Removed workaround for already fixed bug 4217 and removed the call
to SetExplicitPreferredSize(), since the layout system was always able
to calculate good values for PreferredSize.
So for me the call seems to be superfluous.
Fixes#13353
1: Changed CPPUNIT_ASSERT -> CPPUNIT_ASSERT_EQUAL
2: Fixed coding style: Added two lines after including libraries, maintained 80 characters(MAX) in each line and space at comment start
3: Fixed the failure for "BString::Search" test "i != 0 by chaning the testing method(StartsWith->IStartsWith) and changed the string to be searched(sT->st)
Change-Id: I1237d1f2d0e3af7757963cc940bae929f487f088
Since wait_for_objects can wait on sems, threads and ports in addition
to FDs, limiting to RLIMIT_NOFILES as in the select/poll case does not
work. Since space is allocated for the wait objects in kernel memory,
limiting their number to a valid range is still desireable.
The limit is now placed at the sum of max sem, thread and port count
plus RLIMIT_NOFILES.
This also fixes a signed vs. unsigned comparison warning in
check_max_fds introduced in hrev51866.
The debug build of the packagefs kernel addon failed because user_memcpy had no type.
Using the correct headers for user_memcpy and IS_USER_ADDRESS fixed the issue.
because it needs extra space for storing data, and flexible array member o
doesn't do this.
I thought it was...
Change-Id: If64c06827809a4e021581c6adf8e0f198cf47450
It was limited to a uint32 and could for example be overflown by the
slab MemoryManager that uses size_t on a 64 bit system.
This aligns the signature with create_area() that already uses size_t
for the size argument.
Note that the function is currently private, so the impact should be
limited.
A new install should start with an empty syslog. Adding
the syslog path to the set of ignored paths will prevent
an existing syslog file from being copied into the new
install.
Fixes#8373
Change-Id: I79ecf95773a34360185d38ee4ef0c8337b062e7a
Inode::_FindPath() always deletes the inode it uses regardless
the source, including when it's from the HashtableInodeSource.
But the later returned the inode object directly from the hashtable
when it's inside, so it gets deleted bu _FindPath… then reused later on
when searching the hashtable again.
I'm not sure it's the correct solution but at least malloc_debug doesn't
complain anymore.
There is a problem primarily manifesting itself in
the x86_64 build where the packages' names from the
solver are not correlating with the names of the
as-configured depots. There was also a problem
with a local variable being the same name as a function
parameter. This does not fix the underlying problem,
but avoids further anomalies in the HaikuDepot UI.
Possibly related trac #11317, #11674, #13940.
Change-Id: Ic140f114bbe38e59c78760213843bf492ff7a270
DwarfImageDebugInfo:
- In some, but not all cases, gcc5 generates type information where the DIEType
is a child of a namespace rather than of its containing compilation unit.
This needs to be taken into account when building our name lookup table, as
we'll otherwise not find the full definition of such types when attempting to
locate them for corresponding variables. Fixes an issue reported by Axel.
* Show old page table location and provide more feedback
* 16 int32 * 0x10000000 > sizeof(int32), fix to uint32
Change-Id: Ib68c34f5d3c6bfa1da53241e6586c07e4e494750
* calling TRoster::GetShutdownApps() multiple times with the same lists
leads to problems because the lists aren't emptied first.
* instead we watch the user app launches, add them in the fUserApps list,
and let _QuitApps() iterates through the fUserApps list until it is empty.
John's revert of my removal commit dragged back a bunch of cygwin/sunos
cruft, as well as re-adding RegExp.cpp to the host libshared, that we don't
need.
Instead, remove this and add libgnuregex_build to just the tools/keymap
link alongside the FreeBSD gnuregex case.
User can now launch applications in USER_APP_TERMINATION_PHASE
(like when there is a dialog to save an open document). Fixes
Change-Id: I3126e6ab8d2fd76016becb6ce0f726f82ecb7f3b
When the application is starting, it will load in bulk data. As it
does so, it may issue some change notifications to listeners. For
each small change in packages' data, there is a notification. The
system was built to accommodate coalesced changes. This commit will
coalesce those changes during the bulk load for cases where there
are listeners present.
Change-Id: I1cc6668f61917c0068d5559b2bf338ea75d453da
As discussed at #13860.
* Moved "Show only featured packages" from the "Show" menu to a
checkbox in the filter view.
* Moved "Repositories" popup menu into the menu bar.
* Removed the "Featured packages" title.
* Entering a search term doesn't override the "Only featured"
filter anymore (used to disable the "Only featured" filter).
* At this point we want to avoid the user calling the callbacks
in the form of BMediaConnection. Instead we force to use the
BMediaInput and BMediaOutput versions.
The server side has the ability to reject client
requests where the client is no longer supported
because it is presumably too old. This change
will inform the user when this happens and will
prevent the client from attempting further server
communications within this execution of the
application.
While installing packages through pkgman, it will inform if specific
package is already installed. Fixes#12447 : [pkgman] inform if a
package is already installed.
Change-Id: I194bc849c42ba52a696a6cb52d87aebaff530f35
Output-only events (B_EVENT_ERROR, B_EVENT_DISCONNECTED and
B_EVENT_INVALID, with B_EVENT_INVALID masked out before passing down
events) are used to indicate error, so they should not be notified if
the vnode does not have select().
