[Bug 19397] New: Polar ProTrainer 5: IRDA connection to heart rate monitor does not work
http://bugs.winehq.org/show_bug.cgi?id=19397
Summary: Polar ProTrainer 5: IRDA connection to heart rate monitor does not work Product: Wine Version: 1.1.26 Platform: PC OS/Version: Linux Status: UNCONFIRMED Severity: normal Priority: P2 Component: -unknown AssignedTo: wine-bugs@winehq.org ReportedBy: mikko.ronkko@iki.fi
I have installed Polar ProTrainer 5 software version 5.20.130 and attached a USB IrDA dongle from Polar to use with my Polar RS800 heart rate monitor. After installing the dongle and verifying that it can see my heart rate monitor using irdadump, I tried to connect the software and the heart rate monitor. I configured the softaware to use IrDA connection from Options->Preferences->Hardware. (Other options are USB and COM1-COM9) Then I try to transfer data from the heart rate monitor using Tools->Transfer Data. A popup window appears saying "Reading Settings..." and disappears after a while. No data is transferred and no error message is given by the software. The next time I try to connect to the heart rate monitor from the software, I get an error message saying "Communication port couldn't be opened." The next time I try the transfer, I get the right dialog but no data transfer and after that again the error message.
When I used irdadump to test the USB dongle, the red led inside flashed when it was transmitting data. When I try to connect from the software, I do not see this light.
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #1 from Mikko Rönkkö mikko.ronkko@iki.fi 2009-07-20 15:49:13 --- (In reply to comment #0)
When I used irdadump to test the USB dongle, the red led inside flashed when it was transmitting data. When I try to connect from the software, I do not see this light.
Actually, the indicator light on the dongle flashes when I try to connect with the heart rate monitor from the software. The above is not correct.
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #2 from Mikko Rönkkö mikko.ronkko@iki.fi 2009-07-20 15:52:57 --- Created an attachment (id=22484) --> (http://bugs.winehq.org/attachment.cgi?id=22484) Output on the console with WINEDEBUG=warn+all
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #3 from Juan Lang juan_lang@yahoo.com 2009-07-20 17:34:57 --- (In reply to comment #2)
Output on the console with WINEDEBUG=warn+all
Please don't do that until we ask you to. Please just attach the console output with no WINEDEBUG flags.
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #4 from Mikko Rönkkö mikko.ronkko@iki.fi 2009-07-21 01:15:12 --- Created an attachment (id=22493) --> (http://bugs.winehq.org/attachment.cgi?id=22493) Output on the console wihtout WINEDEBUG
This output was created with the IrDA dongle not attached to the computer. However, the software behaves identically and to me the output on console seems to be exactly the same as with the dongle attached.
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #5 from Juan Lang juan_lang@yahoo.com 2009-07-21 18:07:28 --- Here's a clue:
fixme:winsock:convert_af_w2u unhandled Windows address family 26
From winsock.h, addess family 26 is AF_IRDA. I can send a quick patch for that
fixme, but I expect it'll just lead to failure later.
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #6 from Juan Lang juan_lang@yahoo.com 2009-07-21 18:08:27 --- Created an attachment (id=22516) --> (http://bugs.winehq.org/attachment.cgi?id=22516) AF_IRDA map patch
Does this help?
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #7 from Mikko Rönkkö mikko.ronkko@iki.fi 2009-07-22 04:43:38 --- (In reply to comment #6)
Created an attachment (id=22516)
--> (http://bugs.winehq.org/attachment.cgi?id=22516) [details]
AF_IRDA map patch
Does this help?
I downloaded the sources from git, applied the patch, compiled, and installed. The messages to console have changed a bit, but the behavior of the software seems exactly the same.
I will attach the logs without and with winedebug next.
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #8 from Mikko Rönkkö mikko.ronkko@iki.fi 2009-07-22 04:47:49 --- Created an attachment (id=22520) --> (http://bugs.winehq.org/attachment.cgi?id=22520) Output on the console without WINEDEBUG after applying the patch
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #9 from Mikko Rönkkö mikko.ronkko@iki.fi 2009-07-22 04:54:47 --- Created an attachment (id=22522) --> (http://bugs.winehq.org/attachment.cgi?id=22522) Output on the console with WINEDEBUG=warn+all after applying the patch
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #10 from Juan Lang juan_lang@yahoo.com 2009-07-23 10:25:48 --- (In reply to comment #9)
Output on the console with WINEDEBUG=warn+all after applying the patch
Again, please don't do that unless we ask. Usually these don't tell us anything useful.
Could you post a WINEDEBUG=+winsock log with the patch attached? I'm curious if my patch had anything to do with the problem at all. Thanks.
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #11 from Mikko Rönkkö mikko.ronkko@iki.fi 2009-07-23 14:01:12 --- Created an attachment (id=22563) --> (http://bugs.winehq.org/attachment.cgi?id=22563) Output on the console with WINEDEBUG=+winsock after applying the patch
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #12 from Juan Lang juan_lang@yahoo.com 2009-07-23 15:20:31 --- Created an attachment (id=22565) --> (http://bugs.winehq.org/attachment.cgi?id=22565) IRDA bind patch
Okay, here's another patch to try. Could you apply it, in addition to the other patch, and attach another +winsock log?
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #13 from Mikko Rönkkö mikko.ronkko@iki.fi 2009-07-24 00:45:00 --- Created an attachment (id=22570) --> (http://bugs.winehq.org/attachment.cgi?id=22570) Output on the console with WINEDEBUG=+winsock after applying the second patch
Still no noticeable changes in the behavior of the software.
http://bugs.winehq.org/show_bug.cgi?id=19397
Juan Lang juan_lang@yahoo.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Component|-unknown |winsock
--- Comment #14 from Juan Lang juan_lang@yahoo.com 2009-07-24 10:27:19 --- (In reply to comment #13)
Still no noticeable changes in the behavior of the software.
The log is rather different, there are now lots of winsock calls, finally ending in: fixme:winsock:ws_sockaddr_ws2u Unknown address family 26, return NULL.
This confirms my hunch that it's a winsock bug, IRDA isn't implemented in winsock yet.
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #15 from Mikko Rönkkö mikko.ronkko@iki.fi 2009-07-24 10:45:57 --- (In reply to comment #14)
This confirms my hunch that it's a winsock bug, IRDA isn't implemented in winsock yet.
