1 / 14

Detecting Access Anomalies in Programs with Critical Sections

Detecting Access Anomalies in Programs with Critical Sections . Anne Dinning Edith Schonberg 2010. 4. 3 OSLab. Access Anomaly. At least one access is a write The two tasks do not synchronize between the accesses. i = 2. i = 1. doall i = 1, N … X := i + 7 endall Y := X * 3.

edison
Télécharger la présentation

Detecting Access Anomalies in Programs with Critical Sections

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. Detecting Access Anomalies in Programs with Critical Sections Anne Dinning Edith Schonberg 2010. 4. 3 OSLab

  2. Access Anomaly • At least one access is a write • The two tasks do not synchronize between the accesses i = 2 i = 1 doall i = 1, N … X := i + 7 endall Y := X * 3 X := i + 7 X := i + 7 8 = 1 + 7 9 = 2 + 7 Y := X * 3 i = 1: 24 = 8 * 3 i = 2: 27 = 9 * 3 nondeterministic result

  3. task-create (3) X := Y lock (L) lock (L) lock (L) Y := Y + i Y := Y + i Y := Y + i unlock (L) unlock (L) unlock (L) X := i task-terminate ( ) POEG (Partial Order execution Graph) Vertices - parallel operations (task create/terminate, lock/unlock) - an instruction block Edges - execution ordering of parallel operations

  4. On-the-fly detection • Advantage • diagnostic precision • space efficient (cf. traced-based methods) • Restriction • SISE property (single input, single execution) • A single execution instance is sufficient to determine the existence of an anomaly for a given input.

  5. On-the-fly detection using the POEG • Information • Which blocks are ordered (tag) • What variables are accessed by each block • Access history • Readers (X), Writers (X) (a variable X) • Only the most recent accesses Readers (X) Writers (X) The maximum parallelism Of the program At most of a single tag • guaranteed to detect at least one anomaly

  6. procedure Check-Read(X) for all a in Writers(X) do if not Ordered(current-tag, a) then report Access Anomaly endfor end procedure procedure Check-Write(X) for all a in Readers(X) U Writers(X) do if not Ordered(current-tag, a) then report Access Anomaly endfor end procedure Check for Read and Write Events procedure Subtract-Read(X) for all a in Readers(X) do if Ordered(current-tag, a) then delete a from Readers(X) endfor add current-tag to Readers(X) end procedure procedure Subtract-Write(X) Readers(X) = Ø Writers(X) = {current-tag} end procedure Subtraction for Read and Write Events

  7. Problems with POEG representation task-create (3) X := Y lock (L) lock (L) lock (L) Y := Y + i Y := Y + i Y := Y + i unlock (L) unlock (L) unlock (L) X := i task-terminate ( )

  8. r L L w Properties of unlock-lock edges • Accesses inside critical sections protected by the same lock L never appear as access anomalies • Accesses that occur before and after critical sections, as well as accesses within critical sections, are ordered by these edges. L L r w

  9. Lock Covers task-create (3) X := Y • Lock Cover (LC) • the set of the locks (a block B) lock (L) lock (L) lock (L) Y := L Y + i Y := L Y + i Y :=L Y + i unlock (L) unlock (L) unlock (L) X := i task-terminate ()

  10. Anomaly detection using Lock Covers • Access History Readers (X) Writers (X) The maximum parallelism of the program At most of a single tag Readers (X) Writers (X) The maximum parallelism of the program At most of a single tag CS-Readers (X) CS-Writers (X) <tag, lock cover> <tag, lock cover>

  11. Detection using LC task-create (3) <tag> <tag, lc> 1 X := Y 1 lock (L) 2 3 1 lock (L) lock (L) 5 4 1 3 4 7 6 5 3 2 5 6 7 1 3 Y := L Y + i Y := L Y + i Y :=L Y + i 5 1 7 8 3 Access History (Y) unlock (L) unlock (L) unlock (L) 8 1-3 1-5 5-8 7-8 X := i X := i Report Access Anomaly task-terminate ()

  12. B1 B3 X := L1 B2 B4 X := L1, L2 X := L2 X := L2 Nested lock Hidden Anomaly

  13. Non-SISE S := 1 doalli = 1, 2 A[i] := … lock(L) j := S S := S + 1 unlock(L) if (j = 1) then A[1] := … endall A[1] := A[2] := lock (L) lock (L) j := S S := S + 1 j := S S := S + 1 unlock (L) unlock (L) A[1] := Static analysis approach, based on program slicing

  14. False Anomalies A[j] := … lock(L) enqueue(j) Unlock(L) lock(L) k := dequeue(j) Unlock(L) … := A[k] A[j] := lock (L) lock (L) k := dequeue(j) enqueue(j) unlock (L) unlock (L) := A[k] * j and k are private variables

More Related