* Re: [PATCH net-next] bridge: multicast router list manipulation
From: David Miller @ 2010-04-27 23:51 UTC (permalink / raw)
To: shemminger; +Cc: mirqus, herbert, netdev
In-Reply-To: <20100427164412.791e4999@nehalam>
From: Stephen Hemminger <shemminger@vyatta.com>
Date: Tue, 27 Apr 2010 16:44:12 -0700
> Code that doesn't need rcu for traversal should not use it.
> It just confuses things and implies that rcu_read_lock is held
> which it is not in this code.
Fair enough, I've reverted by change.
This would have been helped if either the patch was split up into
two pieces (one straight hlist macro conversion, another removing
the RCU tag) _or_ the commit message explained why the RCU tag
could be elided.
^ permalink raw reply
* Re: [PATCH net-next] bridge: multicast router list manipulation
From: Stephen Hemminger @ 2010-04-27 23:44 UTC (permalink / raw)
To: David Miller; +Cc: mirqus, herbert, netdev
In-Reply-To: <20100427.162811.233419535.davem@davemloft.net>
On Tue, 27 Apr 2010 16:28:11 -0700 (PDT)
David Miller <davem@davemloft.net> wrote:
> From: Stephen Hemminger <shemminger@vyatta.com>
> Date: Tue, 27 Apr 2010 16:25:30 -0700
>
> > On Wed, 28 Apr 2010 01:11:52 +0200
> > Michał Mirosław <mirqus@gmail.com> wrote:
> >
> >> 2010/4/27 Stephen Hemminger <shemminger@vyatta.com>:
> >> > I prefer that the hlist be only accessed through the hlist macro
> >> > objects. Explicit twiddling of links (especially with RCU) exposes
> >> > the code to future bugs.
> >> [...]
> >> > - port->rlist.pprev = h;
> >> > - port->rlist.next = p;
> >> > - rcu_assign_pointer(*h, &port->rlist);
> >> > - if (p)
> >> > - p->pprev = &port->rlist.next;
> >> > + hlist_for_each_entry(p, n, &br->router_list, rlist) {
> >>
> >> Shouldn't this be hlist_for_each_entry_rcu?
> >>
> >
> > This code should already be protected by br->multicast_lock
>
> But the adds et al. use RCU already, I think we should be consistent
> one way or another.
>
> I've already made Michał's suggested change to net-next-2.6, if you
> think _rcu() isn't necessary then trim it from all the hlist calls
> in this function.
Code that doesn't need rcu for traversal should not use it.
It just confuses things and implies that rcu_read_lock is held
which it is not in this code.
^ permalink raw reply
* Re: [PATCH v2] gianfar: Wait for both RX and TX to stop
From: David Miller @ 2010-04-27 23:44 UTC (permalink / raw)
To: afleming; +Cc: netdev
In-Reply-To: <1272408703-20931-1-git-send-email-afleming@freescale.com>
From: Andy Fleming <afleming@freescale.com>
Date: Tue, 27 Apr 2010 17:51:43 -0500
> When gracefully stopping the controller, the driver was continuing if
> *either* RX or TX had stopped. We need to wait for both, or the
> controller could get into an invalid state.
>
> Signed-off-by: Andy Fleming <afleming@freescale.com>
Applied, thanks.
^ permalink raw reply
* Re: [PATCH] RCU: don't turn off lockdep when find suspicious rcu_dereference_check() usage
From: David Miller @ 2010-04-27 23:42 UTC (permalink / raw)
To: paulmck
Cc: miles.lane, ebiederm, vgoyal, eparis, laijs, mingo, peterz,
linux-kernel, nauman, eric.dumazet, netdev, jens.axboe,
guijianfeng, lizf, johannes
In-Reply-To: <20100427233111.GF2424@linux.vnet.ibm.com>
From: "Paul E. McKenney" <paulmck@linux.vnet.ibm.com>
Date: Tue, 27 Apr 2010 16:31:11 -0700
> Thank you both! I have added Eric's Acked-by and Miles's Tested-by.
How can you? It's in net-next-2.6 already :-)
^ permalink raw reply
* Re: [PATCH] RCU: don't turn off lockdep when find suspicious rcu_dereference_check() usage
From: Paul E. McKenney @ 2010-04-27 23:31 UTC (permalink / raw)
To: Miles Lane
Cc: Eric W. Biederman, Vivek Goyal, Eric Paris, Lai Jiangshan,
Ingo Molnar, Peter Zijlstra, LKML, nauman, eric.dumazet, netdev,
Jens Axboe, Gui Jianfeng, Li Zefan, Johannes Berg
In-Reply-To: <g2ya44ae5cd1004271058p9f73fba1v3f611693e81869c3@mail.gmail.com>
On Tue, Apr 27, 2010 at 01:58:30PM -0400, Miles Lane wrote:
> On Tue, Apr 27, 2010 at 12:22 PM, Paul E. McKenney
> <paulmck@linux.vnet.ibm.com> wrote:
> > On Mon, Apr 26, 2010 at 09:27:44PM -0700, Paul E. McKenney wrote:
> >> On Mon, Apr 26, 2010 at 11:35:10AM -0700, Eric W. Biederman wrote:
> >> > "Paul E. McKenney" <paulmck@linux.vnet.ibm.com> writes:
> >> >
> >> > > Eric Dumazet traced these down to a commit from Eric Biederman.
> >> > >
> >> > > If I don't hear from Eric Biederman in a few days, I will attempt a
> >> > > patch, but it would be more likely to be correct coming from someone
> >> > > with a better understanding of the code. ;-)
> >> >
> >> > I already replied.
> >> >
> >> > http://lkml.org/lkml/2010/4/21/420
> >>
> >> You did indeed!!! This experience is giving me an even better appreciation
> >> of the maintainers' ability to keep all their patches straight!
> >>
> >> I will put together something based on your suggestion.
> >
> > How about the following?
> >
> > Thanx, Paul
> >
> > ------------------------------------------------------------------------
> >
> > commit 85fa42bd568ab99c375f018761ae6345249942cd
> > Author: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
> > Date: Mon Apr 26 21:40:05 2010 -0700
> >
> > net: suppress RCU lockdep false positive in twsk_net()
> >
> > Calls to twsk_net() are in some cases protected by reference counting
> > as an alternative to RCU protection. Cases covered by reference counts
> > include __inet_twsk_kill(), inet_twsk_free(), inet_twdr_do_twkill_work(),
> > inet_twdr_twcal_tick(), and tcp_timewait_state_process(). RCU is used
> > by inet_twsk_purge(). Locking is used by established_get_first()
> > and established_get_next(). Finally, __inet_twsk_hashdance() is an
> > initialization case.
> >
> > It appears to be non-trivial to locate the appropriate locks and
> > reference counts from within twsk_net(), so used rcu_dereference_raw().
> >
> > Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
> >
> > diff --git a/include/net/inet_timewait_sock.h b/include/net/inet_timewait_sock.h
> > index 79f67ea..a066fdd 100644
> > --- a/include/net/inet_timewait_sock.h
> > +++ b/include/net/inet_timewait_sock.h
> > @@ -224,7 +224,9 @@ static inline
> > struct net *twsk_net(const struct inet_timewait_sock *twsk)
> > {
> > #ifdef CONFIG_NET_NS
> > - return rcu_dereference(twsk->tw_net);
> > + return rcu_dereference_raw(twsk->tw_net); /* protected by locking, */
> > + /* reference counting, */
> > + /* initialization, or RCU. */
> > #else
> > return &init_net;
> > #endif
> >
>
> Worked for me. Thanks!
Thank you both! I have added Eric's Acked-by and Miles's Tested-by.
Thanx, Paul
^ permalink raw reply
* Re: [net-next-2.6 PATCH 7/7] ixgbevf: use DMA API instead of PCI DMA functions
From: David Miller @ 2010-04-27 23:30 UTC (permalink / raw)
To: jeffrey.t.kirsher; +Cc: netdev, gospo, nicholasx.d.nunley
In-Reply-To: <20100427231048.26173.83950.stgit@localhost.localdomain>
From: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Date: Tue, 27 Apr 2010 16:10:50 -0700
> From: Nick Nunley <nicholasx.d.nunley@intel.com>
>
> Signed-off-by: Nicholas Nunley <nicholasx.d.nunley@intel.com>
> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Applied.
^ permalink raw reply
* Re: [net-next-2.6 PATCH 6/7] ixgbe: use DMA API instead of PCI DMA functions
From: David Miller @ 2010-04-27 23:30 UTC (permalink / raw)
To: jeffrey.t.kirsher; +Cc: netdev, gospo, nicholasx.d.nunley
In-Reply-To: <20100427231025.26173.91168.stgit@localhost.localdomain>
From: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Date: Tue, 27 Apr 2010 16:10:27 -0700
> From: Nick Nunley <nicholasx.d.nunley@intel.com>
>
> Signed-off-by: Nicholas Nunley <nicholasx.d.nunley@intel.com>
> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Applied.
^ permalink raw reply
* Re: [net-next-2.6 PATCH 5/7] ixgb: use DMA API instead of PCI DMA functions
From: David Miller @ 2010-04-27 23:30 UTC (permalink / raw)
To: jeffrey.t.kirsher; +Cc: netdev, gospo, nicholasx.d.nunley
In-Reply-To: <20100427231001.26173.17582.stgit@localhost.localdomain>
From: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Date: Tue, 27 Apr 2010 16:10:03 -0700
> From: Nick Nunley <nicholasx.d.nunley@intel.com>
>
> Signed-off-by: Nicholas Nunley <nicholasx.d.nunley@intel.com>
> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Applied.
^ permalink raw reply
* Re: [net-next-2.6 PATCH 4/7] igbvf: use DMA API instead of PCI DMA functions
From: David Miller @ 2010-04-27 23:30 UTC (permalink / raw)
To: jeffrey.t.kirsher; +Cc: netdev, gospo, nicholasx.d.nunley
In-Reply-To: <20100427230942.26173.47559.stgit@localhost.localdomain>
From: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Date: Tue, 27 Apr 2010 16:09:44 -0700
> From: Nick Nunley <nicholasx.d.nunley@intel.com>
>
> Signed-off-by: Nicholas Nunley <nicholasx.d.nunley@intel.com>
> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Applied.
^ permalink raw reply
* Re: [net-next-2.6 PATCH 3/7] igb: convert igb from using PCI DMA functions to using DMA API functions
From: David Miller @ 2010-04-27 23:30 UTC (permalink / raw)
To: jeffrey.t.kirsher; +Cc: netdev, gospo, alexander.h.duyck
In-Reply-To: <20100427230923.26173.83374.stgit@localhost.localdomain>
From: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Date: Tue, 27 Apr 2010 16:09:25 -0700
> From: Alexander Duyck <alexander.h.duyck@intel.com>
>
> This patch makes it so that igb now uses the DMA API functions instead of
> the PCI API functions. To do this the pci_dev pointer that was in the
> rings has been replaced with a device pointer, and as a result all
> references to [tr]x_ring->pdev have been replaced with [tr]x_ring->dev.
>
> This patch is based of of work originally done by Nicholas Nunley.
> Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Applied.
^ permalink raw reply
* Re: [net-next-2.6 PATCH 2/7] e1000e: use DMA API instead of PCI DMA functions
From: David Miller @ 2010-04-27 23:30 UTC (permalink / raw)
To: jeffrey.t.kirsher; +Cc: netdev, gospo, nicholasx.d.nunley
In-Reply-To: <20100427230903.26173.89549.stgit@localhost.localdomain>
From: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Date: Tue, 27 Apr 2010 16:09:05 -0700
> From: Nick Nunley <nicholasx.d.nunley@intel.com>
>
> Signed-off-by: Nicholas Nunley <nicholasx.d.nunley@intel.com>
> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Applied.
^ permalink raw reply
* Re: [net-next-2.6 PATCH 1/7] e1000: use DMA API instead of PCI DMA functions
From: David Miller @ 2010-04-27 23:30 UTC (permalink / raw)
To: jeffrey.t.kirsher; +Cc: netdev, gospo, nicholasx.d.nunley
In-Reply-To: <20100427230836.26173.59671.stgit@localhost.localdomain>
From: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Date: Tue, 27 Apr 2010 16:08:45 -0700
> From: Nick Nunley <nicholasx.d.nunley@intel.com>
>
> Signed-off-by: Nicholas Nunley <nicholasx.d.nunley@intel.com>
> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Applied.
^ permalink raw reply
* Re: [PATCH net-next] bridge: multicast router list manipulation
From: David Miller @ 2010-04-27 23:28 UTC (permalink / raw)
To: shemminger; +Cc: mirqus, herbert, netdev
In-Reply-To: <20100427162530.04deb551@nehalam>
From: Stephen Hemminger <shemminger@vyatta.com>
Date: Tue, 27 Apr 2010 16:25:30 -0700
> On Wed, 28 Apr 2010 01:11:52 +0200
> Michał Mirosław <mirqus@gmail.com> wrote:
>
>> 2010/4/27 Stephen Hemminger <shemminger@vyatta.com>:
>> > I prefer that the hlist be only accessed through the hlist macro
>> > objects. Explicit twiddling of links (especially with RCU) exposes
>> > the code to future bugs.
