Merge branch 'for-davem' of git://git.kernel.org/pub/scm/linux/kernel…

…/git/bwh/sfc-next

Ben Hutchings says:

====================
1. More workarounds for TX queue flush failures that can occur during
   interface reconfiguration.
2. Fix spurious failure of a firmware request running during a system
   clock change, e.g. ntpd started at the same time as driver load.
3. Fix inconsistent statistics after a firmware upgrade.
4. Fix a variable (non-)initialisation in offline self-test that can
   make it more disruptive than intended.
5. Fix a race that can (at least) cause an assertion failure.
6. Miscellaneous cleanup.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>

drivers/net/ethernet/sfc/efx.c

@@ -106,8 +106,8 @@ static struct workqueue_struct *reset_workqueue;
  *
  * This is only used in MSI-X interrupt mode
  */
-static unsigned int separate_tx_channels;
-module_param(separate_tx_channels, uint, 0444);
+static bool separate_tx_channels;
+module_param(separate_tx_channels, bool, 0444);
 MODULE_PARM_DESC(separate_tx_channels,
 		 "Use separate channels for TX and RX");
 
@@ -160,8 +160,8 @@ static unsigned int rss_cpus;
 module_param(rss_cpus, uint, 0444);
 MODULE_PARM_DESC(rss_cpus, "Number of CPUs to use for Receive-Side Scaling");
 
-static int phy_flash_cfg;
-module_param(phy_flash_cfg, int, 0644);
+static bool phy_flash_cfg;
+module_param(phy_flash_cfg, bool, 0644);
 MODULE_PARM_DESC(phy_flash_cfg, "Set PHYs into reflash mode initially");
 
 static unsigned irq_adapt_low_thresh = 8000;
@@ -2279,7 +2279,7 @@ int efx_reset(struct efx_nic *efx, enum reset_type method)
 	netif_info(efx, drv, efx->net_dev, "resetting (%s)\n",
 		   RESET_TYPE(method));
 
-	netif_device_detach(efx->net_dev);
+	efx_device_detach_sync(efx);
 	efx_reset_down(efx, method);
 
 	rc = efx->type->reset(efx, method);
@@ -2758,7 +2758,7 @@ static int efx_pm_freeze(struct device *dev)
 	if (efx->state != STATE_DISABLED) {
 		efx->state = STATE_UNINIT;
 
-		netif_device_detach(efx->net_dev);
+		efx_device_detach_sync(efx);
 
 		efx_stop_all(efx);
 		efx_stop_interrupts(efx, false);

drivers/net/ethernet/sfc/efx.h

@@ -163,4 +163,17 @@ extern void efx_link_status_changed(struct efx_nic *efx);
 extern void efx_link_set_advertising(struct efx_nic *efx, u32);
 extern void efx_link_set_wanted_fc(struct efx_nic *efx, u8);
 
+static inline void efx_device_detach_sync(struct efx_nic *efx)
+{
+	struct net_device *dev = efx->net_dev;
+
+	/* Lock/freeze all TX queues so that we can be sure the
+	 * TX scheduler is stopped when we're done and before
+	 * netif_device_present() becomes false.
+	 */
+	netif_tx_lock(dev);
+	netif_device_detach(dev);
+	netif_tx_unlock(dev);
+}
+
 #endif /* EFX_EFX_H */

drivers/net/ethernet/sfc/ethtool.c

@@ -816,6 +816,9 @@ static int efx_ethtool_reset(struct net_device *net_dev, u32 *flags)
 /* MAC address mask including only MC flag */
 static const u8 mac_addr_mc_mask[ETH_ALEN] = { 0x01, 0, 0, 0, 0, 0 };
 
