cnxsoft has quit [Read error: Connection reset by peer]
leviathan has quit [Remote host closed the connection]
cnxsoft has joined #linux-sunxi
Net147 has quit [Ping timeout: 260 seconds]
f0xx has quit [Quit: terminated!]
<dan0_0>
Buddy_Zhang1: Do any of Xunlong's Linux images do wear leveling/bad block detection for MicroSD cards or the onboard eMMC?
Net147 has joined #linux-sunxi
f0xx has joined #linux-sunxi
<MoeIcenowy>
dan0_0: this is done by SD/MMC controller on the card
leviathan has joined #linux-sunxi
<MoeIcenowy>
nothing can be done by the CPU
<MoeIcenowy>
the controller hid many details
<dan0_0>
Mmm
<dan0_0>
Can anything be done to stave off corruption?
<MoeIcenowy>
buy a better card
<dan0_0>
Beyond just making the root partition read only
<MoeIcenowy>
or maybe you can try to do trim operation on it
<MoeIcenowy>
fstrim /
<dan0_0>
lol, I've only encountered it with bad cards iirc, but just fearful about putting a Opi+ 2e into production and having its eMMC fail/corrupt and take data with it
<MoeIcenowy>
what's the model of the eMMC on it?
<dan0_0>
Planning to bundle an evo 32gb that has been benched with each one
<dan0_0>
one sec, I'll go lock
<dan0_0>
*look
<MoeIcenowy>
the CPU or CPU card controller can do nothing about card corruption
<MoeIcenowy>
trim operation can do a little work on decreasing corruption
<MoeIcenowy>
but it's the most what can CPU do
<MoeIcenowy>
and you should hope your card supports trim ;-)
<MoeIcenowy>
(although my $5 Kingston ones support it ;-)
<dan0_0>
I should go test trim support on mSD and eMMC...
Mr__Anderson has joined #linux-sunxi
<KotCzarny>
yeah, limiting writes helps
<dan0_0>
wrt eMMC on this OrangePi+ 2e, its a Samsung 601 KLMAG2WEPD-B031, with extra Samsung brand explosions thrown in for fun :P
<MoeIcenowy>
NAND storage gained instability while increasing capacity ;-)
<dan0_0>
mmm
<MoeIcenowy>
this card is a TLC ;-)
<MoeIcenowy>
(but currently usually MicroSD cards are TLC ;-)
f0xx has quit [Quit: terminated!]
<MoeIcenowy>
(and I think Samsung can do good hardware FTL ;-)
<KotCzarny>
lol
<KotCzarny>
emmc in opi+2e is tlc?
<BenG83_PB>
MoeIcenowy, is there a mainline driver for RTL8723xx?
f0xx has joined #linux-sunxi
<MoeIcenowy>
for BS there's one
LargePrime has quit [Ping timeout: 255 seconds]
<MoeIcenowy>
oh I mean mainline-capable out-of-tree driver
<MoeIcenowy>
for AU the rtl8xxxu driver can support it ;-)
<BenG83_PB>
BSP kernel has both a 8723bs and cs, but I accidently used the bs driver and that made no difference so far
<BenG83_PB>
is there an audiocodec wip tree somewhere?
<BenG83_PB>
for A64?
<MoeIcenowy>
I remember codekipper said that he will do it ;-)
<dan0_0>
Yea, I guess my main worry is idiot users killing power repeatedly and inducing corruption
<MoeIcenowy>
P.S. after my DE2 driver for V3s is merged I will go to make a kernel-side driver for Pinebook display pipeline ;-)
<MoeIcenowy>
BenG83_PB: thanks for your info, I will try it ;-)
LargePrime has joined #linux-sunxi
<MoeIcenowy>
oh thr 8723bs driver do not recognize 8723cs
<MoeIcenowy>
Foud unrecognized device 0xb703 for vendor 0x24c
leviathancn has joined #linux-sunxi
<KotCzarny>
moeicenowy: add id to the source
<MoeIcenowy>
KotCzarny: trying
<MoeIcenowy>
wil try to make firmware binaries for 8723cs
<MoeIcenowy>
yes the current 8723bs driver do not work for cs.