>> [...]
>> > - port->rlist.pprev = h;
>> > - port->rlist.next = p;
>> > - rcu_assign_pointer(*h, &port->rlist);
>> > - if (p)
>> > - p->pprev = &port->rlist.next;
>> > + hlist_for_each_entry(p, n, &br->router_list, rlist) {
>>
>> Shouldn't this be hlist_for_each_entry_rcu?
>>
>
> This code should already be protected by br->multicast_lock
But the adds et al. use RCU already, I think we should be consistent
one way or another.
I've already made Michał's suggested change to net-next-2.6, if you
think _rcu() isn't necessary then trim it from all the hlist calls
in this function.
^ permalink raw reply
* Re: [PATCH net-next] bridge: multicast router list manipulation
From: David Miller @ 2010-04-27 23:27 UTC (permalink / raw)
To: mirqus; +Cc: shemminger, herbert, netdev
In-Reply-To: <y2ne2d7436a1004271611h5dc5f184gcbbc6748839c3c58@mail.gmail.com>
From: Michał Mirosław <mirqus@gmail.com>
Date: Wed, 28 Apr 2010 01:11:52 +0200
> 2010/4/27 Stephen Hemminger <shemminger@vyatta.com>:
>> I prefer that the hlist be only accessed through the hlist macro
>> objects. Explicit twiddling of links (especially with RCU) exposes
>> the code to future bugs.
> [...]
>> - port->rlist.pprev = h;
>> - port->rlist.next = p;
>> - rcu_assign_pointer(*h, &port->rlist);
>> - if (p)
>> - p->pprev = &port->rlist.next;
>> + hlist_for_each_entry(p, n, &br->router_list, rlist) {
>
> Shouldn't this be hlist_for_each_entry_rcu?
I think so, I've committed the following to net-next-2.6 to fix this,
thanks!
bridge: Use hlist_for_each_entry_rcu() in br_multicast_add_router()
Noticed by Michał Mirosław.
Signed-off-by: David S. Miller <davem@davemloft.net>
---
net/bridge/br_multicast.c | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/net/bridge/br_multicast.c b/net/bridge/br_multicast.c
index fcba313..e29c9b7 100644
--- a/net/bridge/br_multicast.c
+++ b/net/bridge/br_multicast.c
@@ -1045,7 +1045,7 @@ static void br_multicast_add_router(struct net_bridge *br,
struct net_bridge_port *p;
struct hlist_node *n, *last = NULL;
- hlist_for_each_entry(p, n, &br->router_list, rlist) {
+ hlist_for_each_entry_rcu(p, n, &br->router_list, rlist) {
if ((unsigned long) port >= (unsigned long) p) {
hlist_add_before_rcu(n, &port->rlist);
return;
--
1.7.0.4
^ permalink raw reply related
* Re: [PATCH net-next] bridge: multicast router list manipulation
From: Stephen Hemminger @ 2010-04-27 23:25 UTC (permalink / raw)
To: Michał Mirosław; +Cc: Herbert Xu, David S. Miller, netdev
In-Reply-To: <y2ne2d7436a1004271611h5dc5f184gcbbc6748839c3c58@mail.gmail.com>
On Wed, 28 Apr 2010 01:11:52 +0200
Michał Mirosław <mirqus@gmail.com> wrote:
> 2010/4/27 Stephen Hemminger <shemminger@vyatta.com>:
> > I prefer that the hlist be only accessed through the hlist macro
> > objects. Explicit twiddling of links (especially with RCU) exposes
> > the code to future bugs.
> [...]
> > - port->rlist.pprev = h;
> > - port->rlist.next = p;
> > - rcu_assign_pointer(*h, &port->rlist);
> > - if (p)
> > - p->pprev = &port->rlist.next;
> > + hlist_for_each_entry(p, n, &br->router_list, rlist) {
>
> Shouldn't this be hlist_for_each_entry_rcu?
>
This code should already be protected by br->multicast_lock
^ permalink raw reply
* Re: [PATCH net-next v2] cxgb4: set skb->rxhash
From: David Miller @ 2010-04-27 23:22 UTC (permalink / raw)
To: dm; +Cc: netdev
In-Reply-To: <1272409856-12535-1-git-send-email-dm@chelsio.com>
From: Dimitris Michailidis <dm@chelsio.com>
Date: Tue, 27 Apr 2010 16:10:56 -0700
> Implement the ->set_flags ethtool method to control NETIF_F_RXHASH and
> set skb->rxhash to the HW calculated hash accordingly.
>
> Follow Eric Dumazet's suggestion and use the hash value raw.
>
> Signed-off-by: Dimitris Michailidis <dm@chelsio.com>
Applied, thanks Dimitris.
^ permalink raw reply
* [PATCH net-next v2] cxgb4: set skb->rxhash
From: Dimitris Michailidis @ 2010-04-27 23:10 UTC (permalink / raw)
To: netdev; +Cc: Dimitris Michailidis
Implement the ->set_flags ethtool method to control NETIF_F_RXHASH and
set skb->rxhash to the HW calculated hash accordingly.
Follow Eric Dumazet's suggestion and use the hash value raw.
Signed-off-by: Dimitris Michailidis <dm@chelsio.com>
---
drivers/net/cxgb4/cxgb4_main.c | 15 ++++++++++++++-
drivers/net/cxgb4/sge.c | 7 ++++++-
drivers/net/cxgb4/t4_msg.h | 1 +
3 files changed, 21 insertions(+), 2 deletions(-)
diff --git a/drivers/net/cxgb4/cxgb4_main.c b/drivers/net/cxgb4/cxgb4_main.c
index 5f582db..1bad500 100644
--- a/drivers/net/cxgb4/cxgb4_main.c
+++ b/drivers/net/cxgb4/cxgb4_main.c
@@ -1711,6 +1711,18 @@ static int set_tso(struct net_device *dev, u32 value)
return 0;
}
+static int set_flags(struct net_device *dev, u32 flags)
+{
+ if (flags & ~ETH_FLAG_RXHASH)
+ return -EOPNOTSUPP;
+
+ if (flags & ETH_FLAG_RXHASH)
+ dev->features |= NETIF_F_RXHASH;
+ else
+ dev->features &= ~NETIF_F_RXHASH;
+ return 0;
+}
+
static struct ethtool_ops cxgb_ethtool_ops = {
.get_settings = get_settings,
.set_settings = set_settings,
@@ -1741,6 +1753,7 @@ static struct ethtool_ops cxgb_ethtool_ops = {
.get_wol = get_wol,
.set_wol = set_wol,
.set_tso = set_tso,
+ .set_flags = set_flags,
.flash_device = set_flash,
};
@@ -3203,7 +3216,7 @@ static int __devinit init_one(struct pci_dev *pdev,
netdev->features |= NETIF_F_SG | NETIF_F_TSO | NETIF_F_TSO6;
netdev->features |= NETIF_F_IP_CSUM | NETIF_F_IPV6_CSUM;
- netdev->features |= NETIF_F_GRO | highdma;
+ netdev->features |= NETIF_F_GRO | NETIF_F_RXHASH | highdma;
netdev->features |= NETIF_F_HW_VLAN_TX | NETIF_F_HW_VLAN_RX;
netdev->vlan_features = netdev->features & VLAN_FEAT;
diff --git a/drivers/net/cxgb4/sge.c b/drivers/net/cxgb4/sge.c
index 65d91c4..e804ffc 100644
--- a/drivers/net/cxgb4/sge.c
+++ b/drivers/net/cxgb4/sge.c
@@ -1524,6 +1524,8 @@ static void do_gro(struct sge_eth_rxq *rxq, const struct pkt_gl *gl,
skb->truesize += skb->data_len;
skb->ip_summed = CHECKSUM_UNNECESSARY;
skb_record_rx_queue(skb, rxq->rspq.idx);
+ if (rxq->rspq.netdev->features & NETIF_F_RXHASH)
+ skb->rxhash = (__force u32)pkt->rsshdr.hash_val;
if (unlikely(pkt->vlan_ex)) {
struct port_info *pi = netdev_priv(rxq->rspq.netdev);
@@ -1565,7 +1567,7 @@ int t4_ethrx_handler(struct sge_rspq *q, const __be64 *rsp,
if (unlikely(*(u8 *)rsp == CPL_TRACE_PKT))
return handle_trace_pkt(q->adap, si);
- pkt = (void *)&rsp[1];
+ pkt = (const struct cpl_rx_pkt *)rsp;
csum_ok = pkt->csum_calc && !pkt->err_vec;
if ((pkt->l2info & htonl(RXF_TCP)) &&
(q->netdev->features & NETIF_F_GRO) && csum_ok && !pkt->ip_frag) {
@@ -1583,6 +1585,9 @@ int t4_ethrx_handler(struct sge_rspq *q, const __be64 *rsp,
__skb_pull(skb, RX_PKT_PAD); /* remove ethernet header padding */
skb->protocol = eth_type_trans(skb, q->netdev);
skb_record_rx_queue(skb, q->idx);
+ if (skb->dev->features & NETIF_F_RXHASH)
+ skb->rxhash = (__force u32)pkt->rsshdr.hash_val;
+
pi = netdev_priv(skb->dev);
rxq->stats.pkts++;
diff --git a/drivers/net/cxgb4/t4_msg.h b/drivers/net/cxgb4/t4_msg.h
index fdb1174..7a981b8 100644
--- a/drivers/net/cxgb4/t4_msg.h
+++ b/drivers/net/cxgb4/t4_msg.h
@@ -503,6 +503,7 @@ struct cpl_rx_data_ack {
};
struct cpl_rx_pkt {
+ struct rss_header rsshdr;
u8 opcode;
#if defined(__LITTLE_ENDIAN_BITFIELD)
u8 iff:4;
--
1.5.4
^ permalink raw reply related
* Re: [PATCH net-next] bridge: multicast router list manipulation
From: Michał Mirosław @ 2010-04-27 23:11 UTC (permalink / raw)
To: Stephen Hemminger; +Cc: Herbert Xu, David S. Miller, netdev
In-Reply-To: <20100427101311.2f445227@nehalam>
2010/4/27 Stephen Hemminger <shemminger@vyatta.com>:
> I prefer that the hlist be only accessed through the hlist macro
> objects. Explicit twiddling of links (especially with RCU) exposes
> the code to future bugs.
[...]
> - port->rlist.pprev = h;
> - port->rlist.next = p;
> - rcu_assign_pointer(*h, &port->rlist);
> - if (p)
> - p->pprev = &port->rlist.next;
> + hlist_for_each_entry(p, n, &br->router_list, rlist) {
Shouldn't this be hlist_for_each_entry_rcu?
> + if ((unsigned long) port >= (unsigned long) p) {
> + hlist_add_before_rcu(n, &port->rlist);
> + return;
> + }
> + last = n;
> + }
> +
> + if (last)
> + hlist_add_after_rcu(last, &port->rlist);
> + else
> + hlist_add_head_rcu(&port->rlist, &br->router_list);
> }
>
[...]
