Histogram test failure on master branch
I'm seeing a new failure on the master branch: 883/2538 Test #977: run-boost_histogram-histogram_ostream_ascii_test ......................................***Failed 0.00 sec /home/runner/work/boost/boost/libs/histogram/test/histogram_ostream_ascii_test.cpp(67): test 'str(h).c_str() == expected' ('BEGIN histogram(integer(0, 3, options=underflow | overflow))END' == 'BEGIN histogram(integer(0, 3, options=underflow | overflow)) +------------+ -1 0 | | 0 -1 |=== | 1 3 | ======== | 2 2 | ====== | 3 0 | | +------------+ END') failed in function 'void run_tests() [with Tag = boost::histogram::dynamic_tag]' 1 error detected. Histogram's master hasn't changed for 5 months, so this is probably caused by something else. The CI run is https://github.com/boostorg/boost/actions/runs/19686425029/job/56392904495 and it's being done in response to an Iterator change on master, although the change itself is non-functional.
That's weird and an oddly specific failure. I confirm that code has been stable for a long time, and this is the second test of two almost identical tests run in a single function. It looks like the ASCII histogram is not drawn by the ostream operator, but the previous test in that function passed, which is basically doing the same thing, generating a little ASCII histogram drawing. Since the error is not showing the ASCII histogram drawing, either both should fail or none of them. -----Ursprüngliche Nachricht----- Von: boost@lists.boost.org <boost@lists.boost.org> Gesendet: Mittwoch, 26. November 2025 13:49 An: 'Boost developers' mailing list' <boost@lists.boost.org> Cc: Peter Dimov <pdimov@gmail.com> Betreff: [boost] Histogram test failure on master branch I'm seeing a new failure on the master branch: 883/2538 Test #977: run-boost_histogram-histogram_ostream_ascii_test ......................................***Failed 0.00 sec /home/runner/work/boost/boost/libs/histogram/test/histogram_ostream_ascii_te st.cpp(67): test 'str(h).c_str() == expected' ('BEGIN histogram(integer(0, 3, options=underflow | overflow))END' == 'BEGIN histogram(integer(0, 3, options=underflow | overflow)) +------------+ -1 0 | | 0 -1 |=== | 1 3 | ======== | 2 2 | ====== | 3 0 | | +------------+ END') failed in function 'void run_tests() [with Tag = boost::histogram::dynamic_tag]' 1 error detected. Histogram's master hasn't changed for 5 months, so this is probably caused by something else. The CI run is https://github.com/boostorg/boost/actions/runs/19686425029/job/56392904495 and it's being done in response to an Iterator change on master, although the change itself is non-functional. _______________________________________________ Boost mailing list -- boost@lists.boost.org To unsubscribe send an email to boost-leave@lists.boost.org https://lists.boost.org/mailman3/lists/boost.lists.boost.org/ Archived at: https://lists.boost.org/archives/list/boost@lists.boost.org/message/EMOXYVMD ZM4ZCSUULEBU77WII4LKMGLJ/
On 26 Nov 2025 16:27, Hans Dembinski via Boost wrote:
That's weird and an oddly specific failure. I confirm that code has been stable for a long time, and this is the second test of two almost identical tests run in a single function.
It looks like the ASCII histogram is not drawn by the ostream operator, but the previous test in that function passed, which is basically doing the same thing, generating a little ASCII histogram drawing.
Since the error is not showing the ASCII histogram drawing, either both should fail or none of them.
FWIW, on my machine (Ubuntu 24.04, gcc 13, which is similar to the GHA image) the test passes. I noticed a valgrind warning that indicates use of uninitialized data, but I don't know if this is the culprit. See here: https://github.com/boostorg/histogram/pull/418
-----Ursprüngliche Nachricht----- Von: boost@lists.boost.org <boost@lists.boost.org> Gesendet: Mittwoch, 26. November 2025 13:49 An: 'Boost developers' mailing list' <boost@lists.boost.org> Cc: Peter Dimov <pdimov@gmail.com> Betreff: [boost] Histogram test failure on master branch
I'm seeing a new failure on the master branch:
883/2538 Test #977: run-boost_histogram-histogram_ostream_ascii_test ......................................***Failed 0.00 sec /home/runner/work/boost/boost/libs/histogram/test/histogram_ostream_ascii_te st.cpp(67): test 'str(h).c_str() == expected' ('BEGIN histogram(integer(0, 3, options=underflow | overflow))END' == 'BEGIN histogram(integer(0, 3, options=underflow | overflow)) +------------+ -1 0 | | 0 -1 |=== | 1 3 | ======== | 2 2 | ====== | 3 0 | | +------------+ END') failed in function 'void run_tests() [with Tag = boost::histogram::dynamic_tag]' 1 error detected.
