First try to allocate a page below 2GB.
If it fails, try to allocate a physical page at the address
returned by get_next_virtual_address so that we get an address
that is available in both virtual and physical space.
Finally try to allocate any available page. At this point
it's highly likely that it will be above 0x88000000 so it can be
used for trampoline code.
Change-Id: I3100a3e09efcd938bc630a52dcce74dc5f24beb5
Reviewed-on: https://review.haiku-os.org/c/haiku/+/4878
Tested-by: Commit checker robot <no-reply+buildbot@haiku-os.org>
Reviewed-by: Adrien Destugues <pulkomandy@gmail.com>
Change-Id: I899f288f451dd4c413403cd0e14a855f68355050
Reviewed-on: https://review.haiku-os.org/c/haiku/+/4866
Tested-by: Commit checker robot <no-reply+buildbot@haiku-os.org>
Reviewed-by: Fredrik Holmqvist <fredrik.holmqvist@gmail.com>
Reviewed-by: Alex von Gluck IV <kallisti5@unixzen.com>
All the regions allocated via platform_allocate_region() are mapped to a
high virtual address so there's no need to do identity mapping for them
any more.
Copy kernel_args to a region allocated via platform_allocate_region() so it
will be mapped similarly.
Identity mapping is still needed for trampoline code for jumping to the kernel
Change-Id: I844a7a789b440a38521db49adc077bb77e658ddf
Reviewed-on: https://review.haiku-os.org/c/haiku/+/4865
Tested-by: Commit checker robot <no-reply+buildbot@haiku-os.org>
Reviewed-by: Fredrik Holmqvist <fredrik.holmqvist@gmail.com>
Reviewed-by: Alex von Gluck IV <kallisti5@unixzen.com>
A few invalid ones were introduced in 2071dc2
Change-Id: Icbe5bb4aafd9d6e03ec4175c10eccdb2e65465d7
Reviewed-on: https://review.haiku-os.org/c/haiku/+/4874
Reviewed-by: Alex von Gluck IV <kallisti5@unixzen.com>
Tested-by: Commit checker robot <no-reply+buildbot@haiku-os.org>
EFI runtime service memory areas are remapped to a high virtual address
allocated with get_next_virtual_address() to avoid any address space
collision.
They are also added to virtual_ranges_to_keep so that the kernel knows
that these ranges should be reserved.
Change-Id: I63e1a2450fceff3b2404154907a2cf59dd1b373b
Reviewed-on: https://review.haiku-os.org/c/haiku/+/4863
Tested-by: Commit checker robot <no-reply+buildbot@haiku-os.org>
Reviewed-by: Fredrik Holmqvist <fredrik.holmqvist@gmail.com>
Remove the dependency on identity mapping for the initial page tables
used in MapEarly / _EarlyQuery.
The initial page tables have to satisfy the following conditions instead:
* the inital page tables shall be allocated in the same region as the initial page directory
* this region shall be mapped contiguously
* the page tables shall be mapped at a higher address than the page directory
Change-Id: I34854ef93826aa48f5d90582b52905e553f3d275
Reviewed-on: https://review.haiku-os.org/c/haiku/+/4857
Tested-by: Commit checker robot <no-reply+buildbot@haiku-os.org>
Reviewed-by: Adrien Destugues <pulkomandy@gmail.com>
UART is mapped to kernel virtual address range since commit f5f10a9ec
so it does not overlap with user address range any more.
Change-Id: If8e18904fcddcbef1ccdc91e526d0e581f61e085
Reviewed-on: https://review.haiku-os.org/c/haiku/+/4856
Tested-by: Commit checker robot <no-reply+buildbot@haiku-os.org>
Reviewed-by: Adrien Destugues <pulkomandy@gmail.com>
vendor_id shall come after the bitfields
Move hpet_address to separate struct definition so we can apply
the correct packed flags.
see also: https://wiki.osdev.org/HPET
Change-Id: Iced005846fedd4b895910e9b61137d5349db5b41
Reviewed-on: https://review.haiku-os.org/c/haiku/+/4859
Reviewed-by: Jérôme Duval <jerome.duval@gmail.com>
Tested-by: Commit checker robot <no-reply+buildbot@haiku-os.org>
* We decided to not create our own custom define checks
and go with the standard ones for riscv64.
* clang doesn't define __RISCV64__ either, so this moves
the clang build of riscv64 a bit further
Change-Id: I2a6c3751168a898c1617b32f46055a9ba1609e2b
Reviewed-on: https://review.haiku-os.org/c/haiku/+/4861
Reviewed-by: Alex von Gluck IV <kallisti5@unixzen.com>
Tested-by: Commit checker robot <no-reply+buildbot@haiku-os.org>
This make it look much better on haiku with dark theme.
