[PATCH] misc size-shrinkage; merges

Bernhard Fischer rep.nop at aon.at
Wed Aug 30 07:14:05 UTC 2006


On Tue, Aug 29, 2006 at 05:12:32PM -0400, Rob Landley wrote:
>On Tuesday 29 August 2006 3:28 pm, Bernhard Fischer wrote:
>> On Tue, Aug 29, 2006 at 03:06:22PM -0400, Rob Landley wrote:
>> 
>> >to figure out why svn 16007 broke the build.  (archival/tar.c now breaks 
>> >while building on 4.0.3, but the description of the checkin says that it 
>just 
>> 
>> This change:
>> 
>http://www.busybox.net/cgi-bin/viewcvs.cgi/trunk/busybox/Rules.mak?rev=15761&r1=15701&r2=15761
>> 
>> added -Werror (not sure why you added it, but ok).
>
>I added it because we had no warnings, to try to discourage more warnings from 
>being introduced.

Yes, but please make sure that no -release version does contain it.
>
>Unfortunately, different versions of gcc produce enough different warnings 
>that I've had to disable it for gcc 4.1.0 (some of the warnings that produces 
>are _stupid_, I.E. "if (x) blah=42; else blah=99;", afterwards it'll warn 
>that blah could be used uninitialized even when there is NO path it can go 
>through to get there that won't initialize it...)
>
>But still, the patch added a new warning (-Winline), and -Werror had been in 
>the tree for a while, and an attempt to build what was about to be checked in 
>was apparently not made.

Unfortunately, i didn't try trunk with gcc-4.0 and my branch that indeed
was checked with several 3.x and 4.x didn't contain -Werror.

Still, IMO it's better to be sure that explicit inline's are handled
correctly (as in the setjmp cases) than to fail for some users in
possible obscure ways.
>
>I've fixed up most of it now.  e2fsprogs is toast but that's just more 
>incentive to replace it...



More information about the busybox mailing list