* [PATCH 1/3 net-next] tg3: Fix jumbo loopback test on 5719
@ 2012-03-05 0:48 Michael Chan
2012-03-05 0:48 ` [PATCH 2/3 net-next] tg3: Add memory barriers to sync BD data Michael Chan
2012-03-05 1:39 ` [PATCH 1/3 net-next] tg3: Fix jumbo loopback test on 5719 David Miller
0 siblings, 2 replies; 6+ messages in thread
From: Michael Chan @ 2012-03-05 0:48 UTC (permalink / raw)
To: davem; +Cc: netdev
Loopback on 9K packet fails because the chip has a DMA limit of 4K. The
loopback test logic uses a single BD for simplicity. Fix it by reducing
the jumbo packet size to the DMA limit.
Signed-off-by: Michael Chan <mchan@broadcom.com>
Reviewed-by: Benjamin Li <benli@broadcom.com>
Reviewed-by: Matt Carlson <mcarlson@broadcom.com>
---
drivers/net/ethernet/broadcom/tg3.c | 14 +++++++++++---
1 files changed, 11 insertions(+), 3 deletions(-)
diff --git a/drivers/net/ethernet/broadcom/tg3.c b/drivers/net/ethernet/broadcom/tg3.c
index 76f33d5..4b3243b 100644
--- a/drivers/net/ethernet/broadcom/tg3.c
+++ b/drivers/net/ethernet/broadcom/tg3.c
@@ -11727,6 +11727,10 @@ static int tg3_run_loopback(struct tg3 *tp, u32 pktsz, bool tso_loopback)
} else {
num_pkts = 1;
data_off = ETH_HLEN;
+
+ if (tg3_flag(tp, USE_JUMBO_BDFLAG) &&
+ tx_len > VLAN_ETH_FRAME_LEN)
+ base_flags |= TXD_FLAG_JMB_PKT;
}
for (i = data_off; i < tx_len; i++)
@@ -11857,6 +11861,10 @@ static int tg3_test_loopback(struct tg3 *tp, u64 *data, bool do_extlpbk)
{
int err = -EIO;
u32 eee_cap;
+ u32 jmb_pkt_sz = 9000;
+
+ if (tp->dma_limit)
+ jmb_pkt_sz = tp->dma_limit - ETH_HLEN;
eee_cap = tp->phy_flags & TG3_PHYFLG_EEE_CAP;
tp->phy_flags &= ~TG3_PHYFLG_EEE_CAP;
@@ -11900,7 +11908,7 @@ static int tg3_test_loopback(struct tg3 *tp, u64 *data, bool do_extlpbk)
data[0] |= TG3_STD_LOOPBACK_FAILED;
if (tg3_flag(tp, JUMBO_RING_ENABLE) &&
- tg3_run_loopback(tp, 9000 + ETH_HLEN, false))
+ tg3_run_loopback(tp, jmb_pkt_sz + ETH_HLEN, false))
data[0] |= TG3_JMB_LOOPBACK_FAILED;
tg3_mac_loopback(tp, false);
@@ -11925,7 +11933,7 @@ static int tg3_test_loopback(struct tg3 *tp, u64 *data, bool do_extlpbk)
tg3_run_loopback(tp, ETH_FRAME_LEN, true))
data[1] |= TG3_TSO_LOOPBACK_FAILED;
if (tg3_flag(tp, JUMBO_RING_ENABLE) &&
- tg3_run_loopback(tp, 9000 + ETH_HLEN, false))
+ tg3_run_loopback(tp, jmb_pkt_sz + ETH_HLEN, false))
data[1] |= TG3_JMB_LOOPBACK_FAILED;
if (do_extlpbk) {
@@ -11943,7 +11951,7 @@ static int tg3_test_loopback(struct tg3 *tp, u64 *data, bool do_extlpbk)
tg3_run_loopback(tp, ETH_FRAME_LEN, true))
data[2] |= TG3_TSO_LOOPBACK_FAILED;
if (tg3_flag(tp, JUMBO_RING_ENABLE) &&
- tg3_run_loopback(tp, 9000 + ETH_HLEN, false))
+ tg3_run_loopback(tp, jmb_pkt_sz + ETH_HLEN, false))
data[2] |= TG3_JMB_LOOPBACK_FAILED;
}
--
1.7.1
^ permalink raw reply related [flat|nested] 6+ messages in thread
* [PATCH 2/3 net-next] tg3: Add memory barriers to sync BD data
2012-03-05 0:48 [PATCH 1/3 net-next] tg3: Fix jumbo loopback test on 5719 Michael Chan
@ 2012-03-05 0:48 ` Michael Chan
2012-03-05 0:48 ` [PATCH 3/3 net-next] tg3: Fix poor tx performance on 57766 after MTU change Michael Chan
2012-03-05 1:39 ` [PATCH 2/3 net-next] tg3: Add memory barriers to sync BD data David Miller
2012-03-05 1:39 ` [PATCH 1/3 net-next] tg3: Fix jumbo loopback test on 5719 David Miller
1 sibling, 2 replies; 6+ messages in thread
From: Michael Chan @ 2012-03-05 0:48 UTC (permalink / raw)
To: davem; +Cc: netdev
for weak memory model architectures to ensure that the chip will DMA
valid BD data.
