Kernel 2.6.0test2 LSB test results (w/o attachments)

SUMMARY

Here are the results of LSB test suite run on the latest 2.6 kernel
(2.6.0test2):

            |    TOTALS   Succeeded   Warnings Unresolved Unsupported NotInUse|
Section     | Expect Actual       Failed      FIP    Uninitiated  Untested    |
____________|_________________________________________________________________|
            |                                                                 |
ANSI.os   F |  1244   1244   1093     0   12    0    19     0     0    4  116 |
ANSI.os   M |  1244   1244     86     0    0    0     0     0     0    0 1158 |
POSIX.os  F |  1600   1600   1309     3    5    2    23     0    72    7  179 |
POSIX.os  M |  1600   1600      0     0    0    0     0     0     0    0 1600 |
LI18NUX2K.L1|  1152   1021   1012     2    1    2     0     0     0    4    0 |
LSB.fhs     |   251    251    194     1    0   10     0     0    17    6   23 |
LSB.os    F |   785    785    683     0    2    1     4     0    24   15   56 |
LSB.os    M |   785    785      0     0    0    0     0     0     0    0  785 |
LSB.usersgroups|    94     94     84     0    3    0     0     0     0    1    6
 |
TOTAL       |  8755   8624   4461     6   23   15    46     0   113   37 3923 |
____________|_________________________________________________________________|


Which compare favorably with the results for kernel 2.5.69:

            |    TOTALS   Succeeded   Warnings Unresolved Unsupported NotInUse|
Section     | Expect Actual       Failed      FIP    Uninitiated  Untested    |
____________|_________________________________________________________________|
            |                                                                 |
ANSI.os   F |  1244   1244   1086     0   12    0    26     0     0    4  116 |
ANSI.os   M |  1244   1244     81     0    0    0     5     0     0    0 1158 |
POSIX.os  F |  1600   1600   1300    11    5    2    24     0    72    7  179 |
POSIX.os  M |  1600   1600      0     0    0    0     0     0     0    0 1600 |
LI18NUX2K.L1|  1152   1021   1011     3    1    2     0     0     0    4    0 |
LSB.fhs     |   251    251    194     1    0   10     0     0    17    6   23 |
LSB.os    F |   785    785    666     0    2    1     3    18    24   15   56 |
LSB.os    M |   785    785      0     0    0    0     0     0     0    0  785 |
LSB.usersgroups|    94     94     82     0    3    0     2     0     0    1    6
 |
TOTAL       |  8755   8624   4420    15   23   15    60    18   113   37 3923 |
____________|_________________________________________________________________|


But is still not up-to-par with the results from 2.4.20:

            |    TOTALS   Succeeded   Warnings Unresolved Unsupported NotInUse|
Section     | Expect Actual       Failed      FIP    Uninitiated  Untested    |
____________|_________________________________________________________________|
            |                                                                 |
ANSI.os   F |  1244   1244   1112     0   12    0     0     0     0    4  116 |
ANSI.os   M |  1244   1244     86     0    0    0     0     0     0    0 1158 |
POSIX.os  F |  1600   1600   1333     1    5    2     1     0    72    7  179 |
POSIX.os  M |  1600   1600      0     0    0    0     0     0     0    0 1600 |
LI18NUX2K.L1|  1152   1021   1012     2    1    2     0     0     0    4    0 |
LSB.fhs     |   251    251    194     1    0   10     0     0    17    6   23 |
LSB.os    F |   785    785    685     1    2    1     1     0    24   15   56 |
LSB.os    M |   785    785      0     0    0    0     0     0     0    0  785 |
LSB.usersgroups|    94     94     84     0    3    0     0     0     0    1    6
 |
TOTAL       |  8755   8624   4506     5   23   15     2     0   113   37 3923 |
____________|_________________________________________________________________|


OBSERVATIONS

* Kernel 2.6.0test2 is getting better.  There was only one additional failure
  (/tset/POSIX.os/procprim/alarm/T.alarm 5), but 44 additional unresolved
  tests; otherwise the results are identical.  The additional unresolved tests
  appear to be mostly related to loopback filesystem issues, though I've thus
  far been unable to analyze them much more deeply than that.

* I tried starting a discussion regarding the alarm() failure on the Linux
  Kernel Mailing List.  The first portion of the thread, including the test
  cases I wrote, etc, can be viewed here:
 
 http://groups.google.com/groups?hl=en&lr=&ie=UTF-8&oe=UTF-8&threadm=988b.1TH.17%40gated-at.bofh.it&rnum=1&prev=/groups%3Fq%3D2.5%2Bkernel%2Bregression%2Bin%2Balarm()%2Bsyscall%2Bbehaviour%26hl%3Den%26lr%3D%26ie%3DUTF-8%26oe%3DUTF-8%26selm%3D988b.1TH.17%2540gated-at.bofh.it%26rnum%3D1

  The thread continues here:
  http://groups.google.com/groups?hl=en&lr=&ie=UTF-8&oe=UTF-8&threadm=20030715015303.GA4845%40kvasir.austin.ibm.com.lucky.linux.kernel&rnum=4&prev=/groups%3Fq%3D2.5%2Bkernel%2Bregression%2Bin%2Balarm()%2Bsyscall%2Bbehaviour%26hl%3Den%26lr%3D%26ie%3DUTF-8%26oe%3DUTF-8%26selm%3D20030715015303.GA4845%2540kvasir.austin.ibm.com.lucky.linux.kernel%26rnum%3D4

  Basically, George Anzinger believes the failures to be the fault of libc and
  not the kernel, though the failures obviously did not occur in the 2.4 series
  of kernels


I'd be interested in hearing any ideas or feedback about these failures (loopback and alarm(), specifically).


Paul "Wes" Hofmann
IBM Linux Technology Center
[email protected]

PS: Another email should be available shortly with related attachments