Geek asks and answers

Memory Alignment - Sparc (Sun) compiler, Intel (Linux) g ++ compiler, Intel MVC compiler (Windows)

My question has two parts.

  • I noticed that memalign (block_size, bytes) on sparc (sun) with the cc compiler does not check that the bytes have a power of 2, and not memalign on intel linux and _aligned_malloc on the windows with the mvsc compiler, which check that the bytes are power 2 .

This is a mistake in the sun, is there a patch or do I need to write memalign myself.

  • In addition, I have a structure (not my code):

    typedef struct CLHLockStruct {
    volatile CLHLockNode CACHE_ALIGN *Tail ;
    volatile CLHLockNode CACHE_ALIGN *MyNode[N_THREADS] ;
    volatile CLHLockNode CACHE_ALIGN *MyPred[N_THREADS] ;
} CLHLockStruct;

I am compiling under MVSC (visual studio 2008):

    CACHE_LINE_SIZE = 64
    CACHE_ALIGN = __declspec(align(CACHE_LINE_SIZE)) 
    N_THREADS = 8
    sizeof(CLHLockStruct)=192

The code was originally written for the sparc architecture, and I'm trying to port it to MVSC without changing a lot of code.

memalign (CACHE_LINE_SIZE, sizeof (CLHLockStruct)), _aligned_malloc, , sizeof (CLHLockStruct) 2, , , 2.

?

EDIT1

, 2?

EDIT2

, _aligned_malloc malloc: , block_size, , 2?

+5
2

memalign (CACHE_LINE_SIZE, sizeof (CLHLockStruct)), _aligned_malloc, , sizeof (CLHLockStruct) 2,

_aligned_malloc(CACHE_LINE_SIZE,sizeof(CLHLockStruct))? , . _aligned_malloc , , - , memalign, . .

+4

1: POSIX memalign, . , blockize . ? , Win POSIX .

2: SO , aligned_malloc .

3: (: Google) , , .

+1

More articles:


All Articles