Öffentliche Aktivitäten der HicknHack Software GmbH

initial commit
added a proper readme
* fixed a set of missing braces wrecking stuff
added a proper travis config
made the travis ci button show the correct branch
added rubycritic
* improved a discovered smell
added rspec and spec directory
* removed minitest and the test directory
fixed factories path
simplified some date time calculations
* adjusted time log booking accordingly
added a invalid interval exception for fit_in_bounds
* changed order of arguments to fit the interval spirit
use proper exceptions instead of just text messages
timediff now returns always a positive
round_interval now expects an actual interval instead of start and stop
small refactoring
added a semi-proper time infinity
added an eql? method to the infinity class
removed the delta arguments from the limits_from_overlapping methods method
* fixed it not working without delta
added first tests
added a exception while checking the limits if a record is inside the interval in question
* fixed tests for limits_from_overlapping_interval
fixed support load path
* fixed chronos not loading redmine patches
* added chronos requirement to test
* added a proper factory for timelog
changed patches relying on chronos namespace

Andreas Reischuck

Github Push to arBmind/redmine_hourglass

Bugfix: prevent multiple dialog templates from using same html element id
fix asset hot reloading during development
improve form handling a bit (still messy)

Andreas Reischuck

Github Push to arBmind/redmine_hourglass

Feature: improve ticket completion filters to ignore #

Andreas Reischuck

Github Push to arBmind/redmine_hourglass

Bugfix: Do not allow project and issue to diverge
Feature: improve ticket completion filters to ignore #

Andreas Reischuck

Github Push to arBmind/redmine_hourglass

Feature: improve ticket completion filters to ignore #

Andreas Reischuck

Github Push to arBmind/redmine_hourglass

Bugfix: Do not allow project and issue to diverge

Andreas Reischuck

Github Push to arBmind/redmine_hourglass

improve form handling a bit (still messy)

Andreas Reischuck

Github Push to arBmind/redmine_hourglass

fix asset hot reloading during development

Andreas Reischuck

Github Push to arBmind/redmine_hourglass

Bugfix: prevent multiple dialog templates from using same html element id

Andreas Reischuck

Github Push to arBmind/redmine_hourglass

Feature: improve ticket completion filters to ignore #

Andreas Reischuck

Github Push to arBmind/redmine_hourglass

Bugfix: Do not allow project and issue to diverge

Andreas Reischuck

Github Push to arBmind/redmine_hourglass

fix failing tests (4)
* tracker update responds with json content instead of no content
* avoid using tracker description which is not in Redmine 4.2.9

Andreas Reischuck

Github Push to arBmind/redmine_hourglass

fix failing tests (3)
* tracker update responds with json content instead of no content
* avoid using tracker description which is not in Redmine 4.2.9

Andreas Reischuck

Github Push to arBmind/redmine_hourglass

fix failing tests
* tracker update responds with json content instead of no content
* avoid using tracker description which is not in Redmine 4.2.9

Andreas Reischuck

Github Push to arBmind/redmine_hourglass

fix failing tests
* tracker update responds with json content instead of no content
* avoid using tracker description which is not in Redmine 4.2.9

Andreas Reischuck

Github Push to arBmind/redmine_hourglass

fix failing tests because update now responds with time tracker instead of no content

Andreas Reischuck

Github Push to arBmind/redmine_hourglass

Bugfix: Do not allow project and issue to diverge
Bugfix: prevent multiple dialog templates from using same html element id
fix asset hot reloading during development
improve form handling a bit (still messy)

