The mask/value pair we track in the irange is a bit fickle in that it
can sometimes contradict the bitmask inherent in the range. This can
happen when a series of calculations yield a combination such as:
[3, 1000] MASK 0xfffffffe VALUE 0x0
The mask/value above implies that the lowest bit is a known 0, which
would exclude the 3 in the range. At one time we tried keeping mask
and ranges 100% consistent, but the performance penalty was too high
(5% in VRP). Also, it's unclear whether the intersection of two
incompatible known bits should make the whole range undefined, or
just the contradicting bits. This is all documented in
irange::get_bitmask(). We could revisit both of these assumptions
in the future.
In this testcase IPA ends up with a range where the lower 2 bits are
expected to be 0, but the range is [1,1].
[irange] long int [1, 1] MASK 0xfffffffffffffffc VALUE 0x0
This causes irange::union_bitmask() to think an update occurred, when
no semantic change happened, thus triggering an assert in IPA-cp. We
could get rid of the assert, but it's cleaner to make
irange::{union,intersect}_bitmask always tell the truth. Beside, the
ranger's cache also depends on union being truthful.
PR ipa/110753
gcc/ChangeLog:
* value-range.cc (irange::union_bitmask): Return FALSE if updated
bitmask is semantically equivalent to the original mask.
(irange::intersect_bitmask): Same.
(irange::get_bitmask): Add comment.
gcc/testsuite/ChangeLog:
* gcc.dg/tree-ssa/pr110753.c: New test.
---
gcc/testsuite/gcc.dg/tree-ssa/pr110753.c | 15 +++++++++++++++
gcc/value-range.cc | 18 ++++++++++++++++++
2 files changed, 33 insertions(+)
create mode 100644 gcc/testsuite/gcc.dg/tree-ssa/pr110753.c
diff --git a/gcc/testsuite/gcc.dg/tree-ssa/pr110753.c
b/gcc/testsuite/gcc.dg/tree-ssa/pr110753.c
new file mode 100644
index 00000000000..aa02487e2a7
--- /dev/null
+++ b/gcc/testsuite/gcc.dg/tree-ssa/pr110753.c
@@ -0,0 +1,15 @@
+// { dg-do compile }
+// { dg-options "-O2" }
+
+int a, b, c;
+static int d(long e, long f) { return f == 0 || e && f == 1 ?: f; }
+int g(void) {static int t; return t;}
+static void h(long e) {
+ b = e - 1;
+ a = d(b || d(e, 8), g());
+}
+int tt;
+void i(void) {
+ c = (__SIZE_TYPE__)&tt;
+ h(c);
+}
diff --git a/gcc/value-range.cc b/gcc/value-range.cc
index 2abf57bcee8..76f88d91046 100644
--- a/gcc/value-range.cc
+++ b/gcc/value-range.cc
@@ -1876,6 +1876,8 @@ irange::get_bitmask () const
//
// 3 is in the range endpoints, but is excluded per the known 0 bits
// in the mask.
+ //
+ // See also the note in irange_bitmask::intersect.
irange_bitmask bm = get_bitmask_from_range ();
if (!m_bitmask.unknown_p ())
bm.intersect (m_bitmask);
@@ -1915,10 +1917,18 @@ irange::intersect_bitmask (const irange &r)
return false;
irange_bitmask bm = get_bitmask ();
+ irange_bitmask save = bm;
if (!bm.intersect (r.get_bitmask ()))
return false;
m_bitmask = bm;
+
+ // Updating m_bitmask may still yield a semantic bitmask (as
+ // returned by get_bitmask) which is functionally equivalent to what
+ // we originally had. In which case, there's still no change.
+ if (save == get_bitmask ())
+ return false;
+
if (!set_range_from_bitmask ())
normalize_kind ();
if (flag_checking)
@@ -1938,10 +1948,18 @@ irange::union_bitmask (const irange &r)
return false;
irange_bitmask bm = get_bitmask ();
+ irange_bitmask save = bm;
if (!bm.union_ (r.get_bitmask ()))
return false;
m_bitmask = bm;
+
+ // Updating m_bitmask may still yield a semantic bitmask (as
+ // returned by get_bitmask) which is functionally equivalent to what
+ // we originally had. In which case, there's still no change.
+ if (save == get_bitmask ())
+ return false;
+
// No need to call set_range_from_mask, because we'll never
// narrow the range. Besides, it would cause endless recursion
// because of the union_ in set_range_from_mask.
--
2.41.0
