IRC log of #cubox of Thu 11 Jan 2018. All times are in CET < Back to index

00:48 VoidBoy> YO
04:33 Alisa21> Who want fuck me? I'm naked now, here - http://2018sexxx.com
09:23 Ke> suihkulokki: so the status ended up being: people are working on something, but no fix or workaround is expected in next debian kernel?
10:02 jnettlet[m]> sorry about the spam guys. I am trying to get admin rights moved over to my account so I can deal with it.
10:02 jnettlet[m]> seems like freenode is just a spambot haven since the new year.
10:03 jnettlet[m]> Ke: I will push an ATF branch later that has the fixes supplied by ARM
10:04 Ke> thanks
10:04 Ke> though I was referring to the debian kernel boot problem that has probably no connection to spectre
10:06 Ke> also the kernels do not have meltdown fixes yet either
10:06 Ke> well right now my mcbin is only in uptime testing anyway
10:08 jnettlet[m]> Ke: I am pretty sure that Artox has posted the patch for that. Mainline had messed up some clock settings
10:08 Ke> I mostly care about the mainline
10:09 Ke> I was not meaning to complain about solid run support
10:09 Ke> since this is ARM issue mainly
10:11 suihkulokki> Ke: For the boot problem, It'll get fixed with updated DT - and the second part by disabling omap_rng module build temporarily
10:12 Ke> suihkulokki: assuming I am lazy, will things just work out automatically for me, if I use sid/buster?
10:13 suihkulokki> Ke: today no.. in a few weeks yes
10:13 Ke> that's fine
10:25 topi`> I need to give a hard reset to a USB dongle which is connected to a remote hummingboard. Can anyone remind me how this is done via /sys/class/regulator ?
10:25 topi`> I need to disable +5V from usb_otg and then bring it back
11:12 jnettlet[m]> topi`: under /sys/devices run a find for *usb-otg-vbus*
11:12 jnettlet[m]> then go into the regulator device and echo disabled > state
11:40 topi`> permission denied
11:45 topi`> the node seems to be r--r--r--
11:47 topi`> fixed, now Input/output error
11:47 topi`> kernel is 4.9.39 (bleeding edge, eh)
11:48 Ke> I think the main cause for readonly could be missing write handler
11:48 topi`> did you have some specific patch against 4.x to enable those regulators?
11:48 topi`> Ke: I agree
11:48 topi`> or maybe the thing is not declared in the .dts
11:57 jnettlet[m]> topi`: I have a patch in my kernels to support that. mainline will not carry it.
12:01 topi`> jnettlet[m]: is the patch in some of the github kernel repos?
12:01 topi`> can you find me the right commit id ;)
12:02 jnettlet[m]> not at the moment, bouncing between things. It is probably the latest patch to the regulator driver directory
12:04 topi`> OK, I'll check
12:05 topi`> linux-stalbe or linux-imx6?
12:39 topi`> found a patch that touches regulator/core.c
12:39 topi`> funny that Russell didn't want it in mainline
13:02 jnettlet[m]> mainline feels that controls to the regulators should be exposed by the devices they are attached to.
13:03 jnettlet[m]> yet the USB sub-system removed access to the regulators years ago.
13:04 jnettlet[m]> personally with all the hotplug support in the kernel I see no reason why you shouldn't just be able to shut off regulators at will
13:06 Ke> well here topi` had a use case
13:06 Ke> could be a build time option or something
13:07 Ke> but sufficiently many build time options make a kernep just a lisp interpreter
13:11 Ke> https://modmymods.com/modmymods-raspberry-pi-mini-water-cooling-kit-mod-0171.html could work for mcbin, is just extremely expensive
13:21 jnettlet[m]> Ke: do you have space for a 120mm fan in your case? If so then look at the CoolerMaster seidon line. It is a prefilled and sealed, closed system you buy and install. you can get them between $25 and $50
13:24 Ke> I think the problem was that the cooling block would not fit on top of the SoC
13:25 Ke> 62 x 62 x 36.4 mm (2.4 x 2.4 x 1.4 inch)
15:18 topi`> usb_otg regulator works now
15:18 topi`> 1 patch needed