October 18, 2006

Linux Test Project October Release Announcement

Anonymous Reader writes "The Linux Test Project test suite has been released. The latest version of the testsuite contains 2900+ tests for the Linux OS. Our web site also contains other information such as:
- A Linux test tools matrix
- Technical papers
- How To's on Linux testing
- Code coverage analysis tool.

Release Highlights:
Code Cleanups by Yi Xu, Jeff Burke, Darrick Wong, Mike Frysinger, Michael Reed

We encourage the community to post results to ltp-results@lists.sf.net,
and patches, new tests, or comments/questions to ltp-list@lists.sf.net

See ChangeLog Below
-A fix for Bug 23587 where the connectathon test failed on linux client with cifs mount to Windows2003 R2 server

-For Linux, 3 is a valide value for the scheduler, as found in the /linux/includ/sched.h file. For testing and invalid_policy, the invalid_policy value should not be 3

Ssem_post/5.-1.c According to its intention, #3, call sleep(1) and then alarm(1). Moving sleep before alarm causes the test to pass sem_post/8-1.c Although step 3 intended, The children lock the semaphore. * Make sure the two children are waiting." Without that caeratn piece of cde which explicity make children wat. the test will PASS and the children are waiting. Sem_unlink_1_1 and sem_unline_2_1 "Sem_unlink" is too long for the name of a semaphore for certain architectures
5-1.c, 8-1.c, 1-1.c, 2-1.c

-When calling pthread_attr_setscope, PHREAD_SCOPE_PROCESS is not supported by linux, change it to PTHREAD_SCOPE_SYSTEM and the test will pass
20-1.c, 21-1.c, 21-2.c

-Removed the ":" after 'v' in the getopts line because it is not using optarg.

-Added code to disable tests that will not run on kernels below 2.6.16
faccessat01.c, fchmodat01.c, fchownat01.c, fstatat01.c, futimesat01.c, linkat01.c, mknod01.c, openat01.c, readlinkat01.c, renameat01.c, symlinkat01.c, tee01.c, unlinkat01.c, vmsplice01.c

-When calling pthread_attr_setscope, PHREAD_SCOPE_PROCESS is not supported by linux, change it to PHTREAD-SCOPE_SYSTEM for testing and then the test passes.
22-1.c, 22-2.c

-A patch by David Stevens that fixes: 1) Removes signedness warning by changing the type of valsize from int to socklen_t 2) Correct but in ancillary data - sorce data is unit8_t, memcopy size is "sizeof(int)"; this results in garbage and TBROK on PPC64

-A fix for bugs 27174 and 27177. This fixes the problem of reading HugePages_Free
hugemmap01.c, hugemmap04.c

-Correcting error messages

-This patch by Darrick Wong fixes complier warnings and overflow problems related to the use of large number #defines on some architectures.

-When calling pthread_attr_setscope, PTHREAD_SCOPE_PROCESS is not supported by linux, change it to PTHREAD_SCOPE_SYSTEM for testing
15-1.c, 15-2.c

-A fix for bug #27618 that addresses two issues: 1) PAGE_SIZE which was normally 4096, but on the machine it found to 64k ! 2) The size of the file, offset passed. If those values, happen to be on the *Page boundary*, mmap would be happy. But in our case, it was not !

-Changing include to include "posixtest.h"

-A patch by Darrick Wong that adds a set of rudimentary IPMI tests to pounder. They check that the in-kernel IPMI driver can access the machine's BMC/SP (if there is one), query it for status and check for various IPMI 2.0 features.
default-tests.tar.gz, ipmitool, ipmitool"

Link: ltp.sf.net

Click Here!