**** BEGIN LOGGING AT Tue Jun 26 03:00:02 2018 Jun 26 03:19:48 After installing python2.7, pyversions -i (or -s) still gives only python2.5 Jun 26 04:36:17 how would pyversions know which versions are installed? Jun 26 04:38:01 not that it couldn't tell, but prolly knowing how it works helps spotting the "problem" Jun 26 06:22:08 It occurs to me that fax and SMS and maybe even Telex can be considered a form of electronic mail. Jun 26 06:22:29 messaging. mail is also a form of messaging Jun 26 06:23:32 Yes. Jun 26 06:25:45 Has the “People’s Republic of China” truly had USA DARPA Internetwork connectivity only since 1994, like Wikipedia says? Jun 26 06:30:43 Presumably you could always have just used dial-up. Jun 26 06:31:13 (using a service provider in a different country, if none existed within China) Jun 26 06:32:13 The .tw WHOIS server says that some .tw domain names, at least asus.com.tw , have existed since the 1980s, which suggests that the Republic of China had USA DARPA Internetwork connectivity before both the World Wide Web existed (project was proposed in 1989) and before the PRC had USA DARPA Internetwork connectivity. Jun 26 06:32:34 Wikipedia says: > China has been on the internet intermittently since May 1989 and on a permanent basis since 20 April 1994. Jun 26 06:33:02 I don't think that's particularly late tbh Jun 26 06:33:47 taiwan != china (especially in 90s) Jun 26 06:34:01 Yeah, but he's talking about China, not Taiwain Jun 26 06:34:08 Taiwan = Republic of China Jun 26 06:34:14 China = People's Republic of China Jun 26 06:34:17 but it was special case Jun 26 06:34:21 unless he misspoke. Jun 26 06:34:24 they had some independence Jun 26 06:34:47 that's why in 90s taiwan != china Jun 26 06:36:23 Apparently internet was only in this country (New Zealand) in 1989 too. Jun 26 06:36:44 at a particular university, not available in general. Jun 26 06:37:42 but in any case, we still had phonelines, so people could have dialed up themselves to offshore ISPs beforehand. Jun 26 06:39:58 Mainland China was the original Republic of China from 1912 to 1949, when mainland China became the People’s Republic of China and Taiwan became the current Republic of China. Jun 26 06:40:38 yes, but it still had some independence till some event in 90s/early 2000 Jun 26 06:40:45 when you say "The Republic of China", do you mean by name, or by some other sort of identity of the people in control? Jun 26 06:42:09 I'm not particularly familiar with the history, but statements like that seem kind of vague. Presumably it should involve something more than which government used which name. Jun 26 06:43:29 However, references to the province is now rare since the Taiwan Provincial Government has largely been dissolved and its functions transferred to the central government or county governments since 1997 Jun 26 06:44:10 so, as i've said, till '97 it was a province of china, but rules were more lax then Jun 26 06:49:57 ha, just got someone excited for maemo/n900 Jun 26 06:50:18 (said person was planning to get a android phone to do -development- on it) Jun 26 06:55:34 Maxdamantus: By name. The original Republic of China, governed by the Kuomintang and led by Chiang Kai-shek, lost the Chinese Civil War to the Communist Party of China, led by Mao Tse-tung, who established the People’s Republic of China in mainland China in 1949, which is why 1949‑10‑01 is National Day in the PRC. The Kuomintang fled to Taiwan and established Taiwan as the current Republic of China. The PRC has never controlled Taiwan. The ROC Jun 26 06:55:34 continued to represent “China” in the United Nations until 1971, when the ROC lost its seat to the PRC. Since then, Taiwan/ROC is the both the largest country by population and the largest economy outside of the UN. Jun 26 07:02:06 The One China Policy is the party line of the PRC. Jun 26 07:26:33 buZz: The Nokia N900 can somewhat run Android as the host OS and some computers that ship with Android can also somewhat run GNU+Linux as the host OS. Even the Nexus One from 2010 can run Debian GNU+Linux as the host OS but without cellular telephony capability when I last checked years ago. Jun 26 07:26:51 brolin_empey: thats a huge benefit Jun 26 07:27:07 nonworking cellular a) makes battery lasts longer b) increases privacy of the device Jun 26 07:27:19 and c) prevents harrassment while working on something Jun 26 07:27:31 just buy a tablet? Jun 26 07:27:39 KotCzarny: no clue what you're saying Jun 26 07:27:48 android without cellular Jun 26 07:27:52 which