I am writing a program in C which is using a few processes, semaphores and mapped memory. When I map the memory and then the program fails in the middle of progress so it cant get to the stage when memory is released, the program is stucked and I have to kill it (Ctrl+C).
The problem is that when I fix the bug and run the program again - it calls the shared memory error or semaphores error and program is terminated. I can fix this problem only by restarting the whole OS.
Is there another way, how to "deallocate" allocated memory after unexpected error?
FYI: ipcs doesn't show this allocated memory nor the semaphores used.
EDIT: I had to tag just one "right" answer, but I would like to thank you all for the ideas. The result is that after the problem occurs, deleting everything except pulse... in /dev/shm folder is the solution.
POSIX shared memory doesn't have a specific command-line tool. But it is typically mapped into the /dev/shm tree where you can manage the segments with classic file manipulation tools.
Grapsus's comment is correct for POSIX shared memory. Forget the ipcrm, it's SysV only. Dig around in /dev/shm and remove the file that "represents" your shared memory slab.
You probably should also put in a signal handler to remove the shared memory upon a kill. It won't work if you are going to kill with SIG_KILL (9), but it will work with most of the lesser kills. Once the signal handler is in place, regular kills will call the handler which can then programmatically remove the shared memory before process shutdown.
Related
I am writing a c terminal program that runs until the user terminates it with Ctrl+C. Think something like ping or top.
My program allocates to the heap but starts no other threads or processes. Should I be handling SIGINT and freeing any allocated memory before exit or is leaving it to the OS better practice?
The short answer is yes given your context, which is a normal exit situation. In an abnormal exit situation, then the short answer is absolutely no.
If you are concerned that your program is leaking memory during its execution, which is a bad thing in the sense that it slows your program execution, then you can keep track of the memory that you allocate and then free it before you exit. Then you can run your program with valgrind and if valgrind complains about blocks that weren't free'd, then you will know you have some type of leak. The location of the allocation will help you know if the leak is of any importance.
If you exit anyway, you don't need to release any resources. The OS will take care of it just fine, and there is no benefit in doing it manually.
Note that free() is not async-safe, so you would definitely have to do the actual freeing in the main thread, and not in the handler. But don't do that, unless you want to do other things than exit().
Use SIGINT handlers for things like resetting the terminal (e.g. with ncurses), or saving critical state.
In my program written with C and C++, I will new an object to fulfill the task, then delete the object.
At the moment after new object but before delete object, if the user presses ctrl+c to break the process, that will cause delete not to be called and a memory leak occurs.
What should I do to avoid this situation?
Also, if the memory was reclaimed by the OS, what about the opened files? Are they closed by the OS or should I close them manualy?
In a virtual-memory-based system, all memory is returned to the OS when a process is terminated, regardless of whether it was freed explicitly in the application code. The same might not be true of other resources, however, which you may want to free cleanly. In which case, you need to provide a custom signal handler for the SIGINT signal (which is received on Ctrl+C), see e.g. http://linux.die.net/man/2/sigaction.
Pressing CtrlC will send a SIGINT to the process, which by default does a mostly-orderly shutdown, including tearing down the memory manager and releasing all allocated heap and stack. If you need to perform other tasks then you will need to install a SIGINT handler and perform those tasks yourself.
If you allocated any SYSV Shared Memory Segments using shmget(2) then you must clean up after yourself with shmctl(2).
If you allocated any POSIX Shared Memory Segments using shm_open(3) then you must clean up after yourself with shm_unlink(3).
Both SYSV and POSIX shared memory segments persist past process termination. You can see what persists using the ipcs(1) tool.
Of course, if you haven't used any SYSV or POSIX shared memory segments, then this is all just noise. :)
You are subscribing to a rather common misconception that heap blocks that are not freed, but still accessible at the time a program exists are leaks. This is not true. Leaked blocks are those which no pointer still references, hence they can't be freed.
Through the years of playing with (and breaking) lots of perfectly good kernels, I have never managed to sufficiently break a virtual memory manager to the point where it no longer reclaimed the entire address space of a process once it exited. Unless you are working with a kernel clearly marked as 'new and experimental', you will have better luck winning the lottery than finding a system that doesn't employ an effective virtual memory manager.
