Some intel drivers crash when areaScale shader with "large" array
is used. This adds a "reduced register" version of the areaScale
shader. We still use the first version of the shader for other
drivers and switch between the 2 implementations with a runtime
detection.
Change-Id: I1860f898c03b40a600eb1b41f7262719382a7171
Reviewed-on: https://gerrit.libreoffice.org/30571
Reviewed-by: Tomaž Vajngerl <quikee@gmail.com>
Tested-by: Tomaž Vajngerl <quikee@gmail.com>
This reverts commit 81e3f5f2fdc9c573c83a37009080e4bb974c7955; had failed to
notice that there are already (Scheduler const & overloads of) copy member
functions (which MSVC complains about, but Clang didn't).
since we only care about appending to this container, and then
traversing it (normally once). So reduce the re-allocations that
std::vector requires
Change-Id: I206a7b82d9eefc1fa3762c4a03e7b5e21136951f
Reviewed-on: https://gerrit.libreoffice.org/30706
Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
So we don't have to specify the source and destination type as often.
Change-Id: Id9e286417a1cb246d163cbc3c536b231a4a92624
Reviewed-on: https://gerrit.libreoffice.org/30700
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
Tested-by: Noel Grandin <noel.grandin@collabora.co.uk>
At least on my macOS 10.12.1, PTHREAD_STACK_MIN is 8192 (so stacksize was
substantially smaller than 12MB), and without this change there were randomly
looking failures in e.g. JunitTest_chart2_unoapi with a -fsanitize=address
build.
Change-Id: Icfe989a0e5097a9a0ae76c6e0f6ffcca18271245
for tab cycling, given the problem of tdf#103783
regression from...
commit fd40e3ea09a481a2e97452d2244ecdfd6c3508c1
Author: Caolán McNamara <caolanm@redhat.com>
Date: Tue Nov 1 16:34:22 2016 +0000
lock tab-cycling inside the WB_DIALOGCONTROL widget
activate the color selector in the toolbar in the sidebar panel, press tab
to cycle between its widgets, when it hits the bottom it will
not return to the first widget, but escape out of the floating
window into the next sibling of the toolbar
but only check the Parent, not RealParent
Change-Id: I9577a0a9f3a622373c195e77ca43d17511b4154d
Project: help 578e1e320daaa82b2d3b8c0eeea84bf704967939
tdf#103397 - added example for Transpose function
Added a simple example to the help for transpose() function
Added tablehead style to example rows/columns
Change-Id: I464983d480cb237a53e033005057a1faef0b9193
Reviewed-on: https://gerrit.libreoffice.org/30334
Reviewed-by: Olivier Hallot <olivier.hallot@edx.srv.br>
Tested-by: Olivier Hallot <olivier.hallot@edx.srv.br>
These belonged to extra galleries, introduced in
49d5bdcf4d516b21721f1ed6d313ea4a170d39eb
I missed to remove these in
8dc38d9ded92960eda46343b67da8a8b95404b5a
Change-Id: I32c649d56053b32f2c2993aeda51425d7586c9b7
Reviewed-on: https://gerrit.libreoffice.org/30698
Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Samuel Mehrbrodt <Samuel.Mehrbrodt@cib.de>
In case our xref table doesn't have an entry for "free" object types,
then the table size won't provide a valid id for a next object. That
resulted in creating all new objects with the same ID.
With this, our verifier at least can see the new signature when
appending one to a signed PDF 1.6 file.
Change-Id: Iac39a400706cfcd23dd814d2b81cb8b950c69fc6
Reviewed-on: https://gerrit.libreoffice.org/30704
Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Miklos Vajna <vmiklos@collabora.co.uk>
Don't give up signing just because PDF 1.4 trailer is missing, provided
that PDF 1.5 xref stream is available.
Change-Id: I03360d428346537583a4398aa3a94b195b428713
Reviewed-on: https://gerrit.libreoffice.org/30703
Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Miklos Vajna <vmiklos@collabora.co.uk>
This will need cross-reference stream write support, just don't crash
for now.
Change-Id: Id48c131b22d4ed96174693f3e96b14c273d596a8
Reviewed-on: https://gerrit.libreoffice.org/30702
Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Miklos Vajna <vmiklos@collabora.co.uk>
...if that should ever happen. Also means that
ToolBarManager::RemoveControllers will now only be called with m_bDispose being
false.
Change-Id: If17d60f291cc8dad19d6f6e33d39e08c7f8ad7aa
No need to mention licensing info of OxygenOffice-related
extensions that were dropped in
8dc38d9ded92960eda46343b67da8a8b95404b5a
Change-Id: I12b862433b62154078d88f6debac8ecae749a751
Reviewed-on: https://gerrit.libreoffice.org/30699
Reviewed-by: Michael Meeks <michael.meeks@collabora.com>
Tested-by: Michael Meeks <michael.meeks@collabora.com>
Follow up to 8dc38d9ded92960eda46343b67da8a8b95404b5a
Here we drop some --help descriptions and the rest of
removed extensions from the Jenkins config file
Change-Id: Ie1520adf2c48c2dc714b00c53ba0055f1deb3c62
Reviewed-on: https://gerrit.libreoffice.org/30697
Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Miklos Vajna <vmiklos@collabora.co.uk>
The signature manager always creates an XML helper, and optionally creates a
PDF helper as well. Both of them initialize xmlsec, and when the signature
manager is deleted, there are two de-inits, leading to an assertion failure in
xmlsec.
