[lsb-discuss] Please help check the failures.(LSB 3.1 runtime test)

Stew Benedict stewb at linux-foundation.org
Tue Jul 24 04:33:30 PDT 2007


On Tue, 24 Jul 2007, jie dong wrote:

> The test is based on LSB 3.1 spec.
> 
> There are some failures in runtime test.
> 
> 
> Take "fopen" for example, the output messages are as below:
> st_atime field on ./fopen-t.4.2 was not marked for update
> st_atime was 1185158235; should not be less than 1185158236
> st_ctime field on ./fopen-t.4.2 was not marked for update
> st_ctime was 1185158235; should not be less than 1185158236
> st_mtime field on ./fopen-t.4.2 was not marked for update
> st_mtime was 1185158235; should not be less than 1185158236
> 

Is the partition perhaps mounted with "noatime"? This will cause failures 
similar to these.

> 
> 
> As for "setvbuf",
> buffer buf1 contained wrong character (b) at position 0
> 
> 
> 
> "sort":
> Can't interpret blank character.
> 
> 

I don't think I've seen these. Perhaps someone else has run across 
failures with these.

> 
> 
> "ttyname":
> If the implementation detects that fildes is not a valid file
> descriptor: a call to ttyname_r(fildes, name, namesize) shall fail
> and return EBADF.
> Posix Ref: Component TTYNAME_R
> Assertion 9945-1:1996 4.7.2.4-2(C)
> The system does not support the PTHREAD_SCOPE_PROCESS scope, using
> PTHREAD_SCOPE_SYSTEM
> The system does not support the PTHREAD_SCOPE_PROCESS scope, using
> PTHREAD_SCOPE_SYSTEM
> The system does not support the PTHREAD_SCOPE_PROCESS scope, using
> PTHREAD_SCOPE_SYSTEM
> The system does not support the PTHREAD_SCOPE_PROCESS scope, using
> PTHREAD_SCOPE_SYSTEM
> The system does not support the PTHREAD_SCOPE_PROCESS scope, using
> PTHREAD_SCOPE_SYSTEM
> The system does not support the PTHREAD_SCOPE_PROCESS scope, using
> PTHREAD_SCOPE_SYSTEM
> The system does not support the PTHREAD_SCOPE_PROCESS scope, using
> PTHREAD_SCOPE_SYSTEM
> The system does not support the PTHREAD_SCOPE_PROCESS scope, using
> PTHREAD_SCOPE_SYSTEM
> ttyname_r, expected 9, got 25
> ttyname_r, expected 9, got 25
> ttyname_r, expected 9, got 25
> ttyname_r, expected 9, got 25
> ttyname_r, expected 9, got 25
> ttyname_r, expected 9, got 25
> ttyname_r, expected 9, got 25
> ttyname_r, expected 9, got 25
> 

This one may be:


http://sourceware.org/cgi-bin/cvsweb.cgi/libc/sysdeps/unix/sysv/linux/ttyname_r.c.diff?r1=1.17&r2=1.18&cvsroot=glibc
http://sourceware.org/cgi-bin/cvsweb.cgi/libc/sysdeps/unix/sysv/linux/ttyname.c.diff?r1=1.20&r2=1.21&cvsroot=glibc

We had to patch glibc on Mandriva for a similar issue (when I worked for 
them) - more info here:

http://www.linux-foundation.org/en/TestPilot31

As a general reporting hint, it would be good to report the actual test 
path and test number. I was able to find the code for the setvbuf and sort 
tests, but the full path would have made it faster. (tjreport should 
display this)

ie:

/tset/ANSI.os/streamio/setvbuf_X/T.setvbuf_X 3
/tset/LI18NUX2K.L1/utils/sort/T.sort 3

-- 
Stew Benedict
The Linux Foundation



More information about the lsb-discuss mailing list