Wednesday, 2015-08-05

*** gahseyfan <gahseyfan!> has joined #minnowboard00:17
gahseyfantrying to write a protocol driver for an SPI device (small 2-line character-based display) and wondering if anyone can point me at a good example of something similar.00:19
prpplaguegahseyfan: most of the SPI displays already have nice kernel drivers as part of the FBTFT project -
prpplaguegahseyfan: might check there first00:22
prpplaguegahseyfan: otherwise read up on the spidev character device driver00:22
gahseyfanprpplague: thanks for the link, hadn't seen that project.  poking around their code now.  My device is a NewHaven character display  I don't see any characters currently supported by FBTFT.  I'm a little confused about how spidev is supposed to be used.  If you go the spidev route does that mean you put your protocol driver in userspace?00:55
prpplaguegahseyfan: that is correct00:55
prpplaguegahseyfan: but you can use spidev as a easy way to understand how to write a character driver that uses spi00:56
gahseyfanprpplague: cool, and I've already successfully made a board file that brings up the spi device (  right now I have a hack (that actually, works, i.e. it can print arbitrary text on the screen) with a kobject that creates some knobs in /sys.  Is this not how you are supposed to use spidev?01:08
yoctiBug 7267: normal, Medium, 1.9, dvhart, ACCEPTED , unable to switch SPI SS to active-high in a board file01:08
gahseyfanprpplague: I think I'm confused when you say a character driver that uses spi.  spi dev is a character driver itself, ya?  And you can write a user-space protocol driver that uses the character driver api (but is not a character driver)01:23
gahseyfanprpplague: s/spi dev/spidev01:23
prpplaguespidev is in of itself a simple example of a character driver01:24
prpplagueyou can use it as a guide for creating other character drivers that need to use the SPI interface01:24
*** tjt263 <tjt263!~RAPSCA77I@unaffiliated/tjt263> has joined #minnowboard01:51
*** Abhishek__ <Abhishek__!uid26899@gateway/web/> has joined #minnowboard02:41
*** aholler_ <aholler_!> has joined #minnowboard03:49
*** aholler <aholler!> has quit IRC03:52
*** dvhart <dvhart!~dvhart@> has joined #minnowboard03:55
*** nicksydney <nicksydney!> has quit IRC04:03
*** Abhishek__ <Abhishek__!uid26899@gateway/web/> has quit IRC04:50
gahseyfanprpplague: so I think ad7879-spi.c and ad7879.c are close to what I need to make for my character lcd driver with sys entries.  Does that look right?04:55
*** dvhart <dvhart!~dvhart@> has quit IRC05:12
gahseyfanprpplague: after reading through spi.h, I think I am understanding things a bit more05:31
*** emeb <emeb!> has quit IRC05:56
*** deubeuliou <deubeuliou!~David@> has joined #minnowboard07:26
*** wierdw <wierdw!8ad2de32@gateway/web/freenode/ip.> has joined #minnowboard07:30
wierdwIs anyone still up?07:30
wierdwI need help disabling the SPI write protect on my minnowboard Max07:31
wierdwI foolishly set it, and now cannot unset it.07:31
wierdwI have serial console access to the device07:32
*** bluelightning <bluelightning!~paul@pdpc/supporter/professional/bluelightning> has joined #minnowboard08:09
wierdwIs there a known issue with MBM x64 firmware .080 that prevents the firmware flash from being un-write protected from the system setup menu?08:11
tbrworst case you can still flash it directly. just need the right ref voltage level08:26
wierdwAll I did was set the protect value to enable. That's it08:42
wierdwIt wont set to disable again.08:42
wierdwIt goes through the motions, but checking in the setup afterward shows still protected08:43
wierdwI have not attempted to flash the firmware at all.08:43
*** calculus <calculus!~calculus@gentoo/user/calculus> has quit IRC08:45
*** calculus <calculus!~calculus@gentoo/user/calculus> has joined #minnowboard08:45
wierdwI would rather not have to wire up the SPI header and strap resistors on it just to get the stupid chip to unlock08:46
wierdwI CANT have been the only person to experience this issue.08:49
*** RzR <RzR!~RzR@unaffiliated/rzr> has quit IRC08:50
*** calculus <calculus!~calculus@gentoo/user/calculus> has quit IRC08:53
*** calculus <calculus!> has joined #minnowboard08:54
*** calculus <calculus!~calculus@gentoo/user/calculus> has joined #minnowboard08:54
wierdwIs there a way I can disable it from a booted console linux, with something like flasher? Or failing that, is there an .efi application I can run to forcibly turn off the protection?08:54
