eebrah has quit [Remote host closed the connection]
eebrah has joined #linux-rockchip
Tony__ has joined #linux-rockchip
<Tony__>
naobsd, Hi. mmc_bus_get(host) get a wrong host name.
<Tony__>
drivers/mmc/core/core.c sdmmc_rescan
<Tony__>
naobsd, can you give me some advice ?
<Tony__>
Anybody else ?
mrcan_ has joined #linux-rockchip
<Tony__>
Hi, mrcan.
<Tony__>
My are ask you a question ?
mrcan has quit [Ping timeout: 264 seconds]
rperier has quit [Remote host closed the connection]
rperier has joined #linux-rockchip
<naobsd>
Tony__: what's "host name"?
<Tony__>
mmc host name.
<Tony__>
always is mmc0
<Tony__>
naobsd.
<naobsd>
Tony__: always? when both CONFIG_SDMMC0_RK29 and CONFIG_SDMMC1_RK29 are enabled?
<Tony__>
no, I mean that even if I disable platform_device_register(&device_sdmmc0)
<naobsd>
Tony__: how many sdmmc controller is enabled?
<naobsd>
if you only enable one controller, it's mmc0, first instance.
<Tony__>
naobsd, so wild.
<naobsd>
Tony__: sorry? I cannot understand
<Tony__>
okay, It seem that I fond some thing.
eebrah has quit [Remote host closed the connection]
<Tony__>
sdmmc_rescan_try_freq has a bug.
eebrah has joined #linux-rockchip
<Tony__>
naobsd, there is diff file :https://drive.google.com/file/d/0B6NGoH-iSqrKdFV5VjRpUzk0dmM/view?usp=sharing
<naobsd>
4.4.2 seems fine
<Tony__>
Yeah, My answer is same with your.
<naobsd>
btw
<naobsd>
here is community channel, not rockchip official support channel
<naobsd>
please don't expect "someone should answer"
<naobsd>
you can post question freely, but please don't call everyone at random...
<Tony__>
naobsd, okay. sorry.
<Tony__>
I just , i just can't describe the question that I met clear. I'm sorry. I do have a question.
<naobsd>
it's ok to post question
<naobsd>
but please don't expect that everyone can answer
<Tony__>
okay, okay. i know. thank you.
<Tony__>
My question is about mmc controler that is different than before.
<Tony__>
and ...
<rperier>
hi all
<Tony__>
I try to disable MMC0, but from dmesg I see that mmc_attach_sdio use "mmc0" only.
<Tony__>
I think that it should be "mmc1".
<Tony__>
because I DO disable "mmc0".
<naobsd>
I explained, 0 is number of first instance, not hardware unit number
<Tony__>
But I find that it is "mm1" at Android 4.4 kernel.
<Tony__>
On 4.4 kernel , I disable MMC0 that MMC0 can be used. and
<Tony__>
I disable MMC1, that MMC1 can be used.
<naobsd>
you said you got mmc0 when MMC0 is disabled
<Tony__>
every thing is logical.
<naobsd>
then, what's the problem...
<Tony__>
The problem is that both MMC0 MMC1 can't be used when i only disable MMC0.
<Tony__>
I'm sorry, my question is clear ?
<Tony__>
this time.
<Tony__>
there is log:https://drive.google.com/file/d/0B6NGoH-iSqrKeC14TlhWZTJLMW8/view?usp=sharing
<Tony__>
Can find this line:drivers/mmc/core/core.c...2015.. ===== mmc_rescan Begin....[mmc1]
<Tony__>
The is dmesg on Android 4.4.
arokux has quit [Remote host closed the connection]
arokux has joined #linux-rockchip
<Tony__>
this is that dmesg from androd 4.2:https://drive.google.com/file/d/0B6NGoH-iSqrKb3djSGFwMHBRSTg/view?usp=sharing
<Tony__>
disable MMC0 only.
<Tony__>
The last one has a problem, upload again:https://drive.google.com/file/d/0B6NGoH-iSqrKNDRVMUN0R051cDA/view?usp=sharing
<Tony__>
So I think that sdmmc_rescan maybe has a bug which about host name.
