1 / 60

ECE 4100/6100 Advanced Computer Architecture Lecture 14 Multiprocessor and Memory Coherence

ECE 4100/6100 Advanced Computer Architecture Lecture 14 Multiprocessor and Memory Coherence. Prof. Hsien-Hsin Sean Lee School of Electrical and Computer Engineering Georgia Institute of Technology. Bus-based shared memory. P. P. P. $. $. $. Memory. Fully-connected shared memory

alijah
Télécharger la présentation

ECE 4100/6100 Advanced Computer Architecture Lecture 14 Multiprocessor and Memory Coherence

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. ECE 4100/6100Advanced Computer ArchitectureLecture 14 Multiprocessor and Memory Coherence Prof. Hsien-Hsin Sean Lee School of Electrical and Computer Engineering Georgia Institute of Technology

  2. Bus-based shared memory P P P $ $ $ Memory Fully-connected shared memory (Dancehall) Distributed shared memory P P P P P $ $ $ $ $ Memory Memory Interconnection Network Interconnection Network Memory Memory Memory Hierarchy in a Multiprocessor Shared cache P P P Cache Memory

  3. Cache Coherency • Closest cache level is private • Multiple copies of cache line can be present across different processor nodes • Local updates • Lead to incoherent state • Problem exhibits in both write-through and writeback caches • Bus-based  globally visible • Point-to-point interconnect  visible only to communicated processor nodes

  4. Rd? Rd? X= -100 X= -100 X= -100 Example (Writeback Cache) P P P Cache Cache Cache X= 505 Memory X= -100

  5. Rd? X= 505 X= -100 X= 505 Example (Write-through Cache) P P P Cache Cache Cache X= -100 X= 505 Memory X= -100

  6. Defining Coherence • An MP is coherent if the results of any execution of a program can be reconstructed by a hypothetical serial order Implicit definition of coherence • Write propagation • Writes are visible to other processes • Write serialization • All writes to the same location are seen in the same order by all processes (to “all” locations called write atomicity) • E.g., w1 followed by w2 seen by a read from P1, will be seen in the same order by all reads by other processors Pi

  7. A=1 B=2 T1 A=1 A=1 B=2 B=2 T2 A=1 A=1 B=2 B=2 T3 B=2 A=1 A=1 A=1 B=2 B=2 T3 B=2 B=2 A=1 A=1 See A’s update before B’s See B’s update before A’s Sounds Easy? A=0 B=0 P0 P1 P2 P3

  8. Bus Snooping based on Write-Through Cache • All the writes will be shown as a transaction on the shared bus to memory • Two protocols • Update-based Protocol • Invalidation-based Protocol

  9. Bus Snooping (Update-based Protocol on Write-Through cache) P P P Cache Cache Cache X= -100 X= 505 X= 505 Memory Bus transaction • Each processor’s cache controller constantly snoops on the bus • Update local copies upon snoop hit X= -100 X= 505 Bus snoop

  10. X= 505 Bus Snooping (Invalidation-based Protocol on Write-Through cache) P P P Load X Cache Cache Cache X= -100 X= 505 Memory Bus transaction • Each processor’s cache controller constantly snoops on the bus • Invalidate local copies upon snoop hit X= -100 X= 505 Bus snoop

  11. BusWr / --- PrRd / BusRd Processor-initiated Transaction Bus-snooper-initiated Transaction PrWr / BusWr A Simple Snoopy Coherence Protocol for a WT, No Write-Allocate Cache PrWr / BusWr PrRd / --- Valid Invalid Observed / Transaction

  12. How about Writeback Cache? • WB cache to reduce bandwidth requirement • The majority of local writes are hidden behind the processor nodes • How to snoop? • Write Ordering

  13. Cache Coherence Protocols for WB caches • A cache has an exclusive copy of a line if • It is the only cache having a valid copy • Memory may or may not have it • Modified (dirty) cache line • The cache having the line is the owner of the line, because it must supply the block

  14. update update Cache Coherence Protocol(Update-based Protocol on Writeback cache) P P P Store X Cache Cache Cache X= 505 X= 505 X= -100 X= -100 X= -100 X= 505 Memory Bus transaction • Update data for all processor nodes who share the same data • For a processor node keeps updating the memory location, a lot of traffic will be incurred