Andreas Reischuck

Github Push to hicknhack-software/Qt-Creator

Android: Fetch latest command line tools for "Set Up SDK"
Don't install outdated ones.
https: //developer.android.com/studio#command-line-tools-only
Change-Id: Ifd9c6a2e90a63c6113c859910365689f19b3a532
Reviewed-by: Jarek Kobus
TextEditor: Optimize unindent backspace behavior
Only remove more than one character when the cursor is inside the
indentation in the beginning of the line. In all other cases always only
remove one character.
Fixes: QTCREATORBUG-30725
Change-Id: I973101a95768cdd8b1a318972f53423eb72eb157
Reviewed-by: Christian Stenger
Reviewed-by: André Hartmann
Reviewed-by:
Android: Give Android 14.0 the code name
The name "UpsideDownCake" for Android 14.0 has been revealed a while
ago. Align its package presentation to Android Studio.
Task-number: QTCREATORBUG-30711
Change-Id: I7ca42ca73c7bdf5b7732bab2d2bd36f7a730ccb7
Reviewed-by: Jarek Kobus
TextEditor: Fix calculating block highlight background color
Fixes: QTCREATORBUG-30649
Change-Id: I9a213ecdf4d58ed6531014c99bbdedac8ac9ef00
Reviewed-by: Christian Stenger
SquishTests: Autodetect cdb
Creator automatically detects the debuggers anyway and then can't handle
the fixed ID.
Change-Id: I732f86649b02016be2ca513c896a80ce254d721f
Reviewed-by: Christian Stenger
Reviewed-by:
SquishTests: Avoid using occurrence values for OutputPaneToggleButtons
The occurrence count can't be understood without looking at
the running application and can even change at runtime.
Change-Id: I34677fd2fe2a0a8a7f2cdcd9f37ec938efe1a5ea
Reviewed-by:
Reviewed-by: Christian Stenger
SquishTests: Avoid using occurrence value for IconButton
Change-Id: I0a4d9497801dbd39c2c735cac760c5dc6a0298fd
Reviewed-by:
Reviewed-by: Christian Stenger
SquishTests: Avoid using occurrence value for SideDiffEditorWidgets
Change-Id: I7c519347a25b0c541b019dd1cae65df4a74c8826
Reviewed-by: Christian Stenger
Reviewed-by:
SquishTests: Remove duplicate object
Change-Id: I7ceed8ef40684323b8404c3fe7c529fbd5b7be74
Reviewed-by: Robert Löhning
SquishTests: Make sure that button in Designer is renamed properly
This misused typeLines() to close the dialog by pressing "Return".
Although the entire name of the button should be selected at first,
I witnessed the new name being appended to the old one.
The new code uses replaceEditorContent() to explicit select the old
name and replace it with the new name. After that, it explicitly
presses "Return" to close the dialog.
Change-Id: I5ec946f57000396b70428e08ab795a1ee3eba91a
Reviewed-by:
Reviewed-by: Christian Stenger
Add an environment variable for disabling atomic save
There are situations where e.g. anti virus software leads to saving files
failing. Provide a way to disable atomic save for debugging purposes.
Task-number: QTCREATORBUG-30728
Change-Id: I1f3df009f89f6c27b2e0c9f91869b2a75f5575d8
Reviewed-by: Orgad Shaneh
TextEditor: fix backspace on block start
Amends ec88d279a8d52189685008c8fe46371d13d2505f
Change-Id: I7d83b5ffa617f8d45acc6998ce674b1afcf3a31b
Reviewed-by: Christian Stenger
LanguageClient: fix compile with latest Qt dev
Change-Id: Ifc6b5cfacb1e4e559646304b4ab5889eeb85bf3f
Reviewed-by: hjk
Debugger: Execute additional attach commands for LLDB
This allows setting _Preferences > Debugger > GDB > Additional Attach
Commands_ with the content:
process handle SIGSEGV --pass true --stop false --notify true
which results in the debugger not stopping in assembler for Java runtime
exceptions.
Task-number: QTCREATORBUG-29928
Change-Id: I8edf0d79dcccfe7ddb27502edb122fb5fb2c646c
Reviewed-by: Axel Spoerl
Reviewed-by: hjk
QmlJSEditor: Avoid crash on temporary documents
Fixes: QTCREATORBUG-30739
Change-Id: I62872fc23529674d5fca6dcb6255872eab12a9da
Reviewed-by: David Schulz
QtSupport: Fix Qt Abi detection after Android initial SDK setup
If QtVersion::qtAbis() is called while the Android SDK is not yet set
up, AndroidQtVersion::detectQtAbis() returns an empty Abi list. After
the Android SDK setup, the Abi detection is repeated.
The recently introduced use of std::optional for Abis prevented that
"second chance" for Abis detection. After the Android SDK setup, Qt
Creator now had to first be restarted in order to have functioning
Android Qt Kits.
This change adds an Android specific hack in QtVersion::qtAbis() which
gives AndroidQtVersion a "second chance" to detect Qt Abis.
Amends: ee4f6ca1ae06ea932789fb41fdf8b9427ad27478
Fixes: QTCREATORBUG-30568
Change-Id: I5d8a4fa7081d80015375c76305bb559d890b5f79
Reviewed-by:
Reviewed-by: Christian Kandeler
Update qbs submodule to HEAD of 2.3 branch
Change-Id: I032c554de5920d9a44d8f6ca544ba3cb61fa7bae
Reviewed-by: Christian Stenger
Reviewed-by:
Examples: Re-add the clear button for the search bar
when something is typed there. Got lost during redesign.
Change-Id: I6a765aa542243d2e20e00f6e965dbfbd26a12803
Reviewed-by: Alessandro Portale
Reviewed-by:
CMakePM: Fix crash with specific CMakePresets.json
The code was not taking into consideration an empty QStringList.
Fixes: QTCREATORBUG-30755
Change-Id: Ifb9f5dc3ffefd94d44b16b161bc04d381f3f6786
Reviewed-by: Alessandro Portale
Terminal: Fix TERM env variable
Some (Fedora) OS/Distros set a default value for TERM such as "dumb", which then breaks
command line tools such as "clear" which try to figure out what terminal they are
running in. Therefore we have to force-set our own TERM value.
Fixes: QTCREATORBUG-30737
Change-Id: I50ac86f17c92dc7bc477cc59ed43ec22025593fb
Reviewed-by: Christian Stenger

