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, 9 Jan 2014 12:58:16 +0200
From:	Sagi Grimberg <sagig@...lanox.com>
To:	"Nicholas A. Bellinger" <nab@...erainc.com>,
	target-devel <target-devel@...r.kernel.org>
CC:	linux-scsi <linux-scsi@...r.kernel.org>,
	linux-kernel <linux-kernel@...r.kernel.org>,
	"Martin K. Petersen" <martin.petersen@...cle.com>,
	Christoph Hellwig <hch@....de>, Hannes Reinecke <hare@...e.de>,
	Or Gerlitz <ogerlitz@...lanox.com>,
	Nicholas Bellinger <nab@...ux-iscsi.org>
Subject: Re: [PATCH 01/14] target: Add DIF related base definitions

On 1/8/2014 10:36 PM, Nicholas A. Bellinger wrote:
> From: Nicholas Bellinger <nab@...ux-iscsi.org>
>
> This patch adds DIF related definitions to target_core_base.h
> that includes enums for target_prot_op + target_prot_type +
> target_prot_version + target_guard_type + target_pi_error.
>
> Also included is struct se_dif_v1_tuple, along with changes
> to struct se_cmd, struct se_dev_attrib, and struct se_device.
>
> Also, add new se_subsystem_api->[init,free]_prot() callers
> used by target core code to setup backend specific protection
> information after the device has been configured.
>
> Enums taken from Sagi Grimberg's original patch.
>
> Cc: Martin K. Petersen <martin.petersen@...cle.com>
> Cc: Christoph Hellwig <hch@....de>
> Cc: Hannes Reinecke <hare@...e.de>
> Cc: Sagi Grimberg <sagig@...lanox.com>
> Cc: Or Gerlitz <ogerlitz@...lanox.com>
> Signed-off-by: Nicholas Bellinger <nab@...ux-iscsi.org>
>
> target: more defs
>
> Signed-off-by: Nicholas Bellinger <nab@...ux-iscsi.org>
> ---
>   include/target/target_core_backend.h |    2 ++
>   include/target/target_core_base.h    |   59 ++++++++++++++++++++++++++++++++++
>   2 files changed, 61 insertions(+)
>
> diff --git a/include/target/target_core_backend.h b/include/target/target_core_backend.h
> index 39e0114..930f30d 100644
> --- a/include/target/target_core_backend.h
> +++ b/include/target/target_core_backend.h
> @@ -41,6 +41,8 @@ struct se_subsystem_api {
>   	unsigned int (*get_io_opt)(struct se_device *);
>   	unsigned char *(*get_sense_buffer)(struct se_cmd *);
>   	bool (*get_write_cache)(struct se_device *);
> +	int (*init_prot)(struct se_device *);
> +	void (*free_prot)(struct se_device *);
>   };
>   
>   struct sbc_ops {
> diff --git a/include/target/target_core_base.h b/include/target/target_core_base.h
> index 45412a6..15f402c 100644
> --- a/include/target/target_core_base.h
> +++ b/include/target/target_core_base.h
> @@ -166,6 +166,7 @@ enum se_cmd_flags_table {
>   	SCF_COMPARE_AND_WRITE		= 0x00080000,
>   	SCF_COMPARE_AND_WRITE_POST	= 0x00100000,
>   	SCF_CMD_XCOPY_PASSTHROUGH	= 0x00200000,
> +	SCF_PROT			= 0x00400000,
>   };
>   
>   /* struct se_dev_entry->lun_flags and struct se_lun->lun_access */
> @@ -414,6 +415,45 @@ struct se_tmr_req {
>   	struct list_head	tmr_list;
>   };
>   
> +enum target_prot_op {
> +	TARGET_PROT_NORMAL,
> +	TARGET_PROT_READ_INSERT,
> +	TARGET_PROT_WRITE_INSERT,
> +	TARGET_PROT_READ_STRIP,
> +	TARGET_PROT_WRITE_STRIP,
> +	TARGET_PROT_READ_PASS,
> +	TARGET_PROT_WRITE_PASS,
> +};
> +
> +enum target_prot_type {
> +	TARGET_DIF_TYPE0_PROT,
> +	TARGET_DIF_TYPE1_PROT,
> +	TARGET_DIF_TYPE2_PROT,
> +	TARGET_DIF_TYPE3_PROT,
> +};
> +
> +enum target_prot_version {
> +	TARGET_DIF_V1 = 1,
> +	TARGET_DIF_V2 = 2,
> +};
> +
> +enum target_guard_type {
> +	TARGET_DIX_GUARD_CRC = 1,
> +	TARGET_DIX_GUARD_IP = 2,
> +};
> +
> +enum target_pi_error {
> +	TARGET_GUARD_CHECK_FAILED = 0x1,
> +	TARGET_APPTAG_CHECK_FAILED = 0x2,
> +	TARGET_REFTAG_CHECK_FAILED = 0x3,
> +};
> +
> +struct se_dif_v1_tuple {
> +	__be16			guard_tag;
> +	__be16			app_tag;
> +	__be32			ref_tag;
> +};
> +
>   struct se_cmd {
>   	/* SAM response code being sent to initiator */
>   	u8			scsi_status;
> @@ -498,6 +538,20 @@ struct se_cmd {
>   
>   	/* Used for lun->lun_ref counting */
>   	bool			lun_ref_active;
> +
> +	/* DIF related members */
> +	enum target_prot_op	prot_op;
> +	enum target_prot_type	prot_type;
> +	enum target_guard_type	bg_type;
> +	u16			bg_seed;
> +	u16			reftag_seed;
> +	u32			apptag_seed;
> +	u32			prot_length;
> +	struct scatterlist	*t_prot_sg;
> +	unsigned int		t_prot_nents;
> +	bool			prot_interleaved;
> +	enum target_pi_error	pi_err;
> +	u32			block_num;
>   };

Some of these guys are unreferenced... I figured these should provide 
necessary info both for the transport and the backstores.

Regarding prot_interleaved, I don't remember if we agreed to allow 
backstores to store the protection interleaved with the data, which 
seems to make some sense in pSCSI.
Anyway, I added this flag simply because some transports support it 
(iSER) and it might be useful to avoid de-interleaving + re-interleaving 
the buffers.
In my toy example I modified fileio to support both storing 
data+protection interleaved and storing protection in a seperate 
file.protection
(without verify - left it for the transport) and let the user choose via 
configfs (protection_handover).

I think we should hear more opinions here.

>   
>   struct se_ua {
> @@ -609,6 +663,9 @@ struct se_dev_attrib {
>   	int		emulate_tpws;
>   	int		emulate_caw;
>   	int		emulate_3pc;
> +	enum target_prot_type pi_prot_type;
> +	enum target_prot_version pi_prot_version;
> +	enum target_guard_type pi_guard_type;
>   	int		enforce_pr_isids;
>   	int		is_nonrot;
>   	int		emulate_rest_reord;
> @@ -739,6 +796,8 @@ struct se_device {
>   	/* Linked list for struct se_hba struct se_device list */
>   	struct list_head	dev_list;
>   	struct se_lun		xcopy_lun;
> +	/* Protection Information */
> +	int			prot_length;
>   };
>   
>   struct se_hba {

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