Best Regards,
Michał Mirosław
^ permalink raw reply
* [net-next-2.6 PATCH 7/7] ixgbevf: use DMA API instead of PCI DMA functions
From: Jeff Kirsher @ 2010-04-27 23:10 UTC (permalink / raw)
To: davem; +Cc: netdev, gospo, Nicholas Nunley, Jeff Kirsher
In-Reply-To: <20100427230836.26173.59671.stgit@localhost.localdomain>
From: Nick Nunley <nicholasx.d.nunley@intel.com>
Signed-off-by: Nicholas Nunley <nicholasx.d.nunley@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/ixgbevf/ixgbevf_main.c | 68 +++++++++++++++++++-----------------
1 files changed, 35 insertions(+), 33 deletions(-)
diff --git a/drivers/net/ixgbevf/ixgbevf_main.c b/drivers/net/ixgbevf/ixgbevf_main.c
index f484161..0870740 100644
--- a/drivers/net/ixgbevf/ixgbevf_main.c
+++ b/drivers/net/ixgbevf/ixgbevf_main.c
@@ -139,15 +139,15 @@ static void ixgbevf_unmap_and_free_tx_resource(struct ixgbevf_adapter *adapter,
{
if (tx_buffer_info->dma) {
if (tx_buffer_info->mapped_as_page)
- pci_unmap_page(adapter->pdev,
+ dma_unmap_page(&adapter->pdev->dev,
tx_buffer_info->dma,
tx_buffer_info->length,
- PCI_DMA_TODEVICE);
+ DMA_TO_DEVICE);
else
- pci_unmap_single(adapter->pdev,
+ dma_unmap_single(&adapter->pdev->dev,
tx_buffer_info->dma,
tx_buffer_info->length,
- PCI_DMA_TODEVICE);
+ DMA_TO_DEVICE);
tx_buffer_info->dma = 0;
}
if (tx_buffer_info->skb) {
@@ -416,10 +416,10 @@ static void ixgbevf_alloc_rx_buffers(struct ixgbevf_adapter *adapter,
bi->page_offset ^= (PAGE_SIZE / 2);
}
- bi->page_dma = pci_map_page(pdev, bi->page,
+ bi->page_dma = dma_map_page(&pdev->dev, bi->page,
bi->page_offset,
(PAGE_SIZE / 2),
- PCI_DMA_FROMDEVICE);
+ DMA_FROM_DEVICE);
}
skb = bi->skb;
@@ -442,9 +442,9 @@ static void ixgbevf_alloc_rx_buffers(struct ixgbevf_adapter *adapter,
bi->skb = skb;
}
if (!bi->dma) {
- bi->dma = pci_map_single(pdev, skb->data,
+ bi->dma = dma_map_single(&pdev->dev, skb->data,
rx_ring->rx_buf_len,
- PCI_DMA_FROMDEVICE);
+ DMA_FROM_DEVICE);
}
/* Refresh the desc even if buffer_addrs didn't change because
* each write-back erases this info. */
@@ -536,16 +536,16 @@ static bool ixgbevf_clean_rx_irq(struct ixgbevf_q_vector *q_vector,
rx_buffer_info->skb = NULL;
if (rx_buffer_info->dma) {
- pci_unmap_single(pdev, rx_buffer_info->dma,
+ dma_unmap_single(&pdev->dev, rx_buffer_info->dma,
rx_ring->rx_buf_len,
- PCI_DMA_FROMDEVICE);
+ DMA_FROM_DEVICE);
rx_buffer_info->dma = 0;
skb_put(skb, len);
}
if (upper_len) {
- pci_unmap_page(pdev, rx_buffer_info->page_dma,
- PAGE_SIZE / 2, PCI_DMA_FROMDEVICE);
+ dma_unmap_page(&pdev->dev, rx_buffer_info->page_dma,
+ PAGE_SIZE / 2, DMA_FROM_DEVICE);
rx_buffer_info->page_dma = 0;
skb_fill_page_desc(skb, skb_shinfo(skb)->nr_frags,
rx_buffer_info->page,
@@ -1721,9 +1721,9 @@ static void ixgbevf_clean_rx_ring(struct ixgbevf_adapter *adapter,
rx_buffer_info = &rx_ring->rx_buffer_info[i];
if (rx_buffer_info->dma) {
- pci_unmap_single(pdev, rx_buffer_info->dma,
+ dma_unmap_single(&pdev->dev, rx_buffer_info->dma,
rx_ring->rx_buf_len,
- PCI_DMA_FROMDEVICE);
+ DMA_FROM_DEVICE);
rx_buffer_info->dma = 0;
}
if (rx_buffer_info->skb) {
@@ -1737,8 +1737,8 @@ static void ixgbevf_clean_rx_ring(struct ixgbevf_adapter *adapter,
}
if (!rx_buffer_info->page)
continue;
- pci_unmap_page(pdev, rx_buffer_info->page_dma, PAGE_SIZE / 2,
- PCI_DMA_FROMDEVICE);
+ dma_unmap_page(&pdev->dev, rx_buffer_info->page_dma,
+ PAGE_SIZE / 2, DMA_FROM_DEVICE);
rx_buffer_info->page_dma = 0;
put_page(rx_buffer_info->page);
rx_buffer_info->page = NULL;
@@ -2445,7 +2445,8 @@ void ixgbevf_free_tx_resources(struct ixgbevf_adapter *adapter,
vfree(tx_ring->tx_buffer_info);
tx_ring->tx_buffer_info = NULL;
- pci_free_consistent(pdev, tx_ring->size, tx_ring->desc, tx_ring->dma);
+ dma_free_coherent(&pdev->dev, tx_ring->size, tx_ring->desc,
+ tx_ring->dma);
tx_ring->desc = NULL;
}
@@ -2490,8 +2491,8 @@ int ixgbevf_setup_tx_resources(struct ixgbevf_adapter *adapter,
tx_ring->size = tx_ring->count * sizeof(union ixgbe_adv_tx_desc);
tx_ring->size = ALIGN(tx_ring->size, 4096);
- tx_ring->desc = pci_alloc_consistent(pdev, tx_ring->size,
- &tx_ring->dma);
+ tx_ring->desc = dma_alloc_coherent(&pdev->dev, tx_ring->size,
+ &tx_ring->dma, GFP_KERNEL);
if (!tx_ring->desc)
goto err;
@@ -2561,8 +2562,8 @@ int ixgbevf_setup_rx_resources(struct ixgbevf_adapter *adapter,
rx_ring->size = rx_ring->count * sizeof(union ixgbe_adv_rx_desc);
rx_ring->size = ALIGN(rx_ring->size, 4096);
- rx_ring->desc = pci_alloc_consistent(pdev, rx_ring->size,
- &rx_ring->dma);
+ rx_ring->desc = dma_alloc_coherent(&pdev->dev, rx_ring->size,
+ &rx_ring->dma, GFP_KERNEL);
if (!rx_ring->desc) {
hw_dbg(&adapter->hw,
@@ -2623,7 +2624,8 @@ void ixgbevf_free_rx_resources(struct ixgbevf_adapter *adapter,
vfree(rx_ring->rx_buffer_info);
rx_ring->rx_buffer_info = NULL;
- pci_free_consistent(pdev, rx_ring->size, rx_ring->desc, rx_ring->dma);
+ dma_free_coherent(&pdev->dev, rx_ring->size, rx_ring->desc,
+ rx_ring->dma);
rx_ring->desc = NULL;
}
@@ -2935,10 +2937,10 @@ static int ixgbevf_tx_map(struct ixgbevf_adapter *adapter,
tx_buffer_info->length = size;
tx_buffer_info->mapped_as_page = false;
- tx_buffer_info->dma = pci_map_single(adapter->pdev,
+ tx_buffer_info->dma = dma_map_single(&adapter->pdev->dev,
skb->data + offset,
- size, PCI_DMA_TODEVICE);
- if (pci_dma_mapping_error(pdev, tx_buffer_info->dma))
+ size, DMA_TO_DEVICE);
+ if (dma_mapping_error(&pdev->dev, tx_buffer_info->dma))
goto dma_error;
tx_buffer_info->time_stamp = jiffies;
tx_buffer_info->next_to_watch = i;
@@ -2964,13 +2966,13 @@ static int ixgbevf_tx_map(struct ixgbevf_adapter *adapter,
size = min(len, (unsigned int)IXGBE_MAX_DATA_PER_TXD);
tx_buffer_info->length = size;
- tx_buffer_info->dma = pci_map_page(adapter->pdev,
+ tx_buffer_info->dma = dma_map_page(&adapter->pdev->dev,
frag->page,
offset,
size,
- PCI_DMA_TODEVICE);
+ DMA_TO_DEVICE);
tx_buffer_info->mapped_as_page = true;
- if (pci_dma_mapping_error(pdev, tx_buffer_info->dma))
+ if (dma_mapping_error(&pdev->dev, tx_buffer_info->dma))
goto dma_error;
tx_buffer_info->time_stamp = jiffies;
tx_buffer_info->next_to_watch = i;
@@ -3311,14 +3313,14 @@ static int __devinit ixgbevf_probe(struct pci_dev *pdev,
if (err)
return err;
- if (!pci_set_dma_mask(pdev, DMA_BIT_MASK(64)) &&
- !pci_set_consistent_dma_mask(pdev, DMA_BIT_MASK(64))) {
+ if (!dma_set_mask(&pdev->dev, DMA_BIT_MASK(64)) &&
+ !dma_set_coherent_mask(&pdev->dev, DMA_BIT_MASK(64))) {
pci_using_dac = 1;
} else {
- err = pci_set_dma_mask(pdev, DMA_BIT_MASK(32));
+ err = dma_set_mask(&pdev->dev, DMA_BIT_MASK(32));
if (err) {
- err = pci_set_consistent_dma_mask(pdev,
- DMA_BIT_MASK(32));
+ err = dma_set_coherent_mask(&pdev->dev,
+ DMA_BIT_MASK(32));
if (err) {
dev_err(&pdev->dev, "No usable DMA "
"configuration, aborting\n");
^ permalink raw reply related
* [net-next-2.6 PATCH 6/7] ixgbe: use DMA API instead of PCI DMA functions
From: Jeff Kirsher @ 2010-04-27 23:10 UTC (permalink / raw)
To: davem; +Cc: netdev, gospo, Nicholas Nunley, Jeff Kirsher
In-Reply-To: <20100427230836.26173.59671.stgit@localhost.localdomain>
From: Nick Nunley <nicholasx.d.nunley@intel.com>
Signed-off-by: Nicholas Nunley <nicholasx.d.nunley@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/ixgbe/ixgbe_ethtool.c | 42 +++++++++++---------
drivers/net/ixgbe/ixgbe_main.c | 77 ++++++++++++++++++++-----------------
2 files changed, 64 insertions(+), 55 deletions(-)
diff --git a/drivers/net/ixgbe/ixgbe_ethtool.c b/drivers/net/ixgbe/ixgbe_ethtool.c
index dfbfe35..dc7fd5b 100644
--- a/drivers/net/ixgbe/ixgbe_ethtool.c
+++ b/drivers/net/ixgbe/ixgbe_ethtool.c
@@ -1458,8 +1458,8 @@ static void ixgbe_free_desc_rings(struct ixgbe_adapter *adapter)
struct ixgbe_tx_buffer *buf =
&(tx_ring->tx_buffer_info[i]);
if (buf->dma)
- pci_unmap_single(pdev, buf->dma, buf->length,
- PCI_DMA_TODEVICE);
+ dma_unmap_single(&pdev->dev, buf->dma,
+ buf->length, DMA_TO_DEVICE);
if (buf->skb)
dev_kfree_skb(buf->skb);
}
@@ -1470,22 +1470,22 @@ static void ixgbe_free_desc_rings(struct ixgbe_adapter *adapter)
struct ixgbe_rx_buffer *buf =
&(rx_ring->rx_buffer_info[i]);
if (buf->dma)
- pci_unmap_single(pdev, buf->dma,
+ dma_unmap_single(&pdev->dev, buf->dma,
IXGBE_RXBUFFER_2048,
- PCI_DMA_FROMDEVICE);
+ DMA_FROM_DEVICE);
if (buf->skb)
dev_kfree_skb(buf->skb);
}
}
if (tx_ring->desc) {
- pci_free_consistent(pdev, tx_ring->size, tx_ring->desc,
- tx_ring->dma);
+ dma_free_coherent(&pdev->dev, tx_ring->size, tx_ring->desc,
+ tx_ring->dma);
tx_ring->desc = NULL;
}
if (rx_ring->desc) {
- pci_free_consistent(pdev, rx_ring->size, rx_ring->desc,
- rx_ring->dma);
+ dma_free_coherent(&pdev->dev, rx_ring->size, rx_ring->desc,
+ rx_ring->dma);
rx_ring->desc = NULL;
}
@@ -1520,8 +1520,9 @@ static int ixgbe_setup_desc_rings(struct ixgbe_adapter *adapter)
tx_ring->size = tx_ring->count * sizeof(union ixgbe_adv_tx_desc);
tx_ring->size = ALIGN(tx_ring->size, 4096);
- if (!(tx_ring->desc = pci_alloc_consistent(pdev, tx_ring->size,
- &tx_ring->dma))) {
+ tx_ring->desc = dma_alloc_coherent(&pdev->dev, tx_ring->size,
