Matthew Fernandez [Sat, 23 Jul 2022 15:42:22 +0000 (08:42 -0700)]
sfdpgen: remove unimplemented 'DistanceMatrix_restrict_filtering'
Matthew Fernandez [Sat, 23 Jul 2022 15:39:08 +0000 (08:39 -0700)]
sfdpgen maximal_…: remove unused 'randomize' parameter
Matthew Fernandez [Sat, 23 Jul 2022 15:34:37 +0000 (08:34 -0700)]
sfdpgen: make 'UniformStressSmoother_new' a static function
This function is not used outside of its containing file.
Matthew Fernandez [Sat, 23 Jul 2022 15:33:57 +0000 (08:33 -0700)]
sfdpgen UniformStressSmoother_new: remove unused 'flag' parameter
Matthew Fernandez [Sat, 23 Jul 2022 15:33:02 +0000 (08:33 -0700)]
sfdpgen UniformStressSmoother_new: remove unused 'x' parameter
Matthew Fernandez [Sat, 23 Jul 2022 15:31:47 +0000 (08:31 -0700)]
sfdpgen UniformStressSmoother_new: remove unused 'dim' parameter
Matthew Fernandez [Sat, 23 Jul 2022 15:27:30 +0000 (08:27 -0700)]
sfdpgen stress_model: remove unused 'A' parameter
Matthew Fernandez [Sat, 23 Jul 2022 15:22:58 +0000 (08:22 -0700)]
sfdpgen: make 'get_stress' static
This function is unused outside of its containing file and is only used when
`DEBUG_PRINT` is defined.
Matthew Fernandez [Sat, 23 Jul 2022 15:21:28 +0000 (08:21 -0700)]
sfdpgen get_stress: remove 'weighted' parameter that is always 1
Matthew Fernandez [Sat, 23 Jul 2022 15:20:17 +0000 (08:20 -0700)]
sfdpgen get_stress: remove unused 'data' parameter
Matthew Fernandez [Sat, 23 Jul 2022 15:18:15 +0000 (08:18 -0700)]
sfdpgen jacobi: remove unused 'flag' parameter
Matthew Fernandez [Thu, 28 Jul 2022 02:47:23 +0000 (02:47 +0000)]
Merge branch 'smattr/
45dc48f2-fa91-42ed-8530-
8bcaf9fdb303' into 'main'
neatogen compiler warning squashing
See merge request graphviz/graphviz!2771
Matthew Fernandez [Sat, 23 Jul 2022 01:16:47 +0000 (18:16 -0700)]
neatogen call_tri: remove unused 'dim' parameter
Matthew Fernandez [Sat, 23 Jul 2022 01:10:55 +0000 (18:10 -0700)]
neatogen CMDS_orthog: remove unused 'g' parameter
Matthew Fernandez [Sat, 23 Jul 2022 01:06:16 +0000 (18:06 -0700)]
neatogen constrained_majorization_new_with_gaps: remove unused 'ndim' parameter
Note that the call to this function from
`constrained_majorization_new_with_gaps` also seems confused about the meaning
of its parameters. But no attempt is made to correct this.
Matthew Fernandez [Sat, 23 Jul 2022 01:03:59 +0000 (18:03 -0700)]
neatogen: remove unused 'constrained_majorization_gradient_projection'
This function has seemingly never been used.
Matthew Fernandez [Sat, 23 Jul 2022 00:56:25 +0000 (17:56 -0700)]
neatogen computeHierarchyBoundaries: remove unused 'n' parameter
Matthew Fernandez [Sat, 23 Jul 2022 00:45:45 +0000 (17:45 -0700)]
neatogen mkConstraintG: remove unused 'g' parameter
Matthew Fernandez [Sat, 23 Jul 2022 00:40:26 +0000 (17:40 -0700)]
neatogen bfs_bounded: remove unused 'n' parameter
Matthew Fernandez [Sat, 23 Jul 2022 00:32:16 +0000 (17:32 -0700)]
neatogen initLayout: remove unused 'graph' parameter
Matthew Fernandez [Sat, 23 Jul 2022 00:29:56 +0000 (17:29 -0700)]
neatogen voronoi: remove unused 'triangulate' parameter
Matthew Fernandez [Wed, 27 Jul 2022 15:32:13 +0000 (15:32 +0000)]
Merge branch 'smattr/
5fa804c5-668e-4a54-b373-
00bc933032d5' into 'main'
more compiler warning squashing
See merge request graphviz/graphviz!2770
Matthew Fernandez [Fri, 22 Jul 2022 01:45:23 +0000 (18:45 -0700)]
smyrna on_attrAddBtn_clicked: use const type to squash a -Wcast-qual warning
Unfortunately we cannot also do the same for `attr_name` in this function
because it is passed to a Graphviz function that accepts a non-const pointer.
