Re^3: wait for threads to end in parallel

Syntax. async is shorter, but can only pass args via capture.

Not so. async is identical to threads->create(), except that it is invoked as a function rather than a class method:

# 'async' is a function alias for the 'threads->create()' method
sub async (&;@)
{
    unshift(@_, 'threads');
    # Use "goto" trick to avoid pad problems from 5.8.1 (fixed in 5.8.
+2)
    goto &create;
}
[download]

So async \&sub, $arg, 1, 'fred'; or even async{ ... } $arg, 1, 'fred';

are identical in effect to threads->create( \&sub, $arg, 1, 'fred' );

Examine what is said, not who speaks -- Silence betokens consent -- Love the truth but pardon error.

"Science is about questioning the status quo. Questioning authority".

In the absence of evidence, opinion is indistinguishable from prejudice.

"I'd rather go naked than blow up my ass"

Comment on Re^3: wait for threads to end in parallel Select or Download Code

Replies are listed 'Best First'.
Re^4: wait for threads to end in parallel by ikegami (Patriarch) on Mar 16, 2010 at 16:47 UTC
Ah yes, I forgot about that. Revised answer: Syntax. `async` is shorter, and it simply pass args via capture. That's good, because passing args to a thread as arguments is known to be problematic. Without args: `async(\&process); async { process(); }; thread->create(\&process);` [download] With args: `async(\&process, $arg); # Safe?? thread->create(\&process, $arg); # Safe?? async { process($arg); }; thread->create(sub { process($arg) });` [download] — The stack isn't ref-counted, and this causes issues when thread creation attempts to clone the variables. I can't remember the exact scenario, but I can look it up if you want more info.	[reply] [d/l] [select]
Re^5: wait for threads to end in parallel by BrowserUk (Patriarch) on Mar 16, 2010 at 17:11 UTC
The stack isn't ref-counted, and this causes issues when thread creation attempts to clone the variables. I can't remember the exact scenario, but I can look it up if you want more info. If you have a moments at some point, I would be interested to read about it. Thanks. Examine what is said, not who speaks -- Silence betokens consent -- Love the truth but pardon error. "Science is about questioning the status quo. Questioning authority". In the absence of evidence, opinion is indistinguishable from prejudice. "I'd rather go naked than blow up my ass"	[reply]
Re^6: wait for threads to end in parallel by ikegami (Patriarch) on Mar 16, 2010 at 18:09 UTC
Ah, there it is: RT 70602. The problem was a noisy memory leak. Apparently, it's fixed in 5.12.	[reply]