+ &tx_ring->dma, GFP_KERNEL);
+ if (!(tx_ring->desc)) {
ret_val = 2;
goto err_nomem;
}
@@ -1563,8 +1564,8 @@ static int ixgbe_setup_desc_rings(struct ixgbe_adapter *adapter)
tx_ring->tx_buffer_info[i].skb = skb;
tx_ring->tx_buffer_info[i].length = skb->len;
tx_ring->tx_buffer_info[i].dma =
- pci_map_single(pdev, skb->data, skb->len,
- PCI_DMA_TODEVICE);
+ dma_map_single(&pdev->dev, skb->data, skb->len,
+ DMA_TO_DEVICE);
desc->read.buffer_addr =
cpu_to_le64(tx_ring->tx_buffer_info[i].dma);
desc->read.cmd_type_len = cpu_to_le32(skb->len);
@@ -1593,8 +1594,9 @@ static int ixgbe_setup_desc_rings(struct ixgbe_adapter *adapter)
rx_ring->size = rx_ring->count * sizeof(union ixgbe_adv_rx_desc);
rx_ring->size = ALIGN(rx_ring->size, 4096);
- if (!(rx_ring->desc = pci_alloc_consistent(pdev, rx_ring->size,
- &rx_ring->dma))) {
+ rx_ring->desc = dma_alloc_coherent(&pdev->dev, rx_ring->size,
+ &rx_ring->dma, GFP_KERNEL);
+ if (!(rx_ring->desc)) {
ret_val = 5;
goto err_nomem;
}
@@ -1661,8 +1663,8 @@ static int ixgbe_setup_desc_rings(struct ixgbe_adapter *adapter)
skb_reserve(skb, NET_IP_ALIGN);
rx_ring->rx_buffer_info[i].skb = skb;
rx_ring->rx_buffer_info[i].dma =
- pci_map_single(pdev, skb->data, IXGBE_RXBUFFER_2048,
- PCI_DMA_FROMDEVICE);
+ dma_map_single(&pdev->dev, skb->data,
+ IXGBE_RXBUFFER_2048, DMA_FROM_DEVICE);
rx_desc->read.pkt_addr =
cpu_to_le64(rx_ring->rx_buffer_info[i].dma);
memset(skb->data, 0x00, skb->len);
@@ -1775,10 +1777,10 @@ static int ixgbe_run_loopback_test(struct ixgbe_adapter *adapter)
ixgbe_create_lbtest_frame(
tx_ring->tx_buffer_info[k].skb,
1024);
- pci_dma_sync_single_for_device(pdev,
+ dma_sync_single_for_device(&pdev->dev,
tx_ring->tx_buffer_info[k].dma,
tx_ring->tx_buffer_info[k].length,
- PCI_DMA_TODEVICE);
+ DMA_TO_DEVICE);
if (unlikely(++k == tx_ring->count))
k = 0;
}
@@ -1789,10 +1791,10 @@ static int ixgbe_run_loopback_test(struct ixgbe_adapter *adapter)
good_cnt = 0;
do {
/* receive the sent packets */
- pci_dma_sync_single_for_cpu(pdev,
+ dma_sync_single_for_cpu(&pdev->dev,
rx_ring->rx_buffer_info[l].dma,
IXGBE_RXBUFFER_2048,
- PCI_DMA_FROMDEVICE);
+ DMA_FROM_DEVICE);
ret_val = ixgbe_check_lbtest_frame(
rx_ring->rx_buffer_info[l].skb, 1024);
if (!ret_val)
diff --git a/drivers/net/ixgbe/ixgbe_main.c b/drivers/net/ixgbe/ixgbe_main.c
index 5258b3d..7724fb2 100644
--- a/drivers/net/ixgbe/ixgbe_main.c
+++ b/drivers/net/ixgbe/ixgbe_main.c
@@ -266,15 +266,15 @@ static void ixgbe_unmap_and_free_tx_resource(struct ixgbe_adapter *adapter,
{
if (tx_buffer_info->dma) {
if (tx_buffer_info->mapped_as_page)
- pci_unmap_page(adapter->pdev,
+ dma_unmap_page(&adapter->pdev->dev,
tx_buffer_info->dma,
tx_buffer_info->length,
- PCI_DMA_TODEVICE);
+ DMA_TO_DEVICE);
else
- pci_unmap_single(adapter->pdev,
+ dma_unmap_single(&adapter->pdev->dev,
tx_buffer_info->dma,
tx_buffer_info->length,
- PCI_DMA_TODEVICE);
+ DMA_TO_DEVICE);
tx_buffer_info->dma = 0;
}
if (tx_buffer_info->skb) {
@@ -721,10 +721,10 @@ static void ixgbe_alloc_rx_buffers(struct ixgbe_adapter *adapter,
bi->page_offset ^= (PAGE_SIZE / 2);
}
- bi->page_dma = pci_map_page(pdev, bi->page,
+ bi->page_dma = dma_map_page(&pdev->dev, bi->page,
bi->page_offset,
(PAGE_SIZE / 2),
- PCI_DMA_FROMDEVICE);
+ DMA_FROM_DEVICE);
}
if (!bi->skb) {
@@ -743,9 +743,9 @@ static void ixgbe_alloc_rx_buffers(struct ixgbe_adapter *adapter,
- skb->data));
bi->skb = skb;
- bi->dma = pci_map_single(pdev, skb->data,
+ bi->dma = dma_map_single(&pdev->dev, skb->data,
rx_ring->rx_buf_len,
- PCI_DMA_FROMDEVICE);
+ DMA_FROM_DEVICE);
}
/* Refresh the desc even if buffer_addrs didn't change because
* each write-back erases this info. */
@@ -886,16 +886,17 @@ static bool ixgbe_clean_rx_irq(struct ixgbe_q_vector *q_vector,
*/
IXGBE_RSC_CB(skb)->dma = rx_buffer_info->dma;
else
- pci_unmap_single(pdev, rx_buffer_info->dma,
+ dma_unmap_single(&pdev->dev,
+ rx_buffer_info->dma,
rx_ring->rx_buf_len,
- PCI_DMA_FROMDEVICE);
+ DMA_FROM_DEVICE);
rx_buffer_info->dma = 0;
skb_put(skb, len);
}
if (upper_len) {
- pci_unmap_page(pdev, rx_buffer_info->page_dma,
- PAGE_SIZE / 2, PCI_DMA_FROMDEVICE);
+ dma_unmap_page(&pdev->dev, rx_buffer_info->page_dma,
+ PAGE_SIZE / 2, DMA_FROM_DEVICE);
rx_buffer_info->page_dma = 0;
skb_fill_page_desc(skb, skb_shinfo(skb)->nr_frags,
rx_buffer_info->page,
@@ -937,9 +938,10 @@ static bool ixgbe_clean_rx_irq(struct ixgbe_q_vector *q_vector,
skb = ixgbe_transform_rsc_queue(skb, &(rx_ring->rsc_count));
if (adapter->flags2 & IXGBE_FLAG2_RSC_ENABLED) {
if (IXGBE_RSC_CB(skb)->dma) {
- pci_unmap_single(pdev, IXGBE_RSC_CB(skb)->dma,
+ dma_unmap_single(&pdev->dev,
+ IXGBE_RSC_CB(skb)->dma,
rx_ring->rx_buf_len,
- PCI_DMA_FROMDEVICE);
+ DMA_FROM_DEVICE);
IXGBE_RSC_CB(skb)->dma = 0;
}
if (rx_ring->flags & IXGBE_RING_RX_PS_ENABLED)
@@ -3154,9 +3156,9 @@ static void ixgbe_clean_rx_ring(struct ixgbe_adapter *adapter,
rx_buffer_info = &rx_ring->rx_buffer_info[i];
if (rx_buffer_info->dma) {
- pci_unmap_single(pdev, rx_buffer_info->dma,
+ dma_unmap_single(&pdev->dev, rx_buffer_info->dma,
rx_ring->rx_buf_len,
- PCI_DMA_FROMDEVICE);
+ DMA_FROM_DEVICE);
rx_buffer_info->dma = 0;
}
if (rx_buffer_info->skb) {
@@ -3165,9 +3167,10 @@ static void ixgbe_clean_rx_ring(struct ixgbe_adapter *adapter,
do {
struct sk_buff *this = skb;
if (IXGBE_RSC_CB(this)->dma) {
- pci_unmap_single(pdev, IXGBE_RSC_CB(this)->dma,
+ dma_unmap_single(&pdev->dev,
+ IXGBE_RSC_CB(this)->dma,
rx_ring->rx_buf_len,
- PCI_DMA_FROMDEVICE);
+ DMA_FROM_DEVICE);
IXGBE_RSC_CB(this)->dma = 0;
}
skb = skb->prev;
@@ -3177,8 +3180,8 @@ static void ixgbe_clean_rx_ring(struct ixgbe_adapter *adapter,
if (!rx_buffer_info->page)
continue;
if (rx_buffer_info->page_dma) {
- pci_unmap_page(pdev, rx_buffer_info->page_dma,
- PAGE_SIZE / 2, PCI_DMA_FROMDEVICE);
+ dma_unmap_page(&pdev->dev, rx_buffer_info->page_dma,
+ PAGE_SIZE / 2, DMA_FROM_DEVICE);
rx_buffer_info->page_dma = 0;
}
put_page(rx_buffer_info->page);
@@ -4403,8 +4406,8 @@ int ixgbe_setup_tx_resources(struct ixgbe_adapter *adapter,
tx_ring->size = tx_ring->count * sizeof(union ixgbe_adv_tx_desc);
tx_ring->size = ALIGN(tx_ring->size, 4096);
- tx_ring->desc = pci_alloc_consistent(pdev, tx_ring->size,
- &tx_ring->dma);
+ tx_ring->desc = dma_alloc_coherent(&pdev->dev, tx_ring->size,
+ &tx_ring->dma, GFP_KERNEL);
if (!tx_ring->desc)
goto err;
@@ -4474,7 +4477,8 @@ int ixgbe_setup_rx_resources(struct ixgbe_adapter *adapter,
rx_ring->size = rx_ring->count * sizeof(union ixgbe_adv_rx_desc);
rx_ring->size = ALIGN(rx_ring->size, 4096);
- rx_ring->desc = pci_alloc_consistent(pdev, rx_ring->size, &rx_ring->dma);
+ rx_ring->desc = dma_alloc_coherent(&pdev->dev, rx_ring->size,
+ &rx_ring->dma, GFP_KERNEL);
if (!rx_ring->desc) {
DPRINTK(PROBE, ERR,
@@ -4535,7 +4539,8 @@ void ixgbe_free_tx_resources(struct ixgbe_adapter *adapter,
vfree(tx_ring->tx_buffer_info);
tx_ring->tx_buffer_info = NULL;
- pci_free_consistent(pdev, tx_ring->size, tx_ring->desc, tx_ring->dma);
+ dma_free_coherent(&pdev->dev, tx_ring->size, tx_ring->desc,
+ tx_ring->dma);
tx_ring->desc = NULL;
}
@@ -4572,7 +4577,8 @@ void ixgbe_free_rx_resources(struct ixgbe_adapter *adapter,
vfree(rx_ring->rx_buffer_info);
rx_ring->rx_buffer_info = NULL;
- pci_free_consistent(pdev, rx_ring->size, rx_ring->desc, rx_ring->dma);
+ dma_free_coherent(&pdev->dev, rx_ring->size, rx_ring->desc,
+ rx_ring->dma);
rx_ring->desc = NULL;
}
@@ -5442,10 +5448,10 @@ static int ixgbe_tx_map(struct ixgbe_adapter *adapter,
tx_buffer_info->length = size;
tx_buffer_info->mapped_as_page = false;
- tx_buffer_info->dma = pci_map_single(pdev,
+ tx_buffer_info->dma = dma_map_single(&pdev->dev,
skb->data + offset,
- size, PCI_DMA_TODEVICE);
- if (pci_dma_mapping_error(pdev, tx_buffer_info->dma))
+ size, DMA_TO_DEVICE);
+ if (dma_mapping_error(&pdev->dev, tx_buffer_info->dma))
goto dma_error;
tx_buffer_info->time_stamp = jiffies;
tx_buffer_info->next_to_watch = i;
@@ -5478,12 +5484,12 @@ static int ixgbe_tx_map(struct ixgbe_adapter *adapter,
size = min(len, (uint)IXGBE_MAX_DATA_PER_TXD);
tx_buffer_info->length = size;
- tx_buffer_info->dma = pci_map_page(adapter->pdev,
+ tx_buffer_info->dma = dma_map_page(&adapter->pdev->dev,
frag->page,
offset, size,
- PCI_DMA_TODEVICE);
+ DMA_TO_DEVICE);
tx_buffer_info->mapped_as_page = true;
- if (pci_dma_mapping_error(pdev, tx_buffer_info->dma))
+ if (dma_mapping_error(&pdev->dev, tx_buffer_info->dma))
goto dma_error;
tx_buffer_info->time_stamp = jiffies;
tx_buffer_info->next_to_watch = i;
@@ -6061,13 +6067,14 @@ static int __devinit ixgbe_probe(struct pci_dev *pdev,
if (err)
return err;
- if (!pci_set_dma_mask(pdev, DMA_BIT_MASK(64)) &&
- !pci_set_consistent_dma_mask(pdev, DMA_BIT_MASK(64))) {
+ if (!dma_set_mask(&pdev->dev, DMA_BIT_MASK(64)) &&
+ !dma_set_coherent_mask(&pdev->dev, DMA_BIT_MASK(64))) {
pci_using_dac = 1;
} else {
- err = pci_set_dma_mask(pdev, DMA_BIT_MASK(32));
+ err = dma_set_mask(&pdev->dev, DMA_BIT_MASK(32));
if (err) {
- err = pci_set_consistent_dma_mask(pdev, DMA_BIT_MASK(32));
