it to the shared/ directory.
UFS2, BFS, XFS, EXT2 and EXFAT are adjusted.
Change-Id: I493e37a1e7d3ae24251469f82befd985a3c1dbdd
Reviewed-on: https://review.haiku-os.org/c/haiku/+/2489
Reviewed-by: Adrien Destugues <pulkomandy@gmail.com>
For historical reasons, the package kit has an "url" field that is not
actually meant to be used as an URL. Rename it in the API and user
facing output as "identifier" to make it clear what the file is used
for. This change preserves the "url" key in on-disk and online storage
(hpkr files, stored settings, etc) in an attempt to not break anything.
Fix one remaining misuse of the "url" field as an URL in
get_package_dependencies.
Add an unit test showing that BUrl does parse "tab" URIs properly (there
is just a protocol and a path segment).
Change-Id: I339ce526e5798d42d78ae650855d7e988dbb4a1a
Reviewed-on: https://review.haiku-os.org/c/haiku/+/2542
Reviewed-by: waddlesplash <waddlesplash@gmail.com>
We attempted to make it a 32bit char type, but that is actually not allowed
unless we also make uint_least16_t a 32bit type. And even then, we
wouldn't be allowed to store or handle values wider than 16bit.
Comply more closely to the standard. As a result, mbtoc16r is not
implemented. c16rtomb is implemented by casting the char to 32bit, which
isn't really correct either (I think you're supposed to be able to feed
the two halves of a > 16bit codepoint in two separate calls and get a
meaningful result out?)
Related to #15990 but we may want an actual implementation?
Change-Id: If8198675c27dd2aa412bc44d12d3df4e31d3e8c7
Reviewed-on: https://review.haiku-os.org/c/haiku/+/2623
Reviewed-by: waddlesplash <waddlesplash@gmail.com>
This reverts hrev54120 and instead adds the commonly supported
MAP_NORESERVE flag to request overcommit.
Using PROT_NONE for overcommit is problematic as the protection of
individual pages can still be changed via mprotect to make them
accessible, but that won't change the commitment. An application
using such a pattern may then unexpectedly run into out of memory
conditions on random writes into the address space.
With MAP_NORESERVE the overcommit can explicitly be requested by
applications that want to reserve address space without producing
memory pressure.
Change-Id: Id213d2245c5e23103e8e0857f7902e0cd8a2c65d
Reviewed-on: https://review.haiku-os.org/c/haiku/+/2611
Reviewed-by: waddlesplash <waddlesplash@gmail.com>
Reviewed-by: Jérôme Duval <jerome.duval@gmail.com>
Rename MovePageRange to Adopt and group it with Resize/Rebase as it
covers the third, middle cut case.
Implement VMAnonymousCache::Adopt() to actually adopt swap pages. This
has to recreate swap blocks instead of taking them over from the source
cache as the cut offset or base offset between the caches may not be
swap block aligned. This means that adoption may fail due to memory
shortage in allocating the swap blocks.
For the middle cut case it is therefore now possible to have the adopt
fail in which case the previous cache restore logic is applied. Since
the readoption of the pages from the second cache can fail for the same
reason, there is a slight chance that we can't restore and lose pages.
For now, just panic in such a case and add a TODO to free memory and
retry.
Change-Id: I9a661f00c8f03bbbea2fe6dee90371c68d7951e6
Reviewed-on: https://review.haiku-os.org/c/haiku/+/2588
Reviewed-by: waddlesplash <waddlesplash@gmail.com>
* Adds VMCache::MovePageRange() and VMCache::Rebase() to facilitate
this.
Applied on top of hrev45098 and rebased with the hrev45564 page_num_t to
off_t change included.
Change-Id: Ie61bf43696783e3376fb4144ddced3781aa092ba
Reviewed-on: https://review.haiku-os.org/c/haiku/+/2581
Reviewed-by: waddlesplash <waddlesplash@gmail.com>
- Use gcc builtin
- Define as a static inline function in the .h so no function call overhead is needed
- Keep the function in libroot for backwards compatibility
- Remove a duplicate implementation in the freebsd compatibility layer
gcc2 does not document the builtin, but it is in fact already available
there as well.
Fixes#3281.
Change-Id: I94f8a2548637aa70e85febbfab06f07c1a427005
Reviewed-on: https://review.haiku-os.org/c/haiku/+/2605
Reviewed-by: waddlesplash <waddlesplash@gmail.com>
This code comes from an old Be Newsletter and since then the API
received the addition of SetMouseEventMask. In several places the
MouseDownThread was misused: it would spawn a new thread on every mouse
click and not clear the previous one. This could for example lead to
BSpinner skipping values if you clicked it at the right speed.
There are functional changes in BSpinner, before it updated for the
first time 100ms after mouse down, and then as you moved the mouse
around the button, now it activates immediately on first click and then
every 200ms (which may be a bit short). In other places, no functional
changes intended.
Change-Id: Ie600dc68cbb87d1e237633953e5189918bf36575
Reviewed-on: https://review.haiku-os.org/c/haiku/+/2599
Reviewed-by: waddlesplash <waddlesplash@gmail.com>
Also remove the MINGW support, as it was far too incomplete.
This *should* work under case-sensitive NTFS, but instead,
it seems #14963 occurs. So perhaps there is a GCC bug
related to case-sensitive vs. case-insensitivity after all.
It was not specified as such before C11, but that's only because there
was no C standard way to do it until then.
Fixes#15955.
