I am creating a child process which executes a batch file
success = CodebenderccAPI::CreateProcess(
NULL,
(LPWSTR)command.c_str(), // command line
NULL, // process security attributes
NULL, // primary thread security attributes
TRUE, // Inherit pipe handles from parent process
CREATE_NEW_CONSOLE, // creation flags
NULL, // use parent's environment
current_dir, // use the plugin's directory
&si, // __in, STARTUPINFO pointer
&pi); // __out, receives PROCESS_INFORMATION
The above code was tested and works as it should in various machines but recently failed in a Windows 7 machine and I am still trying to figure out why...
I have ensured that the user has administrative permissions to the folder where the batch file is installed as well as that the batch file is in the working directory. Moreover I disabled the antivirus (to verify that it didn't cause the problem) and tried again with no sucess. CreateProcess always fails with error code 2: ERROR_FILE_NOT_FOUND.
Any ideas what could cause that failure?
Finally I managed to find out what was causing CreateProcess failure.
Opened cmd and cd to the folder where batch file was located, then ran the batch file without any problem. After this, I navigated to the folder through file system and double clicked the batch file to run it. It failed with error message "Windows cannot find the_path_to the batch_file. Make sure you have typed the name correctly and try again."
According to this post the above error is related with the COMSPEC entry. Checked its value in registry and found that it was different from the default value. Updated its value to the default and problem was solved!
Related
I have a bunch of unit tests which contain code like:
File file = new File("src/main/java/com/pany/Foo.java");
assertTrue("Missing file: " + file.getAbsolutePath(), file.exists());
This test is suddenly failing when running it with Maven Surefire and -DforkCount=0. With -DforkCount=1, it works.
Things I tried so far:
The file does exist. Windows Explorer, command line (copy & paste), text editors, Cygwin can all find it and show the contents. That's why I think it's not a permission problem.
It's not modified by the unit tests or anything else. Git shows no modifications for the last two months.
I've checked the file system, it's clean.
I've tried other versions of Java 8, namely 8u171 and 8u181. Same problem.
I've run Maven from within Cygwin and the command prompt. Same result.
Reboot :-) No effect :-(
More details:
When I see this problem, I start to see the "The forked VM terminated without properly saying goodbye. VM crash or System.exit called?" in other projects. That's why I tried forkCount=0 which often helps in this case to find out why the forked VM crashed.
This has started recently, maybe around the October 2018 update of Windows 10. Before that, the builds were rock solid for about three years. My machine was switched to Windows 10 late 2017, I think.
I'm using Maven 3.6 and can't easily try an older version because of an important bug that was fixed with it. I did see the VM crash above with Maven 3.5.2 as well.
It's always the same files which fail (so it's stable).
ulimit (from Cygwin) says:
$ ulimit -a
core file size (blocks, -c) unlimited
data seg size (kbytes, -d) unlimited
file size (blocks, -f) unlimited
open files (-n) 256
pipe size (512 bytes, -p) 8
stack size (kbytes, -s) 2032
cpu time (seconds, -t) unlimited
max user processes (-u) 256
virtual memory (kbytes, -v) unlimited
I'm wondering if the "open files" limit of 256 only applied to Cygwin processes or whether that's something which Cygwin reads from Windows.
Let me know if you need anything else. I'm running out of ideas what I could try.
Update 1
Bernhard asked me to print absolute names. My answer was that I was already using absolute names but I was wrong. The actual code was:
File file = new File("src/main/java/com/pany/Foo.java");
if (!file.exists()) {
log.debug("Missing file {}", file.getAbsolutePath());
... fail ...
}
... do something with file...
I have now changed this to:
File file = new File("src/main/java/com/pany/Foo.java").getAbsoluteFile();
if (!file.exists()) {
log.debug("Missing file {}", file);
}
and that fixed the problem. I just can't understand why.
When Maven creates a forked VM to run the tests with Surefire, then it can change the current directory. So in this case, it would make sense that the tests work when forked but fail when running in the same VM (since the VM was created in the root folder of the multi-module build). But why is making the path absolute before the call to exists() fixing the issue?
Some background. Each process has a notion of "current directory". When started from the command line, then it's the directory in which the command was executed. When started from the UI, it's usually the folder in which the program (the .exe file) is.
In the command prompt or BASH, you can change this folder with cd for the process which runs the command prompt.
When Maven builds a multi-module project, it has to change this for each module (so that the relative path src/main/java/ always points to the right place). Unfortunately, Java doesn't have a "set current directory" method anywhere. You can only specify one when creating a new process and you can modify the system property user.dir.
That's why new File("a").exists() and new File("a").getAbsoluteFile().exists() work differently.