+ err = dma_set_coherent_mask(&pdev->dev,
+ DMA_BIT_MASK(32));
if (err) {
dev_err(&pdev->dev, "No usable DMA "
"configuration, aborting\n");
^ permalink raw reply related
* [net-next-2.6 PATCH 5/7] ixgb: use DMA API instead of PCI DMA functions
From: Jeff Kirsher @ 2010-04-27 23:10 UTC (permalink / raw)
To: davem; +Cc: netdev, gospo, Nicholas Nunley, Jeff Kirsher
In-Reply-To: <20100427230836.26173.59671.stgit@localhost.localdomain>
From: Nick Nunley <nicholasx.d.nunley@intel.com>
Signed-off-by: Nicholas Nunley <nicholasx.d.nunley@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/ixgb/ixgb_main.c | 70 +++++++++++++++++++++++-------------------
1 files changed, 39 insertions(+), 31 deletions(-)
diff --git a/drivers/net/ixgb/ixgb_main.c b/drivers/net/ixgb/ixgb_main.c
index 3cf7951..d58ca6b 100644
--- a/drivers/net/ixgb/ixgb_main.c
+++ b/drivers/net/ixgb/ixgb_main.c
@@ -368,16 +368,22 @@ ixgb_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
if (err)
return err;
- if (!(err = pci_set_dma_mask(pdev, DMA_BIT_MASK(64))) &&
- !(err = pci_set_consistent_dma_mask(pdev, DMA_BIT_MASK(64)))) {
- pci_using_dac = 1;
+ pci_using_dac = 0;
+ err = dma_set_mask(&pdev->dev, DMA_BIT_MASK(64));
+ if (!err) {
+ err = dma_set_coherent_mask(&pdev->dev, DMA_BIT_MASK(64));
+ if (!err)
+ pci_using_dac = 1;
} else {
- if ((err = pci_set_dma_mask(pdev, DMA_BIT_MASK(32))) ||
- (err = pci_set_consistent_dma_mask(pdev, DMA_BIT_MASK(32)))) {
- pr_err("No usable DMA configuration, aborting\n");
- goto err_dma_mask;
+ err = dma_set_mask(&pdev->dev, DMA_BIT_MASK(32));
+ if (err) {
+ err = dma_set_coherent_mask(&pdev->dev,
+ DMA_BIT_MASK(32));
+ if (err) {
+ pr_err("No usable DMA configuration, aborting\n");
+ goto err_dma_mask;
+ }
}
- pci_using_dac = 0;
}
err = pci_request_regions(pdev, ixgb_driver_name);
@@ -673,7 +679,8 @@ ixgb_setup_tx_resources(struct ixgb_adapter *adapter)
txdr->size = txdr->count * sizeof(struct ixgb_tx_desc);
txdr->size = ALIGN(txdr->size, 4096);
- txdr->desc = pci_alloc_consistent(pdev, txdr->size, &txdr->dma);
+ txdr->desc = dma_alloc_coherent(&pdev->dev, txdr->size, &txdr->dma,
+ GFP_KERNEL);
if (!txdr->desc) {
vfree(txdr->buffer_info);
netif_err(adapter, probe, adapter->netdev,
@@ -762,7 +769,8 @@ ixgb_setup_rx_resources(struct ixgb_adapter *adapter)
rxdr->size = rxdr->count * sizeof(struct ixgb_rx_desc);
rxdr->size = ALIGN(rxdr->size, 4096);
- rxdr->desc = pci_alloc_consistent(pdev, rxdr->size, &rxdr->dma);
+ rxdr->desc = dma_alloc_coherent(&pdev->dev, rxdr->size, &rxdr->dma,
+ GFP_KERNEL);
if (!rxdr->desc) {
vfree(rxdr->buffer_info);
@@ -883,8 +891,8 @@ ixgb_free_tx_resources(struct ixgb_adapter *adapter)
vfree(adapter->tx_ring.buffer_info);
adapter->tx_ring.buffer_info = NULL;
- pci_free_consistent(pdev, adapter->tx_ring.size,
- adapter->tx_ring.desc, adapter->tx_ring.dma);
+ dma_free_coherent(&pdev->dev, adapter->tx_ring.size,
+ adapter->tx_ring.desc, adapter->tx_ring.dma);
adapter->tx_ring.desc = NULL;
}
@@ -895,12 +903,11 @@ ixgb_unmap_and_free_tx_resource(struct ixgb_adapter *adapter,
{
if (buffer_info->dma) {
if (buffer_info->mapped_as_page)
- pci_unmap_page(adapter->pdev, buffer_info->dma,
- buffer_info->length, PCI_DMA_TODEVICE);
+ dma_unmap_page(&adapter->pdev->dev, buffer_info->dma,
+ buffer_info->length, DMA_TO_DEVICE);
else
- pci_unmap_single(adapter->pdev, buffer_info->dma,
- buffer_info->length,
- PCI_DMA_TODEVICE);
+ dma_unmap_single(&adapter->pdev->dev, buffer_info->dma,
+ buffer_info->length, DMA_TO_DEVICE);
buffer_info->dma = 0;
}
@@ -966,7 +973,8 @@ ixgb_free_rx_resources(struct ixgb_adapter *adapter)
vfree(rx_ring->buffer_info);
rx_ring->buffer_info = NULL;
- pci_free_consistent(pdev, rx_ring->size, rx_ring->desc, rx_ring->dma);
+ dma_free_coherent(&pdev->dev, rx_ring->size, rx_ring->desc,
+ rx_ring->dma);
rx_ring->desc = NULL;
}
@@ -990,10 +998,10 @@ ixgb_clean_rx_ring(struct ixgb_adapter *adapter)
for (i = 0; i < rx_ring->count; i++) {
buffer_info = &rx_ring->buffer_info[i];
if (buffer_info->dma) {
- pci_unmap_single(pdev,
+ dma_unmap_single(&pdev->dev,
buffer_info->dma,
buffer_info->length,
- PCI_DMA_FROMDEVICE);
+ DMA_FROM_DEVICE);
buffer_info->dma = 0;
buffer_info->length = 0;
}
@@ -1300,9 +1308,10 @@ ixgb_tx_map(struct ixgb_adapter *adapter, struct sk_buff *skb,
WARN_ON(buffer_info->dma != 0);
buffer_info->time_stamp = jiffies;
buffer_info->mapped_as_page = false;
- buffer_info->dma = pci_map_single(pdev, skb->data + offset,
- size, PCI_DMA_TODEVICE);
- if (pci_dma_mapping_error(pdev, buffer_info->dma))
+ buffer_info->dma = dma_map_single(&pdev->dev,
+ skb->data + offset,
+ size, DMA_TO_DEVICE);
+ if (dma_mapping_error(&pdev->dev, buffer_info->dma))
goto dma_error;
buffer_info->next_to_watch = 0;
@@ -1341,10 +1350,9 @@ ixgb_tx_map(struct ixgb_adapter *adapter, struct sk_buff *skb,
buffer_info->time_stamp = jiffies;
buffer_info->mapped_as_page = true;
buffer_info->dma =
- pci_map_page(pdev, frag->page,
- offset, size,
- PCI_DMA_TODEVICE);
- if (pci_dma_mapping_error(pdev, buffer_info->dma))
+ dma_map_page(&pdev->dev, frag->page,
+ offset, size, DMA_TO_DEVICE);
+ if (dma_mapping_error(&pdev->dev, buffer_info->dma))
goto dma_error;
buffer_info->next_to_watch = 0;
@@ -1962,10 +1970,10 @@ ixgb_clean_rx_irq(struct ixgb_adapter *adapter, int *work_done, int work_to_do)
cleaned = true;
cleaned_count++;
- pci_unmap_single(pdev,
+ dma_unmap_single(&pdev->dev,
buffer_info->dma,
buffer_info->length,
- PCI_DMA_FROMDEVICE);
+ DMA_FROM_DEVICE);
buffer_info->dma = 0;
length = le16_to_cpu(rx_desc->length);
@@ -2088,10 +2096,10 @@ ixgb_alloc_rx_buffers(struct ixgb_adapter *adapter, int cleaned_count)
buffer_info->skb = skb;
buffer_info->length = adapter->rx_buffer_len;
map_skb:
- buffer_info->dma = pci_map_single(pdev,
+ buffer_info->dma = dma_map_single(&pdev->dev,
skb->data,
adapter->rx_buffer_len,
- PCI_DMA_FROMDEVICE);
+ DMA_FROM_DEVICE);
rx_desc = IXGB_RX_DESC(*rx_ring, i);
rx_desc->buff_addr = cpu_to_le64(buffer_info->dma);
^ permalink raw reply related
* [net-next-2.6 PATCH 4/7] igbvf: use DMA API instead of PCI DMA functions
From: Jeff Kirsher @ 2010-04-27 23:09 UTC (permalink / raw)
To: davem; +Cc: netdev, gospo, Nicholas Nunley, Jeff Kirsher
In-Reply-To: <20100427230836.26173.59671.stgit@localhost.localdomain>
From: Nick Nunley <nicholasx.d.nunley@intel.com>
Signed-off-by: Nicholas Nunley <nicholasx.d.nunley@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/igbvf/netdev.c | 74 +++++++++++++++++++++++---------------------
1 files changed, 38 insertions(+), 36 deletions(-)
diff --git a/drivers/net/igbvf/netdev.c b/drivers/net/igbvf/netdev.c
index 9ac3f94..7012e3d 100644
--- a/drivers/net/igbvf/netdev.c
+++ b/drivers/net/igbvf/netdev.c
@@ -164,10 +164,10 @@ static void igbvf_alloc_rx_buffers(struct igbvf_ring *rx_ring,
buffer_info->page_offset ^= PAGE_SIZE / 2;
}
buffer_info->page_dma =
- pci_map_page(pdev, buffer_info->page,
+ dma_map_page(&pdev->dev, buffer_info->page,
buffer_info->page_offset,
PAGE_SIZE / 2,
- PCI_DMA_FROMDEVICE);
+ DMA_FROM_DEVICE);
}
if (!buffer_info->skb) {
@@ -178,9 +178,9 @@ static void igbvf_alloc_rx_buffers(struct igbvf_ring *rx_ring,
}
buffer_info->skb = skb;
- buffer_info->dma = pci_map_single(pdev, skb->data,
+ buffer_info->dma = dma_map_single(&pdev->dev, skb->data,
bufsz,
- PCI_DMA_FROMDEVICE);
+ DMA_FROM_DEVICE);
}
/* Refresh the desc even if buffer_addrs didn't change because
* each write-back erases this info. */
@@ -268,25 +268,25 @@ static bool igbvf_clean_rx_irq(struct igbvf_adapter *adapter,
prefetch(skb->data - NET_IP_ALIGN);
buffer_info->skb = NULL;
if (!adapter->rx_ps_hdr_size) {
- pci_unmap_single(pdev, buffer_info->dma,
+ dma_unmap_single(&pdev->dev, buffer_info->dma,
adapter->rx_buffer_len,
- PCI_DMA_FROMDEVICE);
+ DMA_FROM_DEVICE);
buffer_info->dma = 0;
skb_put(skb, length);
goto send_up;
}
if (!skb_shinfo(skb)->nr_frags) {
- pci_unmap_single(pdev, buffer_info->dma,
+ dma_unmap_single(&pdev->dev, buffer_info->dma,
adapter->rx_ps_hdr_size,
- PCI_DMA_FROMDEVICE);
+ DMA_FROM_DEVICE);
skb_put(skb, hlen);
}
if (length) {
- pci_unmap_page(pdev, buffer_info->page_dma,
+ dma_unmap_page(&pdev->dev, buffer_info->page_dma,
PAGE_SIZE / 2,
- PCI_DMA_FROMDEVICE);
+ DMA_FROM_DEVICE);
buffer_info->page_dma = 0;
skb_fill_page_desc(skb, skb_shinfo(skb)->nr_frags,
@@ -369,15 +369,15 @@ static void igbvf_put_txbuf(struct igbvf_adapter *adapter,
{
if (buffer_info->dma) {
if (buffer_info->mapped_as_page)
- pci_unmap_page(adapter->pdev,
+ dma_unmap_page(&adapter->pdev->dev,
buffer_info->dma,
buffer_info->length,
- PCI_DMA_TODEVICE);
+ DMA_TO_DEVICE);
else
- pci_unmap_single(adapter->pdev,
+ dma_unmap_single(&adapter->pdev->dev,
buffer_info->dma,
buffer_info->length,
- PCI_DMA_TODEVICE);
+ DMA_TO_DEVICE);
buffer_info->dma = 0;
}
if (buffer_info->skb) {
@@ -438,8 +438,8 @@ int igbvf_setup_tx_resources(struct igbvf_adapter *adapter,
tx_ring->size = tx_ring->count * sizeof(union e1000_adv_tx_desc);
tx_ring->size = ALIGN(tx_ring->size, 4096);