*** sjolley <sjolley!~sjolley@> has quit IRC09:05
*** sjolley1 <sjolley1!sjolley@nat/intel/x-oxlqwxtorrjrycvi> has joined #minnowboard09:05
*** calculus <calculus!~calculus@gentoo/user/calculus> has quit IRC09:17
*** calculus <calculus!> has joined #minnowboard09:18
*** calculus <calculus!> has quit IRC09:18
*** calculus <calculus!~calculus@gentoo/user/calculus> has joined #minnowboard09:18
wierdwOk, I have determined the problem. There is an issue with the .080 firmware, when the serial console is the only input device.10:12
wierdwOr rather, the setup utility will not commit changes via the serial console.10:12
wierdwneeds to be done with an actual keyboard10:13
wierdwflash is now unlocked.10:13
tbrwarthog9: ^^^^10:23
*** hrw <hrw!~hrw@> has joined #minnowboard10:55
*** hrw <hrw!~hrw@> has quit IRC11:00
*** hrw <hrw!~hrw@> has joined #minnowboard11:01
*** wierdw <wierdw!8ad2de32@gateway/web/freenode/ip.> has quit IRC11:04
*** hrw <hrw!~hrw@> has quit IRC12:16
*** nicksydney <nicksydney!> has joined #minnowboard13:45
*** wierdw <wierdw!8ad2de32@gateway/web/freenode/ip.> has joined #minnowboard14:12
wierdwIs there anything special I should know about getting Linux Mint 17.1 going on MBM?14:13
wierdwI have created a UEFI bootable mint installer stick, and grub2 loads OK-- however, shortly after the desktop starts to load, one of the blue lights on the MBM goes off, and the screen goes dark.14:14
*** Abhishek__ <Abhishek__!uid26899@gateway/web/> has joined #minnowboard14:18
givemefive911wierdw: do you have a big enough power supply..14:46
*** prpplague <prpplague!~prpplague@2602:306:bce4:b80:89fd:86ea:889:cc4c> has quit IRC14:52
wierdwGivemefive: Yes.14:54
wierdwI have a 12vDC@1.5A supply driving a 15W max DC-DC down converter, which gives me the 5vdc for the board14:55
aholler_doesn't sound so14:56
wierdwTo rule out insufficient supply, I have attached a much beefier 19vDC @ 3.7A supply, (this is still in voltage input spec for the DC-DC down converter)14:56
wierdwTHere are no sata hdds attached at this time14:57
aholler_maybe the dc-dc cant't handle 2A.14:58
wierdwUnless 5v@3A (15W) is insufficient supply?14:58
wierdwAccording to the wiki, that should be sufficient14:59
wierdwI can run linux just fine in console mode15:00
wierdwas soon as I start X though-- desktop starts to display for about 5 seconds, then system powers off15:00
aholler_wierdw: maybe the dc-dc puts 2A of the 15w on the 12v-side into heat ;)15:01
*** aholler_ is now known as aholler15:02
wierdwIt isnt hot to the touch, so I doubt that15:02
wierdwand 2W heat would be quite noticeable15:02
wierdwspec sheet for the converter lists 90+% conversion efficiency15:03
wierdwFor reference, this is the converter in question:15:07
wierdwIt specifically states 3A on the 5v side15:09
ahollermaybe the name of the company on the plate mathes the company in the descriptions the same way like the other writings on the plate ;)15:10
wierdwI have a multimeter, if you want me to measure inline amperage on the 5v side15:10
ahollernot necessary, and it likely won't help. maybe it's really something else15:12
wierdwI can go hunt down an actual 5vdc wall wart of sufficient amperage today to completely rule out the DC-DC down converter, but I wanted the 12v source, specifically so I could drive a power hungry HDD at some point.15:13
wierdwIt may well be insufficient supply. It has been acting kinda wonky even in the UEFI shell15:14
wierdwI was personally leaning more toward the known HDMI->DVI converter "Issues" MBM has15:16
wierdwWas wondering if forcing the x session to use a lower resolution might be profitable15:17
wierdwOr if Mint is trying to use some unsupported feature of the intel GPU series to do compositing15:17
ahollerjust in case of, I don't have one, so I'm not really a good conversation partner on that topic. it just sound very unlikely that some sw error puts of the mbm in your case15:18
wierdwI dunno about that.  I have seen similar behavior in other hardware with mint. Specifically, an ancient rackmount server I inherited for free. It has a weaksauce version of an already weaksauce GPU baked in. Mint tries to enable features of the full version of that GPU to do compositing, which crashes the while thing after about 1 second.15:19
wierdwHave to pass a special argument to the x driver to not do compositing.15:20
wierdwThis is my first experience with MBM, and the documentation is... not that great15:21