I assume that this means that fixing the issue will take some time and for now the users of Polar ProTrainer should look for other means to transfer data between the HRM and computer than IrDA under Wine. Is this correct?
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #16 from Juan Lang juan_lang@yahoo.com 2009-07-24 13:35:04 --- (In reply to comment #15)
I assume that this means that fixing the issue will take some time and for now the users of Polar ProTrainer should look for other means to transfer data between the HRM and computer than IrDA under Wine. Is this correct?
I can't help you with that decision. Obviously it doesn't work right now, that's what this bug is about. I have a handful more patches that implement a bit more, I'll attach them and ask for another +winsock log with them applied.
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #17 from Juan Lang juan_lang@yahoo.com 2009-07-24 13:36:53 --- Created an attachment (id=22579) --> (http://bugs.winehq.org/attachment.cgi?id=22579) Patch 1/4: Test for linux/irda.h
First of four patches to be applied on top of the existing two patches.
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #18 from Juan Lang juan_lang@yahoo.com 2009-07-24 13:38:05 --- Created an attachment (id=22580) --> (http://bugs.winehq.org/attachment.cgi?id=22580) Patch 2/4: Add af_irda.h
2nd of 4 patches. This will cause compile warnings due to a redefinition of IAS_MAX_USER_STRING, but the warning is harmless. (I'll fix it before submitting it to winehq.)
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #19 from Juan Lang juan_lang@yahoo.com 2009-07-24 13:40:18 --- Created an attachment (id=22581) --> (http://bugs.winehq.org/attachment.cgi?id=22581) Patch 3/4: Implement ws_sockaddr_ws2u for AF_IRDA
3rd patch of 4. This should address the problem seen in the log.
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #20 from Juan Lang juan_lang@yahoo.com 2009-07-24 13:43:28 --- Created an attachment (id=22582) --> (http://bugs.winehq.org/attachment.cgi?id=22582) Patch 4/4: Implement ws_sockaddr_u2ws for AF_IRDA
Final patch. This wasn't in the log, I'm just guessing it'll be needed too.
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #21 from Juan Lang juan_lang@yahoo.com 2009-07-24 13:47:46 --- Could you apply the 4 patches I just attached on top of the prior two and attach a +winsock log to this bug? You'll need to run autoconf after patch 1/4 is applied, in order to regenerate configure. You'll also need to rerun configure once you've done that. Assuming that patch is named 0003-Test-for-linux-irda.h.txt, you'd want to do something like:
patch -p1 < 0003-Test-for-linux-irda.h.txt autoconf ./configure make
You can defer rebuilding wine (running make) until all the patches are applied if you like.
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #22 from Mikko Rönkkö mikko.ronkko@iki.fi 2009-07-24 14:03:16 --- (In reply to comment #21)
Could you apply the 4 patches I just attached on top of the prior two and attach a +winsock log to this bug? You'll need to run autoconf after patch 1/4 is applied, in order to regenerate configure. You'll also need to rerun configure once you've done that. Assuming that patch is named 0003-Test-for-linux-irda.h.txt, you'd want to do something like:
patch -p1 < 0003-Test-for-linux-irda.h.txt autoconf ./configure make
I started applying the patches and compiling before reading this message. I downloaded the patches to /tmp/patch and then did
git apply /tmp/patch/* configure make depend && make
I think that this does the same as the commands that you posted, but I hope that you correct me if this is wrong. (Make depend is probably not necessary.)
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #23 from Juan Lang juan_lang@yahoo.com 2009-07-24 14:11:40 --- (In reply to comment #22)
I think that this does the same as the commands that you posted, but I hope that you correct me if this is wrong. (Make depend is probably not necessary.)
You're correct that this is the same, and that make depend is probably not necessary. Thanks :)
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #24 from Juan Lang juan_lang@yahoo.com 2009-07-24 14:13:57 --- (In reply to comment #22)
I started applying the patches and compiling before reading this message. I downloaded the patches to /tmp/patch and then did
git apply /tmp/patch/* configure make depend && make
I think that this does the same as the commands that you posted, but I hope that you correct me if this is wrong. (Make depend is probably not necessary.)
Actually, I just caught one problem: you need to run autoconf before running configure. The patch named 1/4 doesn't update configure, just configure.ac, and it needs autoconf to be run to regenerate configure.
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #25 from Mikko Rönkkö mikko.ronkko@iki.fi 2009-07-24 15:29:59 --- Created an attachment (id=22587) --> (http://bugs.winehq.org/attachment.cgi?id=22587) Output on the console with WINEDEBUG=+winsock after applying patches 3-6
The log looks a lot different now. However, the application does not seem to behave any differently. It seems that the software starts data transfer, but the data transfer dialog, which shows text "Reading settings", does not show any progress. I interrupted this log since it seems to start repeating the same thing. I will let the software run a while and post if there are any changes.
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #26 from Juan Lang juan_lang@yahoo.com 2009-07-24 15:54:41 --- (In reply to comment #25)
Created an attachment (id=22587)
--> (http://bugs.winehq.org/attachment.cgi?id=22587) [details]
Output on the console with WINEDEBUG=+winsock after applying patches 3-6
This is in the log: fixme:winsock:ws_sockaddr_ws2u Unknown address family 26, return NULL.
This probably means configure didn't detect <linux/irda.h> on your system, or you didn't run autoconf. Could you tell me if 'grep IRDA include/config.h' produces any output? If it's empty, <linux/irda.h> wasn't found. It should say '#define HAVE_LINUX_IRDA_H 1'.
If it's empty, could you also tell me what 'grep irda config.log' returns?
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #27 from Mikko Rönkkö mikko.ronkko@iki.fi 2009-07-24 16:04:55 --- (In reply to comment #26)
(In reply to comment #25)
Created an attachment (id=22587)
--> (http://bugs.winehq.org/attachment.cgi?id=22587) [details] [details]
Output on the console with WINEDEBUG=+winsock after applying patches 3-6
This is in the log: fixme:winsock:ws_sockaddr_ws2u Unknown address family 26, return NULL.
This probably means configure didn't detect <linux/irda.h> on your system, or you didn't run autoconf. Could you tell me if 'grep IRDA include/config.h' produces any output? If it's empty, <linux/irda.h> wasn't found. It should say '#define HAVE_LINUX_IRDA_H 1'.