- tx_ring->desc = pci_alloc_consistent(pdev, tx_ring->size,
- &tx_ring->dma);
+ tx_ring->desc = dma_alloc_coherent(&pdev->dev, tx_ring->size,
+ &tx_ring->dma, GFP_KERNEL);
if (!tx_ring->desc)
goto err;
@@ -480,8 +480,8 @@ int igbvf_setup_rx_resources(struct igbvf_adapter *adapter,
rx_ring->size = rx_ring->count * desc_len;
rx_ring->size = ALIGN(rx_ring->size, 4096);
- rx_ring->desc = pci_alloc_consistent(pdev, rx_ring->size,
- &rx_ring->dma);
+ rx_ring->desc = dma_alloc_coherent(&pdev->dev, rx_ring->size,
+ &rx_ring->dma, GFP_KERNEL);
if (!rx_ring->desc)
goto err;
@@ -549,7 +549,8 @@ void igbvf_free_tx_resources(struct igbvf_ring *tx_ring)
vfree(tx_ring->buffer_info);
tx_ring->buffer_info = NULL;
- pci_free_consistent(pdev, tx_ring->size, tx_ring->desc, tx_ring->dma);
+ dma_free_coherent(&pdev->dev, tx_ring->size, tx_ring->desc,
+ tx_ring->dma);
tx_ring->desc = NULL;
}
@@ -574,13 +575,13 @@ static void igbvf_clean_rx_ring(struct igbvf_ring *rx_ring)
buffer_info = &rx_ring->buffer_info[i];
if (buffer_info->dma) {
if (adapter->rx_ps_hdr_size){
- pci_unmap_single(pdev, buffer_info->dma,
+ dma_unmap_single(&pdev->dev, buffer_info->dma,
adapter->rx_ps_hdr_size,
- PCI_DMA_FROMDEVICE);
+ DMA_FROM_DEVICE);
} else {
- pci_unmap_single(pdev, buffer_info->dma,
+ dma_unmap_single(&pdev->dev, buffer_info->dma,
adapter->rx_buffer_len,
- PCI_DMA_FROMDEVICE);
+ DMA_FROM_DEVICE);
}
buffer_info->dma = 0;
}
@@ -592,9 +593,10 @@ static void igbvf_clean_rx_ring(struct igbvf_ring *rx_ring)
if (buffer_info->page) {
if (buffer_info->page_dma)
- pci_unmap_page(pdev, buffer_info->page_dma,
+ dma_unmap_page(&pdev->dev,
+ buffer_info->page_dma,
PAGE_SIZE / 2,
- PCI_DMA_FROMDEVICE);
+ DMA_FROM_DEVICE);
put_page(buffer_info->page);
buffer_info->page = NULL;
buffer_info->page_dma = 0;
@@ -2104,9 +2106,9 @@ static inline int igbvf_tx_map_adv(struct igbvf_adapter *adapter,
buffer_info->time_stamp = jiffies;
buffer_info->next_to_watch = i;
buffer_info->mapped_as_page = false;
- buffer_info->dma = pci_map_single(pdev, skb->data, len,
- PCI_DMA_TODEVICE);
- if (pci_dma_mapping_error(pdev, buffer_info->dma))
+ buffer_info->dma = dma_map_single(&pdev->dev, skb->data, len,
+ DMA_TO_DEVICE);
+ if (dma_mapping_error(&pdev->dev, buffer_info->dma))
goto dma_error;
@@ -2127,12 +2129,12 @@ static inline int igbvf_tx_map_adv(struct igbvf_adapter *adapter,
buffer_info->time_stamp = jiffies;
buffer_info->next_to_watch = i;
buffer_info->mapped_as_page = true;
- buffer_info->dma = pci_map_page(pdev,
+ buffer_info->dma = dma_map_page(&pdev->dev,
frag->page,
frag->page_offset,
len,
- PCI_DMA_TODEVICE);
- if (pci_dma_mapping_error(pdev, buffer_info->dma))
+ DMA_TO_DEVICE);
+ if (dma_mapping_error(&pdev->dev, buffer_info->dma))
goto dma_error;
}
@@ -2644,16 +2646,16 @@ static int __devinit igbvf_probe(struct pci_dev *pdev,
return err;
pci_using_dac = 0;
- err = pci_set_dma_mask(pdev, DMA_BIT_MASK(64));
+ err = dma_set_mask(&pdev->dev, DMA_BIT_MASK(64));
if (!err) {
- err = pci_set_consistent_dma_mask(pdev, DMA_BIT_MASK(64));
+ err = dma_set_coherent_mask(&pdev->dev, DMA_BIT_MASK(64));
if (!err)
pci_using_dac = 1;
} else {
- err = pci_set_dma_mask(pdev, DMA_BIT_MASK(32));
+ err = dma_set_mask(&pdev->dev, DMA_BIT_MASK(32));
if (err) {
- err = pci_set_consistent_dma_mask(pdev,
- DMA_BIT_MASK(32));
+ err = dma_set_coherent_mask(&pdev->dev,
+ DMA_BIT_MASK(32));
if (err) {
dev_err(&pdev->dev, "No usable DMA "
"configuration, aborting\n");
^ permalink raw reply related
* [net-next-2.6 PATCH 3/7] igb: convert igb from using PCI DMA functions to using DMA API functions
From: Jeff Kirsher @ 2010-04-27 23:09 UTC (permalink / raw)
To: davem; +Cc: netdev, gospo, Alexander Duyck, Jeff Kirsher
In-Reply-To: <20100427230836.26173.59671.stgit@localhost.localdomain>
From: Alexander Duyck <alexander.h.duyck@intel.com>
This patch makes it so that igb now uses the DMA API functions instead of
the PCI API functions. To do this the pci_dev pointer that was in the
rings has been replaced with a device pointer, and as a result all
references to [tr]x_ring->pdev have been replaced with [tr]x_ring->dev.
This patch is based of of work originally done by Nicholas Nunley.
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/igb/igb.h | 2 -
drivers/net/igb/igb_ethtool.c | 8 ++-
drivers/net/igb/igb_main.c | 109 +++++++++++++++++++++--------------------
3 files changed, 61 insertions(+), 58 deletions(-)
diff --git a/drivers/net/igb/igb.h b/drivers/net/igb/igb.h
index 7d288cc..096a526 100644
--- a/drivers/net/igb/igb.h
+++ b/drivers/net/igb/igb.h
@@ -186,7 +186,7 @@ struct igb_q_vector {
struct igb_ring {
struct igb_q_vector *q_vector; /* backlink to q_vector */
struct net_device *netdev; /* back pointer to net_device */
- struct pci_dev *pdev; /* pci device for dma mapping */
+ struct device *dev; /* device pointer for dma mapping */
dma_addr_t dma; /* phys address of the ring */
void *desc; /* descriptor ring memory */
unsigned int size; /* length of desc. ring in bytes */
diff --git a/drivers/net/igb/igb_ethtool.c b/drivers/net/igb/igb_ethtool.c
index 1b8fd7f..f2ebf92 100644
--- a/drivers/net/igb/igb_ethtool.c
+++ b/drivers/net/igb/igb_ethtool.c
@@ -1394,7 +1394,7 @@ static int igb_setup_desc_rings(struct igb_adapter *adapter)
/* Setup Tx descriptor ring and Tx buffers */
tx_ring->count = IGB_DEFAULT_TXD;
- tx_ring->pdev = adapter->pdev;
+ tx_ring->dev = &adapter->pdev->dev;
tx_ring->netdev = adapter->netdev;
tx_ring->reg_idx = adapter->vfs_allocated_count;
@@ -1408,7 +1408,7 @@ static int igb_setup_desc_rings(struct igb_adapter *adapter)
/* Setup Rx descriptor ring and Rx buffers */
rx_ring->count = IGB_DEFAULT_RXD;
- rx_ring->pdev = adapter->pdev;
+ rx_ring->dev = &adapter->pdev->dev;
rx_ring->netdev = adapter->netdev;
rx_ring->rx_buffer_len = IGB_RXBUFFER_2048;
rx_ring->reg_idx = adapter->vfs_allocated_count;
@@ -1604,10 +1604,10 @@ static int igb_clean_test_rings(struct igb_ring *rx_ring,
buffer_info = &rx_ring->buffer_info[rx_ntc];
/* unmap rx buffer, will be remapped by alloc_rx_buffers */
- pci_unmap_single(rx_ring->pdev,
+ dma_unmap_single(rx_ring->dev,
buffer_info->dma,
rx_ring->rx_buffer_len,
- PCI_DMA_FROMDEVICE);
+ DMA_FROM_DEVICE);
buffer_info->dma = 0;
/* verify contents of skb */
diff --git a/drivers/net/igb/igb_main.c b/drivers/net/igb/igb_main.c
index 919e363..9d042fe 100644
--- a/drivers/net/igb/igb_main.c
+++ b/drivers/net/igb/igb_main.c
@@ -350,7 +350,7 @@ static int igb_alloc_queues(struct igb_adapter *adapter)
goto err;
ring->count = adapter->tx_ring_count;
ring->queue_index = i;
- ring->pdev = adapter->pdev;
+ ring->dev = &adapter->pdev->dev;
ring->netdev = adapter->netdev;
/* For 82575, context index must be unique per ring. */
if (adapter->hw.mac.type == e1000_82575)
@@ -364,7 +364,7 @@ static int igb_alloc_queues(struct igb_adapter *adapter)
goto err;
ring->count = adapter->rx_ring_count;
ring->queue_index = i;
- ring->pdev = adapter->pdev;
+ ring->dev = &adapter->pdev->dev;
ring->netdev = adapter->netdev;
ring->rx_buffer_len = MAXIMUM_ETHERNET_VLAN_SIZE;
ring->flags = IGB_RING_FLAG_RX_CSUM; /* enable rx checksum */
@@ -1398,15 +1398,15 @@ static int __devinit igb_probe(struct pci_dev *pdev,
return err;
pci_using_dac = 0;
- err = pci_set_dma_mask(pdev, DMA_BIT_MASK(64));
+ err = dma_set_mask(&pdev->dev, DMA_BIT_MASK(64));
if (!err) {
- err = pci_set_consistent_dma_mask(pdev, DMA_BIT_MASK(64));
+ err = dma_set_coherent_mask(&pdev->dev, DMA_BIT_MASK(64));
if (!err)
pci_using_dac = 1;
} else {
- err = pci_set_dma_mask(pdev, DMA_BIT_MASK(32));
+ err = dma_set_mask(&pdev->dev, DMA_BIT_MASK(32));
if (err) {
- err = pci_set_consistent_dma_mask(pdev, DMA_BIT_MASK(32));
+ err = dma_set_coherent_mask(&pdev->dev, DMA_BIT_MASK(32));
if (err) {
dev_err(&pdev->dev, "No usable DMA "
"configuration, aborting\n");
@@ -2080,7 +2080,7 @@ static int igb_close(struct net_device *netdev)
**/
int igb_setup_tx_resources(struct igb_ring *tx_ring)
{
- struct pci_dev *pdev = tx_ring->pdev;
+ struct device *dev = tx_ring->dev;
int size;
size = sizeof(struct igb_buffer) * tx_ring->count;
@@ -2093,9 +2093,10 @@ int igb_setup_tx_resources(struct igb_ring *tx_ring)
tx_ring->size = tx_ring->count * sizeof(union e1000_adv_tx_desc);
tx_ring->size = ALIGN(tx_ring->size, 4096);
- tx_ring->desc = pci_alloc_consistent(pdev,
- tx_ring->size,
- &tx_ring->dma);
+ tx_ring->desc = dma_alloc_coherent(dev,
+ tx_ring->size,
+ &tx_ring->dma,
+ GFP_KERNEL);
if (!tx_ring->desc)
goto err;
@@ -2106,7 +2107,7 @@ int igb_setup_tx_resources(struct igb_ring *tx_ring)
err:
vfree(tx_ring->buffer_info);
- dev_err(&pdev->dev,
+ dev_err(dev,
"Unable to allocate memory for the transmit descriptor ring\n");
return -ENOMEM;
}
@@ -2230,7 +2231,7 @@ static void igb_configure_tx(struct igb_adapter *adapter)
**/
int igb_setup_rx_resources(struct igb_ring *rx_ring)
{
- struct pci_dev *pdev = rx_ring->pdev;
+ struct device *dev = rx_ring->dev;
int size, desc_len;
size = sizeof(struct igb_buffer) * rx_ring->count;