+#define IP4_ADDR_FULL_MASK	((__force __be32)~0)
+#define PORT_FULL_MASK		((__force __be16)~0)
+
 static int efx_ethtool_get_class_rule(struct efx_nic *efx,
 				      struct ethtool_rx_flow_spec *rule)
 {
@@ -865,12 +868,12 @@ static int efx_ethtool_get_class_rule(struct efx_nic *efx,
 			&spec, &proto, &ip_entry->ip4dst, &ip_entry->pdst,
 			&ip_entry->ip4src, &ip_entry->psrc);
 		EFX_WARN_ON_PARANOID(rc);
-		ip_mask->ip4src = ~0;
-		ip_mask->psrc = ~0;
+		ip_mask->ip4src = IP4_ADDR_FULL_MASK;
+		ip_mask->psrc = PORT_FULL_MASK;
 	}
 	rule->flow_type = (proto == IPPROTO_TCP) ? TCP_V4_FLOW : UDP_V4_FLOW;
-	ip_mask->ip4dst = ~0;
-	ip_mask->pdst = ~0;
+	ip_mask->ip4dst = IP4_ADDR_FULL_MASK;
+	ip_mask->pdst = PORT_FULL_MASK;
 	return rc;
 }
 
@@ -971,7 +974,7 @@ static int efx_ethtool_set_class_rule(struct efx_nic *efx,
 
 	/* Check for unsupported extensions */
 	if ((rule->flow_type & FLOW_EXT) &&
-	    (rule->m_ext.vlan_etype | rule->m_ext.data[0] |
+	    (rule->m_ext.vlan_etype || rule->m_ext.data[0] ||
 	     rule->m_ext.data[1]))
 		return -EINVAL;
 
@@ -986,16 +989,16 @@ static int efx_ethtool_set_class_rule(struct efx_nic *efx,
 			    IPPROTO_TCP : IPPROTO_UDP);
 
 		/* Must match all of destination, */
-		if ((__force u32)~ip_mask->ip4dst |
-		    (__force u16)~ip_mask->pdst)
+		if (!(ip_mask->ip4dst == IP4_ADDR_FULL_MASK &&
+		      ip_mask->pdst == PORT_FULL_MASK))
 			return -EINVAL;
 		/* all or none of source, */
-		if ((ip_mask->ip4src | ip_mask->psrc) &&
-		    ((__force u32)~ip_mask->ip4src |
-		     (__force u16)~ip_mask->psrc))
+		if ((ip_mask->ip4src || ip_mask->psrc) &&
+		    !(ip_mask->ip4src == IP4_ADDR_FULL_MASK &&
+		      ip_mask->psrc == PORT_FULL_MASK))
 			return -EINVAL;
 		/* and nothing else */
-		if (ip_mask->tos | rule->m_ext.vlan_tci)
+		if (ip_mask->tos || rule->m_ext.vlan_tci)
 			return -EINVAL;
 
 		if (ip_mask->ip4src)

drivers/net/ethernet/sfc/falcon.c

@@ -1792,6 +1792,7 @@ const struct efx_nic_type falcon_a1_nic_type = {
 	.remove_port = falcon_remove_port,
 	.handle_global_event = falcon_handle_global_event,
 	.prepare_flush = falcon_prepare_flush,
+	.finish_flush = efx_port_dummy_op_void,
 	.update_stats = falcon_update_nic_stats,
 	.start_stats = falcon_start_nic_stats,
 	.stop_stats = falcon_stop_nic_stats,
@@ -1834,6 +1835,7 @@ const struct efx_nic_type falcon_b0_nic_type = {
 	.remove_port = falcon_remove_port,
 	.handle_global_event = falcon_handle_global_event,
 	.prepare_flush = falcon_prepare_flush,
+	.finish_flush = efx_port_dummy_op_void,
 	.update_stats = falcon_update_nic_stats,
 	.start_stats = falcon_start_nic_stats,
 	.stop_stats = falcon_stop_nic_stats,