It does say that
mronkko@leonardo:~/wine-git$ grep IRDA include/config.h #define HAVE_LINUX_IRDA_H 1
I did run autoconf. Here are the compile commands from history
492 autoconf 493 ./configure 494 nice -n 15 make 495 sudo make install
If it's empty, could you also tell me what 'grep irda config.log' returns?
mronkko@leonardo:~/wine-git$ grep irda config.log configure:6448: checking for linux/irda.h ac_cv_header_linux_irda_h=yes
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #28 from Juan Lang juan_lang@yahoo.com 2009-07-24 17:51:53 --- (In reply to comment #27)
It does say that
mronkko@leonardo:~/wine-git$ grep IRDA include/config.h #define HAVE_LINUX_IRDA_H 1
Odd. I'll attach a couple more patches that'll just convince me better whether this is compiled correctly.
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #29 from Juan Lang juan_lang@yahoo.com 2009-07-24 17:53:37 --- Created an attachment (id=22589) --> (http://bugs.winehq.org/attachment.cgi?id=22589) Patch: Add debug support for IPv6 addresses
Here's a debug patch to apply. It's not directly relevant to IrDA, but the next patch applies after it. Please apply this after all the other patches are applied.
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #30 from Juan Lang juan_lang@yahoo.com 2009-07-24 17:54:34 --- Created an attachment (id=22590) --> (http://bugs.winehq.org/attachment.cgi?id=22590) Patch: Add debug support for IrDA addresses
Please apply this one too, and attach a +winsock log with all these patches applied.
http://bugs.winehq.org/show_bug.cgi?id=19397
Juan Lang juan_lang@yahoo.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|UNCONFIRMED |NEW CC| |juan_lang@yahoo.com Ever Confirmed|0 |1
--- Comment #31 from Juan Lang juan_lang@yahoo.com 2009-07-24 17:56:01 --- Confirming.
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #32 from Mikko Rönkkö mikko.ronkko@iki.fi 2009-07-25 00:58:33 --- (In reply to comment #28)
(In reply to comment #27)
It does say that
mronkko@leonardo:~/wine-git$ grep IRDA include/config.h #define HAVE_LINUX_IRDA_H 1
Odd. I'll attach a couple more patches that'll just convince me better whether this is compiled correctly.
It seems that there is something wrong with the patches. When I apply the patches against version 1.1.26 in the order posted in this bug, I get
mronkko@leonardo:~/wine-git$ git checkout wine-1.1.26 -f HEAD is now at 578aeb8... Release 1.1.26. mronkko@leonardo:~/wine-git$ ls -1 ../attachment.cgi?id=225* ../attachment.cgi?id=22516 ../attachment.cgi?id=22565 ../attachment.cgi?id=22579 ../attachment.cgi?id=22580 ../attachment.cgi?id=22581 ../attachment.cgi?id=22582 ../attachment.cgi?id=22589 ../attachment.cgi?id=22590 mronkko@leonardo:~/wine-git$ ls -1 ../attachment.cgi?id=225* | xargs -L 1 git apply error: include/af_irda.h: already exists in working directory mronkko@leonardo:~/wine-git$ autoconf mronkko@leonardo:~/wine-git$ nice -n 15 ./configure && nice -n 15 make
The problem is in attachment id 22580. I am now compiling with all the patches.
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #33 from Mikko Rönkkö mikko.ronkko@iki.fi 2009-07-25 04:29:04 --- Created an attachment (id=22601) --> (http://bugs.winehq.org/attachment.cgi?id=22601) 22587: Output on the console with WINEDEBUG=+winsock after applying patches 1-8
Here is a log file after applying all the patches. I started with fresh sources and now all the patches are applied properly. During this run I did not place the heart rate monitor in the range of the IrDA dongle. However, I do not think that the ProTrainer oftware was able to notice this.
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #34 from Juan Lang juan_lang@yahoo.com 2009-07-25 07:50:57 --- (In reply to comment #33)
Here is a log file after applying all the patches.
Well, well. We appear to be getting somewhere. Now the log shows this: trace:winsock:WS_getsockopt socket: 0098, level 0xff, name 0x10, ptr 0x1d9e92c, len 297 fixme:winsock:WS_getsockopt Unknown level: 0x000000ff
I believe level 0xff is SOL_IRLMP, and name 0x10 is IRLMP_ENUMDEVICES. I'll have another look Monday or so.
http://bugs.winehq.org/show_bug.cgi?id=19397
kriberg@tihlde.org changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |kriberg@tihlde.org
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #35 from Juan Lang juan_lang@yahoo.com 2009-07-27 11:32:07 --- Created an attachment (id=22644) --> (http://bugs.winehq.org/attachment.cgi?id=22644) Patch: add a few more definitions
First in a two-part patch series to address the getsockopt in the log.
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #36 from Juan Lang juan_lang@yahoo.com 2009-07-27 11:33:29 --- Created an attachment (id=22645) --> (http://bugs.winehq.org/attachment.cgi?id=22645) Patch: implement getsockopt(SOL_IRLMP, IRLMP_ENUMDEVICES)
2nd patch in this series. Could you apply these two on top of all the existing patches and attach a +winsock log?
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #37 from Mikko Rönkkö mikko.ronkko@iki.fi 2009-07-27 11:44:08 --- (In reply to comment #36)
Created an attachment (id=22645)
--> (http://bugs.winehq.org/attachment.cgi?id=22645) [details]
Patch: implement getsockopt(SOL_IRLMP, IRLMP_ENUMDEVICES)
2nd patch in this series. Could you apply these two on top of all the existing patches and attach a +winsock log?
I will test these patches later today (at around midnight, GMT+2)
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #38 from Mikko Rönkkö mikko.ronkko@iki.fi 2009-07-27 16:22:24 --- (In reply to comment #35)
Created an attachment (id=22644)
--> (http://bugs.winehq.org/attachment.cgi?id=22644) [details]
Patch: add a few more definitions
First in a two-part patch series to address the getsockopt in the log.
This patch does not seem to work. I took a fresh copy of release 1.1.26 and applied all patches in the order they were posted.