Looks the same on default because shine is set to white/very bright.
Change-Id: Iecab3477e1fd5a90c447bcc15a1c598450b074a9
Reviewed-on: https://review.haiku-os.org/c/haiku/+/4706
Reviewed-by: Adrien Destugues <pulkomandy@gmail.com>
Tested-by: Commit checker robot <no-reply+buildbot@haiku-os.org>
* The "size" parameter is the size of "out" not "in", and the
return size_t parameter is supposed to always have the total amount
of wchar_ts needed, not how many are actually used.
* In the case where "outSize == 0", we set "requiredSize" and then
return.
Fixes crashes seen in glib2 Unicode collation routines, which
are used in GTK file dialogs.
Thanks to PulkoMandy for glancing at this.
The scrollbar on some areas is misaligned
with the window frame. This change will fix
this. The class that implements the scroll-
view is also shaded by two class definitions.
This is also fixed by breaking the definition
into its own file.
Fixes#16169
Change-Id: I71372906e165672c9b697ab98181ecfb1b09fb43
Reviewed-on: https://review.haiku-os.org/c/haiku/+/4849
Tested-by: Commit checker robot <no-reply+buildbot@haiku-os.org>
Reviewed-by: Adrien Destugues <pulkomandy@gmail.com>
The old versions used pixel-count dimensions, not pixel-span dimensions
as the entire rest of the API does, which was probably a mistake
in the first place. The new APIs using BSize always use pixel-span
dimensions.
The old versions will shortly be deprecated.
Discovered while working on Xlibe.
Change-Id: I604a5ac6e0588420ff0b667d9193d60ac27b92c6
The USB < 3.0 init code has this, and the XHCI spec indicates that
software is responsible for dealing with this still, indeed.
Noticed by rudolfc when testing some USB audio devices in #1045.
The current style probably has the correct style flags portion. Getting
a new style in that case is unnecessary, and problematic with some font
families that have several fonts mapping to the same style flag
(SemiBold, Bold and ExtraBold are all B_BOLD_FACE, for example): the
chosen one might be different than the one already set.
Fixes#17229
Change-Id: Ie8e73627cc0547df9d624a8fac07f521fc983d06
Reviewed-on: https://review.haiku-os.org/c/haiku/+/4416
Tested-by: Commit checker robot <no-reply+buildbot@haiku-os.org>
Reviewed-by: Adrien Destugues <pulkomandy@gmail.com>
-detect if model > HD600
-apply change by calling ATOM_TRANSMITTER_ACTION_BL_BRIGHTNESS_CONTROL and ATOM_TRANSMITTER_ACTION_LCD_BLON
Author : Anarchos <sylvain_kerjean@hotmail.com>
Change-Id: I30b9dd66c71e05427b827fc3ef2c7fa6c498646f
Reviewed-on: https://review.haiku-os.org/c/haiku/+/2786
Tested-by: Commit checker robot <no-reply+buildbot@haiku-os.org>
Reviewed-by: Alex von Gluck IV <kallisti5@unixzen.com>
Reviewed-by: Adrien Destugues <pulkomandy@gmail.com>
By updating $fdt_addr_r we can simplify boot-arm.scr.txt.
Now we can explicitly pass the device tree address to bootefi instead of
relying on a fallback inside U-Boot.
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Change-Id: Ifceb574aca2814605a2e4e58e9187773817c5df2
Reviewed-on: https://review.haiku-os.org/c/haiku/+/2970
Reviewed-by: Adrien Destugues <pulkomandy@gmail.com>
Reviewed-by: Alex von Gluck IV <kallisti5@unixzen.com>
Tested-by: Commit checker robot <no-reply+buildbot@haiku-os.org>
Change-Id: Iff9e4198aca706097889faf51e9559fe551126ad
Reviewed-on: https://review.haiku-os.org/c/haiku/+/4782
Tested-by: Commit checker robot <no-reply+buildbot@haiku-os.org>
Reviewed-by: Alex von Gluck IV <kallisti5@unixzen.com>
check that the correct BIOS is found with the device and PCI ids.
make the allocated rom area bigger like Linux, otherwise page faults occur.
Change-Id: Iae6ff1c8dfd81ea16c20e73d511f9a0a96f9b071
Reviewed-on: https://review.haiku-os.org/c/haiku/+/4844
Reviewed-by: Alex von Gluck IV <kallisti5@unixzen.com>
