no reply patch

[Brian Aker]

Hi!

Add this to the code:

#if !defined(__GNUC__) || (__GNUC__ == 2 && __GNUC_MINOR__ < 96)
#define __builtin_expect(x, expected_value) (x)
#endif

#define likely(x)       __builtin_expect((x),1)
#define unlikely(x)     __builtin_expect((x),0)


Taken from the linux kernel headers.

Cheers,
	-Brian



On Feb 22, 2008, at 5:07 AM, dormando wrote:

> Committed this to trunk.
>
> I didn't add the unlikely() stuff yet, because we haven't explicitly
> dropped support for gcc-2.95, and I'm not familiar enough with its
> portability to say if it'd work for sun or not.
>
> After this release though, we should probably drop support for  
> gcc-2.95,
> which means freebsd4 users (*cough*) really ought to build memcached
> with a newer gcc. Although it'd be great if we could ditch it *for*
> release 1.2.5, so speak up if you care.
>
> -Dormando
>
> Tomash Brechko wrote:
>> On Tue, Feb 19, 2008 at 00:44:26 -0800, dormando wrote:
>>> For sanity I'm going to have to change it to look like this:
>>>
>>>    if (c->noreply) {
>>>        if (settings.verbose > 1)
>>>            fprintf(stderr, ">%d NOREPLY %s\n", c->sfd, str);
>>>        c->noreply = false;
>>>        conn_set_state(c, conn_read);
>>>        return;
>>>    }
>>
>> Ok, no objections.
>>
>>
>>> ...and for both of these, the test should be 'if unlikely(c- 
>>> >noreply)' -
>>
>> It hardly would make any difference, we are not in a tight loop.  You
>> could also write unlikely(settings.verbose > 1), but I'd say one
>> should save such optimizations to places there they do make  
>> noticeable
>> speedup.  OTOH, I guess GCC's __builtin_expect() may suggest the
>> compiler to reverse the order of branches.
>>
>>
>>> This also needs some more careful documentation warning about the
>>> occasional bad line potentially throwing clients out of whack. Since
>>> this is my pet peeve I'll add it myself.
>>
>> As was mentioned before, the problem is not limited to 'noreply'  
>> only.
>> Send to memcached "SET KEY\r\nVALUE\r\n" (in upper case), and enjoy
>> two replies instead of one.  With 'norepy' it's just one reply  
>> instead
>> of zero, same problem.  And this can only happen for malformed
>> command, where error reply is send before the command has been parsed
>> (memcached couldn't parse it).
>>
>>

--
_______________________________________________________
Brian "Krow" Aker, brian at tangent.org
Seattle, Washington
http://krow.net/                     <-- Me
http://tangent.org/                <-- Software
http://exploitseattle.com/    <-- Fun
_______________________________________________________
You can't grep a dead tree.