mnemoc changed the topic of #linux-sunxi to: Allwinner/sunxi development discussion - did you try looking at our wiki? - Don't ask to ask. Just ask and wait! - - Logs at
Worf has quit [Quit: Konversation terminated!]
* atsampson rgghs as he fails for the second time today to CC a patch to linux-sunxi...
<atsampson> mripard_: I've put an updated version of the pcDuino3 Nano patch here:
naobsd has joined #linux-sunxi
ninolein has quit [Ping timeout: 264 seconds]
ninolein has joined #linux-sunxi
naobsd has quit [Quit: naobsd]
naobsd has joined #linux-sunxi
cnxsoft has joined #linux-sunxi
wenbin has joined #linux-sunxi
kurain has joined #linux-sunxi
adj_ has quit [Quit: Leaving]
egbert has joined #linux-sunxi
egbert_ has quit [Ping timeout: 255 seconds]
mitchell has quit [Ping timeout: 272 seconds]
vishnup has quit [Quit: Leaving]
petr_ has quit [Quit: No Ping reply in 180 seconds.]
petr has joined #linux-sunxi
p1u3sch1_ has joined #linux-sunxi
p1u3sch1 has quit [Ping timeout: 276 seconds]
mozzwald has joined #linux-sunxi
mozzwald has left #linux-sunxi [#linux-sunxi]
TheSeven has quit [Ping timeout: 265 seconds]
TheSeven has joined #linux-sunxi
wenbin has quit [Ping timeout: 240 seconds]
mmarker has quit [Ping timeout: 255 seconds]
JohnDoe_71Rus has joined #linux-sunxi
zombu2 has quit [Ping timeout: 255 seconds]
konradoo77 has joined #linux-sunxi
konradoo77 has quit [Ping timeout: 272 seconds]
reinforce has joined #linux-sunxi
FreezingAlt has quit [Ping timeout: 264 seconds]
kurain has quit [Ping timeout: 264 seconds]
igraltist has quit [Remote host closed the connection]
kurain has joined #linux-sunxi
FreezingAlt has joined #linux-sunxi
hramrach has quit [Ping timeout: 250 seconds]
<plaes> ssvb: doesn't it need also CONFIG_VIDEO_LCD_PANEL_LVDS=y
<plaes> duh.. nevermind
hramrach has joined #linux-sunxi
HeavyMetal has quit [Ping timeout: 245 seconds]
diego71_ has joined #linux-sunxi
diego71 has quit [Ping timeout: 265 seconds]
diego71_ has quit [Ping timeout: 245 seconds]
diego71 has joined #linux-sunxi
HeHoPMaJIeH has joined #linux-sunxi
domidumont has joined #linux-sunxi
domidumont has quit [Remote host closed the connection]
domidumont has joined #linux-sunxi
diego71 has quit [Ping timeout: 256 seconds]
diego71 has joined #linux-sunxi
selfbg has joined #linux-sunxi
hehopmajieh__ has joined #linux-sunxi
f15h has joined #linux-sunxi
HeHoPMaJIeH has quit [Remote host closed the connection]
sehraf has joined #linux-sunxi
cubear has joined #linux-sunxi
<mripard_> atsampson: I won't review or merge any patch that has not been sent to the mailing lists
<oliv3r_> mripard_: good morning!
<mripard_> oliv3r_: morning
focus has quit [Remote host closed the connection]
focus has joined #linux-sunxi
<oliv3r_> mnemoc: how do you keep the mirror/* forks up to date of linux-sunxi?
oliv3r_ is now known as oliv3r
souther has quit [Ping timeout: 272 seconds]
souther has joined #linux-sunxi
<oliv3r> ijc: I see in the new u-boot defconfigs that the A20 olinuxino lime1 uses sunxi_gmac; does it matter that while it uses that port, there is not a gmac PHY connected?
souther has quit [Ping timeout: 245 seconds]
souther has joined #linux-sunxi
souther has quit [Remote host closed the connection]
Gokhan_TR has joined #linux-sunxi
souther has joined #linux-sunxi
<ijc> oliv3r: Not sure, that's one for Hans I think.
<wens> morning
<oliv3r> ijc: yeah figured as much :)
<oliv3r> but he's not around :(
<oliv3r> wens: hi!
* wens @ Schiphol now
<oliv3r> wens: wow, welcome to NL :D
<wens> :D
<oliv3r> wens: what are your plans for this week? fosdem isnt' for 5 more day s:p
<mripard_> wens: welcome to the port-addicted part of the EU
<mripard_> s/port/pot/ :)
gianMOD has joined #linux-sunxi
<oliv3r> LOL i was just typing, what about port-addicted what? we don't drink that much port, that's the french!
<wens> oliv3r: will be in Rotterdam til wednesday, then onto brussels, back in amsterdam next monday
<mripard_> oliv3r: yeah, sure, and the hookers are here for the french too
<wens> though a colleague mentioned rotterdam isn't that interesting
<oliv3r> mripard_: port/wine, it's all the same!
<oliv3r> wens: rotterdam is ok for 1 day :p
pmattern has joined #linux-sunxi
pmattern has quit [Changing host]
pmattern has joined #linux-sunxi
<wens> oliv3r: so i was told
<wens> and amsterdam for 4 days?
wickwire has joined #linux-sunxi
reinforce has quit [Quit: Leaving.]
