**** BEGIN LOGGING AT Wed Feb 13 02:59:58 2013 Feb 13 22:59:23 bluelightning: do you happen to meet hpa? Feb 13 22:59:41 ant_home: I have met him on a couple of occasions Feb 13 22:59:59 does he know about Yocto/OE? Feb 13 23:00:04 ant_home: didn't talk with him much though Feb 13 23:00:10 he's aware of it Feb 13 23:00:34 Darren has worked most closely with him I think Feb 13 23:00:48 (dvhart) Feb 13 23:00:49 ok Feb 13 23:01:08 I have one doubt, about klibc.bbclass Feb 13 23:01:48 I'm building since months with a modified class, started when fixing armv4 Feb 13 23:02:04 now, old version still produces smaller binaries Feb 13 23:02:19 as documented http://dev.man-online.org/man1/klcc/ Feb 13 23:03:19 if I let in place Feb 13 23:03:23 export CFLAGS="" Feb 13 23:03:23 export CPPFLAGS="" Feb 13 23:03:23 export LDFLAGS="" Feb 13 23:03:50 as in old version, we save some 20% Feb 13 23:04:16 but.. Feb 13 23:04:23 the diff is like Feb 13 23:04:25 -export CFLAGS="-O2 -pipe -g -feliminate-unused-debug-types -I/oe/oe-core/build/tmp-eglibc/sysroots/collie/lib/klibc/include -I/oe/oe-core/build/tmp-eglibc/sysroots/collie/lib/klibc/include/bits32" Feb 13 23:04:25 +export CFLAGS="-I/oe/oe-core/build/tmp-eglibc/sysroots/collie/lib/klibc/include -I/oe/oe-core/build/tmp-eglibc/sysroots/collie/lib/klibc/include/bits32" Feb 13 23:04:53 former is without zapping the flags Feb 13 23:05:00 O2 Feb 13 23:05:03 -g Feb 13 23:05:26 in fact, our std toolchain flags Feb 13 23:05:56 out tc also adds Feb 13 23:06:02 export LDFLAGS="-Wl,-O1 -Wl,--hash-style=gnu -Wl,--as-needed" Feb 13 23:06:24 both binaries are sane, fwiw Feb 13 23:07:16 I'm perplexed about the 'ideal' solution... Feb 13 23:08:26 I don't have any immediate suggestions Feb 13 23:08:43 reading klcc manual I'd tend to kep on resetting the flags Feb 13 23:08:45 though it's not hard to imagine our default flags aren't appropriate for a "lean" bootloader Feb 13 23:09:06 if the binary works I can't see the harm in it Feb 13 23:09:29 the point is, in the past more debug flags were added Feb 13 23:09:41 not only -g Feb 13 23:09:55 that in the past, not actually Feb 13 23:10:47 with xz/lzma we can tolerate the extra size Feb 13 23:13:08 hmm Feb 13 23:13:29 unfortunately I have to go... I have to get up at 5 AM tomorrow Feb 13 23:13:36 ouch Feb 13 23:13:36 catch you later Feb 13 23:13:47 sure, take your time **** ENDING LOGGING AT Thu Feb 14 02:59:59 2013