title: "SDHI; refactor reset into proper reset and hw_reset"
team: IO
key: b408bad6-edcd-11ea-942b-eb603d86bc58
status: Active
assignee: Wolfram

bsp39x:

upstream:
 - torvalds: 6e7d4de10890545b00f7d7c33310f1555963d592 # mmc: renesas_sdhi: move wrong 'hw_reset' to 'reset'
 - torvalds: 34e3211e5492756d4e59ef54f9ab1750ec7050ec # Revert "mmc: tmio: fix reset operation"
 - torvalds: a8c83064e73382d6bb069bd51c56074d1e4f4eec # mmc: tmio: remove indirection of 'hw_reset' callback
 - torvalds: 576146ea4ed986f44a4406ec32a5d3363216b3d2 # mmc: tmio: factor out common parts of the reset routine
 - torvalds: 5b0739d76227fd5a3f02f014385bfa9c86e0404b # mmc: tmio: don't reset whole IP core when tuning fails
 - torvalds: 510bfe58936f7d79d4c377bcc7425244d879b1ff # mmc: tmio: remove indirection of 'execute_tuning' callback

comments:
 - "Shimoda-san reported that after IP core reset, some card initialization needs to be done again"
 - "Wolfram then evaluated that the usage of reset and hw_reset in the SDHI driver is different from what the MMC core expects"
 - "a proper refactoring was needed"
 - "v1: https://patchwork.kernel.org/project/linux-renesas-soc/list/?series=335877&state=*"
 - "v1 merged"
 - "regression when re-inserting SD cards found, investigating"
 - "also, we should probably reset the device in probe()"
 - "BSP team suggests to use SRCR3 register SD-IFx bit"
 - "regressions fixed: https://patchwork.kernel.org/project/linux-renesas-soc/list/?series=378775&state=*"
 - "a better fix for the regression: https://patchwork.kernel.org/project/linux-renesas-soc/patch/20201125204953.3344-1-wsa+renesas@sang-engineering.com/"
 - "prototype for reset via SRCR3 sent internally; unusual delay needed, asking HW team for confirmation"
 - "no response so far, asked HW team again"
 - "response: we don't know why :)"
 - "patch series rfc v2: https://patchwork.kernel.org/project/linux-renesas-soc/list/?series=436517&state=*"