| Message ID | Yr6hObntpK6Fo9bj@tucnak |
|---|---|
| State | New |
| Headers | show |
| Series | wide-int: Fix up wi::shifted_mask [PR106144] | expand |
Jakub Jelinek <jakub@redhat.com> writes: > Hi! > > As the following self-test testcase shows, wi::shifted_mask sometimes > doesn't create canonicalized wide_ints, which then fail to compare equal > to canonicalized wide_ints with the same value. > In particular, wi::mask (128, false, 128) gives { -1 } with len 1 and prec 128, > while wi::shifted_mask (0, 128, false, 128) gives { -1, -1 } with len 2 > and prec 128. > The problem is that the code is written with the assumption that there are > 3 bit blocks (or 2 if start is 0), but doesn't consider the possibility > where there are 2 bit blocks (or 1 if start is 0) where the highest block > isn't present. In that case, there is the optional block of negate ? 0 : -1 > elts, followed by just one elt (either one from the if (shift) or just > negate ? -1 : 0) and the rest is implicit sign-extension. > Only if end < prec there is 1 or more bits above it that have different bit > value and so we need to emit all the elts till end and then one more elt. > > if (end == prec) would work too, because we have: > if (width > prec - start) > width = prec - start; > unsigned int end = start + width; > so end is guaranteed to be end <= prec, dunno what is preferred. > > Bootstrapped/regtested on x86_64-linux and i686-linux, ok for trunk? > > 2022-07-01 Jakub Jelinek <jakub@redhat.com> > > PR middle-end/106144 > * wide-int.cc (wi::shifted_mask): If end >= prec, return right after > emitting element for shift or if shift is 0 first element after start. > (wide_int_cc_tests): Add tests for equivalency of wi::mask and > wi::shifted_mask with 0 start. OK, thanks, but could you also remove the "end < prec" condition from: else if (end < prec) val[i++] = negate ? -1 : 0; Richard > --- gcc/wide-int.cc.jj 2022-01-11 23:11:23.592273263 +0100 > +++ gcc/wide-int.cc 2022-06-30 20:41:25.506292687 +0200 > @@ -842,6 +842,13 @@ wi::shifted_mask (HOST_WIDE_INT *val, un > val[i++] = negate ? block : ~block; > } > > + if (end >= prec) > + { > + if (!shift) > + val[i++] = negate ? 0 : -1; > + return i; > + } > + > while (i < end / HOST_BITS_PER_WIDE_INT) > /* 1111111 */ > val[i++] = negate ? 0 : -1; > @@ -2583,6 +2590,10 @@ wide_int_cc_tests () > run_all_wide_int_tests <widest_int> (); > test_overflow (); > test_round_for_mask (); > + ASSERT_EQ (wi::mask (128, false, 128), > + wi::shifted_mask (0, 128, false, 128)); > + ASSERT_EQ (wi::mask (128, true, 128), > + wi::shifted_mask (0, 128, true, 128)); > } > > } // namespace selftest > > Jakub
--- gcc/wide-int.cc.jj 2022-01-11 23:11:23.592273263 +0100 +++ gcc/wide-int.cc 2022-06-30 20:41:25.506292687 +0200 @@ -842,6 +842,13 @@ wi::shifted_mask (HOST_WIDE_INT *val, un val[i++] = negate ? block : ~block; } + if (end >= prec) + { + if (!shift) + val[i++] = negate ? 0 : -1; + return i; + } + while (i < end / HOST_BITS_PER_WIDE_INT) /* 1111111 */ val[i++] = negate ? 0 : -1; @@ -2583,6 +2590,10 @@ wide_int_cc_tests () run_all_wide_int_tests <widest_int> (); test_overflow (); test_round_for_mask (); + ASSERT_EQ (wi::mask (128, false, 128), + wi::shifted_mask (0, 128, false, 128)); + ASSERT_EQ (wi::mask (128, true, 128), + wi::shifted_mask (0, 128, true, 128)); } } // namespace selftest
Hi! As the following self-test testcase shows, wi::shifted_mask sometimes doesn't create canonicalized wide_ints, which then fail to compare equal to canonicalized wide_ints with the same value. In particular, wi::mask (128, false, 128) gives { -1 } with len 1 and prec 128, while wi::shifted_mask (0, 128, false, 128) gives { -1, -1 } with len 2 and prec 128. The problem is that the code is written with the assumption that there are 3 bit blocks (or 2 if start is 0), but doesn't consider the possibility where there are 2 bit blocks (or 1 if start is 0) where the highest block isn't present. In that case, there is the optional block of negate ? 0 : -1 elts, followed by just one elt (either one from the if (shift) or just negate ? -1 : 0) and the rest is implicit sign-extension. Only if end < prec there is 1 or more bits above it that have different bit value and so we need to emit all the elts till end and then one more elt. if (end == prec) would work too, because we have: if (width > prec - start) width = prec - start; unsigned int end = start + width; so end is guaranteed to be end <= prec, dunno what is preferred. Bootstrapped/regtested on x86_64-linux and i686-linux, ok for trunk? 2022-07-01 Jakub Jelinek <jakub@redhat.com> PR middle-end/106144 * wide-int.cc (wi::shifted_mask): If end >= prec, return right after emitting element for shift or if shift is 0 first element after start. (wide_int_cc_tests): Add tests for equivalency of wi::mask and wi::shifted_mask with 0 start. Jakub