On Wed, Jun 18, 2008 at 6:28 PM, Michael Kerrisk
<[EMAIL PROTECTED]> wrote:
> On Wed, Jun 18, 2008 at 6:13 PM, Jeff Moyer <[EMAIL PROTECTED]> wrote:
>> "Michael Kerrisk" <[EMAIL PROTECTED]> writes:
>>
>>> On Wed, Jun 18, 2008 at 4:40 PM, Jeff Moyer <[EMAIL PROTECTED]> wrote:
>>>> "Michael Kerrisk" <[EMAIL PROTECTED]> writes:
>>>>
>>>>> On Wed, Jun 18, 2008 at 3:51 PM, Jeff Moyer <[EMAIL PROTECTED]> wrote:
>>>>
>>>>>> Given that there are consumers of the libaio library, I will continue to
>>>>>> maintain it.  So, I don't see a need to tell folks not to use it.
>>>>>
>>>>> I meant to ask...  So yes there are consumers of the library, but I
>>>>> wonder if there are actually consumers (outside the library itself)
>>>>> of those 5 io_*(2) syscall wrappers.  Any idea on that?
>>>>
>>>> I'm not sure I understand your question.  Are you referring to these
>>>> calls?
>>>>
>>>> io_setup
>>>> io_getevents
>>>> io_submit
>>>> io_destroy
>>>> io_cancel
>>>
>>> Yes, I meant: are apps using these wrappers *directly*, rather than
>>> using the POSIX aio* interrfaces (which then invoke 5 aforementioned
>>> wrappers).
>>
>> Yes, database applications, in particular, are more fond of these
>> interfaces than the posix ones.
>
> Thanks for clearing that up.
>
> So in the end it sounds like I'll need to document the aberrant
> behavior of these 5 wrappers, and note that life is different if you
> use syscall(2).

And the text I'll add to each page is something like this:

   RETURN VALUE
       On  success,  io_cancel()  returns  0.   For  the failure
       return, see NOTES.

   NOTES
       Glibc does not provide a wrapper function for this system
       call.

       The  wrapper  provided in libaio for io_cancel() does not
       follow the usual C  library  conventions  for  indicating
       error:  on  error  it returns a negated error number (the
       negative of one of the values listed in ERRORS).  If  the
       system  call  is  invoked via syscall(2), then the return
       value follows the usual  conventions  for  indicating  an
       error:  -1,  with  errno  set  to a (positive) value that
       indicates the error.

Seem okay?

Cheers,

Michael

-- 
Michael Kerrisk
Linux man-pages maintainer; http://www.kernel.org/doc/man-pages/
man-pages online: http://www.kernel.org/doc/man-pages/online_pages.html
Found a bug? http://www.kernel.org/doc/man-pages/reporting_bugs.html

-------------------------------------------------------------------------
Check out the new SourceForge.net Marketplace.
It's the best place to buy or sell services for
just about anything Open Source.
http://sourceforge.net/services/buy/index.php
_______________________________________________
Ltp-list mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/ltp-list

Reply via email to