On Tue, Apr 23, 2019 at 04:50:58PM +0800, Kaiyen Chang wrote:
Existing data command CRC error handling on kernel 4.20 stable branch is non-standard and does not work with some Intel host controllers. Specifically, the assumption that the host controller will continue operating normally after the error interrupt, is not valid. So we suggest cherry-pick the 3 patches listed below to kernel 4.20 stable branch, which can change the driver to handle the error in the same manner as a data CRC error.
4bf7809: mmc: sdhci: Fix data command CRC error handling 869f8a6: mmc: sdhci: Rename SDHCI_ACMD12_ERR and SDHCI_INT_ACMD12ERR af849c8: mmc: sdhci: Handle auto-command errors
All of the patches above have landed on kernel 5.0 stable branch.
4.20 went EoL quite a while ago, you should be using 5.0 at this point.
I've queued these to older stable kernels, thanks.
-- Thanks, Sasha