Out of order read transfer with host memory buffer

Patent metadata

Field	Value
Publication number	US-10642496-B2
Application number	US-201615088692-A
Country	US
Kind code	B2
Filing date	Apr 1, 2016
Priority date	Apr 1, 2016
Publication date	May 5, 2020
Grant date	May 5, 2020

Abstract

A storage device may utilize a host memory buffer for re-ordering commands in a submission queue. Out of order commands in a submission queue that uses host virtual buffers that are not the same size may be difficult to search. Accordingly, commands in a submission queue may be correctly ordered in a host memory buffer before being put into the host virtual buffers. When the commands are in order, the search operation for specific data is improved.

First claim

We claim:

1. A method for data transfer in a memory device, the method comprising:
receiving a read command from a host, wherein the read command specifies segments of data to be read from non-volatile memory in the memory device;
dynamically allocating memory space in a host memory buffer in the host for the read command, wherein a size of the allocated memory space is equal to a total data a transfer size of the read command;
reading the segments of data from the non-volatile memory in an order different from an order specified in the read command;
storing the segments of data in the host memory buffer, wherein each segment of data is stored in a location in the host memory buffer that corresponds to a location of the segment of data in the order specified in the read command, wherein even though the segments of data are read out of order from the non-volatile memory, the segments of data are stored in order in the host memory buffer;
determining whether a last segment of data was read from the non-volatile memory;
in response to determining that the last segment of data was read from the non-volatile memory, performing the following in parallel:
reading the segments of data stored in the host memory buffer in the order in which they are stored in the host memory buffer;
identifying memory locations in a host memory that are designated as a host buffer for data transfers using a scatter gather list (SGL); and
transferring the segments of data from the host memory buffer to the locations in the host buffer identified in the SGL, wherein the segments of data are transferred in the order in which they are stored in the host memory buffer; and
de-allocating the memory space in the host memory buffer.

2. The method of claim 1 , wherein the host memory buffer is a portion allocated from memory on the host.

3. The method of claim 1 , wherein the host buffer is distinct from the host memory buffer.

4. The method of claim 1 , wherein the stored segments of data are transferred to a plurality of host buffers of different sizes.

5. The method of claim 1 , wherein the host buffer and the host memory buffer are stored in the host.

6. The method of claim 1 , wherein each of the plurality of segments of the SGL comprises:
descriptors that include a pointer; and
an indication of the size of the host buffer.

7. The method of claim 1 , wherein the read commands received from the host reference data that is out of order.

8. The method of claim 7 , wherein the data that is out of order is re-ordered by using an offset in the host memory buffer.

9. A host device configured to be connected to a memory device, the host device comprising:
one or more host buffers;
a host memory buffer; and
a controller configured to:
send a read command to a memory device, wherein the read command specifies segments of data to be read from non-volatile memory in the memory device;
store segments of data received from the memory device in the host memory buffer, wherein each segment of data is stored in a location in the host memory buffer that corresponds to a location of the segment of data in the order specified in the read command, wherein even though the segments of data are read out of order from the nonvolatile memory, the segments of data are stored in order in the host memory buffer;
determine whether a last segment of data was stored in the host memory buffer; and
in response to determining that the last segment of data was stored in the host memory buffer, perform the following in parallel:
read the segments of data stored in the host memory buffer in the order in which they are stored in the host memory buffer;
identify locations in the one or more host buffers for data transfers using a scatter gather list (SGL); and
transfer the segments of data from the host memory buffer to the locations in the one or more host buffers that are identified by the memory device, wherein

the segments of data are transferred in the order in which they are stored in the host memory buffer.

10. The host device of claim 9 , wherein the memory device utilizes a non-volatile memory express (NVMe) protocol.

11. The host device of claim 10 , wherein the host memory buffer is storage in the host device according to the NVMe protocol.

12. A system for data transfer in a memory, the system comprising:
a non-volatile memory;
means for receiving a read command from a host, wherein the read command specifies segments of data to be read from the non-volatile memory;
means for allocating memory space in a host memory buffer for the read command, wherein a size of the allocated memory space is equal to a total data transfer size of the read command;
means for reading the segments of data from the non-volatile memory in an order different from an order specified in the read command;
means for storing the segments of data in the host memory buffer, wherein each segment of data is stored in a location in the host memory buffer that corresponds to a location of the segment of data in the order specified in the read command, wherein even though the segments of data are read out of order from the non-volatile memory, the segments of data are stored in order in the host memory buffer;
means for determining whether a last segment of data was read from the non-volatile memory;
means for, in response to determining that the last segment of data was read from the nonvolatile memory, performing the following in parallel:
reading the segments of data stored in the host memory buffer in the order in which they are stored in the host memory buffer;
identifying memory locations in a host memory that are designated as a host buffer for data transfers using a scatter gather list (SGL); and
transferring the segments of data from the host memory buffer to the locations in the host buffer identified in the SGL, wherein the segments of data are transferred in the order in which they are stored in the host memory buffer; and

means for de-allocating the memory space in the host memory buffer.

13. The system of claim 12 , wherein the host buffer is accessed by the host.

14. The system of claim 12 , wherein a size of the host buffer varies based on the read command.

15. The system of claim 12 , wherein the host memory buffer and the host buffer are memories in the host.

16. The system of claim 12 , wherein the read command received from the host references data that is out of order and is re-ordered with the host memory buffer.

17. The method of claim 1 , wherein the non-volatile memory comprises a three-dimensional memory.

18. The host device of claim 9 , wherein the non-volatile memory comprises a three-dimensional memory.

19. The host device of claim 9 , wherein the memory device is embedded in the host device.

20. The system of claim 12 , wherein the non-volatile memory comprises a three-dimensional memory.

Assignees

• Sandisk Technologies Inc

Inventors

• Benisty Shay
• Sharifie Tal

Classifications

• G06F3/0688

  Non-volatile semiconductor memory arrays · CPC title

• G06F3/067

  Distributed or networked storage systems, e.g. storage area networks [SAN], network attached storage [NAS] · CPC title

• G06F3/0685

  Hybrid storage combining heterogeneous device types, e.g. hierarchical storage, hybrid arrays · CPC title

• G06F12/0871

  Allocation or management of cache space · CPC title

• G06F3/0659

  Command handling arrangements, e.g. command buffers, queues, command scheduling · CPC title