Is Parallel Programming Hard, And, If So, What Can You Do About It?

More documents

Recommendations

Info

$TeX op Mac OS X, met teTeX en TeXShop - Nluug$

230 APPENDIX D. READ-COPY UPDATE IMPLEMENTATIONS1 static void2 record_gp_stall_check_time(struct rcu_state *rsp)3 {4 rsp->gp_start = jiffies;5 rsp->jiffies_stall =6 jiffies + RCU_SECONDS_TILL_STALL_CHECK;7 }Figure D.54: record gp stall check time() Codeson for saving the current grace-period number is tocorrectly handle race conditions involving the currentgrace period ending concurrently with the nextinvocation of force_quiescent_state().As noted earlier, lines 53-58 handle the secondand subsequent invocations of force_quiescent_state() in CONFIG_NO_HZ kernels, and all invocationsin non-CONFIG_NO_HZ kernels. Lines 54and 58 invoke rcu_process_dyntick(), which cyclesthrough the CPUs that have still not passedthrough a quiescent state, invoking rcu_implicit_dynticks_qs() on them, which in turn checks to seeif any of these CPUs have passed through dyntickidlestate (if CONFIG_NO_HZ is enabled), checks tosee if we are waiting on any offline CPUs, and finallysends a reschedule IPI to any remaining CPUsnot in the first two groups.1 static void2 check_cpu_stall(struct rcu_state *rsp,3 struct rcu_data *rdp)4 {5 long delta;6 struct rcu_node *rnp;78 delta = jiffies - rsp->jiffies_stall;9 rnp = rdp->mynode;10 if ((rnp->qsmask & rdp->grpmask) && delta >= 0) {11 print_cpu_stall(rsp);12 } else if (rsp->gpnum != rsp->completed &&13 delta >= RCU_STALL_RAT_DELAY) {14 print_other_cpu_stall(rsp);15 }16 }Figure D.55: check cpu stall() Code1 static void print_cpu_stall(struct rcu_state *rsp)2 {3 unsigned long flags;4 struct rcu_node *rnp = rcu_get_root(rsp);56 printk(KERN_ERR7 "INFO: RCU detected CPU %d stall "8 "(t=%lu jiffies)\n",9 smp_processor_id(),10 jiffies - rsp->gp_start);11 dump_stack();12 spin_lock_irqsave(&rnp->lock, flags);13 if ((long)(jiffies - rsp->jiffies_stall) >= 0)14 rsp->jiffies_stall =15 jiffies + RCU_SECONDS_TILL_STALL_RECHECK;16 spin_unlock_irqrestore(&rnp->lock, flags);17 set_need_resched();18 }D.3.9 CPU-Stall DetectionRCU checks for stalled CPUs when the CONFIG_RCU_CPU_STALL_DETECTOR kernel parameter is selected.“Stalled CPUs” are those spinning in thekernel with preemption disabled, which degrades responsetime. These checks are implemented viathe record_gp_stall_check_time(), check_cpu_stall(), print_cpu_stall(), and print_other_cpu_stall() functions, each of which is describedbelow. All of these functions are no-ops when theCONFIG_RCU_CPU_STALL_DETECTOR kernel parameteris not selected.Figure D.54 shows the code for record_gp_stall_check_time(). Line 4 records the currenttime (of the start of the grace period) in jiffies, andlines 5-6 record the time at which CPU stalls shouldbe checked for, should the grace period run on thatlong.FigureD.55showsthecodeforcheck_cpu_stall,which checks to see if the grace period has stretchedon too long, invoking either print_cpu_stall() orprint_other_cpu_stall()inordertoprintaCPUstallwarning message if so.Line 8 computes the number of jiffies since thetime at which stall warnings should be printed,which will be negative if it is not yet time to printFigure D.56: print cpu stall() Codewarnings. Line 9 obtains a pointer to the leafrcu_node structure corresponding to the currentCPU, and line 10 checks to see if the current CPUhas