getBaseAlignment: Use the rowMajor argument for determining matrix strides

The argument version is passed in from above and struct handling ensures that row-majorness gets propagated correctly from one level to the next. If we just look at the type itself and it's embedded in a struct that's declared row-major, we may get the wrong stride.

getBaseAlignment: Use the rowMajor argument for determining matrix strides
a76766a4 · Jason Ekstrand · 79b7046a · a76766a4
Commit a76766a4 authored Jan 13, 2016 by Jason Ekstrand
Hide whitespace changes
Inline Side-by-side

Showing with 1 addition and 1 deletion

linkValidate.cpp glslang/MachineIndependent/linkValidate.cpp +1 -1

No files found.
--- a/glslang/MachineIndependent/linkValidate.cpp
+++ b/glslang/MachineIndependent/linkValidate.cpp
@@ -987,7 +987,7 @@ int TIntermediate::getBaseAlignment(const TType& type, int& size, int& stride, b
    // rules 5 and 7
    if (type.isMatrix()) {
        // rule 5: deref to row, not to column, meaning the size of vector is num columns instead of num rows
-        TType derefType(type, 0, type.getQualifier().layoutMatrix == ElmRowMajor);
+        TType derefType(type, 0, rowMajor);
            
        alignment = getBaseAlignment(derefType, size, dummyStride, std140, rowMajor);
        if (std140)