Matthew Fernandez [Fri, 22 Jul 2022 01:44:30 +0000 (18:44 -0700)]
smyrna safestrdup: accept a const pointer
This will simplify an upcoming change.
Matthew Fernandez [Fri, 22 Jul 2022 01:42:18 +0000 (18:42 -0700)]
smyrna set_refresh_filters: use const types to squash some -Wcast-qual warnings
Unfortunately we cannot also do the same for `attr_name` in this function
because it is passed to a Graphviz function that accepts a non-const pointer.
Matthew Fernandez [Fri, 22 Jul 2022 01:37:02 +0000 (18:37 -0700)]
smyrna doApply: remove unused 'doAll' code path
Matthew Fernandez [Fri, 22 Jul 2022 01:32:52 +0000 (18:32 -0700)]
smyrna doApply: remove unused 'widget' parameter
Note this also makes `widget` unused in `on_attrApplyBtn_clicked` which is a
Glade callback and cannot have its corresponding parameter removed. So we squash
the resulting warning there.
Matthew Fernandez [Fri, 22 Jul 2022 01:31:11 +0000 (18:31 -0700)]
smyrna on_attrAddBtn_clicked: squash -Wunused-parameter warnings
55efb8e562bca6c4090a73712342ac006e133704 that squashes other such warnings in
this file missed that these are also unused in `on_attrAddBtn_clicked`.
Matthew Fernandez [Fri, 22 Jul 2022 01:27:41 +0000 (18:27 -0700)]
smyrna mOpenSlot: squash -Wunused-parameter warnings
bd869436742887a0089f5ad817ecb36a1546987c that squashed other such warnings in
this file, missed that these are also unused in `mOpenSlot`.
Matthew Fernandez [Fri, 22 Jul 2022 01:26:15 +0000 (18:26 -0700)]
smyrna btnToolFit_clicked: use float literals to simplify some code
Matthew Fernandez [Fri, 22 Jul 2022 01:24:25 +0000 (18:24 -0700)]
smyrna btnToolZoomFit_clicked: rephrase unorthodox ternaries
This function was using assignments on the right hand sides of ternary
expressions. This is technically allowed, but leads to very hard to read code.
Matthew Fernandez [Fri, 22 Jul 2022 01:20:15 +0000 (18:20 -0700)]
smyrna toolboxcallbacks: remove unnecessary parens
This also adjusts the location of some variable definitions to C99 style.
Matthew Fernandez [Fri, 22 Jul 2022 01:18:00 +0000 (18:18 -0700)]
smyrna toolboxcallbacks: remove excess white space
Matthew Fernandez [Fri, 22 Jul 2022 01:16:34 +0000 (18:16 -0700)]
smyrna toolboxcallbacks: squash -Wunused-parameter warnings
These functions are used as Glade callbacks, so need to conform to particular
type signatures.
Matthew Fernandez [Fri, 22 Jul 2022 01:14:08 +0000 (18:14 -0700)]
smyrna topviewsettings: squash -Wunused-parameter warnings
These functions are used as Glade callbacks, so need to conform to particular
type signatures.
Matthew Fernandez [Fri, 22 Jul 2022 01:08:04 +0000 (18:08 -0700)]
smyrna load_settings_from_graph: remove unused 'g' parameter
Matthew Fernandez [Fri, 22 Jul 2022 01:05:09 +0000 (18:05 -0700)]
tclpkg tclGdSizeCmd: squash an -Wunused-parameter warning
This parameter cannot easily be removed because this function is used as a
callback and needs to conform to a particular type signature.