not yet passed through a quiescent state andif the grace period has extended too long (in otherwords, if the current CPU is stalled), with line 11invoking print_cpu_stall() if so.Otherwise, lines 12-13 check to see if the graceperiod is still in effect and if it has extended a coupleof jiffies past the CPU-stall warning duration, withline 14 invoking print_other_cpu_stall() if so.Quick Quiz D.53: Why wait the extra couplejiffies on lines 12-13 in Figure D.55?Figure D.56 shows the code for print_cpu_stall().Line 6-11 prints a console message and dumps thecurrent CPU’s stack, while lines 12-17 compute thetimetothenextCPUstallwarning, shouldthegraceperiod stretch on that much additional time.Quick Quiz D.54: What prevents the grace periodfrom ending before the stall warning is printedin Figure D.56?Figure D.57 shows the code for print_other_
D.4. PREEMPTABLE RCU 2311 static void print_other_cpu_stall(struct rcu_state *rsp)2 {3 int cpu;4 long delta;5 unsigned long flags;6 struct rcu_node *rnp = rcu_get_root(rsp);7 struct rcu_node *rnp_cur;8 struct rcu_node *rnp_end;910 rnp_cur = rsp->level[NUM_RCU_LVLS - 1];11 rnp_end = &rsp->node[NUM_RCU_NODES];12 spin_lock_irqsave(&rnp->lock, flags);13 delta = jiffies - rsp->jiffies_stall;14 if (delta < RCU_STALL_RAT_DELAY ||15 rsp->gpnum == rsp->completed) {16 spin_unlock_irqrestore(&rnp->lock, flags);17 return;18 }19 rsp->jiffies_stall = jiffies +20 RCU_SECONDS_TILL_STALL_RECHECK;21 spin_unlock_irqrestore(&rnp->lock, flags);22 printk(KERN_ERR "INFO: RCU detected CPU stalls:");23 for (; rnp_cur < rnp_end; rnp_cur++) {24 if (rnp_cur->qsmask == 0)25 continue;26 cpu = 0;27 for (; cpu grphi - rnp_cur->grplo; cpu++)28 if (rnp_cur->qsmask & (1UL grplo + cpu);30 }31 printk(" (detected by %d, t=%ld jiffies)\n",32 smp_processor_id(),33 (long)(jiffies - rsp->gp_start));34 force_quiescent_state(rsp, 0);35 }Figure D.57: print other cpu stall() Codecpu_stall(), which prints out stall warnings forCPUs other than the currently running CPU.Lines 10 and 11 pick up references to the firstleaf rcu_node structure and one past the last leafrcu_node structure, respectively. Line 12 acquiresthe root rcu_node structure’s lock, and also disablesinterrupts. Line 13 calculates the how long ago theCPU-stall warning time occurred(which will be negativeif it has not yet occurred), and lines 14 and15 check to see if the CPU-stall warning time haspassed and if the grace period has not yet ended,with line 16 releasing the lock (and re-enabling interrupts)and line 17 returning if so.Quick Quiz D.55: Why does print_other_cpu_stall() in Figure D.57 need to check for thegrace period ending when print_cpu_stall() didnot?Otherwise, lines 19 and 20 compute the next timethat CPU stall warnings should be printed (if thegrace period extends that long) and line 21 releasesthe lock and re-enables interrupts. Lines 23-33 printa list of the stalled CPUs, and, finally, line 34 invokesforce_quiescent_state() in order to nudgethe offending CPUs into passing through a quiescentstate.D.3.10 Possible Flaws and ChangesThe biggest possible issue with Hierarchical RCUput forward as of this writing is the fact thatforce_quiescent_state() involves a potentialwalk through all CPUs’ rcu_data structures. Ona machine with thousands of CPUs, this could potentiallyrepresentanexcessiveimpactonschedulinglatency, given that this scan is conducted