maemo tablet are you talking about? Jun 26 07:28:00 i dno why we're discussing android? Jun 26 07:28:08 nobody wants android at all Jun 26 07:28:46 it has its uses Jun 26 07:29:06 beside alarmclock , i havent found any use for android Jun 26 07:29:14 but sure, if you wanna lower standards to it, you can Jun 26 07:29:27 tv boxes Jun 26 07:29:41 kodi runs on linux just fine Jun 26 07:30:10 as long you have cpu power or hw accelerated decoding Jun 26 07:31:30 sure, like a raspi1 Jun 26 07:31:46 i like oranges Jun 26 07:34:41 buZz: I did not say without cellular connectivity, I said without cellular telephony capability. It may have working cellular data connectivity. Jun 26 07:34:56 brolin_empey: thats horrible :P Jun 26 10:26:38 is there anyway to look at serial console output, messages from the kernel, etc, from an ssh connection? Jun 26 10:27:27 cat /proc/kmsg ? Jun 26 10:27:38 (about kernel messages) Jun 26 10:28:05 that may be good enough Jun 26 10:28:06 thanks! Jun 26 10:28:22 i have a remote thing that is oopsing Jun 26 10:28:44 2.6.34 kernel, modern as it comes Jun 26 10:29:41 maybe ttylog too Jun 26 10:29:54 ie. ttylog tty Jun 26 10:30:36 i'd rather not have to install anything else Jun 26 10:30:45 embedded systems are a pain in the ass Jun 26 10:30:49 maybe it's already installed Jun 26 10:30:54 or available in busybox Jun 26 10:31:06 nope Jun 26 10:31:41 also: t=`tty`; setconsole $t Jun 26 10:32:35 http://old.easyco.com/easydocs/linux/man/man8/setconsole.htm Jun 26 10:32:37 i don't have setconsole even on my pc Jun 26 10:32:38 might be what you need Jun 26 10:33:20 /proc/kmsg may be enough Jun 26 10:33:29 now i need to be able to trigger this oops on command Jun 26 10:33:35 because up to now seems completely random Jun 26 10:33:38 in slackware: ./a/kbd-1.15.3-i586-4.txz: /usr/bin/setconsolefont Jun 26 10:33:42 erm. wrong Jun 26 10:33:49 scratch the last comment ;) Jun 26 10:33:51 lol Jun 26 10:34:44 you are right, it might be either very old command or rh specific Jun 26 10:35:52 cat /dev/console might be also a trick Jun 26 10:36:06 easiest to trigger kernel output is swapon/swapoff Jun 26 10:36:53 there's no swap on this thing :) Jun 26 10:37:05 8MB swapfile ? Jun 26 10:37:23 na Jun 26 10:37:32 i'll wait and see what happens with /proc/kmsg Jun 26 10:37:40 i have a remote person too, who can provide me logs when required Jun 26 10:37:44 but i didn't want to bother him Jun 26 10:37:49 so i'm bothering you :D Jun 26 10:38:07 not bothering at all Jun 26 10:38:23 this oops is also ancient Jun 26 10:38:25 writing some script so i can multitask Jun 26 10:38:30 nowadays oopses are more verbose Jun 26 10:38:34 i don't understand a thing Jun 26 10:38:40 depends on kernel options Jun 26 10:38:55 usually i only understand which module it occured Jun 26 10:39:22 it's telling me that it occured on one task that is a program i wrote Jun 26 10:39:42 but at the moment of the oops, this program is only doing mathematical calculations Jun 26 10:39:46 oh, right, dividing by 0 in c program might also trigger oops Jun 26 10:39:47 :) Jun 26 10:39:49 and shm read/write access Jun 26 10:40:00 well, dividing by 0 segfaults, not oops Jun 26 10:40:08 hmm, right Jun 26 10:40:08 doesn't segfault either, does it? Jun 26 10:40:14 simply crashes the process Jun 26 10:40:16 but not the kernel Jun 26 10:40:26 it also complains Kernel panic - not syncing: Fatal exception in interrupt Jun 26 10:40:26 unless it's a kernel module Jun 26 10:40:39 so i'm not sure if the problem is inside an interrupt service routine Jun 26 10:40:41 might be simply corrupting memory Jun 26 10:40:46 or if it's just a consequence of the oops Jun 26 10:41:11 Unable to handle kernel paging request for data at address 0x00000000 Jun 26 10:41:11 Faulting instruction address: 0x400155b0 Jun 26 10:41:11 Oops: Kernel access of bad area, sig: 11 [#1] Jun 26 10:41:15 there's some null dereferencing Jun 26 10:41:20 that's clear Jun 26 10:41:26 YUP Jun 26 10:41:42 segv in kernel code results in oops Jun 26 10:41:54 but where is it? I mean, I think I know which kernel module would be at fault, because this platform uses a driver written by the company that provides us the hardware Jun 26 10:42:02 so i guess it's their fault Jun 26 10:42:10 :> Jun 26 10:42:17 quality vendor code Jun 26 10:42:22 but where in the hell that is that piece of code lies the problem? that's a mistery Jun 26 10:42:23 binary too? Jun 26 10:42:28 no, at least i have the source Jun 26 10:42:38 recompile with lots of printk Jun 26 10:42:44 or use