title: SDHI; implement support for HW busy timeout
team: IO
assignee: Wolfram
key: e2c455e0-1210-11ea-8389-b359bc6084c5
status: Done

upstream:
  - torvalds: 35cdcd1268f654f55e2308767349c44bd25d8009 # mmc: tmio: Add data timeout error detection
  - torvalds: ed2fab9a8229cc70fe03032e48d0ec375df6013e # mmc: renesas_sdhi: Add a condition of cmd/data timeout for retune
  - torvalds: 87e985aea2bef73242cf2a7d2ef81f950d677361 # mmc: renesas_sdhi: enable WAIT_WHILE_BUSY

comments:
  - "currently we use a tasklet to detect busy timeouts"
  - "MMC maintainer says this is OKish, but HW timeout would be much better"
  - "this is all for supporting MMC_CAP_ERASE properly"
  - "MMC_CAP_ERASE has been removed meanwhile, and we set max_busy_timeout, too"
  - "we still should care about handling when data timeouts happen"
  - "RFT v1: https://patchwork.kernel.org/project/linux-renesas-soc/list/?series=388391&state=*"
  - "RFT v2: https://patchwork.kernel.org/project/linux-renesas-soc/list/?series=388435&state=*"
  - "needs some refactoring"
  - "refactored v1: https://patchwork.kernel.org/project/linux-renesas-soc/list/?series=391145&state=*"
  - "refactored series with max_busy_timeout merged"
  - "prototype patches for handling data timeout developed by Shimoda-san and reviewed by Wolfram"
  - "WAIT_WHILE_BUSY needs to be enabled on top of 'data timeout' patches"
  - "data timeout series v1: https://patchwork.kernel.org/project/linux-renesas-soc/list/?series=405671&state=*"
  - "v1 merged"
  - "patch to enable WAIT_WHILE_BUSY sent internally for further testing"
  - "v1: https://patchwork.kernel.org/project/linux-renesas-soc/patch/20210408133420.2900-1-wsa+renesas@sang-engineering.com/"
  - "merged"