  15. update update Cache Coherence Protocol(Update-based Protocol on Writeback cache) P P P Store X Load X Cache Cache Cache X= 505 X= 333 X= 333 X= 505 X= 333 X= 505 Hit ! Memory Bus transaction • Update data for all processor nodes who share the same data • For a processor node keeps updating the memory location, a lot of traffic will be incurred

  16. invalidate invalidate Cache Coherence Protocol(Invalidation-based Protocol on Writeback cache) P P P Store X Cache Cache Cache X= -100 X= -100 X= -100 X= 505 Memory Bus transaction • Invalidate the data copies for the sharing processor nodes • Reduced traffic when a processor node keeps updating the same memory location

  17. Cache Coherence Protocol(Invalidation-based Protocol on Writeback cache) P P P Load X Cache Cache Cache X= 505 X= 505 Miss ! Snoop hit Memory Bus transaction Bus snoop • Invalidate the data copies for the sharing processor nodes • Reduced traffic when a processor node keeps updating the same memory location

  18. Cache Coherence Protocol(Invalidation-based Protocol on Writeback cache) Store X P P P Store X Store X Cache Cache Cache X= 444 X= 505 X= 333 X= 987 X= 505 Memory Bus transaction Bus snoop • Invalidate the data copies for the sharing processor nodes • Reduced traffic when a processor node keeps updating the same memory location

  19. MSI Writeback Invalidation Protocol • Modified • Dirty • Only this cache has a valid copy • Shared • Memory is consistent • One or more caches have a valid copy • Invalid • Writeback protocol: A cache line can be written multiple times before the memory is updated.

  20. MSI Writeback Invalidation Protocol • Two types of request from the processor • PrRd • PrWr • Three types of bustransactions post by cache controller • BusRd • PrRd misses the cache • Memory or another cache supplies the line • BusRd eXclusive (Read-to-own) • PrWr is issued to a line which is not in the Modified state • BusWB • Writeback due to replacement • Processor does not directly involve in initiating this operation

  21. PrRd / --- PrRd / --- PrWr / BusRdX PrRd / BusRd MSI Writeback Invalidation Protocol(Processor Request) PrWr / BusRdX PrWr / --- Modified Shared Invalid Processor-initiated

  22. BusRd / Flush BusRd / --- BusRdX / Flush BusRdX / --- MSI Writeback Invalidation Protocol(Bus Transaction) • Flush data on the bus • Both memory and requestor will grab the copy • The requestor get data by • Cache-to-cache transfer; or • Memory Modified Shared Invalid Bus-snooper-initiated

  23. BusRd / Flush BusRd / --- BusRdX / Flush BusRdX / --- BusRd / Flush MSI Writeback Invalidation Protocol(Bus transaction) Another possible implementation • Another possible, valid implementation • Anticipate no more reads from this processor • A performance concern • Save “invalidation” trip if the requesting cache writes the shared line later Modified Shared Invalid Bus-snooper-initiated

  24. MSI Writeback Invalidation Protocol PrWr / BusRdX PrWr / --- PrRd / --- BusRd / Flush BusRd / --- Modified Shared PrRd / --- BusRdX / Flush BusRdX / --- PrWr / BusRdX Invalid PrRd / BusRd Processor-initiated Bus-snooper-initiated

  25. X=10 S --- --- BusRd Memory S MSI Example P1 P2 P3 Cache Cache Cache Bus BusRd MEMORY X=10 Processor Action State in P2 State in P3 Bus Transaction Data Supplier State in P1 P1 reads X

  26. X=10 X=10 S S BusRd --- --- --- S BusRd BusRd Memory Memory S S MSI Example P1 P2 P3 Cache Cache Cache Bus MEMORY X=10 Processor Action State in P2 State in P3 Bus Transaction Data Supplier State in P1 P1 reads X P3 reads X

  27. --- X=10 I S BusRdX --- --- S --- BusRd BusRd Memory Memory S S MSI Example P1 P2 P3 Cache Cache Cache X=-25 S M X=10 Bus MEMORY X=10 Processor Action State in P2 State in P3 Bus Transaction Data Supplier State in P1 P1 reads X P3 reads X --- M BusRdX Memory I P3 writes X Does not come from memory if having “BusUpgrade”

