linux-kernel - Re: [PATCH V3] mmc: sdhci-msm: Update the software timeout value for sdhc

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]
Message-ID: <7bd56312-96ab-debe-0f80-112af994b233@intel.com>
Date:   Thu, 15 Jul 2021 08:29:37 +0300
From:   Adrian Hunter <adrian.hunter@...el.com>
To:     Shaik Sajida Bhanu <sbhanu@...eaurora.org>, ulf.hansson@...aro.org
Cc:     asutoshd@...eaurora.org, stummala@...eaurora.org,
        vbadigan@...eaurora.org, rampraka@...eaurora.org,
        sayalil@...eaurora.org, sartgarg@...eaurora.org,
        rnayak@...eaurora.org, cang@...eaurora.org,
        pragalla@...eaurora.org, nitirawa@...eaurora.org,
        linux-mmc@...r.kernel.org, linux-kernel@...r.kernel.org,
        linux-arm-msm@...r.kernel.org, agross@...nel.org,
        bjorn.andersson@...aro.org, robh+dt@...nel.org
Subject: Re: [PATCH V3] mmc: sdhci-msm: Update the software timeout value for
 sdhc

On 5/07/21 6:50 pm, Shaik Sajida Bhanu wrote:
> Whenever SDHC run at clock rate 50MHZ or below, the hardware data
> timeout value will be 21.47secs, which is approx. 22secs and we have
> a current software timeout value as 10secs. We have to set software
> timeout value more than the hardware data timeout value to avioid seeing
> the below register dumps.
> 
> [  332.953670] mmc2: Timeout waiting for hardware interrupt.
> [  332.959608] mmc2: sdhci: ============ SDHCI REGISTER DUMP ===========
> [  332.966450] mmc2: sdhci: Sys addr:  0x00000000 | Version:  0x00007202
> [  332.973256] mmc2: sdhci: Blk size:  0x00000200 | Blk cnt:  0x00000001
> [  332.980054] mmc2: sdhci: Argument:  0x00000000 | Trn mode: 0x00000027
> [  332.986864] mmc2: sdhci: Present:   0x01f801f6 | Host ctl: 0x0000001f
> [  332.993671] mmc2: sdhci: Power:     0x00000001 | Blk gap:  0x00000000
> [  333.000583] mmc2: sdhci: Wake-up:   0x00000000 | Clock:    0x00000007
> [  333.007386] mmc2: sdhci: Timeout:   0x0000000e | Int stat: 0x00000000
> [  333.014182] mmc2: sdhci: Int enab:  0x03ff100b | Sig enab: 0x03ff100b
> [  333.020976] mmc2: sdhci: ACmd stat: 0x00000000 | Slot int: 0x00000000
> [  333.027771] mmc2: sdhci: Caps:      0x322dc8b2 | Caps_1:   0x0000808f
> [  333.034561] mmc2: sdhci: Cmd:       0x0000183a | Max curr: 0x00000000
> [  333.041359] mmc2: sdhci: Resp[0]:   0x00000900 | Resp[1]:  0x00000000
> [  333.048157] mmc2: sdhci: Resp[2]:   0x00000000 | Resp[3]:  0x00000000
> [  333.054945] mmc2: sdhci: Host ctl2: 0x00000000
> [  333.059657] mmc2: sdhci: ADMA Err:  0x00000000 | ADMA Ptr:
> 0x0000000ffffff218
> [  333.067178] mmc2: sdhci_msm: ----------- VENDOR REGISTER DUMP
> -----------
> [  333.074343] mmc2: sdhci_msm: DLL sts: 0x00000000 | DLL cfg:
> 0x6000642c | DLL cfg2: 0x0020a000
> [  333.083417] mmc2: sdhci_msm: DLL cfg3: 0x00000000 | DLL usr ctl:
> 0x00000000 | DDR cfg: 0x80040873
> [  333.092850] mmc2: sdhci_msm: Vndr func: 0x00008a9c | Vndr func2 :
> 0xf88218a8 Vndr func3: 0x02626040
> [  333.102371] mmc2: sdhci: ============================================
> 
> So, set software timeout value more than hardware timeout value.
> 
> Signed-off-by: Shaik Sajida Bhanu <sbhanu@...eaurora.org>

A couple of minor nitpicks below.  Fix those and you can add:

Acked-by: Adrian Hunter <adrian.hunter@...el.com>

> ---
> 
> Changes since V2:
> 	- Updated 22 with 22LL to avoid compiler warning as suggested by
> 	  Adrian Hunter.
> 	- Added a check to update software data timeout value if its value is
> 	  less than the calculated hardware data timeout value as suggested
> 	  by Veerabhadrarao Badiganti.
> 
> Changes since V1:
> 	- Moved software data timeout update part to qcom specific file
> 	  as suggested by Veerabhadrarao Badiganti.
> ---
>  drivers/mmc/host/sdhci-msm.c | 18 ++++++++++++++++++
>  1 file changed, 18 insertions(+)
> 
> diff --git a/drivers/mmc/host/sdhci-msm.c b/drivers/mmc/host/sdhci-msm.c
> index e44b7a6..64fb85e 100644
> --- a/drivers/mmc/host/sdhci-msm.c
> +++ b/drivers/mmc/host/sdhci-msm.c
> @@ -2089,6 +2089,23 @@ static void sdhci_msm_cqe_disable(struct mmc_host *mmc, bool recovery)
>  	sdhci_cqe_disable(mmc, recovery);
>  }
>  
> +static void sdhci_msm_set_timeout(struct sdhci_host *host, struct mmc_command *cmd)
> +{
> +

Unnecessary blank line.

> +	u32 count, start = 15;
> +
> +	__sdhci_set_timeout(host, cmd);
> +	count = sdhci_readb(host, SDHCI_TIMEOUT_CONTROL);
> +	/*
> +	 * Update software timeout value if its value is less than hardware data
> +	 * timeout value. Qcom SoC hardware data timeout value was calculated
> +	 * using 4 * MCLK * 2^(count + 13). where MCLK = 1 / host->clock.
> +	 */
> +	if (cmd && cmd->data && host->clock > 400000 && host->clock <= 50000000
> +			&& ((1 << (count + start)) > (10 * host->clock)))

'&&' is better at the end of the previous line

Also fewer parenthesis is more readable e.g.

	if (cmd && cmd->data && host->clock > 400000 &&
	    host->clock <= 50000000 &&
	    (1 << (count + start)) > (10 * host->clock))

> +		host->data_timeout = 22LL * NSEC_PER_SEC;
> +}
> +
>  static const struct cqhci_host_ops sdhci_msm_cqhci_ops = {
>  	.enable		= sdhci_msm_cqe_enable,
>  	.disable	= sdhci_msm_cqe_disable,
> @@ -2438,6 +2455,7 @@ static const struct sdhci_ops sdhci_msm_ops = {
>  	.irq	= sdhci_msm_cqe_irq,
>  	.dump_vendor_regs = sdhci_msm_dump_vendor_regs,
>  	.set_power = sdhci_set_power_noreg,
> +	.set_timeout = sdhci_msm_set_timeout,
>  };
>  
>  static const struct sdhci_pltfm_data sdhci_msm_pdata = {
>