semaphore

[chiew peng]

Hai,
I'm new to multithread programming. Now has some doubt on it. Hope anybody
can help.

How do use the semphore?

Thanks

[Victor Bazarov]

Re: semaphore

"chiew peng" <limchiewpeng@i wow.com.sg> wrote...[color=blue]
> I'm new to multithread programming. Now has some doubt on it. Hope[/color]
anybody[color=blue]
> can help.
>
> How do use the semphore?[/color]

A semaphore is a global variable with two (rarely more) known
states. Threads can gain access to semaphores using some king
of OS mechanism and check the states of the semaphores. But
since C++ doesn't define semaphores, they are off-topic here.

The best place to ask is comp.programmin g.threads.

Victor

[chiew peng]

Re: semaphore

Ok, anyway thanks!

"Victor Bazarov" <v.Abazarov@att Abi.com> wrote in message
news:9_6Na.6718 $NW6.4716@rwcrn sc51.ops.asp.at t.net...[color=blue]
> "chiew peng" <limchiewpeng@i wow.com.sg> wrote...[color=green]
> > I'm new to multithread programming. Now has some doubt on it. Hope[/color]
> anybody[color=green]
> > can help.
> >
> > How do use the semphore?[/color]
>
> A semaphore is a global variable with two (rarely more) known
> states. Threads can gain access to semaphores using some king
> of OS mechanism and check the states of the semaphores. But
> since C++ doesn't define semaphores, they are off-topic here.
>
> The best place to ask is comp.programmin g.threads.
>
> Victor
>
>[/color]

[Alexander Terekhov]

Re: semaphore


Victor Bazarov wrote:[color=blue]
>
> "chiew peng" <limchiewpeng@i wow.com.sg> wrote...[color=green]
> > I'm new to multithread programming. Now has some doubt on it. Hope[/color]
> anybody[color=green]
> > can help.
> >
> > How do use the semphore?[/color]
>
> A semaphore is a global variable with two (rarely more) known
> states. Threads can gain access to semaphores using some king
> of OS mechanism and check the states of the semaphores. But
> since C++ doesn't define semaphores, they are off-topic here.[/color]

They are brain-damaged (as far as THREADING is concerned).
[color=blue]
>
> The best place to ask is comp.programmin g.threads.[/color]

<Forward Inline>

-------- Original Message --------
Newsgroups: comp.programmin g.threads
Subject: Re: semaphore

chiew peng wrote:[color=blue]
>
> Hi,
> I'm new to multithread programming. Now has some doubt on it. Hope anybody
> can help.
>
> How do use the semphore?[/color]

http://terekhov.de/downey03semaphores.pdf

(The Little Book of Semaphores, Copyright (C) 2003 Allen B. Downey)

Now, check out this thread (please follow the embedded links):

http://www.mail-archive.com/boost@lists.boost.org/msg07826.html

(Subject: [boost] no semaphores in boost::thread)

Try to NOT miss the following msgs of mine... I'll copy&paste only one
of them below:

http://www.mail-archive.com/boost@lists.boost.org/msg07506.html

http://www.mail-archive.com/boost@lists.boost.org/msg07519.html

http://www.mail-archive.com/boost@lists.boost.org/msg08431.html

Maciej Sobczak wrote:
[...][color=blue]
> What about providing both (condvars and semas), but with documenting
> known pros and cons?[/color]

Personally, I'd have no problems with some *separate* Boost.Semas (for
things meant to be done by the current POSIX/IPC semaphores: async-
signal-safe unlock operation, memory-isolated synchronization ) library
with the "attached sticker" ala "Intergalac tic Surgeon General's
Warning: DON'T USE IT FOR THREADING!" (or something like that). Well,
that would actually fit rather nicely into "The POSIX++ Vision". ;-)

regards,
alexander.

--

http://terekhov.de/your-next-book