Inherit from tools::WeakBase non-virtually, so that we can use a
static_cast in tools::WeakReference::get instead of a dynamic_cast.
This takes the file-open time from 1m21 to 40s for me.
Add a clang plugin to make sure we don't accidentally end up inheriting
from tools::WeakBase more than once.
Change-Id: I9c7c36403333f99094e1f9d8cce2ecd9200377f9
Reviewed-on: https://gerrit.libreoffice.org/71231
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
Tested-by: Noel Grandin <noel.grandin@collabora.co.uk>
if we're doing a find/insert on a set or a map, it is better to just do
a conditional insert/emplace operation than triggering two lookups.
Change-Id: I80da5097f5a89fe30fa348ce5b6e747c34287a8d
Reviewed-on: https://gerrit.libreoffice.org/70937
Tested-by: Jenkins
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
...to find more bugs like the one addressed in
6340daac7b99c65249363a4bb61c492de31ef5d6 "Revert broken
loplugin:sequentialassign change". What it does is: "Warn when a variable is
referenced from its own initializer. This is not invalid in general (see C++17
[basic.life]), but is at least suspicious." It found one false positive
(addressed with 884ad0d1af88f9985d30ef0dfe92d89e82f8e576 "Split
localProcessFactory function into class with setter and getter") and five true
positives (addressed with e0ccbe72ed6eb0d309ed272a78fd67a512acff5d "Fix use of
variable before its lifetime begins" and
0e335af4d3f044511551fa2ede20911beaee9b41 "Fix uses of variables before their
lifetimes begin").
Change-Id: I4c45cceaa042e93b37ad24a54784c027f6ca1f87
Reviewed-on: https://gerrit.libreoffice.org/70897
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
Look for places we are assigning to the same variable twice
in succession, which means we can simplify that to a single assign
Change-Id: I499d20e28f5595e81e927bef8e1bf364eea8ba91
Reviewed-on: https://gerrit.libreoffice.org/70531
Tested-by: Jenkins
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
...as seen by clang-cl when there are function parameters of function pointer
type involving SAL_CALL
Change-Id: Ie35f00d4e15ca777b14dd5968cdbd97e43bca1a1
Reviewed-on: https://gerrit.libreoffice.org/69789
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
look for places we are doing code like:
Reference<XProperty>(model, css::uno::UNO_QUERY)->getAsProperty()
which might result in a SIGSEGV is the query fails
Change-Id: I5cbdbc9e64bd0bed588297c512bf60cbacb9442e
Reviewed-on: https://gerrit.libreoffice.org/69044
Tested-by: Jenkins
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
In my macOS build, that clang::tooling::runToolOnCodeWithArgs invocation failed
to find headers like cassert and assert.h, which works now with
COMPILER_PLUGINS_TOOLING_ARGS=-isysroot /Applications/Xcode.app/Contents/Developer/Platforms/MacOSX.platform/Developer/SDKs/MacOSX10.14.sdk -isystem /Users/stephan/Software/llvm/inst/include/c++/v1
added to my autogen.input (I build against my Clang trunk libc++ whose headers
are at /Users/stephan/Software/llvm/inst/include/c++/v1).
Change-Id: Idbffa39c9fd4a88743fd498b8f7b6c9c56d7630d
Reviewed-on: https://gerrit.libreoffice.org/69538
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
I'm not seeing as much as I would expect here, mostly because pahole
seems to be having trouble parsing quite a few of our structures, and
consequently producing useless data than I then ignore.
XDash 24bytes -> 20bytes
vcl::font::FeatureDefinition 64bytes -> 56bytes
SvXMLTokenMapEntry 16bytes -> 12bytes
SvXMLItemMapEntry 16bytes -> 12bytes
SwContentAtPos 40bytes -> 32bytes
Change-Id: I74c8b93f74b8352f48ef552d7d4239aa7f4237d4
Reviewed-on: https://gerrit.libreoffice.org/69304
Tested-by: Jenkins
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
Create a new script to scan our codebase for holes in our
structures.
Implemented a couple of the things I found
ScSortKeyState 12bytes -> 8bytes
sc::MultiDataCellState 12bytes -> 8bytes
Change-Id: I139dda36aedf02b7f19be121eb312e5552142b87
Just in case there's a typo or something, without this that source
would not be compiled but shared visitor would not include that
functionality either.
Change-Id: I5d9fad129b83f24fed2613bc071cb33e149c1db3
Reviewed-on: https://gerrit.libreoffice.org/69093
Tested-by: Jenkins
Reviewed-by: Luboš Luňák <l.lunak@collabora.com>
Each plugin currently uses its own recursive AST run, which adds up.
This patch adds another shared plugin which internally contains all
(suitable) plugins and dispatches to them from the same one recursive
run. This patch converts ~25 plugins and for starmath's accessibility.cxx
reduces clang build time from 5.43s to 5.14s (and it's 4.39s without any
plugins). As there are almost 50 more plugins to go, this can theoretically
result in 4.56s final time, although probably not all plugins can be
that easily converted, if at all.
This mostly requires very little change in many plugins (see e.g.
BadStatics), some even work without any functionality change (e.g.
CharRightShift). Traverse* calls require some changes but are often
not that difficult. WalkUp* probably can't be supported, although some
plugins can(?) possibly be adjusted to not rely on them. And of course
some plugins can be left as they are, using their own recursive run.
See description at the top of generator.cxx for description of how to
convert a plugin.
The sharedvisitor.cxx source is generated based on scanning relevant
plugin sources using a clang-based scanner/generator. The generated
source is intentionally included instead of getting always generated,
as the generating currently takes some time, so it should get updated
in git whenever a change in a plugin triggers a source change in it.
Change-Id: Ia0d2e3a5a464659503dbb4ed6c20b6cc89b4de01
Reviewed-on: https://gerrit.libreoffice.org/68026
Tested-by: Jenkins
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
Reviewed-by: Luboš Luňák <l.lunak@collabora.com>
As far as I can tell these do not need to scan everything, so the filtering
(and faster) base plugin is sufficient.
Change-Id: I0912a730bb5ec9670d1faa5c83f573902118cc95
Reviewed-on: https://gerrit.libreoffice.org/68872
Tested-by: Jenkins
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
detect static variables that can be made const.
Thanks to mike kaganski for suggesting this.
Here I introduce a new plugin feature - using markers
in nearby comments to disable the plugin for specific
vars.
Some of this stuff was old debugging code. I removed the stuff
that was older than 5 years.
Change-Id: I6ec7742a7fdadf28fd128b592fcdf6da8257585c
Reviewed-on: https://gerrit.libreoffice.org/68807
Tested-by: Jenkins
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
