http://bugs.winehq.org/show_bug.cgi?id=21938
Summary: Test tool breaks secured wireless on windows 7 x64 Product: Wine Version: 1.1.37 Platform: x86-64 OS/Version: other Status: UNCONFIRMED Severity: blocker Priority: P2 Component: -unknown AssignedTo: wine-bugs@winehq.org ReportedBy: maxim@ipact.nl
Hello,
I'm having a huge problem with your unit test tool to run on windows I have a 64 Bit windows 7 laptop and I have run the test tool on it
Now the laptop can't use the wireless card properly anymore It can connect to open networks but not to secured networks of any type (WEP, WPA, WPA2) I tried all kinds of drivers and that was no help....
I tried the current test tool for version 1.1.40 aswell on my friends laptop, he cant use his wireless now aswell, hes runing windows 7 x64 aswell.
What does this test tool do and can it be reversed?
Cheers.
http://bugs.winehq.org/show_bug.cgi?id=21938
Ben Klein shacklein@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |shacklein@gmail.com
--- Comment #1 from Ben Klein shacklein@gmail.com 2010-03-06 06:58:00 --- (In reply to comment #0)
What does this test tool do and can it be reversed?
What it *doesn't* do is break your host operating system. Your problems most certainly are not related to Wine.
http://bugs.winehq.org/show_bug.cgi?id=21938
--- Comment #2 from Nikolay Sivov bunglehead@gmail.com 2010-03-06 07:10:26 --- (In reply to comment #1)
(In reply to comment #0)
What does this test tool do and can it be reversed?
What it *doesn't* do is break your host operating system. Your problems most certainly are not related to Wine.
Ben, please avoid such comments, you can't be sure about it.
For example I constantly get:
- hard reboot on d3d tests on my XP box; - corrupted tray area after some user32 tests; - misplaced icons in start menu, etc.
Probably some service just stopped after tests or something like that.
Maxim, did you try 64-bit winetest or 32-bit one? Did you run whole suite or some particular part?
http://bugs.winehq.org/show_bug.cgi?id=21938
--- Comment #3 from Ben Klein shacklein@gmail.com 2010-03-06 07:46:42 --- (In reply to comment #2)
(In reply to comment #1)
(In reply to comment #0)
What does this test tool do and can it be reversed?
What it *doesn't* do is break your host operating system. Your problems most certainly are not related to Wine.
Ben, please avoid such comments, you can't be sure about it.
For example I constantly get:
- hard reboot on d3d tests on my XP box;
- corrupted tray area after some user32 tests;
- misplaced icons in start menu, etc.
In none of these cases would a long-term driver issue occur.
Probably some service just stopped after tests or something like that.
I assume that, as the reporter tried changing drivers, that reboots were involved. That should fix any crashed service.
http://bugs.winehq.org/show_bug.cgi?id=21938
--- Comment #4 from Maxim Munnig Schmidt maxim@ipact.nl 2010-03-06 07:49:47 --- Hello mates,
Sorry if i might scare you about your unit tests, i am a windows developer myself and i know how hard it is to get weird bugs reported that you think have nothing to do with your software.
i tried the 32 bit winetest and the 64 bit one, the problem was already occuring on the 32 bit winetest.
Again this is not just one system or a freak occurance, this happened to my friends laptop aswell running the same OS.
http://bugs.winehq.org/show_bug.cgi?id=21938
Paul Vriens Paul.Vriens.Wine@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |Paul.Vriens.Wine@gmail.com
--- Comment #5 from Paul Vriens Paul.Vriens.Wine@gmail.com 2010-03-06 10:13:18 --- Were the test results sent to test.winehq.org? What is the tag you've used?
http://bugs.winehq.org/show_bug.cgi?id=21938
--- Comment #6 from Maxim Munnig Schmidt maxim@ipact.nl 2010-03-06 10:27:21 --- Yes they were, I think I put in Win7x64 or something like that.
they were submitted from IP address: 82.210.117.48
http://bugs.winehq.org/show_bug.cgi?id=21938
--- Comment #7 from Paul Vriens Paul.Vriens.Wine@gmail.com 2010-03-06 10:38:30 --- And when were the test run and do you happen to know what version of the winetest executable? I can't find the results on test.winehq.org.
Another thing you could do is attach the report file to this bugreport (compressed if possible). This file should be in the %TEMP%\wct directory and is named resXXXX.tmp. It could be that you have multiple wct directories (this happens if previous cleanups didn't succeed) so pick the last one.
Btw. if you're network dropped out during the test that obviously means the report will not be sent.
http://bugs.winehq.org/show_bug.cgi?id=21938
--- Comment #8 from Austin English austinenglish@gmail.com 2010-03-06 11:18:00 --- I'd try checking the services set to run automatically or that are disabled, as well as any relevant registry entries.
