http://bugs.winehq.org/show_bug.cgi?id=9356
Summary: Serial communication not working since wine-0.9.33 Product: Wine Version: 0.9.42. Platform: PC OS/Version: Linux Status: UNCONFIRMED Severity: blocker Priority: P1 Component: wine-binary AssignedTo: wine-bugs@winehq.org ReportedBy: aurel@gmx.de
Created an attachment (id=7647) --> (http://bugs.winehq.org/attachment.cgi?id=7647) Logfiles of woring and not workin wine versions
Senn on application Windows Packet Programm the serial communication is not working correctly since winw version 0.9.33 anymore. It worked perfect with version 0.9.32 and before. Actually I am running version 0.9.42. I attached two logs, one with a working version (0.9.32) and one with my latest version 0.9.42. You can see, that there is a problem here: io_control 0x7c IOCTL_SERIAL_GET_TIMEOUTS. At this point the application hangs and you have to kill it.
Regards Aurel
http://bugs.winehq.org/show_bug.cgi?id=9356
Damjan Jovanovic damjan.jov@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |damjan.jov@gmail.com Severity|blocker |normal
--- Comment #1 from Damjan Jovanovic damjan.jov@gmail.com 2007-08-18 08:59:43 --- It's not clear from the logs what changed to break it, and sadly serial port problems cannot be reproduced without hardware. Can you please regression test wine (http://wiki.winehq.org/RegressionTesting)?
By the way, bugs with blocker rating don't get fixed any sooner than the others :-). Rather vote for the bug?
http://bugs.winehq.org/show_bug.cgi?id=9356
--- Comment #2 from Austin English austinenglish@gmail.com 2007-08-18 20:35:30 --- Aurel, please post to bugzilla in the future, not by e-mail.
From e-mail:
Thank you Damjan,
just now I am running the regrssion test here, I will let you know what happened.
Sorry for the blocker rating, but for my application (WPP) it really was a blocker since it didn't start anymore and I have to downgrade to 0.9.32 every time my Fedora yum upgrades to the latest wine version, which is not working anymore for me.
Best regards Aurel
http://bugs.winehq.org/show_bug.cgi?id=9356
--- Comment #3 from Aurel aurel@gmx.de 2007-08-19 09:04:51 --- Created an attachment (id=7672) --> (http://bugs.winehq.org/attachment.cgi?id=7672) Regression testing 2007-08-19 wine version 0.9.34 vs 0.9.35
Contrary to my first comment I found out, that the bug does not appear in wine version 0.9.33, but it must be 0.9.35.
I hope this helps to solve the problem. Let me know if I can do anything else to support here.
Regards Aurel --- aurel@gmx.de www.p-a-p-p.de
http://bugs.winehq.org/show_bug.cgi?id=9356
Aurel aurel@gmx.de changed:
What |Removed |Added ---------------------------------------------------------------------------- Attachment #7647|Logfiles of woring and not |Logfiles of working and not description|workin wine versions |working wine versions
http://bugs.winehq.org/show_bug.cgi?id=9356
Austin English austinenglish@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |julliard@codeweavers.com Keywords| |regression
--- Comment #4 from Austin English austinenglish@gmail.com 2007-08-19 14:54:48 --- Pasting in comments so it's easier to see...sorry for the noise.
f85437c57fb44d511db97edbaa5b1f8abfe75fd3 is first bad commit commit f85437c57fb44d511db97edbaa5b1f8abfe75fd3 Author: Alexandre Julliard julliard@winehq.org Date: Tue Apr 10 22:25:07 2007 +0200
server: Store the opening options in the file descriptor instead of in the individual objects.
Allows moving the FD_FLAG_OVERLAPPED support to the generic get_handle_fd request handling.
:040000 040000 756555c5ba56a1c2a30b5fae58a97cfbecd1a2ea 6f7c193323c077cce8da8b4da0690aedf6a14793 M server
CC'ing author.
http://bugs.winehq.org/show_bug.cgi?id=9356
Aurel aurel@gmx.de changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |aurel@gmx.de
--- Comment #5 from Aurel aurel@gmx.de 2007-09-15 02:40:19 --- anything happening here?