<oliv3r> wens: there's a lot to see in amsterdam, but 4 days, depends onthe size of your wallet; i know some museum's take you hafl a day to get through :)
<mripard_> oliv3r: if you're smoking wine, you're doing it wrong
reinforce has joined #linux-sunxi
domidumont has quit [Ping timeout: 246 seconds]
Gokhan_TR has quit [Ping timeout: 246 seconds]
<wens> oliv3r: sounds awesome
naobsd has quit [Quit: naobsd]
khuey|aw1y is now known as khuey
gianMOD has quit [Read error: No route to host]
gianMOD has joined #linux-sunxi
Froolad has quit [Quit: KVIrc 4.0.2 Insomnia]
<oliv3r> wens: rijks museum, van-gogh museum and sex museum are the most important ones :)
<mnemoc> oliv3r: cron + sanity checks. is it not working?
<oliv3r> amsterdam dungeon is pretty cool :)
<oliv3r> mnemoc: i just pushed 2 new remotes :p
<oliv3r> mnemoc: 3 actually
<oliv3r> mnemoc: u-boot mirror/master, mirror/sunxi, mirror/next
<oliv3r> being denx/master, sunxi-custodian-tree/master, sunxi-custodian-tree/next
gianMOD has quit [Remote host closed the connection]
<mnemoc> oliv3r: give me the remotes to include them
<mnemoc> same if you want nightlies
_massi has joined #linux-sunxi
popolon has joined #linux-sunxi
gianMOD has joined #linux-sunxi
<oliv3r> mnemoc:
<oliv3r> mnemoc: well what I gather from hans's e-mail; we'll retire the entire sunxi branch sooner rather then later; so these 3 are going to be the only active ones
domidumont has joined #linux-sunxi
Andy-D has joined #linux-sunxi
<mnemoc> oliv3r: giveme 30m to finish something and I'll add those mirrors and the corresponding night builds
<oliv3r> mnemoc: no rush; can do it tonight
naobsd has joined #linux-sunxi
<mnemoc> oliv3r: also, if you are aware of linux branches that should be added (or removed) from the nightlies, it would be awesome if you could give me the data :)
khuey is now known as khuey|away
ssvb has quit [Ping timeout: 264 seconds]
naobsd has quit [Quit: naobsd]
WarheadsSE has quit [Ping timeout: 252 seconds]
WarheadsSE has joined #linux-sunxi
FreezingAlt has quit [Ping timeout: 265 seconds]
gianMOD has quit [Remote host closed the connection]
gianMOD_ has joined #linux-sunxi
iamfrankenstein has quit [Quit: iamfrankenstein]
gianMOD_ has quit [Remote host closed the connection]
cubear has quit [Ping timeout: 244 seconds]
FR^2 has joined #linux-sunxi
WarheadsSE has quit [Ping timeout: 245 seconds]
WarheadsSE has joined #linux-sunxi
cubear has joined #linux-sunxi
ricardocrudo has joined #linux-sunxi
kurain has quit [Quit: kurain]
hypothalamus has joined #linux-sunxi
ssvb has joined #linux-sunxi
ganbold___ has joined #linux-sunxi
ganbold__ has quit [Read error: Network is unreachable]
bluse has quit [Read error: Connection reset by peer]
sehraf has quit [Ping timeout: 252 seconds]
dlan_ has quit [Ping timeout: 252 seconds]
dlan has joined #linux-sunxi
dlan has quit [Changing host]
dlan has joined #linux-sunxi
bluse has joined #linux-sunxi
Black_Horseman has quit [Quit: Zwi se logou mou!!!]
Seppoz has joined #linux-sunxi
iamfrankenstein has joined #linux-sunxi
<ssvb> mnemoc: is it possible for to implement some kind of html form for receiving data and storing it in a database?
<plaes> ssvb: pastebin and send link?
<ssvb> plaes: this is needed for the data, which is to be collected and submitted automatically by a robot
JohnDoe_71Rus has quit [Quit: Miranda IM! Smaller, Faster, Easier.]
<ssvb> actually a mailing list might be a perfect solution (if the data is sent there as an e-mail, then it provides full transparency)
<ssvb> filtering out duplicates might be a bit of a problem though
<ssvb> mnemoc: has a mail server running, right?
<ssvb> if we could create an smtp account there for a robot, which would only accept e-mails with a certain subject going to a certain google groups mailing list (to avoid abusing it for spam), then this would be great
<oliv3r> ssvb: no it does not, well it does; but i never finished the configuration
<oliv3r> it is on my todo list :)
<oliv3r> ssvb: but smtp should be working, only the auth db needs to be setup
<oliv3r> mnemoc: i will
<oliv3r> mnemoc: i will go over them in a while
<ssvb> oliv3r: oh, so you are the right guy, good to know :)
<oliv3r> ssvb: right guy for? :)
<ssvb> oliv3r: taking care of the mail server :)
<oliv3r> hahaha, yeah, it's a big WiP
<oliv3r> i started it, but ran out of time
<oliv3r> so postfix should be up and running and postfixadmin is setup and configured
<oliv3r> just adding some user accounts who are allowed to use smtp etc; and a dev account is to be done
<oliv3r> additionally mailman needs to be setup :p
<ssvb> oliv3r: this is needed to add "calling home" ability to :)
cubear has quit [Quit: Leaving]
<oliv3r> ssvb: ah, very cool :)
<mnemoc> ssvb: if someone wants to upload an SDK, e-mail would explote
<mnemoc> and in some cases the uploader wants to remain anonymous
<oliv3r> doesn't the debian bug reporter do the same thing (anonymously)?
