public inbox for gcc-cvs@sourceware.org
help / color / mirror / Atom feed
* [gcc/devel/omp/gcc-12] vect: WORKAROUND vectorizer bug
@ 2022-10-24 16:19 Andrew Stubbs
  0 siblings, 0 replies; only message in thread
From: Andrew Stubbs @ 2022-10-24 16:19 UTC (permalink / raw)
  To: gcc-cvs

https://gcc.gnu.org/g:2a0dc45c854448ec4628dab5c6c08999a5cf3fe5

commit 2a0dc45c854448ec4628dab5c6c08999a5cf3fe5
Author: Andrew Stubbs <ams@codesourcery.com>
Date:   Fri Oct 21 14:19:31 2022 +0100

    vect: WORKAROUND vectorizer bug
    
    This patch disables vectorization of memory accesses to non-default address
    spaces where the pointer size is different to the usual pointer size.  This
    condition typically occurs in OpenACC programs on amdgcn, where LDS memory is
    used for broadcasting gang-private variables between threads. In particular,
    see libgomp.oacc-c-c++-common/private-variables.c
    
    The problem is that the address space information is dropped from the various
    types in the middle-end and eventually it triggers an ICE trying to do an
    address conversion.  That ICE can be avoided by defining
    POINTERS_EXTEND_UNSIGNED, but that just produces wrong RTL code later on.
    
    A correct solution would ensure that all the vectypes have the correct address
    spaces, but I don't have time for that right now.
    
    gcc/ChangeLog:
    
            * tree-vect-data-refs.cc (vect_analyze_data_refs): Workaround an
            address-space bug.

Diff:
---
 gcc/ChangeLog.omp          |  5 +++++
 gcc/tree-vect-data-refs.cc | 16 +++++++++++++++-
 2 files changed, 20 insertions(+), 1 deletion(-)

diff --git a/gcc/ChangeLog.omp b/gcc/ChangeLog.omp
index d7b693ac830..68d5eea01a7 100644
--- a/gcc/ChangeLog.omp
+++ b/gcc/ChangeLog.omp
@@ -1,3 +1,8 @@
+2022-10-24  Andrew Stubbs  <ams@codesourcery.com>
+
+	* tree-vect-data-refs.cc (vect_analyze_data_refs): Workaround an
+	address-space bug.
+
 2022-10-24  Andrew Stubbs  <ams@codesourcery.com>
 
 	* config/gcn/gcn.cc (gcn_init_cumulative_args): Disallow gfx908.
diff --git a/gcc/tree-vect-data-refs.cc b/gcc/tree-vect-data-refs.cc
index 09223baf718..70b671ed94a 100644
--- a/gcc/tree-vect-data-refs.cc
+++ b/gcc/tree-vect-data-refs.cc
@@ -4598,7 +4598,21 @@ vect_analyze_data_refs (vec_info *vinfo, poly_uint64 *min_vf, bool *fatal)
       /* Set vectype for STMT.  */
       scalar_type = TREE_TYPE (DR_REF (dr));
       tree vectype = get_vectype_for_scalar_type (vinfo, scalar_type);
-      if (!vectype)
+
+      /* FIXME: If the object is in an address-space in which the pointer size
+	 is different to the default address space then vectorizing here will
+	 lead to an ICE down the road because the address space information
+	 gets lost.  This work-around fixes the problem until we have a proper
+	 solution.  */
+      tree base_object = DR_REF (dr);
+      tree op = (TREE_CODE (base_object) == COMPONENT_REF
+		 || TREE_CODE (base_object) == ARRAY_REF
+		 ? TREE_OPERAND (base_object, 0) : base_object);
+      addr_space_t as = TYPE_ADDR_SPACE (TREE_TYPE (op));
+      bool addr_space_bug = (!ADDR_SPACE_GENERIC_P (as)
+			     && targetm.addr_space.pointer_mode (as) != Pmode);
+
+      if (!vectype || addr_space_bug)
         {
           if (dump_enabled_p ())
             {

^ permalink raw reply	[flat|nested] only message in thread

only message in thread, other threads:[~2022-10-24 16:19 UTC | newest]

Thread overview: (only message) (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2022-10-24 16:19 [gcc/devel/omp/gcc-12] vect: WORKAROUND vectorizer bug Andrew Stubbs

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).