Histogram's master hasn't changed for 5 months, so this is probably caused by something else.
The CI run is https://github.com/boostorg/boost/actions/runs/19686425029/job/56392904495 and it's being done in response to an Iterator change on master, although the change itself is non-functional.
_______________________________________________ Boost mailing list -- boost@lists.boost.org To unsubscribe send an email to boost-leave@lists.boost.org https://lists.boost.org/mailman3/lists/boost.lists.boost.org/ Archived at: https://lists.boost.org/archives/list/boost@lists.boost.org/message/EMOXYVMD ZM4ZCSUULEBU77WII4LKMGLJ/
_______________________________________________ Boost mailing list -- boost@lists.boost.org To unsubscribe send an email to boost-leave@lists.boost.org https://lists.boost.org/mailman3/lists/boost.lists.boost.org/ Archived at: https://lists.boost.org/archives/list/boost@lists.boost.org/message/PVI62G4J...
Thanks Andrey. I merged your patch to develop. I would say it is likely that this was causing it. -----Ursprüngliche Nachricht----- Von: boost@lists.boost.org <boost@lists.boost.org> Gesendet: Mittwoch, 26. November 2025 16:57 An: Boost developers' mailing list <boost@lists.boost.org> Cc: Andrey Semashev <andrey.semashev@gmail.com> Betreff: [boost] Re: Histogram test failure on master branch On 26 Nov 2025 16:27, Hans Dembinski via Boost wrote:
That's weird and an oddly specific failure. I confirm that code has been stable for a long time, and this is the second test of two almost identical tests run in a single function.
It looks like the ASCII histogram is not drawn by the ostream operator, but the previous test in that function passed, which is basically doing the same thing, generating a little ASCII histogram drawing.
Since the error is not showing the ASCII histogram drawing, either both should fail or none of them.
FWIW, on my machine (Ubuntu 24.04, gcc 13, which is similar to the GHA image) the test passes. I noticed a valgrind warning that indicates use of uninitialized data, but I don't know if this is the culprit. See here: https://github.com/boostorg/histogram/pull/418
-----Ursprüngliche Nachricht----- Von: boost@lists.boost.org <boost@lists.boost.org> Gesendet: Mittwoch, 26. November 2025 13:49 An: 'Boost developers' mailing list' <boost@lists.boost.org> Cc: Peter Dimov <pdimov@gmail.com> Betreff: [boost] Histogram test failure on master branch
I'm seeing a new failure on the master branch:
883/2538 Test #977: run-boost_histogram-histogram_ostream_ascii_test ......................................***Failed 0.00 sec /home/runner/work/boost/boost/libs/histogram/test/histogram_ostream_as cii_te st.cpp(67): test 'str(h).c_str() == expected' ('BEGIN histogram(integer(0, 3, options=underflow | overflow))END' == 'BEGIN histogram(integer(0, 3, options=underflow | overflow)) +------------+ -1 0 | | 0 -1 |=== | 1 3 | ======== | 2 2 | ====== | 3 0 | | +------------+ END') failed in function 'void run_tests() [with Tag = boost::histogram::dynamic_tag]' 1 error detected.
Histogram's master hasn't changed for 5 months, so this is probably caused by something else.
The CI run is https://github.com/boostorg/boost/actions/runs/19686425029/job/5639290 4495 and it's being done in response to an Iterator change on master, although the change itself is non-functional.
