Thu 13 Oct 2016 06:26:46 PM UTC, comment #10:
Good. That does it for me.
|
Thu 13 Oct 2016 02:57:06 PM UTC, comment #9:
Alright, just remove the out-of-date files manually to fix this.
|
Thu 13 Oct 2016 10:10:35 AM UTC, comment #8:
Rik: Your suggestion did not fix it for me.
Summary:
PASS 14810
FAIL 2
XFAIL 38
SKIPPED 43
Also, this case provided a reminder that when files
are moved within the directory structure, make install
typically fails. In this case,
/usr/local/share/octave/4.3.0+/m/linear-algebra/gsvd.m
was stale and had to be removed. I will open a new bug
report for this.
|
Thu 13 Oct 2016 12:28:27 AM UTC, comment #7:
The failures you are seeing are from leftover out-of-date tests. Try a complete clean and re-build.
|
Wed 12 Oct 2016 11:28:29 PM UTC, comment #6:
OK. I thought if i do fresh clone i do not need to update.
+verbose+
PASS 14810
FAIL 2
XFAIL 38
SKIPPED 43
hg id
68e9bdb7cde3 @
octave:1> test libinterp/parse-tree/pt-binop.cc-tst
x = 0;
do_braindead_shortcircuit_evaluation (0);
if (1 | (x = 1))
endif
assert (x, 1);
do_braindead_shortcircuit_evaluation (1);
if (1 | (x = 0))
endif
assert (x, 1);
!!!!! test failed
'do_braindead_shortcircuit_evaluation' undefined near line 4 column 2
octave:2> test libinterp/octave-value/ov-range.cc-tst
x = 0:10;
warning ("off", "Octave:deprecated-function", "local");
save = allow_noninteger_range_as_index ();
warn_state = warning ("query", "Octave:noninteger-range-as-index");
unwind_protect
allow_noninteger_range_as_index (false);
fail ("x(2.1:5)");
assert (x(2:5), 1:4);
allow_noninteger_range_as_index (true);
warning ("off", "Octave:noninteger-range-as-index");
assert (x(2.49:5), 1:3);
assert (x(2.5:5), 2:4);
assert (x(2.51:5), 2:4);
unwind_protect_cleanup
allow_noninteger_range_as_index (save);
warning (warn_state.state, warn_state.identifier);
end_unwind_protect
!!!!! test failed
'allow_noninteger_range_as_index' undefined near line 5 column 9
-verbose-
|
Wed 12 Oct 2016 09:34:32 PM UTC, comment #5:
Try 'hg pull' followed by 'hg update @'. The default bookmark fell behind the actual head of development, but jwe has corrected that now.
|
Wed 12 Oct 2016 09:25:23 PM UTC, comment #4:
My case is exactly like Dimitri's.
With hg id:
54332b6ddeb6 @
|
Wed 12 Oct 2016 07:46:06 PM UTC, comment #3:
Instead of "hg head", which shows the branch heads, what does "hg id" tell you?
If it doesn't match one of the branch heads, why not? Oh, I see, maybe because you are stuck back at the '@' bookmark and that hasn't been properly advanced recently?
|
Wed 12 Oct 2016 07:38:09 PM UTC, comment #2:
Freashly re-clone and re-build:
Dmitri.
--
|
Wed 12 Oct 2016 04:19:13 PM UTC, comment #1:
Can you try a cset greater than or equal to 22612:c05377052b50?
Also, what happens if you just start octave and try
Does that fail in the same way that 'make check' does?
|
Wed 12 Oct 2016 09:17:49 AM UTC, original submission:
On Fedora 24 x64 I now get:
scripts/ode/ode23.m ......................................... PASS 11/36
FAIL 25
scripts/ode/ode45.m ......................................... PASS 11/40
FAIL 29
It appears that the ode update has gotten in an incorrect state
on the default system
|