int __attribute__((__aligned__(8))) a[8]; int __attribute__((__aligned__(8))) b[8]; void test () { a[0] = b[0] + 1; a[1] = b[1] + 2; a[2] = b[2] + 3; a[3] = b[3] + 4; a[4] = b[0] * 3; a[5] = b[2] * 4; a[6] = b[4] * 5; a[7] = b[6] * 7; } should be vectorized using V4SI vectors in two SLP groups so we can vectorize not only the store but also the loads and the add. When using -mavx2 we instead get only the store vectorized (even with cost modeling enabled) because we try vectorizing that first. It might be possible to guide SLP splitting during the SLP build in a similar way how we try commutating operands. So when we figure /home/rguenther/src/gcc3/gcc/testsuite/gcc.dg/vect/bb-slp-subgroups-3.c:12:10: note: Build SLP for _9 = _1 * 3; /home/rguenther/src/gcc3/gcc/testsuite/gcc.dg/vect/bb-slp-subgroups-3.c:12:10: note: get vectype for scalar type (group size 8): int /home/rguenther/src/gcc3/gcc/testsuite/gcc.dg/vect/bb-slp-subgroups-3.c:12:10: note: vectype: vector(8) int /home/rguenther/src/gcc3/gcc/testsuite/gcc.dg/vect/bb-slp-subgroups-3.c:12:10: note: nunits = 8 /home/rguenther/src/gcc3/gcc/testsuite/gcc.dg/vect/bb-slp-subgroups-3.c:12:10: missed: Build SLP failed: different operation in stmt _9 = _1 * 3; /home/rguenther/src/gcc3/gcc/testsuite/gcc.dg/vect/bb-slp-subgroups-3.c:12:10: missed: original stmt _2 = _1 + 1; and see the parent op (the store in this case) cannot be commutated we can see whether matches[] divides the vector with some constraints and whether the other lanes with matches[] == false form a valid SLP operand (we know the == true ones likely would). The results would then be concatenated via a permute node. This should eventually also replace the splitting done in SLP instance analysis (though splitting stores might still be necessary there). The other option is to somehow tackle this with vector size iteration, doing multiple analyses and comparing costs/benefit though it's hard to not compare apples & oranges since the amount of code vectorized will usually differ (as compared to loop vectorization)
Confirmed.
The master branch has been updated by Richard Biener <rguenth@gcc.gnu.org>: https://gcc.gnu.org/g:6390c5047adb75960f86d56582e6322aaa4d9281 commit r12-3893-g6390c5047adb75960f86d56582e6322aaa4d9281 Author: Richard Biener <rguenther@suse.de> Date: Wed Nov 18 09:36:57 2020 +0100 Allow different vector types for stmt groups This allows vectorization (in practice non-loop vectorization) to have a stmt participate in different vector type vectorizations. It allows us to remove vect_update_shared_vectype and replace it by pushing/popping STMT_VINFO_VECTYPE from SLP_TREE_VECTYPE around vect_analyze_stmt and vect_transform_stmt. For data-ref the situation is a bit more complicated since we analyze alignment info with a specific vector type in mind which doesn't play well when that changes. So the bulk of the change is passing down the actual vector type used for a vectorized access to the various accessors of alignment info, first and foremost dr_misalignment but also aligned_access_p, known_alignment_for_access_p, vect_known_alignment_in_bytes and vect_supportable_dr_alignment. I took the liberty to replace ALL_CAPS macro accessors with the lower-case function invocations. The actual changes to the behavior are in dr_misalignment which now is the place factoring in the negative step adjustment as well as handling alignment queries for a vector type with bigger alignment requirements than what we can (or have) analyze(d). vect_slp_analyze_node_alignment makes use of this and upon receiving a vector type with a bigger alingment desire re-analyzes the DR with respect to it but keeps an older more precise result if possible. In this context it might be possible to do the analysis just once but instead of analyzing with respect to a specific desired alignment look for the biggest alignment we can compute a not unknown alignment. The ChangeLog includes the functional changes but not the bulk due to the alignment accessor API changes - I hope that's something good. 2021-09-17 Richard Biener <rguenther@suse.de> PR tree-optimization/97351 PR tree-optimization/97352 PR tree-optimization/82426 * tree-vectorizer.h (dr_misalignment): Add vector type argument. (aligned_access_p): Likewise. (known_alignment_for_access_p): Likewise. (vect_supportable_dr_alignment): Likewise. (vect_known_alignment_in_bytes): Likewise. Refactor. (DR_MISALIGNMENT): Remove. (vect_update_shared_vectype): Likewise. * tree-vect-data-refs.c (dr_misalignment): Refactor, handle a vector type with larger alignment requirement and apply the negative step adjustment here. (vect_calculate_target_alignment): Remove. (vect_compute_data_ref_alignment): Get explicit vector type argument, do not apply a negative step alignment adjustment here. (vect_slp_analyze_node_alignment): Re-analyze alignment when we re-visit the DR with a bigger desired alignment but keep more precise results from smaller alignments. * tree-vect-slp.c (vect_update_shared_vectype): Remove. (vect_slp_analyze_node_operations_1): Do not update the shared vector type on stmts. * tree-vect-stmts.c (vect_analyze_stmt): Push/pop the vector type of an SLP node to the representative stmt-info. (vect_transform_stmt): Likewise. * gcc.target/i386/vect-pr82426.c: New testcase. * gcc.target/i386/vect-pr97352.c: Likewise.
Fixed for GCC 12.