[Bug 1367] Malloc statistics wrong in R4.10

[bugzilla-daemon at rtems.org]

https://www.rtems.org/bugzilla/show_bug.cgi?id=1367

--- Comment #6 from Sebastian Huber <sebastian.huber at embedded-brains.de> 2012-04-11 03:20:47 CDT ---
Currently we have two statistics.  The one from the heap implementation:

typedef struct {
  /**
   * @brief Instance number of this heap.
   */
  uint32_t instance;

  /**
   * @brief Size of the allocatable area in bytes.
   *
   * This value is an integral multiple of the page size.
   */
  uintptr_t size;

  /**
   * @brief Current free size in bytes.
   *
   * This value is an integral multiple of the page size.
   */
  uintptr_t free_size;

  /**
   * @brief Minimum free size ever in bytes.
   *
   * This value is an integral multiple of the page size.
   */
  uintptr_t min_free_size;

  /**
   * @brief Current number of free blocks.
   */
  uint32_t free_blocks;

  /**
   * @brief Maximum number of free blocks ever.
   */
  uint32_t max_free_blocks;

  /**
   * @brief Current number of used blocks.
   */
  uint32_t used_blocks;

  /**
   * @brief Maximum number of blocks searched ever.
   */
  uint32_t max_search;

  /**
   * @brief Total number of successful allocations.
   */
  uint32_t allocs;

  /**
   * @brief Total number of searches ever.
   */
  uint32_t searches;

  /**
   * @brief Total number of suceessful calls to free.
   */
  uint32_t frees;

  /**
   * @brief Total number of successful resizes.
   */
  uint32_t resizes;
} Heap_Statistics;

And the one from malloc:

typedef struct {
    uint32_t    space_available;             /* current size of malloc area */
    uint32_t    malloc_calls;                /* # calls to malloc */
    uint32_t    memalign_calls;              /* # calls to memalign */
    uint32_t    free_calls;
    uint32_t    realloc_calls;
    uint32_t    calloc_calls;
    uint32_t    max_depth;             /* most ever malloc'd at 1 time */
    uintmax_t   lifetime_allocated;
    uintmax_t   lifetime_freed;
} rtems_malloc_statistics_t;

I would like to remove the special case malloc() statistics and use the heap
statistics instead.  This makes the CONFIGURE_MALLOC_STATISTICS option
obsolete.

The malloc() statistics currently have a max_depth, lifetime_allocated, and
lifetime_freed field.  Should we add this to the heap statistics?

-- 
Configure bugmail: https://www.rtems.org/bugzilla/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are watching all bug changes.