Bug: 13965
Output-only events (B_EVENT_ERROR, B_EVENT_DISCONNECTED and
B_EVENT_INVALID, with B_EVENT_INVALID masked out before passing down
events) are used to indicate error, so they should not be notified if
the device does not have Select().
Bug: 13965
Output-only events (B_EVENT_ERROR, B_EVENT_DISCONNECTED and
B_EVENT_INVALID, with B_EVENT_INVALID masked out before passing down
events) are used to indicate error, so they should not be notified if
the filesystem does not explicitly provide an fd_select() override.
Bug: 13965
also check buffer addresses passed through structures. For transfers, first compute
the direction of the transfer, to copy before or after the actual transfer.
TypeHandler:
- Add name field for presentation purposes. Adapt subclasses accordingly.
TypeHandlerRoster:
- Add methods to count and retrieve all type handlers for a given type,
and adjust CreateValueNode to allow for passing in an explicit handler.
Adjust callers accordingly.
VariablesViewState:
- Add helpers to store an explicitly chosen type handler for a node.
TypeHandlerMenuItem:
- ActionMenuItem subclass that takes care of reference management
for its contained type handler.
VariablesView:
- Add context menu for choosing type handlers if applicable. Implement
support for invoking said type handlers in a similar manner to explicit
typecasts.
- Adjust saving/restoring the view state so that hidden nodes are taken
into account as well. This is necessary since it may be the case that
the handler had to be applied to the hidden child rather than the visible
node (i.e. the BMessage handler when applied to a pointer to a BMessage).
All together, these changes allow choosing to switch between views of a type
when the Debugger has multiple handlers for it. For example, for BMessages
this allows switching between displaying the raw underlying structure vs
the decoded message content.
* use MergeObject even if we don't need the merged object: the rule does what
is needed and is standardly used for instance in libroot.so.
* add a grist to source files for the bootstrap glue.
SMAP will generated page faults when the kernel tries to access user pages unless overriden.
If SMAP is enabled, the override instructions are written where needed in memory with
binary "altcodepatches".
Support is enabled by default, might be disabled per safemode setting.
Change-Id: Ife26cd765056aeaf65b2ffa3cadd0dcf4e273a96
This change re-instates the change-log function (TRAC#13827) that was
dropped in earlier changes. Also now-disused API handling logic is
removed. A small code-style fix is made and the version of the
application has been incremented so that this version's logic can later
be detected by the HDS application server.
gcc does re-align the stack in the main() function prologue, however,
we still need to set the right alignment for other cases: thread entry,
.init and .fini code that is executed before main() is called or after
it has returned, and signal handlers which may use a separate stack.
Part of #10509
Tested on x86_64 and x86_gcc2, works on my machine and most likely fixes
that was most likely due to missing lock sync in previous ACPI upgrade.
* Redid some of kallisti5's work as it was easier to start from earlier
version.
* Evaluating 64 bit physical addressing on 32 bit Haiku again as this
was likely the cause of pagefaults (#12377):
"Implemented internal support for full 64-bit addresses that appear in all
Generic Address Structure (GAS) structures. Previously, only the lower 32
bits were used. Affects the use of GAS structures in the FADT and other
tables, as well as the GAS structures passed to the AcpiRead and
AcpiWrite public external interfaces that are used by drivers. Lv Zheng."
* Mostly bugfixes in ACPICA
With user buffer checks and copies about the same for every driver, it seems logical
to avoid duplicate this code.
Sponsored-by: https://liberapay.com/korli
- Some of the recent changes for memory management exposed a race
condition, where the teams window would potentially try to access
already freed objects on quit. Ensure we acquire references to the target
host so this doesn't happen.
AttributeClasses:
- Fix typo in get_attribute_name_classes which resulted in us not
handling DW_AT_linkage_name properly.
- Fix incorrect class specification for DW_AT_default_value due
to inconsistencies in the DWARF documentation (the table of attribute
classes indicates it is only a reference, but the detailed description
indicates it also being possible to be a flag or constant, both of
which gcc outputs).
DebugInfoEntries:
- Add accessor for DIEClassBaseType's inner types.
DwarfImageDebugInfo:
- When looking up types, create a basic target interface to pass on
to the type context. Otherwise, type lookups that required DWARF
expression evaluation would crash.
- When building the type name table, we now recursively walk a class's
inner types, and add them to the list as well. This omission would
cause the debugger to lack the type description for such classes,
and consequently be unable to display their details in the variables
view.
AbstractTable/Table/TreeTable:
- Let BColumnListView take care of deleting columns. The previous
approach was resulting in the columns being leaked.
Debugger:
- Don't detach the UI reference, as TeamDebugger will acquire its own.
GraphicalUserInterface:
- Cleanup of file panel handler.
SourceView:
- Clean up marker manager.
*Roster:
- Clean up registered objects in destructors.
GlobalTypeLookup:
- Clean up hash tables.
NetworkTargetHostInterfaceInfo:
- Fix reference handling for settings.
BListValueNode:
- BList's item count member is an int32. However, when the handler was reading
said variable, it was mistakenly reading the number of bytes corresponding to
to the address size of the architecture instead.This would cause the read to
fail, and consequently BList and/or BObjectList variables to not display
their contents properly on x86-64.
- Fix various cases where OpenHashTables weren't being cleared properly.
- Fix various reference counting errors.
- Simplify FileManager reference handling.
