https://gcc.gnu.org/bugzilla/show_bug.cgi?id=122826
Bug ID: 122826
Summary: OpenMP SIMD call vectorization has incomplete SLP
support for linear() and uniform() clauses
Product: gcc
Version: 16.0
Status: UNCONFIRMED
Severity: normal
Priority: P3
Component: tree-optimization
Assignee: unassigned at gcc dot gnu.org
Reporter: rguenth at gcc dot gnu.org
Target Milestone: ---
vectorizable_simd_clone_call analyzes only arguments of the first call in the
SLP node (actually that of the representative). That should result in
wrong-code for uniform() and linear().
No testcase but I'm testing a workaround (disable multi-lane SLP for those
cases).