Here is my compile script and the output from patching and compiling with it.
mronkko@leonardo:~$ cat compilewine.sh #!/bin/bash
cd wine-git rm -r * git checkout wine-1.1.26 -f
for f in `ls ../attachment.cgi?id=22*` do echo "Applying patch from file $f" git apply $f done
read -p "Press enter to configure and compile"
autoconf ./configure make mronkko@leonardo:~$ ./compilewine.sh Checking out files: 100% (5975/5975), done. HEAD is now at 578aeb8... Release 1.1.26. Applying patch from file ../attachment.cgi?id=22516 Applying patch from file ../attachment.cgi?id=22565 Applying patch from file ../attachment.cgi?id=22579 Applying patch from file ../attachment.cgi?id=22580 Applying patch from file ../attachment.cgi?id=22581 Applying patch from file ../attachment.cgi?id=22582 Applying patch from file ../attachment.cgi?id=22589 Applying patch from file ../attachment.cgi?id=22590 Applying patch from file ../attachment.cgi?id=22644 error: patch failed: include/af_irda.h:41 error: include/af_irda.h: patch does not apply Applying patch from file ../attachment.cgi?id=22645 Press enter to configure and compile^C mronkko@leonardo:~$
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #39 from Juan Lang juan_lang@yahoo.com 2009-07-27 16:34:01 --- (In reply to comment #38)
This patch does not seem to work. I took a fresh copy of release 1.1.26 and applied all patches in the order they were posted.
Okay, I'll upload a fresh set of patches in a few minutes, in case they've gotten out of sync.
http://bugs.winehq.org/show_bug.cgi?id=19397
Juan Lang juan_lang@yahoo.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Attachment #22516|0 |1 is obsolete| | Attachment #22565|0 |1 is obsolete| | Attachment #22579|0 |1 is obsolete| | Attachment #22580|0 |1 is obsolete| | Attachment #22581|0 |1 is obsolete| | Attachment #22582|0 |1 is obsolete| | Attachment #22589|0 |1 is obsolete| | Attachment #22590|0 |1 is obsolete| | Attachment #22644|0 |1 is obsolete| | Attachment #22645|0 |1 is obsolete| |
--- Comment #40 from Juan Lang juan_lang@yahoo.com 2009-07-27 16:36:27 --- Created an attachment (id=22647) --> (http://bugs.winehq.org/attachment.cgi?id=22647) New patchset
This tarball contains all of the previous patches, consolidated a bit. Please apply them all, run autoconf, configure, and attach a +winsock log from the resulting build. Thanks!
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #41 from Mikko Rönkkö mikko.ronkko@iki.fi 2009-07-27 16:59:52 --- OK, it is now compiling.
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #42 from Mikko Rönkkö mikko.ronkko@iki.fi 2009-07-27 18:23:48 --- Created an attachment (id=22650) --> (http://bugs.winehq.org/attachment.cgi?id=22650) Log after the large patch set WINEDEBUG=+winsock
Although there are still no changes in the userinterface, it is clearly getting further. The messages in the console are now different when the heart rate monitor is in range and when it is not. Clearly some data is moving now.
Is there any way to compile wine only partially after applying patches? It takes quite a bit of time to recompile the whole Wine.
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #43 from Juan Lang juan_lang@yahoo.com 2009-07-27 18:32:17 --- (In reply to comment #42)
Is there any way to compile wine only partially after applying patches? It takes quite a bit of time to recompile the whole Wine.
Yes. Start with a wine tree that you've already built. Then apply the patches, run autoconf and configure, and: cd dlls/ws2_32 make This'll just rebuild ws2_32.dll, which is all that's really changed from the stock tree.
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #44 from Juan Lang juan_lang@yahoo.com 2009-07-27 18:46:48 --- (In reply to comment #42)
Although there are still no changes in the userinterface, it is clearly getting further. The messages in the console are now different when the heart rate monitor is in range and when it is not. Clearly some data is moving now.
Could you attach the output from irdadump while running this?
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #45 from Juan Lang juan_lang@yahoo.com 2009-07-27 19:20:29 --- Created an attachment (id=22652) --> (http://bugs.winehq.org/attachment.cgi?id=22652) Patch: store UNIX IrDA addresses in network order
On a hunch, could you apply this on top of the big patchset and try it? I don't have any IrDA devices myself, so I'm having to guess a little at this point.
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #46 from Mikko Rönkkö mikko.ronkko@iki.fi 2009-07-28 13:00:18 --- Created an attachment (id=22663) --> (http://bugs.winehq.org/attachment.cgi?id=22663) Irdadump
Here is irdadump after applying the large set of patches. As far as I understand it, nothing interesting in this file.
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #47 from Juan Lang juan_lang@yahoo.com 2009-07-28 13:11:07 --- (In reply to comment #46)
Here is irdadump after applying the large set of patches. As far as I understand it, nothing interesting in this file.
This looks like a +winsock log, not irdadump's output. Did you attach the wrong file?
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #48 from Mikko Rönkkö mikko.ronkko@iki.fi 2009-07-28 13:25:25 --- (In reply to comment #47)
(In reply to comment #46)
Here is irdadump after applying the large set of patches. As far as I understand it, nothing interesting in this file.
This looks like a +winsock log, not irdadump's output. Did you attach the wrong file?
Here is the content of the file
7:57:38.861765 xid:cmd eec96579 > ffffffff S=6 s=0 (14) 17:57:38.949733 xid:cmd eec96579 > ffffffff S=6 s=1 (14) 17:57:39.037753 xid:cmd eec96579 > ffffffff S=6 s=2 (14) 17:57:39.125744 xid:cmd eec96579 > ffffffff S=6 s=3 (14) 17:57:39.205747 xid:rsp eec96579 < 5ea0c8cb S=6 s=3 Polar RS800 hint=8204 [ PDA/Palmtop IrCOMM ] (28) 17:57:39.213732 xid:cmd eec96579 > ffffffff S=6 s=4 (14) 17:57:39.301756 xid:cmd eec96579 > ffffffff S=6 s=5 (14) 17:57:39.389752 xid:cmd eec96579 > ffffffff S=6 s=* leonardo hint=0400 [ Computer ] (24) 17:57:41.861759 xid:cmd eec96579 > ffffffff S=6 s=0 (14) 17:57:41.941566 xid:rsp eec96579 < 5ea0c8cb S=6 s=0 Polar RS800 hint=8204 [ PDA/Palmtop IrCOMM ] (28) 17:57:41.949753 xid:cmd eec96579 > ffffffff S=6 s=1 (14) 17:57:42.037753 xid:cmd eec96579 > ffffffff S=6 s=2 (14) 17:57:42.125740 xid:cmd eec96579 > ffffffff S=6 s=3 (14) 17:57:42.213754 xid:cmd eec96579 > ffffffff S=6 s=4 (14) 17:57:42.301764 xid:cmd eec96579 > ffffffff S=6 s=5 (14) 17:57:42.389750 xid:cmd eec96579 > ffffffff S=6 s=* leonardo hint=0400 [ Computer ] (24)
This keeps repeating.
