Re: Where is thr_getscheduler

On Wed, 2 Aug 2006, Steve Kargl wrote:

> On Wed, Aug 02, 2006 at 09:42:55AM -0500, Matthew D. Fuller wrote:
>> On Tue, Aug 01, 2006 at 02:38:03PM -0700 I heard the voice of
>> Steve Kargl, and lo! it spake thus:
>>>
>>> If UPDATING had a proper notice,
>>
>> If UPDATING had a notice every time a {library,program} in 7 did
>> something that wasn't supported by 6 (for various values of 7 and 6),
>> it would be a very, very long and very boring file.
>>
>
> Sigh. We go through this every time someone bumps libc's
> version number without bumping the version numbers of
> all other libraries.  There is significant difference
> bewteen changing libgpib.so version number and changing
> libc.so version number.  Changeing libc's version number
> should have been noted in UPDATING.
>
> The version number of libthr should have been bumped
> when David Xu committed his change.

No, it has nothing to do with that change.  There were
other changes to libc that would necessitate bumping
libpthread and libthr version numbers.  If you take a
libc.so.6 from 6-stable, you'll note that it wouldn't
have worked in -current (under the same circumstances)
before the thr_getscheduler() changes.  If you look at
the -current archives you'll see this problem even with
libpthread which doesn't use thr_getscheduler().

> Last time I checked there were several integers between 2 and
> INT_MAX.  Is there some sort of shortage of integers at
> freebsd.org that prevents bumping libthr.so.2 to libthr.so.3?

Hold your horses.  The libraries are going to be bumped,
and this issue is going to be addressed in the near future.
This is -current, so expect some problems like this.

-- 
DE