**** BEGIN LOGGING AT Tue Mar 12 02:59:57 2019 Mar 12 05:25:12 ugh Mar 12 05:25:18 wtf are these f2fs errors? Mar 12 05:26:36 no upstream fix. this has to be a joke Mar 12 05:27:41 oh? Mar 12 05:28:40 https://gist.github.com/neheb/fae589ba656f8917e49a6aa76e4cf3e8 Mar 12 05:29:05 loop0 is overlay... Mar 12 05:33:40 looks like a TRIM issue Mar 12 05:33:56 i don't think emmc supports TRIM Mar 12 05:34:22 * mangix wonders if the bootloader supports ext4 Mar 12 07:49:52 are there any tricks to get more stable NAT speeds on a router? currently my NAT performance measured using iperf 2.0.9 varies by ~15 Mb/s Mar 12 07:50:52 i'm getting values between 520 Mb/s and 530 Mb/s, i'd like it to more be stable, for the testing purposes Mar 12 07:51:15 any idea what may cause such NAT speed variations? Mar 12 08:19:59 hm, switching to the qdisc pfifo / pfifo_fast seems to be a good idea Mar 12 08:36:44 rmilecki: fq_codel is not concerned with throughput Mar 12 08:37:08 so yes Mar 12 08:37:23 great :) Mar 12 08:38:30 speaking of pfifo_fast, i thought OpenWrt patches it out Mar 12 08:40:13 i just added it back for my testing needs Mar 12 08:40:29 got it Mar 12 11:49:11 dedeckeh: ping Mar 12 11:52:32 dedeckeh: How do you feel about having another go with the iproute2/bpf PR? Mar 12 11:55:52 dedeckeh: https://github.com/openwrt/openwrt/pull/1844 Mar 12 12:30:15 ldir:apart from the PKG_RELEASE increase it looks fine to me Mar 12 12:30:39 ldir:if I find some time I will do some testing Mar 12 12:31:00 dedeckeh: hiya - can you remember what the failure mode was with the earlier commit? Mar 12 12:32:37 ldir:it broke the installation of the utilities ip-bridge, ss, nstat in case the ip-full variant was selected Mar 12 12:35:34 ahh yes, I vaguely remember now Mar 12 20:29:24 Could someone help me with my C please. What are the following 2 lines actually doing? Mar 12 20:29:33 #define to_connmark(a) ((struct tcf_connmark_info *)a) Mar 12 20:29:44 struct tcf_connmark_info *ca = to_connmark(a); Mar 12 20:31:26 the full context can be found in net/sched/act_connmark.c Mar 12 20:37:10 I know " struct tcf_connmark_info *ca " means "create a pointer variable called ca that will point to a structure of as defined by tcf_connmark_info" That's the easy bit. Mar 12 20:38:06 " = ((struct tcf_connmark_info *)a)" and set it's value to .... brain fried core dumped Mar 12 20:40:14 oh, is (struct tcf_connmark_info *) just a cast? In essence forcing the value of a to also be interpreted as a pointer to a structure of tcf_connmark_info ? Mar 12 20:40:43 yes, it's just a cast Mar 12 20:40:57 to a struct tcf_connmark_info pointer Mar 12 20:41:52 so this is simply setting ca = a ? Mar 12 20:55:57 ldir:correct Mar 12 20:56:34 with a whole load of pointer type 'overriding' in between. Mar 12 20:56:51 Ok, so my C is very slowly improving. Mar 12 20:57:24 * ldir watches a snail overtake him Mar 12 21:02:31 Thank you everyone - your help is appreciated Mar 12 21:04:14 hasn't seen #define to_foo(bar) before - and notes it's all over the place in the kernel - mainly to container_of. - hmmm something else to look at Mar 12 21:10:14 it's just a macro Mar 12 21:10:36 it will simply replace it with the value from the macro at compile time Mar 12 21:11:23 Ahh! https://radek.io/2012/11/10/magical-container_of-macro/ ok so it's still magical in how it does what it does, but at least I know what it is supposed to do :-) Mar 12 21:11:30 it can either be a simple value "#define ITEM_COUNT 5" or a function #define so_something(a,b,c) printf(a, b, c) Mar 12 21:26:06 Mister_X: oh yeah I get #define I wasn't sure what ((struct tcf_connmark_info *)a) was trying to do - but it's simple (well now it is) Mar 12 22:31:12 dwmw2_gone: Hi, https://lists.infradead.org/mailman/listinfo/openwrt-adm is not accessible for me Mar 12 22:53:31 ubi0 warning: ubi_attach: valid VID header but corrupted EC header at PEB 40 Mar 12 22:53:46 NM, I see another problem, perhaps Mar 12 23:12:10 (bad build) Mar 13 02:34:07 how do i get a new brand added to the dataentry page? **** ENDING LOGGING AT Wed Mar 13 02:59:57 2019