Re: [PATCH 00/12] mmc: core: Improve code for polling and HW busy detect

From: Ulf Hansson <ulf.hansson@linaro.org>
To: "linux-mmc@vger.kernel.org" <linux-mmc@vger.kernel.org>,
	Baolin Wang <baolin.wang7@gmail.com>,
	Ludovic Barre <ludovic.barre@st.com>
Cc: "Adrian Hunter" <adrian.hunter@intel.com>,
	"Wolfram Sang" <wsa+renesas@sang-engineering.com>,
	"Ulf Hansson" <ulf.hansson@linaro.org>,
	"Linus Walleij" <linus.walleij@linaro.org>,
	"Chaotian Jing" <chaotian.jing@mediatek.com>,
	"Shawn Lin" <shawn.lin@rock-chips.com>,
	"Michał Mirosław" <mirq-linux@rere.qmqm.pl>
Subject: Re: [PATCH 00/12] mmc: core: Improve code for polling and HW busy detect
Date: Wed, 19 Feb 2020 00:38:43 +0100	[thread overview]
Message-ID: <CAPDyKFoVKiXk0Cbj2fqbYB9ZtAJL2cO3k-YCJH0Eiyjb0zfUUg@mail.gmail.com> (raw)
In-Reply-To: <20200204085449.32585-1-ulf.hansson@linaro.org>

On Tue, 4 Feb 2020 at 09:55, Ulf Hansson <ulf.hansson@linaro.org> wrote:
>
> There exists several separate variants of polling loops, used to detect when
> the card stop signals busy for various operations, in the mmc core. All of them
> have different issues that needs to be fixed.
>
> The intent with this series, is to address some of these problems, via first
> improving the mmc_poll_for_busy() function, then consolidate code by moving
> more users to it.
>
> While I was working on this, I stumbled over some code here and there, that
> deserved some cleanup, hence I also folded in a couple of patches for this.
>
> So far, I have only managed to extensively test the updated mmc_poll_for_busy()
> function for CMD6 commands. Some tests for erase/trim/discard and for
> HPI+sanitize are needed.
>
> Note that, there are still separate polling loops in the mmc block layer, but
> moving that to mmc_poll_for_busy() involves some additional work. I am looking
> into that as a next step.
>
> Please help review and test!
>
> Kind regards
> Ulf Hansson
>
>
> Ulf Hansson (12):
>   mmc: core: Throttle polling rate for CMD6
>   mmc: core: Drop unused define
>   mmc: core: Extend mmc_switch_status() to rid of __mmc_switch_status()
>   mmc: core: Drop redundant in-parameter to __mmc_switch()
>   mmc: core: Split up mmc_poll_for_busy()
>   mmc: core: Enable re-use of mmc_blk_in_tran_state()
>   mmc: core: Update CMD13 busy check for CMD6 commands
>   mmc: core: Convert to mmc_poll_for_busy() for erase/trim/discard
>   mmc: core: Drop redundant out-parameter to mmc_send_hpi_cmd()
>   mmc: core: Convert to mmc_poll_for_busy() for HPI commands
>   mmc: core: Fixup support for HW busy detection for HPI commands
>   mmc: core: Re-work the error path for the eMMC sanitize command
>
>  drivers/mmc/core/block.c   |  55 +++++--------
>  drivers/mmc/core/core.c    |  53 +------------
>  drivers/mmc/core/mmc.c     |  38 ++++-----
>  drivers/mmc/core/mmc_ops.c | 159 ++++++++++++++++++++++---------------
>  drivers/mmc/core/mmc_ops.h |  13 ++-
>  include/linux/mmc/core.h   |   3 -
>  include/linux/mmc/mmc.h    |  10 +++
>  7 files changed, 157 insertions(+), 174 deletions(-)
>
> --
> 2.17.1
>

FYI, I have queued up this series for next (except patch12 that
deserves another re-spin). I also amended the changelog for patch1,
according the comment from Ludovic.

Feel free to provide additional feedback and test-reports, while we
monitor how this cook in linux-next.

Kind regards
Uffe