linux-kernel - Re: [PATCH V3 1/8] firmware: qcom_scm: ipq5332: add support to pass metadata size

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: <c2dnzfeeg6ohlfum3t473s52udjwtoypnnp5kqx7avpe7jd3ys@lhx4zsal2lms>
Date: Tue, 7 Jan 2025 22:15:50 -0600
From: Bjorn Andersson <andersson@...nel.org>
To: Gokul Sriram Palanisamy <quic_gokulsri@...cinc.com>
Cc: jassisinghbrar@...il.com, robh@...nel.org, krzk+dt@...nel.org, 
	conor+dt@...nel.org, mathieu.poirier@...aro.org, konradybcio@...nel.org, 
	quic_mmanikan@...cinc.com, linux-arm-msm@...r.kernel.org, linux-kernel@...r.kernel.org, 
	devicetree@...r.kernel.org, linux-remoteproc@...r.kernel.org, dmitry.baryshkov@...aro.org, 
	quic_viswanat@...cinc.com, quic_srichara@...cinc.com
Subject: Re: [PATCH V3 1/8] firmware: qcom_scm: ipq5332: add support to pass
 metadata size

On Tue, Jan 07, 2025 at 03:46:40PM +0530, Gokul Sriram Palanisamy wrote:
> From: Manikanta Mylavarapu <quic_mmanikan@...cinc.com>
> 
> IPQ5332 security software running under trustzone requires metadata size.
> With new command support added in TrustZone that includes a size parameter,
> pass metadata size as well. This new command is specific to IPQ5332 SoC.

But function 0x1a is reserved by TZ owners across the company, right?

> 
> Signed-off-by: Manikanta Mylavarapu <quic_mmanikan@...cinc.com>
> Signed-off-by: Gokul Sriram Palanisamy <quic_gokulsri@...cinc.com>
> ---
>  drivers/firmware/qcom/qcom_scm.c | 13 +++++++++++--
>  drivers/firmware/qcom/qcom_scm.h |  1 +
>  2 files changed, 12 insertions(+), 2 deletions(-)
> 
> diff --git a/drivers/firmware/qcom/qcom_scm.c b/drivers/firmware/qcom/qcom_scm.c
> index 72bf87ddcd96..a713788926b0 100644
> --- a/drivers/firmware/qcom/qcom_scm.c
> +++ b/drivers/firmware/qcom/qcom_scm.c
> @@ -580,8 +580,6 @@ int qcom_scm_pas_init_image(u32 peripheral, const void *metadata, size_t size,
>  	int ret;
>  	struct qcom_scm_desc desc = {
>  		.svc = QCOM_SCM_SVC_PIL,
> -		.cmd = QCOM_SCM_PIL_PAS_INIT_IMAGE,
> -		.arginfo = QCOM_SCM_ARGS(2, QCOM_SCM_VAL, QCOM_SCM_RW),
>  		.args[0] = peripheral,

The arginfo and args are tightly coupled, so I'd prefer if you move
assignment of args[0] with the assignment below (and yes duplicate it).

>  		.owner = ARM_SMCCC_OWNER_SIP,
>  	};
> @@ -616,6 +614,17 @@ int qcom_scm_pas_init_image(u32 peripheral, const void *metadata, size_t size,
>  
>  	desc.args[1] = mdata_phys;

I find this rather ugly as well, so please move (and duplicate this) as
well - so we get it all gathered in one (two) place(s).

>  
> +	if (__qcom_scm_is_call_available(__scm->dev, QCOM_SCM_SVC_PIL,
> +					 QCOM_SCM_PAS_INIT_IMAGE_V2)) {
> +		desc.cmd = QCOM_SCM_PAS_INIT_IMAGE_V2;
> +		desc.arginfo =
> +			QCOM_SCM_ARGS(3, QCOM_SCM_VAL, QCOM_SCM_RW, QCOM_SCM_VAL);

Better leave this line unbroken.

Thanks,
Bjorn

> +		desc.args[2] = size;
> +	} else {
> +		desc.cmd = QCOM_SCM_PIL_PAS_INIT_IMAGE;
> +		desc.arginfo = QCOM_SCM_ARGS(2, QCOM_SCM_VAL, QCOM_SCM_RW);
> +	}
> +
>  	ret = qcom_scm_call(__scm->dev, &desc, &res);
>  	qcom_scm_bw_disable();
>  
> diff --git a/drivers/firmware/qcom/qcom_scm.h b/drivers/firmware/qcom/qcom_scm.h
> index e36b2f67607f..9aad9f92517f 100644
> --- a/drivers/firmware/qcom/qcom_scm.h
> +++ b/drivers/firmware/qcom/qcom_scm.h
> @@ -96,6 +96,7 @@ struct qcom_tzmem_pool *qcom_scm_get_tzmem_pool(void);
>  
>  #define QCOM_SCM_SVC_PIL		0x02
>  #define QCOM_SCM_PIL_PAS_INIT_IMAGE	0x01
> +#define QCOM_SCM_PAS_INIT_IMAGE_V2	0x1a
>  #define QCOM_SCM_PIL_PAS_MEM_SETUP	0x02
>  #define QCOM_SCM_PIL_PAS_AUTH_AND_RESET	0x05
>  #define QCOM_SCM_PIL_PAS_SHUTDOWN	0x06
> -- 
> 2.34.1
>