Don't put cruft in your code just to get a perfect score in Valgrind. If you have no real clean up tasks to do other than freeing memory that still has valid references, you don't need to bother. If someone throws a kill -9 to your program, you won't be able to handle it and will see the old behavior repeat.
If you have file descriptors to clean up, shared locks to relinquish, streams to flush or whatever else must happen so other processes don't miss you when you're gone, by all means take care of that. Just don't go adding code that does nothing to solve a non-problem, it just seems silly to do so.
Note
This was originally going to be a comment, but is far too long and SO frowns on writing a novel one comment at a time.
When CTRL+C is pressed in a Linux console, the SIGINT signal is sent to the application which, if the signal has no handler, will terminate the program, returning all memory to the OS. This of course would make it pointless to do any freeing of memory, since all memory will freed once the program exists. However, if you would like to handle the CTRL+C SIGINT signal (maybe to write out some last data to a file or do some other cleanup), you can use the function signal() to install a function to be called when the signal is received. Check out the man page for this function if you want to learn more.
If the process quits, a memory leak will NOT normally occur.
Most of the memory you allocate will be freed on Ctrl+C. If you see memory usage not return to its prior level, it is almost certainly caused by buffered filesystem blocks.
However, you should definitely clean things up, in particular if you have used any other types of resources:
Files created in temporary directories won't be deleted. This includes /dev/shm, leaving such a file could be considered a "memory leak".
System V or posix shared memory segments won't get thrown away when your process quits. If this bothers you, clean them up specifically. Alternatively, clean them up on a subsequent run.
Normally a leak (of a persistent or semi-persistent object e.g. file) doesn't matter if a subsequent run doesn't leak more memory. So cleaning up on a future run is good enough.
Imagine a process running every 5 minutes from "cron", if it crashes on each run and leaves some mess, it's still ok provided each run cleans up the mess from the previous crash.
The OS will reclaim the memory allocated by the process when the process exits as a result of Ctrl-C or any other means.
Suppose a program has memory leaks.
1) When a process dies (normally or segmentation fault), are those leaked memory freed?
2) What about other resources a process holds?
With most modern operating systems (linux, windows from around NT 3.5), yes.
Stack and heap memory is freed and file descriptors are closed on all modern system, I think.
On POSIX systems there are a number of resources that are not freed when a process exits, shared semaphores, message queues and memory segments. These are meant to be persistent between processes, so they simply can't. It is the responsibility of the application to free them.
It could do that e.g with on_exit handlers, but usually there is a simpler way. For memory segments you would typically use shm_unlink after all processes have opened such a segment. The segment then ceases to exist when the last process (and its file descriptor to the segment) is closed.
1) Yes, the memory is freed.
2) Different process model? I don't know what you mean by that, but once a program dies, all the memory that it malloc'd or new'd is then returned to the OS and will be reallocated to another program later.
3) Once a program exits, all allocated memory is returned to the OS, however until the process is wait()ed by another process, there is a small amount of data such as the exit status waiting around for someone to collect it. On linux, I believe, a normal process from bash/init will be waited and cleaned up automatically.
You can safely assume with modern linux systems that the memory will be freed... However... Its not a guaruntee, and certainly not best practice.
I have a number of data structures (trees, queues, lists), created using dynamic allocation routines (malloc, calloc). Under some critical conditions, the program should terminate. Traversing all objects to free their memory takes long time.
Is it safe to avoid traversing all data structures just before program stops? If yes, does it apply to all operating systems and environments (e.g. multiple threads)?
All the memory dynamically allocated by a process is released back to the OS on process termination, doesn't matter intentionally or via a crash. Same happens with files and sockets - ref counts inside the kernel get decremented and resources get released if there are no more references.
An exception to the above might be the shared memory.
When a program (i.e., a process) terminates, all local and heap memories are automatically reclaimed. Note that these memory regions are specific to a process. So, you may skip the traverse and deallocation just before the program termination. However, if the program uses a shared/global memory, then you need to explicitly reclaim that. Finally, it applies, at least, Linux/Unix and Windows. I believe it applies to all modern operating systems.