<naobsd>
I think you don't use rknand.ko in your 4.4 system
<naobsd>
emmc is registered as mmc0 in 4.4, so sdio is mmc1
<naobsd>
in your 4.2 system, rknand.ko is used, emmc is registered as mtd
<naobsd>
only one mmc, it's mmc0
<naobsd>
very logical
<Tony__>
But I suppose that mmc1 will works well if the host name is "mmc1" of sdmmc_rescan_try_freq.
<Tony__>
because mmc0 can works well when I only disable mmc1.
<naobsd>
I just explained number difference
<Tony__>
So I guess That is not because rknand.ko. I just guess. And I will try.
<naobsd>
I don't know how many lines expect sdio is mmc1 in 4.2 kernel source
<naobsd>
you need to fix bugs if you're using sdio as mmc0
<naobsd>
or you can avoid bugs by using sdio as mmc
<naobsd>
or you can avoid bugs by using sdio as mmc1
<naobsd>
please try
<naobsd>
no need to ask to try
<Tony__>
Yes, I will try. Clear my mind.
<naobsd>
btw why are you using 4.2?
<Tony__>
Yes, I'm using 4.2. 4.4 is just demonstration so far.
<Tony__>
demonstration -> *experiment
<Tony__>
I'm clear about rknand.ko before.
<Tony__>
about boot from sdcard.
<Tony__>
But the problem is same as boot from sdcard even if boot from eMMC.
arokux has quit [Remote host closed the connection]
arokux has joined #linux-rockchip
<naobsd>
if you use rknand.ko, eMMC/SD is recognized as mtd, and sdio is mmc0.
<naobsd>
if you disable rknand.ko, eMMC/SD is mmc0, sdio is mmc1
<naobsd>
"eMMC or SD" is not the problem
<Tony__>
I disagree that " if you use rknand.ko, eMMC/SD is recognized as mtd, and sdio is mmc0."
<Tony__>
I used rknand.ko, but my sdio is still mmc1.
<naobsd>
<Tony__> The last one has a problem, upload again:https://drive.google.com/file/d/0B6NGoH-iSqrKNDRVMUN0R051cDA/view?usp=sharing
<naobsd>
you confirmed sdio is mmc0
<Tony__>
Actually , I just disable mmc0. I haven't change sdio.
<Tony__>
So I think that logic is problem.
<naobsd>
you disabled controller number 0, I talked mmcX numbering in linux kernel
<naobsd>
I explained several times
<naobsd>
<naobsd> I explained, 0 is number of first instance, not hardware unit number
<naobsd>
please think logical, when I/you change kernel configuration, do you think hardware unit number(in hardware/document) is really changed? kernel configuration is just for software
<naobsd>
I didn't say hardware will be modified. hardware things, unit number, physical address, etc are always fixed.
<naobsd>
what you asked "mmc0/mmc1" is software thing
<Tony__>
When I disable SDMMC0_RK29, the mmc0(in kernel) will be mmc1(hardware).
<Tony__>
right ?
<naobsd>
right
wildea01 has joined #linux-rockchip
<naobsd>
<Tony__> naobsd, there is diff file :https://drive.google.com/file/d/0B6NGoH-iSqrKdFV5VjRpUzk0dmM/view?usp=sharing
<naobsd>
4.4 kernel uses hardware unit number, not software name"mmc0" "mmc1"
<naobsd>
^I just guessed "host_dev_id is hardware unit number", not confirmed
<Tony__>
damn, I'm a idiot.
<naobsd>
numbering rule may be vary by implelemtation
<naobsd>
^software
<naobsd>
implementation
<naobsd>
it's not true _for everything_
<naobsd>
I just explained mmc numbering in 4.2/4.4 kernel
<naobsd>
you may fix lines which refers "mmc0" "mmc1" name
<naobsd>
or you may avoid bugs by using "mmc1" name... disable rknand and enable SDMMC0_RK29, or change numbering rule for mmc
<Tony__>
yes, I will try. but you are pretty sweet. you told me the truth.
