linux-kernel - Re: [PATCH v2 09/18] firmware: qcom_scm-64: Move SMC register filling to qcom_scm_call

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]

Date:   Fri, 15 Nov 2019 15:57:04 -0800
From:   Stephen Boyd <swboyd@...omium.org>
To:     Elliot Berman <eberman@...eaurora.org>, agross@...nel.org,
        bjorn.andersson@...aro.org, saiprakash.ranjan@...eaurora.org
Cc:     Elliot Berman <eberman@...eaurora.org>, tsoni@...eaurora.org,
        sidgup@...eaurora.org, psodagud@...eaurora.org,
        linux-arm-msm@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH v2 09/18] firmware: qcom_scm-64: Move SMC register filling to qcom_scm_call_smccc

Quoting Elliot Berman (2019-11-12 13:22:45)
> diff --git a/drivers/firmware/qcom_scm-64.c b/drivers/firmware/qcom_scm-64.c
> index 4131093..977654bb 100644
> --- a/drivers/firmware/qcom_scm-64.c
> +++ b/drivers/firmware/qcom_scm-64.c
> @@ -54,6 +54,10 @@ struct qcom_scm_desc {
>         u32 owner;
>  };
>  
> +struct arm_smccc_args {
> +       unsigned long a[8];

Please call it 'args', not just 'a'.

> +};
> +
>  static u64 qcom_smccc_convention = -1;
>  static DEFINE_MUTEX(qcom_scm_lock);
>  
> @@ -95,12 +95,22 @@ static int ___qcom_scm_call_smccc(struct device *dev,
>  {
>         int arglen = desc->arginfo & 0xf;
>         int i;
> -       u64 x5 = desc->args[SMCCC_FIRST_EXT_IDX];
>         dma_addr_t args_phys = 0;
>         void *args_virt = NULL;
>         size_t alloc_len;
>         gfp_t flag = atomic ? GFP_ATOMIC : GFP_KERNEL;
> +       u32 smccc_call_type = atomic ? ARM_SMCCC_FAST_CALL : ARM_SMCCC_STD_CALL;
>         struct arm_smccc_res res;
> +       struct arm_smccc_args smc = {0};
> +
> +       smc.a[0] = ARM_SMCCC_CALL_VAL(
> +               smccc_call_type,
> +               qcom_smccc_convention,
> +               desc->owner,
> +               SMCCC_FUNCNUM(desc->svc, desc->cmd));
> +       smc.a[1] = desc->arginfo;
> +       for (i = 0; i < SMCCC_N_REG_ARGS; i++)
> +               smc.a[i + SMCCC_FIRST_REG_IDX] = desc->args[i];
>  
>         if (unlikely(arglen > SMCCC_N_REG_ARGS)) {
>                 alloc_len = SMCCC_N_EXT_ARGS * sizeof(u64);
> @@ -131,19 +141,18 @@ static int ___qcom_scm_call_smccc(struct device *dev,
>                         return -ENOMEM;
>                 }
>  
> -               x5 = args_phys;
> +               smc.a[SMCCC_LAST_REG_IDX] = args_phys;
>         }
>  
>         if (atomic) {
> -               __qcom_scm_call_do_quirk(desc, &res, x5, ARM_SMCCC_FAST_CALL);
> +               __qcom_scm_call_do_quirk(&smc, &res);
>         } else {
>                 int retry_count = 0;
>  
>                 do {
>                         mutex_lock(&qcom_scm_lock);
>  
> -                       __qcom_scm_call_do_quirk(desc, &res, x5,
> -                                                ARM_SMCCC_STD_CALL);
> +                       __qcom_scm_call_do_quirk(&smc, &res);
>  
>                         mutex_unlock(&qcom_scm_lock);
>  

Maybe we need to restructure this whole function to be a few steps

	setup_and_map_args()
	do_call()
	unmap_args()
	remap_error()

And pass some set of args to those functions. That would probably
provide clarity to this monstrously large function.