+ Adrian On 4 October 2017 at 17:00, Carlo Caione <carlo@xxxxxxxxxxxx> wrote: > Haye, > I'm working on a Cherry-Trail laptop shipping an SDHCI controller on > ACPI [80860F14:03]. > The problem is that on this hardware the controller seems to be > particularly picky about the type of SD cards it likes. In particular: > > - With ultra high speed DDR50 SDHC cards I see no problems > - With ultra high speed SDR104 SDHC cards as soon as the card is inserted I get: > > mmc2: Tuning timeout, falling back to fixed sampling clock > mmc2: new ultra high speed SDR104 SDHC card at address 59b4 > mmcblk2: mmc2:59b4 USD00 15.0 GiB > mmc2: Timeout waiting for hardware interrupt. > mmc2: sdhci: ============ SDHCI REGISTER DUMP =========== > mmc2: sdhci: Sys addr: 0x00000008 | Version: 0x00001002 > mmc2: sdhci: Blk size: 0x00007200 | Blk cnt: 0x00000008 > mmc2: sdhci: Argument: 0x00000000 | Trn mode: 0x0000003b > mmc2: sdhci: Present: 0x01ff0001 | Host ctl: 0x00000017 > mmc2: sdhci: Power: 0x0000000f | Blk gap: 0x00000080 > mmc2: sdhci: Wake-up: 0x00000000 | Clock: 0x00000007 > mmc2: sdhci: Timeout: 0x0000000a | Int stat: 0x00000000 > mmc2: sdhci: Int enab: 0x02ff008b | Sig enab: 0x02ff008b > mmc2: sdhci: AC12 err: 0x00000000 | Slot int: 0x00000000 > mmc2: sdhci: Caps: 0x0568c8b2 | Caps_1: 0x00000807 > mmc2: sdhci: Cmd: 0x0000123a | Max curr: 0x00000000 > mmc2: sdhci: Resp[0]: 0x00000000 | Resp[1]: 0x0077dd7f > mmc2: sdhci: Resp[2]: 0x325b5900 | Resp[3]: 0x00000000 > mmc2: sdhci: Host ctl2: 0x0000800b > mmc2: sdhci: ADMA Err: 0x00000000 | ADMA Ptr: 0x335c9200 > mmc2: sdhci: ============================================ > mmcblk2: error -110 sending status command, retrying > mmcblk2: error -110 sending status command, retrying > mmcblk2: error -110 sending status command, aborting > > reading and writing is painfully slow. > > - With high speed SDHC cards I get a weird behavior where apparently > there is no problem but it's impossible to format the card. mkfs.ext4 > gets stuck at: > > Discarding device blocks: 4096/3778299 > > Have you ever seen something like this or any hint how to debug this mess? Looping in Adrian, as he maintains sdhci. I guess some more useful information would also be about what commit you are running and whether this is a regression or not. Kind regards Uffe -- To unsubscribe from this list: send the line "unsubscribe linux-mmc" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