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #49 from Mikko Rönkkö mikko.ronkko@iki.fi 2009-07-28 13:31:04 --- (In reply to comment #45)
Created an attachment (id=22652)
--> (http://bugs.winehq.org/attachment.cgi?id=22652) [details]
Patch: store UNIX IrDA addresses in network order
On a hunch, could you apply this on top of the big patchset and try it? I don't have any IrDA devices myself, so I'm having to guess a little at this point.
This patch did not seem to produce anything new in the log.
http://bugs.winehq.org/show_bug.cgi?id=19397
Juan Lang juan_lang@yahoo.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Attachment #22652|0 |1 is obsolete| |
--- Comment #50 from Juan Lang juan_lang@yahoo.com 2009-07-28 13:55:47 --- Created an attachment (id=22666) --> (http://bugs.winehq.org/attachment.cgi?id=22666) Patch: Trace result of enumerating IrDA devices
Could you apply this one, then, and attach a +winsock patch with it applied? I'm obsoleting the network order patch since you said it had no impact, but it doesn't really matter if you apply this with or without that one applied. Thanks.
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #51 from Mikko Rönkkö mikko.ronkko@iki.fi 2009-07-28 14:24:00 --- Created an attachment (id=22667) --> (http://bugs.winehq.org/attachment.cgi?id=22667) Log after patch
No visible changes in the software and the log looks the same. When I started running the software initially, I had the IR receiver in the heart rate monitor switched off. This resulted in errors "Resource temporarily unavailable". After switching it on, I started getting "No route to host".
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #52 from Juan Lang juan_lang@yahoo.com 2009-07-28 16:46:55 --- (In reply to comment #51)
No visible changes in the software and the log looks the same. When I started running the software initially, I had the IR receiver in the heart rate monitor switched off. This resulted in errors "Resource temporarily unavailable". After switching it on, I started getting "No route to host".
Was this log generated with the patch "Store UNIX IrDA addresses in network order" applied, or not applied?
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #53 from Mikko Rönkkö mikko.ronkko@iki.fi 2009-07-29 00:04:38 --- (In reply to comment #52)
(In reply to comment #51)
No visible changes in the software and the log looks the same. When I started running the software initially, I had the IR receiver in the heart rate monitor switched off. This resulted in errors "Resource temporarily unavailable". After switching it on, I started getting "No route to host".
Was this log generated with the patch "Store UNIX IrDA addresses in network order" applied, or not applied?
The patch was applied.
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #54 from Juan Lang juan_lang@yahoo.com 2009-07-29 10:14:16 --- (In reply to comment #53)
The patch was applied.
Okay. Any chance of getting a +winsock log without that patch applied, but with the final patch applied?
You can reverse the patch with 'patch -p1 -R < path_to_patch_to_undo'.
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #55 from Mikko Rönkkö mikko.ronkko@iki.fi 2009-07-29 11:36:49 --- Created an attachment (id=22678) --> (http://bugs.winehq.org/attachment.cgi?id=22678) Log after removing a patch
Now it looks a bit different. The model of the HRM appears in the log.
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #56 from Mikko Rönkkö mikko.ronkko@iki.fi 2009-07-29 11:39:06 --- (In reply to comment #55)
Created an attachment (id=22678)
--> (http://bugs.winehq.org/attachment.cgi?id=22678) [details]
Log after removing a patch
Now it looks a bit different. The model of the HRM appears in the log.
Irdadump does not show anything interesting. Apparently the devices are just announcing each other or something similar
16:37:38.203272 xid:rsp 95177b76 < 3f7467a1 S=6 s=5 Polar RS800 hint=8204 [ PDA/Palmtop IrCOMM ] (28) 16:37:38.211430 xid:cmd 95177b76 > ffffffff S=6 s=* leonardo hint=0400 [ Computer ] (24)
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #57 from Juan Lang juan_lang@yahoo.com 2009-07-29 12:00:58 --- Created an attachment (id=22680) --> (http://bugs.winehq.org/attachment.cgi?id=22680) Patch: Copy enumerated address in host order
Could you apply this patch and attach a +winsock log?
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #58 from Mikko Rönkkö mikko.ronkko@iki.fi 2009-07-29 13:12:09 --- Created an attachment (id=22683) --> (http://bugs.winehq.org/attachment.cgi?id=22683) Log after latest patch
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #59 from Juan Lang juan_lang@yahoo.com 2009-07-29 13:16:19 --- (In reply to comment #58)
Log after latest patch
Thanks. I wonder if I've been barking up the wrong tree entirely. Could you back out that patch, and try the next one in its place?
http://bugs.winehq.org/show_bug.cgi?id=19397
Juan Lang juan_lang@yahoo.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Attachment #22680|0 |1 is obsolete| |
--- Comment #60 from Juan Lang juan_lang@yahoo.com 2009-07-29 13:18:06 --- Created an attachment (id=22684) --> (http://bugs.winehq.org/attachment.cgi?id=22684) Patch: Return destination rather than source address when enumerating IrDA devices
Next patch to try. Could you attach a +winsock log with this one?
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #61 from Mikko Rönkkö mikko.ronkko@iki.fi 2009-07-29 13:45:12 --- Created an attachment (id=22685) --> (http://bugs.winehq.org/attachment.cgi?id=22685) Log after latest patch
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #62 from Juan Lang juan_lang@yahoo.com 2009-07-29 13:53:43 --- (In reply to comment #61)
Log after latest patch
Thanks. It *seems* to be getting further, as instead of getting: warn:winsock:wsaErrno errno 113, (No route to host). we're now getting: warn:winsock:wsaErrno errno 107, (Transport endpoint is not connected).
Could you attach the output from irdadump?
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #63 from Mikko Rönkkö mikko.ronkko@iki.fi 2009-07-29 14:01:00 --- Created an attachment (id=22686) --> (http://bugs.winehq.org/attachment.cgi?id=22686) Irdadump
Now there seems to be actually some data moving.