- Fix bug in LocatableDirectory where the directory named '/' would have its
name returned as empty. This would lead to failed lookups for entries already
in the table, and ultimately corrupted the hash table when deleting unused
entries, leading to #13939. This was previously never noticed due to the
entries not being freed properly.
- AbbreviationTable wasn't clearing its entries.
PlaylistTrackCount returns the number of tracks in Playlist,
and PlaylistTrackTitle the title of a numbered track.
Also, remove B_TRANSLATE macros in scripting properties
when compiled in kernel mode, BMemoryIO::WriteAt() and BBufferDataReader::ReadData()
now check the buffer to write to, and eventually call user_memcpy() instead of memcpy().
* avoid assignment in if statements
* use boolean conditions
* use copy_from_user_value/copy_to_user_value instead of user_memcpy
* no functional changes except user buffer addresses check in:
_user_get_next_disk_device_id(), _user_get_disk_system_info(),
_user_get_next_disk_system_info(), _user_find_disk_system()
* remove TODO "Add user address checks and check return values of user_memcpy()!".
In the kernel and command-line tool, don't leak allocated memory,
even if the tool returns an error.
In the command-line tool, also handle memory allocation errors
nicely by giving the user an OOM message if allocation fails.
Fixes CID 1425367 and 1425224.
Signed-off-by: Augustin Cavalier <waddlesplash@gmail.com>
It appears as if there is a typo in printf_size.c; it makes
no sense to compare an unitialized variable in such a way.
Since `info->width` is referenced later on in the variable
assignment, it seems possible that the intended statement
was to place the larger of the values between `info->width`
and `info->prec` into `width`.
Fixes CID 609431.
korli: Patch applied in cddfcf2f87
was incomplete. This patch completes with what upstream did in 2004:
fa5753eede.patch
Signed-off-by: Jérôme Duval <jerome.duval@gmail.com>
Previously dead teams were always returned (as needed for wait/waitpid).
Using waitid it should however be possible to omit the WEXITED flag to
only get stopped (WUNTRACED | WSTOPPED) or continued (WCONTINUED) teams.
Adjust the other caller of the _kern_wait_for_child syscall, used by the
various less specific wait* functions, to always include WEXITED. This
avoids having to special case waitid in the syscall interface or with an
extra flag.
Add a check to ensure that any of these flags is set to avoid the now
possible case of nothing being specified. This fails with B_BAD_VALUE
(EINVAL) as on other systems.
On other systems their values are usually aliased. We defined them as
two different values, but didn't handle WSTOPPED anywhere. The check
is now simply extended to cover both bits.
When wait_for_child was called with a child argument equal to the
current team, the team was locked again to check if it was a child of
itself.
This is now handled the same way as on other systems, where this case
results in ECHILD (and not EINVAL).
DwarfImageDebugInfo:
- Type name/info entries weren't being freed properly.
BaseUnit:
- The array of entries itself was being destroyed, but not the actual
objects. Consequently none of the DIE objects were ever being freed.
CfaRuleSet:
- The register rules array wasn't being freed.
FileManager:
- References to entries weren't being released. Some more work remains to be
done here though, as a subset of these objects still have a non-zero ref
count in the end.
Following recent changes to use libroot_build on Haiku also, it is now
actually impossible to build Haiku components on non-Haiku platforms
(BeOS R5, Dan0, BONE, Zeta), so we can remove any logic related to this.
This is only the first part; still to be removed are:
* SetSubDirSupportedPlatformsBeOSCompatible
* HOST_PLATFORM_BEOS_COMPATIBLE
* TARGET_PLATFORM_BEOS_COMPATIBLE
It was discussed and introduced based on docs that some systems
need this to shutdown properly. I can find no mention of this in
ACPICA or ACPI docs.
This needs to be re-evalutad, as all my shutdowns been successful
after disabling it and I can't locate where this info came from
or if it actually helped.
See 1316462ab0 for original commit
and bug 12306 on current shutdown issue.
As it turns out, using the xattr emulation layer plus "libgnu"
causes some strange mixups at package build time, and so packages
built with it were winding up with no attributes at all.
So I've just bitten the bullet and written a full passthrough layer
to the system attributes. Verified using a full build of haiku.hpkg
this time ... after a lot of painful debugging of symlink mixups.
Hopefully I am finally rid of this plague...
* The global lock, needed for hw <-> os to sync, was a no op.
This means that OS and HW could read and write to same value
at the same time, causing garbled data and crashes. This is easiest
to see in battery status, where battery values and OEM name would
become garbage. It also causes instability and failure to shutdown
* ACPI used 64 bit addressing (which it normally should) on 32 bit.
This would cause page faults as it would jump to incorrect addresses
To quote jscipione (from 95e8362c52),
"Let me tell you a story about a bug" -- though this tale spans a much
lesser time than that one did.
In 5e19679ea3, I enabled libroot_build for
Haiku, instead of using the system libroot as we had before. There were
a number of bugs introduced along with this that I hadn't fixed (and there
may be more after this), but most of the obvious ones (crashes on x86_64...)
were fixed shortly enough.
Attribute usage, though, was a different story. Unlike most of the POSIX
calls in libroot, which were aliasing system functions no matter what the
platform, the attribute calls were not, as they are specific to Haiku.
Initially I had completely forgot about them, and it wasn't until a few days
later when I noticed that I had an "attributes" directory in my generated
that I realized that the "generic" attribute layer was being used on Haiku.
