[00:17] <CIA-18> kaloz * r3958 /branches/buildroot-ng/openwrt/target/linux/aruba-2.6/patches/000-aruba.patch: return the correct system type for Aruba
[00:22] <nbd> damn... the via-rhine driver doesn't implement napi polling
[00:22] <CIA-18> mbm * r3959 /branches/whiterussian/openwrt/package/webif/files/ (usr/lib/webif/webif.sh www/webif.css): clean up IE workaround
[00:22] <nbd> that leads to crappy performance
[00:22] <nbd> i guess i need to fix that
[00:22] Action: [mbm] curses the redmond campus
[00:23] <[mbm]> every time I fix it for firefox or for IE I break the other in some subtle way
[00:25] <[mbm]> btw the javascript on the advanced wireless page doesn't seem to do anything
[01:42] <[mbm]> well, webif now renders correctly in firefox, ie and opera
[01:42] <[mbm]> suppose I'll check safari later
[01:42] Action: [mbm] guesses if there was an issue it would have shown up on one of th other browsers
[01:43] <[mbm]> next step is cleaning up the page and fixing some of the functionality
[02:16] <nbd> the via-rhine driver is really crappy
[02:16] <nbd> xrg_: btw. i did some more testing: copying the packets rather than letting the kernel access everything unaligned does help with the performance
[02:16] <nbd> xrg_: so the rx_copybreak thing is indeed useful
[07:15] <[mbm]> 'lo
[07:27] <CIA-4> nbd * r3960 /branches/buildroot-ng/openwrt/target/linux/rb532-2.6/patches/130-custom_partitions.patch: put a size limit on rb532 cf partitions
[07:33] <CIA-4> nbd * r3961 /branches/buildroot-ng/openwrt/target/linux/rb532-2.6/patches/ (240-via_hack.patch 240-via_rhine_performance.patch): massive via-rhine performance improvements
[14:12] <wigyori> where can i set up a BR2_EXTRA_GCC_CONFIG_OPTIONS for a specific arch? adding it into toolchain/gcc/Config.in doesn't set it
[20:49] <CIA-4> nbd * r3962 / (2 files in 2 dirs): remove some junk
[21:51] <pjf> some time ago I mentioned my ap started sending beacons with bssids made of all 0s; it seems it's caused by changing ap to sta, scanning, and changing back to ap mode; running 'wifi' helps
[21:52] <nbd> funny
[21:54] <pjf> yeah
[21:54] <pjf> and frustrating
[22:52] <CIA-4> nbd * r3963 /packages/ (. libs/ libs/popt/ libs/popt/Makefile): port popt library to buildroot-ng
[23:23] <CIA-4> nbd * r3964 /packages/libs/libbfd/ (. Makefile): add libbfd
[23:37] <{Nico}> nbd: ping
[23:37] <nbd> pong
[23:38] <{Nico}> nbd: i'm working on buildroot-ng, moving ./target/linux/package/* to ./package/
[23:38] <nbd> great
[23:38] <{Nico}> where's the best place to put the "define Package/<pkg>" in Makefiles
[23:39] <CIA-4> nbd * r3965 /branches/buildroot-ng/openwrt/package/rules.mk: use rm -f in package-rebuild and add extra options to the default configure template
[23:39] <{Nico}> top, bottom, before Package/<pkg>/install ?
[23:40] <nbd> i'd say direcly after include $(TOPDIR)/package/rules.mk
[23:40] <[mbm]> {Nico}: just make it consistant with the existing packages
[23:40] <nbd> let's sort it like this:
[23:40] <nbd> define Package/...
[23:40] <nbd> define Build/*
[23:40] <nbd> define Package/*/install
[23:40] <{Nico}> [mbm]: problem is that's not really consistant yet ;)
[23:41] <[mbm]> {Nico}: all the ones I packaged are
[23:41] <[mbm]> which is the same order nbd just gave
[23:41] <nbd> some of mine might not be consistant
[23:41] <{Nico}> ok
[23:42] <{Nico}> i'm porting kmod-fuse and adding user libs and util too
[23:43] <{Nico}> should i put everything in the kmod- package (like kmod-madwifi)?
