IRC logs for #openrisc Friday, 2016-12-16

--- Log opened Fri Dec 16 00:00:29 2016
shorneolofk: cool08:33
shornefyi, I did my gdb testing with or1k test.  I can get all sprs etc and using yao's suggestions I have elliminated a lot of code08:48
olofkAwesome!08:54
olofkRemoving code is the best feeling :)08:54
stekernespecially if it's someone elses code ;)08:59
shorneyeah, hope I didnt break much09:19
shornedejagnu still running09:20
shorneok, something is not good # of unexpected failures        320209:58
shornevs 405 before09:58
shorneok, bad test setup10:43
shorne# of unexpected failures        40610:48
shorneok better10:49
olofkSo just a single extra failing test. That's good12:08
olofkI wonder if we should try gdb->openocd->target to see that everything works.12:09
olofkHmmm... have I been disconnecting a lot lately? A bit worried that my faitful old raspberry pi is about to give up. See a lot of eth0 link down in dmesg14:45
olofkOff-topic, but might still be the right place to ask. I want to do a simple circuit that can light up a LED when two pins are disconnected. Ten years ago I would have been able to do this, but I completely forgot how to even get started with this14:52
olofkI guess I just need a transistor, a battery, a LED and one or two resistors14:53
kc5tjaDepends onhow you approach the problem.15:03
kc5tjaYou could put the transistor across the LED, so that then base is connected to emitter (via a resistor of course!), the transistor shunts power away from the LED, and thus does not light.15:04
kc5tjaWhen that base/emitter junction breaks, the transistor turns off, allowing collector current to flow through the LED and light.15:05
kc5tjaHowever, this is a pretty power-hungry way of doing things; you're always going to be drawing collector current, regardless of the base/emitter connection.15:05
olofkI hope to run it on a coin cell15:09
kc5tjaChoose your collector resistor wisely.  ;)15:40
stekernyou can even solve it without a transistor, just connect the wires over the LED and have the resistor connected between the anode and vdd. You're going to drain the battery pretty quickly that way though ;)15:50
olofkhttp://www.falstad.com/circuit/circuitjs.html?cct=$+1+0.000005+10.634267539816555+43+2+50%0Aw+384+240+384+272+1%0Aw+384+208+384+160+1%0A172+384+112+384+16+0+6+1.5+1.5+0+0+0.5+Collector+Voltage%0Ag+384+272+384+352+0%0At+336+224+384+224+0+1+-1.3368536855467952+0.13264473452982517+400%0Aw+304+224+336+224+1%0Ar+384+112+384+160+0+500%0Ar+192+224+304+224+0+500%0As+192+112+192+176+0+1+false%0Aw+192+176+192+224+0%0Aw+192+112+384+112+0%0A162+496+19215:52
olofkWoah. That was the mother of all URLs15:52
olofkWhat's the typical current consumption of a dimly lit LED? Maybe it's less power consuming to run the LED all the time instead15:58
olofkThere! http://tinyurl.com/hxf4za716:46
shorneolofk: I did try gdb->openocd->target and it does seem to all work16:48
shorneI can access registers16:48
shornealso, I think the extra failure was showing up since I rebased gdb16:49
shorneI can access sprs16:49
olofkThat's great16:50
shornethe main changes I did were in patch 3/5, Ill let those sit for a while16:53
shorneif anyone else wants to test it will be helpful16:53
shorneIf someone wants me to push this all to openrisc/binutils-gdb I can16:53
shornein a few days Ill squash them all and send more patches to the gdb list16:54
olofkshorne: I read through the patches quickly, but haven't given them a proper review17:11
shorneactually, last time I ran tests (for first gdb submission after rebase) it was # of unexpected failures        41117:24
shorneso it seems my recent parches improve things17:24
--- Log closed Sat Dec 17 00:00:31 2016

Generated by irclog2html.py 2.15.2 by Marius Gedminas - find it at mg.pov.lt!