The article on exception safety (http://www.digitalmars.com/d/exception-safe.html) uses this example:

void abc()
{
    Mutex m = new Mutex;

    lock(m);	// lock the mutex
    scope(exit) unlock(m);	// unlock on leaving the scope

    foo();	// do processing
}

However, it seems to me this won't actually work as written. Doesn't any variable used for scoping need to be declared either static or outside the scope it's protecting? As written, it looks to me as though a new local m will be created on the stack each time a thread enters this func, which won't protect anything!

TimK

Reply to Tim,

> The article on exception safety
> (http://www.digitalmars.com/d/exception-safe.html) uses this example:
> 
> void abc()
> {
> Mutex m = new Mutex;
> lock(m);	// lock the mutex
> scope(exit) unlock(m);	// unlock on leaving the scope
> foo();	// do processing
> }
> However, it seems to me this won't actually work as written. Doesn't
> any variable used for scoping need to be declared either static or
> outside the scope it's protecting? As written, it looks to me as
> though a new local m will be created on the stack each time a thread
> enters this func, which won't protect anything!
> 
> TimK
> 

Good point, and I think you are correct, but I think the intent is still clear.

Tim Keating wrote:
> The article on exception safety (http://www.digitalmars.com/d/exception-safe.html) uses this example:
> 
> void abc()
> {
>     Mutex m = new Mutex;
> 
>     lock(m);    // lock the mutex
>     scope(exit) unlock(m);    // unlock on leaving the scope
> 
>     foo();    // do processing
> }
> 
> However, it seems to me this won't actually work as written. Doesn't any variable used for scoping need to be declared either static or outside the scope it's protecting? As written, it looks to me as though a new local m will be created on the stack each time a thread enters this func, which won't protect anything!
> 
> TimK

On win32, it could be a named mutex, with the name hardcoded :P

L.