All of lore.kernel.org
 help / color / mirror / Atom feed
From: Faiz Abbas <faiz_abbas@ti.com>
To: <linux-kernel@vger.kernel.org>, <linux-mmc@vger.kernel.org>
Cc: <ulf.hansson@linaro.org>, <asutoshd@codeaurora.org>,
	<riteshh@codeaurora.org>, <adrian.hunter@intel.com>,
	<venkatg@codeaurora.org>
Subject: Re: [RFC] mmc: cqhci: commit descriptors before setting the doorbell
Date: Tue, 15 Oct 2019 13:25:15 +0530	[thread overview]
Message-ID: <fac9ad28-dbc3-3948-d99c-742420f3e651@ti.com> (raw)
In-Reply-To: <20191014183849.14864-1-faiz_abbas@ti.com>

Hi,

On 15/10/19 12:08 AM, Faiz Abbas wrote:
> Add a write memory barrier to make sure that descriptors are actually
> written to memory before ringing the doorbell.
> 
> Signed-off-by: Faiz Abbas <faiz_abbas@ti.com>
> ---
> 
> This patch fixes a very infrequent ADMA error (1 out of 100 times) that
> I have been seeing after enabling command queuing for J721e.
> Also looking at memory-barriers.txt and this commit[1],
> it looks like we should be doing this before any descriptor write
> followed by a doorbell ring operation. It'll be nice if someone with more
> expertise in memory barriers can comment.
> 
> [1] ad1a1b9cd67a ("scsi: ufs: commit descriptors before setting the
>     doorbell")

So I see that cqhci_readl/writel() use readl/writel_relaxed() which
seems to be causing this issue. Should I just fix this by converting
those to readl/writel with memory barriers instead?

Thanks,
Faiz

WARNING: multiple messages have this Message-ID (diff)
From: Faiz Abbas <faiz_abbas@ti.com>
To: linux-kernel@vger.kernel.org, linux-mmc@vger.kernel.org
Cc: ulf.hansson@linaro.org, asutoshd@codeaurora.org,
	riteshh@codeaurora.org, adrian.hunter@intel.com,
	venkatg@codeaurora.org
Subject: Re: [RFC] mmc: cqhci: commit descriptors before setting the doorbell
Date: Tue, 15 Oct 2019 13:25:15 +0530	[thread overview]
Message-ID: <fac9ad28-dbc3-3948-d99c-742420f3e651@ti.com> (raw)
In-Reply-To: <20191014183849.14864-1-faiz_abbas@ti.com>

Hi,

On 15/10/19 12:08 AM, Faiz Abbas wrote:
> Add a write memory barrier to make sure that descriptors are actually
> written to memory before ringing the doorbell.
> 
> Signed-off-by: Faiz Abbas <faiz_abbas@ti.com>
> ---
> 
> This patch fixes a very infrequent ADMA error (1 out of 100 times) that
> I have been seeing after enabling command queuing for J721e.
> Also looking at memory-barriers.txt and this commit[1],
> it looks like we should be doing this before any descriptor write
> followed by a doorbell ring operation. It'll be nice if someone with more
> expertise in memory barriers can comment.
> 
> [1] ad1a1b9cd67a ("scsi: ufs: commit descriptors before setting the
>     doorbell")

So I see that cqhci_readl/writel() use readl/writel_relaxed() which
seems to be causing this issue. Should I just fix this by converting
those to readl/writel with memory barriers instead?

Thanks,
Faiz

  reply	other threads:[~2019-10-15  7:54 UTC|newest]

Thread overview: 12+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-10-14 18:38 [RFC] mmc: cqhci: commit descriptors before setting the doorbell Faiz Abbas
2019-10-14 18:38 ` Faiz Abbas
2019-10-15  7:55 ` Faiz Abbas [this message]
2019-10-15  7:55   ` Faiz Abbas
2019-10-15 13:45   ` Adrian Hunter
2019-10-16 12:16     ` Faiz Abbas
2019-10-16 12:16       ` Faiz Abbas
2019-10-18  7:22       ` Faiz Abbas
2019-10-18  7:22         ` Faiz Abbas
2019-10-18  7:30         ` Adrian Hunter
2019-10-18 10:58 ` Ulf Hansson
2019-10-18 11:27   ` Faiz Abbas

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=fac9ad28-dbc3-3948-d99c-742420f3e651@ti.com \
    --to=faiz_abbas@ti.com \
    --cc=adrian.hunter@intel.com \
    --cc=asutoshd@codeaurora.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-mmc@vger.kernel.org \
    --cc=riteshh@codeaurora.org \
    --cc=ulf.hansson@linaro.org \
    --cc=venkatg@codeaurora.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body. 
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.