There is some more information regarding SD card support in https://github.com/96boards/bugs/issues/4. I've documented some info I received from engineering.
On Tue, Mar 10, 2015 at 6:54 AM, Koen Kooi koen.kooi@linaro.org wrote:
Op 10 mrt. 2015, om 11:41 heeft Steve Taylor steve.taylor@linaro.org het volgende geschreven:
Hi all,
From: https://www.96boards.org/forums/topic/problem-with-microsd-card/#post-6695
There seems to be a conflict between the microSD card and other USB peripherals (in my particular case, a USB-ethernet dongle). When the USB-ethernet is connected, the kernel reports the following errors
[ 2571.527346] dwmmc_k3 f723e000.dwmmc1: failed to set rate 0Hz [ 2571.547337] dwmmc_k3 f723e000.dwmmc1: failed to set rate 400000Hz [ 2571.570917] dwmmc_k3 f723e000.dwmmc1: failed to set rate 400000Hz [ 2571.573227] dwmmc_k3 f723e000.dwmmc1: failed to set rate 400000Hz [ 2571.588534] dwmmc_k3 f723e000.dwmmc1: failed to set rate 400000Hz [ 2571.590143] dwmmc_k3 f723e000.dwmmc1: failed to set rate 0Hz
And the microSD card is not discovered. Absent the USB-ethernet dongle, it works ok.
I disagree, without the ethernet dongle I don't have reliable networking :)
Has anyone encountered this error,
Yes
and has any suggestion how it could be fixed?
It can be fixed with a proper clock driver by the looks of it. With 4.0.0-rc3 you'll also get tons of these:
Mar 10 10:49:47 hikey kernel: INFO: task kworker/u16:1:49 blocked for more than 120 seconds. Mar 10 10:49:47 hikey kernel: Tainted: G W 4.0.0-rc3 #11 Mar 10 10:49:47 hikey kernel: "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message. Mar 10 10:49:47 hikey kernel: kworker/u16:1 D ffffffc000086c00 0 49 2 0x00000000 Mar 10 10:49:47 hikey kernel: Workqueue: kmmcd mmc_rescan Mar 10 10:49:47 hikey kernel: Call trace: Mar 10 10:49:48 hikey kernel: [<ffffffc000086c00>] __switch_to+0x5c/0x68 Mar 10 10:49:48 hikey kernel: [<ffffffc0007e48b8>] __schedule+0x29c/0x6ac Mar 10 10:49:48 hikey kernel: [<ffffffc0007e4cf8>] schedule+0x30/0x8c Mar 10 10:49:48 hikey kernel: [<ffffffc0007e74d8>] schedule_timeout+0x140/0x1dc Mar 10 10:49:48 hikey kernel: [<ffffffc0007e5774>] wait_for_common+0xac/0x15c Mar 10 10:49:48 hikey kernel: [<ffffffc0007e5834>] wait_for_completion+0x10/0x1c Mar 10 10:49:48 hikey kernel: [<ffffffc0006088d0>] mmc_wait_for_req_done+0x40/0x140 Mar 10 10:49:48 hikey kernel: [<ffffffc000608a58>] mmc_wait_for_cmd+0x58/0x88 Mar 10 10:49:48 hikey kernel: [<ffffffc000613924>] mmc_io_rw_direct_host+0x78/0x138 Mar 10 10:49:48 hikey kernel: [<ffffffc000613e84>] sdio_reset+0x24/0x70 Mar 10 10:49:48 hikey kernel: [<ffffffc00060b5a0>] mmc_rescan+0x260/0x2f0 Mar 10 10:49:48 hikey kernel: [<ffffffc0000c2860>] process_one_work+0x140/0x348 Mar 10 10:49:48 hikey kernel: [<ffffffc0000c3318>] worker_thread+0x140/0x48c Mar 10 10:49:48 hikey kernel: [<ffffffc0000c7afc>] kthread+0xd8/0xf0
I did notice that the uSD card gets detected more often with 3.18 when using UEFI instead of the evil vendor loader. Still not 100% reliable, but a lot better. With UEFI+4.0.0-rc3 the kernel crashes with a uSD inserted.
-- Koen Kooi Builds and Baselines | Release Manager Linaro.org | Open source software for ARM SoCs
Dev mailing list Dev@lists.96boards.org https://lists.96boards.org/mailman/listinfo/dev