Andreas Reischuck

Github Push to hicknhack-software/Qt-Creator

EffectComposer: Toggle "Assign to.." button based on component selection
Task-number: QDS-12148
Change-Id: I1f44df7c6c027c36c09f7c4d74e4dd609542c5b8
Reviewed-by: Qt CI Patch Build Bot
Reviewed-by: Mahmoud Badri
QmlDesigner: Fix assert
Url properties should be set as urls to avoid asserts.
Change-Id: I990e731f4555943ce35443226ab053286c58bd58
Reviewed-by: Mahmoud Badri
Reviewed-by: Qt CI Patch Build Bot
QmlDesigner: Hide Flags combobox's popup when opening the cog menu
Otherwise the cog menu appears below the popup
Change-Id: If3b541bd98b2d771687373c9259ad017349d4215
Reviewed-by: Miikka Heikkinen
QmlDesigner: Hotfix keep qml cache enabled
Task-number: QDS-12264
Change-Id: I4fb58331141c72eedd74b3d7ac8fd8462ae3158f
Reviewed-by: Tim Jenssen
Nanotrace: Support enumeration arguments
Change-Id: I75f7576f28231fc35269cbb3bb46654739cf7f18
Reviewed-by: Qt CI Patch Build Bot
Reviewed-by: Tim Jenssen
Reviewed-by:
NanoTrace: Add missing category constructor
The constructor for the category was missing. Otherwise the CTAD is not
working.
Change-Id: I4d4c2aaba8915fbbade0b8ce4f1310046d748cd1
Reviewed-by: Thomas Hartmann
Reviewed-by: Tim Jenssen
Reviewed-by: Qt CI Patch Build Bot
Sqlite: Add convertToString to Sqlite id
Change-Id: I52d9b6860d24d715fdd6b584ffc2f8345f6b99c4
Reviewed-by: Tim Jenssen
Reviewed-by:
Reviewed-by: Qt CI Patch Build Bot
Nanotrace: If the process is terminated flush the event queues
That terminate handler is only installed if any event queue is
activated. That is only the case if you compile with a special
option.
Change-Id: I4d8661c0fa11d8bb14b49b7b963471f626b3b950
Reviewed-by: Tim Jenssen
Reviewed-by: Qt CI Patch Build Bot
QmlDesigner: Avoid dangling pointer to the stack
Pointer to temporaries create dangling pointer. So we have to copy the
tuples. The leafes still only contain references.
Change-Id: I37900be569283324178e223829f913fb4ccba16e
Reviewed-by: Tim Jenssen
Reviewed-by: Qt CI Patch Build Bot
QmlDesigner: avoid extra widgetInfo() call
widgetInfo() should be cheap but people make expensive stuff inside the
call. So we should try to avoid to many calls.
Change-Id: I66b0a5431471732dca934fc03efb307e2617fd65
Reviewed-by: Qt CI Patch Build Bot
Reviewed-by: Thomas Hartmann
Reviewed-by: Tim Jenssen
QmlDesigner: Remove unneeded break
Change-Id: I051bf442870ce1ec429565bcbf15068d22eb16a9
Reviewed-by: Qt CI Patch Build Bot
Reviewed-by: Tim Jenssen
QmlDesigner: Add lite designer to qmldesigner project management
Task-number: QDS-12102
Change-Id: I3abea3a5b32128db984a9b0c49b3044a42d9f72d
Reviewed-by: Qt CI Patch Build Bot
Reviewed-by: Tim Jenssen
Nanotrace: Fix dangling tuple
The tuple are constructed on the stack, so we cannot hold references to
it.
Change-Id: Ia02bf21516c97eb67eca8fd56e68dbd10878ca9e
Reviewed-by: Qt CI Patch Build Bot
Reviewed-by: Tim Jenssen
Nanotrace: Fix dangling reference
As EventQueueData was derived from EventQueue the order of
initialization was inverted. So we got dangling pointer. Merging
EventQueueData into EventQueue fixes the dangling references.
Change-Id: I3d9cee492304132fa612d6d19324656df597ec99
Reviewed-by: Tim Jenssen
Reviewed-by: Qt CI Patch Build Bot
Reviewed-by:
QmlDesigner: Fix missing BreakTemplateDeclarations
Change-Id: I00c042e5e0a04705e3c2773dca5499d9bfa32949
Reviewed-by: Qt CI Patch Build Bot
Reviewed-by: Tim Jenssen
Reviewed-by:
Nanotrace: Use begin and end slices instead of duraction slice
If an exception is thrown or the program is abortet there is still the
begin of a slice. Because a duration slice is only wirtten at the end we
get no slice.
Change-Id: I8b05e9ae045517a503c2a6f00936b05715b812ca
Reviewed-by: Qt CI Patch Build Bot
Reviewed-by: Tim Jenssen
Nanotrace: Fix unsed arguments
The arguments must be set because we otherwise access old data.
Change-Id: I18e83297b80558b112af266bd50779bef8da9e94
Reviewed-by: Qt CI Patch Build Bot
Reviewed-by: Tim Jenssen
Reviewed-by:
Doc: Auto-generate navigation links based on "All Topics"
This does not change what you see as values of \previouspage
and \nextpage in the source files. However, when you build the
docs, QDoc replaces those values with links to the next and
previous page from the list in qtdesignstudio-toc.qdoc because
that page contains "\title All Topics".
So, with this change, you need to keep the list in "All Topics"
up to date.
Change-Id: I0996e383cbd60b7f4297524e4141a2d7af6fae13
Reviewed-by: Qt CI Patch Build Bot
Reviewed-by:
Reviewed-by: Mats Honkamaa
QmlDesigner: Fix fly mode camera rotation direction
Flip x-rotation direction when camera is upside down to keep camera
rotation natural.
Fixes: QDS-12186
Change-Id: Ia9f166cafa27a09a4dd5a72dbb8ec28f00239564
Reviewed-by:
Reviewed-by: Qt CI Patch Build Bot
Reviewed-by: Mahmoud Badri
QmlDesigner: Workaround issues with QJSValue in Qt 6.7.x
This forks code from Qt to fix an issue with JSValue.
We must skip QJSValues when iterating properties
since copying or constructing the QVariant crashes.
Task-number: QDS-12263
Change-Id: I8db701cee94dccd7fe254c9473dc25b87f8a04f8
Reviewed-by: Tim Jenssen