Regards Aurel
http://bugs.winehq.org/show_bug.cgi?id=9356
--- Comment #6 from Vladimir Kadlec vladimirkadlec@gmail.com 2007-09-15 12:57:36 --- Created an attachment (id=8085) --> (http://bugs.winehq.org/attachment.cgi?id=8085) Log for wine 0.9.44 (ubuntu binary)
http://bugs.winehq.org/show_bug.cgi?id=9356
Vladimir Kadlec vladimirkadlec@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |vladimirkadlec@gmail.com
--- Comment #7 from Vladimir Kadlec vladimirkadlec@gmail.com 2007-09-15 13:04:59 --- Hello, I can confirm, that the serial communication works with wine 0.9.33 and doesn't work with wine 0.9.44. The problematic application is HACtronic, version 1.60, http://www.ciclosport.de.
I'm not able to do a regression test at the moment. I attached a logfile for wine 0.9.44 (ubuntu binary). The application hangs, as described.
I'm using hardware on USB, the serial port is emulated through ftdi_sio kernel module. I have created link ~/.wine/dosdevices/com1 -> /dev/ttyUSB0
Regards
http://bugs.winehq.org/show_bug.cgi?id=9356
--- Comment #8 from Lei Zhang thestig@google.com 2007-09-18 16:28:42 --- Is serial communication broken for just for your app or is it broken for all apps?
Is there anyway of testing this without extra hardware?
http://bugs.winehq.org/show_bug.cgi?id=9356
--- Comment #9 from Vladimir Kadlec vladimirkadlec@gmail.com 2007-09-18 17:00:57 --- I don't know, if it's broken in other apps, I use only Hactronic. I can test some other, if you tell me, what app should I try.
You can test it, even without the hardware (I checked it). Download Hactronic 1.60: http://nlp.fi.muni.cz/~xkadlec/hac/hactronic160.zip
It was downloadable through ciclosport.de, so I believe, there is no problem, that I mirror it. There is a new version 1.83, which uses different code for serial communication, which doesn't work for me, even under windows, so use version 1.60 for the test.
Before running, create a link to the existing (unused) serial device, e.g. ln -s /dev/ttyS1 ~/.wine/dosdevices/com1
No installation required, just run the .exe, then menu File/Connecting (or the icon with watches). When you click "Cancel" the app hangs under wine 0.9.44 (see the attached log), under 0.9.33 does not hang.
Hope this helps.
http://bugs.winehq.org/show_bug.cgi?id=9356
Lei Zhang thestig@google.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|UNCONFIRMED |NEW Ever Confirmed|0 |1
--- Comment #10 from Lei Zhang thestig@google.com 2007-09-18 18:07:06 --- Hactronic hangs with the latest wine when I try to connect to the non-existent device. With wine 0.9.33, I can hit cancel.
When I get home I'll try my x10 serial interface under Wine and see if it is affected as well. http://www.x10.com/products/x10_cm17a.htm
http://bugs.winehq.org/show_bug.cgi?id=9356
Aurel aurel@gmx.de changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |REMIND
--- Comment #11 from Aurel aurel@gmx.de 2007-10-16 15:31:08 --- Hello,
anything happening here??? Anybody takes care for this bug?
Waiting for a reaction for two months now...
Regards Aurel
http://bugs.winehq.org/show_bug.cgi?id=9356
James Hawkins truiken@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|RESOLVED |REOPENED Resolution|REMIND |
--- Comment #12 from James Hawkins truiken@gmail.com 2007-10-16 15:53:47 --- There are 3000 other bugs to be fixed, so don't be surprised if your bug takes a while to get fixed. Changing it to remind will make sure no one looks at it though..
http://bugs.winehq.org/show_bug.cgi?id=9356
--- Comment #13 from Aurel aurel@gmx.de 2007-11-22 15:49:20 --- Three months now, no reaction except some comments. How many bugs to be fixed before this one? 2999? May I help somehow? I really would appreciate it!
http://bugs.winehq.org/show_bug.cgi?id=9356
--- Comment #14 from Lei Zhang thestig@google.com 2007-11-24 22:05:35 --- well, I ended up not looking at this bug because I didn't have a computer with a serial port...
now that I do, I tried out Hactronic and here's what I found:
In NtReadFile(), with the old code, we would try to do a read(), which returns 0, and NtReadFile() returns. NtReadFile() gets called repeatedly until I hit cancel, or presumably until when it actually reads some data.