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #64 from Juan Lang juan_lang@yahoo.com 2009-07-29 15:28:57 --- (In reply to comment #63)
Now there seems to be actually some data moving.
Agreed, so the last patch appears to be correct. From irdadump:
8:58:48.831478 i:cmd > ca=2e pf=1 nr=0 ns=0 LM slsap=5b dlsap=00 CONN_CMD (6) 18:58:48.863541 i:rsp < ca=2e pf=1 nr=1 ns=0 LM slsap=00 dlsap=5b CONN_RSP (5) 18:58:48.863603 i:cmd > ca=2e pf=1 nr=1 ns=1 LM slsap=5b dlsap=00 GET_VALUE_BY_CLASS: "HRM" "IrDA:TinyTP:LsapSel" (29) 18:58:48.907545 i:rsp < ca=2e pf=1 nr=2 ns=1 LM slsap=00 dlsap=5b GET_VALUE_BY_CLASS: Success Integer: 43 (15) 18:58:48.907625 i:cmd > ca=2e pf=1 nr=2 ns=2 LM slsap=5b dlsap=00 DISC (6) 18:58:48.938557 rr:rsp < ca=2e pf=1 nr=3 (2) 18:58:48.938606 i:cmd > ca=2e pf=1 nr=2 ns=3 LM slsap=5c dlsap=43 CONN_CMD TTP credits=16 (7) 18:58:48.971811 i:rsp < ca=2e pf=1 nr=4 ns=2 LM slsap=43 dlsap=5c CONN_RSP TTP credits=1 (7)
However the socket complains that it's not connected: trace:winsock:WSASendTo socket 0088, wsabuf 0x1d9ea28, nbufs 1, flags 0, to (nil), tolen 0, ovl (nil), func (nil) trace:winsock:WSASendTo fd=35, options=0 warn:winsock:wsaErrno errno 107, (Transport endpoint is not connected).
Looking at the log above, it appears as though there's a connection made (CONN_CMD), then it's disconnected (DISC), and the subsequent connect request (CONN_CMD) is the last thing done.
I'm frankly a bit out of my league at this point. I've never worked with IrDA before, and I don't have any devices to test with myself. I'll send in my patch set since it appears to help the situation a bit, but I'll have to hope that someone with more knowledge in the area steps in.
Out of curiosity, did this USB dongle come from Polar? Is there a different one you can try?
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #65 from Mikko Rönkkö mikko.ronkko@iki.fi 2009-07-29 15:42:16 --- (In reply to comment #64)
Out of curiosity, did this USB dongle come from Polar? Is there a different one you can try?
The one I am testing with is from Polar. It is the version that comes with the heart rate monitor. I have also another dongle, but it is not set up with the computer that I am testing with. It does not "work out off the box", so it will take some time before I can test it. Probably next week.
Thanks.
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #66 from Juan Lang juan_lang@yahoo.com 2009-07-30 13:39:50 --- You may also want to try reading through the Linux IrDA troubleshooting hints here: http://www.hpl.hp.com/personal/Jean_Tourrilhes/IrDA/#checking
This one in particular might be worth trying: echo 1 > /proc/sys/net/irda/max_tx_window and try again. I think it's a long shot, but it can't hurt to try, right?
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #67 from Mikko Rönkkö mikko.ronkko@iki.fi 2009-07-30 13:46:58 --- (In reply to comment #66)
You may also want to try reading through the Linux IrDA troubleshooting hints here: http://www.hpl.hp.com/personal/Jean_Tourrilhes/IrDA/#checking
This one in particular might be worth trying: echo 1 > /proc/sys/net/irda/max_tx_window and try again. I think it's a long shot, but it can't hurt to try, right?
I will try this and also with the other IrDA dongle, but it will be next Wednesday before I can do that. I feel good about this and believe that the software will eventually work.
http://bugs.winehq.org/show_bug.cgi?id=19397
Ken Sharp kennybobs@o2.co.uk changed:
What |Removed |Added ---------------------------------------------------------------------------- Keywords| |hardware
http://bugs.winehq.org/show_bug.cgi?id=19397
Juan Lang juan_lang@yahoo.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Attachment #22647|0 |1 is obsolete| | Attachment #22666|0 |1 is obsolete| | Attachment #22684|0 |1 is obsolete| |
--- Comment #68 from Juan Lang juan_lang@yahoo.com 2009-08-04 19:26:50 --- Created an attachment (id=22816) --> (http://bugs.winehq.org/attachment.cgi?id=22816) Patchset, resynced
This patchset isn't functionally different from the patches already attached, but it's resynced against current git.
http://bugs.winehq.org/show_bug.cgi?id=19397
Erki Hallingu erkiha@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |erkiha@gmail.com
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #69 from Juan Lang juan_lang@yahoo.com 2009-08-11 11:39:33 --- I sent in the current patchset, starting here: http://www.winehq.org/pipermail/wine-patches/2009-August/076871.html This is just to make it easier to test. Assuming the patches get accepted, it should be possible to use vanilla Wine 1.1.28, once it's released.
http://bugs.winehq.org/show_bug.cgi?id=19397
Juan Lang juan_lang@yahoo.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Attachment #22816|0 |1 is obsolete| |
--- Comment #70 from Juan Lang juan_lang@yahoo.com 2009-08-13 10:51:25 --- (From update of attachment 22816) The patchset was committed to git today. The committed patches fixed a couple bugs, and one of them could have impacted the application. Would you mind retrying with current git, or with 1.1.28 when it comes out?
http://bugs.winehq.org/show_bug.cgi?id=19397
Fredrik Eliasson bz@avro.se changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |bz@avro.se
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #71 from Mikko Rönkkö mikko.ronkko@iki.fi 2009-08-17 12:26:14 --- (In reply to comment #70)
(From update of attachment 22816 [details]) The patchset was committed to git today. The committed patches fixed a couple bugs, and one of them could have impacted the application. Would you mind retrying with current git, or with 1.1.28 when it comes out?
I will try to test this during the next couple of days.
http://bugs.winehq.org/show_bug.cgi?id=19397
Mikko Rönkkö mikko.ronkko@iki.fi changed:
What |Removed |Added ---------------------------------------------------------------------------- Platform|PC |All Version|1.1.26 |1.1.33 OS/Version|Linux |All
--- Comment #72 from Mikko Rönkkö mikko.ronkko@iki.fi 2009-11-29 16:48:27 --- I just tested this bug with Wine 1.1.33 on OS X 10.6.2 and get the same error message as in comment number #5
fixme:winsock:convert_af_w2u unhandled Windows address family 26
Are the patches still valid for this version of Wine?