Andreas Reischuck

Github Push to hicknhack-software/Qt-Qbs

Loader: Fix look-up of qbs properties from provider probes
... by setting the provider scope on the values instead of the item.
Task-number: QBS-1742
Change-Id: I8084e3f8529335c3d045c057ba70a3c3b091efa6
Reviewed-by: Ivan Komissarov
GitHub actions: fix macOS 14.2 job
...by bumping minimum macOS version.
Change-Id: I3097cacb6782d262e623bb16851f5d5cec616ee0
Reviewed-by: Christian Kandeler
Qt: fix minimum macOS/iOS versions for Qt 6.7.1
Change-Id: I3338a3b6dbb669abab63f94c5ed9e9440a4de04d
Reviewed-by: Christian Kandeler
GitHub actions: fix doc job
Change-Id: I9d462f5a960cf79ad6ffa8a74689a6e3e27d6425
Reviewed-by: Christian Kandeler
doc: fixes for PkgConfigProbe
Change-Id: Ia8c0c51fad0a33d2e870c71d45868242e3a12238
Reviewed-by: Christian Kandeler
Update changelog
Change-Id: Icedbd322ffd8bd6fbc69dde2ff27ff93d3308273
Reviewed-by: Christian Kandeler
Do not produce project dependencies for Visual Studio projects.
Visual Studio calls qbs which resolves all dependencies on it's own.
remove buggy msbuild filters
They generate a lot of useless folders, that are totally annoying.
It would be nice to use filters to represent the subfolder structure in visual studio, but this would be a full rewrite anyways.
prevent the plugins path from doubling up. (feels hacky, but works for now)
There are two call sites. If the path comes from the build environment of QtCreator the path is wong.
enable support for c++2a and c++2b on msvc
added AdditionalOptions to msbuild project
* adds /std option to improve msvc auto completion
added Github workflow
* added release creation
added moc includes to speed up compilation
add support for object targets
* extended cpp module compilers to link object files from dependencies
drop unused member
clang_tidy allow suspicious includes for moc_ based inculde
added support for extra qt install dir

dependabot[bot]