with interruptsdisabled.Should this become a problem in real life, one fixis to maintain separate force_quiescent_state()sequencing on a per-leaf-rcu_node basis as well asthe current per-rcu_state ->signaled state variable.This would allow incremental forcing of quiescentstates on a per-leaf-rcu_node basis, greatlyreducing the worst-case degradation of schedulinglatency.In the meantime, those caring deeply aboutscheduling latency can limit the number of CPUsin the system or use the preemptable RCU implementation.D.4 Preemptable RCUThe preemptable RCU implementation is unusual inthat it permits read-side critical sections to be preemptedand to be blocked waiting for locks. However,it does not handle general blocking (for example,via the wait_event() primitive): if you needthat, you should instead use SRCU, which is describedin Appendix D.1. In contrast to SRCU, preemptibleRCU only permits blocking within primitivesthat are both subject to priority inheritanceand non-blocking in a non-CONFIG_PREEMPT kernel.This ability to acquire blocking locks and to be preemptedwithin RCU read-side critical sections is requiredfor the aggressive real-time capabilities providedby Ingo Molnar’s -rt patchset. However, theinitial preemptible RCU implementation [McK05c]had some limitations, including:1. Its read-side primitives cannot be calledfrom within non-maskable interrupt (NMI) orsystems-management interrupt handlers.2. Itsread-sideprimitives usebothatomic instructionsand memory barriers, both of which haveexcessive overhead.3. It does no priority boosting of RCU read-sidecritical sections [McK07d].The new preemptible RCU implementation thataccepted into the 2.6.26 Linux kernel removes these
Page 1 and 2:
Is Parallel Programming Hard, And,
Page 3 and 4:
Contents1 Introduction 11.1 Histori
Page 5 and 6:
CONTENTSv6 Locking 676.1 Staying Al
Page 7 and 8:
CONTENTSviiB Synchronization Primit
Page 9 and 10:
CONTENTSixE.7.1 Introduction to Pre
Page 11 and 12:
PrefaceThe purpose of this book is
Page 13 and 14:
Chapter 1IntroductionParallel progr
Page 15 and 16:
1.2. PARALLEL PROGRAMMING GOALS 3CP
Page 17 and 18:
1.3. ALTERNATIVES TO PARALLEL PROGR
Page 19 and 20:
1.4. WHAT MAKES PARALLEL PROGRAMMIN
Page 21 and 22:
1.5. GUIDE TO THIS BOOK 9other hand
Page 23 and 24:
Chapter 2Hardware and its HabitsMos
Page 25:
2.1. OVERVIEW 13Therefore, as shown
Page 28 and 29:
16 CHAPTER 2. HARDWARE AND ITS HABI
Page 30 and 31:
18 CHAPTER 2. HARDWARE AND ITS HABI
Page 32 and 33:
20 CHAPTER 3. TOOLS OF THE TRADE1 p
Page 34 and 35:
22 CHAPTER 3. TOOLS OF THE TRADE1 p
Page 36 and 37:
24 CHAPTER 3. TOOLS OF THE TRADE1.1
Page 38 and 39:
26 CHAPTER 3. TOOLS OF THE TRADEQui
Page 40 and 41:
28 CHAPTER 3. TOOLS OF THE TRADE
Page 42 and 43:
30 CHAPTER 4. COUNTING1 atomic_t co
Page 44 and 45:
32 CHAPTER 4. COUNTING4.2.3 Eventua
Page 46 and 47:
34 CHAPTER 4. COUNTINGvanish when t
Page 48 and 49:
36 CHAPTER 4. COUNTINGper-thread va
Page 50 and 51:
38 CHAPTER 4. COUNTING1 unsigned lo
Page 52 and 53:
Page 54 and 55:
42 CHAPTER 4. COUNTING1 #define THE
Page 56 and 57:
Page 58 and 59:
46 CHAPTER 4. COUNTINGReadsAlgorith
Page 60 and 61:
48 CHAPTER 5. PARTITIONING AND SYNC
Page 62 and 63:
Page 64 and 65:
Page 66 and 67:
Page 68 and 69:
Page 70 and 71:
Page 72 and 73:
Page 74 and 75:
Page 76 and 77:
Page 78 and 79:
Page 80 and 81:
68 CHAPTER 6. LOCKING1 int delete(i
Page 82 and 83:
