On Mon, Mar 22, 2010 at 3:06 AM, Shi Weihua <[email protected]> wrote:
> Hi Garrett
>
> at 2010-3-13 15:13, Garrett Cooper wrote:
>> On Fri, Mar 12, 2010 at 10:27 PM, Garrett Cooper <[email protected]> wrote:
>>> On Fri, Mar 12, 2010 at 7:49 PM, Shi Weihua <[email protected]> wrote:
>>>> An error message occured when i tested proc01.
>>>> ---------
>>>> proc01      1  TFAIL  :  read failed: /proc/acpi/event: errno=EAGAIN(11): 
>>>> Resource temporarily unavailable
>>>> ---------
>>>>
>>>> If "service acpid start", this message will be turned to:
>>>> ---------
>>>> proc01      0  TINFO  :  /proc/acpi/event: known issue: errno=EBUSY(16): 
>>>> Device or resource busy
>>>> ---------
>>>>
>>>> I think we should to add a check of acpi daemon's status in 
>>>> testcases/kernel/fs/proc/proc01.c.
>>>
>>>    No, that's making things too complicated and won't scale as every
>>> distro has developed a different means for specifying the status for a
>>> service. Why not just add EAGAIN for that /proc node?
>>
>>     And FWIW the error condition is correct in my opinion if kernel
>> acpi hasn't been enabled, which was the case here.
>
> I think we should to give a warning message to LTP users.
> How do you think the following patch?
>
> diff --git a/testcases/kernel/fs/proc/proc01.c 
> b/testcases/kernel/fs/proc/proc01.c
> index 2d4a5fd..62d44b1 100644
> --- a/testcases/kernel/fs/proc/proc01.c
> +++ b/testcases/kernel/fs/proc/proc01.c
> @@ -101,6 +101,7 @@ const Mapping known_issues[] = {
>     {"read", "/proc/fs/nfsd/filehandle", EINVAL},
>     {"read", "/proc/fs/nfsd/.getfs", EINVAL},
>     {"read", "/proc/fs/nfsd/.getfd", EINVAL},
> +    {"read", "/proc/acpi/event", EAGAIN}, //EAGAIN will be returned if acpi 
> daemon disabled
>     {"", "", 0}
>  };

    Looks good minus the C++ style comment ;).
Thanks again!
-Garrett

------------------------------------------------------------------------------
Download Intel&#174; Parallel Studio Eval
Try the new software tools for yourself. Speed compiling, find bugs
proactively, and fine-tune applications for parallel performance.
See why Intel Parallel Studio got high marks during beta.
http://p.sf.net/sfu/intel-sw-dev
_______________________________________________
Ltp-list mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/ltp-list

Reply via email to