<ssvb> mnemoc: maybe a reasonable size limit can be enforced? also uploading this data should be of course optional and always done with the user's consent
<oliv3r> mail is standard limited to 10 mb anyway
Seppoz has quit [Remote host closed the connection]
<ssvb> oliv3r: we want to have the device SID (or just its hash) as part of the submission in order to filter out duplicates, but the user name or other personal information is surely not required
Seppoz has joined #linux-sunxi
<oliv3r> ssvb: you do know that SID is not guaranteed to be unique, especially with a lot of chips being stuck at '00000000000000'
<oliv3r> so get ready for a lot of 00000 submissiosn
fr4z has joined #linux-sunxi
<ssvb> oliv3r: it's easy, the 00000 submissions can be just ignored :)
<ssvb> oliv3r: realistically, I have a bunch of sunxi devices:
<ssvb> oliv3r: and only one of them has zero SID
<oliv3r> ssvb: BUT! someone mailed me the other week, wanting to invest time into figuring out how ot write it
<oliv3r> AND with the u-boot dump from allwinner a while ago, we saw how the chips can get programmed
<ssvb> oliv3r: if we can do this, then it is in fact excellent! we can store the device id there and lookup dtb files based on it
<oliv3r> not sure if this works on all gens; but it gives us a good indication
<oliv3r> but it requires that the efuse vdd to be connected etc, so who knows, needs a lot of experimentation
<ssvb> SID is currently used to distinguish between A13 and A10s
<ssvb> so whoever is programming the initial value there, hopefully knows what he is doing
<oliv3r> :p
<oliv3r> expect a mess
<libv> typical.
<ssvb> duplicated SIDs do not matter for the statistics collection, it just means that we are going to ignore some of the data with no other negative consequences
<ssvb> first come, first served :)
<ssvb> but if SIDs can be in fact re-programmed, then some mess can be indeed expected
reinforce has quit [Quit: Leaving.]
cnxsoft has quit [Quit: cnxsoft]
<libv> actually, that's not too bad, we can at least override it ourselves and make it match what allwinner expects it to be when needs be
<libv> but yes, the device makers will mess this up everywhere
vbmithr has joined #linux-sunxi
ricardocrudo has quit [Ping timeout: 245 seconds]
<oliv3r> i don't know who programs them now
<oliv3r> i think it's mostly aw themselves
<oliv3r> during factory testing
<jemk> could somebody with a10s/a31 boards please check the ve version there?
<jemk> i would like to know if they are the same as on a13/a31s
<jemk> this should be able to easily read the version:
selfbg has quit [Quit: Leaving]
<ssvb> hramrach: ^
<ssvb> jemk: a10s and a31 (non-s variant) don't seem to be very popular around here
<oliv3r> mripard_: how well tested is your ssd1306 (not 07) code with regards to module unloading?
ricardocrudo has joined #linux-sunxi
<mripard_> oliv3r: what is your real question.
<oliv3r> mripard_:
<oliv3r> :)
<oliv3r> my 1309 behaves just like the 1306, e.g. some init stuff(that seems to be fine, the panel works etc) but when i rmmod, i get that
<mripard_> the code and the whole logs would hep
<oliv3r> i'll get you the whole shebang
<oliv3r> mripard_:
<plaes> oliv3r: wrong copyright year?
<oliv3r> plaes: it's an old driver?
<mripard_> oliv3r: logs please
<oliv3r> mripard_:
<oliv3r> sorry, took a while to produce them :)
<oliv3r> had to reboot the board :)
<oliv3r> as you can see, i only added a .init function for the 1309, going by what the 1306 did
<oliv3r> which only calls the write_cmd() on the i2c bus
<mripard_> what is it trying to access when it crashes?
<oliv3r> not sure what you mean
<oliv3r> i unbind vtcon, i unload fbcon; i unlead ssd1307fb
<oliv3r> and uppon unloading ssd1307, it just segfaults a
<oliv3r> lsmod shows it as active still, with ahuge number (e.g. -1?)
<mripard_> what I mean is : what pointer dereference makes it crash?
<oliv3r> fb_deferred_io_cleanup(info);
<oliv3r> so i'm guessing 'info'?
<oliv3r> info is obtained via i2c_get_clientdata(client); which seems to work for the line above, unregister_framebuff(info);
<oliv3r> *info is struct fb_info
<oliv3r> all happening in ssd1307fb_remove() (line 764)
<mripard_> I don't really want a guess :)
<mripard_> info is a structure
<oliv3r> the guess was about me not knowing for sure what you where referring to
<oliv3r> a pointer to a structure!
<mripard_> there's presumably a pointer somewhere in that structure that is dereferenced and breaks it
<mripard_> my question is: which one
<oliv3r> okay, i'll digg into that
<oliv3r> the crashdump doesn't show that :(
<oliv3r> but it's possible that this is an old bug?
<oliv3r> e.g. not something I did :)
<mripard_> yeah, of course it's possible.
reinforce has joined #linux-sunxi
<oliv3r> mripard_: cool; then i'll figure it out :)
<oliv3r> just to aid in finding it quicker, does the fact that the dump doesn't go any deeper supposed to help me where it goes wrong? e.g. fb_deferred_io_cleanup()? or it won't show anything deeper because of some other reason?
