The diff-pair placer creates DIFF_PAIRs on the stack, which have
VIAs in them, which have HOLEs in them. The HOLEs get put in the
NODE's index, and we then crash later when doing a collision after
the stack variable has gone out of scope.
The stack variable is also copied while doing a search for "best",
and without the operator= definitions the "best" and the "original"
both think they own the HOLE.
Fixes https://gitlab.com/kicad/code/kicad/-/issues/14852
This was causing intermittent shove hiccups/freezes as RULE_RESOLVER::GetClearance() could return a bogus clearance value, taking
a cached value for an item that does not exist anymore. If the allocator/stack accidentally reclaimed such item's address - we were getting
very nasty and difficult to reproduce misbehaviours of the shove algorithm. Hopefully this patch fixes this. More info in the comments.
Note to self - I'm still not fully convinced I want PNS::ITEM pointers as the cache key, probably a unique counter would be better here.
Don't open-code knockout text shape generation in several different
places.
Make sure triangulated knockout text gets clearance added when
specified.
Collapse duplicated footprint text item plot routine (they're no
longer any different from plotting pcb text items).
(cherry picked from commit c71cf21e2f)
using a non Kicad font
PCB_PLUGIN::formatRenderCache() was considering all shapes in cache are polygons,
but this is not true: some items can be stroke shapes.
From stable branch
Fixes#14804https://gitlab.com/kicad/code/kicad/-/issues/14804
Also removes auto-closing of blocking dialogs when inserting
symbols from the Symbol Editor or Chooser, preventing the bug
that QuasiModel was removed for earlier.
Also fixes a non-initialized parent pointer when the TextBox
Properties dialog was called from DrawShape().
Fixes https://gitlab.com/kicad/code/kicad/-/issues/14777
(cherry picked from commit 5a9ed66cfd)
The layer combobox color swatch was clipped when the combobox was
closed when the dialog was launched. Setting the minimum width
resolved the issue. This may have been a GTK only issue.
(cherry picked from commit a13a49f45a)
Operator[] was creating an empty entry in the connectivity map,
fooling us into thinking the item had already been added.
Worryingly, this bug has been in there since 2017; a recent
change in teardrops just happened to expose it.
Fixes https://gitlab.com/kicad/code/kicad/-/issues/14781
(cherry picked from commit 7e51077992)
(cherry picked from commit 0d94532df9)
The root cause was not removing the original (pre-dragging) via from the world, comparing the ITEM::Parent() pointers of the items was only hiding the main issue.
This also fixes the PNSViaCollisions test in the qa/ suite.
The previous test didn't handle is-self-tests between a hole and
its override in a different NODE. When calculating the pushout
force we don't remove the via (and its hole) from the current
node until after the calculation, so we end up checking for
collisions between the new (cloned) via's hole and the original
hole in the root node.
Fixes https://gitlab.com/kicad/code/kicad/-/issues/14795
(cherry picked from commit b442d769fd)
Operator[] was creating an empty entry in the connectivity map,
fooling us into thinking the item had already been added.
Worryingly, this bug has been in there since 2017; a recent
change in teardrops just happened to expose it.
Fixes https://gitlab.com/kicad/code/kicad/-/issues/14781
(cherry picked from commit 7e51077992)
The layer combobox color swatch was clipped when the combobox was
closed when the dialog was launched. Setting the minimum width
resolved the issue. This may have been a GTK only issue.
(cherry picked from commit a13a49f45a)
Version 7 text alignment changed subtly for stroke fonts from version 6.
Additionally, the output has been different between screen and plotting,
leading to offset text in plotted output relative to the text shown on
screen.
This introduces a fudge factor in FONT::getLinePositions to correct the
offset in the plotting output relative to v6.
This also changes the SCH_PAINTER and PCB_PAINTER to correct the
relative offsets between GAL and PLOTTER classes. The source of these
offsets is atm unclear.
Fixes https://gitlab.com/kicad/code/kicad/issues/14755
(cherry picked from commit 0de24bfd59)
If you drag a track out and connect to a pad then press the hotkey
to switch the track posture it may not switch because the track
posture selected when the pad was reached is different than the one
that the mouse trail placer drew. As the aspect switching is done
in the mouse trail placer it does not know that the line placer
already override the track.
To correct this, in LINE_PLACER::FlipPosture() if the aspect is not
currently manually forced then copy the current aspect from the
line placer to m_mouseTrailPlacer before calling the FlipPosture()
method of m_mouseTrailPlacer. This will only be done once per
dragging session as once the aspect is manually forced it remains
manually forced. This seems to fix the issue with no undesirable
side effects.
