From mboxrd@z Thu Jan 1 00:00:00 1970 From: Jeremy McNicoll Subject: Re: [PATCH 5/5] sdhci: Add quirk for delayed IRQ ACK Date: Tue, 22 Nov 2016 21:23:46 -0800 Message-ID: References: <1479863388-23678-1-git-send-email-jeremymc@redhat.com> <1479863388-23678-6-git-send-email-jeremymc@redhat.com> <20161123113647.0b3deac0@xhacker> <20161123121209.3623e14b@xhacker> Mime-Version: 1.0 Content-Type: text/plain; charset=windows-1252; format=flowed Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <20161123121209.3623e14b@xhacker> Sender: devicetree-owner-u79uwXL29TY76Z2rM5mHXA@public.gmane.org To: Jisheng Zhang Cc: Jeremy McNicoll , linux-arm-msm-u79uwXL29TY76Z2rM5mHXA@public.gmane.org, linux-soc-u79uwXL29TY76Z2rM5mHXA@public.gmane.org, devicetree-u79uwXL29TY76Z2rM5mHXA@public.gmane.org, linux-mmc-u79uwXL29TY76Z2rM5mHXA@public.gmane.org, andy.gross-QSEj5FYQhm4dnm+yROfE0A@public.gmane.org, sboyd-sgV2jX0FEOL9JmXXK+q4OQ@public.gmane.org, robh-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org, arnd-r2nGTMty4D4@public.gmane.org, bjorn.andersson-QSEj5FYQhm4dnm+yROfE0A@public.gmane.org, riteshh-sgV2jX0FEOL9JmXXK+q4OQ@public.gmane.org List-Id: linux-mmc@vger.kernel.org On 2016-11-22 8:12 PM, Jisheng Zhang wrote: > On Tue, 22 Nov 2016 19:48:56 -0800 Jeremy McNicoll wrote: > >> On 2016-11-22 7:36 PM, Jisheng Zhang wrote: >>> On Tue, 22 Nov 2016 17:09:48 -0800 >>> Jeremy McNicoll wrote: >>> >>>> On msm8992 it has been observed that IRQs were not getting >>>> ACK'd correctly when clocked at speeds greater than 400KHz. >>>> >>>> Signed-off-by: Jeremy McNicoll >>>> --- >>>> drivers/mmc/host/sdhci-msm.c | 7 +++++++ >>>> drivers/mmc/host/sdhci.c | 12 ++++++++++-- >>>> drivers/mmc/host/sdhci.h | 2 ++ >>>> 3 files changed, 19 insertions(+), 2 deletions(-) >>>> >>>> diff --git a/drivers/mmc/host/sdhci-msm.c b/drivers/mmc/host/sdhci-msm.c >>>> index 1fcda96..459003c 100644 >>>> --- a/drivers/mmc/host/sdhci-msm.c >>>> +++ b/drivers/mmc/host/sdhci-msm.c >>>> @@ -1303,6 +1303,13 @@ static int sdhci_msm_probe(struct platform_device *pdev) >>>> CORE_VENDOR_SPEC_CAPABILITIES0); >>>> } >>>> >>>> + /* Enable delayed IRQ handling workaround on 8992 */ >>>> + if (core_major == 1 && core_minor == 0x3e) { >>>> + /* Add 40us delay in interrupt handler when operating >>>> + * at initialization frequency of 400KHz. */ >>>> + host->quirks2 |= SDHCI_QUIRK2_SLOW_INT_CLR; >>>> + } >>>> + >>>> /* Setup IRQ for handling power/voltage tasks with PMIC */ >>>> msm_host->pwr_irq = platform_get_irq_byname(pdev, "pwr_irq"); >>>> if (msm_host->pwr_irq < 0) { >>>> diff --git a/drivers/mmc/host/sdhci.c b/drivers/mmc/host/sdhci.c >>>> index 5911f98..c1aae22 100644 >>>> --- a/drivers/mmc/host/sdhci.c >>>> +++ b/drivers/mmc/host/sdhci.c >>>> @@ -2703,11 +2703,19 @@ static irqreturn_t sdhci_irq(int irq, void *dev_id) >>>> result = IRQ_WAKE_THREAD; >>>> } >>>> >>>> - if (intmask & SDHCI_INT_CMD_MASK) >>>> + if (intmask & SDHCI_INT_CMD_MASK) { >>>> + if ((host->quirks2 & SDHCI_QUIRK2_SLOW_INT_CLR) && (host->clock <= 400000)) { >>>> + udelay(40); >>>> + } >>>> sdhci_cmd_irq(host, intmask & SDHCI_INT_CMD_MASK); >>>> + } >>>> >>>> - if (intmask & SDHCI_INT_DATA_MASK) >>>> + if (intmask & SDHCI_INT_DATA_MASK) { >>>> + if ((host->quirks2 & SDHCI_QUIRK2_SLOW_INT_CLR) && (host->clock <= 400000)) { >>>> + udelay(40); >>>> + } >>>> sdhci_data_irq(host, intmask & SDHCI_INT_DATA_MASK); >>>> + } >>>> >>>> if (intmask & SDHCI_INT_BUS_POWER) >>>> pr_err("%s: Card is consuming too much power!\n", >>>> diff --git a/drivers/mmc/host/sdhci.h b/drivers/mmc/host/sdhci.h >>>> index c055e24..5f8301e 100644 >>>> --- a/drivers/mmc/host/sdhci.h >>>> +++ b/drivers/mmc/host/sdhci.h >>>> @@ -24,6 +24,8 @@ >>>> * Controller registers >>>> */ >>>> >>>> +#define SDHCI_QUIRK2_SLOW_INT_CLR (1<<5) >>> >>> IIRC, new quirk isn't allowed now. >>> >> >> Why not? > > IIRC, mmc subsystem will behave as a lib in the long run, so the community > and developers call for no new quirk. For your case, we may need to handle > the udelay in sdhci-msm.c, export sdhci_irq as a helper function? > Thanks for the details and suggestions. I am going to wait until the US Thanksgiving (and cyber Monday) is over before I do any more. -jeremy -- To unsubscribe from this list: send the line "unsubscribe devicetree" in the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org More majordomo info at http://vger.kernel.org/majordomo-info.html