Perhaps someone that has windows 7 in a vm could snapshot it and check the registry for differences?
http://bugs.winehq.org/show_bug.cgi?id=21938
--- Comment #9 from Paul Vriens Paul.Vriens.Wine@gmail.com 2010-03-06 11:32:58 --- Maxim, is your box by any chance called LaptopMax? I found this in a run from January 22nd.
If so, does this mean this issue was not present in January but is now?
I did see some references to wireless security not working anymore after running the tests on a Dutch site (http://tweakers.net/meuktracker/22358/wine-1137.html).
http://bugs.winehq.org/show_bug.cgi?id=21938
--- Comment #10 from Maxim Munnig Schmidt maxim@ipact.nl 2010-03-06 12:34:41 --- Yes my box is called LaptopMax, Indeed I had run it months ago.
The problem started back then. I didn't care to much because I thought it was a Microsoft bug in the new windows 7 and I didn't need wireless.
Now i do need wireless though, Microsoft didn't fix the bug with any updates and rerunning the test tool had no success.
I'm sorry i can't find any wct directories in my temp folders.
The dutch reference your gave is indeed me asking if anyone else got the bug. another person replied that he experienced the same thing, i just mailed him asking if he found a way to fix it.
I checked all my services, 2 were disabled (routing and remote access, blue tooth helper)
I put them to automatic started them, even restarted the box, no fix.
I then manually started all services that had the start setting to "manual"
No fix.
Because I wanted to make sure this was a wine test error i re-run the wine tests on a other computer, the problem occurred there to.
http://bugs.winehq.org/show_bug.cgi?id=21938
--- Comment #11 from Henri Verbeet hverbeet@gmail.com 2010-03-06 12:59:57 --- I don't suppose there's any helpful information in the event log?
It would help a lot if you could narrow it down to a specific test, but I understand that may not be reasonably possible. As a guess, since it seems specific to secured networks, I'd be suspicious of any cryptography / certificate related tests, like e.g. the crypt32 ones.
http://bugs.winehq.org/show_bug.cgi?id=21938
--- Comment #12 from Maxim Munnig Schmidt maxim@ipact.nl 2010-03-06 17:10:01 --- Thanks so far for the ideas.
I exported my eventlog for January the 22nd and 23th. You can download it here: http://213.163.66.200/eventlog.zip
Just extract anywhere and open your windows event logger, go to Windows logs then hit open saved log.
I browsed through it and found nothing that catches my eye, although i am not a expert on this field.
I found something about failed authentication but I don't think it has anything to do with wireless?
I do have the correct password for the wireless ofcource ;) I tried connecting to my home network and the network at my workspace, both won't work.
Cheers!
http://bugs.winehq.org/show_bug.cgi?id=21938
Dmitry Timoshkov dmitry@codeweavers.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Severity|blocker |normal
--- Comment #13 from Dmitry Timoshkov dmitry@codeweavers.com 2010-03-07 02:46:33 --- Sounds like a Windows driver bug.
http://bugs.winehq.org/show_bug.cgi?id=21938
--- Comment #14 from Maxim Munnig Schmidt maxim@ipact.nl 2010-03-07 04:28:35 --- (In reply to comment #13)
Sounds like a Windows driver bug.
I already tried 3 different drivers for my Intel 5300 AGN card.
http://bugs.winehq.org/show_bug.cgi?id=21938
--- Comment #15 from Paul Vriens Paul.Vriens.Wine@gmail.com 2010-03-07 10:04:05 --- I've reproduced this with my Vista Business laptop (actually a partition I use solely for running winetest).
Will do some more tests (I'm using partimage for quick reverts) and see if the registry (before and after) shows some differences. Can't promise when though.
http://bugs.winehq.org/show_bug.cgi?id=21938
--- Comment #16 from Maxim Munnig Schmidt maxim@ipact.nl 2010-03-07 10:10:11 --- Wow! reproducing is the first path to success. keep us updated.
http://bugs.winehq.org/show_bug.cgi?id=21938
Nikolay Sivov bunglehead@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|UNCONFIRMED |NEW Component|-unknown |testcases Ever Confirmed|0 |1
--- Comment #17 from Nikolay Sivov bunglehead@gmail.com 2010-03-07 13:29:46 --- (In reply to comment #15)
I've reproduced this with my Vista Business laptop (actually a partition I use solely for running winetest).
This deserves to be confirmed then. Thanks, Paul.
http://bugs.winehq.org/show_bug.cgi?id=21938
--- Comment #18 from Paul Vriens Paul.Vriens.Wine@gmail.com 2010-03-08 02:26:40 --- Ran a diff on before and after registry settings. Don't know which test messes this up but the fix (to get secured wireless back) is to edit the registry.