The latter will use new File(System.getProperty("user.dir"), "a") to determine the path and the former will use the Windows API function _wgetdcwd (docs) which in turn uses a field of the Windows process to get the current directory - in our case, that's always the folder in which Maven was originally started because Java doesn't update the field in the process when someone changes user.dir and Maven can only change this property to "simulate" changing folders.
WinNTFileSystem_md.c calls fileToNTPath(). That's defined in io_util_md.c and calls pathToNTPath(). For relative paths, it will call currentDirLength() which calls currentDir() which calls _wgetdcwd().
See also:
https://github.com/openjdk-mirror/jdk7u-jdk/blob/jdk7u6-b08/src/windows/native/java/io/WinNTFileSystem_md.c
https://github.com/openjdk-mirror/jdk7u-jdk/blob/jdk7u6-b08/src/windows/native/java/io/io_util_md.c
and here is the place where the Surefire plugin modifies the Property user.dir: https://github.com/apache/maven-surefire/blob/56d41b4c903b6c134c5e1a2891f9f08be7e5039f/maven-surefire-common/src/main/java/org/apache/maven/plugin/surefire/AbstractSurefireMojo.java#L1060
When not forking, it's copied into the current VM's System properties: https://github.com/apache/maven-surefire/blob/56d41b4c903b6c134c5e1a2891f9f08be7e5039f/maven-surefire-common/src/main/java/org/apache/maven/plugin/surefire/AbstractSurefireMojo.java#L1133
So I have checked via printing out the system properties with some simple tests.
During the tests via maven-surefire-plugin the user.dir will be changed to the root of the appropriate module in a multi module build.
But as I mentioned already there is a system property available basedir which can be used to correctly handle the location for tests which needs to access them via File...The basedir is pointed to the location of the pom.xml of the appropriate module.
But unfortunately the basedir property is not set by IDEA IntelliJ during the test run.
But this can be solved by a setup like this:
private String basedir;
#Before
public void before() {
this.basedir = System.getProperty("basedir", System.getProperty("user.dir", "Need to think about the default value"));
}
#Test
public void testX() {
File file = new File(this.basedir, "src/main/java/com/pany/Foo.java");
assertTrue("Missing file: " + file.getAbsolutePath(), file.exists());
}
This will work in Maven Surefire with -DforkCount=0 as well as -DforkCount=1 and in IDE's (checked only IDEA IntelliJ).
And yes that is an issue in Maven Surefire plugin changing the user.dir.
We might convince IDE's to support the basedir property as well ?
Aaron, we develop the Surefire. We can help you if you provide the path for this:
assertTrue("Missing file: " + file.getAbsolutePath(), file.exists());
Pls post the actual path, expected path and basedir where your POM resides.
The theory would not help here. We are testing all the spectrum of JDKs 7-12 but we do not have the combination Cygwin+Windows which must be considered.
The code setting user.dir in Surefire you mentioned exists a decade.
I have a Windows service (running on Windows Vista+), which needs to launch a normal GUI application. For some reason, I need to run the GUI application on non-default desktop (of course, since it's launched by the service process, it runs in Session 0, WinSta0, but not on the Default desktop).
The code looks like this.
// create new desktop
hDesktop = CreateDesktop(NEW_DESKTOP, 0, 0 ,0,
DESKTOP_SWITCHDESKTOP | DESKTOP_WRITEOBJECTS |
DESKTOP_READOBJECTS | DESKTOP_ENUMERATE |
DESKTOP_CREATEWINDOW | DESKTOP_CREATEMENU|
DESKTOP_HOOKCONTROL, &sa);
// create process of the normal GUI application,
// running on the new desktop, not the default one
STARTUPINFO si;
ZeroMemory(&si, sizeof(si));
si.cb = sizeof(si);
si.lpDesktop = NEW_DESKTOP;
ZeroMemory(&processInfo,sizeof(processInfo));
BOOL bRet = CreateProcess(NULL, &commandLine, NULL, NULL,
FALSE, 0, NULL, NULL, &si, &processInfo);
TCreateProcess(...) method returns successfully, but the GUI application exits immediately after launching. With help from Gflags, I got following information.
LdrpInitializeRoutines - Error: Init routine 7595D711 for DLL "C:\Windows\system32\USER32.dll" failed during DLL_PROCESS_ATTACH
_LdrpInitialize - ERROR: Process initialization failed with status 0xc0000142
LdrpInitializationFailure - ERROR: Process initialization failed with status 0xc0000142
I searched for a while, seems it's related to security issue. I tried to grant all desktop related rights to current user, but it didn't help.
One thing may help. I noticed that there are a few DLLs were not loaded when running with SYSTEM account, the first one is uxTheme.dll.