I attempted a fix for this in 5e19679ea3,
thinking that would clear the problem up, but I didn't actually run a test
beyond seeing that my BuildConfig had been updated properly. In fact,
BuildSetup was hard-wired to not even pass that definition through on
Haiku, and so that commit had in effect caused nothing.
My initial "fix" of just changing BuildSetup then caused a build failure,
as while libroot_build itself compiled, it ran into errors whenever attributes
were used, because in letting the real libroot's attribute calls shine
through, I had bypassed libroot_build's FD emulation/shim layer.
Then I tried and failed at three separate attempts to solve this with code:
- a version of the "fs_attr_...h" interface for Haiku. This proved possible
in theory, but in practice I would need to reimplement a lot of attribute
handling code in it, because all I had access to from there was syscalls.
- a version of "fs_attr_untyped" that bypassed its reimplementations of
the "fs*attr" functions for the libroot ones, only using the FD shim layer.
This proved possibly not even theoretically possible because it would have
caused preprocessor hell in some of the build headers, and also assumptions
about how attributes are read were totally different.
- a completely new "fs_attr_haiku" that was a completely new interface to
the fs*attr functions. This proved practically impossible because of the
need to include structures from the system libroot to call out to readdir,
etc. that attempts to solve would also have caused preprocessor hell.
Then I realized that the Linux xattr emulation library, which I'd used
as a reference when attempting the first solution, was shipped by default
as a system library in all builds of Haiku ... and so I could just tell
fs_attr_untyped to use the Linux xattr handler, and then link against libgnu.
So that is how I arrived at this strange and decidedly unorthodox solution
to a problem of my own creation.
Per #10267, "Most (ported) third-party software should be removed
from the Haiku source repository."
Since HaikuPorts already has the ncompress package, this file
should no longer exist.
Signed-off-by: Augustin Cavalier <waddlesplash@gmail.com>
Original patch missed modifying the "minimum" definition and the
src/bin/Jamfile, so I took care of that.
In MediaPlay.cpp, media_play returns non-zero value on failure but forgets to return zero on success.
Coverity CID: 1422617
Signed-off-by: Augustin Cavalier <waddlesplash@gmail.com>
set_var doesn't return any value yet its return type is ULONG, and Coverity freaks out about that. This patch turns that function into a void function.
Coverity CID: 603433
Signed-off-by: Augustin Cavalier <waddlesplash@gmail.com>
TeamWindow:
- Properly delete stack frame selection entries in destructor.
TeamDebugger:
- The reference to TeamDebugInfo wasn't correctly initialized to own,
leading to TeamDebugInfo never being destroyed.
FunctionInstance:
- Also clear source code reference in destructor.
ImageDebugInfo:
- Release references to specific infos in destructor. These are acquired
on our behalf when loading the image info, but were never released,
leading to the latter never being freed.
DwarfManager:
- Release references to files in destructor.
FileManager:
- Release entries in various destructors.
Somewhat improves the situation in #13800, but is still far from complete,
as a thorough review of this nature hasn't ever really been done.
TeamDebugger:
- In several cases, BReferences on thread handlers after a call to
_GetThreadHandler() weren't correctly initialized to account for
that function implicitly acquiring a reference on our behalf, leading
to the reference count to spiral out of control, and consequently the
handler would never be freed. Found while attempting to address
#13800, likely more to come.
- Recent changes to the build system appear to make the assumption
that the GNU regex APIs are universally available. This isn't the
case on FreeBSD, which requires libgnuregex to provide that
functionality. This broke the host keymap build.
* Update the work status view with package state changes from
MainWindow instead of PackageListView, because PackageListView
does not always show all packages that are going to be installed
(e.g. dependencies that are not shown when a search filter is
active).
* Fixes the bug that prevented pending/download status to show up
for dependencies when filters are used.
* During package population, when the repository is being
refreshed, thousands of package change messages arrive
in MainWindow while the PackageInfo fields are filled with
data.
Every such message caused a full generation of a package
list (including applying filters to all known packages),
which is quite slow, and caused the freezing from bug
#13823. The list was only used temporarily to figure out
if the package should be visible now or not.
* Instead of generating a new package list, we now only
match the single updated PackageInfo against the current
filters.
* Fixes#13823
* To reduce CPU usage even more, it would be better to not
generate at all, or not subscribe to, package changes while the
list is being initially populated, and/or to coalesce all the
changes to each package info into one update message (instead
of one for title, one for prominence, etc).
Deferring subscription requires some additional thought on when
to subscribe then, and how to avoid leaving a "hole" where
changes would go unnoticed.
It was needed on macOS for a time when BUrl used regexes for parsing.
Now it does not, and so we can remove libshared's RegExp from build
libshared, and thus also libgnuregex.
scripting properties:
* Info, Grid, and Stick (true/false)
* CopyImage
* MakeSquare
* Zoom (1-100)
Decreasing pixel size from the menu will half the size
until it's less than 32.
fixes#13884
Signed-off-by: Adrien Destugues <pulkomandy@pulkomandy.tk>
Fixes Coverity issue CID 1340122.
Also resolves two TODO items in the code: allowing the deletion
of a selection of cookies, and deleting all cookies from a site
when there are no specific cookies selected.
Signed-off-by: Adrien Destugues <pulkomandy@pulkomandy.tk>
The feature gives user ability to choose the position of notifications
out of Follow Deskbar, Lower Right, Lower Left, Upper Right and Upper
Left. Fixes#9749 - Notification_Server: add the ability to choose the
position of notifications (easy).
