I have a problem with netbeans, and i can't find a solution on google.
When i try to save a file, the pointer shows loading icon, and linux wait's indefinitely.
At this moment i can't even close netbeans, and have to run kill or pkill command.
Restarting netbeans or computer doesn't change that.
If i want to be able to save a file again, I have to kill netbeans, and remove the cache
(rm -rf ~/.cache/netbeans/8.0.2/*)
If I run strace -p , it gives me :
wait4(-1,
after removing cache and restarting netbeans, i can save file only if I don't change too much lines (about 10 lines are enough to crash netbeans again)
i have no idea about what is happening, i never had this problem before.
I think it could be a svn problem, or a disk size problem but I am absolutely not sure of that
This computer is my work computer, so i prefer not reinstall netbeans if you can help me to find out what's happening.
I am working on linux (centos 7) with netbeans 8.0.2
ps : this problem is on 1 file only, i can work fine with the rest of my project's files (of course that's the file I need right now :( )
ps2 :
I found some more details.
fuser ~/.netbeans/8.0.2/lock gave me a pid of a netbeans child process (his ppid is netbeans pid).
strace -p gives me :
futex(0xa280a0c, FUTEX_WAIT, , NULL
So it seems a thread can't release futex?
Does that seems possible to you?
If yes how can I find if it's a netbean's bug or a system problem?
So after a few days looking for a solution, i may have found what's happening.
It seems the problem doesn't come from netbeans but from centOS.
Apparently, CentOS 6.6 and 7.X have an error in the futex management (a default is missing in a switch).
https://groups.google.com/forum/#!topic/mechanical-sympathy/QbmpZxp6C64
I don't know if the person who posted this is good in system administration, but after i found this post, i have found some more saying the same thing.
I don't use the solved button right know because i'm waiting for the administrator to look at this, but I post what I found if it can help someone else.
Related
The Ubuntu 16.04 freeze by itself especially when 2-3 instance of Pycharm is actively running.I wanted to know the exact cause of the OS freeze and found linux-crashdump.
The linux-crashdump was installed using the link [kernel-crash-dump][1]
[1]: https://help.ubuntu.com/lts/serverguide/kernel-crash-dump.html "kernel-crash-dump"
It states that after installation and rebooting the system
cat /proc/cmdline
BOOT_IMAGE=/vmlinuz-3.2.0-17-server root=/dev/mapper/PreciseS-root ro
crashkernel=384M-2G:64M,2G-:128M
But cat /proc/cmdline is not showing the parameter crashkernel added to it.
I am not able to edit the /proc/cmdline in anyway.
If I add crashkernel=384M-2G:64M,2G-:128M then the error appears.
"/proc/cmdline" E667: Fsync failed
WARNING: Original file may be lost or damaged
don't quit the editor until the file is successfully written!
I am not able to move further, as per the documentation of kernel-crash-dump after the installation, I rebooted the system, but still the crashkernel parameter in the /proc/cmdline is not appearing.
I didn't get completely what the problem was but as far as I understood you had problems with crashkernel. I think that it would help if you just remove it completely and reinstall it again, because you have lost some of the files of crashkernel and that's why it's not responding. Hope it helps :)
I'm running mint 17.2 kde.
I installed intellij CE by running idea.sh in ~/Downloads/
Then I moved intellij to /usr/lib/. Now I get this when I try to open intellij from the Application Launcher:
KDEInit could not launch '/home/******/Downloads/idea-IC-143.1821.5/bin/idea.sh'
I know I could just move it back, but I would really prefer to point Application Launcher in the right direction.
I've already checked out .kde/share/config and didn't find any mention of idea.sh, jetbrains, or intellij that seemed like a relevant entry.
Is there somewhere else that Application Launcher entries are saved? Is it possible to remove entries then add new ones from the application itself?
I ended up grepping through my entire home directory for files that used the full path posted in the error message.
grep -rnw '/home/******/' -e ".*/home/******/Downloads/idea-IC-143.1821.5/bin/idea.sh.*"
This returned jetbrains-idea-ce.desktop in my .gnome, .kde, and .local directories, all of which I modified. I could launch intellij from the Application Launcher after this.
I am from Mac background trying my hand at installing perforce client visual(P4V) on my linux box.For this I download the correct version here and untar the files.