<mripard_> it probably crashes in the function itself, and not in another one
<oliv3r> good, cause fb_deferred_io_cleanup isn't too big
<plaes> every sysfs entry needs to be documented under Do.../ABI ?
<mripard_> yes
<plaes> ok, I added a "voltage" entry for sun4i-lradc driver
<plaes> so one doesn't need to tear down the tablet and hook up multimeter to determine values..
<plaes> or should I just add debug option to module?
<oliv3r> plaes: well it's something you usually don't want in the code, so should be fully optional imo
<oliv3r> since it's a dev-setting if anything
<plaes> well, it isn't that much code.. :P
<plaes> anyway, I'll cook up a RFC patch and shoot it to mailing list..
<mripard_> plaes: if that is going to be used and exposed as an ADC, then you want to move it to IIO.
<oliv3r> oh that's a good trick
<oliv3r> use it as an adc, and as an input :)
konradoo77 has joined #linux-sunxi
<plaes> well, creating a dual driver (iio/input) is currently way over my skills
<mripard_> it's a good opportunity to learn then :)
domidumont has quit [Ping timeout: 246 seconds]
naobsd has joined #linux-sunxi
<plaes> btw, is there any other way to "open" input device other than load evdev and then `cat /dev/event/inputX` ?
<mripard_> what do you mean?
<plaes> the sysfs attribute only shows voltage when input device is opened
Andy-D has quit [Ping timeout: 276 seconds]
<mripard_> that doesn't sound that reliable :)
<mripard_> but to open it, you have to call open() on the file
<mripard_> so any program that does so will do
arokux has quit [Remote host closed the connection]
<plaes> there's no device node when input-evdev is not loaded
<plaes> though, it might hide somewhere under sys or proc
hehopmajieh__ has quit [Quit: Konversation terminated!]
<mripard_> well, yeah
<mripard_> evdev provides the userspace interface, so that doesn't sound that weird
paulk-collins has joined #linux-sunxi
Skaag has quit [Quit: Leaving.]
orly_owl has quit [Quit: leaving]
tomboy64 has quit [Ping timeout: 250 seconds]
VargaD_ has joined #linux-sunxi
VargaD has quit [Ping timeout: 264 seconds]
VargaD_ is now known as VargaD
Zboonet has joined #linux-sunxi
FreezingAlt has joined #linux-sunxi
konradoo77 has quit [Ping timeout: 252 seconds]
orly_owl has joined #linux-sunxi
orly_owl has quit [Client Quit]
Zboonet has quit [Ping timeout: 264 seconds]
Skaag has joined #linux-sunxi
Skaag has quit [Max SendQ exceeded]
Skaag has joined #linux-sunxi
Skaag has quit [Max SendQ exceeded]
Skaag has joined #linux-sunxi
Skaag has quit [Max SendQ exceeded]
Skaag has joined #linux-sunxi
Skaag has quit [Max SendQ exceeded]
Skaag has joined #linux-sunxi
Andy-D has joined #linux-sunxi
konradoo77 has joined #linux-sunxi
libcg has joined #linux-sunxi
akaizen_ has quit [Remote host closed the connection]
akaizen has joined #linux-sunxi
afaerber_ has joined #linux-sunxi
akaizen has quit [Ping timeout: 245 seconds]
afaerber has quit [Ping timeout: 265 seconds]
bonbons has joined #linux-sunxi
<mnemoc> oliv3r: u-boot-denx is f*ed up :<
<mnemoc> Fetching denx
<mnemoc> error: Unable to find ecb091a9035a9c89a57b7fbd6cdf1db1c5ee9a80 under
<mnemoc> Cannot obtain needed object ecb091a9035a9c89a57b7fbd6cdf1db1c5ee9a80
<mnemoc> while processing commit c8766fad52167df22150efcf96f9ce1a7cf50b14.
cubear has joined #linux-sunxi
FR^2 has quit [Quit: Connection reset by peer]
afaerber_ is now known as afaerber
tomboy64 has joined #linux-sunxi
<plaes> how do you get the v2 into subject line when sending pages with `git send-email ..` ?
<plaes> add custom subject via --subject?
<plaes> or via --annotate?
<mnemoc> plaes: git format-patch -v2
<plaes> ahh
* plaes has lots to learn..
<mnemoc> we all do
<plaes> argh.. I actually forgot to test the keys :S
f15h has quit [Remote host closed the connection]
libcg has quit [Quit: libcg]
FDCX_ has joined #linux-sunxi
setkeh has quit [Ping timeout: 245 seconds]
setkeh has joined #linux-sunxi
ricardocrudo has quit [Ping timeout: 265 seconds]
gianMOD has joined #linux-sunxi
konradoo77 has quit [Ping timeout: 264 seconds]
_massi has quit [Remote host closed the connection]
konradoo77 has joined #linux-sunxi
igraltist has joined #linux-sunxi
gianMOD has quit [Remote host closed the connection]
Skaag has quit [Quit: Leaving.]
iamfrankenstein has quit [Quit: iamfrankenstein]
iamfrankenstein has joined #linux-sunxi
ricardocrudo has joined #linux-sunxi
gianMOD has joined #linux-sunxi
<paulk-collins> between u-boot-sunxi and mainstream u-boot, I get a diff of 9152 Mb available to the system -- I suspect that's the memory reserved for the framebuffer by mainstream u-boot
<paulk-collins> which is claimed in mainstream linux but is not in linux-sunxi 3.4, correct?
akaizen has joined #linux-sunxi
wickwire has quit [Read error: Connection reset by peer]
el3 has joined #linux-sunxi
f15h has joined #linux-sunxi
Skaag has joined #linux-sunxi
Skaag has quit [Max SendQ exceeded]
Skaag has joined #linux-sunxi
Skaag has quit [Max SendQ exceeded]
Skaag has joined #linux-sunxi
Skaag has quit [Max SendQ exceeded]
Skaag has joined #linux-sunxi
pmattern has quit [Quit: Genug für heute.]
