develooper Front page | perl.perl5.porters | Postings from January 2001

Re: Does perl really need to use sigsetjmp? (18% performance hit)

Thread Previous | Thread Next
From:
nick
Date:
January 11, 2001 14:43
Subject:
Re: Does perl really need to use sigsetjmp? (18% performance hit)
Message ID:
E14GqOG-0001M2-00@roam1
Nick Ing-Simmons <nik@tiuk.ti.com> writes:
>>Solaris/sparc
>>$ jmp
>>Loop overhead: 22 ns/iter
>>setjmp(jb): 32 ns/call
>>sigsetjmp(sjb, 0): 11795 ns/call
>>sigsetjmp(sjb, 1): 11856 ns/call
>
>Many thanks. 
>
>It was a thought ... - more #if-ery will be required.

Perhaps not. 

The attached patch makes Linux work on the original 
die-out-of signal handler with basic setjmp().

So I _think_ we can just use setjmp() everywhere.
The only area of worry is if there is some machine which has working 
sigsetjmp() but no sigprocmask(). 

So when it makes it to a snapshot I would appreciate testing on 
"older" or "awkward" platforms.

I will try and cook up a non-interactive test... 

-- 
Nick Ing-Simmons


Thread Previous | Thread Next


nntp.perl.org: Perl Programming lists via nntp and http.
Comments to Ask Bjørn Hansen at ask@perl.org | Group listing | About