Anyone has any idea why it doesn't work with non-default desktop, while working well with default desktop?
Thanks.
You need to use CrateDesktopEx and increase the size of heap -- the default is not sufficient even for the notepad.
We use Inno Setup(version 5.4.2) as the packaging tool to generate our installer. While upgrading our software from older version to current version,
we try to overwrite existing binaries/drivers. This is often leading to issues as some monitoring software like 'HP ArcSight Logger/Connector', 'SplunkUniversalForwarder' etc. are holding file handles on our binaries and casuing overwrite to fail. Our installer shows beolow popup message on encountering this issue each time.
"C:\Windows\System32\drivers\xxx.sys
An error occurred while trying to replace the existing file:
DeleteFile failed; code 5.
Access is denied.
Click Retry to try again, Ignore to skip this file(not recommended), or Abort to cancel installation"
Interestingly, even after above getting above pop-up, we were able to rename xxx.sys to xxx.sys.old manully. We used to suggest cusotmers to rename
xxx.sys to xxx.sys.old and 'Retry' the installation. After renaming, upgrade used to complete without any issues.
Questions
1) Is it possible to rename xxx.sys through program always, when we hit this issue.
2) Any process to reproduce the DeleteFile failed; code 5 issue?
Using the installer you can rename the existing file and install the new file using the correct file name with help from the windows registry.Remember to ask the user to reboot their computer to complete the install.Another option would be to either use the installer to rename the file and again using the window's registry to delete the unneeded file.
Window's registry allows you to delete or rename files on reboot as part of an install.
Reference to use of PendingFileRenameOperations:
http://support.microsoft.com/kb/181345
Example to rename from microsoft support:
The syntax used is (without quotes):
"\??\source file !\??\target file"
HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Control\Session Manager
Value name: PendingFileRenameOperations Data type : REG_MULTI_SZ
Value data: \??\c:\temp\win32k.sys !\??\c:\winnt\system32\win32k.s
The same command can be used to delete a file as well although requires the use of nulls.
http://www.pcreview.co.uk/forums/pendingfilerenameoperations-delete-file-t1715654.html
Same example used to delete a file:
HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Control\Session Manager
Value name: PendingFileRenameOperations Data type : REG_MULTI_SZ
Value data: \??\c:\temp\win32k.sys\0\0\0
I hope this helps you!
I am trying to launch an external application from within my Win32 application but it's failing. Following is the code I am using:
HINSTANCE instance = ShellExecute(NULL, _T("open"), _T("loader.exe"), NULL, NULL, SW_SHOWNORMAL);
if((int)instance <= 32)
{
_cprintf("Error = 0x%X\n", GetLastError());
return 0;
}
The instance value I get is 0x00000002 and GetLastError returns 0x2. The same code works when I try to launch other applications like iTunes.exe or cmd.exe. Does it has anything to do with external application? By the way, win32 application and loader.exe application are located in the same folder.
Any help would be highly appreciated.
Farooq-
Well, error 0x2 is ERROR_FILE_NOT_FOUND
Looks like it can't find "loader.exe"
Error 2 is "File not found":
http://msdn.microsoft.com/en-us/library/ms681382(v=vs.85).aspx
I'm guessing it can't find loader.exe.
Put loader.exe somewhere in the search path, or provide the full path. That is how to avoid this file not found error. Windows error codes are all documented on MSDN.
I am currently trying to debug my Crash-Handler, it is an external executable that gets called when my program finds an unhandled structured exception. I recompiled the crash-handler executable, and when I tested, I got a message: "Program too big to fit in memory".
After digging around for a bit, I found that this was being caused by the CreateProcess call within my unhandled exception filter. I found this post that says that this error message indicates that the executable is corrupted, however if I invoke the CrashHandler.exe from the command line, I get no such error.
Other Information:
I have tried rebuilding my
application and the crash-handler
multiple times in both debug and
release mode.
I have tried giving the running thread 2 orders of magnitude more stack space.
I tested the same CrashHandler.exe in another application that was already using it, and there were no problems.
I tried using a previous version of the exe that worked before, but with the same result.
Restarting the system.
My Call to CreateProcess:
//Fire off the Handler
//HandlerArgs = "CrashHandler.exe arg1 arg2 arg3 etc..."
if(CreateProcess(NULL, HandlerArgs, NULL, NULL, TRUE, CREATE_PRESERVE_CODE_AUTHZ_LEVEL | CREATE_SUSPENDED,
NULL, NULL, &StartupInfo, &NewProcessHandle))
Turns out that one of my post-build hooks was copying over the exe from The source control repository, and the file I had in the source control repository was actually the pdb. While testing I was copying directly to my running folder, and then the hook would copy the "corrupted" exe over again.