*** xes has quit IRC | 00:08 | |
*** xes has joined #maemo-ssu | 00:08 | |
parazyd | freemangordon: the kernel is in /root/ of the image | 00:28 |
---|---|---|
parazyd | you're supposed to do ubootmenu yourself | 00:28 |
parazyd | through maemo | 00:29 |
parazyd | if there's another way, let me know :) | 00:29 |
*** RedW has joined #maemo-ssu | 00:48 | |
*** RedM has quit IRC | 00:48 | |
*** RedW has quit IRC | 01:51 | |
*** RedW has joined #maemo-ssu | 01:52 | |
*** NishanthMenon has joined #maemo-ssu | 02:23 | |
*** RedW has quit IRC | 03:13 | |
*** RedW has joined #maemo-ssu | 03:14 | |
*** Guest64392 is now known as Milhouse | 03:50 | |
*** Milhouse has joined #maemo-ssu | 03:51 | |
*** RedW has quit IRC | 06:35 | |
*** RedW has joined #maemo-ssu | 06:35 | |
*** RedW has quit IRC | 06:52 | |
*** RedW has joined #maemo-ssu | 06:53 | |
*** DocScrutinizer05 has quit IRC | 07:30 | |
*** DocScrutinizer05 has joined #maemo-ssu | 07:30 | |
kerio | DocScrutinizer05: that devuan image has a partition table ._. | 09:20 |
kerio | ok admittedly i could use my 1gb microsd but | 09:23 |
freemangordon | kerio: uImage won't boot | 09:24 |
kerio | nooooooooooooou D; | 09:24 |
kerio | why not | 09:24 |
kerio | i thought it was the latest pali linux4 | 09:24 |
freemangordon | no idea, all I get is a black screen | 09:24 |
freemangordon | yep, but it is compiled by some custom build tools :) | 09:25 |
freemangordon | and no, it is not the latest | 09:25 |
kerio | rip | 09:25 |
freemangordon | that one https://github.com/pali/linux-n900/tree/v4.7-rc4-n900 is the latest | 09:25 |
kerio | surely https://github.com/pali/linux-n900/tree/master | 09:27 |
kerio | there's no master D; | 09:28 |
freemangordon | or even https://github.com/freemangordon/linux-n900/tree/v4.7-rc5-n900 ;) | 09:28 |
freemangordon | but... all those add stuff we don;t really need atm for devuan | 09:29 |
freemangordon | so, vanilla upstream with custom config is the right way IMO | 09:29 |
DocScrutinizer05 | FYI: please see http://termbin.com/meyp | 10:24 |
DocScrutinizer05 | (it's a tad annoying to 'sync' *three* IRC channels) | 10:25 |
kerio | DocScrutinizer05: to be fair | 10:26 |
kerio | "modem turned into usb keyboard" would make headlines | 10:26 |
kerio | unprecedented attack vector ;o | 10:27 |
DocScrutinizer05 | indeed :-P | 10:27 |
DocScrutinizer05 | particularly when you consider that there's per default no USB kbd at all in this platform | 10:28 |
DocScrutinizer05 | and even if you could turn the modem into a fully working kbd and get it accepted by the system, you still gained zilch since there's no console per default | 10:29 |
DocScrutinizer05 | not to think of you *seeing* anything from such nonexistent console, even if it existed | 10:29 |
kerio | oh really | 10:30 |
DocScrutinizer05 | I mean, it's not like the xserver would usually autodetect all input devices and activate them, on a platform like Neo900 | 10:30 |
kerio | what if i also present a usb video card | 10:30 |
kerio | :D | 10:30 |
kerio | ok actually | 10:30 |
DocScrutinizer05 | that might work :-) | 10:30 |
kerio | ftdi usb serial | 10:31 |
kerio | although arguably that's the other way around | 10:31 |
kerio | DocScrutinizer05: usb NIC and rely on network autoconf? | 10:32 |
DocScrutinizer05 | if your system already has a driver for that USB NIC... :-D | 10:33 |
kerio | DocScrutinizer05: what about rethinking the connection between system and modem | 10:34 |
kerio | can't you use ethernet | 10:34 |
DocScrutinizer05 | no | 10:35 |
kerio | :( | 10:35 |
kerio | i like ethernet | 10:35 |
DocScrutinizer05 | the modem has USB and UART | 10:35 |
DocScrutinizer05 | the OMAP has no genuine ethernet interface anyway | 10:35 |
kerio | that's where you attach the USB NIC :D | 10:36 |
DocScrutinizer05 | LOL | 10:36 |
kerio | are all the modem functions available through uart | 10:36 |
