starpu.suppr 3.9 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149
  1. {
  2. don't care about cache hit stats
  3. Helgrind:Race
  4. fun:_starpu_msi_cache_hit
  5. ...
  6. }
  7. {
  8. don't care about cache miss stats
  9. Helgrind:Race
  10. fun:_starpu_msi_cache_miss
  11. ...
  12. }
  13. {
  14. known race, but not problematic in practice, see comment in _starpu_tag_clear
  15. Helgrind:LockOrder
  16. ...
  17. fun:_starpu_tag_free
  18. fun:_starpu_htbl_clear_tags
  19. ...
  20. fun:_starpu_tag_clear
  21. fun:starpu_shutdown
  22. ...
  23. }
  24. {
  25. There is actually no race on current_mode, because the mode can not change unexpectedly, until _starpu_notify_data_dependencies() is called further down. Valgrind can not know about such software rwlock.
  26. Helgrind:Race
  27. fun:_starpu_release_data_on_node
  28. fun:_starpu_push_task_output
  29. ...
  30. }
  31. {
  32. We do not care about races on profiling statistics
  33. Helgrind:Race
  34. fun:_starpu_worker_get_status
  35. fun:_starpu_worker_reset_profiling_info_with_lock
  36. ...
  37. }
  38. {
  39. This is racy, but since we'll always put the same values, this is not a problem.
  40. Helgrind:Race
  41. fun:_starpu_codelet_check_deprecated_fields
  42. ...
  43. }
  44. {
  45. This is racy, but we don't care, it's only a statistic
  46. Helgrind:Race
  47. fun:starpu_task_nsubmitted
  48. ...
  49. }
  50. {
  51. This is racy, but we don't care, it's only a statistic
  52. Helgrind:Race
  53. fun:starpu_task_nready
  54. ...
  55. }
  56. {
  57. fscanf error
  58. Memcheck:Cond
  59. ...
  60. fun:fscanf
  61. fun:_starpu_load_bus_performance_files
  62. ...
  63. }
  64. {
  65. mc / handle locking order. We always take handle before mc, except when we tidy memory, where we use trylock, but helgrind doesn't handle that :/ https://bugs.kde.org/show_bug.cgi?id=243232
  66. Helgrind:LockOrder
  67. ...
  68. fun:_starpu_spin_trylock
  69. fun:lock_all_subtree
  70. ...
  71. }
  72. {
  73. mc / handle locking order2
  74. Helgrind:LockOrder
  75. ...
  76. fun:_starpu_spin_lock
  77. fun:_starpu_memchunk_recently_used
  78. ...
  79. }
  80. {
  81. mc / handle locking order3
  82. Helgrind:LockOrder
  83. ...
  84. fun:_starpu_spin_lock
  85. fun:_starpu_request_mem_chunk_removal
  86. ...
  87. }
  88. {
  89. mc / handle locking order4
  90. Helgrind:LockOrder
  91. ...
  92. fun:_starpu_spin_lock
  93. fun:_starpu_allocate_interface
  94. ...
  95. }
  96. {
  97. TODO1: This is temporary. It perhaps does not pose problem because only the worker takes this mutex. Fixing this will require changing the scheduler interface, so that the schedulers themselves take the scheduling lock, not the caller. Filter it out for now, so we can see other races more easily.
  98. Helgrind:LockOrder
  99. fun:pthread_mutex_lock
  100. fun:starpu_pthread_mutex_lock
  101. fun:simple_worker_pull_task
  102. ...
  103. }
  104. {
  105. TODO1: This is temporary. It perhaps does not pose problem because only the worker takes this mutex. Fixing this will require changing the scheduler interface, so that the schedulers themselves take the scheduling lock, not the caller. Filter it out for now, so we can see other races more easily.
  106. Helgrind:LockOrder
  107. fun:pthread_mutex_lock
  108. fun:starpu_pthread_mutex_lock
  109. fun:_starpu_sched_component_worker_lock_scheduling
  110. fun:simple_worker_pull_task
  111. ...
  112. }
  113. {
  114. TODO2: This is temporary. It perhaps does not pose problem because only the worker takes this mutex. Fixing this will require changing the scheduler interface, so that the schedulers themselves take the scheduling lock, not the caller. Filter it out for now, so we can see other races more easily.
  115. Helgrind:LockOrder
  116. fun:pthread_mutex_lock
  117. fun:starpu_pthread_mutex_lock
  118. fun:_starpu_sched_component_lock_worker
  119. fun:simple_worker_pull_task
  120. ...
  121. }
  122. {
  123. TODO2: This is temporary. It perhaps does not pose problem because only the worker takes this mutex. Fixing this will require changing the scheduler interface, so that the schedulers themselves take the scheduling lock, not the caller. Filter it out for now, so we can see other races more easily.
  124. Helgrind:LockOrder
  125. fun:pthread_mutex_lock
  126. fun:starpu_pthread_mutex_lock
  127. fun:_starpu_sched_component_lock_worker
  128. fun:_starpu_sched_component_worker_lock_scheduling
  129. fun:simple_worker_pull_task
  130. ...
  131. }