Compressed victim cache

1. A method for caching a block of data, comprising:

receiving a request to store the block;

determining whether an active cache is able to expand;

if the active cache is able to expand:

expanding the active cache at runtime to obtain an expanded active cache, wherein expanding the active cache comprises increasing the size of the active cache, and

storing the block in the expanded active cache;

if the active cache is not able to expand:

determining whether evictable blocks are present in a victim cache, wherein the victim cache is structured to comprise a most recently used (MRU) evictable portion and a most frequently used (MFU) evictable portion, wherein the MRU evictable portion comprises blocks that previously had at most one active reference and currently have no active references, and the MFU evictable portion comprises blocks that previously had at least two active references and currently have no active references;

if evictable blocks are present in the victim cache:

evicting a sufficient number of the evictable blocks from the victim cache to facilitate storing the block in the active cache,

wherein the evictable blocks in the victim cache are compressed.

2. The method of claim 1 , wherein the active cache comprises blocks that have at least one active reference.

3. The method of claim 1 , wherein the victim cache supports multiple block sizes.

4. The method of claim 1 , wherein determining whether to expand the cache comprises determining whether a memory state of a system in which the cache resides is low.

5. The method of claim 1 , wherein the active cache and the victim cache are located in a single memory.

6. The method of claim 1 , wherein evicting the sufficient number of the evictable blocks from the victim cache comprises:

evicting evictable blocks in a MRU evictable portion of the victim cache; and

evicting evictable blocks in a MFU evictable portion of the victim cache only if the sufficient number of evictable blocks is not present in the MRU evictable portion.

7. A system, comprising:

a memory, wherein the memory comprises an active cache and a victim cache, wherein the victim cache is structured to comprise a most recently used (MRU) evictable portion and a most frequently used (MFU) evictable portion, wherein the MRU evictable portion comprises blocks that previously had at most one active reference and currently have no active references, and the MFU evictable portion comprises blocks that previously had at least two active references and currently have no active references,

and wherein the system is configured to:

receive a request to store a block of data;

determine whether the active cache is able to expand;

if the active cache is able to expand:

expand the active cache at runtime to obtain an expanded active cache, wherein expanding the active cache comprises increasing the size of the active cache, and

store the block in the expanded active cache;

if the active cache is not able to expand:

determine whether evictable blocks are present in the victim cache;

if evictable blocks are present in the victim cache:

determine whether a total size of the evictable blocks is greater than or equal to a size of the block;

evict a sufficient number of the evictable blocks from the victim cache to facilitate storing the block in the active cache,

wherein the evictable blocks in the victim cache are compressed.

8. The system of claim 7 , wherein a size of the block is within a range of 512 bytes and 128 K.

9. The system of claim 7 , wherein the active cache comprises blocks that have at least one active reference.

10. The system of claim 7 , wherein determining whether to expand the cache comprises determining whether a memory state of a system in which the cache resides is low.

11. The system of claim 7 , wherein the active cache comprises blocks that have at least two active references.

12. The system of claim 7 , wherein evicting the sufficient number of the evictable blocks from the victim cache comprises:

evicting evictable blocks in a MRU evictable portion of the victim cache; and

evicting evictable blocks in a MFU evictable portion of the victim cache only if the sufficient number of evictable blocks is not present in the MRU evictable portion.

13. A computer readable medium comprising instructions for performing a method for caching a block of data, the method comprising:

receiving a request for the block from a process;

determining whether the block is in an active cache;

returning the block to the process, if the block is in the active cache;

if the block is not in the active cache:

determining whether the block is in a victim cache, wherein the victim cache is structured to comprise a most recently used (MRU) evictable portion and a most frequently used (MFU) evictable portion, wherein the MRU evictable portion comprises blocks that previously had at most one active reference and currently have no active references, and the MFU evictable portion comprises blocks that previously had at least two active references and currently have no active references;

placing the block in the active cache and returning the block to the process, if the block is in the victim cache;

obtaining the block from persistent storage, if the block is not in the victim cache, wherein the block is in a compressed form, when the block is located in the victim cache.

14. The computer readable medium of claim 13 , wherein the active cache and the victim cache are located in a single memory.

15. The computer readable medium of claim 13 , wherein placing the block in the active cache comprises uncompressing the block prior to placing the block in the active cache.

16. The computer readable medium of claim 13 , wherein returning the block to the process comprises returning a reference to the block.