drivers/net/ethernet/sfc/io.h

@@ -22,22 +22,21 @@
  *
  * Notes on locking strategy:
  *
- * Most CSRs are 128-bit (oword) and therefore cannot be read or
- * written atomically.  Access from the host is buffered by the Bus
- * Interface Unit (BIU).  Whenever the host reads from the lowest
- * address of such a register, or from the address of a different such
- * register, the BIU latches the register's value.  Subsequent reads
- * from higher addresses of the same register will read the latched
- * value.  Whenever the host writes part of such a register, the BIU
- * collects the written value and does not write to the underlying
- * register until all 4 dwords have been written.  A similar buffering
- * scheme applies to host access to the NIC's 64-bit SRAM.
+ * Many CSRs are very wide and cannot be read or written atomically.
+ * Writes from the host are buffered by the Bus Interface Unit (BIU)
+ * up to 128 bits.  Whenever the host writes part of such a register,
+ * the BIU collects the written value and does not write to the
+ * underlying register until all 4 dwords have been written.  A
+ * similar buffering scheme applies to host access to the NIC's 64-bit
+ * SRAM.
  *
- * Access to different CSRs and 64-bit SRAM words must be serialised,
- * since interleaved access can result in lost writes or lost
- * information from read-to-clear fields.  We use efx_nic::biu_lock
- * for this.  (We could use separate locks for read and write, but
- * this is not normally a performance bottleneck.)
+ * Writes to different CSRs and 64-bit SRAM words must be serialised,
+ * since interleaved access can result in lost writes.  We use
+ * efx_nic::biu_lock for this.
+ *
+ * We also serialise reads from 128-bit CSRs and SRAM with the same
+ * spinlock.  This may not be necessary, but it doesn't really matter
+ * as there are no such reads on the fast path.
  *
  * The DMA descriptor pointers (RX_DESC_UPD and TX_DESC_UPD) are
  * 128-bit but are special-cased in the BIU to avoid the need for
@@ -204,20 +203,6 @@ static inline void efx_reado_table(struct efx_nic *efx, efx_oword_t *value,
 	efx_reado(efx, value, reg + index * sizeof(efx_oword_t));
 }
 
-/* Write a 32-bit CSR forming part of a table, or 32-bit SRAM */
-static inline void efx_writed_table(struct efx_nic *efx, efx_dword_t *value,
-				       unsigned int reg, unsigned int index)
-{
-	efx_writed(efx, value, reg + index * sizeof(efx_oword_t));
-}
-
-/* Read a 32-bit CSR forming part of a table, or 32-bit SRAM */
-static inline void efx_readd_table(struct efx_nic *efx, efx_dword_t *value,
-				   unsigned int reg, unsigned int index)
-{
-	efx_readd(efx, value, reg + index * sizeof(efx_dword_t));
-}
-
 /* Page-mapped register block size */
 #define EFX_PAGE_BLOCK_SIZE 0x2000
 

drivers/net/ethernet/sfc/mcdi.c

@@ -22,7 +22,7 @@
  **************************************************************************
  */
 
-#define MCDI_RPC_TIMEOUT       10 /*seconds */
+#define MCDI_RPC_TIMEOUT       (10 * HZ)
 
 #define MCDI_PDU(efx)							\
 	(efx_port_num(efx) ? MC_SMEM_P1_PDU_OFST : MC_SMEM_P0_PDU_OFST)
@@ -120,7 +120,7 @@ static void efx_mcdi_copyout(struct efx_nic *efx, u8 *outbuf, size_t outlen)
 static int efx_mcdi_poll(struct efx_nic *efx)
 {
 	struct efx_mcdi_iface *mcdi = efx_mcdi(efx);
-	unsigned int time, finish;
+	unsigned long time, finish;
 	unsigned int respseq, respcmd, error;
 	unsigned int pdu = FR_CZ_MC_TREG_SMEM + MCDI_PDU(efx);
 	unsigned int rc, spins;
@@ -136,7 +136,7 @@ static int efx_mcdi_poll(struct efx_nic *efx)
 	 * and poll once a jiffy (approximately)
 	 */
 	spins = TICK_USEC;
-	finish = get_seconds() + MCDI_RPC_TIMEOUT;
+	finish = jiffies + MCDI_RPC_TIMEOUT;
 
 	while (1) {
 		if (spins != 0) {
@@ -146,7 +146,7 @@ static int efx_mcdi_poll(struct efx_nic *efx)
 			schedule_timeout_uninterruptible(1);
 		}
 