With the new code, read() returns EAGAIN, so we keep going until we call poll() later in NtReadFile() with an infinite timeout. Since there's no hardware connected to the serial port, the poll() call hangs.
http://bugs.winehq.org/show_bug.cgi?id=9356
--- Comment #15 from Aurel aurel@gmx.de 2007-11-25 06:42:17 --- So what does that mean now? Will we never get a bug fix, because you haven't got any hardware to test? Is there a workaround? Can we do the testing? You change the code and we tell you if it works? In my opinion it can't be that much, since it worked perfect up to wine 0.9.34. What if you just change the timeout to some seconds instead of infinite for the polling in NtReadFile()?
http://bugs.winehq.org/show_bug.cgi?id=9356
--- Comment #16 from Damjan Jovanovic damjan.jov@gmail.com 2007-11-25 23:32:12 --- Created an attachment (id=9347) --> (http://bugs.winehq.org/attachment.cgi?id=9347) set the fd back to blocking if necessary
This part of the patch that causes the regression looks suspicious:
- /* set the fd back to blocking if necessary */ - if (options & (FILE_SYNCHRONOUS_IO_ALERT | FILE_SYNCHRONOUS_IO_NONALERT)) - fcntl( unix_fd, F_SETFL, 0 );
I'm guessing in the new code it isn't set to blocking, leading to EAGAIN on reads. If that's the case, the attached patch should fix it. Please try it and report back.
(Bloody serial ports. We really need regression tests for the them.)
http://bugs.winehq.org/show_bug.cgi?id=9356
Klementi teamloser@netzero.net changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |teamloser@netzero.net
http://bugs.winehq.org/show_bug.cgi?id=9356
--- Comment #17 from Aurel aurel@gmx.de 2007-11-28 16:32:49 --- Hi Damjan,
your patch works perfectly for me, I just tried with wine-0.9.49! Thank you very much!
Is it possible to add the code to the next wine release, because normally I am using RPMs for my Fedora 8 system, which are installed by yum automatically.
Best regards Aurel
http://bugs.winehq.org/show_bug.cgi?id=9356
--- Comment #18 from Damjan Jovanovic damjan.jov@gmail.com 2007-11-28 23:21:30 --- I'm glad it works, I'll submit it to wine-patches right away.
http://bugs.winehq.org/show_bug.cgi?id=9356
--- Comment #19 from Damjan Jovanovic damjan.jov@gmail.com 2007-11-29 08:54:41 --- My patch was rejected: http://www.winehq.org/pipermail/wine-devel/2007-November/060847.html
If anybody else wants to work on it, feel free. I'm not in the mood.
http://bugs.winehq.org/show_bug.cgi?id=9356
Aurel aurel@gmx.de changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |julliard@winehq.org
http://bugs.winehq.org/show_bug.cgi?id=9356
--- Comment #20 from Aurel aurel@gmx.de 2007-12-02 12:10:03 --- So that means they rather come up with non working serial ports (since 0.9.35 now) than having a work around as long as nobody takes care for fixing the EAGAIN thing. Unfortunately I am not able to fix it myself, I really hope somebody will.
http://bugs.winehq.org/show_bug.cgi?id=9356
--- Comment #21 from Vitaliy Margolen vitaliy@kievinfo.com 2007-12-02 19:26:58 --- (In reply to comment #20)
So that means they rather come up with non working serial ports (since 0.9.35 now) than having a work around as long as nobody takes care for fixing the EAGAIN thing. Unfortunately I am not able to fix it myself, I really hope somebody will.
You are exactly correct, that's the way Wine development works. It has to be fixed "right" or not at all.
http://bugs.winehq.org/show_bug.cgi?id=9356
Vitaliy Margolen vitaliy@kievinfo.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Component|_obsolete_binary |ntdll
http://bugs.winehq.org/show_bug.cgi?id=9356
Austin English austinenglish@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Target Milestone|--- |1.0.0
--- Comment #22 from Austin English austinenglish@gmail.com 2008-01-26 05:03:03 --- Nominating for 1.0.
http://bugs.winehq.org/show_bug.cgi?id=9356
--- Comment #23 from Aurel aurel@gmx.de 2008-03-24 13:54:26 --- Hi Guys,
seven months gone now, no bug fix of that serial communication which was working fine in former wine versions, so I guess it shouldn't be that hard to do it. Only things I've seen here were a lot of bla bla replies and they were really quick! Wow!
Sorry, but that's not what I call a great job, in fact its really sad. I ask myself why reporting those bugs if nobody cares for them.
Lets switch back to windoze, they have got that thing working perfectly!