70 CHAPTER 7. DATA OWNERSHIP
Page 84 and 85:
72 CHAPTER 8. DEFERRED PROCESSINGfo
Page 86 and 87:
74 CHAPTER 8. DEFERRED PROCESSINGth
Page 88 and 89:
76 CHAPTER 8. DEFERRED PROCESSING
Page 90 and 91:
78 CHAPTER 8. DEFERRED PROCESSINGfi
Page 92 and 93:
80 CHAPTER 8. DEFERRED PROCESSINGti
Page 94 and 95:
82 CHAPTER 8. DEFERRED PROCESSINGNo
Page 96 and 97:
84 CHAPTER 8. DEFERRED PROCESSING12
Page 98 and 99:
Page 100 and 101:
88 CHAPTER 8. DEFERRED PROCESSINGvo
Page 102 and 103:
90 CHAPTER 8. DEFERRED PROCESSINGLi
Page 104 and 105:
92 CHAPTER 8. DEFERRED PROCESSINGpe
Page 106 and 107:
94 CHAPTER 8. DEFERRED PROCESSINGCa
Page 108 and 109:
96 CHAPTER 8. DEFERRED PROCESSINGTh
Page 110 and 111:
Page 112 and 113:
Page 114 and 115:
Page 116 and 117:
104 CHAPTER 8. DEFERRED PROCESSINGs
Page 118 and 119:
106 CHAPTER 8. DEFERRED PROCESSINGo
Page 120 and 121:
108 CHAPTER 8. DEFERRED PROCESSING
Page 122 and 123:
110 CHAPTER 9. APPLYING RCU1 struct
Page 124 and 125:
112 CHAPTER 9. APPLYING RCU
Page 126 and 127:
114 CHAPTER 10. VALIDATION: DEBUGGI
Page 128 and 129:
116 CHAPTER 11. DATA STRUCTURES
Page 130 and 131:
118 CHAPTER 12. ADVANCED SYNCHRONIZ
Page 132 and 133:
Page 134 and 135:
Page 136 and 137:
Page 138 and 139:
Page 140 and 141:
Page 142 and 143:
Page 144 and 145:
Page 146 and 147:
Page 148 and 149:
Page 150 and 151:
138 CHAPTER 13. EASE OF USEFigure 1
Page 152 and 153:
140 CHAPTER 13. EASE OF USE
Page 154 and 155:
142 CHAPTER 14. TIME MANAGEMENT
Page 156 and 157:
144 CHAPTER 15. CONFLICTING VISIONS
Page 158 and 159:
Page 160 and 161:
Page 162 and 163:
Page 164 and 165:
Page 166 and 167:
154 APPENDIX A. IMPORTANT QUESTIONS
Page 168 and 169:
156 APPENDIX A. IMPORTANT QUESTIONS
Page 170 and 171:
158 APPENDIX B. SYNCHRONIZATION PRI
Page 172 and 173:
160 APPENDIX B. SYNCHRONIZATION PRI
Page 174 and 175:
162 APPENDIX C. WHY MEMORY BARRIERS
Page 176 and 177:
Page 178 and 179:
Page 180 and 181:
Page 182 and 183:
Page 184 and 185:
Page 186 and 187:
Page 188 and 189:
Page 190 and 191:
Page 192 and 193: 180 APPENDIX C. WHY MEMORY BARRIERS
Page 194 and 195: 182 APPENDIX C. WHY MEMORY BARRIERS
Page 196 and 197: 184 APPENDIX D. READ-COPY UPDATE IM
Page 256 and 257: 244 APPENDIX E. FORMAL VERIFICATION
Page 284 and 285: 272 APPENDIX F. ANSWERS TO QUICK QU
Page 292 and 293:
280 APPENDIX F. ANSWERS TO QUICK QU
Page 294 and 295:
Page 296 and 297:
Page 298 and 299:
Page 300 and 301:
Page 302 and 303:
Page 304 and 305:
Page 306 and 307:
Page 308 and 309:
Page 310 and 311:
Page 312 and 313:
Page 314 and 315:
Page 316 and 317:
Page 318 and 319:
Page 320 and 321:
Page 322 and 323:
Page 324 and 325:
Page 326 and 327:
Page 328 and 329:
Page 330 and 331:
Page 332 and 333:
Page 334 and 335:
Page 336 and 337:
Page 338 and 339:
Page 340 and 341:
Page 342 and 343:
330 APPENDIX G. GLOSSARY(2) A physi
Page 344 and 345:
332 APPENDIX G. GLOSSARYnear by. Th
Page 346 and 347:
334 APPENDIX G. GLOSSARY
Page 348 and 349:
336 BIBLIOGRAPHY[But97]USA, March 2
Page 350 and 351:
338 BIBLIOGRAPHY[HMB06][Hol03][HP95
Page 352 and 353:
340 BIBLIOGRAPHY[McK06] Paul E. McK
Page 354 and 355:
342 BIBLIOGRAPHYtor. Software - Pra
Page 356 and 357:
344 BIBLIOGRAPHY[UoC08][VGS08]Berke
Page 358:
346 APPENDIX H. CREDITSH.4 Original
show all

Is Parallel Programming Hard, And, If So, What Can You Do About It?

Create successful ePaper yourself

Delete template?

Save as template?