-		time = get_seconds();
+		time = jiffies;
 
 		rmb();
 		efx_readd(efx, &reg, pdu);
@@ -158,7 +158,7 @@ static int efx_mcdi_poll(struct efx_nic *efx)
 		    EFX_DWORD_FIELD(reg, MCDI_HEADER_RESPONSE))
 			break;
 
-		if (time >= finish)
+		if (time_after(time, finish))
 			return -ETIMEDOUT;
 	}
 
@@ -207,7 +207,9 @@ static int efx_mcdi_poll(struct efx_nic *efx)
 	return 0;
 }
 
-/* Test and clear MC-rebooted flag for this port/function */
+/* Test and clear MC-rebooted flag for this port/function; reset
+ * software state as necessary.
+ */
 int efx_mcdi_poll_reboot(struct efx_nic *efx)
 {
 	unsigned int addr = FR_CZ_MC_TREG_SMEM + MCDI_STATUS(efx);
@@ -223,6 +225,11 @@ int efx_mcdi_poll_reboot(struct efx_nic *efx)
 	if (value == 0)
 		return 0;
 
+	/* MAC statistics have been cleared on the NIC; clear our copy
+	 * so that efx_update_diff_stat() can continue to work.
+	 */
+	memset(&efx->mac_stats, 0, sizeof(efx->mac_stats));
+
 	EFX_ZERO_DWORD(reg);
 	efx_writed(efx, &reg, addr);
 
@@ -250,7 +257,7 @@ static int efx_mcdi_await_completion(struct efx_nic *efx)
 	if (wait_event_timeout(
 		    mcdi->wq,
 		    atomic_read(&mcdi->state) == MCDI_STATE_COMPLETED,
-		    msecs_to_jiffies(MCDI_RPC_TIMEOUT * 1000)) == 0)
+		    MCDI_RPC_TIMEOUT) == 0)
 		return -ETIMEDOUT;
 
 	/* Check if efx_mcdi_set_mode() switched us back to polled completions.
@@ -1216,7 +1223,7 @@ int efx_mcdi_flush_rxqs(struct efx_nic *efx)
 
 	rc = efx_mcdi_rpc(efx, MC_CMD_FLUSH_RX_QUEUES, (u8 *)qid,
 			  count * sizeof(*qid), NULL, 0, NULL);
-	WARN_ON(rc > 0);
+	WARN_ON(rc < 0);
 
 	kfree(qid);
 

drivers/net/ethernet/sfc/net_driver.h

@@ -200,6 +200,7 @@ struct efx_tx_queue {
 	/* Members shared between paths and sometimes updated */
 	unsigned int empty_read_count ____cacheline_aligned_in_smp;
 #define EFX_EMPTY_COUNT_VALID 0x80000000
+	atomic_t flush_outstanding;
 };
 
 /**
@@ -907,6 +908,7 @@ static inline unsigned int efx_port_num(struct efx_nic *efx)
  * @remove_port: Free resources allocated by probe_port()
  * @handle_global_event: Handle a "global" event (may be %NULL)
  * @prepare_flush: Prepare the hardware for flushing the DMA queues
+ * @finish_flush: Clean up after flushing the DMA queues
  * @update_stats: Update statistics not provided by event handling
  * @start_stats: Start the regular fetching of statistics
  * @stop_stats: Stop the regular fetching of statistics
@@ -954,6 +956,7 @@ struct efx_nic_type {
 	void (*remove_port)(struct efx_nic *efx);
 	bool (*handle_global_event)(struct efx_channel *channel, efx_qword_t *);
 	void (*prepare_flush)(struct efx_nic *efx);
+	void (*finish_flush)(struct efx_nic *efx);
 	void (*update_stats)(struct efx_nic *efx);
 	void (*start_stats)(struct efx_nic *efx);
 	void (*stop_stats)(struct efx_nic *efx);