Re: [eigen]

[Benoit Jacob <jacob.benoit.1@xxxxxxxxx>]

I see what you mean, here we're hitting a limitation of the C and C++
languages, that there is for each type a single "sizeof" constant that
governs various separate things. The same issue is also why empty
objects have nonzero "sizeof" which leads to various issues too. Feel
free to submit idea to the C/C++ standardization groups ;)

Benoit

2009/10/12 Rohit Garg <rpg.314@xxxxxxxxx>:
> This is surprising. I didn't know that the compiler would pad
> everything to the next highest multiple of 4. Even if one aligns a
> float3 to 16 bytes, the last 4 bytes should be usable on stack for
> other variables. It's not like you typedef'ed a float3 and aligned it.
> Then if you declare an array of such type, I can understand that
> compiler will pad the w component to make all elements of that array
> aligned.
>
> Then why this behavior?
>
> On Mon, Oct 12, 2009 at 9:17 PM, Benoit Jacob <jacob.benoit.1@xxxxxxxxx> wrote:
>> 2009/10/12 Rohit Garg <rpg.314@xxxxxxxxx>:
>>> I am not sure regarding the round up of mem size bit. Why can't you
>>> have a float3 array aligned at 16 byte boundary?
>>
>> Because that makes its sizeof() increase to 16 bytes instead of 12, so
>> that incurs a +33% memory overhead for everybody when you allocate an
>> array of N such vectors --- and even so, this isn't quite optimal,
>> e.g. this doesn't give you a very easy way to vectorize the product of
>> 3x3 matrices with 3-vectors. For Vector5f, the situation is different,
>> the vectorization can always work (you can always fit a packet) but
>> the memory overhead is even bigger: 32/20 = 1.6 so it's a +60% memory
>> overhead.
>>
>> Have a look at the example program (attached):
>>
>> #include<iostream>
>>
>> template<int N> struct foo
>> {
>>  __attribute__((aligned(16))) float f[N];
>>  foo()
>>  {
>>    std::cout << "sizeof(foo<" << N << ">) is " << sizeof(foo)
>>              << " instead of " << N*sizeof(float) << std::endl;
>>  }
>> };
>>
>> int main()
>> {
>>  foo<3>();
>>  foo<5>();
>> }
>>
>>
>> Output:
>>
>> $ g++ sizeof_aligned.cpp -o s && ./s
>> sizeof(foo<3>) is 16 instead of 12
>> sizeof(foo<5>) is 32 instead of 20
>>
>>
>> Cheers,
>> Benoit
>>
>
>
>
> --
> Rohit Garg
>
> http://rpg-314.blogspot.com/
>
> Senior Undergraduate
> Department of Physics
> Indian Institute of Technology
> Bombay
>
>
>