Short answer: yes. In any modern OS, memory is private to each process, and once the process exits, all memory is reclaimed by the OS (unless the OS itself is broken).
You don't have to free() all your dynamically-allocated memory before terminating the program. The operating system releases all the memory that was owned by the process anyway. It also closes any network connections that you had open.
In my program written with C and C++, I will new an object to fulfill the task, then delete the object.
At the moment after new object but before delete object, if the user presses ctrl+c to break the process, that will cause delete not to be called and a memory leak occurs.
What should I do to avoid this situation?
Also, if the memory was reclaimed by the OS, what about the opened files? Are they closed by the OS or should I close them manualy?
In a virtual-memory-based system, all memory is returned to the OS when a process is terminated, regardless of whether it was freed explicitly in the application code. The same might not be true of other resources, however, which you may want to free cleanly. In which case, you need to provide a custom signal handler for the SIGINT signal (which is received on Ctrl+C), see e.g. http://linux.die.net/man/2/sigaction.
Pressing CtrlC will send a SIGINT to the process, which by default does a mostly-orderly shutdown, including tearing down the memory manager and releasing all allocated heap and stack. If you need to perform other tasks then you will need to install a SIGINT handler and perform those tasks yourself.
If you allocated any SYSV Shared Memory Segments using shmget(2) then you must clean up after yourself with shmctl(2).
If you allocated any POSIX Shared Memory Segments using shm_open(3) then you must clean up after yourself with shm_unlink(3).
Both SYSV and POSIX shared memory segments persist past process termination. You can see what persists using the ipcs(1) tool.
Of course, if you haven't used any SYSV or POSIX shared memory segments, then this is all just noise. :)
You are subscribing to a rather common misconception that heap blocks that are not freed, but still accessible at the time a program exists are leaks. This is not true. Leaked blocks are those which no pointer still references, hence they can't be freed.
Through the years of playing with (and breaking) lots of perfectly good kernels, I have never managed to sufficiently break a virtual memory manager to the point where it no longer reclaimed the entire address space of a process once it exited. Unless you are working with a kernel clearly marked as 'new and experimental', you will have better luck winning the lottery than finding a system that doesn't employ an effective virtual memory manager.
Don't put cruft in your code just to get a perfect score in Valgrind. If you have no real clean up tasks to do other than freeing memory that still has valid references, you don't need to bother. If someone throws a kill -9 to your program, you won't be able to handle it and will see the old behavior repeat.
If you have file descriptors to clean up, shared locks to relinquish, streams to flush or whatever else must happen so other processes don't miss you when you're gone, by all means take care of that. Just don't go adding code that does nothing to solve a non-problem, it just seems silly to do so.
Note
This was originally going to be a comment, but is far too long and SO frowns on writing a novel one comment at a time.
When CTRL+C is pressed in a Linux console, the SIGINT signal is sent to the application which, if the signal has no handler, will terminate the program, returning all memory to the OS. This of course would make it pointless to do any freeing of memory, since all memory will freed once the program exists. However, if you would like to handle the CTRL+C SIGINT signal (maybe to write out some last data to a file or do some other cleanup), you can use the function signal() to install a function to be called when the signal is received. Check out the man page for this function if you want to learn more.
If the process quits, a memory leak will NOT normally occur.
Most of the memory you allocate will be freed on Ctrl+C. If you see memory usage not return to its prior level, it is almost certainly caused by buffered filesystem blocks.
However, you should definitely clean things up, in particular if you have used any other types of resources:
Files created in temporary directories won't be deleted. This includes /dev/shm, leaving such a file could be considered a "memory leak".
System V or posix shared memory segments won't get thrown away when your process quits. If this bothers you, clean them up specifically. Alternatively, clean them up on a subsequent run.
Normally a leak (of a persistent or semi-persistent object e.g. file) doesn't matter if a subsequent run doesn't leak more memory. So cleaning up on a future run is good enough.
Imagine a process running every 5 minutes from "cron", if it crashes on each run and leaves some mess, it's still ok provided each run cleans up the mess from the previous crash.
The OS will reclaim the memory allocated by the process when the process exits as a result of Ctrl-C or any other means.