Fixes: https://gitlab.com/kicad/code/kicad/-/issues/12369
(cherry picked from commit 85f4f0d6c8)
Fixes rotating bitmaps with "rotate counterclockwise",
"rotate clockwise" and "move exactly" commands.
Only rotation by 0, 90, 180 and 270 degrees is supported
for bitmaps. Other values are rounded to the nearest 90
degree multiple.
There is a different bug that this commit doesn't address,
which is that BITMAP_BASE::Rotate() got its CW/CCW logic
reversed ("m_image->Rotate( false )" should rotate CW, but
does not).
Fixes https://gitlab.com/kicad/code/kicad/issues/14197
(cherry picked from commit 45b94a4b3e)
Rely on DRC for text readability instead.
(Fixes Sentry KICAD-R4 on Mac, and the assert when opening
Preferences from a non-PCBNew-module window.)
(cherry picked from commit 245c3ba6e4)
A sheetpath is required to correctly resolve text variables.
Depending on currentSheet is rife with bugs.
There are many places where we do *not* want to be prepending
field names to the field values, such as netlisting,
building PDF hypertext menus, etc.
Also, Find/Replace needs to work on unresolved text, as
that's what we're going to display (and if replace nuked
your variable references you wouldn't be happy).
(cherry picked from commit b41d446f58)
Settings should be initialized on start-up. This removes the option of
lazily loading the settings from file and instead requires all settings
needed to be loaded on KiFACE start before requesting data from the
settings object
(cherry picked from commit e6ab9a88ce)
Adds QA checks to copper sliver tests. Adds the following checks:
- Dot product between two arms (quickly avoids checks for >90°)
- Checks the sliver is convex (area test)
- Eliminates minor slivers with angles that are approximately 0 and ones
with the opposite side width beneath a configurable level
- Updates Clipper2 to fix a couple of jagged edges on inflate
- Adds simplify during zone fill inflation to limit jaggies
Fixes https://gitlab.com/kicad/code/kicad/issues/14549
(cherry picked from commit f7f52d77e4)
In particular, when a CN_ITEM is freed set all its anchor's
item pointers to NULL. (The anchors have a separate
lifecycle due to being std::shared_ptrs.)
Hopefully fixes Sentry KICAD-KV.
(cherry picked from commit 0eac5c6748)
Our basic logic for whether we have a hole or an outline in a footprint
checks if there is copper pads outside of the closed edge. This misses
the case where the footprint is used to align multiple holes but has no
copper. Since we do not support multiple outlines in a board, we can
safely assume that multiple, closed edges in a footprint represent holes
and not outlines
Fixes https://gitlab.com/kicad/code/kicad/issues/14468
(cherry picked from commit 4b91c0c8ee)
The spacemouse driver, particularly on Mac, is extremely unstable and
causes crashes even when not being used. This places the interface
behind an advanced config flag to ensure that users can affirmatively
opt-in to the potential for crashes
(cherry picked from commit 2cd854af14)
DRC exclusions were originally written following the C++
pragma model (ie: allow this violation here). However, the
"exclusion" terminology we used in the GUI suggests a model
model where the exclusions go away when the violation no
longer exists.
Fixes https://gitlab.com/kicad/code/kicad/issues/14351
About 1/3 of callers to Parent() don't care if they get the immediate
parent or not, about 1/3 want only the immediate parent, and about 1/3
want the hole parent's Parent().
I had earlier changed PNS::ITEM::HOLE to override Parent() and return
the hole parent's Parent(), but since the callers are pretty evenly
split I've reverted that and added BoardItem() for callers who want the
eventual BOARD_ITEM (whether a direct parent or a grandparent).
(Also removes a dead routine so I didn't have to figure out which of
the two it wanted....)
(cherry picked from commit 6f0d963683)
(While we don't currently have holes without a parent pad or via, you
could imagine using HOLEs in the future to represent a hole in a first-class
board outline, so better not to build this assumption in.)
(cherry picked from commit b18bf09208)
- LINEs now own their vias optionally (depending whether a LINE with its VIA belongs to a node)
- get rid of unique_ptr on the internal NODE::Add API (and also made it private)
- make sure stack pointers are not added to the node
- use rvalue reference for NODE::Add/NODE::Replace
Manually rebased by Jeff Young <jeff@rokeby.ie> 5 April 2023
(cherry picked from commit d961cdce3f)
Rebased by Jeff Young <jeff@rokeby.ie> 5 April 2023
(Also includes a bug-fix for highlighting collisions with edge-cut
items.)