Signed-off-by: Michael Chan <mchan@broadcom.com>
Reviewed-by: Benjamin Li <benli@broadcom.com>
Reviewed-by: Matt Carlson <mcarlson@broadcom.com>
---
drivers/net/ethernet/broadcom/tg3.c | 9 +++++++++
1 files changed, 9 insertions(+), 0 deletions(-)
diff --git a/drivers/net/ethernet/broadcom/tg3.c b/drivers/net/ethernet/broadcom/tg3.c
index 4b3243b..0da93db 100644
--- a/drivers/net/ethernet/broadcom/tg3.c
+++ b/drivers/net/ethernet/broadcom/tg3.c
@@ -5928,6 +5928,9 @@ next_pkt_nopost:
/* Refill RX ring(s). */
if (!tg3_flag(tp, ENABLE_RSS)) {
+ /* Sync BD data before updating mailbox */
+ wmb();
+
if (work_mask & RXD_OPAQUE_RING_STD) {
tpr->rx_std_prod_idx = std_prod_idx &
tp->rx_std_ring_mask;
@@ -6970,6 +6973,9 @@ static netdev_tx_t tg3_start_xmit(struct sk_buff *skb, struct net_device *dev)
skb_tx_timestamp(skb);
netdev_sent_queue(tp->dev, skb->len);
+ /* Sync BD data before updating mailbox */
+ wmb();
+
/* Packets are ready, update Tx producer idx local and on card. */
tw32_tx_mbox(tnapi->prodmbox, entry);
@@ -11763,6 +11769,9 @@ static int tg3_run_loopback(struct tg3 *tp, u32 pktsz, bool tso_loopback)
tnapi->tx_prod++;
+ /* Sync BD data before updating mailbox */
+ wmb();
+
tw32_tx_mbox(tnapi->prodmbox, tnapi->tx_prod);
tr32_mailbox(tnapi->prodmbox);
--
1.7.1
^ permalink raw reply related [flat|nested] 6+ messages in thread
* [PATCH 3/3 net-next] tg3: Fix poor tx performance on 57766 after MTU change
2012-03-05 0:48 ` [PATCH 2/3 net-next] tg3: Add memory barriers to sync BD data Michael Chan
@ 2012-03-05 0:48 ` Michael Chan
2012-03-05 1:39 ` David Miller
2012-03-05 1:39 ` [PATCH 2/3 net-next] tg3: Add memory barriers to sync BD data David Miller
1 sibling, 1 reply; 6+ messages in thread
From: Michael Chan @ 2012-03-05 0:48 UTC (permalink / raw)
To: davem; +Cc: netdev
GRC reset causes the read DMA engine to go into a mode that breaks up
requests into 256 bytes. A PHY reset is required to bring it back to
the normal mode.
Signed-off-by: Michael Chan <mchan@broadcom.com>
---
drivers/net/ethernet/broadcom/tg3.c | 10 ++++++++--
1 files changed, 8 insertions(+), 2 deletions(-)
diff --git a/drivers/net/ethernet/broadcom/tg3.c b/drivers/net/ethernet/broadcom/tg3.c
index 0da93db..a8490a4 100644
--- a/drivers/net/ethernet/broadcom/tg3.c
+++ b/drivers/net/ethernet/broadcom/tg3.c
@@ -12267,7 +12267,7 @@ static inline void tg3_set_mtu(struct net_device *dev, struct tg3 *tp,
static int tg3_change_mtu(struct net_device *dev, int new_mtu)
{
struct tg3 *tp = netdev_priv(dev);
- int err;
+ int err, reset_phy = 0;
if (new_mtu < TG3_MIN_MTU || new_mtu > TG3_MAX_MTU(tp))
return -EINVAL;
@@ -12290,7 +12290,13 @@ static int tg3_change_mtu(struct net_device *dev, int new_mtu)
tg3_set_mtu(dev, tp, new_mtu);
- err = tg3_restart_hw(tp, 0);
+ /* Reset PHY, otherwise the read DMA engine will be in a mode that
+ * breaks all requests to 256 bytes.