Github Push to hicknhack-software/Qt-llvm-project

dependabot[bot]

Github Push to hicknhack-software/Qt-llvm-project

dependabot[bot]

Github Push to hicknhack-software/Qt-llvm-project

dependabot[bot]

Github Push to hicknhack-software/Qt-llvm-project

[compiler-rt] making getrandom call blocking. (#78340)
except when `GRND_NONBLOCK` is present in the flags.
[libc++] Fix vector_aligned_tag (#76611)
Co-authored-by: Nikolas Klauser
[Clang] Fix dependency of SourceLocExpr. (#78436)
SourceLocExpr that may produce a function name are marked dependent so that the non-instantiated
name of a function does not get evaluated.
In GH78128, the name('s size) is used as
template argument to a `DeclRef` that is not otherwise dependent, and therefore cached and not transformed when the function is
instantiated, leading to 2 different values existing at the same time for the same function.
Fixes #78128
[gn] attempt to port 8dfc67d6724e (__assertion_handler)
[clangd] Handle an expanded token range that ends in the `eof` token in TokenBuffer::spelledForExpanded() (#78092)
Such ranges can legitimately arise in the case of invalid code, such as
a declaration missing an ending brace.

Fixes https://github.com/clangd/clangd/issues/1559
[CodeGen] Support start/stop in CodeGenPassBuilder (#70912)
Add `-start/stop-before/after` support for CodeGenPassBuilder.
Part of #69879.
[Path] Fix off-by-one in finding filename for win style paths (#78055)
This fixes a crash where `path::parent_path` causes an invalid access on
a string upon receiving a path that consists of a single colon.

On Windows machine, with runtime checks enabled build, upon `clang -I:
test.cc` produces:
```
Assertion failed: Index < Length && "Invalid index!", file llvm\include\llvm/ADT/StringRef.h, line 232
...
#6 0x00007ff7816201eb `anonymous namespace'::parent_path_end llvm\lib\Support\Path.cpp:144:0
#7 0x00007ff781620135 llvm::sys::path::parent_path(class llvm::StringRef, enum llvm::sys::path::Style) llvm\lib\Support\Path.cpp:470:0
```

Ideally, we can look for the last colon starting from the last
character, but we can instead start from second to last, and handle
empty paths by abusing `0 - 1 == npos`.
[Clang] NFC: Move Arm type attributes to separate trailing object. (#78424)
This decouples the Arm type attributes from other bits, which means
the data will only be allocated when a function uses these Arm
attributes.

The first patch adds the bit `HasArmTypeAttributes` to
`FunctionTypeBitfields`, which grows from 62 bits to 63 bits.

In the second patch, I've moved this bit (`HasArmTypeAttributes`) to
`FunctionTypeExtraBitfields`, because it looks like the bits in
`FunctionTypeBitfields` are precious and we really don't want that
struct
to grow beyond 64 bits.

I've split this out into two patches to explain the rationale, but those
can be squashed before merging.
DAG: Fix chain mismanagement in SoftenFloatRes_FP_EXTEND (#74558)
[AsmPrinter] Fix gcc -Wparentheses warning [NFC]
Without this gcc warned
../lib/CodeGen/AsmPrinter/DwarfDebug.cpp:3585:70: warning: suggest parentheses around '&&' within '||' [-Wparentheses]
3584 | ((&Current == &AccelDebugNames) &&
| ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
3585 | (Unit.getUnitDie().getTag() != dwarf::DW_TAG_type_unit)) &&
| ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^~
3586 | "Kind is CU but TU is being processed.");
| ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
../lib/CodeGen/AsmPrinter/DwarfDebug.cpp:3589:70: warning: suggest parentheses around '&&' within '||' [-Wparentheses]
3588 | ((&Current == &AccelTypeUnitsDebugNames) &&
| ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
3589 | (Unit.getUnitDie().getTag() == dwarf::DW_TAG_type_unit)) &&
| ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^~
3590 | "Kind is TU but CU is being processed.");
| ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
[AMDGPU] Use alias info to relax waitcounts for LDS DMA (#74537)
LDA DMA loads increase VMCNT and a load from the LDS stored must wait on
this counter to only read memory after it is written. Wait count
insertion pass does not track memory dependencies, it tracks register
dependencies. To model the LDS dependency a pseudo register is used in
the scoreboard, acting like if LDS DMA writes it and LDS load reads it.