  28. BusRd --- --- --- S BusRd BusRd Memory Memory S S --- --- M S BusRdX BusRd P3 Cache S I MSI Example P1 P2 P3 Cache Cache Cache S X=-25 --- I X=-25 M S Bus MEMORY X=-25 X=10 Processor Action State in P2 State in P3 Bus Transaction Data Supplier State in P1 P1 reads X P3 reads X Memory P3 writes X P1 reads X

  29. X=-25 S BusRd --- --- S --- BusRd BusRd Memory Memory S S --- --- S S S M BusRd BusRd BusRdX P3 Cache Memory S I S MSI Example P1 P2 P3 Cache Cache Cache X=-25 S X=-25 S M Bus MEMORY X=10 X=-25 Processor Action State in P2 State in P3 Bus Transaction Data Supplier State in P1 P1 reads X P3 reads X Memory P3 writes X P1 reads X P2 reads X

  30. MESI Writeback Invalidation Protocol • To reduce two types of unnecessary bus transactions • BusRdX that snoops and converts the block from S to M when only you are the sole owner of the block • BusRd that gets the line in S state when there is no sharers (that lead to the overhead above) • Introduce the Exclusive state • One can write to the copy without generating BusRdX • Illinois Protocol: Proposed by Pamarcos and Patel in 1984 • Employed in Intel, PowerPC, MIPS

  31. PrWr / --- PrRd, PrWr / --- PrRd / --- PrWr / BusRdX PrWr / BusRdX PrRd / BusRd (not-S) PrRd / --- PrRd / BusRd (S) MESI Writeback Invalidation ProtocolProcessor Request (Illinois Protocol) Exclusive Modified Invalid Shared S: Shared Signal Processor-initiated

  32. BusRd / Flush Or ---) BusRdX / --- BusRd / Flush BusRdX / Flush BusRd / Flush* BusRdX / Flush* MESI Writeback Invalidation ProtocolBus Transactions (Illinois Protocol) • Whenever possible, Illinois protocol performs $-to-$ transfer rather than having memory to supply the data • Use a Selection algorithm if there are multiple suppliers (Alternative: add an O state or force update memory) • Most of the MESI implementations simply write to memory Exclusive Modified Invalid Shared Bus-snooper-initiated Flush*: Flush for data supplier; no action for other sharers

  33. BusRdX / --- BusRd / Flush BusRdX / Flush BusRd / Flush* BusRdX / Flush* MESI Writeback Invalidation Protocol(Illinois Protocol) PrWr / --- PrRd, PrWr / --- PrRd / --- Exclusive Modified BusRd / Flush (or ---) PrWr / BusRdX PrWr / BusRdX PrRd / BusRd (not-S) Invalid Shared PrRd / --- S: Shared Signal Processor-initiated PrRd / BusRd (S) Bus-snooper-initiated Flush*: Flush for data supplier; no action for other sharers

  34. Add one additional state ─ Owner state Similar to Shared state The O state processor will be responsible for supplying data (copy in memory may be stale) Employed by Sun UltraSparc AMD Opteron In dual-core Opteron, cache-to-cache transfer is done through a system request interface (SRI) running at full CPU speed CPU0 CPU1 L2 L2 System Request Interface Crossbar Mem Controller Hyper- Transport MOESI Protocol

  35. Implication on Multi-Level Caches • How to guarantee coherence in a multi-level cache hierarchy • Snoop all cache levels? • Intel’s 8870 chipset has a “snoop filter” for quad-core • Maintaining inclusion property • Ensure data in the outer level must be present in the inner level • Only snoop the outermost level (e.g. L2) • L2 needs to know L1 has write hits • Use Write-Through cache • Use Write-back but maintain another “modified-but-stale” bit in L2

  36. Inclusion Property • Not so easy … • Replacement: Different bus observes different access activities, e.g. L2 may replace a line frequently accessed in L1 • Split L1 caches: Imagine all caches are direct-mapped. • Different cache line sizes

  37. Inclusion Property • Use specific cache configurations • E.g., DM L1 + bigger DM or set-associative L2 with the same cache line size • Explicitly propagate L2 action to L1 • L2 replacement will flush the corresponding L1 line • Observed BusRdX bus transaction will invalidate the corresponding L1 line • To avoid excess traffic, L2 maintains an Inclusion bit for filtering (to indicate in L1 or not)

