forked from mirrors/linux
		
	 59f4d59b25
			
		
	
	
		59f4d59b25
		
	
	
	
	
		
			
			E830 adds hardware support to prevent the VF from overflowing the PF mailbox with VIRTCHNL messages. E830 will use the hardware feature (ICE_F_MBX_LIMIT) instead of the software solution ice_is_malicious_vf(). To prevent a VF from overflowing the PF, the PF sets the number of messages per VF that can be in the PF's mailbox queue (ICE_MBX_OVERFLOW_WATERMARK). When the PF processes a message from a VF, the PF decrements the per VF message count using the E830_MBX_VF_DEC_TRIG register. Signed-off-by: Paul Greenwalt <paul.greenwalt@intel.com> Reviewed-by: Alexander Lobakin <aleksander.lobakin@intel.com> Tested-by: Rafal Romanowski <rafal.romanowski@intel.com> Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
		
			
				
	
	
		
			412 lines
		
	
	
	
		
			14 KiB
		
	
	
	
		
			C
		
	
	
	
	
	
			
		
		
	
	
			412 lines
		
	
	
	
		
			14 KiB
		
	
	
	
		
			C
		
	
	
	
	
	
| // SPDX-License-Identifier: GPL-2.0
 | |
| /* Copyright (c) 2018, Intel Corporation. */
 | |
| 
 | |
| #include "ice_common.h"
 | |
| #include "ice_vf_mbx.h"
 | |
| 
 | |
| /**
 | |
|  * ice_aq_send_msg_to_vf
 | |
|  * @hw: pointer to the hardware structure
 | |
|  * @vfid: VF ID to send msg
 | |
|  * @v_opcode: opcodes for VF-PF communication
 | |
|  * @v_retval: return error code
 | |
|  * @msg: pointer to the msg buffer
 | |
|  * @msglen: msg length
 | |
|  * @cd: pointer to command details
 | |
|  *
 | |
|  * Send message to VF driver (0x0802) using mailbox
 | |
|  * queue and asynchronously sending message via
 | |
|  * ice_sq_send_cmd() function
 | |
|  */
 | |
| int
 | |
| ice_aq_send_msg_to_vf(struct ice_hw *hw, u16 vfid, u32 v_opcode, u32 v_retval,
 | |
| 		      u8 *msg, u16 msglen, struct ice_sq_cd *cd)
 | |
| {
 | |
| 	struct ice_aqc_pf_vf_msg *cmd;
 | |
| 	struct ice_aq_desc desc;
 | |
| 
 | |
| 	ice_fill_dflt_direct_cmd_desc(&desc, ice_mbx_opc_send_msg_to_vf);
 | |
| 
 | |
| 	cmd = &desc.params.virt;
 | |
| 	cmd->id = cpu_to_le32(vfid);
 | |
| 
 | |
| 	desc.cookie_high = cpu_to_le32(v_opcode);
 | |
| 	desc.cookie_low = cpu_to_le32(v_retval);
 | |
| 
 | |
| 	if (msglen)
 | |
| 		desc.flags |= cpu_to_le16(ICE_AQ_FLAG_RD);
 | |
| 
 | |
| 	return ice_sq_send_cmd(hw, &hw->mailboxq, &desc, msg, msglen, cd);
 | |
| }
 | |
| 
 | |
| static const u32 ice_legacy_aq_to_vc_speed[] = {
 | |
| 	VIRTCHNL_LINK_SPEED_100MB,	/* BIT(0) */
 | |
| 	VIRTCHNL_LINK_SPEED_100MB,
 | |
| 	VIRTCHNL_LINK_SPEED_1GB,
 | |
| 	VIRTCHNL_LINK_SPEED_1GB,
 | |
| 	VIRTCHNL_LINK_SPEED_1GB,
 | |
| 	VIRTCHNL_LINK_SPEED_10GB,
 | |
| 	VIRTCHNL_LINK_SPEED_20GB,
 | |
| 	VIRTCHNL_LINK_SPEED_25GB,
 | |
| 	VIRTCHNL_LINK_SPEED_40GB,
 | |
| 	VIRTCHNL_LINK_SPEED_40GB,
 | |
| 	VIRTCHNL_LINK_SPEED_40GB,
 | |
| };
 | |
| 
 | |
| /**
 | |
|  * ice_conv_link_speed_to_virtchnl
 | |
|  * @adv_link_support: determines the format of the returned link speed
 | |
|  * @link_speed: variable containing the link_speed to be converted
 | |
|  *
 | |
|  * Convert link speed supported by HW to link speed supported by virtchnl.
 | |
|  * If adv_link_support is true, then return link speed in Mbps. Else return
 | |
|  * link speed as a VIRTCHNL_LINK_SPEED_* casted to a u32. Note that the caller
 | |
|  * needs to cast back to an enum virtchnl_link_speed in the case where
 | |
|  * adv_link_support is false, but when adv_link_support is true the caller can
 | |
|  * expect the speed in Mbps.
 | |
|  */
 | |
| u32 ice_conv_link_speed_to_virtchnl(bool adv_link_support, u16 link_speed)
 | |
| {
 | |
| 	/* convert a BIT() value into an array index */
 | |
| 	u32 index = fls(link_speed) - 1;
 | |
| 
 | |
| 	if (adv_link_support)
 | |
| 		return ice_get_link_speed(index);
 | |
| 	else if (index < ARRAY_SIZE(ice_legacy_aq_to_vc_speed))
 | |
| 		/* Virtchnl speeds are not defined for every speed supported in
 | |
| 		 * the hardware. To maintain compatibility with older AVF
 | |
| 		 * drivers, while reporting the speed the new speed values are
 | |
| 		 * resolved to the closest known virtchnl speeds
 | |
| 		 */
 | |
| 		return ice_legacy_aq_to_vc_speed[index];
 | |
| 
 | |
| 	return VIRTCHNL_LINK_SPEED_UNKNOWN;
 | |
| }
 | |
| 
 | |
| /* The mailbox overflow detection algorithm helps to check if there
 | |
|  * is a possibility of a malicious VF transmitting too many MBX messages to the
 | |
|  * PF.
 | |
|  * 1. The mailbox snapshot structure, ice_mbx_snapshot, is initialized during
 | |
|  * driver initialization in ice_init_hw() using ice_mbx_init_snapshot().
 | |
|  * The struct ice_mbx_snapshot helps to track and traverse a static window of
 | |
|  * messages within the mailbox queue while looking for a malicious VF.
 | |
|  *
 | |
|  * 2. When the caller starts processing its mailbox queue in response to an
 | |
|  * interrupt, the structure ice_mbx_snapshot is expected to be cleared before
 | |
|  * the algorithm can be run for the first time for that interrupt. This
 | |
|  * requires calling ice_mbx_reset_snapshot() as well as calling
 | |
|  * ice_mbx_reset_vf_info() for each VF tracking structure.
 | |
|  *
 | |
|  * 3. For every message read by the caller from the MBX Queue, the caller must
 | |
|  * call the detection algorithm's entry function ice_mbx_vf_state_handler().
 | |
|  * Before every call to ice_mbx_vf_state_handler() the struct ice_mbx_data is
 | |
|  * filled as it is required to be passed to the algorithm.
 | |
|  *
 | |
|  * 4. Every time a message is read from the MBX queue, a tracking structure
 | |
|  * for the VF must be passed to the state handler. The boolean output
 | |
|  * report_malvf from ice_mbx_vf_state_handler() serves as an indicator to the
 | |
|  * caller whether it must report this VF as malicious or not.
 | |
|  *
 | |
|  * 5. When a VF is identified to be malicious, the caller can send a message
 | |
|  * to the system administrator.
 | |
|  *
 | |
|  * 6. The PF is responsible for maintaining the struct ice_mbx_vf_info
 | |
|  * structure for each VF. The PF should clear the VF tracking structure if the
 | |
