Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Already on GitHub? Sign in to your account

Failing tests on Windows8. #711

Closed
txdv opened this Issue Feb 15, 2013 · 9 comments

Comments

Projects
None yet
7 participants
Contributor

txdv commented Feb 15, 2013

I have some tests failing on windows:

=============================================================
[%   8|+  15|-   0]: pipe_connect_to_file
`pipe_connect_to_file` failed: exit code 3
Output from process `pipe_connect_to_file`:
Assertion failed in test\test-pipe-connect-error.c on line 56: uv_last_error(uv_
default_loop()).code == UV_ENOTSOCK || uv_last_error(uv_default_loop()).code ==
UV_ECONNREFUSED
=============================================================
[%  76|+ 139|-   1]: fs_file_loop
`fs_file_loop` failed: exit code 3
Output from process `fs_file_loop`:
Assertion failed in test\test-fs.c on line 529: uv_last_error(loop).code == UV_E
LOOP
=============================================================
[%  86|+ 156|-   2]: fs_event_watch_file_twice
`fs_event_watch_file_twice` failed: exit code 3
Output from process `fs_event_watch_file_twice`:
Assertion failed in test\test-fs-event.c on line 255: 0 == uv_fs_event_init(loop
, watchers + 0, path, fail_cb, 0)
=============================================================
[%  91|+ 163|-   3]: fs_readdir_file
`fs_readdir_file` failed: exit code 3
Output from process `fs_readdir_file`:
Assertion failed in test\test-fs.c on line 1662: uv_last_error(loop).code == UV_
ENOTDIR
=============================================================
[%  94|+ 168|-   4]: threadpool_cancel_getaddrinfo
`threadpool_cancel_getaddrinfo` failed: exit code 3
Output from process `threadpool_cancel_getaddrinfo`:
Assertion failed in test\test-threadpool-cancel.c on line 90: 0 == uv_cancel((uv
_req_t*) req)
=============================================================
[%  95|+ 168|-   5]: threadpool_cancel_work
`threadpool_cancel_work` failed: exit code 3
Output from process `threadpool_cancel_work`:
Assertion failed in test\test-threadpool-cancel.c on line 90: 0 == uv_cancel((uv
_req_t*) req)
=============================================================
[%  95|+ 168|-   6]: threadpool_cancel_fs
`threadpool_cancel_fs` failed: exit code 3
Output from process `threadpool_cancel_fs`:
Assertion failed in test\test-threadpool-cancel.c on line 90: 0 == uv_cancel((uv
_req_t*) req)
=============================================================
[%  96|+ 168|-   7]: threadpool_cancel_single
`threadpool_cancel_single` failed: exit code 1
Output from process `threadpool_cancel_single`:
Failed to cancel a work req in 5,000 iterations, giving up.
=============================================================
[%  99|+ 173|-   8]: dlerror
`dlerror` failed: exit code 3
Output from process `dlerror`:
Assertion failed in test\test-dlerror.c on line 48: strstr(msg, dlerror_desc) !=
 NULL
=============================================================
[% 100|+ 173|-   9]: Done.

I have left the stats out.

Contributor

schlamar commented Mar 8, 2013

I'm seeing the following errors, too:

  • fs_file_loop
  • threadpool_cancel_getaddrinfo
  • threadpool_cancel_work
  • threadpool_cancel_fs
  • threadpool_cancel_single
  • dlerror
Contributor

bnoordhuis commented Mar 8, 2013

The threadpool_cancel_* tests are expected to fail, they're testing functionality that hasn't been implemented yet in uv-win. The other two should pass, I think.

Contributor

schlamar commented Mar 8, 2013

And loop_stop is failing occasionally:

>Release\run-tests.exe loop_stop

`loop_stop` failed: exit code 3
Output from process `loop_stop`:
Assertion failed in test\test-loop-stop.c on line 65: prepare_called == 3
=============================================================

See the issue on pyuv for more details: saghul/pyuv#85

Contributor

schlamar commented Mar 11, 2013

My errors are on Windows 7, BTW.

I'm seeing failing

  • loop_stop
  • timer_huge_timeout
  • timer_huge_repeat
  • fs_file_loop
  • fs_symlink_dir
    ...and the four threadpool_cancel_*

Also, in run-benchmarks:

  • tcp_multi_accept{2,4,8} fail, assert on benchmark-multi-accept.c line 123
  • async{2,4,8} fail with timeout (but async1 passes)
  • async_pummel_{4,8} fail with timeout
  • spawn fails with timeout
  • thread_create fails with timeout
    A bunch of run-benchmarks.exe processes remain listed in Task Manager, and consume CPU.

Windows 7, release x64 EXE.

ksubox commented Jul 11, 2013

I confirm same issues, especially with async_pummel.
When I test in virtual machine with 2 cores async_pummel_1 works well, but async_pummel_2 takes loooong time, so I added uv_sleep(0) after uv_async_send(handle) and it started work fast, but async_pummel_[48] work veeery slow. I hope this symptoms will help.

Windows 8, mingw gcc 4.8.1 x64

Member

piscisaureus commented Jul 11, 2013

loop_stop is expected to fail in v0.10 but not master. The threadpool_cancel tests always fail for the moment.

The others are real issues. Unfortunately I'm extremely busy with many different things atm, and I would be very grateful if one of you could dig into the the failing tests and see what's going on.

My ususal strategy is to pick a single failing test and return it outside of the test harness. The way to do this (for tests that require no helpers atleast) is by running run-tests.exe loop_stop loop_stop. Subsequently I'd set a breakpoint on the failing assert and figure out what's going on. Fix the bug or test, rinse, repeat with another failure.

Contributor

schlamar commented Feb 26, 2014

This one is new

`tcp_close_accept` failed: exit code 3
Output from process `tcp_close_accept`:
Assertion failed in test\test-tcp-close-accept.c on line 102: 0 == uv_read_start((uv_stream_t*) &tcp_check, alloc_cb, read_cb)
Contributor

saghul commented Sep 19, 2014

I'm going to close this since many of the failures have already been fixed in v1.x. Please open a new issue with fresh failures if need be.

@saghul saghul closed this Sep 19, 2014

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment