linux-kernel - Re: [PATCH 3/3] scsi: ufs: add quirk to support host reset only

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:   Thu, 27 May 2021 14:31:37 +0800
From:   Can Guo <cang@...eaurora.org>
To:     jongmin jeong <jjmin.jeong@...sung.com>
Cc:     jejb@...ux.ibm.com, martin.petersen@...cle.com,
        alim.akhtar@...sung.com, avri.altman@....com, beanhuo@...ron.com,
        adrian.hunter@...el.com, linux-scsi@...r.kernel.org,
        linux-kernel@...r.kernel.org
Subject: Re: [PATCH 3/3] scsi: ufs: add quirk to support host reset only

On 2021-05-27 11:09, jongmin jeong wrote:
> samsung ExynosAuto SoC has two types of host controller interface to
> support the virtualization of UFS Device.
> One is the physical host(PH) that the same as conventaional UFSHCI,
> and the other is the virtual host(VH) that support data transfer 
> function only.
> 
> In this structure, the virtual host does support host reset handler 
> only.
> This patch calls the host reset handler when abort or device reset 
> handler
> has occured in the virtual host.
> 
> Change-Id: I3f07e772415a35fe1e7374e02b3c37ef0bf5660d
> Signed-off-by: jongmin jeong <jjmin.jeong@...sung.com>
> ---
>  drivers/scsi/ufs/ufshcd.c | 7 +++++++
>  drivers/scsi/ufs/ufshcd.h | 6 ++++++
>  2 files changed, 13 insertions(+)
> 
> diff --git a/drivers/scsi/ufs/ufshcd.c b/drivers/scsi/ufs/ufshcd.c
> index 4787e40c6a2d..9d1912290f87 100644
> --- a/drivers/scsi/ufs/ufshcd.c
> +++ b/drivers/scsi/ufs/ufshcd.c
> @@ -6826,6 +6826,9 @@ static int ufshcd_eh_device_reset_handler(struct
> scsi_cmnd *cmd)
>  	u8 resp = 0xF, lun;
>  	unsigned long flags;
> 
> +	if (hba->quirks & UFSHCD_QUIRK_BROKEN_RESET_HANDLER)
> +		return ufshcd_eh_host_reset_handler(cmd);
> +
>  	host = cmd->device->host;
>  	hba = shost_priv(host);
> 
> @@ -6972,6 +6975,10 @@ static int ufshcd_abort(struct scsi_cmnd *cmd)
>  	host = cmd->device->host;
>  	hba = shost_priv(host);
>  	tag = cmd->request->tag;
> +
> +	if (hba->quirks & UFSHCD_QUIRK_BROKEN_RESET_HANDLER)
> +		return ufshcd_eh_host_reset_handler(cmd);
> +

Unless you are not using runtime PM. Otherwise, when abort happens
to SSU cmd (sent from pm ops), your change will lead to a live lock,
because ufshcd_eh_host_reset_handler() invokes error handling and
flushes it, error handling calls runtime_pm_get_sync(), which flushes
pm ops, while pm ops is blocked by SSU cmd.

To be on the safe side, you should move these codes after below
check in ufshcd_abort(), right before sending task abort TMRs.

  /*
   * Task abort to the device W-LUN is illegal. When this command
   * will fail, due to spec violation, scsi err handling next step
   * will be to send LU reset which, again, is a spec violation.
   * To avoid these unnecessary/illegal steps, first we clean up
   * the lrb taken by this cmd and mark the lrb as in_use, then
   * queue the eh_work and bail.
   */
   if (lrbp->lun == UFS_UPIU_UFS_DEVICE_WLUN) {
     ...

     goto out;
   }

Thanks,

Can Guo.

>  	lrbp = &hba->lrb[tag];
>  	if (!ufshcd_valid_tag(hba, tag)) {
>  		dev_err(hba->dev,
> diff --git a/drivers/scsi/ufs/ufshcd.h b/drivers/scsi/ufs/ufshcd.h
> index 0ab4c296be32..82a9c6889978 100644
> --- a/drivers/scsi/ufs/ufshcd.h
> +++ b/drivers/scsi/ufs/ufshcd.h
> @@ -581,6 +581,12 @@ enum ufshcd_quirks {
>  	 * support interface configuration.
>  	 */
>  	UFSHCD_QUIRK_SKIP_INTERFACE_CONFIGURATION	= 1 << 16,
> +
> +	/*
> +	 * This quirk needs to be enabled if the host controller support
> +	 * host reset handler only.
> +	 */
> +	UFSHCD_QUIRK_BROKEN_RESET_HANDLER		= 1 << 17,
>  };
> 
>  enum ufshcd_caps {