|  * VF is reset. When a VF is shut down and brought back up, we will then
 | |
|  * assume that the new VF is not malicious and may report it again if we
 | |
|  * detect it again.
 | |
|  *
 | |
|  * 7. The function ice_mbx_reset_snapshot() is called to reset the information
 | |
|  * in ice_mbx_snapshot for every new mailbox interrupt handled.
 | |
|  */
 | |
| #define ICE_RQ_DATA_MASK(rq_data) ((rq_data) & PF_MBX_ARQH_ARQH_M)
 | |
| /* Using the highest value for an unsigned 16-bit value 0xFFFF to indicate that
 | |
|  * the max messages check must be ignored in the algorithm
 | |
|  */
 | |
| #define ICE_IGNORE_MAX_MSG_CNT	0xFFFF
 | |
| 
 | |
| /**
 | |
|  * ice_mbx_reset_snapshot - Reset mailbox snapshot structure
 | |
|  * @snap: pointer to the mailbox snapshot
 | |
|  */
 | |
| static void ice_mbx_reset_snapshot(struct ice_mbx_snapshot *snap)
 | |
| {
 | |
| 	struct ice_mbx_vf_info *vf_info;
 | |
| 
 | |
| 	/* Clear mbx_buf in the mailbox snaphot structure and setting the
 | |
| 	 * mailbox snapshot state to a new capture.
 | |
| 	 */
 | |
| 	memset(&snap->mbx_buf, 0, sizeof(snap->mbx_buf));
 | |
| 	snap->mbx_buf.state = ICE_MAL_VF_DETECT_STATE_NEW_SNAPSHOT;
 | |
| 
 | |
| 	/* Reset message counts for all VFs to zero */
 | |
| 	list_for_each_entry(vf_info, &snap->mbx_vf, list_entry)
 | |
| 		vf_info->msg_count = 0;
 | |
| }
 | |
| 
 | |
| /**
 | |
|  * ice_mbx_traverse - Pass through mailbox snapshot
 | |
|  * @hw: pointer to the HW struct
 | |
|  * @new_state: new algorithm state
 | |
|  *
 | |
|  * Traversing the mailbox static snapshot without checking
 | |
|  * for malicious VFs.
 | |
|  */
 | |
| static void
 | |
| ice_mbx_traverse(struct ice_hw *hw,
 | |
| 		 enum ice_mbx_snapshot_state *new_state)
 | |
| {
 | |
| 	struct ice_mbx_snap_buffer_data *snap_buf;
 | |
| 	u32 num_iterations;
 | |
| 
 | |
| 	snap_buf = &hw->mbx_snapshot.mbx_buf;
 | |
| 
 | |
| 	/* As mailbox buffer is circular, applying a mask
 | |
| 	 * on the incremented iteration count.
 | |
| 	 */
 | |
| 	num_iterations = ICE_RQ_DATA_MASK(++snap_buf->num_iterations);
 | |
| 
 | |
| 	/* Checking either of the below conditions to exit snapshot traversal:
 | |
| 	 * Condition-1: If the number of iterations in the mailbox is equal to
 | |
| 	 * the mailbox head which would indicate that we have reached the end
 | |
| 	 * of the static snapshot.
 | |
| 	 * Condition-2: If the maximum messages serviced in the mailbox for a
 | |
| 	 * given interrupt is the highest possible value then there is no need
 | |
| 	 * to check if the number of messages processed is equal to it. If not
 | |
| 	 * check if the number of messages processed is greater than or equal
 | |
| 	 * to the maximum number of mailbox entries serviced in current work item.
 | |
| 	 */
 | |
| 	if (num_iterations == snap_buf->head ||
 | |
| 	    (snap_buf->max_num_msgs_mbx < ICE_IGNORE_MAX_MSG_CNT &&
 | |
| 	     ++snap_buf->num_msg_proc >= snap_buf->max_num_msgs_mbx))
 | |
| 		*new_state = ICE_MAL_VF_DETECT_STATE_NEW_SNAPSHOT;
 | |
| }
 | |
| 
 | |
| /**
 | |