@@ -2245,8 +2246,10 @@ int igb_setup_rx_resources(struct igb_ring *rx_ring)
rx_ring->size = rx_ring->count * desc_len;
rx_ring->size = ALIGN(rx_ring->size, 4096);
- rx_ring->desc = pci_alloc_consistent(pdev, rx_ring->size,
- &rx_ring->dma);
+ rx_ring->desc = dma_alloc_coherent(dev,
+ rx_ring->size,
+ &rx_ring->dma,
+ GFP_KERNEL);
if (!rx_ring->desc)
goto err;
@@ -2259,8 +2262,8 @@ int igb_setup_rx_resources(struct igb_ring *rx_ring)
err:
vfree(rx_ring->buffer_info);
rx_ring->buffer_info = NULL;
- dev_err(&pdev->dev, "Unable to allocate memory for "
- "the receive descriptor ring\n");
+ dev_err(dev, "Unable to allocate memory for the receive descriptor"
+ " ring\n");
return -ENOMEM;
}
@@ -2636,8 +2639,8 @@ void igb_free_tx_resources(struct igb_ring *tx_ring)
if (!tx_ring->desc)
return;
- pci_free_consistent(tx_ring->pdev, tx_ring->size,
- tx_ring->desc, tx_ring->dma);
+ dma_free_coherent(tx_ring->dev, tx_ring->size,
+ tx_ring->desc, tx_ring->dma);
tx_ring->desc = NULL;
}
@@ -2661,15 +2664,15 @@ void igb_unmap_and_free_tx_resource(struct igb_ring *tx_ring,
{
if (buffer_info->dma) {
if (buffer_info->mapped_as_page)
- pci_unmap_page(tx_ring->pdev,
+ dma_unmap_page(tx_ring->dev,
buffer_info->dma,
buffer_info->length,
- PCI_DMA_TODEVICE);
+ DMA_TO_DEVICE);
else
- pci_unmap_single(tx_ring->pdev,
+ dma_unmap_single(tx_ring->dev,
buffer_info->dma,
buffer_info->length,
- PCI_DMA_TODEVICE);
+ DMA_TO_DEVICE);
buffer_info->dma = 0;
}
if (buffer_info->skb) {
@@ -2740,8 +2743,8 @@ void igb_free_rx_resources(struct igb_ring *rx_ring)
if (!rx_ring->desc)
return;
- pci_free_consistent(rx_ring->pdev, rx_ring->size,
- rx_ring->desc, rx_ring->dma);
+ dma_free_coherent(rx_ring->dev, rx_ring->size,
+ rx_ring->desc, rx_ring->dma);
rx_ring->desc = NULL;
}
@@ -2777,10 +2780,10 @@ static void igb_clean_rx_ring(struct igb_ring *rx_ring)
for (i = 0; i < rx_ring->count; i++) {
buffer_info = &rx_ring->buffer_info[i];
if (buffer_info->dma) {
- pci_unmap_single(rx_ring->pdev,
+ dma_unmap_single(rx_ring->dev,
buffer_info->dma,
rx_ring->rx_buffer_len,
- PCI_DMA_FROMDEVICE);
+ DMA_FROM_DEVICE);
buffer_info->dma = 0;
}
@@ -2789,10 +2792,10 @@ static void igb_clean_rx_ring(struct igb_ring *rx_ring)
buffer_info->skb = NULL;
}
if (buffer_info->page_dma) {
- pci_unmap_page(rx_ring->pdev,
+ dma_unmap_page(rx_ring->dev,
buffer_info->page_dma,
PAGE_SIZE / 2,
- PCI_DMA_FROMDEVICE);
+ DMA_FROM_DEVICE);
buffer_info->page_dma = 0;
}
if (buffer_info->page) {
@@ -3480,7 +3483,7 @@ static inline bool igb_tx_csum_adv(struct igb_ring *tx_ring,
struct sk_buff *skb, u32 tx_flags)
{
struct e1000_adv_tx_context_desc *context_desc;
- struct pci_dev *pdev = tx_ring->pdev;
+ struct device *dev = tx_ring->dev;
struct igb_buffer *buffer_info;
u32 info = 0, tu_cmd = 0;
unsigned int i;
@@ -3531,7 +3534,7 @@ static inline bool igb_tx_csum_adv(struct igb_ring *tx_ring,
break;
default:
if (unlikely(net_ratelimit()))
- dev_warn(&pdev->dev,
+ dev_warn(dev,
"partial checksum but proto=%x!\n",
skb->protocol);
break;
@@ -3565,7 +3568,7 @@ static inline int igb_tx_map_adv(struct igb_ring *tx_ring, struct sk_buff *skb,
unsigned int first)
{
struct igb_buffer *buffer_info;
- struct pci_dev *pdev = tx_ring->pdev;
+ struct device *dev = tx_ring->dev;
unsigned int len = skb_headlen(skb);
unsigned int count = 0, i;
unsigned int f;
@@ -3578,9 +3581,9 @@ static inline int igb_tx_map_adv(struct igb_ring *tx_ring, struct sk_buff *skb,
/* set time_stamp *before* dma to help avoid a possible race */
buffer_info->time_stamp = jiffies;
buffer_info->next_to_watch = i;
- buffer_info->dma = pci_map_single(pdev, skb->data, len,
- PCI_DMA_TODEVICE);
- if (pci_dma_mapping_error(pdev, buffer_info->dma))
+ buffer_info->dma = dma_map_single(dev, skb->data, len,
+ DMA_TO_DEVICE);
+ if (dma_mapping_error(dev, buffer_info->dma))
goto dma_error;
for (f = 0; f < skb_shinfo(skb)->nr_frags; f++) {
@@ -3600,12 +3603,12 @@ static inline int igb_tx_map_adv(struct igb_ring *tx_ring, struct sk_buff *skb,
buffer_info->time_stamp = jiffies;
buffer_info->next_to_watch = i;
buffer_info->mapped_as_page = true;
- buffer_info->dma = pci_map_page(pdev,
+ buffer_info->dma = dma_map_page(dev,
frag->page,
frag->page_offset,
len,
- PCI_DMA_TODEVICE);
- if (pci_dma_mapping_error(pdev, buffer_info->dma))
+ DMA_TO_DEVICE);
+ if (dma_mapping_error(dev, buffer_info->dma))
goto dma_error;
}
@@ -3617,7 +3620,7 @@ static inline int igb_tx_map_adv(struct igb_ring *tx_ring, struct sk_buff *skb,
return ++count;
dma_error:
- dev_err(&pdev->dev, "TX DMA map failed\n");
+ dev_err(dev, "TX DMA map failed\n");
/* clear timestamp and dma mappings for failed buffer_info mapping */
buffer_info->dma = 0;
@@ -5059,7 +5062,7 @@ static bool igb_clean_tx_irq(struct igb_q_vector *q_vector)
!(rd32(E1000_STATUS) & E1000_STATUS_TXOFF)) {
/* detected Tx unit hang */
- dev_err(&tx_ring->pdev->dev,
+ dev_err(tx_ring->dev,
"Detected Tx Unit Hang\n"
" Tx Queue <%d>\n"
" TDH <%x>\n"
@@ -5138,7 +5141,7 @@ static inline void igb_rx_checksum_adv(struct igb_ring *ring,
if (status_err & (E1000_RXD_STAT_TCPCS | E1000_RXD_STAT_UDPCS))
skb->ip_summed = CHECKSUM_UNNECESSARY;
- dev_dbg(&ring->pdev->dev, "cksum success: bits %08X\n", status_err);
+ dev_dbg(ring->dev, "cksum success: bits %08X\n", status_err);
}
static void igb_rx_hwtstamp(struct igb_q_vector *q_vector, u32 staterr,
@@ -5193,7 +5196,7 @@ static bool igb_clean_rx_irq_adv(struct igb_q_vector *q_vector,
{
struct igb_ring *rx_ring = q_vector->rx_ring;
struct net_device *netdev = rx_ring->netdev;
- struct pci_dev *pdev = rx_ring->pdev;
+ struct device *dev = rx_ring->dev;
union e1000_adv_rx_desc *rx_desc , *next_rxd;
struct igb_buffer *buffer_info , *next_buffer;
struct sk_buff *skb;
@@ -5233,9 +5236,9 @@ static bool igb_clean_rx_irq_adv(struct igb_q_vector *q_vector,
cleaned_count++;
if (buffer_info->dma) {
- pci_unmap_single(pdev, buffer_info->dma,
+ dma_unmap_single(dev, buffer_info->dma,
rx_ring->rx_buffer_len,
- PCI_DMA_FROMDEVICE);
+ DMA_FROM_DEVICE);
buffer_info->dma = 0;
if (rx_ring->rx_buffer_len >= IGB_RXBUFFER_1024) {
skb_put(skb, length);
@@ -5245,8 +5248,8 @@ static bool igb_clean_rx_irq_adv(struct igb_q_vector *q_vector,
}
if (length) {
- pci_unmap_page(pdev, buffer_info->page_dma,
- PAGE_SIZE / 2, PCI_DMA_FROMDEVICE);
+ dma_unmap_page(dev, buffer_info->page_dma,
+ PAGE_SIZE / 2, DMA_FROM_DEVICE);
buffer_info->page_dma = 0;
skb_fill_page_desc(skb, skb_shinfo(skb)->nr_frags,
@@ -5354,12 +5357,12 @@ void igb_alloc_rx_buffers_adv(struct igb_ring *rx_ring, int cleaned_count)
buffer_info->page_offset ^= PAGE_SIZE / 2;
}
buffer_info->page_dma =
- pci_map_page(rx_ring->pdev, buffer_info->page,
+ dma_map_page(rx_ring->dev, buffer_info->page,
buffer_info->page_offset,
PAGE_SIZE / 2,
- PCI_DMA_FROMDEVICE);
- if (pci_dma_mapping_error(rx_ring->pdev,
- buffer_info->page_dma)) {
+ DMA_FROM_DEVICE);
+ if (dma_mapping_error(rx_ring->dev,
+ buffer_info->page_dma)) {
buffer_info->page_dma = 0;
rx_ring->rx_stats.alloc_failed++;
goto no_buffers;
@@ -5377,12 +5380,12 @@ void igb_alloc_rx_buffers_adv(struct igb_ring *rx_ring, int cleaned_count)
buffer_info->skb = skb;
}
if (!buffer_info->dma) {
- buffer_info->dma = pci_map_single(rx_ring->pdev,
+ buffer_info->dma = dma_map_single(rx_ring->dev,
skb->data,
bufsz,
- PCI_DMA_FROMDEVICE);
- if (pci_dma_mapping_error(rx_ring->pdev,
- buffer_info->dma)) {
+ DMA_FROM_DEVICE);
+ if (dma_mapping_error(rx_ring->dev,
+ buffer_info->dma)) {
buffer_info->dma = 0;
rx_ring->rx_stats.alloc_failed++;
goto no_buffers;
^ permalink raw reply related
* [net-next-2.6 PATCH 2/7] e1000e: use DMA API instead of PCI DMA functions
From: Jeff Kirsher @ 2010-04-27 23:09 UTC (permalink / raw)
To: davem; +Cc: netdev, gospo, Nicholas Nunley, Jeff Kirsher
In-Reply-To: <20100427230836.26173.59671.stgit@localhost.localdomain>
From: Nick Nunley <nicholasx.d.nunley@intel.com>
Signed-off-by: Nicholas Nunley <nicholasx.d.nunley@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/e1000e/ethtool.c | 30 +++++++------
drivers/net/e1000e/netdev.c | 99 +++++++++++++++++++++---------------------
2 files changed, 66 insertions(+), 63 deletions(-)
diff --git a/drivers/net/e1000e/ethtool.c b/drivers/net/e1000e/ethtool.c
index 983493f..7f9915c 100644
--- a/drivers/net/e1000e/ethtool.c
+++ b/drivers/net/e1000e/ethtool.c
@@ -1069,10 +1069,10 @@ static void e1000_free_desc_rings(struct e1000_adapter *adapter)
if (tx_ring->desc && tx_ring->buffer_info) {
for (i = 0; i < tx_ring->count; i++) {
if (tx_ring->buffer_info[i].dma)
- pci_unmap_single(pdev,
+ dma_unmap_single(&pdev->dev,
tx_ring->buffer_info[i].dma,
tx_ring->buffer_info[i].length,
- PCI_DMA_TODEVICE);
+ DMA_TO_DEVICE);
if (tx_ring->buffer_info[i].skb)
dev_kfree_skb(tx_ring->buffer_info[i].skb);
}
@@ -1081,9 +1081,9 @@ static void e1000_free_desc_rings(struct e1000_adapter *adapter)
if (rx_ring->desc && rx_ring->buffer_info) {
for (i = 0; i < rx_ring->count; i++) {
if (rx_ring->buffer_info[i].dma)
- pci_unmap_single(pdev,
+ dma_unmap_single(&pdev->dev,
