- Store whether or not the use of the horizontal scrollbar is desired
on the class itself. If the CLV was set to use the horizontal scrollbar,
and then asked to lay itself out while hidden, it would incorrectly assume
the horizontal scrollbar wasn't in use, and consequently repositioned its
views such that the horizontal scrollbar and outline view overlapped.
- The cdrecord port was using the wrong path for searching for SCSI
devices. This led to it failing to find SATA CD drives. Updated
package by Chris Roberts.
- When a job needs to go dependent on another job, we no longer use recursion
to manage the execution stack. Instead the job is simply marked as waiting
and we execute other jobs with no dependencies in the meantime. When a job
completes, all dependents are moved back onto the unscheduled list and
executed as needed.
- Adjustments to ResolveValueNodeJob to handle the now asynchronous nature
of waiting.
- If the node we're looking at is a compound node, retrieve its location
and show that as the value with an indicator to clarify that it's an object.
- Minor tweaks to tooltip format.
- Style cleanups.
- When resolving its value, CStringValueNode now sets its node child's
address to the address of the string buffer rather than the location of
the originating pointer, which allows things like Inspect to pick that up.
- TableCellContextMenuTracker now supports menus that don't have a settings
submenu, since some variables won't have renderer settings but will still
have context actions.
- Add _GetContextActionsForNode() to retrieve the list of contextual actions
available for a given model node. Currently this is only adds an action
to inspect the memory address of the highlighted value, but will be extended
for other actions later.
- If we're asked to generate an entry for a tag we don't recognize,
return an error so don't then attempt to add a null/invalid entry into the
compilation unit's entry list and later crash dereferencing it.
- When asking for signed/unsigned, IntegerFormatter was previously
treating all value types as int64, which would result in much larger
than expected values in the variable list in some cases. Inspect the actual
integer type of the variable and adjust the format string accordingly in
order to deal with that.
- If a string column was exactly the correct pixel width for the strings
contained within it, they would skip attempting to truncate the string.
However, the truncated string was always used for final drawing, with
the end result that the fields would be drawn blank. This would sometimes
manifest itself in Debugger where things like ID columns wound wind up
showing no data until one resized the column.
- When binary searching functions in the source entry list,
comparing by name and location alone isn't sufficient, since
templates will match those for different instances, Fixes a crash on
exit where the wrong function would get removed from the list, while
the one we actually wanted to remove was still in the list, but then
had its source code cleared. This would later crash the comparison
function due to not being able to get its source location.
- When an image was unloaded, its corresponding image info was never
removed from TeamDebugInfo's list, leading to the latter containing
a deleted object, resulting in various random crashes.
- When attempting to unwind the call frame, we now search for the appropriate
FDE in both .debug_frame and .eh_frame. This mirrors gdb's behavior and
works around the ever-changing whims of the gcc developers as to which
section the requisite FDE/CIE resides in.
- The buffer that the debugger used to retrieve messages from
the debug port was slightly too small for the largest of the message
data structs (currently 1100 bytes), causing some types of debug events
to get truncated. This resulted in image creation/deletion events being
received with a truncated image_info struct, which would result in several
fields being returned with random values, most notably the text/data base
and size fields. Consequently, searching those images for an address within
them would fail, leading to #8709. It's possible but not yet confirmed
that this bug is also responsible for #8710, need to test further.
- The version of instantiate_object() that returns the image id from which
the object was reinstantiated was not correctly returning it in all
circumstances, only if it had to find/load the add-on itself. This caused
problems for BShelf since the latter relies on that image id in order to
determine what image to unload when replicants are removed. To remedy this,
introduce an additional version of find_instantiation_func() that returns
the image id in which the instantiation function was found, and make use
of it in instantiate_object() in order to also be able to return the image
id in that case.
- repnz movsb turns out to not actually be a legal instruction,
resulting in various strings being copied incorrectly, leading to
random crashes in various places. Rework to use loop instead.
Thanks to Alex Smith for helping review changes and offering
improvements.
- Minor cleanups.
- Fixes#8650 properly.
- Replace arch_cpu_user_strlcpy() and arch_cpu_user_memset() with x86 assembly
versions. These correctly handle the fault handler, which had broken again
on gcc4 for the C versions, causing stack corruption in certain error cases.
The other architectures will still need to have corresponding asm variants
added in order for them to not run into the same issue though.
- Automatic whitespace cleanup.
- Remove no longer exposed show/hide time setting. The methodology for
saving it on exit was broken anyways, since it relied on the current
show/hide state of the time view, which would be hidden if the deskbar
was currently in Autohide mode.
Our more recent build of gcc4 appears to have switched to using
.eh_frame for almost all useful call frame information when built
with debugging. Use a somewhat crude heuristic (size) to determine
if the .debug_frame section we've been given might actually be of use
or not (assuming it exists at all, this was inconsistent in my tests.
Sometimes apps had no .debug_frame at all, other times it was present
but was only roundabouts 100 bytes).
Fixes ticket #8508.