<karlp>
ah, this explains why I had some commentary that the wifi on my board (on sdio) was mmc0 but others referred to it as mmc1.
<karlp>
thanks for having this discussion!
<karlp>
(I don't think it's logical at all, renumbering things on physical ports, but that's just me)
<Tony__>
I think that the RK' code broken the logical.
<Tony__>
I think that the RK's code broken the logical.
<Tony__>
Yes, It works fine when I change the correct logical.
<Tony__>
Yeah!!!
<naobsd>
1st HDD is always get "sda" regardless of physical connection
<karlp>
naobsd: harddrives yes, but not usb for instance, nor graphics cards
<karlp>
and I think the harddrives was a mistake
<naobsd>
same for ethX
Tony__ has quit [Quit: 离开]
<karlp>
not any more
<karlp>
enp0s20u1u3 != eth0
<naobsd>
I know some distro uses bus/device related naming
<karlp>
it's what makes sense :)
<naobsd>
but I think many people dislike that name :)
<karlp>
anyway, regardless of my opinion, I now understand it better for the mmc
<naobsd>
I said that mmcX numbering rule is not _true for everything_
Tony_ has quit [Remote host closed the connection]
<naobsd>
btw I think usb hub number and usb device address (0-127) is also "cannot fix" thing
<naobsd>
what happen if usb ethere is connected to usb hub?
<naobsd>
if many hubs are connected, it gets longer name?
<karlp>
yep
<karlp>
which is what's expected if you change where it's plugged in...
<naobsd>
port0-port3-port2-...
<karlp>
anyway, out of schope for this conversation :)
<naobsd>
I just have interest :)
<karlp>
enp0s25 is my onboard ethernet, enp0s20u1u3 is a usb ethernet dongle on a docking station. enp0s20u3 is the same usb dongle directly into my laptop, enp0s20u4u1 is the same dongle on a hub on the laptop
<naobsd>
enp0s20u1u3 == pci bus 0 slot(device) 20 usb root port 1 next hub port 3 ?
<karlp>
something like that.
<naobsd>
very ... interesting ;)
<karlp>
s20u3 is usb3 on the laptop itself, s20uX are all the different root ports on the laptop itself.
<naobsd>
I sometimes but surely want fixed name...
<karlp>
fixed like what?
<naobsd>
for example, if I have 2 USB ethernet
<naobsd>
ah, well
<naobsd>
ethernet may not be good example, some ethX distro fixes X by udev and MAC address
<naobsd>
anyway
<naobsd>
for example, I want "this one is eth0" "this one is eth1"
<naobsd>
there is uuid for disk
<naobsd>
there is mac address for network
<naobsd>
recent distro will fix name as much as possible, so it shouldn't be a problem for most case
<naobsd>
probably name like enp0s20u1u3 is good
<naobsd>
e.g. I can exchange USB ethernet dongle easily when it's broken
Tony_ has joined #linux-rockchip
<Tony_>
In my opinion, the removable device can just use devName+$number.
<Tony_>
But the unremovable device such as mmc controler of CPU should be hard code.
naobsd has quit [Quit: Page closed]
Celia_ has joined #linux-rockchip
Astralix1 has left #linux-rockchip [#linux-rockchip]
Astralix has joined #linux-rockchip
Celia_ has quit [Remote host closed the connection]
Tony__ has joined #linux-rockchip
mazen has joined #linux-rockchip
mazen has left #linux-rockchip [#linux-rockchip]
Tony__ has quit [Quit: Page closed]
<karlp>
mmind00: My gut wouldbe that internal pullups for i2c should be disabled by default, though I don't care much for Doug's chromebook bug as a reason, that's just a problemwith their hardware, period.
<karlp>
on micros i use the pullups are either absent or off by default.