_______________________________________________ Boost mailing list -- boost@lists.boost.org To unsubscribe send an email to boost-leave@lists.boost.org https://lists.boost.org/mailman3/lists/boost.lists.boost.org/ Archived at: https://lists.boost.org/archives/list/boost@lists.boost.org/message/EM OXYVMD ZM4ZCSUULEBU77WII4LKMGLJ/
_______________________________________________ Boost mailing list -- boost@lists.boost.org To unsubscribe send an email to boost-leave@lists.boost.org https://lists.boost.org/mailman3/lists/boost.lists.boost.org/ Archived at: https://lists.boost.org/archives/list/boost@lists.boost.org/message/PV I62G4J5GMZNP5CMVIFE2CHA6PY3R35/
_______________________________________________ Boost mailing list -- boost@lists.boost.org To unsubscribe send an email to boost-leave@lists.boost.org https://lists.boost.org/mailman3/lists/boost.lists.boost.org/ Archived at: https://lists.boost.org/archives/list/boost@lists.boost.org/message/3PH7XNY2...
On 28 Nov 2025 13:11, hans.dembinski@gmail.com wrote:
Thanks Andrey. I merged your patch to develop. I would say it is likely that this was causing it.
It's probably worth including in the release then.
-----Ursprüngliche Nachricht----- Von: boost@lists.boost.org <boost@lists.boost.org> Gesendet: Mittwoch, 26. November 2025 16:57 An: Boost developers' mailing list <boost@lists.boost.org> Cc: Andrey Semashev <andrey.semashev@gmail.com> Betreff: [boost] Re: Histogram test failure on master branch
On 26 Nov 2025 16:27, Hans Dembinski via Boost wrote:
That's weird and an oddly specific failure. I confirm that code has been stable for a long time, and this is the second test of two almost identical tests run in a single function.
It looks like the ASCII histogram is not drawn by the ostream operator, but the previous test in that function passed, which is basically doing the same thing, generating a little ASCII histogram drawing.
Since the error is not showing the ASCII histogram drawing, either both should fail or none of them.
FWIW, on my machine (Ubuntu 24.04, gcc 13, which is similar to the GHA image) the test passes.
I noticed a valgrind warning that indicates use of uninitialized data, but I don't know if this is the culprit. See here:
https://github.com/boostorg/histogram/pull/418
-----Ursprüngliche Nachricht----- Von: boost@lists.boost.org <boost@lists.boost.org> Gesendet: Mittwoch, 26. November 2025 13:49 An: 'Boost developers' mailing list' <boost@lists.boost.org> Cc: Peter Dimov <pdimov@gmail.com> Betreff: [boost] Histogram test failure on master branch
I'm seeing a new failure on the master branch:
883/2538 Test #977: run-boost_histogram-histogram_ostream_ascii_test ......................................***Failed 0.00 sec /home/runner/work/boost/boost/libs/histogram/test/histogram_ostream_as cii_te st.cpp(67): test 'str(h).c_str() == expected' ('BEGIN histogram(integer(0, 3, options=underflow | overflow))END' == 'BEGIN histogram(integer(0, 3, options=underflow | overflow)) +------------+ -1 0 | | 0 -1 |=== | 1 3 | ======== | 2 2 | ====== | 3 0 | | +------------+ END') failed in function 'void run_tests() [with Tag = boost::histogram::dynamic_tag]' 1 error detected.
Histogram's master hasn't changed for 5 months, so this is probably caused by something else.
The CI run is https://github.com/boostorg/boost/actions/runs/19686425029/job/5639290 4495 and it's being done in response to an Iterator change on master, although the change itself is non-functional.
_______________________________________________ Boost mailing list -- boost@lists.boost.org To unsubscribe send an email to boost-leave@lists.boost.org https://lists.boost.org/mailman3/lists/boost.lists.boost.org/ Archived at: https://lists.boost.org/archives/list/boost@lists.boost.org/message/EM OXYVMD ZM4ZCSUULEBU77WII4LKMGLJ/
_______________________________________________ Boost mailing list -- boost@lists.boost.org To unsubscribe send an email to boost-leave@lists.boost.org https://lists.boost.org/mailman3/lists/boost.lists.boost.org/ Archived at: https://lists.boost.org/archives/list/boost@lists.boost.org/message/PV I62G4J5GMZNP5CMVIFE2CHA6PY3R35/
_______________________________________________ Boost mailing list -- boost@lists.boost.org To unsubscribe send an email to boost-leave@lists.boost.org https://lists.boost.org/mailman3/lists/boost.lists.boost.org/ Archived at: https://lists.boost.org/archives/list/boost@lists.boost.org/message/3PH7XNY2...
participants (3)
-
Andrey Semashev -
hans.dembinski＠gmail.com -
Peter Dimov