(cherry picked from commit bfbda978b8)
- prerequisite for holes as first class objects code
- introduced the OWNABLE_ITEM interface that defines the owning container (NODE/other ITEM/ITEM_SET)
- simplified the ITEM_SET interface
- constified a lot of APIs (ownership/JOINT management) as a side effect
Rebased and cleaned up by Jeff Young <jeff@rokeby.ie> 5 April 2023
- some STL cover types removed
(cherry picked from commit 1283c4713f)
This method shadows the C++ method, and doesn't properly handle updating
the duplicated object's KIID to be unique.
Also add tests to ensure the KIIDs are unique after duplication.
Fixes https://gitlab.com/kicad/code/kicad/issues/14460
(Cherry-picked from 5d0b5e16ff)
pickSingleItem() had a two-attempt algorithm, but the second attempt
was never made because we'd check for no hits before throwing out the
aAvoidItems.
The second attempt would fail anyway because QueryHoverItems() ignored
the aUseClearance flag when not idle, and didn't factor in the actual
clearance.
(cherry picked from commit d040690f60)
Substantial elements following a divot should be at least as far in each
cardinal direction from the origin point in order to be considered
substantial. This catches cases where the "substantial" element is
actually a straight segment away from the divot
Fixes https://gitlab.com/kicad/code/kicad/issues/14130
(cherry picked from commit 914b5a4d21)
Set their parent to nullptr before deleting (as the board may have
already been deleted).
Likely fix for Sentry KICAD-79.
(cherry picked from commit 13e79af8a0)
ShowBoardSetup can be called multiple times from multiple locations,
resulting in duplicate board setup dialog boxes. This is confusing and
should be prevented. We use the established patter from the symbol
picker to ensure that only one copy of the dialog is ever opened at a
time
(cherry picked from commit 7695d510d6)
Previously, it was not always the edited footprint, but a footprint
selected from the library tree.
It was especially annoying when the footprint was loaded from the board editor:
The footprint was not the loaded footprint, but its copy from library.
From master branch
This allows custom rules to be authored without firing the auto-generated
rule from Board Setup > Silk Item Clearance (which should be only for silk-
to-silk clearances).
Fixes https://gitlab.com/kicad/code/kicad/issues/14417
(cherry picked from commit 0266d03f79)
Adds the option to simplify the output of Clipper ops that remove minor
detours from the output lines. These detours are not substantive, so
removing them speeds up the rest of the system by removing unimportant
vertices. This also prevents the introduction of inadvertant concave
points when unioning two, closely-sized rounded shapes
Software like Solidworks and other CAD treat STEP labels as unique entries.
When you import multi STEP files into the same project, it'll start deduplicating your design by those name.
So two completely unrelated PCBs with the same "PCB" name will result in one being replaced by another.
When drawing the ratsnest, we need to be careful to avoid accessing
CONN_ITEMs that have been changed by the underlying model. Checking for
dirty items instead of valid items will prevent us looking at data that
are out of date
Fixes https://gitlab.com/kicad/code/kicad/issues/14265
(cherry picked from commit 283770fc45)
Includes changing the class hierarchy so that dimension objects inherit
from PCB_TEXT rather than containing a PCB_TEXT member variable.
(cherry picked from commit 23accffc6d)
The logic to handle divots needs to account for fully nested, same net
zones. The process of subtracting led to us considering the inner zone
to be a zone knock-out (negative polygon). To avoid this, we need to
check if the inner, higher priority zone has any connection to the outer
zone. If it does not, then we can treat it as an isolated zone without
worrying about divots to the outer zone.
(cherry picked from commit 9146e38af4)
We report bitmaps as being 'on' a copper layer because they are
associated with it but they are not actually physical parts so should
not be included in the DRC checks
Additionally, Bitmaps effective shape starts are the top left corner
rather than the center (unclear why that was there in the first place)
Fixes https://gitlab.com/kicad/code/kicad/issues/14065
(cherry picked from commit 1b8b216ac8)
Rounded line ends when deflating leads to microscopic jags in the
outline that are not visible but add substantial computation time and
minor error when computing polygon offsets.
Instead, the chamfer deflate method prevents these jagged lines by
clipping angles < 90° by the error level. This does not impact deflate
calls where we explicitly require the angles to be maintained
(cherry picked from commit af10878954)
Jeff Young
Tomasz Wlostowski
David Holdeman
Marek Roszko
Mike Williams
jean-pierre charras
Wayne Stambaugh
Seth Hillbrand
Jon Evans
Alex
Ian McInerney
Salvador E. Tropea
Steve Bollinger
qu1ck
Martin Thierer
Lucas Dumont
Ian McInerney
Roberto Fernandez Bautista
Salvador E. Tropea
mitxela