This patch adds 8 more pseudo registers to use for independent LDS
locations if we can prove they are disjoint using alias analysis.

Fixes: SWDEV-433427
[AMDGPU][NFC] Rename feature FP8Insts to FP8ConversionInsts (#78439)
[clang][dataflow] Use `Formula::isLiteral()` in a couple more places. (#78404)
[clang][dataflow] Consider `CXXDefaultInitExpr` to be an "original record ctor". (#78423)
The CFG doesn't contain a CFGElement for the
`CXXDefaultInitExpr::getInit()`, so
it makes sense to consider the `CXXDefaultInitExpr` to be the expression
that
originally constructs the object.
[clang][ASTImporter] Improve structural equivalence of overloadable operators. (#72242)
Operators that are overloadable may be parsed as `CXXOperatorCallExpr`
or as `UnaryOperator` (or `BinaryOperator`). This depends on the context
and can be different if a similar construct is imported into an existing
AST. The two "forms" of the operator call AST nodes should be detected
as equivalent to allow AST import of these cases.

This fix has probably other consequences because if a structure is
imported that has `CXXOperatorCallExpr` into an AST with an existing
similar structure that has `UnaryOperator` (or binary), the additional
data in the `CXXOperatorCallExpr` node is lost at the import (because
the existing node will be used). I am not sure if this can cause
problems.
[LV][AArch64] LoopVectorizer allows scalable frem instructions (#76247)
LoopVectorizer is aware when a target can replace a scalable frem
instruction with a vector library call for a given VF and it returns the
relevant cost. Otherwise, it returns an invalid cost (as previously).

Add test that check costs on AArch64, when there is no vector library
available and when there is (with and without tail-folding).

NOTE: Invoking CostModel directly (not through LV) would still return
invalid costs.
[AMDGPU] Add mark last scratch load pass (#75512)
[flang] Allow user to define free via BIND(C) (#78428)
A user defining and using free/malloc via BIND(C) would previously cause
flang to crash when generating LLVM IR with error "redefinition of
symbol named 'free'". This was caused by flang codegen not expecting to
find a mlir::func::FuncOp definition of these function and emitting a
new mlir::LLVM::FuncOp that later conflicted when translating the
mlir::func::FuncOp.
[AMDGPU][True16] Support V_FLOOR_F16. (#78446)
[gn build] Port 1d286ad59b90

dependabot[bot]

Github Push to hicknhack-software/Qt-llvm-project

[compiler-rt] making getrandom call blocking. (#78340)
except when `GRND_NONBLOCK` is present in the flags.
[libc++] Fix vector_aligned_tag (#76611)
Co-authored-by: Nikolas Klauser
[Clang] Fix dependency of SourceLocExpr. (#78436)
SourceLocExpr that may produce a function name are marked dependent so that the non-instantiated
name of a function does not get evaluated.
In GH78128, the name('s size) is used as
template argument to a `DeclRef` that is not otherwise dependent, and therefore cached and not transformed when the function is
instantiated, leading to 2 different values existing at the same time for the same function.
Fixes #78128
[gn] attempt to port 8dfc67d6724e (__assertion_handler)
[clangd] Handle an expanded token range that ends in the `eof` token in TokenBuffer::spelledForExpanded() (#78092)
Such ranges can legitimately arise in the case of invalid code, such as
a declaration missing an ending brace.

Fixes https://github.com/clangd/clangd/issues/1559
[CodeGen] Support start/stop in CodeGenPassBuilder (#70912)
Add `-start/stop-before/after` support for CodeGenPassBuilder.
Part of #69879.
[Path] Fix off-by-one in finding filename for win style paths (#78055)
This fixes a crash where `path::parent_path` causes an invalid access on
a string upon receiving a path that consists of a single colon.

On Windows machine, with runtime checks enabled build, upon `clang -I:
test.cc` produces:
```
Assertion failed: Index < Length && "Invalid index!", file llvm\include\llvm/ADT/StringRef.h, line 232
...
#6 0x00007ff7816201eb `anonymous namespace'::parent_path_end llvm\lib\Support\Path.cpp:144:0
#7 0x00007ff781620135 llvm::sys::path::parent_path(class llvm::StringRef, enum llvm::sys::path::Style) llvm\lib\Support\Path.cpp:470:0
```

Ideally, we can look for the last colon starting from the last
character, but we can instead start from second to last, and handle
empty paths by abusing `0 - 1 == npos`.
[Clang] NFC: Move Arm type attributes to separate trailing object. (#78424)
This decouples the Arm type attributes from other bits, which means
the data will only be allocated when a function uses these Arm
attributes.

