sfc: Allow driver to cope with a lower number of VIs than it needs for RSS

Previously, the driver would refuse to load if it couldn't secure
enough VIs from the MC to fulfill its RSS requirements.
This was causing probe to fail on later functions in
configurations where we'd run out of VIs, such as having many
VFs.

This change allows the driver to load with fewer VIs, down to a
minimum of 2. A warning will be printed saying that RSS
requirements were not met, possibly affecting performance.

efx->max_tx_channels needs to be set to avoid going down the
failure path in efx_probe_nic() immediately in the loop after the
probe() NIC-type function.
Also, Set rc=ENOSPC when bombing out of efx_probe_nic due to lack
of VIs.
Signed-off-by: Shradha Shah <sshah@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

drivers/net/ethernet/sfc/ef10.c

@@ -295,11 +295,11 @@ static int efx_ef10_probe(struct efx_nic *efx)
 	/* We can have one VI for each 8K region.  However, until we
 	 * use TX option descriptors we need two TX queues per channel.
 	 */
-	efx->max_channels =
-		min_t(unsigned int,
+	efx->max_channels = min_t(unsigned int,
 				  EFX_MAX_CHANNELS,
 				  efx_ef10_mem_map_size(efx) /
 				  (EFX_VI_PAGE_SIZE * EFX_TXQ_TYPES));
+	efx->max_tx_channels = efx->max_channels;
 	if (WARN_ON(efx->max_channels == 0))
 		return -EIO;
 
@@ -824,11 +824,13 @@ static int efx_ef10_dimension_resources(struct efx_nic *efx)
 {
 	struct efx_ef10_nic_data *nic_data = efx->nic_data;
 	unsigned int uc_mem_map_size, wc_mem_map_size;
-	unsigned int min_vis, pio_write_vi_base, max_vis;
+	unsigned int min_vis = max(EFX_TXQ_TYPES,
+				   efx_separate_tx_channels ? 2 : 1);
+	unsigned int channel_vis, pio_write_vi_base, max_vis;
 	void __iomem *membase;
 	int rc;
 
-	min_vis = max(efx->n_channels, efx->n_tx_channels * EFX_TXQ_TYPES);
+	channel_vis = max(efx->n_channels, efx->n_tx_channels * EFX_TXQ_TYPES);
 
 #ifdef EFX_USE_PIO
 	/* Try to allocate PIO buffers if wanted and if the full
@@ -862,11 +864,11 @@ static int efx_ef10_dimension_resources(struct efx_nic *efx)
 	 * page size is >4K).  So we may allocate some extra VIs just
 	 * for writing PIO buffers through.
 	 *
-	 * The UC mapping contains (min_vis - 1) complete VIs and the
+	 * The UC mapping contains (channel_vis - 1) complete VIs and the
 	 * first half of the next VI.  Then the WC mapping begins with
 	 * the second half of this last VI.
 	 */
-	uc_mem_map_size = PAGE_ALIGN((min_vis - 1) * EFX_VI_PAGE_SIZE +
+	uc_mem_map_size = PAGE_ALIGN((channel_vis - 1) * EFX_VI_PAGE_SIZE +
 				     ER_DZ_TX_PIOBUF);
 	if (nic_data->n_piobufs) {
 		/* pio_write_vi_base rounds down to give the number of complete
@@ -881,7 +883,7 @@ static int efx_ef10_dimension_resources(struct efx_nic *efx)
 	} else {
 		pio_write_vi_base = 0;
 		wc_mem_map_size = 0;
-		max_vis = min_vis;
+		max_vis = channel_vis;
 	}
 
 	/* In case the last attached driver failed to free VIs, do it now */
@@ -893,6 +895,23 @@ static int efx_ef10_dimension_resources(struct efx_nic *efx)
 	if (rc != 0)
 		return rc;
 
+	if (nic_data->n_allocated_vis < channel_vis) {
+		netif_info(efx, drv, efx->net_dev,
+			   "Could not allocate enough VIs to satisfy RSS"
+			   " requirements. Performance may not be optimal.\n");
+		/* We didn't get the VIs to populate our channels.
+		 * We could keep what we got but then we'd have more
+		 * interrupts than we need.
+		 * Instead calculate new max_channels and restart
+		 */
+		efx->max_channels = nic_data->n_allocated_vis;
+		efx->max_tx_channels =
+			nic_data->n_allocated_vis / EFX_TXQ_TYPES;
+
+		efx_ef10_free_vis(efx);
+		return -EAGAIN;
+	}
+
 	/* If we didn't get enough VIs to map all the PIO buffers, free the
 	 * PIO buffers
 	 */

drivers/net/ethernet/sfc/efx.c

@@ -115,9 +115,9 @@ static struct workqueue_struct *reset_workqueue;
  *
  * This is only used in MSI-X interrupt mode
  */
-static bool separate_tx_channels;
-module_param(separate_tx_channels, bool, 0444);
-MODULE_PARM_DESC(separate_tx_channels,
+bool efx_separate_tx_channels;
+module_param(efx_separate_tx_channels, bool, 0444);
+MODULE_PARM_DESC(efx_separate_tx_channels,
 		 "Use separate channels for TX and RX");
 