Take care Aurel
http://bugs.winehq.org/show_bug.cgi?id=9356
Austin English austinenglish@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Keywords| |patch
http://bugs.winehq.org/show_bug.cgi?id=9356
Andrius andriusmao@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |andriusmao@gmail.com
--- Comment #24 from Andrius andriusmao@gmail.com 2008-04-24 11:40:14 --- Aurel, don't be impatient, it's nominated for wine 1.0, which will be released in june, wine developers do a good job, there are several bugs, remember that wine is ALPHA and you shouldn't compare with windows's compatibility (now).
is there much difference between the cause of this bug and of bug #8783 ?
http://bugs.winehq.org/show_bug.cgi?id=9356
--- Comment #25 from Alexandre Julliard julliard@winehq.org 2008-05-06 10:59:40 --- I can't find hactronic 1.60, and 1.83 is working fine for me as far as I can tell without a device to talk to.
http://bugs.winehq.org/show_bug.cgi?id=9356
igor igor_u@ukr.net changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |igor_u@ukr.net
--- Comment #26 from igor igor_u@ukr.net 2008-05-06 12:01:20 --- I use COM port for automobile diagnostics. After 0.9.33 does not work. In house conditions for check of working capacity I use phone Samsung C100 and COM a cable. After 0.9.33 does not work.
http://bugs.winehq.org/show_bug.cgi?id=9356
--- Comment #27 from Vladimir Kadlec vladimirkadlec@gmail.com 2008-05-06 13:51:41 --- (In reply to comment #25)
I can't find hactronic 1.60, and 1.83 is working fine for me as far as I can tell without a device to talk to.
You can download hactronic 1.60 again from: http://nlp.fi.muni.cz/~xkadlec/hac/hactronic160.zip Sorry for inconvenience.
http://bugs.winehq.org/show_bug.cgi?id=9356
Alexandre Julliard julliard@winehq.org changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|REOPENED |RESOLVED Resolution| |FIXED
--- Comment #28 from Alexandre Julliard julliard@winehq.org 2008-05-07 07:05:01 --- (In reply to comment #27)
You can download hactronic 1.60 again from: http://nlp.fi.muni.cz/~xkadlec/hac/hactronic160.zip
Thanks. It should work now.
http://bugs.winehq.org/show_bug.cgi?id=9356
Dan Kegel dank@kegel.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |dank@kegel.com
--- Comment #29 from Dan Kegel dank@kegel.com 2008-05-07 07:56:53 --- Here's the patch he committed: http://www.winehq.org/pipermail/wine-cvs/2008-May/043429.html
http://bugs.winehq.org/show_bug.cgi?id=9356
--- Comment #30 from Andrius andriusmao@gmail.com 2008-05-08 09:37:18 --- I've tested with an old windows program: "terminal", it still doesn't recognise any COM port.
http://bugs.winehq.org/show_bug.cgi?id=9356
Alexandre Julliard julliard@winehq.org changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|RESOLVED |CLOSED
--- Comment #31 from Alexandre Julliard julliard@winehq.org 2008-05-09 12:54:35 --- Closing bugs fixed in 1.0-rc1.
http://bugs.winehq.org/show_bug.cgi?id=9356
--- Comment #32 from igor igor_u@ukr.net 2008-05-10 00:47:28 --- (In reply to comment #31)
Closing bugs fixed in 1.0-rc1.
Thanks.
http://bugs.winehq.org/show_bug.cgi?id=9356
--- Comment #33 from Austin English austinenglish@gmail.com 2008-05-11 17:22:03 --- Has anyone tested this to make sure it works? A user is reporting it's not, so someone should make sure it works for them before opening a new bug...
http://bugs.winehq.org/show_bug.cgi?id=9356
--- Comment #34 from Vladimir Kadlec vladimirkadlec@gmail.com 2008-05-12 03:20:28 --- (In reply to comment #33) I confirm, that under wine-1.0-rc1 (ubuntu binary) HACtronic 1.60 doesn't hang. The patch (or other changes) fixed serial communication also in HACtronic 1.83 and CicloTour 3.03. Thank you.
http://bugs.winehq.org/show_bug.cgi?id=9356
--- Comment #35 from Aurel aurel@gmx.de 2008-05-19 13:47:11 --- Hi guys,
WPP is now working perfectly again under wine-1.0-rc1, source code compiled myself, since there is no binary available yet for my Fedora 8.
Thanks everybody for working on it.
Regards Aurel