US5953538A
Expired / Lapsed
US
⏱️ 0.0 years remaining
Method and apparatus providing DMA transfers between devices coupled to different host bus bridges
Assignee: ImaginOn
•
Filing: 1996-11-12
•
Granted: 1999-09-14
•
Expires: 2016-11-12
Abstract
Abstract A multiprocessor having improved bus efficiency is shown to include a number of processing units and a memory coupled to a system bus. Also coupled to the system bus are at least one I/O bridge systems. A method for improving partial cache line writes from I/O devices to the central processing units incorporates cache coherency protocol and an enhanced invalidation scheme to ensure atomicity which minimizing the bus utilization. In addition, a method for allowing peer-to-peer communication between I/O devices coupled to the system bus via different I/O bridges includes a command and address space configuration that allows for communication without the involvement of any central processing device. Interrupt performance is improved through the storage of an interrupt data structure in main memory. The I/O bridges maintain the data structure, and when the CPU is available the interrupts can be accessed by a fast memory read; thereby reducing the requirement of I/O reads for interrupt handling.
Inventors
- 👤 Samuel Hammond Duncan
- 👤 Craig Durand Keefer
- 👤 Thomas Adam McLaughlin
- 👤 Paul Michael Guglielmi
Bibliographic Details
Application Number
—
Jurisdiction
US
Data Source
google_patents
Last Synchronized
2026-11-08 15:12
CPC Classifications
G; G06; G06F; G06F12/00; G06F12/02; G06F12/08; G06F12/0802; G06F12/0806; G06F12/0815; G06F12/0831; G06F13/00; G06F13/38; G06F13/40; G06F13/4004; G06F13/4027
Claims Analysis
Total Claims: 22
Claim 1
Independent Claim
190 words
A method for providing communication between a first external device coupled to a first external bus and a second external device coupled to a second external bus, said first external bus coupled to a multiprocessor computer system bus via a first bridge device, said second external bus coupled to said multi-processor system bus via a second bridge device, said method comprising the steps of: said first external device issuing a first access request associated with a portion of a memory space of said multi-processor computer system to said first bridge device, said portion of said memory space allocated to accesses for said second external device, said first access request including a first address of said portion of said memory space; said first bridge device remapping said first address to a second address before issuing a second access request, the second access request being on said computer system bus; and said second bridge device detecting said access request on said bus as an access request for said second external device, said second bridge device mapping said second address to said first address, and forwarding said request to said second external device.
Claim 2
Independent Claim
190 words
A method for providing communication between a first external device coupled to a first external bus and a second external device coupled to a second external bus, said first external bus coupled to a multiprocessor computer system bus via a first bridge device, said second external bus coupled to said multi-processor system bus via a second bridge device, said method comprising the steps of: said first external device issuing a first access request associated with a portion of a memory space of said multi-processor computer system to said first bridge device, said portion of said memory space allocated to accesses for said second external device, said first access request including a first address of said portion of said memory space; said first bridge device remapping said first address to a second address before issuing a second access request, the second access request being on said computer system bus; and said second bridge device detecting said access request on said bus as an access request for said second external device, said second bridge device mapping said second address to said first address, and forwarding said request to said second external device.
Claim 3
Dependent Claim
23 words
The method according to claim 1 wherein said access request is a direct access request to an address space of said memory space.
Claim 4
Dependent Claim
45 words
The method according to claim 1, wherein said first and second bridge devices each include at least two queues, for queuing input and output data, respectively, and wherein said method further comprises the step of reordering commands stored in said queues for prevention of deadlock.
Claim 5
Dependent Claim
39 words
The method according to claim 3, wherein said step of reordering commands includes the steps of allowing a subset of write operations received from said system bus to bypass a subset of read operations received from said system bus.
Claim 6
Dependent Claim
33 words
The method according to claim 3, wherein said step of reordering commands includes the steps of allowing DMA read return data to bypass posted data from a write initiated on said system bus.