<ssvb> paulk-collins: yes
Skaag has quit [Client Quit]
gianMOD has quit [Remote host closed the connection]
<el3> Anyone here has experience in building kernel for A20 device?
<vovcia> el3: me
<el3> vovcia, okey I cannot use the mainline kernel for A20. did I understand that right?
<vovcia> el3: yes
<el3> that meaning I do not use mainline uboot eighter?
<vovcia> hmm dunno
<libv> have either of you tried looking at our wiki?
<el3> vovcia, okey does not matter. I use the sunxi branch.
<el3> libv, yes I have.
konradoo77 has quit [Ping timeout: 256 seconds]
<vovcia> el3: im building kernel from working packages in arch
<vovcia> el3: just to add some features
<vovcia> el3: it was quite difficult to get from scratch
<igraltist> vovcia: i have since weekend the mainline kernel 3.19-rc5 running with mainline u-boot
<igraltist> so even kvm is working
<igraltist> only wifi is not proper working in moment
<igraltist> Linux jaschtschik 3.19.0-rc5-3 #8 SMP Sun Jan 25 20:34:25 CET 2015 armv7l Allwinner sun7i (A20) Family GNU/Linux
<igraltist> in moment i try to include the sound and sunxi-ss
<vovcia> hmmm igraltist let mi try it
<el3> vovcia, okey. I built a kernel for pcduino3, and debian. But after a few minutes I get some kernel errors and crash. So I want to ask you, do I have to remove alot of drivers from the kernel, to avoid crash.. I use some gpio pins for leds, and use g_mass_storage to make it act like a usb device
trollinger has joined #linux-sunxi
<el3> and wifi
<el3> only need that actually
<vovcia> el3: might be stupid questiot, but have You tried stall=0 with that module?
<igraltist> wifi is working but not proper as ap
<el3> vovcia, yes I use stall=0 on that
<vovcia> igraltist: what device do You have?
<igraltist> cubietruck
<vovcia> oh me too can You share configs? :)
<igraltist> jo
<vovcia> el3: what kind of errors do You get?
<vovcia> el3: You have serial console or sth to see messages?
Skaag has joined #linux-sunxi
Skaag has quit [Max SendQ exceeded]
<el3> vovcia, or maybe would another device be better. I only need: 1. it to act like a usb device, 2. wifi, 3. couple of gpio.
<el3> vovcia, yes wait. I will start it up
Skaag has joined #linux-sunxi
<igraltist> vovcia: i think you have to use u-boot mainline
<igraltist> i just painless replace the sunxi-u-boot
<vovcia> el3: You might also try the mainline kernel igraltist is talking about
Skaag has quit [Client Quit]
<vovcia> igraltist: u-boot from git or 2015.01 ?
<vovcia> ill go with git
Skaag has joined #linux-sunxi
<igraltist> vovcia: yes
<igraltist> its in german but the commands you can read
<el3> vovcia, igraltist , I will try the mainline before further questions. See if that helps. Thanks for advice thank you guys
gianMOD has joined #linux-sunxi
libcg has joined #linux-sunxi
FDCX_ has quit [Ping timeout: 245 seconds]
konradoo77 has joined #linux-sunxi
interrobangd has joined #linux-sunxi
<igraltist> what are the node numbers for nand devices?
<vovcia> igraltist: You mean major for nand? 93
<vovcia> igraltist: nand = 93, 0 ; nanda = 93, 1 and so on
<igraltist> ok thanks
<igraltist> it try the nand driver but i dont get devices
reinforce has quit [Quit: Leaving.]
Cooper_ has joined #linux-sunxi
<Cooper_> I've managed to install Gentoo on my PcDuino3 Nano (A20). Runs fine for the most part but when I tell Gentoo to (re)compile a package, doesn't matter which, I occasionally get an 'Illegal Instruction' or process terminated by signal 4 (SIGILL).
<Cooper_> I've seen this behaviour with all software on the board compiled using GCC 4.8.4 and 4.9.2.
FreezingDroid has joined #linux-sunxi
<ssvb> Cooper_: create a coredump and look into it with gdb
<Cooper_> It happens far less often when I run make single-threaded.
<ssvb> just to see what kind of instruction is that
<Cooper_> The problem is that it happens, like, 8 script levels deep away in Gentoo's build thing.
<Cooper_> I have no idea how I can trap this.
<ssvb> ulimit -c unlimited
<Cooper_> Ah. That will make it always dump core?
bengal has joined #linux-sunxi
<ssvb> and you should get something like "core dumped" message when this happens and also a nice core dump file
<ssvb> yes, when it crashes
<Cooper_> Excellent. I'll try that as soon as I get a chance (it's currently half-way done building tex)
FreezingAlt has quit [Ping timeout: 244 seconds]
<ssvb> it is either some rare and strange instruction, which is not supported by your CPU (if the CFLAGS are wrong), or it is a general reliability problem
<ssvb> looking into a coredump with gdb may provide some insights
<ssvb> Cooper_: you are not overclocking the hardware, right?
