linux-kernel - Re: [PATCH v3] scsi: ufs: ufs-pci: Correct check for RESET DSM

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: <4b889b61-5aa0-892f-0808-352545193e47@intel.com>
Date:   Fri, 29 Jul 2022 11:02:17 +0300
From:   Adrian Hunter <adrian.hunter@...el.com>
To:     Daniil Lunev <dlunev@...omium.org>,
        Bart Van Assche <bvanassche@....org>
Cc:     Avri Altman <avri.altman@....com>, Bean Huo <beanhuo@...ron.com>,
        "James E.J. Bottomley" <jejb@...ux.ibm.com>,
        "Martin K. Petersen" <martin.petersen@...cle.com>,
        linux-kernel@...r.kernel.org, linux-scsi@...r.kernel.org
Subject: Re: [PATCH v3] scsi: ufs: ufs-pci: Correct check for RESET DSM

On 28/07/22 04:18, Daniil Lunev wrote:
> dsm_fns is a bitmap, and it is 0-indexed according to the check in
> __intel_dsm function. But common initialization was checking it as if it
> was 1-indexed. This patch corrects the discrepancy. This change won't
> break any existing calls to the function, since before the change both
> bits 0 and 1 were checked and needed to be set.
> 
> Signed-off-by: Daniil Lunev <dlunev@...omium.org>

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

> 
> ---
> 
> Changes in v3:
> * Commit message typos
> * Args indentation
> 
> Changes in v2:
> * Make __INTEL_DSM_SUPPORTED a function
> * use `1u` instead of `1` in shift operator
> 
>  drivers/ufs/host/ufshcd-pci.c | 17 +++++++++++++----
>  1 file changed, 13 insertions(+), 4 deletions(-)
> 
> diff --git a/drivers/ufs/host/ufshcd-pci.c b/drivers/ufs/host/ufshcd-pci.c
> index 04166bda41daa..b9572fcc3bda5 100644
> --- a/drivers/ufs/host/ufshcd-pci.c
> +++ b/drivers/ufs/host/ufshcd-pci.c
> @@ -24,7 +24,7 @@ struct ufs_host {
>  	void (*late_init)(struct ufs_hba *hba);
>  };
>  
> -enum {
> +enum intel_ufs_dsm_func_id {
>  	INTEL_DSM_FNS		=  0,
>  	INTEL_DSM_RESET		=  1,
>  };
> @@ -42,6 +42,15 @@ static const guid_t intel_dsm_guid =
>  	GUID_INIT(0x1A4832A0, 0x7D03, 0x43CA,
>  		  0xB0, 0x20, 0xF6, 0xDC, 0xD1, 0x2A, 0x19, 0x50);
>  
> +static bool __intel_dsm_supported(struct intel_host *host,
> +				  enum intel_ufs_dsm_func_id fn)
> +{
> +	return fn < 32 && fn >= 0 && (host->dsm_fns & (1u << fn));
> +}
> +
> +#define INTEL_DSM_SUPPORTED(host, name) \
> +	__intel_dsm_supported(host, INTEL_DSM_##name)
> +
>  static int __intel_dsm(struct intel_host *intel_host, struct device *dev,
>  		       unsigned int fn, u32 *result)
>  {
> @@ -71,7 +80,7 @@ static int __intel_dsm(struct intel_host *intel_host, struct device *dev,
>  static int intel_dsm(struct intel_host *intel_host, struct device *dev,
>  		     unsigned int fn, u32 *result)
>  {
> -	if (fn > 31 || !(intel_host->dsm_fns & (1 << fn)))
> +	if (!__intel_dsm_supported(intel_host, fn))
>  		return -EOPNOTSUPP;
>  
>  	return __intel_dsm(intel_host, dev, fn, result);
> @@ -300,7 +309,7 @@ static int ufs_intel_device_reset(struct ufs_hba *hba)
>  {
>  	struct intel_host *host = ufshcd_get_variant(hba);
>  
> -	if (host->dsm_fns & INTEL_DSM_RESET) {
> +	if (INTEL_DSM_SUPPORTED(host, RESET)) {
>  		u32 result = 0;
>  		int err;
>  
> @@ -342,7 +351,7 @@ static int ufs_intel_common_init(struct ufs_hba *hba)
>  		return -ENOMEM;
>  	ufshcd_set_variant(hba, host);
>  	intel_dsm_init(host, hba->dev);
> -	if (host->dsm_fns & INTEL_DSM_RESET) {
> +	if (INTEL_DSM_SUPPORTED(host, RESET)) {
>  		if (hba->vops->device_reset)
>  			hba->caps |= UFSHCD_CAP_DEEPSLEEP;
>  	} else {