Change-Id: Ied7b7fd94988ed7724460917aebc859b74eaa585
Reviewed-on: https://review.haiku-os.org/c/haiku/+/2558
Reviewed-by: waddlesplash <waddlesplash@gmail.com>
add a WMI Asus driver, to control keyboard backlight brightness.
Change-Id: Ib86f70b4a407178b0a1f532269387a55915cc460
Reviewed-on: https://review.haiku-os.org/c/haiku/+/2485
Reviewed-by: Fredrik Holmqvist <fredrik.holmqvist@gmail.com>
Reviewed-by: Adrien Destugues <pulkomandy@gmail.com>
This allows to use FileDescriptorCloser as unique pointer for file descriptor.
Change-Id: I4c768fafba6ed35658b2fdb075b9b547f53bc8da
Reviewed-on: https://review.haiku-os.org/c/haiku/+/2495
Reviewed-by: Jérôme Duval <jerome.duval@gmail.com>
The code in input_server was pretty much all set for this, but there was
no way to configure the extra buttons. Add them to the mouse view in
Input preferences (up to 5 buttons are handled now)
Define a new B_MOUSE_BUTTON(n) macro to generate the bitmask for a given
button (numbered from 1).
Change-Id: I9091082277937d89b08464ff474e7bbb5db82401
Reviewed-on: https://review.haiku-os.org/c/haiku/+/180
Reviewed-by: waddlesplash <waddlesplash@gmail.com>
BeOS didn't support transparent views. As documented in the Be Book,
SetViewColor(B_TRANSPARENT_COLOR) only effect is to not fill the
invalidated areas with the view color before calling Draw() (it avoids
flickering, especially when combined with B_FULL_UPDATE_ON_RESIZE).
A previous change made B_TRANSPARENT_COLOR actually make the view
transparent (that is, additionally to the above, the underlying view is
drawn before the transparent children), but it creates compatibility
issues.
In order to keep the API compatible with BeOS, the new behavior is now
enabled explicitly using the B_TRANSPARENT_VIEW flag. This also opens
for future developments like allowing a view color with an alpha
channel (not supported yet).
Adjust programs that require transparent views.
Fixes#15744, #15745.
Helps with #15645.
Change-Id: I529574ea23db0a23579521b263bc8d572775e35a
Reviewed-on: https://review.haiku-os.org/c/haiku/+/2275
Reviewed-by: Adrien Destugues <pulkomandy@gmail.com>
Scroll bars should look and work identically to before on
HaikuControlLook.
Add DrawScrollBarButton() and DrawScrollBarThumb() and
DrawScrollBarBorder() methods. These methods are used to draw scroll
bars in a generic way so that they can be drawn differently by alternative
control look's (e.g. BeControlLook). Also it gives us back drawing of
scroll bar knobs. However the knob setting is not exposed in the
interface in this commit.
These methods are in addition to the 2 existing DrawScrollBarBackground()
methods that draw the scroll bar background. One draws the area above and
below the thumb and the other is called by the first to actually draw the
area.
The rest of the drawing besides the backgrounds was being done in
BScrollBar before. To draw the scroll bar arrows and thumb we were recyling
other ControlLook methods, while this worked well enough on HaikuControlLook
it wasn't flexible enough for alternative control looks.
DrawScrollBarButton() is used to draw the four scroll buttons and is
typically (so far) used in combination with DrawArrowShape().
DrawScrollBarThumb() draws the scroll bar thumb.
DrawScrollBarBorder() draws a 1px border around the entire scroll bar,
potentially B_KEYBOARD_NAVIGATION_COLOR if focused (although this is
feature not currently used.)
Draw unscrollable scroll bars as if they were disabled including the
buttons with their arrow shapes, background, and thumb.
Add FBC backwords compatibility macros in ControlLook.cpp
Change-Id: I9237c5ce45d17d674785111d51de951e5686306b
Reviewed-on: https://review.haiku-os.org/c/haiku/+/351
Reviewed-by: Adrien Destugues <pulkomandy@gmail.com>
type is acpi_handle, to be used with the ACPI bus manager
Change-Id: Ibbdd81a21bdd57fc651f7a7238e3676033204857
Reviewed-on: https://review.haiku-os.org/c/haiku/+/2456
Reviewed-by: Rene Gollent <rene@gollent.com>
ACPI_DEVICE_HID_ITEM is now optional, instead of an empty string.
Change-Id: I352ffaaad377659f650a0b8c0d56e40a68b739c3
Reviewed-on: https://review.haiku-os.org/c/haiku/+/2420
Reviewed-by: Jérôme Duval <jerome.duval@gmail.com>
to find a node in the hierarchy.
Change-Id: Iee858f21ce134569bf25fccbef9fe18ea8787e9d
Reviewed-on: https://review.haiku-os.org/c/haiku/+/2419
Reviewed-by: Jérôme Duval <jerome.duval@gmail.com>
That's a bad idea. The drag and drop tracking could get confused enough
to crash app_server (possibly because of calling InitiateDrag multiple
times?)
Fixes#14983.
Change-Id: Id43bfcbfc24b1adb8f6e9fff587c6df9b62910f2
Reviewed-on: https://review.haiku-os.org/c/haiku/+/2413
Reviewed-by: Alex von Gluck IV <kallisti5@unixzen.com>
Reviewed-by: waddlesplash <waddlesplash@gmail.com>
Previous version of the patch was broken by the EFI refactoring.
Change-Id: I6dd125100b22b2461c531bfd8f81b3dd28e2b751
Reviewed-on: https://review.haiku-os.org/c/haiku/+/2409
Reviewed-by: Jérôme Duval <jerome.duval@gmail.com>