 /* This is the weight assigned to each of the (per-channel) virtual
@@ -1391,7 +1391,7 @@ static int efx_probe_interrupts(struct efx_nic *efx)
 		unsigned int n_channels;
 
 		n_channels = efx_wanted_parallelism(efx);
-		if (separate_tx_channels)
+		if (efx_separate_tx_channels)
 			n_channels *= 2;
 		n_channels += extra_channels;
 		n_channels = min(n_channels, efx->max_channels);
@@ -1418,13 +1418,16 @@ static int efx_probe_interrupts(struct efx_nic *efx)
 			efx->n_channels = n_channels;
 			if (n_channels > extra_channels)
 				n_channels -= extra_channels;
-			if (separate_tx_channels) {
-				efx->n_tx_channels = max(n_channels / 2, 1U);
+			if (efx_separate_tx_channels) {
+				efx->n_tx_channels = min(max(n_channels / 2,
+							     1U),
+							 efx->max_tx_channels);
 				efx->n_rx_channels = max(n_channels -
 							 efx->n_tx_channels,
 							 1U);
 			} else {
-				efx->n_tx_channels = n_channels;
+				efx->n_tx_channels = min(n_channels,
+							 efx->max_tx_channels);
 				efx->n_rx_channels = n_channels;
 			}
 			for (i = 0; i < efx->n_channels; i++)
@@ -1450,7 +1453,7 @@ static int efx_probe_interrupts(struct efx_nic *efx)
 
 	/* Assume legacy interrupts */
 	if (efx->interrupt_mode == EFX_INT_MODE_LEGACY) {
-		efx->n_channels = 1 + (separate_tx_channels ? 1 : 0);
+		efx->n_channels = 1 + (efx_separate_tx_channels ? 1 : 0);
 		efx->n_rx_channels = 1;
 		efx->n_tx_channels = 1;
 		efx->legacy_irq = efx->pci_dev->irq;
@@ -1624,7 +1627,8 @@ static void efx_set_channels(struct efx_nic *efx)
 	struct efx_tx_queue *tx_queue;
 
 	efx->tx_channel_offset =
-		separate_tx_channels ? efx->n_channels - efx->n_tx_channels : 0;
+		efx_separate_tx_channels ?
+		efx->n_channels - efx->n_tx_channels : 0;
 
 	/* We need to mark which channels really have RX and TX
 	 * queues, and adjust the TX queue numbers if we have separate
@@ -1653,18 +1657,35 @@ static int efx_probe_nic(struct efx_nic *efx)
 	if (rc)
 		return rc;
 
-	/* Determine the number of channels and queues by trying to hook
-	 * in MSI-X interrupts. */
+	do {
+		if (!efx->max_channels || !efx->max_tx_channels) {
+			netif_err(efx, drv, efx->net_dev,
+				  "Insufficient resources to allocate"
+				  " any channels\n");
+			rc = -ENOSPC;
+			goto fail1;
+		}
+
+		/* Determine the number of channels and queues by trying
+		 * to hook in MSI-X interrupts.
+		 */
 		rc = efx_probe_interrupts(efx);
 		if (rc)
 			goto fail1;
 
 		efx_set_channels(efx);
 
+		/* dimension_resources can fail with EAGAIN */
 		rc = efx->type->dimension_resources(efx);
-	if (rc)
+		if (rc != 0 && rc != -EAGAIN)
 			goto fail2;
 
+		if (rc == -EAGAIN)
+			/* try again with new max_channels */
+			efx_remove_interrupts(efx);
+
+	} while (rc == -EAGAIN);
+
 	if (efx->n_channels > 1)
 		netdev_rss_key_fill(&efx->rx_hash_key,
 				    sizeof(efx->rx_hash_key));

drivers/net/ethernet/sfc/efx.h

@@ -35,6 +35,7 @@ void efx_xmit_done(struct efx_tx_queue *tx_queue, unsigned int index);
 int efx_setup_tc(struct net_device *net_dev, u8 num_tc);
 unsigned int efx_tx_max_skb_descs(struct efx_nic *efx);
 extern unsigned int efx_piobuf_size;
+extern bool efx_separate_tx_channels;
 
 /* RX */
 void efx_set_default_rx_indir_table(struct efx_nic *efx);

drivers/net/ethernet/sfc/falcon.c

@@ -2371,6 +2371,7 @@ static int falcon_probe_nic(struct efx_nic *efx)
 
 	efx->max_channels = (efx_nic_rev(efx) <= EFX_REV_FALCON_A1 ? 4 :
 			     EFX_MAX_CHANNELS);
+	efx->max_tx_channels = efx->max_channels;
 	efx->timer_quantum_ns = 4968; /* 621 cycles */
 
 	/* Initialise I2C adapter */

drivers/net/ethernet/sfc/net_driver.h

@@ -972,6 +972,7 @@ struct efx_nic {
 	unsigned next_buffer_table;
 
 	unsigned int max_channels;
+	unsigned int max_tx_channels;
 	unsigned n_channels;
 	unsigned n_rx_channels;
 	unsigned rss_spread;

drivers/net/ethernet/sfc/siena.c

@@ -262,6 +262,7 @@ static int siena_probe_nic(struct efx_nic *efx)
 	}
 
 	efx->max_channels = EFX_MAX_CHANNELS;
+	efx->max_tx_channels = EFX_MAX_CHANNELS;
 
 	efx_reado(efx, &reg, FR_AZ_CS_DEBUG);
 	efx->port_num = EFX_OWORD_FIELD(reg, FRF_CZ_CS_PORT_NUM) - 1;