Method and program product for avoiding cache congestion by offsetting addresses while allocating memory
View Patent ↗A method of allocating memory operates to avoid overlapping hot spots in cache that can ordinarily cause cache thrashing. This method includes steps of determining a spacer size, reserving a spacer block of memory from a memory pool, and allocating memory at a location following the spacer block. In an alternative embodiment, the spacer size is determined randomly in a range of allowable spacer size. In other alternative embodiments, spacers are allocated based upon size of a previously allocated memory block.
1. A method of dynamically allocating memory of a computer system operable when a program running on the computer system requests allocation of a requested memory block from a memory pool comprising steps of:
testing a size of the requested memory block to determine if the size is representable as N times two raised to the power M times a cache line size of the computer system, where N is an integer equal to or greater than 1, and M is an integer greater than one; and, if the size of the requested memory block is so representable, further comprising:
determining a spacer size determined as a random spacer size within a predetermined range of allowable spacer size,
reserving a spacer block of memory from the memory pool, the spacer block being of the spacer size; and
allocating the memory block, adjacent to the spacer block, from the memory pool.
2. The method of claim 1 , wherein the integer M is parameterized such that it may be adjusted to optimize performance.
3. The method of claim 2 wherein the range of permissible spacer sizes is adjustable by a system administrator.