linux-arm-msm.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Bjorn Andersson <bjorn.andersson@linaro.org>
To: Elliot Berman <eberman@codeaurora.org>
Cc: saiprakash.ranjan@codeaurora.org, agross@kernel.org,
	tsoni@codeaurora.org, sidgup@codeaurora.org,
	psodagud@codeaurora.org, linux-arm-msm@vger.kernel.org,
	linux-kernel@vger.kernel.org
Subject: Re: [PATCH 09/17] firmware: qcom_scm-64: Improve SMC convention detection
Date: Thu, 7 Nov 2019 11:18:46 -0800	[thread overview]
Message-ID: <20191107191846.GA3907604@builder> (raw)
In-Reply-To: <1572917256-24205-10-git-send-email-eberman@codeaurora.org>

On Mon 04 Nov 17:27 PST 2019, Elliot Berman wrote:

> - Use enum to describe SMC convention.
> - Improve SMC convention detection to use __qcom_scm_is_call_available
>   instead of circumventing qcom_scm_call_smccc.
> - Improve SMC convention detection to check that SMCCC-32 works, instead
>   of just assuming it does of SMCCC-64 does not.

I was about to tell you that your list represent individual changes, but
I think you should rewrite the commit message instead. Something like:

"""
Improve the calling convention detection to use
__qcom_scm_is_call_available() and not blindly assume 32-bit mode if
the checks fails.
"""

> 
> Signed-off-by: Elliot Berman <eberman@codeaurora.org>
> ---
>  drivers/firmware/qcom_scm-64.c | 42 ++++++++++++++++++++++++------------------
>  1 file changed, 24 insertions(+), 18 deletions(-)
> 
> diff --git a/drivers/firmware/qcom_scm-64.c b/drivers/firmware/qcom_scm-64.c
> index f79b0dc..2579246 100644
> --- a/drivers/firmware/qcom_scm-64.c
> +++ b/drivers/firmware/qcom_scm-64.c
> @@ -58,7 +58,13 @@ struct arm_smccc_args {
>  	unsigned long a[8];
>  };
>  
> -static u64 qcom_smccc_convention = -1;
> +enum qcom_smc_convention {
> +	SMC_CONVENTION_UNKNOWN,
> +	SMC_CONVENTION_ARM_32,
> +	SMC_CONVENTION_ARM_64,
> +};
> +
> +static enum qcom_smc_convention qcom_smc_convention = SMC_CONVENTION_UNKNOWN;
>  static DEFINE_MUTEX(qcom_scm_lock);
>  
>  #define QCOM_SCM_EBUSY_WAIT_MS 30
> @@ -103,7 +109,9 @@ static int ___qcom_scm_call_smccc(struct device *dev,
>  
>  	smc.a[0] = ARM_SMCCC_CALL_VAL(
>  		atomic ? ARM_SMCCC_FAST_CALL : ARM_SMCCC_STD_CALL,
> -		qcom_smccc_convention,

Use a local variable instead of using a ternary operator in the middle
of the arguments.

> +		(qcom_smc_convention == SMC_CONVENTION_ARM_64) ?
> +			ARM_SMCCC_SMC_64 :
> +			ARM_SMCCC_SMC_32,

Here SMC_CONVENTION_UNKNOWN would mean ARM_SMCCC_SMC_32...

>  		desc->owner,
>  		SMCCC_FUNCNUM(desc->svc, desc->cmd));
>  	smc.a[1] = desc->arginfo;
> @@ -117,7 +125,7 @@ static int ___qcom_scm_call_smccc(struct device *dev,
>  		if (!args_virt)
>  			return -ENOMEM;
>  
> -		if (qcom_smccc_convention == ARM_SMCCC_SMC_32) {
> +		if (qcom_smc_convention == SMC_CONVENTION_ARM_32) {

...but here it would mean ARM_SMCCC_SMC_64.

>  			__le32 *args = args_virt;
>  
>  			for (i = 0; i < SMCCC_N_EXT_ARGS; i++)
> @@ -583,19 +591,17 @@ int __qcom_scm_qsmmu500_wait_safe_toggle(struct device *dev, bool en)
>  
>  void __qcom_scm_init(void)
>  {
> -	u64 cmd;
> -	struct arm_smccc_res res;
> -	u32 function = SMCCC_FUNCNUM(QCOM_SCM_SVC_INFO, QCOM_SCM_INFO_IS_CALL_AVAIL);
> -
> -	/* First try a SMC64 call */
> -	cmd = ARM_SMCCC_CALL_VAL(ARM_SMCCC_FAST_CALL, ARM_SMCCC_SMC_64,
> -				 ARM_SMCCC_OWNER_SIP, function);
> -
> -	arm_smccc_smc(cmd, QCOM_SCM_ARGS(1), cmd & (~BIT(ARM_SMCCC_TYPE_SHIFT)),
> -		      0, 0, 0, 0, 0, &res);
> -
> -	if (!res.a0 && res.a1)
> -		qcom_smccc_convention = ARM_SMCCC_SMC_64;
> -	else
> -		qcom_smccc_convention = ARM_SMCCC_SMC_32;
> +	qcom_smc_convention = SMC_CONVENTION_ARM_64;
> +	if (__qcom_scm_is_call_available(NULL, QCOM_SCM_SVC_INFO,
> +			QCOM_SCM_INFO_IS_CALL_AVAIL) == 1)
> +		goto out;
> +
> +	qcom_smc_convention = SMC_CONVENTION_ARM_32;
> +	if (__qcom_scm_is_call_available(NULL, QCOM_SCM_SVC_INFO,
> +			QCOM_SCM_INFO_IS_CALL_AVAIL) == 1)
> +		goto out;
> +
> +	qcom_smc_convention = SMC_CONVENTION_UNKNOWN;

If above two tests can be considered reliable I would suggest that you
fail hard here instead.