|  * ice_mbx_detect_malvf - Detect malicious VF in snapshot
 | |
|  * @hw: pointer to the HW struct
 | |
|  * @vf_info: mailbox tracking structure for a VF
 | |
|  * @new_state: new algorithm state
 | |
|  * @is_malvf: boolean output to indicate if VF is malicious
 | |
|  *
 | |
|  * This function tracks the number of asynchronous messages
 | |
|  * sent per VF and marks the VF as malicious if it exceeds
 | |
|  * the permissible number of messages to send.
 | |
|  */
 | |
| static int
 | |
| ice_mbx_detect_malvf(struct ice_hw *hw, struct ice_mbx_vf_info *vf_info,
 | |
| 		     enum ice_mbx_snapshot_state *new_state,
 | |
| 		     bool *is_malvf)
 | |
| {
 | |
| 	/* increment the message count for this VF */
 | |
| 	vf_info->msg_count++;
 | |
| 
 | |
| 	if (vf_info->msg_count >= ICE_ASYNC_VF_MSG_THRESHOLD)
 | |
| 		*is_malvf = true;
 | |
| 
 | |
| 	/* continue to iterate through the mailbox snapshot */
 | |
| 	ice_mbx_traverse(hw, new_state);
 | |
| 
 | |
| 	return 0;
 | |
| }
 | |
| 
 | |
| /**
 | |
|  * ice_mbx_vf_dec_trig_e830 - Decrements the VF mailbox queue counter
 | |
|  * @hw: pointer to the HW struct
 | |
|  * @event: pointer to the control queue receive event
 | |
|  *
 | |
|  * This function triggers to decrement the counter
 | |
|  * MBX_VF_IN_FLIGHT_MSGS_AT_PF_CNT when the driver replenishes
 | |
|  * the buffers at the PF mailbox queue.
 | |
|  */
 | |
| void ice_mbx_vf_dec_trig_e830(const struct ice_hw *hw,
 | |
| 			      const struct ice_rq_event_info *event)
 | |
| {
 | |
| 	u16 vfid = le16_to_cpu(event->desc.retval);
 | |
| 
 | |
| 	wr32(hw, E830_MBX_VF_DEC_TRIG(vfid), 1);
 | |
| }
 | |
| 
 | |
| /**
 | |
|  * ice_mbx_vf_clear_cnt_e830 - Clear the VF mailbox queue count
 | |
|  * @hw: pointer to the HW struct
 | |
|  * @vf_id: VF ID in the PF space
 | |
|  *
 | |
|  * This function clears the counter MBX_VF_IN_FLIGHT_MSGS_AT_PF_CNT, and should
 | |
|  * be called when a VF is created and on VF reset.
 | |
|  */
 | |
| void ice_mbx_vf_clear_cnt_e830(const struct ice_hw *hw, u16 vf_id)
 | |
| {
 | |
| 	u32 reg = rd32(hw, E830_MBX_VF_IN_FLIGHT_MSGS_AT_PF_CNT(vf_id));
 | |
| 
 | |
| 	wr32(hw, E830_MBX_VF_DEC_TRIG(vf_id), reg);
 | |
| }
 | |
| 
 | |
| /**
 | |
|  * ice_mbx_vf_state_handler - Handle states of the overflow algorithm
 | |
|  * @hw: pointer to the HW struct
 | |
|  * @mbx_data: pointer to structure containing mailbox data
 | |
|  * @vf_info: mailbox tracking structure for the VF in question
 | |
|  * @report_malvf: boolean output to indicate whether VF should be reported
 | |
|  *
 | |
|  * The function serves as an entry point for the malicious VF
 | |
|  * detection algorithm by handling the different states and state
 | |
|  * transitions of the algorithm:
 | |
|  * New snapshot: This state is entered when creating a new static
 | |
|  * snapshot. The data from any previous mailbox snapshot is
 | |
|  * cleared and a new capture of the mailbox head and tail is
 | |
|  * logged. This will be the new static snapshot to detect
 | |
|  * asynchronous messages sent by VFs. On capturing the snapshot