<Cooper_> Not that I'm aware of.
<ssvb> checking may be a good idea
<Cooper_> cpuinfo says both cores running at 1820MHz
<Cooper_> Euhh... BogoMIPS
<Cooper_> From dmesg: [ 0.479501] [cpu_freq] INF:sunxi_cpufreq_initcall, get cpu frequency from sysconfig, max freq: 912MHz, min freq: 60MHz
<Cooper_> Oh, I'm running this as a server, so I've done my best to turn the Mali chip mostly off.
<Cooper_> From my uboot boot.cmd: setenv bootargs console=tty0 disp.screen0_output_mode=1280x720p50 console=ttyS0,115200 root=/dev/mmcblk0p2 sunxi_g2d_mem_reserve=0 sunxi_no_mali_mem_reserve sunxi_fb_mem_reserve=16 consoleblank=0 rootwait panic=10
<Cooper_> I unconfigured mali from the kernel aswell, obviously.
<vovcia> Cooper_: did You run some memtests?
<Cooper_> Removed a host of other things aswell. Wondering if I maybe overdid it somehow with regards to the SMP capabilities... Since these problems tend to pop up more often when I run make -j3
<Cooper_> No, not yet.
cubear has quit [Quit: Leaving]
<vovcia> Cooper_: try it
<Cooper_> Will do.
<Cooper_> Does it matter that I don't have the mali module on here anymore?
<vovcia> it doesnt matter
<Cooper_> Ah, says so on the page.
<Cooper_> Srry
<ssvb> well, disabling mali in the kernel does not gain you much
<Cooper_> 30 megs of additional free memory. At least when I combine removing the module with those kernel params.
<ssvb> mali has mmu and does dynamic memory allocation if needed
<ssvb> how did you count 30 megs? and why this saving is attibuted to mali?
<ssvb> if you don't even load the mali module, nothing is wasted
<Cooper_> I looked at /proc/memfree directly after boot before and after those kernel params.
<Cooper_> Noticed the difference in the MemFree value.
tomcheng76 has quit [Read error: Connection reset by peer]
<ssvb> I believe it has something to do with sunxi_g2d_mem_reserve=0 and sunxi_fb_mem_reserve=16
<Cooper_> Likely.
<Cooper_> I don't think memory itself has a problem. I've rebuilt GCC and the kernel repeatedly, all without incident. My problems tend to arise when compilation is over and the files get moved to their final destination.
<Cooper_> I would consider the mere compilation of those enormous things a decent test of the memory on the board. (I'll still run the test though. Doesn't hurt to check)
<ssvb> Cooper_: is a *lot* more sensitive to DRAM misconfiguration than GCC compilation
<ssvb> Cooper_: also wrong DVFS voltages may be causing problems
<Cooper_> Hehehe. I'll run it once I see this compilation die on me (which, of course, it doesn't yet).
tomcheng76 has joined #linux-sunxi
<Cooper_> What can I grep dmesg for to see those voltages?
<Cooper_> I see a number reported by 'axp20'
<Cooper_> Or should I use meminfo for this?
Skaag has quit [Quit: Leaving.]
<ssvb> Cooper_: the dvfs voltages are specified in the fex file
<Cooper_> I've used the default Linaro_pcduino3 fex file.
<Cooper_> Even though this is the Nano.
<Cooper_> Linksprite. Sorry
<ssvb> Cooper_: cubieboard2/cubietruck had a rather nasty problem with this stuff earlier -
konradoo77 has quit [Ping timeout: 256 seconds]
fr4z has quit [Remote host closed the connection]
<ssvb> Cooper_: basically, this all had been identified by
steev has quit [Remote host closed the connection]
ojn has quit [Remote host closed the connection]
<Cooper_> Does it make sense to run that memtester while a compile is active, or should they just stay out of eachother's way?
<ssvb> if you don't have the mali module now, it will not work
<Cooper_> Ah.
<ssvb> and the generic memtester is not much better than GCC compilation
<vovcia> ssvb: mali module is essential for proper memory testing?
<Cooper_> So I'll just wait until this compilation fails on me and take it from there.
konradoo77 has joined #linux-sunxi
hypothalamus has quit [Read error: Connection reset by peer]
<ssvb> vovcia: CPU and Mali both stressing RAM at the same time is more efficient than just CPU alone
sehraf has joined #linux-sunxi
<vovcia> i see
<Cooper_> It's easy to miss in the description on the Hardware Reliability Tests page - doesn't require userland driver but does require kernel module.
<Cooper_> I was wondering, I'm currently running the kernel from the stable 3.4 branch. If I were to go to a later version or even mainline, the wiki suggests I'd lose SMP but that page was last updated in march of last year. Is that still an issue?
<ssvb> Cooper_: the mainline kernel has SMP support
FreezingDroid has quit [Ping timeout: 264 seconds]
Skaag has joined #linux-sunxi
<Cooper_> Nice. I'll give that a go after I get this sufficiently stable.
