Programming :: Thread Stack Using Pthread_attr_setstackaddr()?
Sep 22, 2010
When I set the stack base address of the child thread using the POSIX library function "pthread_attr_setstackaddr()", I am unable to access the memory contents of its parent. The data-structures that are created on the HEAP of its parent using malloc() are either getting destroyed or unaccessible when moving to the context of the child thread. These data-structures are being passed as an argument to the child thread.Even if I make these variables global then also it is not working.pthread_attr_setstacksize(tattr, ...);stackbase = (void *) malloc(...);pthread_attr_setstackaddr(tattr, stackbase);But when I create the child thread without setting its stack base address using that pthread_attr_setstackaddr(), then it is able to access the parent's memory contents.
I seem to only be able to set my stack size on my linux server to 15000. If I increase it to 20000 I get a Segmentation Fault. how I can get the linux OS to increase the stack size? Code: threadRet |= pthread_attr_setstacksize( &m_ThreadAttributes, 15000 );
Why the thread stack size can not be changed after calling pthread_attr_setstacksize & pthread_create in a dynamic library? Detail: I write a file thread_factory.c and plan to build it and produce a dynamic library (libthread_factory.so) In the thread_factory.c , there is a routine
[Code]....
And after this, there is application, it will call fct_thread_create(STACK_SIZE_256KB), and then call pthread_attr_getstacksize(), but the stack size return always be a fixed value 0xa01000. (I tried this on Fedora12) But if I build the application source code with the file thread_factory.c directly, the stack size return is right as my expect. I checked the source code of glibc about the routine pthread_create() as below:
i have an application that launches several pthreads, i know that the default size used by Linux is 8Mb for each pthread. However i would like to optimize the total memory usage by my application by decreasing the default stack size of each pthread to the needed resources. My questions:
- Are there any rules to set the pthread stack size. - How to compute the memory needed by each thread. - Is the malloc call inside a thread counted from the stack size of the same pthread?
Desperate to reduce RAM usage of my tiny VPS running Ubuntu 9.04 and Apache2.2.11, here I saw that: On Linux, each child process will use 8MB of memory by default. This is probably unnecessary. You can decrease the overall memory used by Apache by setting ThreadStackSize used by Apache by setting ThreadStackSize to 1MB in.
So I tried to give the suggestion a try. But when I append: ThreadStackSize 1000000 in my /etc/apache2/httpd.conf <IfModule mpm_prefork_module> directive, and restarted apache, it failed with this message: Invalid command 'ThreadStackSize', perhaps misspelled or defined by a module not included in the server configuration
So I figured out that the relevant modules are neither enabled nor available on apache2. Now I am wondering whether there is a way to decrease the ThreadStackSize without the need to compile apache from source? If not, what should I do?
I am going to use "pthread_setaffinity_np" to bind a thread to a specific core. My application has two threads. I have used mutex to assign a specific id to each thread and then bind that thread to a core different from another core. but it seems that the os assigns both thread to one core.What should I do to bind each thread to a specific core?
A simple TCP based chat server could allow users to use any TCP client (telnet, for example) to communicate with each other. For this question you should consider a single process, single thread server that can support exactly 2 clients at once, the server simply forwards whatever is sent from one client to the other (in both directions). Your server must not insist on any specific ordering of messages as soon as something is sent from one client it is immediately forwarded to the other client. As soon as either client terminates the connection the server can exit
If I issue the following and the process doesn't quit, is there a way to view what it's doing with the SIGTERM signal on the stack? Is this done via the pstack command?kill -s SIGTERM <PID>
Why is it in Linux that there is a stack size set by default? And why is it so small? (My system is set to 8192 kbytes.) And why is there a default limit on the stack size when the max memory and virtual memory size are, by default, unlimited? (Aren't they both fed from the same place ultimately?)
Reason I ask: I want to use recursive functions in my programming a lot more. Problem is, if the language (or implementation) doesn't happen to support tail-call recursion, then I can be pretty well certain that the first huge problem that gets thrown at my function is going to kill my program because the stack size limit is going to be quickly reached. Obviously, I can change the stack size limit for my own computers, but it doesn't feel so great knowing that most of the people who copy and execute my code will have probably have overlooked this. Anyway, does anyone know: is this small default stack size limit just one of those historical artifacts, or is there some technical reason for it?
i have just modify tcp.c file in/usr/src/linux/net/ ipv4 location.Now should i compile the complete kernel?if not then how to compile that net/ipv4 package or etc.
What is the quickest way to get the thread id(tid) of the current thread in linux? Specifically I use fedora 13. ...and I am talking about inside a C++ program.
