Efficient context save/restore during hardware decompression of DEFLATE encoded data

Patent metadata

Field	Value
Publication number	US-9374106-B2
Application number	US-201314012382-A
Country	US
Kind code	B2
Filing date	Aug 28, 2013
Priority date	Aug 28, 2013
Publication date	Jun 21, 2016
Grant date	Jun 21, 2016

Abstract

A hardware accelerator receives a request to decompress a data stream that includes multiple deflate blocks and multiple deflate elements compressed according to block-specific compression configuration information. The hardware accelerator identifies a commit point that is based upon an interruption of a first decompression session of the data stream and corresponds to one of the deflate blocks. As such, the hardware accelerator configures a decompression engine based upon the corresponding deflate block's configuration information and, in turn, recommences decompression of the data stream at an input bit location corresponding to the commit point.

First claim

The invention claimed is:

1. A method utilizing an information handling system for resuming decompression of a data stream, the method comprising:
receiving a request to decompress a data stream that includes a plurality of deflate blocks, wherein each of the plurality of deflate blocks includes block-specific compression configuration information and a plurality of deflate elements compressed according to the block-specific compression configuration information;
identifying a commit point that is based upon an interruption of a first decompression session of the data stream and corresponds to one of the plurality of deflate blocks, wherein the commit point corresponds to a committed deflate element of the plurality of deflate elements included in a corresponding deflate block that was processed during the first decompression session to generate one or more first output bytes, the processing further comprising:
inserting one or more of the one or more first output bytes into an output word;
determining that one of the inserted one or more output bytes is a final one of the one or more first output bytes; and
generating context state data that indicates the commit point corresponds to the committed deflate element;
configuring a decompression engine based upon the corresponding deflate block's block-specific compression information; and
commencing a second decompression session of the data stream, using the configured decompression engine, at an input bit location corresponding to the commit point.

2. The method of claim 1 , wherein the plurality of deflate blocks is compressed according to a deflate compression algorithm that includes Huffman encoding, and wherein each of the plurality of deflate blocks is selected from a group consisting of a static deflate block, a dynamic deflate block, and a non-compressed block.

3. The method of claim 1 , wherein the configuring of the decompression engine further comprises:
loading the one or more first output bytes into the decompression engine as output history bytes; and
using the loaded one or more first output bytes during the second decompression session to generate one or more second output bytes corresponding to one or more subsequent deflate elements that are subsequent to the committed deflate element in the data stream.

4. The method of claim 1 , further comprising:
determining that the committed deflate element is a last deflate element included in the corresponding deflate block; and
including an end of block indicator in the context state data.

5. The method of claim 1 , wherein the data stream corresponds to a first logical partition executing on a host system that interfaces to a hardware accelerator, the method further comprising:
decompressing the data stream during the first decompression session on the hardware accelerator;
receiving a different request at the hardware accelerator from the host system to decompress a different data stream corresponding to a second logical partition executing on the host system;
interrupting the first decompression session at the hardware accelerator in response to the received different request; and
decompressing the different data stream prior to commencing the second decompression session of the data stream.

6. An information handling system comprising:
one or more processors;
a memory coupled to at least one of the processors; and
a hardware accelerator coupled to at least one of the processors, wherein the hardware accelerator performs a set of actions comprising:
receiving a request to decompress a data stream that includes a plurality of deflate blocks, wherein each of the plurality of deflate blocks includes block-specific compression configuration information and a plurality of deflate elements compressed according to the block-specific compression configuration information;
identifying a commit point that is based upon an interruption of a first decompression session of the data stream and corresponds to one of the plurality of deflate blocks, wherein the commit point corresponds to a committed deflate element of the plurality of deflate elements included in a corresponding deflate block that was processed during the first decompression session to generate one or more first output bytes, the processing further comprising:
inserting one or more of the one or more first output bytes into an output word;
determining that one of the inserted one or more output bytes is a final one of the one or more first output bytes; and
generating context state data that indicates the commit point corresponds to the committed deflate element;
configuring a decompression engine based upon the corresponding deflate block's block-specific compression information; and
commencing a second decompression session of the data stream, using the configured decompression engine, at an input bit location corresponding to the commit point.

7. The information handling system of claim 6 , wherein the plurality of deflate blocks is compressed according to a deflate compression algorithm that includes Huffman encoding, and wherein each of the plurality of deflate blocks is selected from a group consisting of a static deflate block, a dynamic deflate block, and a non-compressed block.

8. The information handling system of claim 6 , wherein the set of actions performed by the hardware accelerator further comprise:
loading the one or more first output bytes into the decompression engine as output history bytes; and
using the loaded one or more first output bytes during the second decompression session to generate one or more second output bytes corresponding to one or more subsequent deflate elements that are subsequent to the committed deflate element in the data stream.

9. The information handling system of claim 6 , wherein the set of actions performed by the hardware accelerator further comprise:
determining that the committed deflate element is a last deflate element included in the corresponding deflate block; and
including an end of block indicator in the context state data.

10. The information handling system of claim 6 , wherein the data stream corresponds to a first logical partition executing on the information handling system, the set of actions performed by the hardware accelerator further comprising:
decompressing the data stream during the first decompression session;
receiving a different request from the information handling system to decompress a different data stream corresponding to a second logical partition executing on the host system;
interrupting the first decompression session in response to the received different request; and
decompressing the different data stream prior to commencing the second decompression session of the data stream.

11. A computer program product stored in a non-transitory computer readable storage medium, comprising computer program code that, when executed by an information handling system, causes the information handling system to perform actions comprising:
receiving a request to decompress a data stream that includes a plurality of deflate blocks, wherein each of the plurality of deflate blocks includes block-specific compression configuration information and a plurality of deflate elements compressed according to the block-specific compression configuration information;
identifying a commit point that is based upon an interruption of a first decompression session of the data stream and corresponds to one of the plurality of deflate blocks, wherein the commit point corresponds to a committed deflate element of the plurality of deflate elements included in a corresponding deflate block that was processed during the first decompression sessi

Assignees

• IBM

Inventors

• Agarwal Kanak B
• Jamsek Damir A
• Martin Andrew K
• Rieke Reiner
• Vogt Joerg-Stephan
• Von Boehn Gunnar

Classifications

• H03M7/3086 Primary

  employing a sliding window, e.g. LZ77 · CPC title

• G06F17/30153

  Physics · mapped topic

• G06F17/30598

  Physics · mapped topic

• H03M7/40 Primary

  Conversion to or from variable length codes, e.g. Shannon-Fano code, Huffman code, Morse code · CPC title

• G06F16/24568

  Data stream processing; Continuous queries · CPC title