Re: bzr push into mysql-6.0-falcon branch (olav:2742 to 2743) Bug#37622

[Joerg Bruehe]

Hi Olav !


Olav Sandstaa wrote:
> Joerg Bruehe wrote:
>> Have you tried that patch on some platforms which are not Solaris 9, 
>> or used a different compiler on Solaris 9 ?
> 
> Yes, I ran configure and compiled it on at least two other platforms 
> than Solaris 9: Solaris 10 with Sun Studio compiler and Linux with gcc.
> 
> Based on the running pushbuild 
> (https://intranet.mysql.com/secure/pushbuild/showdir.pl?dir=bzr_mysql-6.0-falcon 
> ) it seems like it is compiling without problems on several Linux and 
> Windows platforms.

Good to hear.

> 
>> We once had problems with a Makefile macro whose final continuation 
>> line was empty, as yours of "INCLUDES" will be in all cases where 
>> "FALCON_SPARC_ASSEMBLY" is not set.
>>
>> I scanned my archive and found that was in the 5.0 community tree, fix 
>> committed by me on 2007-11-15 for 5.0.51.
>> Our central host "production" failed on that.
>>
>> The fix was to reorder the elements on the lines so that the last line 
>> would always have some contents.
> 
> Thanks for the information. Do you want me to make an updated patch 
> where I have re-ordered the elements in the "INCLUDES" to avoid that 
> this breaks something? Or should we "wait-and-see" if this still is a 
> problem for some platform?

I cannot say I "want", but IMO it would be a safeguard against us 
hitting a "make" that has this problem.
I assume it may be version-specific, or rely on sufficient blank lines 
following, or whatever - all I know is that we hit it once, and solved 
it by reordering.

So yes, I would prefer to see that changed: "better safe than sorry".


Thanks,
Jörg

-- 
Joerg Bruehe,  MySQL Build Team,  joerg@stripped   ([phone]
Sun Microsystems GmbH,   Sonnenallee 1,   D-85551 Kirchheim-Heimstetten
Geschaeftsfuehrer: Thomas Schroeder, Wolfgang Engels, Dr. Roland Boemer
Vorsitzender des Aufsichtsrates: Martin Haering     Muenchen: HRB161028