Signed-off-by: Adrien Destugues <pulkomandy@pulkomandy.tk>
Previously I had intended to take the simpler route and just lock the
already-inserted team before setting the io_context (as in prior commits),
but after hearing some reports from users that some other seemingly
unrelated KDLs had possibly cleared up after the first iteration of
that fix, I decided to go with this route instead.
Now we do not insert the team into the parent and hash and send the
notification until just before the team's main thread is actually started;
i.e. we now initialize not only io_context but also the team's args, VM
address space, and user data (and if creation of any of these fails
we do not inset the team into the hash at all.)
Since the team structure was not locked at all while this initialization
was taking place, any number of race-dependent bugs could have been
caused by this on multicore systems.
* Further improves the logging and provides some
basic performance numbers.
* Moves the bulk-load logic out of the data-model
class.
* Introduces a state-machine for the bulk-load
process so that it will be more easily able to be
shifted to non-blocking IO when the HTTP libraries
can do that.
* Implements concurrent loading of the bulk-data to
hopefully improve lead time for icons and meta-data.
* Loads data to a temporary file and then moves to the
final location in order to avoid partially written
data in the cache.
* Handles situations where no network is available;
prevents attempt to access the network.
* Allows bulk-load processes to be cancelled when the
application quits.
* Introduces command-line arguments to help simulate
scenarios to help with testing performance and
network absence.
* Implements ordered insert and binary search in the
'List' class + basic unit test.
The KDL in #13861 was apparently caused by a NULL io_context. I read over
the Team code and have applied the following fixes here as a result:
1. A NULL io_context is entirely possible, as implied by the NULL check
before deinitializing it in Team's destructor, so account for that.
2. Team objects have a class-wide lock which must be acquired before
directly accessing class members, so use that here.
First part of the fix for #13861.
calloc, when multiplying to find the true size of memory desired
by a user, will occasionally overflow if the numbers are too
large. This commit makes calloc ensure that the multiplication
does not overflow by dividing the result by one of the numbers,
and checking to ensure that the result is equal to the other
number. If there is overflow, it is treated as an OOM error.
Signed-off-by: Augustin Cavalier <waddlesplash@gmail.com>
I wrote this back in 2010 as my first driver project.
Reasons to remove it:
* The license is GPL
* Current WiMAX hardware is generally WIFI based.
* It controlled the hardware, but never worked
for network access since we need SSL certs and stuff
which vendors weren't too open with.
* WiMAX kind of died (at least in the US)
I left the wwan directory, it would be a nice spot
for CDMA / GSM dongle drivers.
Implemented by just not defining the functions at all and letting
the Haiku ones shine through. In the future, though, we should wrap
them properly.
(Between the switch to libroot_build and this commit, libroot_build
was using the generic attribute emulation layer on Haiku ... whoops.)
I didn't notice this in the previous commit because apparently GCC2
just links against libroot's versions of them. On GCC5, however,
the version from libroot_build was used even for calls from libroot itself,
which led to infinite loops and then stack overflows.
So instead we must have the "syscall" functions in libroot_build shadow
the real ones by being named differently, which I did by changing their
prefix from "_kern" to "_kernbuild" via preprocessor macros.
Since the build syscalls.h is now substantially different than the non-
build one (and has not been synchronized in nearly a decade anyway),
I've just stripped out all the syscall defns except for the ones actually used
in the build.
Thanks to kallisti5 for helping me debug and test.
Previously we just used the system libroot, which of course meant
that when libroot's ABI changed, the build broke. Now we use the full
libroot_build that we do on non-Haiku platforms. The logic for "BeOS-compatible
but not Haiku" does not really apply anymore, so it has been gutted where
appropriate (and libhaikucompat has been decoupled from the build.)
The only caveat here is the change to Errors.h -- we really should be using
the system's one where I included the one from the tree, but for whatever
reason, GCC2 refused to handle the #include_next properly.
Fixes the build breakage of Haiku-on-Haiku by my prior commits (sorry).
* The GPL license here was a relic from the past to help
control YellowTab's use of Haiku source code. The intent
at the time was to relicense MIT. Some GPL parts remained.
* This is going to get harder and harder to correct over time.
* This code makes Haiku *not* MIT. Lets correct that.
* I reached out to everyone who made changes to these files and
got personal approvals to relicense MIT.
* If I missed somebody, *and* you feel as though the MIT license
doesn't meet your needs, please contact Haiku, Inc. and we will
work to revert your changes.
Marcus Overhagen EMAIL - MIT OK 11/29/17
Stephan Aßmus EMAIL - MIT OK 11/28/17
Axel Dörfler IRC - MIT OK 11/29/17
Ryan Leavengood KEYBASE - MIT OK 11/28/17
Michael Lotz IRC - MIT OK 11/29/17
Adrien Destugues EMAIL - MIT OK 11/28/17
Joachim Seemer EMAIL - MIT OK 11/28/17
Jonas Sundström EMAIL - MIT OK 11/29/17
François Revol IRC - MIT OK 11/29/17
Jérôme Duval EMAIL - MIT OK 11/29/17
Oliver Tappe EMAIL - MIT OK 11/29/17
Dziadek EMAIL - MIT OK 11/29/17
Philippe Saint-Pierre EMAIL - MIT OK 12/6/17
Philippe Houdoin EMAIL - MIT OK 12/8/17
Dario Casalinuovo IRC - MIT OK 12/13/17
We need to handle a case when node size is small reasonably it can be merged with another node or push data from other node to this node.
