| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
|
| |
For creating/destroying nodes dynamically.
Change-Id: I6328bc2f71d30c43b54f10a8f5667c3ac70f0153
Reviewed-by: Paul Lemire <paul.lemire@kdab.com>
|
|
|
|
|
| |
Change-Id: I0dcbe1a8d805d027f1322fcab160b1df737d72d5
Reviewed-by: Paul Lemire <paul.lemire@kdab.com>
|
|
|
|
|
| |
Change-Id: I886de4405c422ef464657636fb214ac500a8364b
Reviewed-by: Paul Lemire <paul.lemire@kdab.com>
|
|
|
|
|
| |
Change-Id: I27607880f6bf411e55cbf2416e6da36fa38e84ef
Reviewed-by: Paul Lemire <paul.lemire@kdab.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Now that we also have QNodeCreatedChange/QNodeDestroyedChange we need
to protect against casting to incorrect types.
It may well make sense to have a helper on QBackendNode that calls a
specific virtual for the different change types. This would be
analogous to QObject::event() calling e.g. resizeEvent() with an
argument properly cast to the right type.
Change-Id: I5313ba43c39b6e47f3fdf1f8f26f3e6d57a5677c
Reviewed-by: Paul Lemire <paul.lemire@kdab.com>
|
|
|
|
|
|
|
|
| |
Added some todo's to investigate folding the scene updates into the
tree traversal used to build the creation/destruction events.
Change-Id: Ia05bca989bc5f44d29e85d9a09c3d25cf32f0776
Reviewed-by: Paul Lemire <paul.lemire@kdab.com>
|
|
|
|
|
|
|
| |
...and not using cloning.
Change-Id: I768f2bfc93b8a2654535056bc918aae1c3a89fa1
Reviewed-by: Paul Lemire <paul.lemire@kdab.com>
|
|
|
|
|
|
|
|
| |
Need to do similar for destruction. Scope to optimize delivery of
creation events noted in a TODO.
Change-Id: I1076824f7e8a6048cdf3060d6a9df5f06361b8ef
Reviewed-by: Paul Lemire <paul.lemire@kdab.com>
|
|
|
|
|
|
|
|
|
| |
This can then be used by the QAspectEngine to create creation changes
at startup and by QNodePrivate to generate creation changes for dynamic
objects at runtime.
Change-Id: Idf13fc249de7238a8c5803634a825bd86bc3c52a
Reviewed-by: Paul Lemire <paul.lemire@kdab.com>
|
|
|
|
|
| |
Change-Id: I13435b356a2466daa37e5913b4889f7be233591e
Reviewed-by: Paul Lemire <paul.lemire@kdab.com>
|
|
|
|
|
| |
Change-Id: If2d290c15f8d7ffaed40083ecc6a5daf5686d274
Reviewed-by: Paul Lemire <paul.lemire@kdab.com>
|
|
|
|
|
|
|
|
| |
This will allow to try out the no cloning approach on hotter code paths
without readign the environment variable every time.
Change-Id: I5cd4532ef942496640b09984784eb80e2203ebce
Reviewed-by: Paul Lemire <paul.lemire@kdab.com>
|
|
|
|
|
|
|
|
| |
Taking care not to do this recursively. One event can take care of
notifying the backend of an entire sub-tree being destroyed.
Change-Id: Iabf849317943b982eebac72e201e117c2f174026
Reviewed-by: Paul Lemire <paul.lemire@kdab.com>
|
|
|
|
|
| |
Change-Id: I3b753b529f7cae5b839620598a8414fb78589451
Reviewed-by: Paul Lemire <paul.lemire@kdab.com>
|
|
|
|
|
|
|
|
|
|
| |
Now non-window sized viewports work, too.
Task-number: QTBUG-52498
Change-Id: Icc4aba3f93d78b4266a3b565156fe8799797baef
Reviewed-by: Volker Krause <volker.krause@kdab.com>
Reviewed-by: Antti Määttä <antti.maatta@theqtcompany.com>
Reviewed-by: Sean Harmer <sean.harmer@kdab.com>
|
|
|
|
|
|
|
|
|
|
|
|
| |
Also improve the alpha to coverage handlign in the helpers by using
a single function rather than two.
This removes GL errors about bad glFace enum values (caused by bad
handlign of alpha to coverage in switch statement, which led to a
FrontFace state being created with bad data).