Fix the problem by moving the duplicated xmlsec init to the signature manager.
This has the additional benefit that general security-related code no longer
has to talk to the XML helper, it can use the signature manager, which feels
more natural. (What viewing a certificate had to do with XML?)
Change-Id: If6a6bc433636445f3782849a367d4a7ac0be7688
Reviewed-on: https://gerrit.libreoffice.org/30695
Reviewed-by: Miklos Vajna <vmiklos@collabora.co.uk>
Tested-by: Jenkins <ci@libreoffice.org>
It was intended this way from day one, but the implementation
didn't live up to that promise...
Change-Id: I231ddc8923fdd8a205127c6a3214dd93f13378b9
Reviewed-on: https://gerrit.libreoffice.org/30678
Reviewed-by: Kohei Yoshida <libreoffice@kohei.us>
Tested-by: Kohei Yoshida <libreoffice@kohei.us>
The XSecController::setX509Certificate() overload that took a
sal_Int32 nSecurityEnvironmentIndex was called in just one place, and
-1 was always passed for the nSecurityEnvironmentIndex.
Change-Id: Ic37b7053c90a6fbb8b4e5b6485483d34c605b965
Vim users: No reason to panic. This does not mean that Emacs users
would suddenly start to split existing (or new) code lines always
before column 100. The default value for fill-column is 70, and that
hasn't forced code lines edited in Emacs to be shorter than that
earlier either.
The primary intent of using a fill-column of 100 (instead of the
default 70) is that when you edit some long (multi-line) comment
block, and you want to reformat ("fill", "reflow") that comment (using
the fill-paragraph command, bound to Meta-Q), lines will be filled up
to column 100, and not just 70, which in most cases would look quite
short.
Unless I am strongly advised not to, I will start adding this to the
mode lines in source files if I remember, as I happen to edit some
comment block in them.
Change-Id: Icfb93dbb22b2db7190fdc9c8ee9518d08e73c7a8
When the user uses .uno:SignPDF to open a PDF file in read-only mode,
then the infobar now has a "Sign Document" button, which is a shortcut
for the File -> Digital Signatures -> Digital Signatures menu item.
Change-Id: I2953a5cb1252c5ff1c395d45c54db1a97b39433e
xmlsecurity is such a mess. Too many different classes that actually
have more or less identical life-time, with names that smell of
delusions of grandeur. "Manager", "security framework controller" etc
for stuff that actually exist only during the execution of a simple
dialog. And then a "helper" class that actually in on a higher level
than a "framework controller". But oh well.
Change-Id: I86e461b1bc91a0d8f5b7fb9f13a5be201729df1e
instead of returning a Primitive2DContainer from each method which we
are then going to immediately append to another container, pass down a
single container by reference which we can append to
Change-Id: I0f28a499d2ec54f7111a7044c30099767aa079e1
Reviewed-on: https://gerrit.libreoffice.org/30258
Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
which led to discovering some unused parameters
Change-Id: Ie8f67a47aa9ec409637da32ef9bc457952397772
Reviewed-on: https://gerrit.libreoffice.org/30681
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
Tested-by: Noel Grandin <noel.grandin@collabora.co.uk>
The bug document has a normal table, then its C1 cell starts with a
nested table, which is floating. The problem is that converting the
nested table to a textframe invalidates the start text range of the C1
cell in the outer table we store, so the conversion of the outer table
from text to table fails.
This never worked, so to avoid the regression just don't convert inner
floating tables to textframes when they're anchored at the cell start.
A more general fix in the future can be addressing the actual
invalidation of the cell start/end text ranges, and then this specific
fix will not be necessary anymore.
Change-Id: I12cefa41977cf719b07b0fb3ef9ec423c17ef3b1
Reviewed-on: https://gerrit.libreoffice.org/30685
Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Miklos Vajna <vmiklos@collabora.co.uk>
The underlying design is that there's a single JavaInteractionHandler
instance owned by a JavaContext, and that JavaContext installed in
Desktop::Main (desktop/source/app/app.cxx).
This patch ensures that no additional JavaContext is created in
SfxOfficeDispatch::dispatch*() functions unless they are used without
preinstalled JavaContext.
Thanks to Stephan Bergmann for guidance!
Change-Id: I2569df221067a5b9bf1f6cd5d8f69b561316a170
Reviewed-on: https://gerrit.libreoffice.org/30529
Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