kerio | or is that too slow | 10:36 |
DocScrutinizer05 | OMAP-USB-->USB/NIC--ethernet-->NIC/USB-->modem | 10:36 |
kerio | indeed :D | 10:37 |
DocScrutinizer05 | all available through UART but too slow for LTE and HSPA | 10:37 |
kerio | anyway, surely there's a way to lock down specific usb ports | 10:37 |
kerio | in linux | 10:37 |
DocScrutinizer05 | there is, for sure | 10:37 |
DocScrutinizer05 | pabs3 just mentioned usbguard | 10:38 |
kerio | usb mass storage and trick uboot into booting some other kernel? | 10:38 |
kerio | usb keyboard into sysrq+b, then usb mass storage and uboot script | 10:38 |
DocScrutinizer05 | https://dkopecek.github.io/usbguard/ | 10:39 |
DocScrutinizer05 | I invited paulfertser and pabs3 to join here, but I dizbt they will | 10:42 |
DocScrutinizer05 | doubt* | 10:42 |
*** pabs3 has joined #maemo-ssu | 10:42 | |
pabs3 | usbguard is more for situations where you have external USB ports anything can be plugged into. | 10:47 |
pabs3 | for non-pluggable USB stuff I think an in-kernel or in-devicetree whitelist is the way to go | 10:48 |
DocScrutinizer05 | yep | 11:10 |
DocScrutinizer05 | a dedicated kernel driver module that only matches/binds to that particular USB interface and modem device type, and that gets loaded during boot and thus occupied the usb controller for that bus before any generic USB driver could get its grubby fingers at it | 11:13 |
DocScrutinizer05 | nothing wrong with a dedicated FOSS kernel driver for a particular detail of a particular hw platform | 11:22 |
DocScrutinizer05 | could even get pretty generic when you hand in USB address and device's VID/PID as module parameters | 11:23 |
DocScrutinizer05 | err well, when we make it a kernel driver for the USB *controller* then the USB address is pretty much irrelevant since there's only this one USB device supposed to be hooked up to this controller, and this particular controller driver could probably not handle other devices concurrently to the 'locked' device anyway | 11:26 |
DocScrutinizer05 | rather it needs to be the controller address then, I guess | 11:26 |
DocScrutinizer05 | OMAP for example has one musb-hdrc OTG controller and two host controllers | 11:27 |
DocScrutinizer05 | I'm too much of a DT noob to know if the controller could get assigned our special kernel driver in DT | 11:29 |
DocScrutinizer05 | freemangordon: pali: ^^^ | 11:35 |
DocScrutinizer05 | we might want a special "USB"driver for the modem anyway, optimizing out any delays from (useless) ENUM, and maybe special handling for USB bus suspend | 11:43 |
DocScrutinizer05 | to optimize power management and performance, we need to minimize latency of that particular USB bus to enter and leave suspended state | 11:44 |
DocScrutinizer05 | and possibly even shortcut ENUM completely, too | 11:45 |
*** RedW has quit IRC | 12:21 | |
*** RedW has joined #maemo-ssu | 12:21 | |
*** pyllyukko has joined #maemo-ssu | 12:54 | |
*** jon_y_ is now known as jon_y | 14:28 | |
*** DrCode has quit IRC | 17:53 | |
*** DrCode has joined #maemo-ssu | 18:11 | |
*** Pali has joined #maemo-ssu | 18:16 | |
*** RedW has quit IRC | 19:38 | |
*** RedW has joined #maemo-ssu | 19:38 | |
*** povbot_ has joined #maemo-ssu | 21:43 | |
*** ChanServ sets mode: +v povbot_ | 21:43 | |
*** gregoa_ has joined #maemo-ssu | 21:47 | |
*** jon_y has quit IRC | 21:51 | |
*** povbot has quit IRC | 21:51 | |
*** gregoa has quit IRC | 21:51 | |
*** XDS2010 has quit IRC | 21:54 | |
*** XDS2010 has joined #maemo-ssu | 21:57 | |
*** xes has quit IRC | 21:57 | |
*** XDS2010 has quit IRC | 21:59 | |
*** XDS2010 has joined #maemo-ssu | 22:00 | |
*** xes has joined #maemo-ssu | 22:24 | |
*** gregoa_ is now known as gregoa | 22:51 | |
*** Pali has quit IRC | 23:17 |
Generated by irclog2html.py 2.15.1 by Marius Gedminas - find it at mg.pov.lt!