OpenCL Parallel compilation only works inside Blender. When using cycles in a different setup (standaline or other software) it failed compiling kernels as they don't have the appropriate Python API and command line arguments.
This change introduces a `running_inside_blender` debug flag, that triggers out of process compilation of the kernels. Compilation still happens in subthread that enabled the preview kernels and compilation of the kernels during BVH building
Reviewed By: brecht
Differential Revision: https://developer.blender.org/D5439
This patch stitches the vertices along patch edges so that cracks can
no longer form when applying subdivision or displacement a mesh.
Subpatches are now formed in a way that ensures vertex indices along
subpatch edges are equal for adjacent subpatches. A mapping of vertices
along patch edges is built to preform stitching. Overall performance is
roughly the same, some gains were made in splitting, but some was lost
in stitching.
This fixes:
- T49049 (cracks between patches from material and uv seams)
- T49048 (discontinuous normals with true displacement)
Reviewers: sergey, brecht
Differential Revision: https://developer.blender.org/D3692
This will be used by Optix to help the compiler figure out scoping. It is not
used by other devices currently, but worth testing if it helps there too.
Ref D5363
Before there were two options: Paste to original layer called "Paste" and Paste to active layer called "Paste & Merge"
Now, by default the paste is in active layer and the "Paste & Merge" has been renamed "Paste".
For old "Paste", now is called "Paste by Layer" and it's not the default value anymore.
Note: Minor edits to add icons not present in Differential revision.
Differential Revision: https://developer.blender.org/D5591
x64 builds with WITH_CYCLES_OPTIMIZED_KERNEL_SSE2 not defined
since SSE2 is the lower bar for x64 cpus. Turning the architecture
logging related if into the last if in the architecture detection
chain, which will never execute unless you turn off all kernels
in de debug flags.
Reviewers: brecht
Differential Revision: https://developer.blender.org/D5579
Use a single loop to iterate over all lights, reducing divergence and amount
of code to generate. Moving ray intersection calls out of conditionals will
also help the Optix compiler.
Ref D5363
The old layout of `PointerRNA` was confusing for historic reasons:
```
typedef struct PointerRNA {
struct {
void *data;
} id;
struct StructRNA *type;
void *data;
} PointerRNA;
```
This patch updates it to:
```
typedef struct PointerRNA {
struct ID *owner_id;
struct StructRNA *type;
void *data;
} PointerRNA;
```
Throughout the code base `id.data` was replaced with `owner_id`.
Furthermore, many explicit pointer type casts were added which
were implicit before. Some type casts to `ID *` were removed.
Reviewers: brecht, campbellbarton
Differential Revision: https://developer.blender.org/D5558
The object color property is added as an additional output in
the Object Info node.
Reviewers: brecht
Differential Revision: https://developer.blender.org/D5554
The fract function in OpenCL does more than just return the fraction.
It also writes the floor to the second argument. Which wasn't put
in consideration.
Instead, we use a simple `a - floor(a)` like the Math node.
Reviewers: brecht
Differential Revision: https://developer.blender.org/D5553
The Volume Info node provides the Color, Desnity, Flame, and Temperature
of smoke domains.
Reviewers: brecht
Differential Revision: https://developer.blender.org/D5551
The White Noise node hashes the input and returns a random number in the
range [0, 1]. The input can be a 1D, 2D, 3D, or a 4D vector.
Reviewers: brecht, JacquesLucke
Differential Revision: https://developer.blender.org/D5550
Add Multiply, Divide, Project, Reflect, Distance, Length, Scale, Snap,
Floor, Ceil, Modulo, Fraction, Absolute, Minimum, and Maximum operators
to the Vector Math node. The Value output has been removed from operators
whose output is a vector, and the other way around. All of those removals
has been handled properly in versioning code.
The patch doesn't include tests for the new operators. Tests will be added
in a later patch.
Reviewers: brecht, JacquesLucke
Differential Revision: https://developer.blender.org/D5523
Previously eigens internal include order somehow implicitly provided
M_PI and friends. The recent eigen version bump broke this implicit
behaviour, better to be explicit that we need the math defines for MSVC.
The clamp option is implemented using graph expansion, where a Clamp node
named "clamp" is added and connected to the output. So the final result
is actually from the node "clamp".
Reviewers: brecht
Differential Revision: https://developer.blender.org/D5540
When calling `MEM_guarded_mallocN_aligned` with an alignment of 4,
a pointer that was returned that is 4 byte but not 8 byte aligned.
When freeing this pointer, `MEM_guarded_freeN` thinks that it is an
illegal pointer, because it asserts that `((intptr_t)memh) & 0x7 == 0`.
The fix is to always use at least 8 byte alignment.
Reviewers: brecht
Differential Revision: https://developer.blender.org/D5529
- Implement dynamic inputs. The second input is now unavailable in single
operand math operators.
- Reimplemenet the clamp option using graph expansion for Cycles.
- Clean up code and unify naming between Blender and Cycles.
- Remove unused code.
Reviewers: brecht
Differential Revision: https://developer.blender.org/D5481
This code is needed to implement the Voxel Remesher as well as other features like a better remesh modifier with filters and CSG operations.
Done by Martin Felke and Pablo Dobarro
Reviewed By: brecht
Differential Revision: https://developer.blender.org/D5364
This could happen e.g. when changing smoke type from flow to domain or
connecting a volume shader with to a domain without an actual flow type
around.
Fixes T58569, T68359
Reviewers: brecht
Maniphest Tasks: T58569, T68359
Differential Revision: https://developer.blender.org/D5478
If the option is enabled, the output is clamped to the target range.
The target range is [To Min, To Max]. The option is enabled by default.
The clamp option is implemented in EEVEE by linking to the `clamp_value`
GLSL function. And it is implemented in Cycles using a graph expand
function.
Reviewers: brecht, JacquesLucke
Differential Revision: https://developer.blender.org/D5477
This patch adds a new node that clamps a value between a maximum and
a minimum values.
Reviewers: brecht
Differential Revision: https://developer.blender.org/D5476
This patch adds a new Map Range node that linearly remaps an input
value from a range to another. This node is similar to the compositor's
Map Range node.
Reviewers: brecht, JacquesLucke
Differential Revision: https://developer.blender.org/D5471
We can add more fine grained checks for when these flags are supported so
that adding asan flags manually still has all the workarounds, but for now
compiling succesfully is more important.