Is there a library/system call that will return to me what CPU core a thread of execution is running on? I've looked for a bit on the net already and also in /usr/include and couldn't find one.
getcpu() and sched_getcpu() are two that I found, but when I include the appropriate header files (linux/getcpu.h, and sys/sched.h respectively), gcc says getcpu.h doesn't exist and the linker complains it cannot find the implementation in sched.h.
I'm sure I'm doing something really stupid or overthinking...
What I am doing is running an OpenMP application and specifying a list of cpu cores to run on with GOMP_CPU_AFFINITY. I want to make sure that each core is getting the same number of threads.
**UPDATE**
It seems that this code works, but only with glibc >= 2.6, and my machine has 2.5:
I have a program which is uses sigaction to register for a SIGIO signal (for incoming data on a fd) with an appropriate event handler. I also create a new detached thread 'B' that does some work with the received data. Normally the thread B runs properly. But when my event handler is called (because a there is new incoming data), after the event is handled, the thread B is not called immediately. There is a noticeable delay of the order of many seconds before it is scheduled again .During this delay, my program is doing nothing.
What am i doing wrong? Is there someway i can run thread B as soon as the event is handled (and assuming no other work is to be carried out)?
I have a process and two threads running in it. Thread1 is my GUI-Thread (using FLTK) and Thread2 is running at the first 10 sec. after start.Now, I would like to make Thread2 "nicer" to Thread1, because otherwise GUI stucks sometimes. But I didn't find a solution.Is it possible to define the "niceness" of threads inside a process (to each other)? And if so, how do I have to do this?
I have done some searching around the internet and this site, but I haven't found a good way to count the context switches on a thread in a c++ program I am running. I need to know if it get swapped off of a CPU for correct timing.
I have a timer thread, and when it expires, it sets a flag. The only problem is, if any higher priority thread comes in between, then the timer is not expired in real time.Thus i want to set a highest priority to my thread. Now, i know 2 algos, which can make my thread in real time are: SCHED_FIFO and SCHED_RR.
So, here are my queries: 1) Which scheduling algo is best suited for this purpose? 2) Is it guaranted to work in real time if i use that algo (you suggest in 1) and set the max_priority by getting the maximum settable priority for that particular algorithm using, int sched_get_priority_max(int policy);
I have created an error logger thread in my application using pthread_create(), I want this thread to have a lower priority than my application's priority so that the application would run properly. Is there an easy way to do this? I know that we can change priority for thread created with policies FIFO or RR, since I don't set the policy for my application, I don't know what to set for my thread.
I want to suspend/resume a thread. The library I am using is pthread.h.I am also running my application on linux.Is there any function in pthread. let me suspend a thread temporary?I have read a document in which it was mentioned thatthread does not support suspend/resume
I am new to thread programming. I need to generate one thread in one process using posix thread. Which scheduling and priority do i need to use? I want to generate the thread with the lowest priority. As i know there are 3 scheduling policy available SCHED_FIFO, SCHED_RR or SCHED_OTHAR.
This is a client server issue where the server creates a new thread for listening to a new client, now when the thread starts execution and during its execution the server needs to pass it an object of some X class, how's that to be done?
I am trying to pass an array to thread program. Its printing the val[0] value but its not printing value at val[1]. Can somebody help me with this problem.
Rather than get in hot (or even slightly-above-comfortable temperature) water for reviving a closed (ca. 2007) thread without moderator permission or acquaintance of the OP personally or by proxy, I'll post this alternate solution (approach?) along with one positive real-world example illustrating how mine is easier to write and is equally reliable, if not altogether better. The thread is here. Their solution:
Code: if [ "$myfilesize" eq "0" ];then echo "$file: the file size is zero"; else echo "$file: the file size is NOT zero"; fi (and this hangs on the output of a stat -c %s command). "Hangs" is probably a most appropriate word. The two outputs, using two different stat options, the same file, and three marginally-different command-line syntaxes I just tried gave me these: First, using their approach, I got this:
Code: tjsfa070929-094.jpg.error: the file size is NOT zero Even though stat gave me this Code: >> stat -c %s $file
[Code]...
From where I sit, getting the file type string instead of an integer, which can be equally (if not often more) accurate, makes for an easier time of setting the conditional and, maybe only in terms of a few ticks, a faster script from start to finish. I think my method is better, and proven so to the limits of my knowledge and ability here in this OP.
I hope someone can help me with this. I need to call the same thread multiple times, but only run one thread at a time. So what I do is create the thread which runs then drops out. Run the main loop and after some time pthread_create the same thread again. So in pseudo code
[Code]...
I always presumed if I run a thread and it reaches with a pthread_exit, then the thread was destroyed? However I've found that after I've created the thread several 100 times the programme crashes? So can anyone advise how I call the same thread over and over?
