linux-kernel - Re: [PATCH] mmc: mmc: do not use CMD13 to get status after speed mode switch

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] [day] [month] [year] [list]

Date:	Fri, 13 May 2016 10:42:23 +0800
From:	Chaotian Jing <chaotian.jing@...iatek.com>
To:	Adrian Hunter <adrian.hunter@...el.com>
CC:	Ulf Hansson <ulf.hansson@...aro.org>,
	Matthias Brugger <matthias.bgg@...il.com>,
	Wolfram Sang <wsa+renesas@...g-engineering.com>,
	Kuninori Morimoto <kuninori.morimoto.gx@...esas.com>,
	Masahiro Yamada <yamada.masahiro@...ionext.com>,
	<linux-mmc@...r.kernel.org>, <linux-kernel@...r.kernel.org>,
	<linux-arm-kernel@...ts.infradead.org>,
	<linux-mediatek@...ts.infradead.org>,
	<srv_heupstream@...iatek.com>,
	"Sascha Hauer" <kernel@...gutronix.de>
Subject: Re: [PATCH] mmc: mmc: do not use CMD13 to get status after speed
 mode switch

On Thu, 2016-05-12 at 13:29 +0300, Adrian Hunter wrote:
> On 12/05/16 10:00, Chaotian Jing wrote:
> > On Wed, 2016-05-11 at 10:50 +0300, Adrian Hunter wrote:
> >> On 04/05/16 09:54, Chaotian Jing wrote:
> >>> Per JEDEC spec, it is not recommended to use CMD13 to get card status
> >>> after speed mode switch. below are two reason about this:
> >>> 1. CMD13 cannot be guaranteed due to the asynchronous operation.
> >>> Therefore it is not recommended to use CMD13 to check busy completion
> >>> of the timing change indication.
> >>> 2. After switch to HS200, CMD13 will get response of 0x800, and even the
> >>> busy signal gets de-asserted, the response of CMD13 is aslo 0x800.
> >>>
> >>> this patch drops CMD13 when doing speed mode switch, if host do not
> >>> support MMC_CAP_WAIT_WHILE_BUSY and there is no ops->card_busy(),
> >>> then the only way is to wait a fixed timeout.
> >>
> >> This looks like it should be 3 patches:
> >> 	1. Change __mmc_switch
> >> 	2. Change HS200 and HS400 selection
> >> 	3. Change HS selection
> >>
> >> However there is another problem: card_busy is not the same as busy signal -
> >> see below.  So that needs to be sorted out first.
> >>
> > We should make that card_busy() is the same with busy signal asserted.
> > as you know, if card was not in busy state, all data pins should be high
> > level as it is pull-up by default. so that's no conflict to check card
> > busy signal by DAT0 or DAT0 ~ DAT3.
> 
> Potentially SDIO uses DAT1 for card interrupt, so that is a conflict right
> there.
> 
> Also SDHCI does it backwards (don't ask me why) and considers 0000 to be
> busy, so there's another conflict.
> 
> Some of the language in the SD and SDHCI specifications seems to indicate
> that checking all 4 DAT lines during voltage switch is optional i.e. only 1
> of the lines must be checked.  If that is true then we could change all the
> drivers over to check just DAT0, and expect that to work for both busy
> signalling and SD voltage switch checks.
> 
> So it seems to me card_busy still needs to be sorted out first.

One thing must point out is that the __mmc_switch() is only for MMC
card. SD/SDIO will never use this interface.
by the way, Per JEDEC SD3.0 spec, below is the quote from spec:
"Completion of voltage switch sequence is checked by high level of
DAT[3:0]. Any bit of DAT[3:0] can be checked depends on ability of the
host."
So that the implement of ops->card_busy() can be changed from check
DAT[3:0] to only check DAT[0].

In fact, for SD/SDIO voltage switch, if switch success, all DAT pins are
high level and if switch failed, all data pins are low level.
>