Factor out updating of the source path view into a dedicated function,
and fix some errors that would sometimes result in the text not updating
properly when switching stack frames, particularly if the target frame
didn't have source code available.
- Conditionally enable and disable event masks as needed when we detect
the mouse entering/exiting the deskbar area. This is necessary because
otherwise the app_server appears to eat mouse messages due to the ctrl+cmd+
resize shortcut. Also change some conditionals slightly to better deal with
differing combinations of autohide and/or autoraise.
- Rework Deskbar's handling for the autoraise/autohide cases to not
need event masks. This prevents various problems that would occur
if one happened to be working in another window that overlapped part
of Deskbar, as reported in #8497. Instead, we now use a message filter
to redirect the messages as needed.
Redo the precision levels of various calculations in the memory bar items
so they don't overflow on systems with > 4GB of RAM. Previously one could
see fun results like the kernel using negative amounts of memory on such
systems.
Set an explicit minimum size on the source path view, in order to prevent it
from unduly influencing the initial size constraints of the window. (Belatedly)
fixes the remaining part of #8094.
- SetTopTap() -> SetTopTab().
- When reloading the decor, if the window in question is borderless,
short circuit. Previously, we wouldn't allocate a decorator for such
a window, but would then go through the remaining steps of attempting
to set the focus/top tab, which would obviously crash. Fixes#8500 and
possibly others.
- Add error check that adding tabs to the decorator actually succeeded.
When asked to build via a relative path (i.e. from within a generated subdir),
gcc2 was generating a compilation directory attribute pointing to the generated
subdirectory, as well as entering the actual filename attribute as the full
absolute path to the file. This wasn't checked for, leading to the issue
described in ticket #8498. gcc4 does not appear to have the same behavior,
it generates a relative path + filename in all cases.
NetworkStatus now tracks the status of each interface separately, and
uses that information to make correct comparisons for notifications.
Previously, it would only send notifications using whatever interface
was last in the enumeration list, and if an earlier interface had a higher
configuration level than that, the notification would incorrectly be sent
as occurring on the last interface. We now send notifications for each
interface separately. The status shown by the deskbar icon remains that
of the most highly configured detected interface. Fixes#8468.
AutoConfigLooper now remembers the last media state it received from the device,
and compares both to see if the active state actually changed. Should help with
quality changes where the link didn't actually drop. May also help on ethernet
if the chipset sends spurious link status change notifications.
- Add accessor to PoseView that reports whether the view is currently
typeahead filtering.
- Use said accessor in the file panel's filter in order to detect whether
to tell the latter to cancel filtering vs closing the panel.
Fixes#8140.
If the refresh limits only allowed a single refresh, the menu item constructed
had no BMessage. This caused _UpdateRefreshControls() to crash when attempting
to match the current refresh rate to the item. For simplicity, give it a message
containing the fixed refresh rate just as in all other cases. Should finally
fix#8431.
- Fixes a problem in Web+'s autocompletion handler that would
result in it eating things like workspace switching shortcuts
if Web+'s URL input had focus.
- Always initialize other menu item data member.
- Always use refresh_rate_to_string() to generate item labels.
- When searching for the item matching the selected refresh rate,
look for it by the actual refresh rate stored in the item's message
rather than by string comparisons of the label.
- Don't assume the Other menu item exists, since it doesn't in cases
where the refresh rate constraints only allow a single rate.
- Update copyright year.
Fixes#8431.
Previously, there were two instances of the actual variables, named in
ControlLook.cpp and Utilities.cpp. This caused clashes on gcc4 builds, leading
to various crashes during initialization. Instead we now declare them as
static constants in the header itself as is done for various other interface
color constants elsewhere. Fixes ticket #8358.
- Architecture::CreateStackTrace() now uses the last frame's
PreviousCpuState() as the basis to continue unwinding when passed
a partial trace to continue from, rather than the (incorrect) actual
cpu state of that frame, which would have resulted in the last frame
being duplicated in the trace.
- Renamed variables to be more clear.
Used to store the unwound state of the CPU from this frame,
which will form the actual state of the following frame. Needed
in order to properly support continuation of unwinding from a
partial stack trace.
- Limit invocations of CreateStackTrace() in ThreadHandler to
only unwind the topmost frame since that's all it actually
cares about anyways. Also adjust Step Over to use this
functionality in order to work with the correct frame addresses,
since the CPU frame pointer register isn't entirely what we want
here.
CreateStackTrace() can now optionally be asked to limit the maximum
number of frames it tries to unwind. In conjunction, it can also be
passed an already existing partial stack trace, and be asked to
unwind more frames from it.
When stepping over, check to ensure that we're in the same call
frame that we started in. If not, reinstate the temporary breakpoint
and continue running.
Will be used later in order to determine if a given call obeys the
standard calling convention, in order to determine if we can safely
attempt to retrieve its return value.
When asked to step out, we now store the frame pointer of the current
stack frame. Upon hitting our temporary breakpoint, we then verify that
the current frame pointer is in fact in the previous frame. If not, we
reinstate the breakpoint and continue running.
Rene Gollent