<KotCzarny>
o.o
<MoeIcenowy>
that's what I expected ;-)
<KotCzarny>
i wonder why they haven't invented unified driver yet
<KotCzarny>
as most of the things are the same for all those of revisions
Buddy_Zhang1 has quit [Quit: Page closed]
lkcl has quit [Ping timeout: 260 seconds]
phipli has joined #linux-sunxi
<MoeIcenowy>
modified BSP driver to work on Pinebook linux-next ;-)
<codekipper>
has anybody here every experienced any boards with broken ram?
<codekipper>
for some reason my A20-SOM-EVB is refusing to netboot....uboot complains that kernel fails CRC.
<codekipper>
it's worked as recently as this week and the uImage works with other boards.
<KotCzarny>
codekipper, lower dram freq as a first checkup?
<codekipper>
changing to zImage seemed to get it pass the CRC check but the kernel didn't start
<codekipper>
KotCzarny: thanks...I'll try that...it's a bitch of an sdcard to remove
leviathancn has quit [Remote host closed the connection]
leviathan has quit [Remote host closed the connection]
kristina_ is now known as kristina
BenG83 has joined #linux-sunxi
kristina has quit [Quit: leaving]
hojnikb has joined #linux-sunxi
<hojnikb>
MoeIcenowy: trim won't help with data curruption. If you can even find an sd card that supports trim, it's only going to help with write performance and write amplification. Trim will do nothing for data reliability
<hojnikb>
Trim can actually be disastrous if not implemented right. Not that long ago for example, patriot had huge issues with their phison drives, because trim command was implemented poorly in firmware
<hojnikb>
This translated to huge drive failure and almost always data corruption.
kristina has joined #linux-sunxi
<hojnikb>
same thing with intel and samsung drives under linux; because linux handles trim differently than windows, there were issues and those drives had to be blacklisted.
<hojnikb>
KotCzarny>emmc in opi+2e is tlc?
<hojnikb>
It's MLC actually. 16GB tlc emmc would have atrocious write speeds.
lkcl has joined #linux-sunxi
hojnikb has quit [Quit: Page closed]
<MoeIcenowy>
Samsung says it's "3-bit MLC" ...
<KotCzarny>
marke3ting, hehe
<KotCzarny>
tlc got bad mouthed
<kristina>
anyone got the code to generate and sign a secure boot container for a64 once the security fuse is blown?
<MoeIcenowy>
use local-mac-address from dt (set by U-Boot)
hojnikb has quit [Quit: Page closed]
mossroy has joined #linux-sunxi
menomc is now known as mnemoc
ErwinH has joined #linux-sunxi
codekipper has quit [Quit: Page closed]
<BenG83_PB>
MoeIcenowy, thanks
<MoeIcenowy>
but currently mainline kernel do not support RSB... makes it now impossible to monite battery
<BenG83_PB>
so we run AXP with ATF settings?
<MoeIcenowy>
yes
<BenG83_PB>
ok
<BenG83_PB>
thats something I want to look in anyways
<MoeIcenowy>
my modified ATF enables necessary power
<BenG83_PB>
so that it charges with full 3A if not powered
<MoeIcenowy>
dldo2, fldo1 for lcd panel; gpio0ldo for usb
<BenG83_PB>
e.g. set charging limit on power-off or sth
<MoeIcenowy>
P.S. I think we need a configuration menu in U-Boot for PMIC ;-)
<BenG83_PB>
that would be awesome
<BenG83_PB>
the PB PCB has testpads for CPU_S_URX/TX
<BenG83_PB>
is there anything interesting what one can do with the UART?
<MoeIcenowy>
I think these pins are already used for powering up wifi
<BenG83_PB>
yes I saw that, I was just wondering if that UART has some other special function
IgorPec has quit [Ping timeout: 260 seconds]
scream has joined #linux-sunxi
IgorPec has joined #linux-sunxi
Andy-D has joined #linux-sunxi
leviathancn has joined #linux-sunxi
jernej has joined #linux-sunxi
jernej has quit [Quit: Konversation terminated!]