Signed-off-by: Augustin Cavalier <waddlesplash@gmail.com>
We need to handle a case when node is full, the solution should be split or push data to another node.
Signed-off-by: Augustin Cavalier <waddlesplash@gmail.com>
CopyOnWrite works like this:
* Cache original node
* Allocating new block
* Cache new block to be writable
* Copy original node to new node, and changing.
Also if a node is already be COW-ed it cannot be COW-ed again, it will be changed in-place instead.
InternalCopy does CopyOnWrite all the nodes that we don't need to change anything on them.
Signed-off-by: Augustin Cavalier <waddlesplash@gmail.com>
Some details about the current Journal:
* Journal can only end transaction.
* It holds a transaction id of fs (fCurrentGeneration) that increments each time a transaction starts.
* _TransactionWritten now just printing message.
Signed-off-by: Augustin Cavalier <waddlesplash@gmail.com>
There are 4 new classes, structs:
* CachedExtent, is a AVLTreeNode, caches the extent locating in extent tree, a extent can be free, allocated, metadata or a data extent. It also hold a references count,
that is incremented each time a new extent refer to it (COW) and item data, that is only for allocated extent (NULL for free).
* CachedTreeExtent, is a AVLTree, cache the whole extent tree and has CachedExtent as its node.
* BlockGroup represents the group of extents that represent the region for each type of allocated extent. For example, region for data extents, metada blocks. It
responsible for inserting nodes in CachedTreeExtent.
* And the final, ExtentAllocator it knows how to allocate/deallocate extents, but for now only the allocating is implemented, actually allocating and deallocating works
are already implemented, they are in functions _AddFreeExtent, _AddAllocatedExtent in CachedTreeExtent. However the deallocating is not needed for now, so it will be
finished later then.
Signed-off-by: Augustin Cavalier <waddlesplash@gmail.com>
Add BTree::Path as a attribute so enhance performance, so that everytime we iterate through items it wont search all the root to leaf
again. The Iterator is initialized without rewinding to make more flexible.
Signed-off-by: Augustin Cavalier <waddlesplash@gmail.com>
Remove attribute fCurrentSlot in BTree::Node as it will be handled by Path explicitly. BTree control Path by passing its type in
BTree's method, Path also hold BTree type as its attribute to do some internal actions.
Add constant BTREE_KEY_TYPE_ANY, find search key has this type will success regardless of the found key's type.
Split the the _Find function into Traverse and GetEntry. Traverse will fill in the Path (nodes and slots) along way its finding,
GetEntry will get the item data, item size, key from leaf, if the slot is valid, that we have from Traverse. The _Find function also
check type if is correct and then retrieve. Doing this way there will be more flexible, the "read" flag is not needed as we only
need Path to manipulate tree, and it also enhance the performance at some points, because Path caches all the nodes from root to leaf,
so that we don't have to block_cache_put and block_cache_get after each finding.
Signed-off-by: Augustin Cavalier <waddlesplash@gmail.com>
* Not handle traversing type correctly (looks for the graph).
* Reorder the codes because *slot is uninitialized if type is BTREE_EXACT.
* Incorrect return type: int32 -> status_t
Signed-off-by: Augustin Cavalier <waddlesplash@gmail.com>
The glue code pushed 12 bytes to the stack, breaking the 16-byte stack
alignment requirement. This would be fixed by the main() prologue from
gcc, but all "init" and "fini" code (static/global constructors/ destructors)
would run with a misaligned stack.
This was already fixed for x86_64 in hrev49731. Note that the fix here
is slightly different, the pointer is realigned after it is saved to EBP
and the function epilogue restores it from EBP, so no changes to crtn.S are
needed.
This application tests the BMediaTrack/BMediaFile API and underlying
ffmpeg plugin for video decoding. You can press (or hold) any key to go
through frames in a video. This avoids debugging both MediaPlayer and
the underlying libraries at the same time.
It shows that ReadFrames is getting video frames out of PTS order.
I still don't get what's happening, but doing the cookie parsing at the
same time as the main thread is handling HeadersReceived seems to
trigger a memory corruption, and it will escape all my attempts to debug
it (adding printfs or any other slight change to the code will make it
go away). So just chage the order we do things and hope that's enough to
always avoid it.
As a side effect, HeadersReceived can now rely on the cookies being
already stored in the cookie jar, which I think makes more sense.
I still plan to rewrite the HTTP request code as a proper state machine,
instead of one long Run() function. This would allow to run it in
smaller steps, and thus group multiple requests in a single thread
(triggering them from poll, select, or similar).
Tracker uses KeyInfos.h from it. Previously this wasn't needed
due to the SEARCH_SOURCE added in the Shortcuts preferences,
but as that is now just a UseHeaders, we need it here (if only
for the DoCatalogs.)
i) there was an integer promotion problem in updating the retransmission
timeout : a signed int was being divided by an unsigned int. This was causing
the values to overflow. Thus leading to huge values for timeout which
manifested in the perception of pause in data flow.
ii) for an ack to be recognised as a duplicate ack, the advertised window
must remain same. This was not taken care of in the code so I added it.
Signed-off-by: Augustin Cavalier <waddlesplash@gmail.com>
Helps with #13769 but does not fix it completely (upload gets
farther but still stalls.)