Matthew Fernandez [Tue, 26 Jul 2022 03:00:38 +0000 (03:00 +0000)]
Merge branch 'smattr/
d5eaf8b2-870d-4ef0-9557-
51428f498a1b' into 'main'
misc simplification
See merge request graphviz/graphviz!2768
Matthew Fernandez [Tue, 19 Jul 2022 03:00:30 +0000 (20:00 -0700)]
Pango plugin: remove unnecessary cast
Matthew Fernandez [Tue, 19 Jul 2022 03:00:16 +0000 (20:00 -0700)]
GDK plugin: remove unnecessary casts
Matthew Fernandez [Tue, 19 Jul 2022 02:59:46 +0000 (19:59 -0700)]
GD plugin: remove unnecessary casts
Matthew Fernandez [Tue, 19 Jul 2022 02:49:22 +0000 (19:49 -0700)]
rsvg plugin gvloadimage_rsvg_cairo: remove unnecessary cast
Matthew Fernandez [Tue, 19 Jul 2022 02:46:29 +0000 (19:46 -0700)]
glcomp plugin file_exists: simplify readability test
We do not need to open a file to test if it is readable.
Matthew Fernandez [Tue, 19 Jul 2022 02:42:42 +0000 (19:42 -0700)]
expr extoken_fn: simplify hex parsing loop
Matthew Fernandez [Tue, 19 Jul 2022 02:30:50 +0000 (19:30 -0700)]
expr prformat: replace pointer and extent with central string view abstraction
Matthew Fernandez [Tue, 19 Jul 2022 02:25:57 +0000 (19:25 -0700)]
common: remove some unnecessary casts
Magnus Jacobsson [Mon, 25 Jul 2022 19:26:06 +0000 (19:26 +0000)]
Merge branch 'fix-subgraph-layout-and-rendering' into 'main'
Fix subgraph layout and rendering
Closes #1800
See merge request graphviz/graphviz!2767
Magnus Jacobsson [Tue, 12 Jul 2022 13:15:50 +0000 (15:15 +0200)]
remove GD_drawing(g) NULL guard
This guard is unnecessary since `graph_cleanup` alrady has the
necessary guards. This also ensures that the Agraphinfo_t record and
the graph's label is cleaned away even if the graph for some reason
doesn't have a drawing (can this ever happen?).
Magnus Jacobsson [Tue, 12 Jul 2022 13:04:57 +0000 (15:04 +0200)]
gvFreeLayout: correct comments about reset and root graph
This comment was added in
44b02b91bd156cb832894fb4948bf4bc75ab27cb,
but this commit changed the cleanup to work on the given graph
(possibly a subgraph), not the root graph. The resets mentioned were
removed in
67acd9e27a053a5b80b3d1458f512d472debb434,
b3b7e0d1aea43baa81bf06a27d38447d6958c1b8 and
a71555a4ac650bf91396c77c277697b45eee44e6, although the called
`graph_cleanup` function does its own reset.
Magnus Jacobsson [Wed, 6 Jul 2022 12:56:16 +0000 (14:56 +0200)]
gvFreeLayout: remove incorrect comment about only the root graph having a drawing
This comment was added in
c55b546e3965a8dfd2a66763961c4a8003e433bf,
but both subgraphs and root graphs have a reference to (the same)
drawing through this statement in gvLayoutJobs:
GD_drawing(agroot(g)) = GD_drawing(g);
Magnus Jacobsson [Mon, 18 Jul 2022 13:02:59 +0000 (15:02 +0200)]
twopigen: layout: fix heap-buffer-overflow by storing dimension in root graph
The layout allocates memory based on the dimension of the root graph,
but since the dimension was stored in the subgraph and the dimension
in the root graph defaulted to zero, too little memory was allocated.
An alternative solution would have been to use the dimension of the
subgraph, but this had other implications and the chosen solution is
the same as what the other two layout engines supporting the dim
attribute (neato and sfdp) use. Twopi always uses two dimensions.
This fixes the last memory issue detected by ASan in the
test_subgraph_layout test, which now runs without failures.
Magnus Jacobsson [Mon, 18 Jul 2022 12:56:45 +0000 (14:56 +0200)]
circogen: layout: fix heap-buffer-overflow by storing dimension in root graph
The layout allocates memory based on the dimension of the root graph,
but since the dimension was stored in the subgraph and the dimension
in the root graph defaulted to zero, too little memory was allocated.
An alternative solution would have been to use the dimension of the
subgraph, but this had other implications and the chosen solution is
the same as what the other two layout engines supporting the `dim`
attribute (neato and sfdp) use. Circo always uses two dimensions.
Magnus Jacobsson [Mon, 18 Jul 2022 12:17:04 +0000 (14:17 +0200)]
fdpgen: layout: fix heap-buffer-overflow by storing dimension in root graph
The layout allocates memory based on the dimension of the root graph,
but since the dimension was stored in the subgraph and the dimension
in the root graph defaulted to zero, too little memory was allocated.