And if so I think you should postpone the introduction of the enum until
you actually need it to represent the legacy mode.

Regards,
Bjorn

> +out:
> +	pr_debug("QCOM SCM SMC Convention: %d\n", qcom_smc_convention);
>  }
> -- 
> The Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum,
> a Linux Foundation Collaborative Project
> 

  reply	other threads:[~2019-11-07 19:18 UTC|newest]

Thread overview: 34+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-11-05  1:27 [PATCH 00/17] Restructure, improve target support for qcom_scm driver Elliot Berman
2019-11-05  1:27 ` [PATCH 01/17] firmware: qcom_scm: Rename macros and structures Elliot Berman
2019-11-07 16:22   ` Vinod Koul
2019-11-05  1:27 ` [PATCH 02/17] firmware: qcom_scm: Apply consistent naming scheme to command IDs Elliot Berman
2019-11-06  6:02   ` Bjorn Andersson
2019-11-07 16:42   ` Vinod Koul
2019-11-05  1:27 ` [PATCH 03/17] firmware: qcom_scm: Order functions, definitions by service/command Elliot Berman
2019-11-06  6:05   ` Bjorn Andersson
2019-11-07 16:50   ` Vinod Koul
2019-11-08 22:03   ` Bjorn Andersson
2019-11-05  1:27 ` [PATCH 04/17] firmware: qcom_scm: Remove unused qcom_scm_get_version Elliot Berman
2019-11-06  6:06   ` Bjorn Andersson
2019-11-08  4:36   ` Vinod Koul
2019-11-05  1:27 ` [PATCH 05/17] firmware: qcom_scm-64: Move svc/cmd/owner into qcom_scm_desc Elliot Berman
2019-11-08  6:04   ` Vinod Koul
2019-11-05  1:27 ` [PATCH 06/17] firmware: qcom_scm-64: Add SCM results to descriptor Elliot Berman
2019-11-08  6:09   ` Vinod Koul
2019-11-05  1:27 ` [PATCH 07/17] firmware: qcom_scm-64: Remove qcom_scm_call_do_smccc Elliot Berman
2019-11-05  1:27 ` [PATCH 08/17] firmware: qcom_scm-64: Move SMC register filling to qcom_scm_call_smccc Elliot Berman
2019-11-05  1:27 ` [PATCH 09/17] firmware: qcom_scm-64: Improve SMC convention detection Elliot Berman
2019-11-07 19:18   ` Bjorn Andersson [this message]
2019-11-07 20:20     ` eberman
2019-11-07 23:44       ` Bjorn Andersson
2019-11-05  1:27 ` [PATCH 10/17] firmware: qcom_scm-32: Use SMC arch wrappers Elliot Berman
2019-11-08  0:05   ` Bjorn Andersson
2019-11-05  1:27 ` [PATCH 11/17] firmware: qcom_scm-32: Use qcom_scm_desc in non-atomic calls Elliot Berman
2019-11-05  1:27 ` [PATCH 12/17] firmware: qcom_scm-32: Move SMCCC register filling to qcom_scm_call Elliot Berman
2019-11-05  1:27 ` [PATCH 13/17] firmware: qcom_scm-32: Create common legacy atomic call Elliot Berman
2019-11-05  1:27 ` [PATCH 14/17] firmware: qcom_scm-32: Add device argument to atomic calls Elliot Berman
2019-11-05  1:27 ` [PATCH 15/17] firmware: qcom_scm: Merge legacy and SMCCC conventions Elliot Berman
2019-11-05  1:27 ` [PATCH 16/17] firmware: qcom_scm: Enable legacy calling convention in qcom_scm-64.c Elliot Berman
2019-11-05  1:27 ` [PATCH 17/17] firmware: qcom_scm: Rename -64 -> -smc, remove -32 Elliot Berman
2019-11-08  0:20   ` Bjorn Andersson
2019-11-05  2:14 ` [PATCH 00/17] Restructure, improve target support for qcom_scm driver Sai Prakash Ranjan

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=20191107191846.GA3907604@builder \
    --to=bjorn.andersson@linaro.org \
    --cc=agross@kernel.org \
    --cc=eberman@codeaurora.org \
    --cc=linux-arm-msm@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=psodagud@codeaurora.org \
    --cc=saiprakash.ranjan@codeaurora.org \
    --cc=sidgup@codeaurora.org \
    --cc=tsoni@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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).