The base VMCache class changed to the generic_ types with their
introduction in in *2011* (435c43f591),
but these classes were never properly adapted. These functions should not
be called here (they panic() -- but the base class only returns B_ERROR,
so that is a difference at least.)
Found by Clang's -Woverloaded-virtual.
- When reading the areas note, there was no check in place to ensure that a
matching segment for the address was actually found. This would later lead
to a crash when attempting to read memory from the corresponding (missing)
piece of memory in the core.
- Don't remove .eh_frame in the linker scripts, as this saves little
space in practice, and results in debug binaries of the runtime loader
not being properly debuggable due to missing canonical frame
information.
* Add a note about a workaround that can be used to run
test_app_server, which currently fails to load the right
library libbe_test.so. The bug should be fixed too, but
with this, it can at least be used until then.
The command to get the hub descriptor needs to be a class request and was incorrect causing a Stall error for the hub. The the driver does not handle a stall error yet, causing odd behavior.
Signed-off-by: Jérôme Duval <jerome.duval@gmail.com>
* Expects its config files in /boot/home/test_launch.
* Uses standard I/O, and is always in user mode.
* Also added test_launch_roster command that is able to talk to the test
server like it does to the real thing.
* When opening the screenshot window for a package, set the window
size to the largest dimensions of all the screenshots of this
package. Prevents it from resizing/jumping around when browsing
through the images.
Call newly created BDeskbar methods to adjust default Zoom().
if auto-hide is turned on don't consider Deskbar's frame.
In vertical mode allow windows to zoom on top of Deskbar if
auto-raise or always on top is turned on.
Deskbar can take up a significant area so you can reclaim that
space by turning on auto-raise or always on top.
In horizontal mode ignore auto-raise and always on top settings,
don't allow windows to zoom on top of Deskbar even if these
settings are turned on.
Deskbar doesn't take up very much space and you don't want your
window to get cut-off at the bottom.
Add methods to get and set "Always on top", "Auto raise", and "auto hide"
which are all booleans which control aspects of the Deskbar window to
BDeskbar.
Set the bool to the default value initially. Check if sending the
message succeeds, if so check the reply which also fills out the bool.
Don't check to see if reply succeeded because the bool will only be
overwritten if it did.
Follow the BDeskbar convention Is...() for getter, Set...() for setter
e.g IsAlwaysOnTop() is the getter, SetAlwaysOnTop() is the setter.
Define new message constants to call the newly created methods.
Follow BDeskbar convention: 'gtla' is used for getter, 'stla' for setter.
g/s for getter/setter, tla is an all-lowercase code unique to each
getter/setter pair.
Copy/paste these message constants into BarApp.h unchanged. Replace four
letter codes with imported message constants in BarApp.cpp and
BarWindow.cpp. Much nicer than using bare codes.
The new BDeskbar methods are all handled by TBarApp. The getters send
back a reply message containing the bool while the setters fall through
to existing setter cases.
Zoom() takes up all Desktop area excluding the area
occupied by Deskbar. This is calculated using information
from the BDeskbar class.
Window zooms just as you'd expect -- it takes up fullscreen
minus area taken up by Deskbar. In vertical mode the zoom
area depends on the width of Deskbar, consequently Zoom()
is more space efficient in horizontal mode than vertical
mode because the window doesn't use the area below Deskbar.
If the zoom limits are smaller than the Desktop area, the
zoom area shrinks towards the center of the Desktop not
covered by Deskbar. This is slightly different behavior,
the window insets off-center following Deskbar.
In some scenarios zooming to the non-Deskbar area was too
restrictive. I made an exception that if there is enough
room above or below Deskbar i.e. a short window, then Zoom()
instead insets from the screen edge ignoring Deskbar. Apps
which meet this criteria include DriveSetup and Expander.
* Hashing semantics for the new build repositories are different than
the old ones, so update those (if the x86 build was not broken before
it is now...)
* OptionalPackages has been updated slightly (removed libtool and git_cvs
from the default images, as they are rarely used nowadays and would pull
in a bunch of dependencies we don't really care for either)
* Removed lib:libqrencode from Haiku package requires (qrencode_kdl is a
static library, the userland libqrencode is not used anywhere in the tree,
as far as I can tell)
* Fix build of JPEG2000 translator after update
* Decouple fluidsynth build machinery and remove from image now that it
is no longer used
* Update repository URL in Repositories preflet
This is essentially the replacement for "jam upload-packages" --
it goes through a directory of packages (intended to be one made by
buildmaster), picks the packages out of it that are in the repository
file passed to it, and then hardlinks them to a second directory
as well as updates the repository file and creates a package_repo.
This is what was used to mass-modify the repository files which
will be updated in the next commit.
* Initial support for displaying multiple screenshots for packages
which have more than one. Still rough and unfinished.
Screenshot window now has a toolbar with prev/next buttons and
a busy loading indicator. Switching through the screenshots works.
There's currently a server-side bug which makes all data turn up
15 times in the JSON file, so please don't report a bug about
HaikuDepot showing 15 or 30 screenshots available when it's really
just 1 or 2 :)
Still to be done: toolbar icons instead of text labels; better
handling of screenshot window resizing; maybe thumbnails of
screenshots and preloading other screenshots in the background.
Main window also needs a way to indicate that there are more
screenshots than the one thumbnail, needs some more thought about
how that might look.
