linux-kernel - Re: [PATCH 3/3] mmc: core: fix mmc_sdio_reinit

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Wed, 22 Apr 2020 15:46:43 -0700
From:   Matthias Kaehlcke <mka@...omium.org>
To:     Yong Mao <yong.mao@...iatek.com>
Cc:     Ulf Hansson <ulf.hansson@...aro.org>,
        Chaotian Jing <chaotian.jing@...iatek.com>,
        Matthias Brugger <matthias.bgg@...il.com>,
        linux-mmc@...r.kernel.org, linux-arm-kernel@...ts.infradead.org,
        linux-mediatek@...ts.infradead.org, linux-kernel@...r.kernel.org,
        srv_heupstream@...iatek.com
Subject: Re: [PATCH 3/3] mmc: core: fix mmc_sdio_reinit_card fail issue

Hi Yong,

On Tue, Apr 14, 2020 at 11:40:11AM +0800, Yong Mao wrote:
> From: yong mao <yong.mao@...iatek.com>
> 
> If SDIO device is initialized by UHS mode, it will run with 1.8v power.
> In this mode, mmc_go_idle may not make SDIO device go idle successfully
> in some special SDIO device. And then it can't be re-initialized
> successfully.
> According to the logic in sdio_reset_comm and mmc_sdio_sw_reset,
> invoking mmc_set_clock(host, host->f_min) before mmc_send_io_op_cond
> can make this SDIO device back to right state.
>

The commit message isn't very concise. Suggestion for a better
structure:

mmc: core: reset clock to minimum speed during card reinit

Some buggy (?) SDIO devices don't (consistently?) enter idle mode
through mmc_go_idle() when running in UHS mode. [add rationale why
setting the clock to minimum speed fixes this]


Also the function sdio_reset_comm() mentioned in the commit message
doesn't exist in recent kernels. And mmc_sdio_sw_reset() does not invoke
mmc_send_io_op_cond(), as the commit message appears to claim.

> Signed-off-by: Yong Mao <yong.mao@...iatek.com>
> ---
>  drivers/mmc/core/sdio.c | 1 +
>  1 file changed, 1 insertion(+)
> 
> diff --git a/drivers/mmc/core/sdio.c b/drivers/mmc/core/sdio.c
> index f173cad..dc4dc63 100644
> --- a/drivers/mmc/core/sdio.c
> +++ b/drivers/mmc/core/sdio.c
> @@ -850,6 +850,7 @@ static int mmc_sdio_reinit_card(struct mmc_host *host)
>  
>  	sdio_reset(host);
>  	mmc_go_idle(host);
> +	mmc_set_clock(host, host->f_min);

mmc_sdio_sw_reset() - which is mentioned as reference in the commit
message - sets the clock speed before sdio_reset(). Should this order
be followed here too?