An alternative solution would have been to use the dimension of the
subgraph, but this had other implications and the chosen solution is
the same as what the other two layout engines supporting the `dim`
attribute (neato and sfdp) use.
Magnus Jacobsson [Mon, 18 Jul 2022 15:46:01 +0000 (17:46 +0200)]
gvRender*: render subgraph when given, not always root graph
These functions always rendered the root graph, even if a sub graph
was given. Apart from being incorrect, this caused
heap-buffer-overflow when trying to render a graph consisting of two
subgraph of which only one had a layout. Also, even if both subgraphs
had layouts, they somehow mixed up the bounding boxes of the graphs
causing some nodes and edges not to be rendered since they were
outside the graph bounding box.
Magnus Jacobsson [Wed, 13 Jul 2022 06:17:54 +0000 (08:17 +0200)]
gvLayoutJobs: initialize also root graph if it isn't already
The root graph info is needed by gvLayoutJobs when doing subgraph
layout. This fixes a heap-buffer-overflow detected by ASan in the
test_subgraph_layout test.
Magnus Jacobsson [Tue, 12 Jul 2022 13:52:26 +0000 (15:52 +0200)]
add new test test_subgraph_layout
This test is run in CI with ASan leak detection enabled. It currectly
fails because ASan detects several memory issues in various layout
engines.
Magnus Jacobsson [Mon, 18 Jul 2022 07:53:21 +0000 (09:53 +0200)]
patchwork: fix memory leak by freeing cluster data in graph cleanup
This fixes a memory leak detected by ASan in the test_clusters test,
which now runs without failures
Magnus Jacobsson [Mon, 18 Jul 2022 07:47:15 +0000 (09:47 +0200)]
dotgen cluster: fix memory leaks by freeing data from removed rank leaders
The `build_skeleton` function inserts virtual nodes and edges into the
graph to be used by the clustering algorithms. The `expand_cluster`
function later installs real nodes or sub-clusters and deletes the
virtual nodes and edges by calling the `remove_rank_leaders` function
which removes these from the graph. However, it didn't free the data
allocated for them which caused several memory leaks.
Magnus Jacobsson [Sun, 10 Jul 2022 08:47:22 +0000 (10:47 +0200)]
add new test case test_clusters
Upcoming commits in this series will make changes to subgraph layout
and since clusters are also subgraphs, we want to weed out any memory
issues related to clusters separately.
This test is run in CI with ASan leak detection enabled. It currectly
fails because ASan detects memory leaks in the dot and patchwork
layout engines.
Magnus Jacobsson [Tue, 12 Jul 2022 13:11:25 +0000 (15:11 +0200)]
don't clean away Agraphinfo_t in layout specific cleanups
This cleanup is since
2b3b37c47244f57a9690bff8fb92f77d17ec616c handled
by the general `graph_cleanup` function.
This information is since
44b02b91bd156cb832894fb4948bf4bc75ab27cb
needed by the general layout cleanup function that runs after the
layout-specific cleanup.
This problem was fixed for root graphs in commit
c55b546e3965a8dfd2a66763961c4a8003e433bf, but not for subgraph layout.
This fixes a heap-buffer-overflow detected by ASan in the
test_neatopack test, which now runs without failures.
This commit also reverts commit
291b7a40f742551797e5681acbb8f5662c084822 since the
dot.demo/neatopack.c example now runs without failures.
Fixes https://gitlab.com/graphviz/graphviz/-/issues/1800.
Magnus Jacobsson [Mon, 4 Jul 2022 13:01:10 +0000 (15:01 +0200)]
add new test_neatopack test
Upcoming commits in this series will make changes to gvFreeLayout and
we want to ensure that those changes don't introduce any memory leaks
when run with ASan, since gvFreeLayout handles cleanup of both
subgraphs and the root graph and the test case extracts subgraphs from
the root graph.
This test is run in CI with ASan leak detection enabled. It is more or
less a replica of 'dot.demo/neatopack.c', which is also tested in CI
by 'tests/test_examples.py', but with leak detection disabled. It
currently fails because of a heap-buffer-overflow in the gvFreeLayout.
Towards https://gitlab.com/graphviz/graphviz/-/issues/1800.
Magnus Jacobsson [Wed, 6 Jul 2022 07:46:19 +0000 (09:46 +0200)]
dot.demo/neatopack.c: add missing free of connected components subgraphs
Towards https://gitlab.com/graphviz/graphviz/-/issues/1800.