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #73 from Mikko Rönkkö mikko.ronkko@iki.fi 2009-11-29 16:52:31 --- Just an additional note if it helps. My wine is compiled from git. The version string is
$ wine --version wine-1.1.33-404-gac85305
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #74 from Juan Lang juan_lang@yahoo.com 2009-11-29 17:23:25 --- (In reply to comment #72)
I just tested this bug with Wine 1.1.33 on OS X 10.6.2 and get the same error message as in comment number #5
fixme:winsock:convert_af_w2u unhandled Windows address family 26
The patches only work at Linux at the moment, I'm afraid. There is apparently some support for IrDA in Leopard and newer MacOS via the AppleUSBIrDA driver, but I don't know how to program it (and I don't have a Mac, so I won't be of use to you here.)
Are the patches still valid for this version of Wine?
Wine version 1.1.33 already incorporates these patches, so you don't need to add them to Wine.
http://bugs.winehq.org/show_bug.cgi?id=19397
Vitaliy Margolen vitaliy@kievinfo.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Platform|All |PC Version|1.1.33 |1.1.26 OS/Version|All |Linux
--- Comment #75 from Vitaliy Margolen vitaliy@kievinfo.com 2009-11-29 17:40:21 --- Don't change original reported Wine version.
http://bugs.winehq.org/show_bug.cgi?id=19397
Mikko Rönkkö mikko.ronkko@iki.fi changed:
What |Removed |Added ---------------------------------------------------------------------------- Platform|PC |All Version|1.1.26 |1.1.33 OS/Version|Linux |All
--- Comment #76 from Mikko Rönkkö mikko.ronkko@iki.fi 2009-11-30 02:13:28 ---
(In reply to comment #74)
The patches only work at Linux at the moment, I'm afraid. There is apparently some support for IrDA in Leopard and newer MacOS via the AppleUSBIrDA driver, but I don't know how to program it (and I don't have a Mac, so I won't be of use to you here.)
This is correct: IrDA works in Leopard and newer versions with the AppleUSBIrDA driver. The source code of the driver is available at http://www.opensource.apple.com/source/AppleUSBIrDA/AppleUSBIrDA-131.2.2/ if someone who has a Mac is willing to take a look at the issue.
This is a bit off-topic, but a IrDA driver that is compatible with the Polar USB dongle can be found at http://www.ismarttrain.com/drivers/
http://bugs.winehq.org/show_bug.cgi?id=19397
Nikolay Sivov bunglehead@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Version|1.1.33 |1.1.26
--- Comment #77 from Nikolay Sivov bunglehead@gmail.com 2009-11-30 02:24:21 --- (In reply to comment #76)
(In reply to comment #74)
Mikko, again. Please don't change version field.
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #78 from Mikko Rönkkö mikko.ronkko@iki.fi 2009-11-30 02:31:07 --- (In reply to comment #77)
Mikko, again. Please don't change version field.
Sorry. I did not change it intentionally, but my browser's autofill did based on the previous choice.
http://bugs.winehq.org/show_bug.cgi?id=19397
Dmitry Timoshkov dmitry@codeweavers.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Platform|All |PC OS/Version|All |Linux
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #79 from Mikko Rönkkö mikko.ronkko@iki.fi 2009-11-30 15:32:48 --- I tested the software on Linux again without success. Also limiting the baud rate or transfer window size did not help. (These were suggested at http://www.hpl.hp.com/personal/Jean_Tourrilhes/IrDA/#checking)
While testing, I noted these in these lines kernel log when the socked disconnected:
Nov 30 23:14:54 Leonardo kernel: [ 4126.098252] __ratelimit: 62 callbacks suppressed Nov 30 23:14:54 Leonardo kernel: [ 4126.098258] irttp_data_request: No data, or not connected
One thing that I could test is to run the software in Windows and see how the IrDA dump differs. However, I did not find anything like irdadump for Windows with Google. Any suggestions how to dump the IrDA traffic in windows?
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #80 from Mikko Rönkkö mikko.ronkko@iki.fi 2009-12-02 07:03:39 --- I filed this bug also two CodeWeavers support ticket system
http://www.codeweavers.com/support/tickets/browse/?ticket_id=774202;list=13;...
http://bugs.winehq.org/show_bug.cgi?id=19397
Adrian Rueegsegger ken@codelabs.ch changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |ken@codelabs.ch
--- Comment #81 from Adrian Rueegsegger ken@codelabs.ch 2010-04-25 07:39:00 --- (In reply to comment #79)
I tested the software on Linux again without success. Also limiting the baud rate or transfer window size did not help. (These were suggested at http://www.hpl.hp.com/personal/Jean_Tourrilhes/IrDA/#checking)
While testing, I noted these in these lines kernel log when the socked disconnected:
Nov 30 23:14:54 Leonardo kernel: [ 4126.098252] __ratelimit: 62 callbacks suppressed Nov 30 23:14:54 Leonardo kernel: [ 4126.098258] irttp_data_request: No data, or not connected
I can confirm this with the current (033bb8a...) git version.
One thing that I could test is to run the software in Windows and see how the IrDA dump differs. However, I did not find anything like irdadump for Windows with Google. Any suggestions how to dump the IrDA traffic in windows?
There is a irdadump windows port [1]. Since I don't have access to a Windows Workstation I don't know if the port is any good and can't produce a trace...
[1] - http://www.alanjmcf.me.uk/comms/infrared/Windows irdadump port/
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #82 from wolfgang.schwotzer@gmx.net 2010-07-18 15:51:48 --- Created an attachment (id=29696) --> (http://bugs.winehq.org/attachment.cgi?id=29696) 0001-wine-bugzilla-19397.patch
http://bugs.winehq.org/show_bug.cgi?id=19397
wolfgang.schwotzer@gmx.net changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |wolfgang.schwotzer@gmx.net
--- Comment #83 from wolfgang.schwotzer@gmx.net 2010-07-18 16:03:53 --- (In reply to comment #79) There is another irdadump for window available. It also supports IrDA over USB: http://www.ircomm2k.de/English/index.html
The error reason is: Polar Protrainer 5 for any reason calls WSASendTo with a buffer size of 0. This results in errno 107 when returning from sendmsg. Within WS2_send The patch prevents calling sendmsg and returns by returning 0 (= number of sent bytes). The Polar Protrainer 5 IrDA communication runs perfectly with this patch.
http://bugs.winehq.org/show_bug.cgi?id=19397
mike.evans@cardolan.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |mike.evans@cardolan.com
--- Comment #84 from mike.evans@cardolan.com 2010-08-11 09:19:38 --- (In reply to comment #83)
The Polar Protrainer 5 IrDA communication runs perfectly with this patch.
