公开(公告)号：US11347514B2

公开(公告)日：2022-05-31

申请号：US16277764

申请日：2019-02-15

Applicant: Apple Inc.

Inventor： Deepak Limaye ,  Brian R. Mestan ,  Gideon N. Levinsky

IPC: G06F12/00 ,  G06F9/38 ,  G06F12/0815 ,  G06F12/0891 ,  G06F12/0864

Abstract: Techniques are disclosed relating to filtering access to a content-addressable memory (CAM). In some embodiments, a processor monitors for certain microarchitectural states and filters access to the CAM in states where there cannot be a match in the CAM or where matching entries will not be used even if there is a match. In some embodiments, toggle control circuitry prevents toggling of input lines when filtering CAM access, which may reduce dynamic power consumption. In some example embodiments, the CAM is used to access a load queue to validate that out-of-order execution for a set of instructions matches in-order execution, and situations where ordering should be checked are relatively rare.

公开(公告)号：US11928467B2

公开(公告)日：2024-03-12

申请号：US17473076

申请日：2021-09-13

Applicant: Apple Inc.

Inventor： Brian R. Mestan ,  Gideon N. Levinsky ,  Michael L. Karm

IPC: G06F9/38 ,  G06F9/30 ,  G06F9/32 ,  G06F9/52

CPC classification number: G06F9/30043 ,  G06F9/3004 ,  G06F9/30087 ,  G06F9/321 ,  G06F9/3826 ,  G06F9/3834 ,  G06F9/3842 ,  G06F9/528 ,  G06F2209/521

Abstract: In an embodiment, a processor comprises an atomic predictor circuit to predict whether or not an atomic operation will complete successfully. The prediction may be used when a subsequent load operation to the same memory location as the atomic operation is executed, to determine whether or not to forward store data from the atomic operation to the subsequent load operation. If the prediction is successful, the store data may be forwarded. If the prediction is unsuccessful, the store data may not be forwarded. In cases where an atomic operation has been failing (not successfully performing the store operation), the prediction may prevent the forwarding of the store data and thus may prevent a subsequent flush of the load.

公开(公告)号：US20220091846A1

公开(公告)日：2022-03-24

申请号：US17473076

申请日：2021-09-13

Applicant: Apple Inc.

Inventor： Brian R. Mestan ,  Gideon N. Levinsky ,  Michael L. Karm

IPC: G06F9/30 ,  G06F9/38 ,  G06F9/32 ,  G06F9/52

Abstract: In an embodiment, a processor comprises an atomic predictor circuit to predict whether or not an atomic operation will complete successfully. The prediction may be used when a subsequent load operation to the same memory location as the atomic operation is executed, to determine whether or not to forward store data from the atomic operation to the subsequent load operation. If the prediction is successful, the store data may be forwarded. If the prediction is unsuccessful, the store data may not be forwarded. In cases where an atomic operation has been failing (not successfully performing the store operation), the prediction may prevent the forwarding of the store data and thus may prevent a subsequent flush of the load.

公开(公告)号：US20220083484A1

公开(公告)日：2022-03-17

申请号：US17246311

申请日：2021-04-30

Applicant: Apple Inc.

Inventor： Jeffrey E. Gonion ,  Charles E. Tucker ,  Tal Kuzi ,  Richard F. Russo ,  Mridul Agarwal ,  Christopher M. Tsay ,  Gideon N. Levinsky ,  Shih-Chieh Wen

IPC: G06F13/24 ,  G06F1/26

Abstract: An interrupt delivery mechanism for a system includes and interrupt controller and a plurality of cluster interrupt controllers coupled to respective pluralities of processors in an embodiment. The interrupt controller may serially transmit an interrupt request to respective cluster interrupt controllers, which may acknowledge (Ack) or non-acknowledge (Nack) the interrupt based on attempting to deliver the interrupt to processors to which the cluster interrupt controller is coupled. In a soft iteration, the cluster interrupt controller may attempt to deliver the interrupt to processors that are powered on, without attempting to power on processors that are powered off. If the soft iteration does not result in an Ack response from one of the plurality of cluster interrupt controllers, a hard iteration may be performed in which the powered-off processors may be powered on.

公开(公告)号：US20200264888A1

公开(公告)日：2020-08-20

申请号：US16277764

申请日：2019-02-15

Applicant: Apple Inc.

Inventor： Deepak Limaye ,  Brian R. Mestan ,  Gideon N. Levinsky

IPC: G06F9/38 ,  G06F12/0891 ,  G06F12/0864 ,  G06F12/0815

Abstract: Techniques are disclosed relating to filtering access to a content-addressable memory (CAM). In some embodiments, a processor monitors for certain microarchitectural states and filters access to the CAM in states where there cannot be a match in the CAM or where matching entries will not be used even if there is a match. In some embodiments, toggle control circuitry prevents toggling of input lines when filtering CAM access, which may reduce dynamic power consumption. In some example embodiments, the CAM is used to access a load queue to validate that out-of-order execution for a set of instructions matches in-order execution, and situations where ordering should be checked are relatively rare.