Magnus Jacobsson [Mon, 4 Jul 2022 09:43:14 +0000 (11:43 +0200)]
dot.demo/neatopack.c: add missing initialization
Normally gvLayoutJobs does this initialization, but since we call
ccomps before doing the layout and it expects graphs and nodes to have
'Agraphinfo_t' and 'Anodeinfo_t' records attached already, we must do
this initialization in the user program instead, see
https://graphviz.org/pdf/pack.3.pdf.
Fixes first part of
https://gitlab.com/graphviz/graphviz/-/issues/1800.
Magnus Jacobsson [Tue, 5 Jul 2022 20:45:20 +0000 (22:45 +0200)]
add new test_subgraphs test
Upcoming commits in this series will make changes to gvFreeLayout and
we want to ensure that those changes don't introduce any memory leaks
when run with ASan, since gvFreeLayout handles cleanup of both
subgraphs and the root graph.
This test is run in CI with ASan leak detection enabled.
Towards https://gitlab.com/graphviz/graphviz/-/issues/1800.
Magnus Jacobsson [Sun, 17 Jul 2022 11:13:14 +0000 (13:13 +0200)]
osage_cleanup: add missing cleanup of edges
This fixes a memory leak detected by ASan in the test_engines test,
which now runs without failures
Magnus Jacobsson [Tue, 5 Jul 2022 17:22:31 +0000 (19:22 +0200)]
add new simple test_engines test
Upcoming commits in this series will make changes to gvFreeLayout and
we want to ensure that those changes don't introduce any memory leaks
when run with ASan, since gvFreeLayout calls layout-specific cleanup
functions.
This test is run in CI with ASan leak detection enabled. It currently
fails because of a memory leak in the osage layout.
Towards https://gitlab.com/graphviz/graphviz/-/issues/1800.
Magnus Jacobsson [Tue, 19 Jul 2022 19:13:12 +0000 (21:13 +0200)]
add the fmt library as a test dependency
Magnus Jacobsson [Sun, 6 Mar 2022 13:11:31 +0000 (14:11 +0100)]
add libfmt-dev to Ubuntu 22.04 Dockerfile
It is going to be used in new tests in upcoming commits.
Magnus Jacobsson [Fri, 22 Jul 2022 12:10:19 +0000 (14:10 +0200)]
tests: CMake: change test_common from a static to a shared library
This was the original intention according to the commit message in
commit
a9daedd8bee1b5620a933b612c3335b90bde263d.
Magnus Jacobsson [Wed, 6 Jul 2022 10:42:02 +0000 (12:42 +0200)]
tests/CMakeLists.txt: sort test cases
Matthew Fernandez [Sun, 24 Jul 2022 05:52:16 +0000 (05:52 +0000)]
Merge branch 'smattr/
BA128E20-1C11-4C5C-978E-
6F16C28B3671' into 'main'
rephrase comparators to avoid arithmetic
See merge request graphviz/graphviz!2766
Matthew Fernandez [Fri, 8 Jul 2022 00:24:05 +0000 (17:24 -0700)]
sparse comp_ascend_int: remove unnecessary parens
Matthew Fernandez [Fri, 8 Jul 2022 00:24:05 +0000 (17:24 -0700)]
sfdpgen scomp: remove unnecessary parens
Matthew Fernandez [Fri, 8 Jul 2022 00:24:05 +0000 (17:24 -0700)]
patchwork nodecmp: rephrase comparator to avoid arithmetic
cccb8b1d22a18031fc92d93133c7fa14ef7e1361 fixed an integer overflow in a
`memcmp`-/`strcmp`-like comparator. The same situation exists in the code
touched in this commit. Rather than wait for an edge case to expose an overflow
here, this change makes the same update, removing arithmetic and the consequent
possibility of overflow.
Matthew Fernandez [Fri, 8 Jul 2022 00:24:05 +0000 (17:24 -0700)]
pack cmpf: rephrase comparator to avoid arithmetic
cccb8b1d22a18031fc92d93133c7fa14ef7e1361 fixed an integer overflow in a
`memcmp`-/`strcmp`-like comparator. The same situation exists in the code
touched in this commit. Rather than wait for an edge case to expose an overflow
here, this change makes the same update, removing arithmetic and the consequent
possibility of overflow.
