• Joshua Moore-Oliva (2/2) Aug 14 2009 I am doing research on multi-threaded call stack mechanisms, and in addi...
• Walter Bright (2/8) Aug 14 2009 That's right. It's a layer over the C thread creation mechanism.
• Jarrett Billingsley (13/14) Aug 14 2009 ion to academic papers, I am surveying what mechanisms existing language...
• Brad Roberts (9/17) Aug 14 2009 Answering carefully, neither C, C++ nor D define how threads behave.

Joshua Moore-Oliva <josh chatgris.com> writes:

I am doing research on multi-threaded call stack mechanisms, and in addition to
academic papers, I am surveying what mechanisms existing languages use. Does D
use a mechanism other than the standard C-pthread style "one stack is allocated
on thread creation for each thread, and if that space is exhausted it is
considered an error without recovery"? If so, could you detail your call stack
mechanism?

Thanks, Joshua Moore-Oliva 

Walter Bright <newshound1 digitalmars.com> writes:

Joshua Moore-Oliva wrote:
 I am doing research on multi-threaded call stack mechanisms, and in
 addition to academic papers, I am surveying what mechanisms existing
 languages use. Does D use a mechanism other than the standard
 C-pthread style "one stack is allocated on thread creation for each
 thread, and if that space is exhausted it is considered an error
 without recovery"? If so, could you detail your call stack mechanism?

That's right. It's a layer over the C thread creation mechanism.

Jarrett Billingsley <jarrett.billingsley gmail.com> writes:

On Fri, Aug 14, 2009 at 3:44 PM, Joshua Moore-Oliva<josh chatgris.com> wrot=
e:
 I am doing research on multi-threaded call stack mechanisms, and in addit=

ion to academic papers, I am surveying what mechanisms existing languages u=
se. Does D use a mechanism other than the standard C-pthread style "one sta=
ck is allocated on thread creation for each thread, and if that space is ex=
hausted it is considered an error without recovery"? If so, could you detai=
l your call stack mechanism?

That's pretty much it.

That being said, druntime and Tango's runtime (basically the same
thing..) also allow for the creation of Fibers, which are
cooperatively-scheduled threads of execution, as opposed to
preemptively-scheduled ones.  But they still work pretty much the same
way, by allocating a new stack.

Brad Roberts <braddr bellevue.puremagic.com> writes:

On Fri, 14 Aug 2009, Joshua Moore-Oliva wrote:

 I am doing research on multi-threaded call stack mechanisms, and in 
 addition to academic papers, I am surveying what mechanisms existing 
 languages use. Does D use a mechanism other than the standard C-pthread 
 style "one stack is allocated on thread creation for each thread, and if 
 that space is exhausted it is considered an error without recovery"? If 
 so, could you detail your call stack mechanism?
 
 Thanks, Joshua Moore-Oliva 

Answering carefully, neither C, C++ nor D define how threads behave.  
Both C++ and D intend to but even there I don't think they define the 
specifics of stack usage but rather leave that to the ABI to define.

However, practically speaking.. yes, D follows in the footsteps of C and 
defers that to the pthread libraries for the most part.  See the other 
responses on the thread.

Later,
Brad