Then I cd to the directory
~/Desktop/p4v-2012-blah-blah/bin
I also say
chmod +x p4*
After this i try running p4v (by double clicking) but I dont see anything .The file type is shown as a "text executable" but i dont know why it is not running.
On mac i had done the same thing -just clicked on p4v and the client would show up(where I filled the server address and everything )But not sure what is going wrong here.Can someone give me directions?
FWIW i did check out this link
and I do see the desktop (so GNOME is probably installed ?) .Hence p4v should be able to run.
I know we don't like link-based answers, but there is a rather complete walk-through here: http://www.perforce.com/documentation/tenminute-test-drive-linuxunix
Probably should be on Super User though.
You may have a 64bit Linux and tried to install a x86 P4V or vice versa, which is why the command doesn't run. Check that the OS and app types match.
When I tried to un-install and install an executable foo.exe, Im geting the below error
Error opening file for writing
click abort to stop the installation
I could find that some files were failed to get deleted while un-installation.
And by using process explorer I can see that explorer.exe has still reference to some dlls of foo.exe even after unregistering them.
and killing the explorer.exe and re-start it again while un-installation solves the issue. (thru NSIS installation script).
or by setting rebook ok flag in script and ask user to re-start his machine before installing foo.exe again.
I don't think its a good solution to ask user to reboot his machine after un-installation .
Can anyone help me on how can I make sure that my installation goes fine instead of restarting explorer.exe
For registered shell extensions there is really no clean way to delete the dll's other than forcing a reboot (Or log-off if they are registered in HKCU\Software\Classes)
If that is not good enough you basically have two options:
Rename the .dll and delete it later
<HACKALERT>Inject a thread that calls CoFreeUnusedLibraries() into every process that has your dll loaded</HACKALERT>
Just killing and restarting explorer is not really an option since you don't know if some other program is also using your shell extension.
All of the sudden svn stopped working in cygwin installation on windows xp. when I execute svn binary, nothing happens, svn process does not even show up in the Task Manager. I've reinstalled svn but it did not help (the last resort would be to uninstall cygwin itself). Everything else in cygwin works fine: awk,python,sed,more,less,tail and etc.
here is what is happening ...
mt#s022 ~
$ which svn
/usr/bin/svn
mt#s022 ~
$ svn --version
mt#s022 ~
$ svn status
mt#s022 ~
$ svn info
mt#s022 ~
$
Same problem as well.
No solution, but here's an interesting thing: For me, "curl" and "wget" are also broken in exactly the same way. This makes me wonder if it is one of the underlying networking libraries (libcurl, or similar) that is at the root of the problem (total guess).
EDIT: I have solved this problem. The problem turned out to be what seems to have been a "bad" version of openssl. I used the Cygwin setup program to "Reinstall" OpenSSL. The version number in setup is now "0.9.8n" which seems to have replaced "0.9.8", which I suspect was a bad version that was temporarily distributed, so if you were unlucky enough to update at the wrong time then this is what you got. The new version doesn't seem to be installed automatically just by updating normally. I'm guessing 0.9.8 is seen as the same version as 0.9.8n.
In case this doesn't solve it for others: the key insight was gained by launching svn.exe from Windows Explorer. While it exits silently when run from the bash command line, when run from Windows Explorer it pops up a window complaining: "the procedure entry point pqueue_size could not be located in the dynamic link library cygcrypto-0.9.8.dll". This allowed me to identify OpenSSL as the culprit. The same might be possible for other problems that manifest as commands silently exiting when run from the bash terminal.
This is what I did:
net stop sshd (this is optional)
remove /usr/bin/cygwin1.dll
remove /usr/bin/cygcrypto-0.9.8.dll
run cygwin-setup
reinstall cygwin-base
Hope this helps.
Ran into this problem. It seemed to have started when I added one small package and left everything else as keep. Solution that worked: remove cygwin1.dll, run cygwin's setup.exe, keep everything except base->cygwin. Updating that updated cygwin1.dll, and all was fine again.
I had same issue.
I followed the suggestion above and tried to launch svn from the windows explorer and it turned out that I had a problem with cygwin1.dll itself.
Doing a reinstall of the cygwin package solved my problem
I had the same problem and upgrading/reinstalling the cygwin base classes didn't help me neither...
Until I disabled the "CYGWIN sshd" service.
Therefore, if you run the Cygwin sshd process, stop it first, before upgrading Cygwin.
Good luck!