Matthew Fernandez [Fri, 8 Jul 2022 00:24:05 +0000 (17:24 -0700)]
neatogen cmpIpair: rephrase comparator to avoid arithmetic
cccb8b1d22a18031fc92d93133c7fa14ef7e1361 fixed an integer overflow in a
`memcmp`-/`strcmp`-like comparator. The same situation exists in the code
touched in this commit. Rather than wait for an edge case to expose an overflow
here, this change makes the same update, removing arithmetic and the consequent
possibility of overflow.
Matthew Fernandez [Fri, 8 Jul 2022 00:24:05 +0000 (17:24 -0700)]
neatogen cmpitem: rephrase comparator to avoid arithmetic
cccb8b1d22a18031fc92d93133c7fa14ef7e1361 fixed an integer overflow in a
`memcmp`-/`strcmp`-like comparator. The same situation exists in the code
touched in this commit. Rather than wait for an edge case to expose an overflow
here, this change makes the same update, removing arithmetic and the consequent
possibility of overflow.
Matthew Fernandez [Fri, 8 Jul 2022 00:24:05 +0000 (17:24 -0700)]
fdpgen ijcmpf: rephrase comparator to avoid arithmetic
cccb8b1d22a18031fc92d93133c7fa14ef7e1361 fixed an integer overflow in a
`memcmp`-/`strcmp`-like comparator. The same situation exists in the code
touched in this commit. Rather than wait for an edge case to expose an overflow
here, this change makes the same update, removing arithmetic and the consequent
possibility of overflow.
Matthew Fernandez [Fri, 8 Jul 2022 00:24:05 +0000 (17:24 -0700)]
dotgen edgeposcmpf: rephrase comparator to avoid arithmetic
cccb8b1d22a18031fc92d93133c7fa14ef7e1361 fixed an integer overflow in a
`memcmp`-/`strcmp`-like comparator. The same situation exists in the code
touched in this commit. Rather than wait for an edge case to expose an overflow
here, this change makes the same update, removing arithmetic and the consequent
possibility of overflow.
Matthew Fernandez [Fri, 8 Jul 2022 00:24:05 +0000 (17:24 -0700)]
dotgen nodeposcmpf: rephrase comparator to avoid arithmetic
cccb8b1d22a18031fc92d93133c7fa14ef7e1361 fixed an integer overflow in a
`memcmp`-/`strcmp`-like comparator. The same situation exists in the code
touched in this commit. Rather than wait for an edge case to expose an overflow
here, this change makes the same update, removing arithmetic and the consequent
possibility of overflow.
Matthew Fernandez [Fri, 8 Jul 2022 00:24:05 +0000 (17:24 -0700)]
dotgen ordercmpf: rephrase comparator to avoid arithmetic
cccb8b1d22a18031fc92d93133c7fa14ef7e1361 fixed an integer overflow in a
`memcmp`-/`strcmp`-like comparator. The same situation exists in the code
touched in this commit. Rather than wait for an edge case to expose an overflow
here, this change makes the same update, removing arithmetic and the consequent
possibility of overflow.
Matthew Fernandez [Fri, 8 Jul 2022 00:24:05 +0000 (17:24 -0700)]
common textfont_comparf: rephrase comparator to avoid arithmetic
cccb8b1d22a18031fc92d93133c7fa14ef7e1361 fixed an integer overflow in a
`memcmp`-/`strcmp`-like comparator. The same situation exists in the code
touched in this commit. Rather than wait for an edge case to expose an overflow
here, this change makes the same update, removing arithmetic and the consequent
possibility of overflow.
Matthew Fernandez [Fri, 8 Jul 2022 00:24:05 +0000 (17:24 -0700)]
common increasingrankcmpf: rephrase comparator to avoid arithmetic
cccb8b1d22a18031fc92d93133c7fa14ef7e1361 fixed an integer overflow in a
`memcmp`-/`strcmp`-like comparator. The same situation exists in the code
touched in this commit. Rather than wait for an edge case to expose an overflow
here, this change makes the same update, removing arithmetic and the consequent
possibility of overflow.
Matthew Fernandez [Fri, 8 Jul 2022 00:24:05 +0000 (17:24 -0700)]
common decreasingrankcmpf: rephrase comparator to avoid arithmetic
cccb8b1d22a18031fc92d93133c7fa14ef7e1361 fixed an integer overflow in a
`memcmp`-/`strcmp`-like comparator. The same situation exists in the code
touched in this commit. Rather than wait for an edge case to expose an overflow
here, this change makes the same update, removing arithmetic and the consequent
possibility of overflow.