The first patch adds the bit `HasArmTypeAttributes` to
`FunctionTypeBitfields`, which grows from 62 bits to 63 bits.

In the second patch, I've moved this bit (`HasArmTypeAttributes`) to
`FunctionTypeExtraBitfields`, because it looks like the bits in
`FunctionTypeBitfields` are precious and we really don't want that
struct
to grow beyond 64 bits.

I've split this out into two patches to explain the rationale, but those
can be squashed before merging.
DAG: Fix chain mismanagement in SoftenFloatRes_FP_EXTEND (#74558)
[AsmPrinter] Fix gcc -Wparentheses warning [NFC]
Without this gcc warned
../lib/CodeGen/AsmPrinter/DwarfDebug.cpp:3585:70: warning: suggest parentheses around '&&' within '||' [-Wparentheses]
3584 | ((&Current == &AccelDebugNames) &&
| ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
3585 | (Unit.getUnitDie().getTag() != dwarf::DW_TAG_type_unit)) &&
| ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^~
3586 | "Kind is CU but TU is being processed.");
| ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
../lib/CodeGen/AsmPrinter/DwarfDebug.cpp:3589:70: warning: suggest parentheses around '&&' within '||' [-Wparentheses]
3588 | ((&Current == &AccelTypeUnitsDebugNames) &&
| ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
3589 | (Unit.getUnitDie().getTag() == dwarf::DW_TAG_type_unit)) &&
| ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^~
3590 | "Kind is TU but CU is being processed.");
| ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
[AMDGPU] Use alias info to relax waitcounts for LDS DMA (#74537)
LDA DMA loads increase VMCNT and a load from the LDS stored must wait on
this counter to only read memory after it is written. Wait count
insertion pass does not track memory dependencies, it tracks register
dependencies. To model the LDS dependency a pseudo register is used in
the scoreboard, acting like if LDS DMA writes it and LDS load reads it.

This patch adds 8 more pseudo registers to use for independent LDS
locations if we can prove they are disjoint using alias analysis.

Fixes: SWDEV-433427
[AMDGPU][NFC] Rename feature FP8Insts to FP8ConversionInsts (#78439)
[clang][dataflow] Use `Formula::isLiteral()` in a couple more places. (#78404)
[clang][dataflow] Consider `CXXDefaultInitExpr` to be an "original record ctor". (#78423)
The CFG doesn't contain a CFGElement for the
`CXXDefaultInitExpr::getInit()`, so
it makes sense to consider the `CXXDefaultInitExpr` to be the expression
that
originally constructs the object.
[clang][ASTImporter] Improve structural equivalence of overloadable operators. (#72242)
Operators that are overloadable may be parsed as `CXXOperatorCallExpr`
or as `UnaryOperator` (or `BinaryOperator`). This depends on the context
and can be different if a similar construct is imported into an existing
AST. The two "forms" of the operator call AST nodes should be detected
as equivalent to allow AST import of these cases.

This fix has probably other consequences because if a structure is
imported that has `CXXOperatorCallExpr` into an AST with an existing
similar structure that has `UnaryOperator` (or binary), the additional
data in the `CXXOperatorCallExpr` node is lost at the import (because
the existing node will be used). I am not sure if this can cause
problems.
[LV][AArch64] LoopVectorizer allows scalable frem instructions (#76247)
LoopVectorizer is aware when a target can replace a scalable frem
instruction with a vector library call for a given VF and it returns the
relevant cost. Otherwise, it returns an invalid cost (as previously).

Add test that check costs on AArch64, when there is no vector library
available and when there is (with and without tail-folding).