Claim 7
Independent Claim
194 words
A method of providing a communication link between at least two I/O devices coupled to different I/O busses via at least two corresponding I/O bridges, each of the I/O bridges coupled to communicate over a system bus, wherein a plurality of other devices are also coupled to said system bus, and wherein each of said devices coupled to said bus has associated therewith an identification value, said method comprising the steps of: issuing, by a first one of said I/O devices, a first access request to said corresponding I/O bridge device, said corresponding I/O bridge device issuing a second access request, said second access request being over said system bus to a second one of said I/O devices, said second access request identifying said second one of said I/O devices via an address incorporating said identification value of said device; identifying, at the corresponding bridge device of said second I/O device, said access request as an access request for said second I/O device; and said second I/O bridge formatting said second access request and forwarding said formatted access request to said second I/O device without intervention from other devices coupled to said system bus.
Claim 8
Independent Claim
194 words
A method of providing a communication link between at least two I/O devices coupled to different I/O busses via at least two corresponding I/O bridges, each of the I/O bridges coupled to communicate over a system bus, wherein a plurality of other devices are also coupled to said system bus, and wherein each of said devices coupled to said bus has associated therewith an identification value, said method comprising the steps of: issuing, by a first one of said I/O devices, a first access request to said corresponding I/O bridge device, said corresponding I/O bridge device issuing a second access request, said second access request being over said system bus to a second one of said I/O devices, said second access request identifying said second one of said I/O devices via an address incorporating said identification value of said device; identifying, at the corresponding bridge device of said second I/O device, said access request as an access request for said second I/O device; and said second I/O bridge formatting said second access request and forwarding said formatted access request to said second I/O device without intervention from other devices coupled to said system bus.
Claim 9
Dependent Claim
32 words
The method according to claim 6, wherein said access request is a command indicating that a transfer is to be made between external devices coupled to different I/O busses deadlock avoidance to.
Claim 10
Dependent Claim
211 words
A multi-processor computer system comprising: a plurality of processors coupled via a system bus; at least two bridge devices, coupled to said system bus, each of said bridge devices for communicating with a corresponding external bus, and each respective external bus for coupling a different plurality of devices to an associated one of said at least two bridge devices; and means for transferring data from a first external device coupled to a first one of said respective external busses to a second external device couple to a second one of said respective external busses without intervention by said plurality of processors, wherein said means for transferring data further comprises: a memory address space, said memory address space apportioned into a plurality of subsets, where each one of a group of said plurality of subsets of memory address space corresponds to memory at one of said external devices; and each of said bridge devices further comprising: means for translating a transaction request received from external devices coupled to said bridge into said memory space addresses of said corresponding external device; means for monitoring said system bus to detect transactions issued on said system bus at memory space addresses corresponding to the memory space addresses of the external devices coupled to said bridge.
Claim 11
Dependent Claim
211 words
A multi-processor computer system comprising: a plurality of processors coupled via a system bus; at least two bridge devices, coupled to said system bus, each of said bridge devices for communicating with a corresponding external bus, and each respective external bus for coupling a different plurality of devices to an associated one of said at least two bridge devices; and means for transferring data from a first external device coupled to a first one of said respective external busses to a second external device couple to a second one of said respective external busses without intervention by said plurality of processors, wherein said means for transferring data further comprises: a memory address space, said memory address space apportioned into a plurality of subsets, where each one of a group of said plurality of subsets of memory address space corresponds to memory at one of said external devices; and each of said bridge devices further comprising: means for translating a transaction request received from external devices coupled to said bridge into said memory space addresses of said corresponding external device; means for monitoring said system bus to detect transactions issued on said system bus at memory space addresses corresponding to the memory space addresses of the external devices coupled to said bridge.
Claim 12
Dependent Claim
91 words
The apparatus of claim 8, wherein each of said bridge devices further comprises: a first queue for storing transactions to be forwarded from said external devices coupled to said bridge device to said system bus; a second queue for storing transactions to be forwarded from said system bus to said external devices coupled to said bridge; wherein each of said first and second queues are first in first out design queues; and each queue comprising control means for re-ordering the transactions in said queue to preclude deadlock at said associated bridge.