ErwinH has quit [Remote host closed the connection]
jernej has joined #linux-sunxi
IgorPec has quit [Ping timeout: 240 seconds]
Pepe has quit [Read error: Connection reset by peer]
Pepe has joined #linux-sunxi
IgorPec has joined #linux-sunxi
Net147 has quit [Ping timeout: 240 seconds]
Net147 has joined #linux-sunxi
reinforce has joined #linux-sunxi
ErwinH has joined #linux-sunxi
vinimac has joined #linux-sunxi
lkcl has quit [Ping timeout: 260 seconds]
ErwinH has quit [Ping timeout: 255 seconds]
lurchi__ has joined #linux-sunxi
ErwinH has joined #linux-sunxi
astephanh has quit [Ping timeout: 260 seconds]
ErwinH has quit [Ping timeout: 255 seconds]
phipli has quit [Ping timeout: 255 seconds]
ErwinH has joined #linux-sunxi
astephanh has joined #linux-sunxi
Net147 has quit [Ping timeout: 240 seconds]
ErwinH has quit [Ping timeout: 260 seconds]
Net147 has joined #linux-sunxi
IgorPec4 has joined #linux-sunxi
afaerber has quit [Ping timeout: 240 seconds]
IgorPec2 has joined #linux-sunxi
IgorPec has quit [Ping timeout: 255 seconds]
jernej has quit [Ping timeout: 260 seconds]
IgorPec4 has quit [Ping timeout: 255 seconds]
afaerber has joined #linux-sunxi
fdcx has joined #linux-sunxi
Pepe has quit [Quit: WeeChat 1.4]
Pepe has joined #linux-sunxi
Pepe has quit [Client Quit]
Pepe has joined #linux-sunxi
phipli has joined #linux-sunxi
cptG_ has joined #linux-sunxi
cptG has quit [Ping timeout: 240 seconds]
Net147 has quit [Read error: Connection reset by peer]
|Jeroen| has quit [Quit: dada]
Pepe has quit [Quit: WeeChat 1.4]
Pepe has joined #linux-sunxi
Net147 has joined #linux-sunxi
cnxsoft has quit [Quit: cnxsoft]
vinimac has quit [Quit: Saindo]
BenG83_PB has quit [Quit: Leaving]
Pepe has quit [Quit: WeeChat 1.4]
Pepe has joined #linux-sunxi
Pepe has quit [Client Quit]
Pepe has joined #linux-sunxi
<MoeIcenowy>
apritzel: my Pinebook have 2GiB on PCB
<MoeIcenowy>
and also 2GiB seen in Android
<MoeIcenowy>
it seems to be dual rank DRAM, but our current code supports only one rank
IgorPec2 has quit [Read error: Connection reset by peer]
Pepe has quit [Quit: WeeChat 1.4]
IgorPec has joined #linux-sunxi
chomwitt has joined #linux-sunxi
matthias_bgg has joined #linux-sunxi
dave0x6d has joined #linux-sunxi
scelestic has quit [Quit: leaving]
Pepe has joined #linux-sunxi
matthias_bgg has quit [Quit: Leaving]
Pepe has quit [Quit: WeeChat 1.4]
<willmore>
MoeIcenowy, oh, that's great to hear. Once they stopped using a mailing list for communication, I lost track of that project.
Pepe has joined #linux-sunxi
IgorPec has quit [Ping timeout: 268 seconds]
<MoeIcenowy>
?
lurchi__ has quit [Ping timeout: 240 seconds]
<MoeIcenowy>
apritzel: just change initial dual_rank value from 0 to 1 works.
LargePrime has quit [Ping timeout: 240 seconds]
<willmore>
I used to be on the ath wireless mailing list but they shut that down about a year ago.
<willmore>
So, i've lost track of development on the 9k family and newer.
<KotCzarny>
moeicenowy: would it be possible to add resolution changing in simplefb?
<MoeIcenowy>
impossible
<KotCzarny>
not even via some hack?
<MoeIcenowy>
I think swo
<MoeIcenowy>
I think so
<willmore>
And hojnikb is wrong about trim not helping with corruption. Since we're talking about write corruption due to worn out flash blocks and trim *will* help decrease write amplification and make wear leveling more effective.
<KotCzarny>
hw or simplefb limitation?
chomwitt has quit [Quit: WeeChat 1.0.1]
scelestic has joined #linux-sunxi
<wens_>
KotCzarny: simplefb does not know about hardware
<wens_>
it's simply a framebuffer
wens_ is now known as wens
<wens>
KotCzarny: you can't change the resolution without reconfiguring the hardware
<MoeIcenowy>
theortically the resolution can be changed in U-Boot
<KotCzarny>
wens, ok, let me rephrase my question, would it be possible to make some tool to reconfigure hw and simplefb ?