<ssvb> Cooper_: and the wiki can be fixed, you can even do it yourself :)
<Cooper_> Of course, but I'd have to make sure the data is valid and if there's no SMP there I'm not really interested. Bit of a chicken-egg problem I'll admit.
adj_ has joined #linux-sunxi
steev has joined #linux-sunxi
<Cooper_> Gotcha!
<Cooper_> : /var/tmp/portage/sys-devel/bison-3.0.4/image/usr/share/man/core: ELF 32-bit LSB core file ARM, version 1 (SYSV), SVR4-style, from '/bin/bash /usr/lib/portage/python3.3/ebuild-helpers/ecompressdir /usr/share/man'
<vovcia> i have a question about specifically dd if=u-boot-sunxi-with-spl.bin of=/dev/mmcblk0 bs=1024 seek=8
ojn has joined #linux-sunxi
<vovcia> i have root on sda and /boot on nanda
<vovcia> should i write to mmcblk0?
<vovcia> i mean, i dont have any card in sd slot..
<ssvb> Cooper_: now open it in gdb
<ssvb> Cooper_: read "man gdb" if necessary :)
<Cooper_> I was on my way to do that when... I noticed I didn't have gdb so that's now compiling.
<ssvb> Cooper_: :)
<ssvb> vovcia: what are you trying to do?
<vovcia> ssvb: im trying to install mainline u-boot
<vovcia> ssvb: thanks :)
<vovcia> ssvb: i have built u-boot already
<vovcia> "And now you will have a u-boot-sunxi-with-spl.bin to dd to your sdcard as
<vovcia> usual.
<vovcia> i dont have sdcard.
<ssvb> vovcia: have you built it correctly?
<ssvb> vovcia: also the mainline u-boot does not support nand yet
<vovcia> ssvb: yes i built i correctly
<vovcia> ssvb: i actually tried to boot mainline kernel, but forgot to install mainline u-boot
<vovcia> and guess what it doesnt boot.
<vovcia> ssvb: so, I should get sdcard?
<ssvb> vovcia: if you want to try the mainline u-boot and kernel, then yes
gianMOD has quit [Remote host closed the connection]
iamfrankenstein has quit [Quit: iamfrankenstein]
ojn has quit []
<igraltist> Cooper_: are you have a swap file enabled?
<Cooper_> Yes.
<igraltist> ok
<vovcia> i dont know how to install u-boot :\ U-Boot 2011.09-rc1-00000-gf75abad-dirty (Oct 21 2013 - 18:44:22) Allwinner Technology
<vovcia> CPU: SUNXI Family
<vovcia> Board: A20-Cubietruck
<vovcia> DRAM: 2 GiB
<vovcia> NAND: NB1 : enter NFB_Init
<vovcia> [NAND] nand driver(b) version: 0x2, 0x12, data: 20130526
vovcia has quit [Quit: WeeChat 1.0.1]
vovcia has joined #linux-sunxi
<vovcia> sorry for miss-paste.. i meant
iamfrankenstein has joined #linux-sunxi
<Cooper_> I actually decided to err on the side of caution and created 2 1gig swapfiles - not partitions. I've got SATA harddisks attached, but for the moment all is run off of the SD card and a few tmpfs mounts.
<Cooper_> Hrmm... I hope this core file is going to give me something. I just noticed I compiled my packages with -fomit-frame-pointer.
<ssvb> vovcia: write u-boot to some sd card, insert the sd card, power on your cubietruck
<vovcia> ssvb: i did that and this is the result
<ssvb> vovcia: then there is something wrong with the sd card
<vovcia> ssvb: i mean i did dd uboot to sdcard and i have /boot on nanda and / on sda1
<vovcia> ssvb: maybe i should dd uboot to nanda?
<vovcia> or this sounds like bricking? ;)
bonbons has quit [Quit: Leaving]
<ssvb> vovcia: the sd card has the highest boot priority
<ssvb> vovcia: be sure to clean the sd card and then write u-boot to it as instructed
<igraltist> Cooper_: can you share you cflags? iam using gentoo also
<vovcia> ssvb: ok ill get back in 7 minutes of FEX rescue booting.. :)
gianMOD has joined #linux-sunxi
<Cooper_> igraltist: -O2 -pipe -march=armv7ve -mtune=cortex-a7 -mfpu=neon-vfpv4 -mfloat-abi=hard -ffast-math -ftree-vectorize -fomit-frame-pointer
<Cooper_> igraltist: If you're using gcc 4.8 or earlier, replace armv7ve with armv7a
<igraltist> ok i have gcc
<Cooper_> ... version..?
<igraltist> 4.8.3
<igraltist> typing mistake :D
FlibberTGibbet has joined #linux-sunxi
<Cooper_> Then -march=armv7a
<igraltist> yo
<igraltist> but i dont have -ffast-math -ftree-vectorize -fomit-frame-pointer
<igraltist> other is same
<Cooper_> First two are for compiler-generated neon use.
konradoo87 has joined #linux-sunxi
<Cooper_> Last is to free up a register.
<Cooper_> Letting your compiler optimise your binary by doing things using neon sounds nice, but don't expect a lot from it. Unless hand-crafted it hardly makes a dent is my experience.
<igraltist> maybe i do add this too but later i will switch to hardened
<Cooper_> What board are you using?
<igraltist> cubietruck and you?
<Cooper_> PcDuino3 Nano.
<Cooper_> Setting it up to be my poor man's NAS.
