- Introduce base classes RemoteDebug{Request,Response} and the first set of
implementing subclasses for the various types of debug requests. These
encapsulate the information needed to make a request to the target debugger
interface, as well as a means to flatten/unflatten to a BMessage. These will
be used to marshal the relevant information for transmission across the
eventual transport interface. Not used anywhere yet since there still remain
some requests and responses to implement, among other things.
Architecture:
- Store and provide accessor for the size in bytes of the low-level
debug_cpu_state size of the respective target CPU. Adjust subclasses
to pass in the appropriate size information.
- Split into separate subdirectories for each interface type, as is done in
the target_host_interface subdir. Preparation for the remote interfaces,
which will have quite a few more components than the existing local ones.
Originally, the tmp folder was in /boot/common, where it didn't show
an alert box. Since the move to /system, it simply fell into the general
"is in system directory" case.
Fixes#10173.
Matches the text used elsewhere on the system (and that of the "add-repo"
command.) You can still type a single character to confirm/deny.
Fixes#11260.
The only file in here was Deskbar.h, which just included the "real"
Deskbar.h. Considering nothing in-tree cares about this, and only
a small number of applications at HaikuArchives do, and that Haiku already
broke BeOS source compatibility in lots of other ways, let's just remove
this deprecated cruft.
We aren't really using Bullet itself; there are only two classes
(Quaternion and Vector3) which are from Bullet, and even they were
heavily modified. Furthermore, the Bullet license is really just the Zlib
license, which the Bullet source repository confirms (and you can see
by looking at its text.) It does not require attribution in documentation,
but only in the source code, and we aren't even using a substantial part
of it.
... as we obtained permission from Mark Kilgard to do so (see commit
c3eddfc0682b99f28ddbe33a14dadc40dc5b5f71.) So update AboutSystem and
the licenses directory to reflect that.
This file was introduced with the rationale that bots attempt to scrape
GitHub repositories for license data. GitHub now has an API for this,
and since all its sniffer rules fail because this file is not identical
to any known license, it just marks it "Other" and includes the full text
anyway.
Since this is the case, nobody else wants to look at all the licenses
in a single monolithic file, so we just put the explanatory message
and a reference to the directory containing all the licenses at the top
of the file, and include only the MIT license.
Also, GitHub renders Markdown nicer than it does plain text, so use that
(which matches what we already do for READMEs.)
* TD chains were broken, any transfer above ~1MB would trigger a KDL.
* _LinkDescriptorForPipe() links the last TD of the chain instead of the first.
* Still buggy: the transfer ring can still block and transfers then fail.
* Enable TRB_3_ISP_BIT where appropriate (Interrupt on Short Packet).
* Also fix WriteDescriptorChain() and ReadDescriptorChain() for multiple-TD chains.
There are still asserts in inet_pton4 and 6 anyway.
libgit2's tests expect inet_pton() to fail with EAFNOSUPPORT on unsupported
address families instead of crashing...
Discussed with kallisti5 and PulkoMandy before committing. So far, the
only bugs since korli's recent changes either are minor ones that don't
appear to affect functionality (e.g. #12860), or so bad that nothing works
(so, just as good as if the controller was disabled.) PulkoMandy reports
that his controller works but wasn't whitelisted, as have some other users,
so I'm enabling it by default.
If there turn out to be too many broken devices, we can revert this. Or only
revert it in the beta branch, if needed.
* Not sure if cursors could also have triggered this, but the memory
allocator can now outlive its ServerApp.
* However, this may also reveal cases of memory that is not freed
correctly.
The test in the previous commit had a totally flawed invocation of strcmp().
I probably shouldn't attempt to write C code without consulting the manual.
Even if it had worked, it would have worked on too many cases, including
some volume paths. So I reordered the if statement and added it in there.
This includes FSCopyAttributesAndStats as well as an older version of
FSGetDeskDir, and adds another alias of FSGetTrashDir. Fixes BeIDE
compatibility.
BPoseView::NewFileFromTemplate does not want to copy the times. Fixes#11702.
We have to keep a 2-argument version of this function because it looks like
some third-party applications might call it (e.g. Dockbert), and even
though it's technically private API, it's easy enough to preserve binary
compatibility.
If the pose matches the current type-ahead filter, it will be added
by the node watcher anyway, so this just created duplicate poses
which messed up internal data structures. Fixes#11122.
This is instead of just checking for NVIDIA devices. Also add an entry
for Intel 0x2668 (VirtualBox's HDA controller). Gets HDA a bit further
in VirtualBox, but an acquire_sem() times out later on.
