Hi folks,
I am still trying to hunt down the shm bugs I am experiencing under
high smp load. I reported kernel oopses during swapping.
But I now realized that errors occur without even swapping. My test
program shows inconsistent data on rereading the segment.
I am running the appended test program on an 8GB 8way with
2.3.41-pre2+tmp/smp-2.3.41-A1:
#ipctst 10 666400000 10 30 20&
#vmstat 5
procs memory swap io system cpu
r b w swpd free buff cache si so bi bo in cs us sy id
10 0 0 0 2876192 2320 9804 0 0 3 1 16 183 0 3 97
9 1 1 0 1017356 2320 9804 0 0 0 1 125 62010 0 90 9
9 1 0 0 660392 2320 9804 0 0 0 0 111 61688 1 86 13
5 5 0 0 1036448 2320 9804 0 0 0 0 117 87875 3 85 13
5 5 0 0 507900 2320 9804 0 0 0 0 115 106577 0 64 35
7 3 0 0 223152 2320 9804 0 0 0 0 108 86321 3 79 18
8 2 0 0 3488412 2320 9804 0 0 0 0 113 82936 3 84 14
5 5 0 0 3649372 2320 9804 0 0 0 0 109 102095 2 70 27
*p(0) != p(1458691561)
9 1 0 0 2909576 2320 9804 0 0 0 0 107 79827 2 83 15
8 2 0 0 3618740 2320 9804 0 0 0 0 122 101344 4 74 22
5 5 0 0 3999636 2320 9804 0 0 0 0 110 76780 2 77 21
8 2 0 0 3271184 2320 9804 0 0 0 0 114 62834 1 78 20
# ipcs -um
------ Shared Memory Status --------
segments allocated 9
pages allocated 1464264
pages resident 1245454
pages swapped 0
Swap performance: 0 attempts 0 successes
## vmstat 5
procs memory swap io system cpu
r b w swpd free buff cache si so bi bo in cs us sy id
8 2 0 0 3625700 2332 9808 0 0 1 0 14 6848 2 57 41
8 2 0 0 2554328 2332 9808 0 0 0 0 114 83047 2 83 15
10 0 0 0 3040068 2332 9808 0 0 0 1 111 79825 3 82 16
6 4 0 0 2791848 2332 9808 0 0 0 0 107 85891 3 84 13
*p(0) != p(1623759691)
7 3 0 0 2316364 2332 9808 0 0 0 0 108 99128 2 68 30
7 3 0 0 2316364 2332 9808 0 0 0 0 107 102336 2 67 31
7 3 0 0 2061568 2332 9808 0 0 0 0 108 90342 3 75 21
6 4 0 0 1793400 2332 9808 0 0 0 0 107 87724 2 82 16
*p(-115) != p(1735206285)
*p(116) != p(1737201524)
*p(73) != p(1738074185)
*p(116) != p(1738250356)
*p(-26) != p(1738717414)
*p(29) != p(1738942749)
*p(57) != p(1739057465)
*p(108) != p(1739266412)
*p(-116) != p(1739397516)
*p(-71) != p(1739581881)
*p(-7) != p(1739844089)
*p(84) != p(1740216916)
*p(-115) != p(1740450445)
*p(-89) != p(1740556967)
*p(-50) != p(1740716750)
*p(2) != p(1740929794)
*p(8) != p(1074946056)
*p(71) != p(1075204167)
*p(67) != p(1076236611)
*p(96) != p(1076355424)
*p(-113) != p(1076547983)
*p(-86) != p(1076658602)
*p(-1) != p(1077006847)
*p(53) != p(1077228085)
*p(-118) != p(1077576330)
*p(-58) != p(1077822150)
*p(-33) != p(1077924575)
*p(85) != p(1078408021)
*p(118) != p(1078543222)
*p(-122) != p(1078608774)
*p(-55) != p(1078883273)
*p(-19) != p(1079030765)
*p(16) != p(1079174160)
*p(44) != p(1079288876)
*p(108) != p(1079551084)
*p(-52) != p(1079944396)
*p(-44) != p(1079977172)
*p(-21) != p(1080071403)
*p(44) != p(1080337708)
*p(-116) != p(1080731020)
*p(-63) != p(1080948161)
*p(16) != p(1081271824)
*p(44) != p(1081386540)
*p(57) != p(1081439801)
*p(67) != p(1081480771)
*p(84) != p(1081550420)
*p(-118) != p(1081771658)
5 4 0 0 3458140 2344 9868 0 0 10 0 132 89677 4 78 18
6 3 0 0 2614876 2344 9868 0 0 0 6 111 97286 1 63 36
7 2 0 0 395560 2344 9868 0 0 0 0 106 80362 3 66 31
5 4 0 0 3948976 2344 9868 0 0 0 0 174 126050 2 66 33
8 1 0 0 3852856 2344 9868 0 0 0 0 108 87175 2 72 25
8 1 0 0 3560500 2344 9868 0 0 0 0 107 71922 3 77 20
6 3 0 0 3440356 2344 9868 0 0 0 0 132 105096 2 59 39
7 2 0 0 2969508 2344 9868 0 0 0 0 107 89670 2 65 33
9 0 0 0 2864612 2344 9868 0 0 0 0 108 83785 4 75 21
shmctl IPC_RMID: Invalid argument
9 0 0 0 35372 2344 9868 0 0 0 0 117 79154 4 82 14
7 2 0 0 546528 2344 9868 0 0 0 0 110 80181 2 75 23
7 2 0 0 3737452 2344 9868 0 0 0 0 112 77312 3 81 16
4 5 0 0 3811436 2344 9868 0 0 0 0 111 67178 2 75 23
#
(vmstat wraps at 4GB free, it's an old version)
Greetings
Christoph
--
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/
This archive was generated by hypermail 2b29 : Mon Jan 31 2000 - 21:00:14 EST