<KotCzarny>
without rebooting/rewriting uboot
<wens>
KotCzarny: yes, but that's a hack
<MoeIcenowy>
as U-Boot knows about hardware -- it's U-Boot who sets simplefb device node for Linux
<wens>
KotCzarny: and if you're already reconfiguring the hardware, why not just do a kms driver
<KotCzarny>
wens, kms driver?
<wens>
KotCzarny: drm/kms driver
<KotCzarny>
does it know/use g2d? has some neon accel etc?
<wens>
hmm, also there's no way to signal simplefb for the resolution change
LargePrime has joined #linux-sunxi
<wens>
nope
<wens>
as i said, simplefb is just a framebuffer, or fbdev
<wens>
all the kernel knows about is the chunk of memory, and the resolution and color depth
<wens>
nothing more
<wens>
a chuck of memory is allotted, which the display hardware scans out
<wens>
how you use it from userspace (like with neon accel in xf86-fbturbo) is up to you
<MoeIcenowy>
I think KotCzarny asks whether DRM driver have acceleration
<MoeIcenowy>
DRM driver have also no acceleration, except hardware layer support
<wens>
KotCzarny: why would the drm driver do acceleration? it's just display hardware?
<KotCzarny>
if drm is display rendering manager then it should know how to move bits of screen fast
<MoeIcenowy>
in the situation of sun4i-drm we should better call it a KMS driver
<MoeIcenowy>
as it do not know a lot about "render"
<MoeIcenowy>
it only knows how to put a framebuffer to the TCON ;-)
lurchi__ has joined #linux-sunxi
<wens>
KotCzarny: the display hardware does dma to stream and composite layers or framebuffers together
<wens>
i'm not sure how drm exposes the framebuffers and layers to userspace, but presumably as blocks of memory the application updates
<wens>
so the drm driver does not move bits around
<wens>
KotCzarny: you should read up on KMS
IgorPec has joined #linux-sunxi
LargePrime has quit [Ping timeout: 260 seconds]
LargePrime has joined #linux-sunxi
lkcl has joined #linux-sunxi
phipli has quit [Ping timeout: 255 seconds]
IgorPec has quit [Ping timeout: 260 seconds]
sunxi_fan has joined #linux-sunxi
BenG83_PB has joined #linux-sunxi
sunxi_fan1 has quit [Ping timeout: 260 seconds]
jernej has joined #linux-sunxi
dave0x6d has quit [Quit: Connection closed for inactivity]
Mr__Anderson has quit [Ping timeout: 255 seconds]
<FergusL>
Are you using nmtui to create access points successfully? I've created one but it doesn't activate, Activating it times out after a while
sunxi_fan1 has joined #linux-sunxi
leio has joined #linux-sunxi
sunxi_fan has quit [Ping timeout: 260 seconds]
lurchi__ has quit [Ping timeout: 240 seconds]
vishnup has quit [Ping timeout: 255 seconds]
Mr__Anderson has joined #linux-sunxi
jstein_ has joined #linux-sunxi
IgorPec has joined #linux-sunxi
dave0x6d has joined #linux-sunxi
jstein_ is now known as jstein
leio has quit [Ping timeout: 240 seconds]
leio has joined #linux-sunxi
jstein has quit [Remote host closed the connection]
hojnikb has joined #linux-sunxi
<hojnikb>
willmore: Actually, it doesn't work like that. Again, trim won't help you reduce curruption.
<hojnikb>
Curruption happens when controller doesn't do it's job
mossroy has quit [Quit: Leaving]
<KotCzarny>
or when radon levels in the air are too high
<hojnikb>
doesn't really matter if you're trimming or not. When controller doesn't handle errors properly (and there are LOTS of errors with flash) you get curruption.
