As the comment implies, we don't want to wait here, but low_resource()
waits forever when timeout=0. It doesn't seem to do anything with the
"size" argument at all, so not calling it is just as effective, and so
replace the comment there altogether.
The case where this was most often hit, causing a full system deadlock
in anything relating to memory management, was when the passed "size"
was 0. So now we check for this case explicity at the beginning and
panic() on KDEBUG kernels for it.
This is the cause of #14557. It seems that OpenBSD does the same thing.
I didn't read the spec yet to determine precisely what this means,
so for now, just do what OpenBSD does and bail.
This only seems to happen with a secondary XHCI controller on that
device, so hopefully most USB ports will still work.
The previous version did not account for the position not being a
multiple of the block size, among other problems.
Tested by hand with DiskProbe and then "dd skip=..." to read single
bytes from partial blocks, and validated as correct.
Only one qpair is used for reading, which is rather inefficient.
We currently allocate a bounce buffer for every allocation, which is
also inefficient, due to the fact that we must read an integer multiple
of LBAs.
But it does work, and it is actually reasonably fast, even on an emulated
machine using a spinning-disk-backed NVMe device (88MB/s.)
I wasn't able to get it working in non-packaged, though; the device manager
called supports_device() on a number of PCI devices, but not the NVMe
device, so I have a different version with a hack that grabs the PCI info
manually. I didn't test inside haiku.hpkg yet; perhaps it will work in there.
Previously it was possible that we put() them twice, leading to
assert-failure panics as their ref-counts would go below 0.
Now we get() them once for every bus object that we create, as
the destructors put() them.
Should fix the panic in #15004.
... based on mouse type setting in Mouse Preferences.
for (default) 3 button mouse, nothing changes. #important
for 1 button mouse,
emulate right click by ctrl+click. This replicates Tracker
behavior everwhere
for 2 button mouse,
add todo to emulate middle click by pushing both buttons at
the same time. Unfortunately I was unsucessful getting this
feature to work.
Change-Id: I5377ed5b4967483096f7f185f434bced3f86cdb9
Reviewed-on: https://review.haiku-os.org/c/303
Reviewed-by: Adrien Destugues <pulkomandy@gmail.com>
It is called before them (and is responsible for calling them),
so having it in the middle of the file does not make a lot of sense.
Already done for XHCI. Only one functional change -- removing the
set_dprintf_enabled call. Drivers probably shouldn't spuriously
re-enable that if it was specifically disabled.
Some time back the menu item 'Partition | Initilize' changed to 'Format'.
The format dialog's button label still says 'Initialize'.
Change to 'Format' for consistency.
Change-Id: I6f1e84187a2d4b7c55843f9cd5cf0691f40c11fc
Reviewed-on: https://review.haiku-os.org/c/1405
Reviewed-by: waddlesplash <waddlesplash@gmail.com>
* This isn't perfect, but should be a solid start.
Change-Id: Icb3f8c57b2bc4148d2ac14a71375d3f36c15e13b
Reviewed-on: https://review.haiku-os.org/c/1403
Reviewed-by: Alex von Gluck IV <kallisti5@unixzen.com>
* Add generic riscv64 support.
* Previously, missing architecture syscall.inc's would
result in obsecure unrecognized opcode messages.
* The logic should be identical, but with a friendly error.
Change-Id: I8f4878e99faf0f64465d2a27c3da2dd1e43c4ba5
Reviewed-on: https://review.haiku-os.org/c/1402
Reviewed-by: Alex von Gluck IV <kallisti5@unixzen.com>
Change-Id: I9a8fcd8efdf6f728ee0938d10a4b0e7dfbf2ca7e
Reviewed-on: https://review.haiku-os.org/c/1401
Reviewed-by: Alex von Gluck IV <kallisti5@unixzen.com>
FreeBSD has the names we used before this commit; OpenBSD has names
more similar to the ones used in this commit (though following their
conventions of course.)
This makes the code significantly more readable.
When there is only one disk, automatically select it.
Otherwise, do not select anything to make sure the user does not
accidentally do things to the wrong disk.
Change-Id: I834415bda6bc43692697bdfa9030f35ffc94048f
Reviewed-on: https://review.haiku-os.org/c/1397
Reviewed-by: waddlesplash <waddlesplash@gmail.com>
Code was comparing "/usr/bin/" with "/usr/bin/env" of whatever was
passed. We now only compare the substring.
Change-Id: I0d09589fa83227d88fc8ba458eaf4293e86a041b
This enables the Wacom Bamboo Pen/Connect (CTL-470 Model) to be
recognised by the wacom driver. Pen support and pressure-sensitivity
is also working fine.
Tested on hrev53021.
Change-Id: I165a03e4a60f4cad8537122fdc2992ff291b9889
Reviewed-on: https://review.haiku-os.org/c/1392
Reviewed-by: waddlesplash <waddlesplash@gmail.com>
starsseed told me it worked fine and is actually nice for demos :-)
Another option could be to add a menu entry to open a new window, but
I'm just back from JDLL so I'm quite lazy :p
Change-Id: I91e34517812f84157db7e24f2edff587deabbbe9
This set is only used until the SetupEnvironment script has run, i.e.
for launch_daemon and its initial tasks, which does not need the user
paths. Solves a TODO.
Previously we only rewrote ones pointing to "env" with that.
Change-Id: I775fedc4a3b8e0e590c76f50f4c3fb259c5df32c
Reviewed-on: https://review.haiku-os.org/c/1389
Reviewed-by: Axel Dörfler <axeld@pinc-software.de>
Reviewed-by: waddlesplash <waddlesplash@gmail.com>
The variable is used after a "cd" so it must be absolute.
Change-Id: I8d591b8642179603249ac56a1fcac082f29c870c
Reviewed-on: https://review.haiku-os.org/c/1387
Reviewed-by: waddlesplash <waddlesplash@gmail.com>
We don't reschedule or block in here, so checking for interrupts
is needlessly strict. Sorry for the noise, forgot to remove this
before the previous commit.