<rperier>
Mark Brown has a serious ego problem no? (just read his replies for some reviews... ^^)
<mmind00>
rperier: where did you get that impression? I don't share that sentiment and got along great with Mark in the past
Tony_ has quit [Remote host closed the connection]
<mmind00>
karlp: you could answer to the mail for that ... so I don't have to remember that there also was a response on irc ;-)
rz2k_ has joined #linux-rockchip
<karlp>
sure, even though I'm really just here in the peanut gallery :)
<rperier>
mmind00: I don't know him personnally, but for most of his replies he looks cold with people... I can be wrong of course. However, this is not easy to understand people's feeling and emotion through an email.
ganbold_ has quit [Ping timeout: 255 seconds]
<rperier>
well, that's not a good thing to judge people without knowing them.
<rperier>
karlp: feel free to answer to the discussion on the ML ;)
<rperier>
important thing being that everything is written, no need to remember them
ganbold_ has joined #linux-rockchip
nighty^ has joined #linux-rockchip
eebrah has quit [Quit: leaving]
<karlp>
yeah, I have.
GriefNorth has quit [Ping timeout: 245 seconds]
<mmind00>
rperier: in general don't interpret to much into kernel mails ... especially if "courtesy phrases" are missing
<mmind00>
rperier: for one these people write dozens to hundreds of mails each day, so it comes down to transporting the _necessary_ informations
<rperier>
there is an very important workflow too... you're right...
<mmind00>
rperier: and on the other hand with to much prose around the core stuff you can puzzle non-native speakers from different cultures
<mmind00>
and of course recipients have to read any additional prose too ;-)
<mmind00>
rperier: directly offensive stuff would be different - but short simple "cold" messages are the norm
<rperier>
I really don't care about "courtesy phrases" and "prose" , that's not the point. That's just something like "This is a single, standalone SPI patch - you didn't send it as part of a
<rperier>
series (which is the only reason I read it)." might be misinterpreted.
<rperier>
well, different cultures is also something to take into consideration (which is very important for worlwide project like linux)
<mmind00>
karlp: thanks - especially for the added detail :-)
<karlp>
I just went and checked a few datasheets I had here, the i2c sections of most of them, if they even metnion pull ups, just say that "external pull ups are required"
<karlp>
you can sometimes fake it by using the regular gpio pullups, and iirc the arduino i2c library does that, but the defualt state is definitely not pulled up
<rperier>
anyway, as I said, I don't like to judge people without knowing them, what I said is not a real argument, just a feel (which is wrong apparently, so everything is alright). /me goes drink coffee
<mmind00>
rperier: for getting to know people, the LinuxCon Europe / embedded Linux conference Europe should be somewhat near you ... next year in Dublin ;-)
<rperier>
I started discussions with "free-electrons", perhaps I could work with them. Who knows, next year perhaps. ;) (anyway, I can come to these conferences myself)
<rperier>
(I mean personnally, without a company)
<karlp>
I'd love to go and just hang out at _any_ professional meeting, we don't have any here, and we don't get any companies even offering their self promoting training events in iceland :|
<wildea01>
I'd be up for a conference in iceland!
rz2k_ has quit []
<karlp>
there's a few agencies here trying to promote iceland as a conference destination for being "between europe and america"
<hramrach_>
it's away from both europe and america. about same far away for the usual distance perception of inhabitants of either continent I would guess
<karlp>
apparently some conferences have political infighting if it's always in one groups home bse or somthing, and they get to be "exotic" and go away together if they come to iceland
* karlp
shrugs
bengal has joined #linux-rockchip
eebrah_ has joined #linux-rockchip
eebrah_ has quit [Client Quit]
eebrah__ has joined #linux-rockchip
eebrah__ is now known as eebrah|legit
gb_master has joined #linux-rockchip
Qi999 has joined #linux-rockchip
<rperier>
mmind00: now that initial ethernet support for rk3066 is sent for review, I think I will start to work on phy-usb for rk3066/rk3188. Except if there is something more critical...
<rperier>
(I don't think so, so usb-phy driver looks a good idea)
<mmind00>
rperier: correct I think ... just take what area you like :-)
<rperier>
okay cool, so I will add phy-usb-drv and "rk1000" as next steps on the wiki (the todolist)
mrcan__ has joined #linux-rockchip
mrcan_ has quit [Ping timeout: 264 seconds]
Bludot has joined #linux-rockchip
mrcan__ is now known as mrcan
markm_ has joined #linux-rockchip
markm_ has quit [Client Quit]
wildea01 has quit [Quit: leaving]
<cristian_c>
Hello
<cristian_c>
What methods can I use to know if my bootloader is proprietary or not?