wierdwNearly all of the documentation I have found in the wiki is so very general that it is not helpful for troubleshooting at all.15:22
ahollerI assume at least one of the lights is turned on by hw. so any sw-bug would need to be able to switch of the power, which sounds unlikely.15:24
wierdwI will hunt down a beefy 5vdc wall wart, give it a shot, and report back.15:25
wierdwOr, it causes a triple fault15:25
wierdwThis is an embedded device, not a full PC-- so triple fault may not be handled the same way.15:26
wierdwPower off seems as good a solution (if not better) than hard reset15:26
wierdwAgain, not good documentation to know15:27
wierdwOn PCs with Intel cpus, when you get caught in an unresolvable exception, that trips 3 times, the CPU does a hard reset to get out of it.15:27
ahollerlook at the schematics to see how the leds are wired.15:28
wierdwDepending on how intel decided to handle this on embedded platforms, "power off" is a valid choice from an engineering perspective.15:30
wierdw(Keeps the platform from getting stuck in a boot loop, or other problem that could damage the hardware)15:31
ahollerif one of the leds is hardwired to power and that one goes off, then it's very likely that it's because the recoverable protection in the dc-dc went in action.15:32
wierdwThe power present LED stays on.15:32
wierdwThe system on LED goes off15:32
wierdwand I meant a triple fault exception resulting from incorrect driver parameters being loaded by Mint for the GPU15:33
ahollerso maybe try some other distro before buying some hw.15:34
wierdwI suppose-- I could try giving debian a try. It's listed on the supported OS page15:35
wierdwWell.. Time to download another 4gb disc image15:39
*** prpplague <prpplague!David@nat/intel/x-pepxxpmwdilsxima> has joined #minnowboard15:43
wierdwMight as well grab lubuntu to try as well.15:43
wierdwSounds familiar. Just with Mint instead15:46
wierdwI note he did not get any responses to his query15:48
*** lamego <lamego!~jalamego@> has joined #minnowboard15:49
*** lamego <lamego!~jalamego@> has quit IRC15:59
*** lamego <lamego!~jalamego@> has joined #minnowboard16:00
*** dvhart <dvhart!~dvhart@> has joined #minnowboard16:02
*** emeb <emeb!> has joined #minnowboard16:09
*** deubeuliou <deubeuliou!~David@> has quit IRC16:23
*** diana_olhovik <diana_olhovik!~quassel@> has joined #minnowboard16:48
*** dvhart <dvhart!~dvhart@> has quit IRC16:53
*** meflin_ <meflin_!> has joined #minnowboard16:54
*** meflin <meflin!> has quit IRC16:58
*** bluelightning <bluelightning!~paul@pdpc/supporter/professional/bluelightning> has quit IRC17:05
*** tjt263 <tjt263!~RAPSCA77I@unaffiliated/tjt263> has quit IRC17:12
*** clsulliv <clsulliv!clsulliv@nat/intel/x-ilfvvqndkjpyjiwp> has quit IRC17:13
*** clsulliv <clsulliv!clsulliv@nat/intel/x-xmgozzwxuvgpsqjx> has joined #minnowboard17:14
*** dvhart <dvhart!dvhart@nat/intel/x-kmcfxhkuxbskkrus> has joined #minnowboard17:21
*** wierdw <wierdw!8ad2de32@gateway/web/freenode/ip.> has quit IRC17:41
*** dvhart <dvhart!dvhart@nat/intel/x-kmcfxhkuxbskkrus> has quit IRC17:42
*** meflin_ is now known as meflin18:02
*** dvhart <dvhart!dvhart@nat/intel/x-csxalsyszswyowyg> has joined #minnowboard18:02
*** lamego <lamego!~jalamego@> has left #minnowboard19:03
*** diana_olhovik <diana_olhovik!~quassel@> has quit IRC19:06
*** lamego1 <lamego1!~lamego@> has joined #minnowboard19:08
*** lamego1 <lamego1!~lamego@> has quit IRC19:10
*** lamego1 <lamego1!~lamego@> has joined #minnowboard19:11
*** lamego1 <lamego1!~lamego@> has quit IRC19:13
*** lamego1 <lamego1!~lamego@> has joined #minnowboard19:13
*** clsulliv <clsulliv!clsulliv@nat/intel/x-xmgozzwxuvgpsqjx> has left #minnowboard19:44
*** clsulliv <clsulliv!~clsulliv@> has joined #minnowboard19:52
*** _daniel_ <_daniel_!~daniel@> has quit IRC21:43
*** lamego1 <lamego1!~lamego@> has quit IRC21:51
*** lamego <lamego!~lamego@> has joined #minnowboard21:52
*** Abhishek__ <Abhishek__!uid26899@gateway/web/> has quit IRC22:30
*** lamego <lamego!~lamego@> has quit IRC22:31
*** Yoofie <Yoofie!> has joined #minnowboard22:34
*** xnox <xnox!~xnox@ubuntu/member/xnox> has quit IRC22:57
*** xnox <xnox!~xnox@ubuntu/member/xnox> has joined #minnowboard23:00
*** dvhart <dvhart!dvhart@nat/intel/x-csxalsyszswyowyg> has quit IRC23:26
*** dvhart <dvhart!dvhart@nat/intel/x-losunvvsnhjdomgy> has joined #minnowboard23:54

Generated by 2.11.0 by Marius Gedminas - find it at!