 | |
|  * and depending on whether the number of pending messages in that
 | |
|  * snapshot exceed the watermark value, the state machine enters
 | |
|  * traverse or detect states.
 | |
|  * Traverse: If pending message count is below watermark then iterate
 | |
|  * through the snapshot without any action on VF.
 | |
|  * Detect: If pending message count exceeds watermark traverse
 | |
|  * the static snapshot and look for a malicious VF.
 | |
|  */
 | |
| int
 | |
| ice_mbx_vf_state_handler(struct ice_hw *hw, struct ice_mbx_data *mbx_data,
 | |
| 			 struct ice_mbx_vf_info *vf_info, bool *report_malvf)
 | |
| {
 | |
| 	struct ice_mbx_snapshot *snap = &hw->mbx_snapshot;
 | |
| 	struct ice_mbx_snap_buffer_data *snap_buf;
 | |
| 	struct ice_ctl_q_info *cq = &hw->mailboxq;
 | |
| 	enum ice_mbx_snapshot_state new_state;
 | |
| 	bool is_malvf = false;
 | |
| 	int status = 0;
 | |
| 
 | |
| 	if (!report_malvf || !mbx_data || !vf_info)
 | |
| 		return -EINVAL;
 | |
| 
 | |
| 	*report_malvf = false;
 | |
| 
 | |
| 	/* When entering the mailbox state machine assume that the VF
 | |
| 	 * is not malicious until detected.
 | |
| 	 */
 | |
| 	 /* Checking if max messages allowed to be processed while servicing current
 | |
| 	  * interrupt is not less than the defined AVF message threshold.
 | |
| 	  */
 | |
| 	if (mbx_data->max_num_msgs_mbx <= ICE_ASYNC_VF_MSG_THRESHOLD)
 | |
| 		return -EINVAL;
 | |
| 
 | |
| 	/* The watermark value should not be lesser than the threshold limit
 | |
| 	 * set for the number of asynchronous messages a VF can send to mailbox
 | |
| 	 * nor should it be greater than the maximum number of messages in the
 | |
| 	 * mailbox serviced in current interrupt.
 | |
| 	 */
 | |
| 	if (mbx_data->async_watermark_val < ICE_ASYNC_VF_MSG_THRESHOLD ||
 | |
| 	    mbx_data->async_watermark_val > mbx_data->max_num_msgs_mbx)
 | |
| 		return -EINVAL;
 | |
| 
 | |
| 	new_state = ICE_MAL_VF_DETECT_STATE_INVALID;
 | |
| 	snap_buf = &snap->mbx_buf;
 | |
| 
 | |
| 	switch (snap_buf->state) {
 | |
| 	case ICE_MAL_VF_DETECT_STATE_NEW_SNAPSHOT:
 | |
| 		/* Clear any previously held data in mailbox snapshot structure. */
 | |
| 		ice_mbx_reset_snapshot(snap);
 | |
| 
 | |
| 		/* Collect the pending ARQ count, number of messages processed and
 | |
| 		 * the maximum number of messages allowed to be processed from the
 | |
| 		 * Mailbox for current interrupt.
 | |
| 		 */
 | |
| 		snap_buf->num_pending_arq = mbx_data->num_pending_arq;
 | |
| 		snap_buf->num_msg_proc = mbx_data->num_msg_proc;
 | |
| 		snap_buf->max_num_msgs_mbx = mbx_data->max_num_msgs_mbx;
 | |
| 
 | |
| 		/* Capture a new static snapshot of the mailbox by logging the
 | |
| 		 * head and tail of snapshot and set num_iterations to the tail
 | |
| 		 * value to mark the start of the iteration through the snapshot.
 | |
| 		 */
 | |
| 		snap_buf->head = ICE_RQ_DATA_MASK(cq->rq.next_to_clean +
 | |
| 						  mbx_data->num_pending_arq);
 | |
| 		snap_buf->tail = ICE_RQ_DATA_MASK(cq->rq.next_to_clean - 1);
 | |
