| 16:45 | ngochai | Hi guys, is there a reason why SDcard get corrupted? we have about 25 hummingboard i2ex having this problem even of a readonly rootfs (Ubuntu 14.04) |
| 22:37 | ngochai | hi guys, what is the MicroSD controller that hummingboard use? |
| 22:38 | Catwich | ngochai: what do you mean by controller? |
| 22:38 | Catwich | its part of the SoC |
| 22:38 | Catwich | and supposrts sdxc cards up to UHS-1 afaik |
| 22:39 | ngochai | Catwich, so what's the name? |
| 22:39 | ngochai | and what's Linux driver/module name for it? |
| 22:41 | mk01 | DRIVER=sdhci-esdhc-imx |
| 22:41 | mk01 | OF_NAME=usdhc |
| 22:41 | mk01 | OF_FULLNAME=/soc/aips-bus@02100000/usdhc@02194000 |
| 22:41 | mk01 | OF_COMPATIBLE_0=fsl,imx6q-usdhc |
| 22:41 | mk01 | OF_COMPATIBLE_N=1 |
| 22:41 | mk01 | OF_ALIAS_0=mmc0 |
| 22:41 | mk01 | MODALIAS=of:NusdhcTCfsl,imx6q-usdhc |
| 22:42 | ngochai | mk01, thanks |
| 22:43 | mk01 | welcome |
| 22:43 | mk01 | what are the cards you use ? |
| 22:43 | mk01 | is there any pattern among them ? |
| 22:44 | ngochai | Sandisk Ultra UHS-I 32GB |
| 22:44 | mk01 | ok |
| 22:44 | mk01 | yes |
| 22:45 | mk01 | axact hit |
| 22:45 | ngochai | mk01, ??? |
| 22:45 | mk01 | (card causing issues) |
| 22:45 | ngochai | really, what card should I use then? |
| 22:46 | mk01 | what kernel you use right now? |
| 22:46 | mk01 | the 3.14-*****-sr branch ? |
| 22:46 | ngochai | 3.14.14+ |
| 22:46 | mk01 | upgrade to the new tree |
| 22:47 | mk01 | there was lot of work done to make those cards work |
| 22:47 | ngochai | which new tree? |
| 22:47 | mk01 | https://github.com/SolidRun/linux-fslc |
| 22:49 | ngochai | is there any better card if I don't want to use the new kernel? |
| 22:55 | mk01 | personaly |
| 22:55 | mk01 | I don't have any other UHS cards |
| 22:55 | mk01 | just those three SanDisks for testing |
| 22:56 | mk01 | but general reports / feedbacks do speak about sandisks a lot - jnettlet was also mentioning some ultra-mega-special-speedy SAMSUNG cards |
| 22:57 | mk01 | so perhaps any others are fine (]assumption). |
| 22:57 | ngochai | so those samsung cards are better than the sandisk? |
| 22:57 | mk01 | no I mean - problem reports |
| 22:58 | mk01 | I don't remember other brands being reported (as causing problems) |
| 22:58 | ngochai | ok, thank you mk01 |
| 22:59 | mk01 | if you would go for the new tree, I run 4,8 and 16GB sandisks - in HB1 - HB2ex boxes |
| 22:59 | mk01 | and are fine now |
| 23:00 | ngochai | what's with the new tree, what's fslc by the way? |
| 23:00 | mk01 | fslc is freescale |
| 23:00 | mk01 | this is upstream repo - https://github.com/Freescale/linux-fslc |
| 23:01 | ngochai | ah I see |
| 23:02 | mk01 | basically it is very much safe to switch as it is the same 3.14 tree, just with many fixes on top of FSL + support for wider portfolio of imx6 devices |
| 23:02 | mk01 | (so not just solidrun). |
| 23:04 | mk01 | perhaps |
| 23:04 | mk01 | you could just merge the drivers/mmc commits on top of your 3.14.14+ you use now |
| 23:04 | mk01 | +one or two commits in dtb files |
| 23:05 | ngochai | well that may take a lot of time I imagine |
| 23:06 | ngochai | compiling linux-fslc now |
| 23:08 | mk01 | I was looking at the log difference, it is perhaps up to 10 commits |
| 23:09 | mk01 | from jnettlet and rmk |
| 23:10 | mk01 | but SolidRun/linux-fslc branch is regularly updated (as LTS kernel) - 3.14.14+ for some time not anymore |
| 23:11 | ngochai | alright, thanks |
| 23:11 | mk01 | welcome |