Claim 13
Dependent Claim
30 words
The apparatus of claim 9, wherein said control means of said first queue further comprises means for allowing a subset of write transactions to bypass a subset of read transactions.
Claim 14
Dependent Claim
31 words
The apparatus of claim 9, wherein said control means of said second queue further comprises means for allowing a subset of read return transactions to bypass a subset of write transactions.
Claim 15
Independent Claim
190 words
A method for providing communication between a first external device coupled to a first external bus and a second external device coupled to a second external bus, said first external bus coupled to a multi-processor computer system via a first bridge device, said second external bus coupled to said multi-processor system via a second bridge device, said method comprising the steps of: said first external device issuing an access request to a portion of a memory space of said multi-processor computer system, said portion of said memory space allocated to accesses for said second external device; and said second bridge device detecting said access request on said bus as an access request for said second external device, and forwarding said request to said second external device, wherein said first and second bride devices each include at least two queues, for queuing input and output data, respectively; and reordering commands stored in said queues for prevention of deadlock, wherein said step of reordering commands includes the steps of allowing a subset for write operations received from said system bus to bypass a subset of read operations received from said system bus.
Claim 16
Independent Claim
190 words
A method for providing communication between a first external device coupled to a first external bus and a second external device coupled to a second external bus, said first external bus coupled to a multi-processor computer system via a first bridge device, said second external bus coupled to said multi-processor system via a second bridge device, said method comprising the steps of: said first external device issuing an access request to a portion of a memory space of said multi-processor computer system, said portion of said memory space allocated to accesses for said second external device; and said second bridge device detecting said access request on said bus as an access request for said second external device, and forwarding said request to said second external device, wherein said first and second bride devices each include at least two queues, for queuing input and output data, respectively; and reordering commands stored in said queues for prevention of deadlock, wherein said step of reordering commands includes the steps of allowing a subset for write operations received from said system bus to bypass a subset of read operations received from said system bus.
Claim 17
Independent Claim
184 words
A method for providing communication between a first external device coupled to a first external bus and a second external device coupled to a second external bus, said first external bus coupled to a multi-processor computer system via a first bridge device, said second external bus coupled to said multi-processor system via a second bridge device, said method comprising the steps of: said first external device issuing an access request to a portion of a memory space of said multi-processor computer system, said portion of said memory space allocated to accesses for said second external device; and said second bridge device detecting said access request on said bus as an access request for said second external device, and forwarding said request to said second external device, wherein said first and second bride devices each include at least two queues, for queuing input and output data, respectively; and reordering commands stored in said queues for prevention of deadlock, wherein said step of reordering commands includes the steps of allowing DMA read return data to bypass posted data from a write initiated on said system bus.
Claim 18
Independent Claim
184 words
A method for providing communication between a first external device coupled to a first external bus and a second external device coupled to a second external bus, said first external bus coupled to a multi-processor computer system via a first bridge device, said second external bus coupled to said multi-processor system via a second bridge device, said method comprising the steps of: said first external device issuing an access request to a portion of a memory space of said multi-processor computer system, said portion of said memory space allocated to accesses for said second external device; and said second bridge device detecting said access request on said bus as an access request for said second external device, and forwarding said request to said second external device, wherein said first and second bride devices each include at least two queues, for queuing input and output data, respectively; and reordering commands stored in said queues for prevention of deadlock, wherein said step of reordering commands includes the steps of allowing DMA read return data to bypass posted data from a write initiated on said system bus.
Claim 19
Dependent Claim
207 words
A multi-processor computer system comprising: a plurality of processors coupled together via a system bus; at least two bridge devices, coupled to said system bus, each of said bridge devices for communicating with a corresponding external bus, and each respective external bus for coupling a different plurality of devices to said associated bridge device; and means for transferring data from a first external device coupled to a first one of said respective external busses to a second external device coupled to a second one of said respective external busses without intervention by said plurality of processors; each of said bridge devices further comprising: a first queue for storing transactions to be forearded from said external devices coupled to said bridge device to said system bus; a second queue for storing transactions to be forearded from said system bus to said external devices coupled to said bridge; wherein each of said first and second queues are first in first out design queues; and each queue comprising control means for re-ordering the transactions in said queue to precluded deadlock at said associated bridge, and wherein said control means of said first queue further comprises means for allowing a subset of write transactions to bypass a subset of read transactions.