<KotCzarny>
wihch could be an issue in highlands
<MoeIcenowy>
or when the NAND went to a status that controller cannot handle ;-)
<hojnikb>
Truth be told, if you want absolute reliability, you'll stay clear of anything flash based
<MoeIcenowy>
KotCzarny: I have never heard this kind of info ;-)
<MoeIcenowy>
hojnikb: all storage devices have its disadvantage
<KotCzarny>
my friend who lives in highland area has all kind of fun with radon radiation
<hojnikb>
of course, but you do have to be aware of them
<KotCzarny>
random memory errors on otherwise good hw
<MoeIcenowy>
he needs ECC memory ;-)
<KotCzarny>
hehe
<KotCzarny>
show me a laptop wuth ecc mem ;)
<KotCzarny>
*with
<MoeIcenowy>
where does he live...
<hojnikb>
there are some xeon laptops, that should support ecc
<MoeIcenowy>
does he faced more frequent NAND flash corruption?
<hojnikb>
personally, i had a LOT of issues, when i ran rootfs off flash drives
<hojnikb>
on my NAS
<KotCzarny>
nah, it was long time since i last chatted with him
<hojnikb>
it was a constant nightmare of curruption
<KotCzarny>
but it makes you think that environment in which you live is important for electronics stability
<hojnikb>
went through 5 sticks until i said fuck it and moved everything to good ol harddrive
<MoeIcenowy>
seems not so high...
<MoeIcenowy>
what I though when you says radiation-related memory corruption is somewhere like Tibet ;-)
<KotCzarny>
nah, its most likely related to coal mines and natural deposits of radon in the area
orly_owl has quit [Read error: Connection reset by peer]
hojnikb has quit [Quit: Page closed]
ErwinH has joined #linux-sunxi
<plaes>
ha.. OWA is One Wire Audio...
<MoeIcenowy>
a.k.a. SPDIF.
<plaes>
yup
ErwinH has quit [Ping timeout: 240 seconds]
orly_owl has joined #linux-sunxi
LargePrime has quit [Ping timeout: 240 seconds]
premoboss has joined #linux-sunxi
<KotCzarny>
hojnikb: usb sticks arent made for stability..
lurchi__ has joined #linux-sunxi
<willmore>
hojnikb if the corruption is dues to flash wear out, then trim would have helped as it would have lowered the write amplification and allowed the wear leveling to work. Since this is the type of corruption we were talking about, I continue to assert that trim is beneficial.
<TheLinuxBug>
'usb sticks arent made for stability..' This for sure... I tested capturing video from rtsp and recording to a 64gb USB stick and twice I had it totally corrupt its self and crash the sbc... that just one example though
<TheLinuxBug>
some of the cheaper modules get too hot, especially with some of the cases that are on them
<willmore>
If the corruption is due to firmware bugs or communication errors, then yeah, trim won't help.
LargePrime has joined #linux-sunxi
LargePrime has quit [Ping timeout: 240 seconds]
wzyy2 has quit [Ping timeout: 240 seconds]
LargePrime has joined #linux-sunxi
lkcl has quit [Ping timeout: 260 seconds]
wzyy2 has joined #linux-sunxi
GrimKriegor_ has joined #linux-sunxi
GrimKriegor has quit [Ping timeout: 268 seconds]
BenG83_ has joined #linux-sunxi
BenG83 has quit [Ping timeout: 255 seconds]
lurchi__ has quit [Ping timeout: 255 seconds]
chomwitt has joined #linux-sunxi
hojnikb has joined #linux-sunxi
<hojnikb>
willmore: if controller firmware is so fucked, that it need the help of trim to work properly, then this flash device is useless
<hojnikb>
i agree that trim does potentially help reducing wear on flash, this still doesn't help the fact, that if corruption occur, it was the controllers fault not absence of trim.
<hojnikb>
errors happen all the time on modern flash and if controller can't handle the error, trim sure as help won't either
<hojnikb>
in my opinion having no trim on sd cards is a good thing really.
IgorPec has quit [Ping timeout: 268 seconds]
<hojnikb>
help avoid fucked up implementations from noname emmc makers
premoboss has quit [Ping timeout: 260 seconds]
<hojnikb>
MoeIcenowy: can you tell us when pinebook will be available ?
hojnikb has quit [Quit: Page closed]
Mr__Anderson has quit [Quit: Leaving.]
leviathancn has quit [Remote host closed the connection]
jernej has quit [Ping timeout: 260 seconds]
lurchi__ has joined #linux-sunxi
terra854 has quit [Quit: Connection closed for inactivity]