freebsd stable - Feb 2006 - CFLAGS in libc_r, libpthread and libthr, and MUTEX stuff in kern_mutex.c

Given that STABLE is supposed to be...well...*stable* :-) ...

Could we possibly remove "-D_PTHREADS_INVARIANTS",
"-D_LOCK_DEBUG" and "-g"
from the CFLAGS in the Makefiles for libc_r, libpthread and libthr?  I've
been under the impression that these options were only intended for
testing/debugging under CURRENT.

Granted, it's a simple matter to patch the Makefiles after an update/before
a buildworld, but it would be nice not to have to do so.

On a related note, I was rather surprised to discover recently that in
sys/kern/kern_mutex.c, the MUTEX_WAKE_ALL option is forcibly defined if not
already defined in the user's kernel config.  While the accompanying
comment (in mutex.c) would seem to indicate that there's good reason for
doing so at this time, at the very least, this should be documented in
sys/conf/NOTES as well.
 
Additionally, it should also be made unmistakably clear that certain other
mutex-related options -- namely, ADAPTIVE_GIANT and NO_ADAPTIVE_MUTEXES --
have absolutely no effect whatsoever unless SMP is also defined.  While I
suppose this might be inferred from the fact that they're listed under the
SMP section of NOTES, nonetheless, there is a chance that this relationship
may go unrecognized.

Thank you.

-- 
Conrad J. Sabatier <conrads@cox.net> -- "In Unix veritas"

On Fri, Feb 10, 2006 at 05:00:18PM -0600, Conrad Sabatier
wrote:
> Given that STABLE is supposed to be...well...*stable* :-) ...
> 
> Could we possibly remove "-D_PTHREADS_INVARIANTS",
"-D_LOCK_DEBUG" and "-g"
> from the CFLAGS in the Makefiles for libc_r, libpthread and libthr? 
I've
> been under the impression that these options were only intended for
> testing/debugging under CURRENT.
I think your impressions are wrong.  For example, -g has no effect on
the installed code (since it's stripped), and IIRC
-D_PTHREADS_INVARIANTS doesn't have a run-time penalty either.
> On a related note, I was rather surprised to discover recently that in
> sys/kern/kern_mutex.c, the MUTEX_WAKE_ALL option is forcibly defined if not
> already defined in the user's kernel config.  While the accompanying
> comment (in mutex.c) would seem to indicate that there's good reason
for
> doing so at this time, at the very least, this should be documented in
> sys/conf/NOTES as well.
>  
> Additionally, it should also be made unmistakably clear that certain other
> mutex-related options -- namely, ADAPTIVE_GIANT and NO_ADAPTIVE_MUTEXES --
> have absolutely no effect whatsoever unless SMP is also defined.  While I
> suppose this might be inferred from the fact that they're listed under
the
> SMP section of NOTES, nonetheless, there is a chance that this relationship
> may go unrecognized.
Sounds like you can submit a patch for these.

Kris
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 187 bytes
Desc: not available
Url :
http://lists.freebsd.org/pipermail/freebsd-stable/attachments/20060210/4589a6b8/attachment.bin

On Fri, 10 Feb 2006, Conrad Sabatier wrote:
> Given that STABLE is supposed to be...well...*stable* :-) ...
>
> Could we possibly remove "-D_PTHREADS_INVARIANTS",
"-D_LOCK_DEBUG" and "-g"
> from the CFLAGS in the Makefiles for libc_r, libpthread and libthr? 
I've
> been under the impression that these options were only intended for
> testing/debugging under CURRENT.
<sigh>
Please go search the archives for the last time this
was brought up.  You are under the mistaken assumption
that there is a performance impact.

I think next time I'm in there, I'll just change the names
of the macros or just remove them completely so they're
always enabled :(

-- 
DE