On Sun, Jul 26, 2009 at 1:26 PM, Garrett Cooper<[email protected]> wrote:
> On Mon, Jul 20, 2009 at 8:21 AM, Garrett Cooper<[email protected]> wrote:
>> On Mon, Jul 13, 2009 at 10:32 AM, Garrett Cooper<[email protected]> wrote:
>>> On Mon, Jul 13, 2009 at 9:53 AM, Garrett Cooper<[email protected]> wrote:
>>>> On Mon, Jul 13, 2009 at 3:23 AM, Cyril Hrubis<[email protected]> wrote:
>>>>> Hi!
>>>>> I've seen quite a progress on ltp mailing list. What's the current state? 
>>>>>  I'm
>>>>> back here and would like to help.  (There is quite a lot of patches and 
>>>>> I'm too
>>>>> lazy to apply them all).
>>>>
>>>> Hi Cyril,
>>>>    The last set of patches is almost complete. I need to redo the
>>>> top-level Makefile one, one more time (I think) because I wasn't
>>>> copying over runtests / ChangeLog.
>>>>    runltp was also looking for ltp-pan in the wrong spot (the build
>>>> tree) instead of the install root, so that needs to be fixed too.
>>>
>>> The exact list of items that need to be fixed with some of my patches
>>> (I may or may not have submitted them previously) are listed below.
>>> Some are expected failures, because I completely turned off the mem
>>> section of the tree due to compile errors on my box (missing mm.h
>>> header).
>>
>> [...]
>>
>> Here's the current list on makefile_infra_rework:
>>
>
> [...]
>
>>
>> Again, the memory tests are still disabled because I don't have the
>> mm.h header on my system and the _16 / _64 testcases will get taken
>> care of soon...
>
> And here's the new list as of 1:21pm PST on makefile_infra_rework:

[...]

> Still only the default run, but it's shaping up nicely if you compare
> the amount of missing commands between this version and the previous
> one.
>
> The float Makefiles will be slightly entertaining as I need to study
> the inputs vs the outputs, but it shouldn't take more than 45 mins...
>
> I'm hoping that after this is all said and done I can resolve the
> libmm.h issue in a better way, using my Fedora 10 VM and a bit more C
> coding in the source files, effectively making the tests inert by
> using tst_res (CONF) liberally.
>
> Thanks,
> -Garrett
>

Awesome -- after fixing some more [minor] logic, this is all that
remains from the default set [4.5 hours later]:

App | Tag
mbind01 mbind01
move_pages.sh move_pages01
move_pages.sh move_pages02
move_pages.sh move_pages03
move_pages.sh move_pages04
move_pages.sh move_pages05
move_pages.sh move_pages06
move_pages.sh move_pages07
move_pages.sh move_pages08
move_pages.sh move_pages09
move_pages.sh move_pages10
move_pages.sh move_pages11
memcg_regression_test.sh memcg
filecapstest.sh Filecaps
run_capbounds.sh Cap_bounds
locktests FCNTL_LOCKTESTS
connector_test.sh Connectors
runpwtests.sh POWER_MANAGEMENT

- mbind / move_pages are there strictly because I don't have numa
support in my kernel, but I'm going to change them to properly
autoconf'ed sources soon...
- I'll just need to take a small mop to the rest of the cleanup items...

Thanks,
-Garrett

------------------------------------------------------------------------------
_______________________________________________
Ltp-list mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/ltp-list

Reply via email to