Re: [VOTE] Timing attack safe string comparison function

From: Date: Wed, 12 Feb 2014 08:11:21 +0000
Subject: Re: [VOTE] Timing attack safe string comparison function
References: 1 2 3 4 5 6 7  Groups: php.internals 
Request: Send a blank email to [email protected] to get a copy of this message
Hi all,

On Wed, Feb 12, 2014 at 4:24 PM, Lester Caine <[email protected]> wrote:

> Yasuo Ohgaki wrote:
>
>> We cannot assume word aligned data for strings. This is the reason why
>> there is
>> additional byte by byte comparison. It could be made to align memory to
>> word size,
>> but it requires memory allocation for comparison. I'm not sure which is
>> better, though.
>>
>
> Since PHP is controlling the memory management, there is no reason it
> can't properly align data elements ... based on the target platform. When
> you only had 640k of memory every byte counted, but these days the speed
> gain provided by simply wasting a few bytes and always padding buffers to a
> relevant word boundary is far greater than the space saving. Things have
> changed somewhat since the original C compilers were written. But this is
> better discussed on the 32/64bit optimization thread. Suffice to say that
> Firebird recorded substantial speed improvement once the basic ground rules
> were accepted and unnecessary 'byte related' processing dropped.


Word aligned memory allocation can be done. It may be good for
fragmentation also.
I don't read current memory manager code, so I cannot answer this.

Anyone?

--
Yasuo Ohgaki
[email protected]


Thread (54 messages)

« previous php.internals (#72504) next »