  38. Presence bits, one for each node Modified bit Directory Directory-based Coherence Protocol • Snooping-based protocol • N transactions for an N-node MP • All caches need to watch every memory request from each processor • Not a scalable solution for maintaining coherence in large shared memory systems • Directory protocol • Directory-based control of who has what; • HW overheads to keep the directory (~ # lines * # processors) P P P P $ $ $ $ Interconnection Network Memory

  39. 0 1 0 0 0 1 0 1 1 0 0 0 0 0 0 0 0 1 1 0 0 1 0 0 1 modified bit for each cache block in memory Directory-based Coherence Protocol P P P P P $ $ $ $ $ Interconnection Network Memory C(k) C(k+1) C(k+j) 1 presence bit for each processor, each cache block in memory

  40. 0 0 0 1 0 1 - - - - 0 0 0 0 1 1 1 0 - - - - 1 modified bit for each cache block in memory Directory-based Coherence Protocol (Limited Dir) P0 P13 P1 P14 P15 $ $ $ $ $ Interconnection Network Memory 1 1 0 0 0 0 1 - - - - Presence encoding is NULL or not Encoded Present bits (lg2N), each cache line can reside in 2 processors in this example

  41. P P P P P P $ $ $ $ $ $ Memory Memory Memory Memory Memory Memory Directory Directory Directory Directory Directory Directory Interconnection Network Distributed Directory Coherence Protocol • Centralized directory is less scalable (contention) • Distributed shared memory (DSM) for a large MP system • Interconnection network is no longer a shared bus • Maintain cache coherence (CC-NUMA) • Each address has a “home”

  42. Directory Directory Distributed Directory Coherence Protocol • Stanford DASH (4 CPUs in each cluster, total 16 clusters) • Invalidation-based cache coherence • Directory keeps one of the 3 status of a cache block at its home node • Uncached • Shared (unmodified state) • Dirty P P P P $ $ $ $ Memory Memory Memory Memory Snoop bus Snoop bus Interconnection Network

  43. Directory Directory DASH Memory Hierarchy • Processor Level • Local Cluster Level • Home Cluster Level (address is at home) If dirty, needs to get it from remote node which owns it • Remote Cluster Level P P P P $ $ $ $ Memory Memory Memory Memory Snoop bus Snoop bus Interconnection Network

  44. Go to Home Node Directory Coherence Protocol: Read Miss P Miss Z (read) P P $ $ $ Z Z Home of Z Memory Memory Memory Z 0 0 1 1 1 Interconnection Network Data Z is shared (clean)

  45. Data Request Go to Home Node Respond with Owner Info Directory Coherence Protocol: Read Miss P Miss Z (read) P P $ $ $ Z Z Memory Memory Memory Z 1 0 0 1 1 0 1 Interconnection Network Data Z is Clean, Shared by 3 nodes Data Z is Dirty

  46. Invalidate ACK Go to Home Node Respond w/ sharers ACK Invalidate Directory Coherence Protocol: Write Miss P Miss Z (write) P P $ $ $ Z Z Z Memory Memory Memory Z 1 0 1 0 1 0 1 0 Interconnection Network Write Z can proceed in P0

  47. P1 P1 P2 P2 A=1; Flag = 1; A=1; B=1; while (Flag==0) {}; print A; print B; print A; Memory Consistency Issue • What do you expect for the following codes? Initial values A=0 B=0 Is it possible P2 prints A=0? Is it possible P2 prints A=0, B=1?

  48. Memory Consistency Model • Programmers anticipate certain memory ordering and program behavior • Become very complex When • Running shared-memory programs • A processor supports out-of-order execution • A memory consistency model specifies the legal ordering of memory events when several processors access the shared memory locations

  49. Memory Sequential Consistency (SC) [Leslie Lamport] • An MP is Sequentially Consistent if the result of any execution is the same as if the operations of all the processors were executed in some sequential order, and the operations of each individual processor appear in this sequence in the order specified by its program. • Two properties • Program ordering • Write atomicity (All writes to any location should appear to all processors in the same order) • Intuitive to programmers P P P

  50. P0 P0 P1 P1 P2 P2 P3 P3 A=1 A=1 A=2 A=2 T=A T=A Y=A Y=A U=A U=A Z=A Z=A SC Example T=1 U=2 Y=1 Z=2 T=1 U=2 Y=2 Z=1 Violating Sequential Consistency! (but possible in processor consistency model) Sequentially Consistent

More Related