<Alex1269>
283: this function is not exported to modules. Is it possible to replace it with something ? I with my driver could work as module..
<Alex1269>
410 - it use 32 virtual irqs (A10) to demux gpio irq28. I choose irqs 97-129, is it correct choise?
<Alex1269>
This is final version of sunxi gpio driver...
<mnemoc>
q1) how do others deal with set_irq_flags() ?
<mnemoc>
q2) why making the SUNXI_ macros SUN4I only? the data, sure. but the macro functions? they are harmless and we will have to add data for other SoCs too
<Alex1269>
as I understand - they disable irq in modules...
gimli has quit [Read error: Operation timed out]
<mnemoc>
about the irq base, sounds fair... but i don't know what would be correct, comment this in the commit message so others can suggest a better base if wise
<Alex1269>
q2 - I'm not sure this registers same in other sunxi platforms. And I can't check... For example a13 have only 24 eint gpios
<mnemoc>
the most frustrating bit is that we can't distinguish A10S from A13 programatically yet
<mnemoc>
iirc the pio and eint registers are identical in sun4i and sun5i
<mnemoc>
even part of the same controller
<Alex1269>
Ok. This need to be tested. I found info about a13 eint and can add a table.
<mnemoc>
ignore sun5i/a13 in the first submit
<mnemoc>
we need to import the code to distiguish the different sun5i socs first
<mnemoc>
also, as this is new code, can we avoid typedef?
<Alex1269>
Ok. About set_irq_flags. I've seen a patch exporting this function to modules and a patch to remove this functions from drivers (they say it is useless). I tried to remove it and it doesn't work - gpiolib can't request virtual interrupts
<Alex1269>
enum... I'll remove it :)
<mnemoc>
adding one EXPORT_SYMBOL isn't such a big deal
<Alex1269>
So I'll add it via seaparate patch
<mnemoc>
yes, 1/N
<mnemoc>
(to not break bisecting)
<Alex1269>
What is 1/N ? :)
<mnemoc>
patch 1 of N
<Alex1269>
Ok..
<mnemoc>
git rebase -i is awesome :)
tinti has joined #arm-netbook
<Alex1269>
So I 1) remove enum 2) comment virq base in commit 3) move out SUNXI_ macroses 3) make export patch 1/2 4) make gpio patch 2/2
toxicpsion has quit [Remote host closed the connection]
Avernos_ has joined #arm-netbook
Avernos has quit [Ping timeout: 248 seconds]
GeorgeIoak has joined #arm-netbook
<GeorgeIoak>
another day and some more work ahead! i now have android running with a 15.6" LVDS panel thanks to tom
<GeorgeIoak>
doesn't look like it has the resistive touch screen driver loaded sun4i_ts
<GeorgeIoak>
lsmod (from the UART console) doesn't list it and insmod sun4i_ts says ir can't open sun4i_ts
<Turl>
GeorgeIoak: the kernel probably lacks the module
<Turl>
for insmod you need a full path
<Turl>
eg insmod /system/lib/modules/sun4i_ts.ko
<GeorgeIoak>
i hope it has it, he knew that's what i needed, i'll try the full path
merbzt has quit [Ping timeout: 264 seconds]
<GeorgeIoak>
either this cubie board has network problem or the image does because i can't connect to internet, even fixing the IP
<GeorgeIoak>
still trying to loacte the modules
<GeorgeIoak>
can't use find, what works in android
Adibek_ has quit [Quit: Page closed]
<Turl>
ls
<Turl>
busybox has find if you have it installed
<GeorgeIoak>
busybox doesn't seem to work, not really sure what shell i'm in on the uart console. seems limited whatever it is
<GeorgeIoak>
ls works but it's difficult to use because | more or piping doesn't seem to work
<GeorgeIoak>
with network i can't install a terminal
<Turl>
try running 'mksh'
<GeorgeIoak>
ok, that worked but the damn status messgaes keep coming onto the prompt but at least a step better
<GeorgeIoak>
kkkk
<GeorgeIoak>
system/lib has a bunch of libxxx.so but no .ko files or modules directory
hg_5 has quit [Ping timeout: 246 seconds]
vinifm has joined #arm-netbook
<GeorgeIoak>
i see a /sys/modules directory
<Turl>
/sys is a virtual filesystem created by the kernel, you won't find anything there
<GeorgeIoak>
oh, i found /sys/modules/sun4i_cs0 directory with files in it but no sun4i_ts directory. where should i look for the "real" modules then
<Turl>
I don't know where they are on your image, usually they're on /system/lib/modules
<GeorgeIoak>
he said he was building a standard tablet image. if i ls in the /system/libary directory i see a bunch of those libxxx.so files and the next file is soundfx which is a directpry
<GeorgeIoak>
could i possibly have more curve balls thrown at me??!
<Turl>
haha
<Turl>
if you had busybox you could busybox find / -name "*ko"
<GeorgeIoak>
oh wait, now that i'm in makesh i do have busybox
<GeorgeIoak>
Linux localhost 3.0.8+ #15 PREEMPT Sat Feb 9 23:42:57 CST 2013 armv7l GNU/Linux
<GeorgeIoak>
is what i have, do i just need to find a sun4i_ts.ko for that kernel, copy it over and insmod it?
<ssvb>
techn_: yes, I have seen that. Actually it is the driver for Mali-T604 from ARM Chromebook
<ssvb>
techn_: unfortunately it's a bit of a mess, with one git repository hosted by linaro people, another one hosted by chromium people (the linaro variant oopses the kernel on chromebook)
<ssvb>
techn_: and it has some OMAP remnants left in it, which indicates that it has been hacked in a hurry
<techn_>
ssvb: someone mentioned that mali stuff mainlining has started.. where I can find sources?