Change the data for 'Name' at HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Cryptography\Defaults\Provider Types\Type 001
from 'Microsoft Base Cryptographic Provider v1.0' to 'Microsoft Strong Cryptographic Provider'
My first suspect is the advapi32:crypt tests btw. As we now only have to look for tests that change the above value we can use VM's (much easier testing).
http://bugs.winehq.org/show_bug.cgi?id=21938
--- Comment #19 from Paul Vriens Paul.Vriens.Wine@gmail.com 2010-03-08 02:52:54 --- The culprit is the test_set_provider_ex() test in advapi32/tests/crypt.c
Looks like the tests always writes MS_DEF_PROV_A back while it ignores the initial value.
Fixing the test won't be such an issue ,but what about fixing all those other boxes out there that have run this test once in their lifetime?
It looks like "Microsoft Strong Cryptographic Provider" is the default as of W2K.
http://bugs.winehq.org/show_bug.cgi?id=21938
--- Comment #20 from Maxim Munnig Schmidt maxim@ipact.nl 2010-03-08 03:48:48 --- I can confirm this registry change fixes the problem on my computer.
Good luck fixing all boxes that ran this test.
http://bugs.winehq.org/show_bug.cgi?id=21938
--- Comment #21 from Paul Vriens Paul.Vriens.Wine@gmail.com 2010-03-08 04:04:34 --- I have a fix (will sent it soon) for the tests.
The change is that we write back the original value after the tests are done. I don't think we should write "Microsoft Strong Cryptographic Provider" by default to the registry as someone could have changed his box deliberately.
So this means that broken boxes will remain broken. Maybe it's time to introduce a 'known errors' wiki page (maybe even referenced from test.winehq.org) with information about these kind of potential issues.
http://bugs.winehq.org/show_bug.cgi?id=21938
--- Comment #22 from Paul Vriens Paul.Vriens.Wine@gmail.com 2010-03-08 04:06:10 --- Created an attachment (id=26689) --> (http://bugs.winehq.org/attachment.cgi?id=26689) Potential fix for the tests
Patch sets the registry value back to the original one, once the tests are done.
http://bugs.winehq.org/show_bug.cgi?id=21938
--- Comment #23 from Paul Vriens Paul.Vriens.Wine@gmail.com 2010-03-08 12:29:06 --- Patch sent: http://source.winehq.org/patches/data/59075
http://bugs.winehq.org/show_bug.cgi?id=21938
--- Comment #24 from Nikolay Sivov bunglehead@gmail.com 2010-03-08 12:33:13 --- (In reply to comment #23)
Patch sent: http://source.winehq.org/patches/data/59075
Great job. I think we need an alert on wine-devel and possibly a temporary message on test.winehq.org with link to wine-devel post (need to know proper default value for each system too of course).
http://bugs.winehq.org/show_bug.cgi?id=21938
--- Comment #25 from Paul Vriens Paul.Vriens.Wine@gmail.com 2010-03-08 14:38:50 --- I'll sent out a note wine-devel, once this is in Git, or do you want it sooner?
The default as of W2K is "Microsoft Strong Cryptographic Provider" and these tests are not run on NT4 and lower.
http://bugs.winehq.org/show_bug.cgi?id=21938
--- Comment #26 from Ben Klein shacklein@gmail.com 2010-03-08 15:08:37 --- (In reply to comment #25)
I'll sent out a note wine-devel, once this is in Git, or do you want it sooner?
The default as of W2K is "Microsoft Strong Cryptographic Provider" and these tests are not run on NT4 and lower.
I'd put a message on wine-devel and wine-users now, to let people know that if they have trouble with their wireless authentication to check this value.
I apologise for my earlier catagorical dismissal of this bug. It's probably a good idea to go through all the tests and make sure that there are no Wine-set registry keys being left behind.
http://bugs.winehq.org/show_bug.cgi?id=21938
--- Comment #27 from Paul Vriens Paul.Vriens.Wine@gmail.com 2010-03-15 13:16:24 --- Fixed by commit 0b994d0aee643c59b985613802d3d4aa069f3554
http://bugs.winehq.org/show_bug.cgi?id=21938
Austin English austinenglish@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |FIXED
--- Comment #28 from Austin English austinenglish@gmail.com 2010-03-15 13:22:17 --- Fixed.
http://bugs.winehq.org/show_bug.cgi?id=21938
Alexandre Julliard julliard@winehq.org changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|RESOLVED |CLOSED
--- Comment #29 from Alexandre Julliard julliard@winehq.org 2010-03-19 14:10:57 --- Closing bugs fixed in 1.1.41.