<igraltist> never heard
mmarker has joined #linux-sunxi
konradoo77 has quit [Ping timeout: 245 seconds]
<Cooper_> Using a small patch to the SATA driver I can attach it to a 1-to-5 SATA port multiplier and thus drive 5 harddisks.
<Cooper_> See here:
<vovcia> ssvb: thanks it worked :)
<vovcia> igraltist: i have limited success kernel is trying to tftpboot :))))
<vovcia> igraltist: but it runs and has DHCP so thank You i will try this more tomorrow :)
<vovcia> thanks You all guys bye
<Cooper_> And my gdb build just dumped core so I made the build use just one job and hopefully it will complete this time.
xinj has joined #linux-sunxi
nicksydney has quit [Remote host closed the connection]
konradoo87 has quit [Ping timeout: 256 seconds]
<igraltist> Cooper_ this is nice to add more hdd
<Cooper_> It's what set me off on this mission.
<Cooper_> Hrmm... I'm not seeing the pictures anymore. That's strange.
<Cooper_> Photobucket is having a hissy fit it seems.
bengal has quit [Quit: Leaving]
<el3> igraltist, ARCH=arm CROSS_COMPILE=<toolchain-prefix> LOADADDR=0x40008000 make uImage dtbs <-- is toolchainprefix --> arm-linux-gnueabihf-??
<atsampson> el3: yup (assuming you have arm-linux-gnueabihf-gcc, etc.)
<el3> atsampson, okey I get error : arch/arm/kernel/asm-offsets.c:53:2: error: #error Your compiler is too buggy; it is known to miscompile kernels
<atsampson> Cooper_: I'm working on mainline u-boot/kernel support for the pcDuino v3 Nano, if that's of any use to you
<Cooper_> I noticed your name on the Nano's wiki page. Did a brief browse of the patches that page links to.
<el3> atsampson, I am actually building kernel for pcDuino3 right now :)
<atsampson> el3: you've got a known bad version of GCC, then (early 4.8?)
<Cooper_> Is there anything you can tell me about the process that might not yet be on the page? All I want from the kernel is networking, SATA with PMP support, and preferably USB. Don't care about the rest, including sound and video.
<el3> atsampson, okey probably.. I just typed: #apt-get install gcc-arm-linux-gnueabihf
libcg has quit [Quit: libcg]
<el3> atsampson, what should I type instead?
Black_Horseman has joined #linux-sunxi
<atsampson> el3: I'm using gcc-4.7-arm-linux-gnueabi from (or latest 4.9 built from source)
paulk-collins has quit [Quit: Quitte]
<el3> atsampson, okey thank you
<atsampson> Cooper_: the only problem I've seen so far is GMAC not working reliably at gigabit speeds -- but I've seen other people reporting it (as a lockdep-reported problem with stmmac), so I'm not convinced it's board-specific...
<Cooper_> atsampson: Hrmm... Okay, something to keep an eye on when I venture into these territories. Thanks for the heads-up. Are you using Linus' git, an rc or a regular release?
<Cooper_> (still compiling gdb...)
<atsampson> Cooper_: I'm using mripard's sunxi-next as a base
<Cooper_> atsampson: Ah, okay. Any particular reason you picked that one?
<atsampson> Cooper_: because it's what recommends for development ;-)
<atsampson> (it includes the sunxi patches queued up for the next release)
<Cooper_> atsampson: Lol! Well, that's as good a reason as any. :)
gianMOD has quit [Remote host closed the connection]
f15h has quit [Remote host closed the connection]
gianMOD_ has joined #linux-sunxi
<Cooper_> Finally! It's past midnight here and I've got to be up at 6:30 so tomorrow's gonna be fun, but I've got gdb so let's see what we can see.
iamfrankenstein has quit [Quit: iamfrankenstein]
<Cooper_> This is the output from 'info all-registers':
<Cooper_> There's not much in the way of functions with the frame pointer omitted...
<el3> atsampson, how do I install the toolchain?
gianMOD_ has quit [Ping timeout: 246 seconds]
gianMOD has joined #linux-sunxi
trollinger has quit [Quit: Reicht jetzt erst mal.]
<Cooper_> According to gdb's "layout asm" the instruction executing when things went off was "0x4e658 vminnm.f32 s28, s9, s31"
<Cooper_> A quick google says that's an ARMv8 instruction, mostly...?
gianMOD has quit [Ping timeout: 252 seconds]
<Cooper_> Problem is, I have no clue what to do next. My first suggestion would be to rebuild bash with that frame pointer around so the code block can be inspected.
gianMOD has joined #linux-sunxi
<hramrach> hello
<hramrach> I think my sd uart is not working:
<Cooper_> I really have to go and get some sleep. If anybody has any suggestions on how to go from here, I'm all ears. I'll keep an eye on the logs to see if anybody responds while I'm away, otherwise I'll be back again tomorrow. Thanks for all the help so far!
Cooper_ has quit []
xinj has quit [Ping timeout: 244 seconds]
<ssvb> Cooper_: it's hard to say anything about this vminnm.f32 instruction out of context, it would be very interesting to see the surrounding instructions
<ssvb> Cooper_: if it is indeed emitted by your compiler, then you need to review the CFLAGS you are using
gianMOD has quit [Ping timeout: 264 seconds]
gianMOD has joined #linux-sunxi
gianMOD has quit [Remote host closed the connection]
<el3> how do I install a toolchain that works?