Change-Id: I832d92fcb07134525f421d3a3b4d12e9f5f6428c
Reviewed-by: Paul Lemire <paul.lemire@kdab.com>
|
|
|
|
|
|
|
|
|
|
|
|
| |
This ensures that subclasses of QNodeCreatedChangeBase always pass in
the QNode pointer that allows to correctly set the subjectId to be the
id() of the QNode.
With this change the materials-cpp example is now able to run using the
new node creation mechanism.
Change-Id: I0a5ce4f3e93b485adfdb7534c15493946cfa00c6
Reviewed-by: Paul Lemire <paul.lemire@kdab.com>
|
|
|
|
|
|
|
| |
Also remove last vestiges of blend state separate.
Change-Id: I75d3a97ae8383b5ad1bcdb903ed745cbcda31eec
Reviewed-by: Paul Lemire <paul.lemire@kdab.com>
|
|
|
|
|
|
|
|
| |
The genericstate_p.h has gotten a little large and contains
several classes.
Change-Id: I76126a300afaa4dda4686fb231dd9d8c9cc68fab
Reviewed-by: Paul Lemire <paul.lemire@kdab.com>
|
|
|
|
|
| |
Change-Id: I19965c493c0eac74c9f86a74075c113ba499d34e
Reviewed-by: Paul Lemire <paul.lemire@kdab.com>
|
|
|
|
|
| |
Change-Id: I4de6a94381e67be0b98e4786d057196e3729b501
Reviewed-by: Paul Lemire <paul.lemire@kdab.com>
|
|
|
|
|
| |
Change-Id: I601adac5931ec845f4ac134d24d6b8bc02dc3989
Reviewed-by: Paul Lemire <paul.lemire@kdab.com>
|
|
|
|
|
| |
Change-Id: I5e476276cb7b7f1e12859df708be5fe3077419bc
Reviewed-by: Paul Lemire <paul.lemire@kdab.com>
|
|
|
|
|
| |
Change-Id: Iaa7166d38424b3a987d56c3bf1a1559e9588031a
Reviewed-by: Paul Lemire <paul.lemire@kdab.com>
|
|
|
|
|
| |
Change-Id: If0ffa048913fd8d90e1b53235a1d2a2c6fe2b555
Reviewed-by: Paul Lemire <paul.lemire@kdab.com>
|
|
|
|
|
| |
Change-Id: If7cf27c255d7b307ada564618baa44f07596ecc1
Reviewed-by: Paul Lemire <paul.lemire@kdab.com>
|
|
|
|
|
| |
Change-Id: I569ced3777ee0963e5471d1a8f623d8cfaaa900e
Reviewed-by: Paul Lemire <paul.lemire@kdab.com>
|
|
|
|
|
| |
Change-Id: Ib06ce379ad2aaa6abca4f5d0f6c3301ad0c59577
Reviewed-by: Paul Lemire <paul.lemire@kdab.com>
|
|
|
|
|
| |
Change-Id: I1f7673e8494c2dd43a874c101f0d806dfdb3d02c
Reviewed-by: Paul Lemire <paul.lemire@kdab.com>
|
|
|
|
|
| |
Change-Id: Ie7c2490664a7efc7704b547f9fa7bf92d29653d5
Reviewed-by: Paul Lemire <paul.lemire@kdab.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
We should change the QBlendEquationSeparate class so that it does not
inherit from this one. That would then allow us to simplify this class
to not contain the alpha arguments. Alternatively, we could investigate
folding the separate args version into this by introducing a NotSet
value that we default the alpha arguments to. If they are NotSet then
we use the non-separate blending functions on the backend. If they are
set then we use the separate Rgb + Alpha blending functionality.
Change-Id: I14993b338228d120c6309536b0a286bed9ed870c
Reviewed-by: Paul Lemire <paul.lemire@kdab.com>
|
|
|
|
|
| |
Change-Id: I5abb94ebf2dd66114d4fa6cc5cd783f6725e8a15
Reviewed-by: Paul Lemire <paul.lemire@kdab.com>
|
|
|
|
|
| |
Change-Id: I3c12a75146fc8753f6c8dbfc2026dffa896fb67f
Reviewed-by: Paul Lemire <paul.lemire@kdab.com>
|
|
|
|
|
|
|
|
| |
Will be used to pass renderstate type to the backend so that it can
be cast to the correct payload type.