This concludes my HaikuDepot commits from the coding sprint at
KDC 2017 Toulouse!
* The UI became unresponsive while the PackageListView was filled
with all the packages. This was especially apparent when using
the search function, which clears and refills the view with every
typed character.
* Add a new worker thread with the task of asynchronously filling the
PackageListView. When a new data model is adopted, we hand the
thread a copy of the visible package list. The worker thread then
goes through the list and sends the package infos via BMessage back
to the MainWindow, in batches of 20 infos per message. When the 20
entries were added, it acknowledges this to the worker thread which
will send the next 20 infos (so UI messages can get in between,
keeping it responsive). The lists also get a unique ID so that
model changes while the list is populating will invalidate
previously sent messages (and cause the worker thread to cancel
processing the outdated list).
* Search is much nicer to use this way, staying responsive and
listing packages while typing. Still not perfect since the
PackageListView is still cleared and refilled each time a character
is typed, instead of just narrowing down the already displayed
package set. But that's to be improved on another day...
* Same applies to filling FeaturedPackagesView btw
* Use a hash table to find PackageRows by package name in
PackageListView, instead of doing linear search over all rows.
Improves performance of populating the list, since every
AddPackage() checks if a row for the package already exists.
* Add new WorkStatusView which keeps the user informed about what's
happening. It's a status bar at the bottom of the window which
shows on the left side either a spinning barber pole (for
operations without a progress), or a progress bar (for download
progress). Next to that is a text view showing a descriptive
status text.
* Currently, it will notify of the following operations:
- Repository refresh (barber pole)
- Background packet actions, like preparation of install or uninstall
(barber pole)
- Package downloads, including downloads of dependencies (progress
bar). Status text indicates the name of the package currently
being downloaded (if any), and how many more packages are queued
for download after it (if any).
* Hooks into PackageListView to be notified of package status changes
(such as becoming pending or download progress)
* When the package currently being downloaded is also selected in the
list view, the user sees the progress bar in WorkStatusView
as well as the one in the PackageInfoView, which is redundant. This
still needs a good solution...
* Add a new event "ConfirmedChanges" to PackageProgressListener. It
triggers when a package action was confirmed by the user and is
about to be run. This can be used to e.g. get a full list of
packages about to be installed (i.e. including dependencies)
right before the process of fetching/installing them is started.
* Add a handler for it which looks at PackagesToActivate and sets
all their PackageInfo states to pending (so now, when installing
a package, its dependencies immediately become pending as well).
* When a view is used with the layout system, its initial frame rect
was set to (0, 0, 0, 0), which is a BRect covering 1 pixel in the
top left corner of the window.
Since this a valid rect, it can cause "badly behaved" views to
trigger redraws of themselves and other views during the layout
process, which is ultimately the reason for the HaikuDepot UI
freezing while populating with packages.
The misbehaving view in this case is BTextView. When in read-only
mode, since commit e27a53b2, its GetHeightForWidth() implementation
causes the view to resize (really resizing, not just simulating a
resize) and thus it invalidates itself. This is broken behaviour,
and needs to be fixed in BTextView. Since GetHeightForWidth() is
called during the layout process, all the not-yet-layouted views
have a frame of (0, 0, 0, 0). The invalidation of just the one
BTextView in the layout then hits *all* new views that are being
layouted (because they all occupy the same one pixel in the
corner), and they all get redrawn.
Many view Draw() implementations ignore the update rect, so work
is being done. And even if not, this can cause a lot of traffic
on the app_server link. In a test case with HaikuDepot's
FeaturedPackagesView, adding 300 rows (each containing a BTextView,
among other views) in quick succession caused over 6 million
commands to travel over the app_server link, completely freezing
the UI for a long time.
* The actual problem here is in BTextView::GetHeightForWidth() and
must be fixed there.
However we also put in an extra-fix here because it never makes
sense anyway to try and draw a view that has not yet been layouted.
So we set the initial BView frame to an invalid rect
(0, 0, -1, -1), which will suppress any actual updating, even
when the view actively invalidates itself, as long it doesn't
have a size yet. (The dirty region will always end up empty
then).
* Fixes HaikuDepot UI freezing during package population (caused by
above described behaviour from BTextViews in FeaturedPackagesView).
Might improve performance in other applications using BTextView
with layouting as well.
* Fix use-after-free in the auto-generated JSON listener code (was
calling method of instance after 'delete this')
Will also submit a patch for the code generation script in the
haikudepotserver repo so it won't get lost -- this is just a
temporary fix-up until then.
* Fixes some random occasional crashes
* BarberPole view for indicating that the application is busy while
the amount of work is not known
* Flat looks with scrolling color stripes. Number of stripes and their
colors are freely configurable, so you could add a whole rainbow of
colors to endlessly scroll through if you like.
* When SystemProfiler::_MaybeNotifyProfilerThreadLocked() is called
and the conditions are right, it will lock the thread's scheduler
spinlock and unblock it. Internally, the unblock will enqueue the
thread into the run queue, which causes a ThreadEnqueuedInRunQueue
event for SystemProfiler. Since the conditions haven't changed, it
now went into _MaybeNotifyProfilerThreadLocked again (this time
from the profiler thread context). In there, it will try to lock
the profiler thread's scheduling spinlock, which is already locked
by the other thread (which is firmly sleeping). Deadlock, KDL.
* Before unblocking the profiler thread, unset fWaitingProfilerThread
so that further events will not try to unblock it again.