<cristian_c>
Any ideas?
prop81 has quit [Ping timeout: 246 seconds]
cristian__c has joined #linux-rockchip
mrcan has quit [Ping timeout: 255 seconds]
mrcan has joined #linux-rockchip
mrcan has quit [Changing host]
mrcan has joined #linux-rockchip
cristian_c has quit [Ping timeout: 260 seconds]
cristian__c is now known as cristian_c
cristian_c is now known as Guest56863
Guest56863 is now known as cristian_c
mrcan has quit [Ping timeout: 258 seconds]
Qi999 has quit [Ping timeout: 256 seconds]
cristian__c has joined #linux-rockchip
cristian_c has quit [Ping timeout: 244 seconds]
cristian__c has quit [Ping timeout: 240 seconds]
cristian___c has joined #linux-rockchip
cristian_c has joined #linux-rockchip
cristian___c has quit [Ping timeout: 256 seconds]
cristian__c has joined #linux-rockchip
cristian_c is now known as Guest82547
Guest82547 has quit [Ping timeout: 264 seconds]
cristian__c has quit [Client Quit]
cristian__c has joined #linux-rockchip
cristian__c is now known as cristian_c
cristian__c has joined #linux-rockchip
cristian_c is now known as Guest57470
cristian__c has quit [Client Quit]
cristian__c has joined #linux-rockchip
Guest57470 has quit [Ping timeout: 272 seconds]
cristian__c has quit [Client Quit]
cristian__c has joined #linux-rockchip
cristian__c is now known as cristian_c
AstralixNB1 has left #linux-rockchip [#linux-rockchip]
cristian__c has joined #linux-rockchip
cristian_c is now known as Guest24995
cristian__c is now known as cristian_c
Guest24995 has quit [Ping timeout: 255 seconds]
amstan has joined #linux-rockchip
gb_master has quit [Remote host closed the connection]
<amstan>
mmind00: hey, i see you're up
<mmind00>
amstan: it's only 11pm here ;-)
<amstan>
mmind00: i have a few concerns about the clock phase, notably that there's no precise way to convert between the delay elements(the fine tuning) to degrees
cristian__c has joined #linux-rockchip
<amstan>
the trm states that the delay elements are between 40-80ps, but that's not accurate enough to get a degree out of them
<amstan>
that's fine for tuning, but it's not really compatible with the framework(since it assumes we know the exact degrees)
cristian_c is now known as Guest69443
cristian__c is now known as cristian_c
Guest69443 has quit [Ping timeout: 265 seconds]
cristian__c has joined #linux-rockchip
cristian_c is now known as Guest27191
Guest27191 has quit [Ping timeout: 258 seconds]
cristian__c is now known as cristian_c
cyrozap has quit [Ping timeout: 244 seconds]
cristian__c has joined #linux-rockchip
cristian__c has quit [Read error: Connection reset by peer]
cristian__c has joined #linux-rockchip
cristian_c has quit [Ping timeout: 244 seconds]
cristian__c is now known as cristian_c
cristian_c is now known as Guest93060
cristian__c has joined #linux-rockchip
cristian__c is now known as cristian_c
<mmind00>
amstan: dianders said something similar, but so far I only had passing glances at the whole tuning stuff - so I guess I won't be much help there
Guest93060 has quit [Ping timeout: 264 seconds]
cristian__c has joined #linux-rockchip
cristian_c is now known as Guest72779
Guest72779 has quit [Ping timeout: 245 seconds]
cristian__c is now known as cristian_c
cristian__c has joined #linux-rockchip
cristian_c has quit [Ping timeout: 245 seconds]
bengal has quit [Quit: Leaving]
cristian__c has quit [Client Quit]
cyrozap has joined #linux-rockchip
nighty^ has quit [Quit: Disappears in a puff of smoke]