Will this patch appear in the next release, or is this something I'll need to learn how to apply myself? I'm keen to finally get ProTrainer working fully under Wine.
http://bugs.winehq.org/show_bug.cgi?id=19397
Saulius K. saulius2@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |saulius2@gmail.com
http://bugs.winehq.org/show_bug.cgi?id=19397
wolfgang.schwotzer@gmx.net changed:
What |Removed |Added ---------------------------------------------------------------------------- Attachment #29696|0 |1 is obsolete| |
--- Comment #85 from wolfgang.schwotzer@gmx.net 2010-08-26 14:49:09 --- (From update of attachment 29696) Patch has been rejected. Reason: It would hide potential errors when sending zero bytes on non IrDA stacks.
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #86 from wolfgang.schwotzer@gmx.net 2010-08-26 15:27:51 --- Status update: - The error is still reproducible on Polar ProTrainer V5.40.171 - The data transmition in Polar Pro Trainer has type=SOCK_STREAM. It works like this: - WS2_sendto some bytes - WS2_recvfrom some bytes - 5-7 times: WS2_sendto zero bytes ... - WS2_sendto some bytes - WS2_recvfrom some bytes - 5-7 times: WS2_sendto zero bytes
- The errno ENOTCONN (107) is created in the Linux IrDA stack in net/irda/irttp.c, function irttp_data_request: http://lxr.linux.no/linux+v2.6.35/net/irda/irttp.c#L590 - As already mentioned above (See comment #79) a kernel log entry is written when buffer size is zero: ....irttp_data_request: No data, or not connected - The check for skb->len == 0 is already part of the Linux kernel since IrDA stack has been included in linux-2.2.0 (26-Jan-1999). - The Windows/Linux AF_INET stack returns 0 when sending a buffer with size 0 (Tested on all 32/64bit Windows Versions, Linux 2.6.32) - The Windows AF_IRDA stack returns 0 when sending a buffer with size 0 (Tested on Windows XP)
Next steps: - I've posted a message to the Linux IrDA newsgroup to evtl. get more information about the Linux IrDA stack, waiting for feedback: http://thread.gmane.org/gmane.linux.irda.general/1643 - I will write a wine test for AF_IRDA to get the current behavior on all Windows versions
http://bugs.winehq.org/show_bug.cgi?id=19397
Wolfram Sang wolfram@the-dreams.de changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |wolfram@the-dreams.de
--- Comment #87 from Wolfram Sang wolfram@the-dreams.de 2010-09-28 08:31:21 CDT --- Your mail to irda-users looks very reasonable to me, just the list appears to be dead :( My suggestion would be to resend it and also CC the IRDA maintainer (Samuel Ortiz) and the more generic netdev-list. Also make sure your example-source does not get line-wrapped (it appears to be in the gmane-archive).
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #88 from Wolfram Sang wolfram@the-dreams.de 2010-11-07 15:43:57 CST --- Created an attachment (id=31789) --> (http://bugs.winehq.org/attachment.cgi?id=31789) _Linux kernel_ patch to allow zero byte packets
This _Linux kernel_ (not WINE) patch should take care of the issue at kernel level. Before I post it to the relevant mailing lists for discussion, I'd like to have at least one report saying that this patch really helps :) I created it on top of 2.6.37-rc1 but the irda-layer hasn't been touched much, so it will probably apply to various older kernels, too. So, if someone feels like verifying this patch actually works, that would be of great help! Thanks.
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #89 from mike.evans@cardolan.com 2010-11-08 16:06:05 CST --- (In reply to comment #88)
So, if someone feels like verifying this patch actually works, that would be of great help! Thanks.
I'm going to attempt to test this, but it'll be my first attempt at patching and building a kernel, so it might take me a while to find out how. I'll report back if I get a result (or get stuck).
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #90 from mike.evans@cardolan.com 2010-11-09 16:25:50 CST --- (In reply to comment #89)
I'm going to attempt to test this, but it'll be my first attempt at patching and building a kernel, so it might take me a while to find out how. I'll report back if I get a result (or get stuck).
OK, so patching and building the kernel turned out to be relatively straightforward, and I can confirm that the patch seems to work fine, and I've just transferred the run I did on Sunday from my RS800CX to Protrainer. Thanks for your help here Wolfram (and Wolfgang) - I hope the patch makes it into the kernel.
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #91 from Wolfram Sang wolfram@the-dreams.de 2010-11-09 17:40:12 CST --- Awesome, thanks for testing! Patch sent, let's see what happens:
http://thread.gmane.org/gmane.linux.network/177511
http://bugs.winehq.org/show_bug.cgi?id=19397
--- Comment #92 from Wolfram Sang wolfram@the-dreams.de 2010-11-18 16:23:02 CST --- Good news, the patch made it to the net-tree and will thus be part of the next kernel release (which is 2.6.37). This application should then work out of the box.
http://git.kernel.org/?p=linux/kernel/git/davem/net-2.6.git;a=commitdiff;h=4...
http://bugs.winehq.org/show_bug.cgi?id=19397
Dmitry Timoshkov dmitry@codeweavers.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |INVALID Summary|Polar ProTrainer 5: IRDA |Polar ProTrainer 5: IRDA |connection to heart rate |connection to heart rate |monitor does not work |monitor does not work (not | |a Wine bug)
--- Comment #93 from Dmitry Timoshkov dmitry@codeweavers.com 2010-11-20 05:00:24 CST --- Marking this bug as invalid then.
http://bugs.winehq.org/show_bug.cgi?id=19397
Dmitry Timoshkov dmitry@codeweavers.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|RESOLVED |CLOSED
--- Comment #94 from Dmitry Timoshkov dmitry@codeweavers.com 2010-11-20 05:00:42 CST --- Closing invalid.
-
wine-bugs＠winehq.org