kgdb Jun 26 10:43:44 76513 total Jun 26 10:43:51 that's quite a few lines to read Jun 26 10:44:14 for a start add printk("%s\n",__func); in ever function start Jun 26 10:44:16 ;) Jun 26 10:44:28 s/__func/__func__ Jun 26 10:44:34 first i'm trying to make the oops repeatable Jun 26 10:46:04 if I had to guess, i'd say it should be due to the dma functions Jun 26 10:46:14 because that's the only thing that is happening: lots of dma transfers to userspace Jun 26 10:46:34 but, who knows what else is it doing on the background Jun 26 10:47:05 just add printk i've mentioned to every function with some sed Jun 26 10:47:25 but that's not going to serve much if i don't know how to make it crash Jun 26 10:47:33 it may happen today, or maybe tomorrow Jun 26 10:47:39 i need to make it happen fast enough Jun 26 10:47:53 check logs around time it happened Jun 26 10:48:00 maybe there will be some clue Jun 26 10:48:11 nope, nothing Jun 26 11:00:24 reading the Call Trace and looking at System.map, I may know where's the origin Jun 26 11:00:37 it may be some access to a strange serial device Jun 26 11:00:45 so i'm forcing lots of read/writes through that device, to see if something happens Jun 26 11:01:08 gonna have lunch meanwhile, let's see if it breaks then Jun 26 12:25:06 nothing happened Jun 26 12:25:10 grrr Jun 26 13:28:57 brolin_empey: AFAIK Taiwan was totally unrelated and almost at war with mainland China until maybe 20 or 30 years ago. Now they have somewhat better standing with each other but still are two totally separate states Jun 26 13:29:21 who represented China in UN is another highly intriguing topic Jun 26 13:29:47 Taiwan basically is under protectorate of USA Jun 26 13:31:15 there were times when there were MEGA speakers on the common costline, to bridge the several seamiles with mere sound and annoy the opponent Jun 26 13:31:37 iirc Jun 26 13:31:39 this was true for korea until a few weeks ago Jun 26 13:31:50 yeah Jun 26 13:32:07 they have a common land vorder though Jun 26 13:32:11 border Jun 26 13:32:30 ~wp taiwan Jun 26 13:32:47 ~wiki taiwan Jun 26 13:32:54 At https://en.wikipedia.org/wiki/Taiwan (URL), Wikipedia explains: "{{other uses}} {{redirect|Republic of China|the People's Republic of China|China|other uses|Republic of China (disambiguation)}} {{pp-move-indef|small=yes}} {{pp|small=yes}} {{short description|State in East Asia}} {{EngvarO|date=December 2016}} {{Use dmy dates|date=August 2017}} {{Infobox country | conventional_long_name = Republic of China | common_name = Taiwan | native_name = ... Jun 26 13:40:52 "recently" a president or governor or whatever been elected in Taiwan who followed a strong policy of "one China" in some way, and that led to quite some turmoil Jun 26 13:43:03 to call mainland and Taiwan one country is almost like back when you would have called GDR and Western Germany one country Jun 26 14:06:11 ceene: don't you get a stack trace when your kernel segv? Jun 26 14:08:39 bencoh: yes, I was missing the fact that this kernel is compiled without CONFIG_KALLSYMS Jun 26 14:08:45 so the backtrace was pretty useless Jun 26 14:08:51 but I found the matching System.map file Jun 26 14:09:14 so now I know which driver is responsible, this seem to be triggered by the IRQ service routing of a strange serial driver Jun 26 14:09:29 but I haven't been able to replicate it at will Jun 26 14:09:42 i'm stresssing that serial port as a bitch, with no oops Jun 26 14:14:46 https://en.wikipedia.org/wiki/File:101.portrait.altonthompson.jpg www.linkedin.com/in/joerg-reisenweber ;-) Jun 26 14:16:58 ugh what? linkedin doesn't show my profile to non-"members"? Jun 26 14:18:47 depends on the profile Jun 26 14:19:03 basically that's configurable iirc Jun 26 14:43:22 bencoh: so can you visit http://www.linkedin.com/in/joerg-reisenweber ? Jun 26 14:43:36 aaaand are you a linkedin victim? Jun 26 14:45:43 DocScrutinizer05: it asks me to join Jun 26 14:45:59 that's what I ranted about Jun 26 14:46:00 DocScrutinizer05: and I do have a linkedin profile Jun 26 14:46:04 yeah :) Jun 26 14:46:09 aaah so maybe only for members Jun 26 14:46:36 you can set it to be public as well if you want (or so I recall) Jun 26 14:46:53 I did Jun 26 14:46:59 hmm, strange then Jun 26 14:47:52 wait, maybe they changed their policy Jun 26 14:49:51 even supposedly public, google-indexed profiles still redirect to that page Jun 26 14:49:59 (say, torvalds, for instance) Jun 26 15:06:26 and proxy on startpage gets 403 Jun 26 15:10:49 proxy? **** ENDING LOGGING AT Wed Jun 27 03:00:01 2018