+ */
+ if (GET_ASIC_REV(tp->pci_chip_rev_id) == ASIC_REV_57766)
+ reset_phy = 1;
+
+ err = tg3_restart_hw(tp, reset_phy);
if (!err)
tg3_netif_start(tp);
--
1.7.1
^ permalink raw reply related [flat|nested] 6+ messages in thread
* Re: [PATCH 1/3 net-next] tg3: Fix jumbo loopback test on 5719
2012-03-05 0:48 [PATCH 1/3 net-next] tg3: Fix jumbo loopback test on 5719 Michael Chan
2012-03-05 0:48 ` [PATCH 2/3 net-next] tg3: Add memory barriers to sync BD data Michael Chan
@ 2012-03-05 1:39 ` David Miller
1 sibling, 0 replies; 6+ messages in thread
From: David Miller @ 2012-03-05 1:39 UTC (permalink / raw)
To: mchan; +Cc: netdev
From: "Michael Chan" <mchan@broadcom.com>
Date: Sun, 4 Mar 2012 16:48:13 -0800
> Loopback on 9K packet fails because the chip has a DMA limit of 4K. The
> loopback test logic uses a single BD for simplicity. Fix it by reducing
> the jumbo packet size to the DMA limit.
>
> Signed-off-by: Michael Chan <mchan@broadcom.com>
> Reviewed-by: Benjamin Li <benli@broadcom.com>
> Reviewed-by: Matt Carlson <mcarlson@broadcom.com>
Applied.
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [PATCH 2/3 net-next] tg3: Add memory barriers to sync BD data
2012-03-05 0:48 ` [PATCH 2/3 net-next] tg3: Add memory barriers to sync BD data Michael Chan
2012-03-05 0:48 ` [PATCH 3/3 net-next] tg3: Fix poor tx performance on 57766 after MTU change Michael Chan
@ 2012-03-05 1:39 ` David Miller
1 sibling, 0 replies; 6+ messages in thread
From: David Miller @ 2012-03-05 1:39 UTC (permalink / raw)
To: mchan; +Cc: netdev
From: "Michael Chan" <mchan@broadcom.com>
Date: Sun, 4 Mar 2012 16:48:14 -0800
> for weak memory model architectures to ensure that the chip will DMA
> valid BD data.
>
> Signed-off-by: Michael Chan <mchan@broadcom.com>
> Reviewed-by: Benjamin Li <benli@broadcom.com>
> Reviewed-by: Matt Carlson <mcarlson@broadcom.com>
Applied.
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [PATCH 3/3 net-next] tg3: Fix poor tx performance on 57766 after MTU change
2012-03-05 0:48 ` [PATCH 3/3 net-next] tg3: Fix poor tx performance on 57766 after MTU change Michael Chan
@ 2012-03-05 1:39 ` David Miller
0 siblings, 0 replies; 6+ messages in thread
From: David Miller @ 2012-03-05 1:39 UTC (permalink / raw)
To: mchan; +Cc: netdev
From: "Michael Chan" <mchan@broadcom.com>
Date: Sun, 4 Mar 2012 16:48:15 -0800
> GRC reset causes the read DMA engine to go into a mode that breaks up
> requests into 256 bytes. A PHY reset is required to bring it back to
> the normal mode.
>
> Signed-off-by: Michael Chan <mchan@broadcom.com>
Applied.
^ permalink raw reply [flat|nested] 6+ messages in thread
end of thread, other threads:[~2012-03-05 1:39 UTC | newest]
Thread overview: 6+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2012-03-05 0:48 [PATCH 1/3 net-next] tg3: Fix jumbo loopback test on 5719 Michael Chan
2012-03-05 0:48 ` [PATCH 2/3 net-next] tg3: Add memory barriers to sync BD data Michael Chan
2012-03-05 0:48 ` [PATCH 3/3 net-next] tg3: Fix poor tx performance on 57766 after MTU change Michael Chan
2012-03-05 1:39 ` David Miller
2012-03-05 1:39 ` [PATCH 2/3 net-next] tg3: Add memory barriers to sync BD data David Miller
2012-03-05 1:39 ` [PATCH 1/3 net-next] tg3: Fix jumbo loopback test on 5719 David Miller
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).