Change-Id: I94ccc8bb90425b2f9dd1cb50ceb3de2813d68b0b
Reviewed-by: Paul Lemire <paul.lemire@kdab.com>
|
|
|
|
|
|
|
|
|
|
| |
With this commit, the renderer is able to draw the cylinder-cpp example
with the QT3D_NO_CLONE environment variable set!
Input is not yet working again.
Change-Id: I6df29546559591ed9f9c1d3e4c150227bdeb7798
Reviewed-by: Paul Lemire <paul.lemire@kdab.com>
|
|
|
|
|
| |
Change-Id: Ib6a7ec403342fc67b9055383ba5d07d47a4d7ea7
Reviewed-by: Paul Lemire <paul.lemire@kdab.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Move this out of the creation functor as the peerId is not set at this
point which breaks the setParentId() function. Move this to the
initializeFromPeer() function and have each FrameGraphNode type call
the base class implementation to set the basic stuff up.
With this change, the main loop now runs through using the QT3D_NO_CLONE
env var. However, it does not render again, yet.
Change-Id: Ied12deb4732b5c103e647cc4cfea935114ea31ca
Reviewed-by: Paul Lemire <paul.lemire@kdab.com>
|
|
|
|
|
|
|
|
|
|
| |
To make the id explicit rather than the function extracting it from
the QBackendNode. This is because under the no cloning approach the
QBackendNode does not yet have its peerId() set at this stage. We do
however have it from the change itself.
Change-Id: I2dc1607088b8e8039094755529f6afb127b607f8
Reviewed-by: Sean Harmer <sean.harmer@kdab.com>
|
|
|
|
|
|
|
| |
Also remove obsolete comment.
Change-Id: Ic2b0f77ba318d2edc272fa12dc88ebab868224eb
Reviewed-by: Paul Lemire <paul.lemire@kdab.com>
|
|
|
|
|
| |
Change-Id: I5d430a3ec2d5e18219d5e16fd9819b0ffac1e650
Reviewed-by: Paul Lemire <paul.lemire@kdab.com>
|
|
|
|
|
| |
Change-Id: Ib356c3b960ad7ef9aa937ca85e5f46b69e5cff31
Reviewed-by: Paul Lemire <paul.lemire@kdab.com>
|
|
|
|
|
| |
Change-Id: I2565b933c7319130ba7eee6e9316df2d12f5936e
Reviewed-by: Paul Lemire <paul.lemire@kdab.com>
|
|
|
|
|
|
|
| |
This is now on the clear buffers node.
Change-Id: I0fec56d1596523ddb4121a9a39a5a3a47e071c31
Reviewed-by: Paul Lemire <paul.lemire@kdab.com>
|
|
|
|
|
| |
Change-Id: I70403ff6a7f028800e81b8c7c4f9738c3328f7c5
Reviewed-by: Paul Lemire <paul.lemire@kdab.com>
|
|
|
|
|
| |
Change-Id: I96049adc485f503e66232f0dc8c13b121eb69813
Reviewed-by: Paul Lemire <paul.lemire@kdab.com>
|
|
|
|
|
|
|
|
| |
Added a TODO about rethinking the approach for setting up the
relationship between Textures and TextureImages.
Change-Id: I97f6d4ee5194e2b255c06a0f236f7fdbe265cc5b
Reviewed-by: Paul Lemire <paul.lemire@kdab.com>
|
|
|
|
|
| |
Change-Id: Ia164f44c53047c258c9e256fee20df0c8ecb9496
Reviewed-by: Paul Lemire <paul.lemire@kdab.com>
|
|
|
|
|
|
|
| |
Backend type handles them too.
Change-Id: Ibd87c2e5a464c0953c1b017db21e7831ca4c6681
Reviewed-by: Paul Lemire <paul.lemire@kdab.com>
|
|
|
|
|
| |
Change-Id: If20053bba9c494673e6768ee9bf9b15fb9665bf9
Reviewed-by: Paul Lemire <paul.lemire@kdab.com>
|
|
|
|
|
| |
Change-Id: I0705d14a4d7c5a9ce6df5274ff8e59414f55e695
Reviewed-by: Paul Lemire <paul.lemire@kdab.com>
|