NOTE: Invoking CostModel directly (not through LV) would still return
invalid costs.
[AMDGPU] Add mark last scratch load pass (#75512)
[flang] Allow user to define free via BIND(C) (#78428)
A user defining and using free/malloc via BIND(C) would previously cause
flang to crash when generating LLVM IR with error "redefinition of
symbol named 'free'". This was caused by flang codegen not expecting to
find a mlir::func::FuncOp definition of these function and emitting a
new mlir::LLVM::FuncOp that later conflicted when translating the
mlir::func::FuncOp.
[AMDGPU][True16] Support V_FLOOR_F16. (#78446)
[gn build] Port 1d286ad59b90

dependabot[bot]

Github Push to hicknhack-software/Qt-llvm-project

Andreas Reischuck

Github Push to hicknhack-software/Qt-Creator

Atempt to fix designerr CMake build (3)

Andreas Reischuck

Github Push to hicknhack-software/Qt-Creator

Atempt to fix designerr CMake build (2)

Andreas Reischuck

Github Push to hicknhack-software/ccache

refactor: fs::path-ify TestUtil
enhance: Add wrapper of std::filesystem::is_regular_file
refactor: Convert usage of Util::base_name to std::filesystem
test: Add test of core::rewrite_stderr_to_absolute_paths
refactor: Convert usage of util::real_path to std::filesystem
refactor: Convert usage of Util::get_extension to std::filesystem
refactor: Convert usage of Util::change_extension to std::filesystem
enhance: Add wrapper of std::filesystem::equivalent
refactor: Convert usage of Util::dir_name to std::filesystem
refactor: Convert usage of Util::remove_extension to std::filesystem
fix: Avoid crash for -arch as last compiler option
The bug has been present since 89aa39f394e0fa4935e6cadd7184e372a37f9506
(ccache 3.3).
Fixes #1377.
refactor: Use std::string_view in compopt functions
refactor: Move Util::is_absolute_path_with_prefix to util
chore: Remove time_of_compilation in favor of time_of_invocation
We can compare source/include files against time of invocation just as
well; source files shouldn't be modified when ccache has started anyway.
enhance: Add Args::push_back(std::string&& arg)
refactor: fs::path-ify util::is_dev_null_path
feat: Allow any prefix in prefix_command(_cpp) options
The prefix_command and prefix_command_cpp configuration options require
that the prefixes can be found in PATH.
This commit relaxes this requirement so that a relative prefix can be
used and a prefix program can have arguments.
fix: Disable caching for modified source/include files
On a cache miss, ccache opts out of the direct mode when a "too new"
include file is found. This is done to avoid a race condition in the
direct mode when a header file if modified after ccache is invoked but
before the compiler is executed. Modifying the main source code file is
not a race condition when "run_second_cpp = false", which was the case
at the time the direct mode was introduced. However, the default of
run_second_cpp was changed to true in ccache 3.3, and the potiential
race condition then exists for both the source and the include files.
Fix this by disabling caching completely (not only the direct mode) when
modification of a source/include file is detected.
Closes #1378.
chore: Remove left-over debug comment
feat: Add `--print-log-stats` (#1381)
This adds a "machine readable" variant of `--show-log-stats`, similar to
`--print-stats` as a variant of `--show-stats`.

Andreas Reischuck

Github Push to hicknhack-software/Qt-Creator

Added debug with VisualStudio2022 action for qbs projects
Fix folding issues on save
* restore cursor postion before and after folding
* ensure cursor stays visible
* also update document layout
keep cursor visible after text cut operation
Improved Pretty Printer
* preserve trailing return type function signatures
* added missing comma for variadic argument
* preserve variadic arguments on rewrite
Fixed overdoing argument type expansion on method relocations
* c++ methods carry the scope of the class for arguments and trailing return types
fixed support for trailing return types on function declaration definition link
* also support migrating from and to trailing return types
Suppress scrollToBottom on CompileOutput after navigation
improve indent for qml and javascript
* qml binding assignments with arrow functions indent one level
* qml binding assignments with object literals indent one level
* js object literals indent one level
Fix task positions in outputwindow
* fixed the task offset for tasks created directly by build steps
* no longer clear output window for new build step
* update task navigation by outputwindow to capture tasks from output parsers
* qbs build steps now only report first task for failure
* qbs build steps do not convert from path to string to path uselessly
Feature: Show VCS changes in project tree (2)
* Color nodes of project tree based on the type of VCS change.
# Conflicts:
# src/plugins/coreplugin/iversioncontrol.h
# src/plugins/git/gitplugin.cpp
Add MSVC caret diagnostics output parsing
* assemble the correct tasks and output navigations
Revert "qbs build: Do not build QmlDesigner & friends"
This reverts commit a0f356aed6795b09091472eaaed9872b6abc5514.
* fixes moc fails with forwards
type resolve curly brace constructors
try to fix gha cmake windows build
added long build media playback support
align qml formatting with normal js formatter (Beautifier, Prettier etc.)
insert matching curly brace
* should avoid unwanted empty lines on line break
fix compile debugger without tests
Atempt to fix designerr CMake build (1)