Thank you for the information, Mulyadi and askb, To confirm again,
Since kernel thread by kernel_thread() is created by do_fork() in the kernel, pthread_create() uses clone() system call which eventually calls do_fork() in the kernel, and fork() in user space calls eventually do_fork() as well, all threads/process are manged by task_struct with different configuration(flags). Is this correct? If then, is my understanding that the implementation of do_fork() is by NPTL in linux 2.6 correct? If so, there is no heavy process in the linux in typical. Right? Best Regards, Daniel (Youngwhan) Song On Fri, Jan 15, 2010 at 8:28 PM, Mulyadi Santosa <mulyadi.sant...@gmail.com>wrote: > Hi Daniel... > > Here comes my thoughts.... > > On Sat, Jan 16, 2010 at 8:24 AM, Daniel (Youngwhan) Song > <breadn...@gmail.com> wrote: > > 1. It seems that the concept of thread in the linux kernel is lightweight > > process. Is this right? > > Yes. A bit OOT, in Windows, there's a lighter form than thread called > fiber. But in Linux, thread (NPTL, if we want to be specific) is the > actualization of LWP. > > > 1-1 If it is right, are both kernel thread and user-space thread > (pthread) > > lightweight process? Or does the lightweight process mean only to kernel > > thread? > > IMHO, LWP (Light Weight Process) is a concept which can be applied in > both space. However, there is correlation between user space thread > and kernel space thread. IIRC, in NPTL, the relationship is 1:1 thus > one user space thread is running on top of one kernel thread. (sorry > if I use wrong words, but hopefully you get it) > > Other threading implementation could do it differently e.g many user > space threads inside one thread group are multiplexed on top of one > kernel space thread > > > 2. I guess the instance from fork() is a process in user application, and > > the instance from pthread_create() is a thread. > > Correct > > >The former process is heavy > > weight process having copy-on-write feature and the later thread is > managed > > by pthread library handled in user mode. Is this right? > > Right, with additionaly note: pthread_create eventually calls clone() > with CLONE_VM and CLONE_THREAD flags IIRC. fork() calls clone() too > but without those two flags. So you can say, both essentially some > similar operations (e.g creating new task structs) but the rest is > different (one is creating whole address space, the other is not). > > > Any help would be appreciated. Thanks, > > Hopefully it makes things clearer :) > > -- > regards, > > Mulyadi Santosa > Freelance Linux trainer and consultant > > blog: the-hydra.blogspot.com > training: mulyaditraining.blogspot.com >