[23:43] <[mbm]> if it comes as part of the same tar.gz we probably should just put it in the same package
[23:44] <[mbm]> think that's why we did madwifi that way
[23:44] <nbd> yeah
[23:44] <nbd> if it makes sense to split it up, do it
[23:44] <nbd> but not if all parts are required anyway
[23:45] <{Nico}> ok
[00:00] --- Sat Jun 17 2006
[00:22] <nbd> damn... the via-rhine driver doesn't implement napi polling
[00:22] <CIA-18> mbm * r3959 /branches/whiterussian/openwrt/package/webif/files/ (usr/lib/webif/webif.sh www/webif.css): clean up IE workaround
[00:22] <nbd> that leads to crappy performance
[00:22] <nbd> i guess i need to fix that
[00:22] Action: [mbm] curses the redmond campus
[00:23] <[mbm]> every time I fix it for firefox or for IE I break the other in some subtle way
[00:25] <[mbm]> btw the javascript on the advanced wireless page doesn't seem to do anything
[01:42] <[mbm]> well, webif now renders correctly in firefox, ie and opera
[01:42] <[mbm]> suppose I'll check safari later
[01:42] Action: [mbm] guesses if there was an issue it would have shown up on one of th other browsers
[01:43] <[mbm]> next step is cleaning up the page and fixing some of the functionality
[02:16] <nbd> the via-rhine driver is really crappy
[02:16] <nbd> xrg_: btw. i did some more testing: copying the packets rather than letting the kernel access everything unaligned does help with the performance
[02:16] <nbd> xrg_: so the rx_copybreak thing is indeed useful
[07:15] <[mbm]> 'lo
[07:27] <CIA-4> nbd * r3960 /branches/buildroot-ng/openwrt/target/linux/rb532-2.6/patches/130-custom_partitions.patch: put a size limit on rb532 cf partitions
[07:33] <CIA-4> nbd * r3961 /branches/buildroot-ng/openwrt/target/linux/rb532-2.6/patches/ (240-via_hack.patch 240-via_rhine_performance.patch): massive via-rhine performance improvements
[14:12] <wigyori> where can i set up a BR2_EXTRA_GCC_CONFIG_OPTIONS for a specific arch? adding it into toolchain/gcc/Config.in doesn't set it
[20:49] <CIA-4> nbd * r3962 / (2 files in 2 dirs): remove some junk
[21:51] <pjf> some time ago I mentioned my ap started sending beacons with bssids made of all 0s; it seems it's caused by changing ap to sta, scanning, and changing back to ap mode; running 'wifi' helps
[21:52] <nbd> funny
[21:54] <pjf> yeah
[21:54] <pjf> and frustrating
[22:52] <CIA-4> nbd * r3963 /packages/ (. libs/ libs/popt/ libs/popt/Makefile): port popt library to buildroot-ng
[23:23] <CIA-4> nbd * r3964 /packages/libs/libbfd/ (. Makefile): add libbfd
[23:37] <{Nico}> nbd: ping
[23:37] <nbd> pong
[23:38] <{Nico}> nbd: i'm working on buildroot-ng, moving ./target/linux/package/* to ./package/
[23:38] <nbd> great
[23:38] <{Nico}> where's the best place to put the "define Package/<pkg>" in Makefiles
[23:39] <CIA-4> nbd * r3965 /branches/buildroot-ng/openwrt/package/rules.mk: use rm -f in package-rebuild and add extra options to the default configure template
[23:39] <{Nico}> top, bottom, before Package/<pkg>/install ?
[23:40] <nbd> i'd say direcly after include $(TOPDIR)/package/rules.mk
[23:40] <[mbm]> {Nico}: just make it consistant with the existing packages
[23:40] <nbd> let's sort it like this:
[23:40] <nbd> define Package/...
[23:40] <nbd> define Build/*
[23:40] <nbd> define Package/*/install
[23:40] <{Nico}> [mbm]: problem is that's not really consistant yet ;)
[23:41] <[mbm]> {Nico}: all the ones I packaged are
[23:41] <[mbm]> which is the same order nbd just gave
[23:41] <nbd> some of mine might not be consistant
[23:41] <{Nico}> ok
[23:42] <{Nico}> i'm porting kmod-fuse and adding user libs and util too
[23:43] <{Nico}> should i put everything in the kmod- package (like kmod-madwifi)?
[23:43] <[mbm]> if it comes as part of the same tar.gz we probably should just put it in the same package
[23:44] <[mbm]> think that's why we did madwifi that way
[23:44] <nbd> yeah
[23:44] <nbd> if it makes sense to split it up, do it
[23:44] <nbd> but not if all parts are required anyway
[23:45] <{Nico}> ok
[00:00] --- Sat Jun 17 2006