| 		snap_buf->num_iterations = snap_buf->tail;
 | |
| 
 | |
| 		/* Pending ARQ messages returned by ice_clean_rq_elem
 | |
| 		 * is the difference between the head and tail of the
 | |
| 		 * mailbox queue. Comparing this value against the watermark
 | |
| 		 * helps to check if we potentially have malicious VFs.
 | |
| 		 */
 | |
| 		if (snap_buf->num_pending_arq >=
 | |
| 		    mbx_data->async_watermark_val) {
 | |
| 			new_state = ICE_MAL_VF_DETECT_STATE_DETECT;
 | |
| 			status = ice_mbx_detect_malvf(hw, vf_info, &new_state, &is_malvf);
 | |
| 		} else {
 | |
| 			new_state = ICE_MAL_VF_DETECT_STATE_TRAVERSE;
 | |
| 			ice_mbx_traverse(hw, &new_state);
 | |
| 		}
 | |
| 		break;
 | |
| 
 | |
| 	case ICE_MAL_VF_DETECT_STATE_TRAVERSE:
 | |
| 		new_state = ICE_MAL_VF_DETECT_STATE_TRAVERSE;
 | |
| 		ice_mbx_traverse(hw, &new_state);
 | |
| 		break;
 | |
| 
 | |
| 	case ICE_MAL_VF_DETECT_STATE_DETECT:
 | |
| 		new_state = ICE_MAL_VF_DETECT_STATE_DETECT;
 | |
| 		status = ice_mbx_detect_malvf(hw, vf_info, &new_state, &is_malvf);
 | |
| 		break;
 | |
| 
 | |
| 	default:
 | |
| 		new_state = ICE_MAL_VF_DETECT_STATE_INVALID;
 | |
| 		status = -EIO;
 | |
| 	}
 | |
| 
 | |
| 	snap_buf->state = new_state;
 | |
| 
 | |
| 	/* Only report VFs as malicious the first time we detect it */
 | |
| 	if (is_malvf && !vf_info->malicious) {
 | |
| 		vf_info->malicious = 1;
 | |
| 		*report_malvf = true;
 | |
| 	}
 | |
| 
 | |
| 	return status;
 | |
| }
 | |
| 
 | |
| /**
 | |
|  * ice_mbx_clear_malvf - Clear VF mailbox info
 | |
|  * @vf_info: the mailbox tracking structure for a VF
 | |
|  *
 | |
|  * In case of a VF reset, this function shall be called to clear the VF's
 | |
|  * current mailbox tracking state.
 | |
|  */
 | |
| void ice_mbx_clear_malvf(struct ice_mbx_vf_info *vf_info)
 | |
| {
 | |
| 	vf_info->malicious = 0;
 | |
| 	vf_info->msg_count = 0;
 | |
| }
 | |
| 
 | |
| /**
 | |
|  * ice_mbx_init_vf_info - Initialize a new VF mailbox tracking info
 | |
|  * @hw: pointer to the hardware structure
 | |
|  * @vf_info: the mailbox tracking info structure for a VF
 | |
|  *
 | |
|  * Initialize a VF mailbox tracking info structure and insert it into the
 | |
|  * snapshot list.
 | |
|  *
 | |
|  * If you remove the VF, you must also delete the associated VF info structure
 | |
|  * from the linked list.
 | |
|  */
 | |
| void ice_mbx_init_vf_info(struct ice_hw *hw, struct ice_mbx_vf_info *vf_info)
 | |
| {
 | |
| 	struct ice_mbx_snapshot *snap = &hw->mbx_snapshot;
 | |
| 
 | |
| 	ice_mbx_clear_malvf(vf_info);
 | |
| 	list_add(&vf_info->list_entry, &snap->mbx_vf);
 | |
| }
 | |
| 
 | |
| /**
 | |
|  * ice_mbx_init_snapshot - Initialize mailbox snapshot data
 | |
|  * @hw: pointer to the hardware structure
 | |
|  *
 | |
|  * Clear the mailbox snapshot structure and initialize the VF mailbox list.
 | |
|  */
 | |
| void ice_mbx_init_snapshot(struct ice_hw *hw)
 | |
| {
 | |
| 	struct ice_mbx_snapshot *snap = &hw->mbx_snapshot;
 | |
| 
 | |
| 	INIT_LIST_HEAD(&snap->mbx_vf);
 | |
| 	ice_mbx_reset_snapshot(snap);
 | |
| }
 |