public inbox for gcc-cvs@sourceware.org
help / color / mirror / Atom feed
* [gcc r13-5625] ree: Fix -fcompare-debug issues in combine_reaching_defs [PR108573]
@ 2023-02-01 11:53 Jakub Jelinek
  0 siblings, 0 replies; only message in thread
From: Jakub Jelinek @ 2023-02-01 11:53 UTC (permalink / raw)
  To: gcc-cvs

https://gcc.gnu.org/g:e4473d7cf871c8ddf8f22d105c5af6375ebe37bf

commit r13-5625-ge4473d7cf871c8ddf8f22d105c5af6375ebe37bf
Author: Jakub Jelinek <jakub@redhat.com>
Date:   Wed Feb 1 12:52:52 2023 +0100

    ree: Fix -fcompare-debug issues in combine_reaching_defs [PR108573]
    
    The PR78437 r7-4871 changes made combine_reaching_defs punt on
    WORD_REGISTER_OPERATIONS targets if a setter of smaller than word
    register has wider uses.  This unfortunately breaks -fcompare-debug,
    because if such a use appears only in DEBUG_INSN(s), while all other
    uses aren't wider than the setter, we can REE optimize it without -g
    and not with -g.
    
    Such decisions shouldn't be based on debug instructions.  We could try
    to reset them or adjust in some other way after we decide to perform the
    change, but at least on the testcase which used to fail on riscv64-linux
    the
    (debug_insn 8 7 9 2 (var_location:HI s (minus:HI (subreg:HI (and:DI (reg:DI 10 a0 [160])
                    (const_int 1 [0x1])) 0)
            (subreg:HI (ashiftrt:DI (reg/v:DI 9 s1 [orig:151 l ] [151])
                    (debug_expr:SI D#1)) 0))) "pr108573.c":12:5 -1
         (nil))
    clearly doesn't care about the upper bits and I have hard time imaging how
    could one end up with DEBUG_INSN which actually cares about those upper
    bits.
    
    So, the following patch just ignores uses on DEBUG_INSNs in this case,
    if we run into something where we'd need to do something further later on,
    let's deal with it when we have a testcase for it.
    
    2023-02-01  Jakub Jelinek  <jakub@redhat.com>
    
            PR debug/108573
            * ree.cc (combine_reaching_defs): Don't return false for paradoxical
            subregs in DEBUG_INSNs.
    
            * gcc.dg/pr108573.c: New test.

Diff:
---
 gcc/ree.cc                      |  6 ++++--
 gcc/testsuite/gcc.dg/pr108573.c | 18 ++++++++++++++++++
 2 files changed, 22 insertions(+), 2 deletions(-)

diff --git a/gcc/ree.cc b/gcc/ree.cc
index 98bd80efd7b..413aec7c8eb 100644
--- a/gcc/ree.cc
+++ b/gcc/ree.cc
@@ -875,7 +875,8 @@ combine_reaching_defs (ext_cand *cand, const_rtx set_pat, ext_state *state)
 
 	  for (df_link *use = uses; use; use = use->next)
 	    if (paradoxical_subreg_p (GET_MODE (*DF_REF_LOC (use->ref)),
-				      GET_MODE (SET_DEST (*dest_sub_rtx))))
+				      GET_MODE (SET_DEST (*dest_sub_rtx)))
+		&& !DEBUG_INSN_P (DF_REF_INSN (use->ref)))
 	      return false;
 	}
 
@@ -963,7 +964,8 @@ combine_reaching_defs (ext_cand *cand, const_rtx set_pat, ext_state *state)
 	      rtx dest2 = SET_DEST (*dest_sub_rtx2);
 	      for (use = uses; use; use = use->next)
 		if (paradoxical_subreg_p (GET_MODE (*DF_REF_LOC (use->ref)),
-					  GET_MODE (dest2)))
+					  GET_MODE (dest2))
+		    && !DEBUG_INSN_P (DF_REF_INSN (use->ref)))
 		  break;
 	      if (use)
 		break;
diff --git a/gcc/testsuite/gcc.dg/pr108573.c b/gcc/testsuite/gcc.dg/pr108573.c
new file mode 100644
index 00000000000..5dd18043b2b
--- /dev/null
+++ b/gcc/testsuite/gcc.dg/pr108573.c
@@ -0,0 +1,18 @@
+/* PR debug/108573 */
+/* { dg-do compile } */
+/* { dg-options "-O2 -fcompare-debug" } */
+
+unsigned g;
+
+int bar (void);
+int baz (int);
+
+void
+foo (unsigned short s, long l)
+{
+  unsigned u = bar ();
+  s &= __builtin_add_overflow_p (0, u, 0);
+  s %= g;
+  s -= l >> s;
+  baz (s);
+}

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

only message in thread, other threads:[~2023-02-01 11:53 UTC | newest]

Thread overview: (only message) (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2023-02-01 11:53 [gcc r13-5625] ree: Fix -fcompare-debug issues in combine_reaching_defs [PR108573] Jakub Jelinek

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).