公开(公告)号：US20210349823A1

公开(公告)日：2021-11-11

申请号：US16870358

申请日：2020-05-08

Applicant: Apple Inc.

Inventor： Michael L. Karm ,  Gideon N. Levinsky

IPC: G06F12/0815

Abstract: In one embodiment, a processor includes a write combining buffer that includes a memory having a plurality of entries. The entries may be allocated to committed store operations transmitted by a load/store unit in the processor, and subsequent committed store operations may merge data with previous store memory operations in the buffer if the subsequent committed store operations are to addresses that match addresses of the previous committed store operations within a predefined granularity (e.g. the width of a cache port). The write combining buffer may be configured to retain up to N entries of committed store operations, but may also be configured to write one or more of the entries to the data cache responsive to receiving more than a threshold amount of non-merging committed store operations in the write combining buffer.

公开(公告)号：US20240311319A1

公开(公告)日：2024-09-19

申请号：US18674203

申请日：2024-05-24

Applicant: Apple Inc.

Inventor： Jeffrey E. Gonion ,  Charles E. Tucker ,  Tal Kuzi ,  Richard F. Russo ,  Mridul Agarwal ,  Christopher M. Tsay ,  Gideon N. Levinsky ,  Shih-Chieh Wen ,  Lior Zimet

IPC: G06F13/24 ,  G06F1/26

CPC classification number: G06F13/24 ,  G06F1/26

Abstract: An interrupt delivery mechanism for a system includes and interrupt controller and a plurality of cluster interrupt controllers coupled to respective pluralities of processors in an embodiment. The interrupt controller may serially transmit an interrupt request to respective cluster interrupt controllers, which may acknowledge (Ack) or non-acknowledge (Nack) the interrupt based on attempting to deliver the interrupt to processors to which the cluster interrupt controller is coupled. In a soft iteration, the cluster interrupt controller may attempt to deliver the interrupt to processors that are powered on, without attempting to power on processors that are powered off. If the soft iteration does not result in an Ack response from one of the plurality of cluster interrupt controllers, a hard iteration may be performed in which the powered-off processors may be powered on.

公开(公告)号：US20220083338A1

公开(公告)日：2022-03-17

申请号：US17469504

申请日：2021-09-08

Applicant: Apple Inc.

Inventor： Jeff Gonion ,  John H. Kelm ,  James Vash ,  Pradeep Kanapathipillai ,  Mridul Agarwal ,  Gideon N. Levinsky ,  Richard F. Russo ,  Christopher M. Tsay

IPC: G06F9/30 ,  G06F12/0875 ,  G06F12/02

Abstract: Techniques are disclosed relating to data synchronization barrier operations. A system includes a first processor that may receive a data barrier operation request from a second processor include in the system. Based on receiving that data barrier operation request from the second processor, the first processor may ensure that outstanding load/store operations executed by the first processor that are directed to addresses outside of an exclusion region have been completed. The first processor may respond to the second processor that the data barrier operation request is complete at the first processor, even in the case that one or more load/store operations that are directed to addresses within the exclusion region are outstanding and not complete when the first processor responds that the data barrier operation request is complete.

公开(公告)号：US11256622B2

公开(公告)日：2022-02-22

申请号：US16870358

申请日：2020-05-08

Applicant: Apple Inc.

Inventor： Michael L. Karm ,  Gideon N. Levinsky

IPC: G06F12/08 ,  G06F12/0815 ,  G06F12/0811

Abstract: In one embodiment, a processor includes a write combining buffer that includes a memory having a plurality of entries. The entries may be allocated to committed store operations transmitted by a load/store unit in the processor, and subsequent committed store operations may merge data with previous store memory operations in the buffer if the subsequent committed store operations are to addresses that match addresses of the previous committed store operations within a predefined granularity (e.g. the width of a cache port). The write combining buffer may be configured to retain up to N entries of committed store operations, but may also be configured to write one or more of the entries to the data cache responsive to receiving more than a threshold amount of non-merging committed store operations in the write combining buffer.

公开(公告)号：US11119767B1

公开(公告)日：2021-09-14

申请号：US16906396

申请日：2020-06-19

Applicant: Apple Inc.

Inventor： Brian R. Mestan ,  Gideon N. Levinsky ,  Michael L. Karm

IPC: G06F9/30 ,  G06F9/38 ,  G06F9/32 ,  G06F9/52

Abstract: In an embodiment, a processor comprises an atomic predictor circuit to predict whether or not an atomic operation will complete successfully. The prediction may be used when a subsequent load operation to the same memory location as the atomic operation is executed, to determine whether or not to forward store data from the atomic operation to the subsequent load operation. If the prediction is successful, the store data may be forwarded. If the prediction is unsuccessful, the store data may not be forwarded. In cases where an atomic operation has been failing (not successfully performing the store operation), the prediction may prevent the forwarding of the store data and thus may prevent a subsequent flush of the load.