Matthew Fernandez [Fri, 8 Jul 2022 00:24:05 +0000 (17:24 -0700)]
cgraph agsubnodeseqcmpf: remove unnecessary casts
Matthew Fernandez [Fri, 8 Jul 2022 00:24:05 +0000 (17:24 -0700)]
cgraph agsubnodeidcmpf: remove unnecessary casts
Matthew Fernandez [Fri, 8 Jul 2022 00:24:05 +0000 (17:24 -0700)]
gvcolor cmpf: rephrase comparator to avoid arithmetic
cccb8b1d22a18031fc92d93133c7fa14ef7e1361 fixed an integer overflow in a
`memcmp`-/`strcmp`-like comparator. The same situation exists in the code
touched in this commit. Rather than wait for an edge case to expose an overflow
here, this change makes the same update, removing arithmetic and the consequent
possibility of overflow.
Matthew Fernandez [Fri, 8 Jul 2022 00:24:05 +0000 (17:24 -0700)]
dijkstra cmpf: rephrase comparator to avoid arithmetic
cccb8b1d22a18031fc92d93133c7fa14ef7e1361 fixed an integer overflow in a
`memcmp`-/`strcmp`-like comparator. The same situation exists in the code
touched in this commit. Rather than wait for an edge case to expose an overflow
here, this change makes the same update, removing arithmetic and the consequent
possibility of overflow.
Matthew Fernandez [Fri, 8 Jul 2022 00:24:05 +0000 (17:24 -0700)]
dotgen edgelblcmpfn: remove unnecessary casts
Matthew Fernandez [Fri, 8 Jul 2022 00:24:05 +0000 (17:24 -0700)]
dotgen edgecmp: rephrase comparator to avoid arithmetic
cccb8b1d22a18031fc92d93133c7fa14ef7e1361 fixed an integer overflow in a
`memcmp`-/`strcmp`-like comparator. The same situation exists in the code
touched in this commit. Rather than wait for an edge case to expose an overflow
here, this change makes the same update, removing arithmetic and the consequent
possibility of overflow.
Note that the first comparison in this function appears to be the wrong way
around. But this (mistake?) is left as-is.
Matthew Fernandez [Fri, 8 Jul 2022 00:24:05 +0000 (17:24 -0700)]
dotgen edgecmp: remove unnecessary casts
Matthew Fernandez [Fri, 8 Jul 2022 00:24:05 +0000 (17:24 -0700)]
circogen: remove commented out 'cleanBlocklist'
This was commented out when first committed in the initial revision,
256ef66663ca0c072554ee3f5e7971911031b3c7. It has never been uncommented or used.
Matthew Fernandez [Sat, 23 Jul 2022 01:14:16 +0000 (01:14 +0000)]
Merge branch 'smattr/
ac7d6a52-2008-42b9-8ce4-
e82b996d2211' into 'main'
squash more -Wunused-parameter warnings
See merge request graphviz/graphviz!2765
Matthew Fernandez [Sun, 17 Jul 2022 22:51:06 +0000 (15:51 -0700)]
neatogen getPath: remove unused obstacle parameters
This function ignores obstacles when routing.
Matthew Fernandez [Sun, 17 Jul 2022 22:44:42 +0000 (15:44 -0700)]
dotgen make_flat_adj_edges: remove unused 'P' parameter
Matthew Fernandez [Sun, 17 Jul 2022 22:43:31 +0000 (15:43 -0700)]
dotgen cloneNode: remove unused 'flipped' parameter
Matthew Fernandez [Sun, 17 Jul 2022 22:32:33 +0000 (15:32 -0700)]
dotgen constraining_flat_edge: remove unused 'v' parameter
Matthew Fernandez [Sun, 17 Jul 2022 22:31:08 +0000 (15:31 -0700)]
dotgen getCluster: remove unused 'g' parameter
This also makes `g` in `makeCompoundEdge` unused, so we remove that parameter
too.
Matthew Fernandez [Sun, 17 Jul 2022 22:27:14 +0000 (15:27 -0700)]
dotgen make_interclust_chain: remove unused 'g' parameter
Matthew Fernandez [Sun, 17 Jul 2022 22:25:34 +0000 (15:25 -0700)]
topfish make_coarse_ex_graph: remove unused 'nvtxs' parameter