# US 2017/0286303 A1 · Prefetch mechanism for servicing demand miss

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2017/0286303 A1
- Document ID: US20170286303A1
- Status: Published application
- Kind: A1
- Application number: 15/087,649
- Filed: 2016-03-31
- Published: 2017-10-05
- Earliest claimed: 2016-03-31
- Assignees: QUALCOMM Incorporated
- Applicants: QUALCOMM Incorporated
- Inventors: Khary Jason ALEXANDER; Eric Francis ROBINSON
- Web page: https://patentgrep.com/patent/US20170286303A1

## CPC

- G06F 12/0862: with prefetch
- G06F 12/0811: with multilevel cache hierarchies
- G06F 12/0891: using clearing, invalidating or resetting means
- G06F 12/0897: with two or more cache hierarchy levels (with multilevel cache hierarchies G06F12/0811)
- G06F 2212/6022: Using a prefetch buffer or dedicated prefetch cache
- G06F 2212/62: Details of cache specific to multiprocessor cache arrangements
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)

## Abstract

Systems and methods relate to servicing a demand miss for a cache line in a first cache (e.g., an L1 cache) of a processing system, for example, when none of one or more fill buffers for servicing the demand miss are available. In exemplary aspects, the demand miss is converted to a prefetch operation to prefetch the cache line into a second cache (e.g., an L2 cache), wherein the second cache is a backing storage location for the first cache. Thus, servicing the demand miss is not delayed until a fill buffer becomes available, and once a fill buffer becomes available, the prefetched cache line is returned from the second cache to the available fill buffer.

## Claims

### Claim 1

A method of cache access in a processing system, the method comprising:
determining that there is a first demand miss for a first cache line in a first cache; and
when no fill buffer is available to service the first demand miss, converting the first demand miss into a prefetch operation for prefetching the first cache line into a second cache.

### Claim 2 (depends on 1)

The method of claim 1, further comprising returning the first cache line from the second cache to a fill buffer, when the fill buffer becomes available.

### Claim 3 (depends on 2)

The method of claim 2, further comprising writing the first cache line from the fill buffer to the first cache.

### Claim 4 (depends on 2)

The method of claim 2, further comprising marking the first cache line with a hint to not be replaced or evicted from the second cache until the first cache line is returned to the fill buffer.

### Claim 5 (depends on 1)

The method of claim 1, wherein converting the first demand miss into the prefetch operation comprises holding a first address of the first cache line in a prefetch request queue (PRQ), and forwarding the first address from the PRQ to prefetch mechanisms for prefetching the first cache line into the second cache.

### Claim 6 (depends on 5)

The method of claim 5, comprising releasing the PRQ from holding the first address after forwarding the first address to the prefetch mechanisms.

### Claim 7 (depends on 5)

The method of claim 5, comprising retaining the PRQ for holding the first address until a fill buffer becomes available to service a second demand miss for the first cache line encountered by the first cache.

### Claim 8 (depends on 5)

The method of claim 5, comprising storing prefetch addresses generated by a prefetch engine in the PRQ, when the PRQ is not holding the first address.

### Claim 9 (depends on 1)

The method of claim 1, wherein the first cache is a L1 cache and the second cache is a L2 cache.

### Claim 10 (depends on 1)

The method of claim 1, wherein no fill buffer is available to service the first demand miss when one or more fill buffers of the processing system are busy servicing other demand misses to the first cache.

### Claim 11 (depends on 1)

The method of claim 1, wherein the second cache is a backing storage of the first cache.

### Claim 12

An apparatus comprising:
a first cache;
one or more fill buffers configured to service demand misses to the first cache;
a second cache; and
logic configured to convert a first demand miss for a first cache line in the first cache to a prefetch operation, to prefetch the first cache line into the second cache if none of the one or more fill buffers are available to service the first demand miss.

### Claim 13 (depends on 12)

The apparatus of claim 12, the first cache line is returned from the second cache to a fill buffer of the one or more fill buffers, when the fill buffer becomes available.

### Claim 14 (depends on 13)

The apparatus of claim 13, wherein the first cache line is marked with a hint to not to be replaced or evicted from the second cache until the first cache line is returned to the fill buffer.

### Claim 15 (depends on 12)

The apparatus of claim 12, wherein the logic configured to convert the first demand miss for the first cache line in the first cache comprises a prefetch request queue (PRQ) configured to hold a first address of the first cache line and to forward the first address from the PRQ to prefetch mechanisms configured to prefetch the first cache line into the second cache.

### Claim 16 (depends on 15)

The apparatus of claim 15, wherein the PRQ does not hold the first address after the first address is forwarded to the prefetch mechanisms.

### Claim 17 (depends on 15)

The apparatus of claim 15, wherein the first address is held in the PRQ until a fill buffer becomes available to service a second demand miss for the first cache line encountered by the first cache.

### Claim 18 (depends on 15)

The apparatus of claim 15, wherein the PRQ is further configured to store prefetch addresses generated by a prefetch engine, when the first address is not held in the PRQ.

### Claim 19 (depends on 15)

The apparatus of claim 15, wherein a size of the PRQ is smaller than any one of the one or more fill buffers.

### Claim 20 (depends on 12)

The apparatus of claim 12, wherein the first cache is a L1 cache and the second cache is a L2 cache.

### Claim 21 (depends on 12)

The apparatus of claim 12, wherein the second cache is a backing storage of the first cache.

### Claim 22 (depends on 12)

The apparatus of claim 12, integrated into a device selected from the group consisting of a set top box, a music player, a video player, an entertainment unit, a navigation device, a personal digital assistant (PDA), a fixed location data unit, a computer, a laptop, a tablet, a communications device, and a mobile phone.

### Claim 23

An apparatus comprising:
a first cache;
one or more means for servicing demand misses in the first cache;
a second cache; and
means for converting a first demand miss for a first cache line in the first cache, to a prefetch operation for prefetching the first cache line into the second cache if none of the one or more means for servicing are available.

### Claim 24 (depends on 23)

The apparatus of claim 23, further comprising means for returning the first cache from the second cache to a means for servicing, when the means for servicing becomes available.

### Claim 25 (depends on 23)

The apparatus of claim 23, wherein the means for converting comprises a means for holding a first address of the first cache line and means for prefetching the first cache line from the first address into the second cache.

### Claim 26 (depends on 25)

The apparatus of claim 25, wherein a size of the means for holding is smaller than a size of the means for servicing.

### Claim 27

A non-transitory computer-readable storage medium comprising code, which, when executed by a processor, causes the processor to perform operations for accessing a cache, wherein the non-transitory computer-readable storage medium comprises:
code for determining that there is a first demand miss for a first cache line in a first cache; and
code for converting the first demand miss into a prefetch operation for prefetching the first cache line into a second cache when no fill buffer is available to service the first demand miss.

### Claim 28 (depends on 27)

The non-transitory computer-readable storage medium of claim 27, further comprising code for returning the first cache line from the second cache to a fill buffer, when the fill buffer becomes available.

### Claim 29 (depends on 28)

The non-transitory computer-readable storage medium of claim 28, further comprising code for marking the first cache line with a hint to not to be replaced or evicted from the second cache until the first cache line is returned to the fill buffer.

### Claim 30 (depends on 1)

The non-transitory computer-readable storage medium of claim 1, comprising code for holding a first address of the first cache line in a prefetch request queue (PRQ), and code for forwarding the first address from the PRQ to prefetch mechanisms for prefetching the first cache line into the second cache.
