[lfs-support] LFS-BOOK-7.0: Section 6.22: E2fsprogs - Test failure

Michael Builov mbuilov at gmail.com
Sun Mar 24 03:22:29 PDT 2013


Emerson Yesupatham <yemerson1976 <at> gmail.com> writes:

> 
> 
> Hi Team,
>  
> I am seeing a test fauilure in LFS-Boot-7.0, Section 6.22 (E2fsprogs-1.41.14).
>  
> Out of 107 tests, 1 test got failed.
>  
> I have checked the reference test logs present at
http://www.linuxfromscratch.org/lfs/build-logs/7.0/core2duo/test-logs/085-e2fsprogs,
all the 107 tests are passed here.
> 
>  
> I greped for "fail" in my test log and found the following:
>  
> >>d_loaddump: debugfs load/dump test: ok>>e_brel_bma: block relocation table
using the memory array implementation: skipped>>e_icount_normal: inode counting
abstraction optimized for storing inode counts: failed <<>>diff: expect: No such
file or directory<<>>e_icount_opt: inode counting abstraction optimized for
counting: ok
>  
> Is the failure realted to diff/expect tools? Any inputs please.
>  
> I have ignored this failure and proceeded till 6.41 (Gettext),  so far I have
not faced any issues so wondering whether the above failure will impact in
any of the future sections or I can safely proceed further.
> 
>  
> Thanks,Emerson
> 
> 

Hi, I had a similar problem when building clfs under chroot.
In my configuration two tests of e2fsprogs-1.42.7 have failed:

e_icount_opt: inode counting abstraction optimized for counting: failed
diff: expect: No such file or directory

e_icount_normal: inode counting structure optimized for low counts: failed
diff: expect: No such file or directory

> Is the failure realted to diff/expect tools? Any inputs please.

Check your environment variables and unset EXPECT variable.
I found that the test suite checks results of e_icount_opt test issuing

diff -u $EXPECT e_icount_opt.log

and prefers to use EXPECT from environment over internal one.
In my environment EXPECT was set as EXPECT=expect
That's why you see "diff: expect: No such file or directory"

I don't know how and who has set up this variable, but not me :)


-- Michael





More information about the lfs-support mailing list