Claim 20
Dependent Claim
207 words
A multi-processor computer system comprising: a plurality of processors coupled together via a system bus; at least two bridge devices, coupled to said system bus, each of said bridge devices for communicating with a corresponding external bus, and each respective external bus for coupling a different plurality of devices to said associated bridge device; and means for transferring data from a first external device coupled to a first one of said respective external busses to a second external device coupled to a second one of said respective external busses without intervention by said plurality of processors; each of said bridge devices further comprising: a first queue for storing transactions to be forearded from said external devices coupled to said bridge device to said system bus; a second queue for storing transactions to be forearded from said system bus to said external devices coupled to said bridge; wherein each of said first and second queues are first in first out design queues; and each queue comprising control means for re-ordering the transactions in said queue to precluded deadlock at said associated bridge, and wherein said control means of said first queue further comprises means for allowing a subset of write transactions to bypass a subset of read transactions.
Claim 21
Dependent Claim
208 words
A multi-processor computer system comprising: a plurality of processors coupled together via a system bus; at least two bridge devices, coupled to said system bus, each of said bridge devices for communicating with a corresponding external bus, and each respective external bus for coupling a different plurality of devices to said associated bridge device; and means for transferring data from a first external device coupled to a first one of said respective external busses to a second external device coupled to a second one of said respective external busses without intervention by said plurality of processors; wherein each of said bridge devices further comprises: a first queue for storing transactions to be forearded from said external devices coupled to said bridge device to said system bus; a second queue for storing transactions to be forearded from said system bus to said external devices coupled to said bridge; wherein each of said first and second queues are first in first out design queues; and each queue comprising control means for re-ordering the transactions in said queue to precluded deadlock at said associated bridge, wherein said control means of said second queue further comprises means for allowing a subset of read return transactions to bypass a subset of write transactions.
Claim 22
Dependent Claim
208 words
A multi-processor computer system comprising: a plurality of processors coupled together via a system bus; at least two bridge devices, coupled to said system bus, each of said bridge devices for communicating with a corresponding external bus, and each respective external bus for coupling a different plurality of devices to said associated bridge device; and means for transferring data from a first external device coupled to a first one of said respective external busses to a second external device coupled to a second one of said respective external busses without intervention by said plurality of processors; wherein each of said bridge devices further comprises: a first queue for storing transactions to be forearded from said external devices coupled to said bridge device to said system bus; a second queue for storing transactions to be forearded from said system bus to said external devices coupled to said bridge; wherein each of said first and second queues are first in first out design queues; and each queue comprising control means for re-ordering the transactions in said queue to precluded deadlock at said associated bridge, wherein said control means of said second queue further comprises means for allowing a subset of read return transactions to bypass a subset of write transactions.
References Cited (Backward Citations)
US5379378A
Cited
US5671371A
Cited
US5619726A
Cited
US5754802A
Cited
US5694556A
Cited
US5632021A
Cited
US5673399A
Cited
US5717876A
Cited
US5778197A
Cited
USPTO Annuity & Maintenance Windows
| Annuity Stage | Window Opens | Payment Deadline | Surcharge Period Until | Estimated Fee | Status |
|---|---|---|---|---|---|
| 3.5 Years (Year 4) | 2002-09-13 | 2003-03-15 | 2003-09-14 | $2000.00 | Window Closed / Lapsed |
| 7.5 Years (Year 8) | 2006-09-13 | 2007-03-15 | 2007-09-14 | $3760.00 | Window Closed / Lapsed |
| 11.5 Years (Year 12) | 2010-09-13 | 2011-03-15 | 2011-09-14 | $7700.00 | Window Closed / Lapsed |