rx_ring->buffer_info[i].dma,
- 2048, PCI_DMA_FROMDEVICE);
+ 2048, DMA_FROM_DEVICE);
if (rx_ring->buffer_info[i].skb)
dev_kfree_skb(rx_ring->buffer_info[i].skb);
}
@@ -1163,9 +1163,10 @@ static int e1000_setup_desc_rings(struct e1000_adapter *adapter)
tx_ring->buffer_info[i].skb = skb;
tx_ring->buffer_info[i].length = skb->len;
tx_ring->buffer_info[i].dma =
- pci_map_single(pdev, skb->data, skb->len,
- PCI_DMA_TODEVICE);
- if (pci_dma_mapping_error(pdev, tx_ring->buffer_info[i].dma)) {
+ dma_map_single(&pdev->dev, skb->data, skb->len,
+ DMA_TO_DEVICE);
+ if (dma_mapping_error(&pdev->dev,
+ tx_ring->buffer_info[i].dma)) {
ret_val = 4;
goto err_nomem;
}
@@ -1226,9 +1227,10 @@ static int e1000_setup_desc_rings(struct e1000_adapter *adapter)
skb_reserve(skb, NET_IP_ALIGN);
rx_ring->buffer_info[i].skb = skb;
rx_ring->buffer_info[i].dma =
- pci_map_single(pdev, skb->data, 2048,
- PCI_DMA_FROMDEVICE);
- if (pci_dma_mapping_error(pdev, rx_ring->buffer_info[i].dma)) {
+ dma_map_single(&pdev->dev, skb->data, 2048,
+ DMA_FROM_DEVICE);
+ if (dma_mapping_error(&pdev->dev,
+ rx_ring->buffer_info[i].dma)) {
ret_val = 8;
goto err_nomem;
}
@@ -1556,10 +1558,10 @@ static int e1000_run_loopback_test(struct e1000_adapter *adapter)
for (i = 0; i < 64; i++) { /* send the packets */
e1000_create_lbtest_frame(tx_ring->buffer_info[k].skb,
1024);
- pci_dma_sync_single_for_device(pdev,
+ dma_sync_single_for_device(&pdev->dev,
tx_ring->buffer_info[k].dma,
tx_ring->buffer_info[k].length,
- PCI_DMA_TODEVICE);
+ DMA_TO_DEVICE);
k++;
if (k == tx_ring->count)
k = 0;
@@ -1569,9 +1571,9 @@ static int e1000_run_loopback_test(struct e1000_adapter *adapter)
time = jiffies; /* set the start time for the receive */
good_cnt = 0;
do { /* receive the sent packets */
- pci_dma_sync_single_for_cpu(pdev,
+ dma_sync_single_for_cpu(&pdev->dev,
rx_ring->buffer_info[l].dma, 2048,
- PCI_DMA_FROMDEVICE);
+ DMA_FROM_DEVICE);
ret_val = e1000_check_lbtest_frame(
rx_ring->buffer_info[l].skb, 1024);
diff --git a/drivers/net/e1000e/netdev.c b/drivers/net/e1000e/netdev.c
index 2476f8c..3a71215 100644
--- a/drivers/net/e1000e/netdev.c
+++ b/drivers/net/e1000e/netdev.c
@@ -181,10 +181,10 @@ static void e1000_alloc_rx_buffers(struct e1000_adapter *adapter,
buffer_info->skb = skb;
map_skb:
- buffer_info->dma = pci_map_single(pdev, skb->data,
+ buffer_info->dma = dma_map_single(&pdev->dev, skb->data,
adapter->rx_buffer_len,
- PCI_DMA_FROMDEVICE);
- if (pci_dma_mapping_error(pdev, buffer_info->dma)) {
+ DMA_FROM_DEVICE);
+ if (dma_mapping_error(&pdev->dev, buffer_info->dma)) {
dev_err(&pdev->dev, "RX DMA map failed\n");
adapter->rx_dma_failed++;
break;
@@ -250,11 +250,12 @@ static void e1000_alloc_rx_buffers_ps(struct e1000_adapter *adapter,
adapter->alloc_rx_buff_failed++;
goto no_buffers;
}
- ps_page->dma = pci_map_page(pdev,
- ps_page->page,
- 0, PAGE_SIZE,
- PCI_DMA_FROMDEVICE);
- if (pci_dma_mapping_error(pdev, ps_page->dma)) {
+ ps_page->dma = dma_map_page(&pdev->dev,
+ ps_page->page,
+ 0, PAGE_SIZE,
+ DMA_FROM_DEVICE);
+ if (dma_mapping_error(&pdev->dev,
+ ps_page->dma)) {
dev_err(&adapter->pdev->dev,
"RX DMA page map failed\n");
adapter->rx_dma_failed++;
@@ -279,10 +280,10 @@ static void e1000_alloc_rx_buffers_ps(struct e1000_adapter *adapter,
}
buffer_info->skb = skb;
- buffer_info->dma = pci_map_single(pdev, skb->data,
+ buffer_info->dma = dma_map_single(&pdev->dev, skb->data,
adapter->rx_ps_bsize0,
- PCI_DMA_FROMDEVICE);
- if (pci_dma_mapping_error(pdev, buffer_info->dma)) {
+ DMA_FROM_DEVICE);
+ if (dma_mapping_error(&pdev->dev, buffer_info->dma)) {
dev_err(&pdev->dev, "RX DMA map failed\n");
adapter->rx_dma_failed++;
/* cleanup skb */
@@ -369,10 +370,10 @@ check_page:
}
if (!buffer_info->dma)
- buffer_info->dma = pci_map_page(pdev,
+ buffer_info->dma = dma_map_page(&pdev->dev,
buffer_info->page, 0,
PAGE_SIZE,
- PCI_DMA_FROMDEVICE);
+ DMA_FROM_DEVICE);
rx_desc = E1000_RX_DESC(*rx_ring, i);
rx_desc->buffer_addr = cpu_to_le64(buffer_info->dma);
@@ -446,10 +447,10 @@ static bool e1000_clean_rx_irq(struct e1000_adapter *adapter,
cleaned = 1;
cleaned_count++;
- pci_unmap_single(pdev,
+ dma_unmap_single(&pdev->dev,
buffer_info->dma,
adapter->rx_buffer_len,
- PCI_DMA_FROMDEVICE);
+ DMA_FROM_DEVICE);
buffer_info->dma = 0;
length = le16_to_cpu(rx_desc->length);
@@ -550,12 +551,11 @@ static void e1000_put_txbuf(struct e1000_adapter *adapter,
{
if (buffer_info->dma) {
if (buffer_info->mapped_as_page)
- pci_unmap_page(adapter->pdev, buffer_info->dma,
- buffer_info->length, PCI_DMA_TODEVICE);
+ dma_unmap_page(&adapter->pdev->dev, buffer_info->dma,
+ buffer_info->length, DMA_TO_DEVICE);
else
- pci_unmap_single(adapter->pdev, buffer_info->dma,
- buffer_info->length,
- PCI_DMA_TODEVICE);
+ dma_unmap_single(&adapter->pdev->dev, buffer_info->dma,
+ buffer_info->length, DMA_TO_DEVICE);
buffer_info->dma = 0;
}
if (buffer_info->skb) {
@@ -756,9 +756,9 @@ static bool e1000_clean_rx_irq_ps(struct e1000_adapter *adapter,
cleaned = 1;
cleaned_count++;
- pci_unmap_single(pdev, buffer_info->dma,
+ dma_unmap_single(&pdev->dev, buffer_info->dma,
adapter->rx_ps_bsize0,
- PCI_DMA_FROMDEVICE);
+ DMA_FROM_DEVICE);
buffer_info->dma = 0;
/* see !EOP comment in other rx routine */
@@ -814,13 +814,13 @@ static bool e1000_clean_rx_irq_ps(struct e1000_adapter *adapter,
* kmap_atomic, so we can't hold the mapping
* very long
*/
- pci_dma_sync_single_for_cpu(pdev, ps_page->dma,
- PAGE_SIZE, PCI_DMA_FROMDEVICE);
+ dma_sync_single_for_cpu(&pdev->dev, ps_page->dma,
+ PAGE_SIZE, DMA_FROM_DEVICE);
vaddr = kmap_atomic(ps_page->page, KM_SKB_DATA_SOFTIRQ);
memcpy(skb_tail_pointer(skb), vaddr, l1);
kunmap_atomic(vaddr, KM_SKB_DATA_SOFTIRQ);
- pci_dma_sync_single_for_device(pdev, ps_page->dma,
- PAGE_SIZE, PCI_DMA_FROMDEVICE);
+ dma_sync_single_for_device(&pdev->dev, ps_page->dma,
+ PAGE_SIZE, DMA_FROM_DEVICE);
/* remove the CRC */
if (!(adapter->flags2 & FLAG2_CRC_STRIPPING))
@@ -837,8 +837,8 @@ static bool e1000_clean_rx_irq_ps(struct e1000_adapter *adapter,
break;
ps_page = &buffer_info->ps_pages[j];
- pci_unmap_page(pdev, ps_page->dma, PAGE_SIZE,
- PCI_DMA_FROMDEVICE);
+ dma_unmap_page(&pdev->dev, ps_page->dma, PAGE_SIZE,
+ DMA_FROM_DEVICE);
ps_page->dma = 0;
skb_fill_page_desc(skb, j, ps_page->page, 0, length);
ps_page->page = NULL;
@@ -956,8 +956,8 @@ static bool e1000_clean_jumbo_rx_irq(struct e1000_adapter *adapter,
cleaned = true;
cleaned_count++;
- pci_unmap_page(pdev, buffer_info->dma, PAGE_SIZE,
- PCI_DMA_FROMDEVICE);
+ dma_unmap_page(&pdev->dev, buffer_info->dma, PAGE_SIZE,
+ DMA_FROM_DEVICE);
buffer_info->dma = 0;
length = le16_to_cpu(rx_desc->length);
@@ -1093,17 +1093,17 @@ static void e1000_clean_rx_ring(struct e1000_adapter *adapter)
buffer_info = &rx_ring->buffer_info[i];
if (buffer_info->dma) {
if (adapter->clean_rx == e1000_clean_rx_irq)
- pci_unmap_single(pdev, buffer_info->dma,
+ dma_unmap_single(&pdev->dev, buffer_info->dma,
adapter->rx_buffer_len,
- PCI_DMA_FROMDEVICE);
+ DMA_FROM_DEVICE);
else if (adapter->clean_rx == e1000_clean_jumbo_rx_irq)
- pci_unmap_page(pdev, buffer_info->dma,
+ dma_unmap_page(&pdev->dev, buffer_info->dma,
PAGE_SIZE,
- PCI_DMA_FROMDEVICE);
+ DMA_FROM_DEVICE);
else if (adapter->clean_rx == e1000_clean_rx_irq_ps)
- pci_unmap_single(pdev, buffer_info->dma,
+ dma_unmap_single(&pdev->dev, buffer_info->dma,
adapter->rx_ps_bsize0,
- PCI_DMA_FROMDEVICE);
+ DMA_FROM_DEVICE);
buffer_info->dma = 0;
}
@@ -1121,8 +1121,8 @@ static void e1000_clean_rx_ring(struct e1000_adapter *adapter)
ps_page = &buffer_info->ps_pages[j];
if (!ps_page->page)
break;
- pci_unmap_page(pdev, ps_page->dma, PAGE_SIZE,
- PCI_DMA_FROMDEVICE);
+ dma_unmap_page(&pdev->dev, ps_page->dma, PAGE_SIZE,
+ DMA_FROM_DEVICE);
ps_page->dma = 0;
put_page(ps_page->page);
ps_page->page = NULL;
@@ -3917,10 +3917,11 @@ static int e1000_tx_map(struct e1000_adapter *adapter,
buffer_info->length = size;
buffer_info->time_stamp = jiffies;
buffer_info->next_to_watch = i;
- buffer_info->dma = pci_map_single(pdev, skb->data + offset,
- size, PCI_DMA_TODEVICE);
+ buffer_info->dma = dma_map_single(&pdev->dev,
+ skb->data + offset,
+ size, DMA_TO_DEVICE);
buffer_info->mapped_as_page = false;
- if (pci_dma_mapping_error(pdev, buffer_info->dma))
+ if (dma_mapping_error(&pdev->dev, buffer_info->dma))
goto dma_error;
len -= size;
@@ -3952,11 +3953,11 @@ static int e1000_tx_map(struct e1000_adapter *adapter,
buffer_info->length = size;
buffer_info->time_stamp = jiffies;
buffer_info->next_to_watch = i;
- buffer_info->dma = pci_map_page(pdev, frag->page,
+ buffer_info->dma = dma_map_page(&pdev->dev, frag->page,
offset, size,
- PCI_DMA_TODEVICE);
+ DMA_TO_DEVICE);
buffer_info->mapped_as_page = true;
- if (pci_dma_mapping_error(pdev, buffer_info->dma))
+ if (dma_mapping_error(&pdev->dev, buffer_info->dma))
goto dma_error;
len -= size;
@@ -5050,16 +5051,16 @@ static int __devinit e1000_probe(struct pci_dev *pdev,
return err;
pci_using_dac = 0;
- err = pci_set_dma_mask(pdev, DMA_BIT_MASK(64));
+ err = dma_set_mask(&pdev->dev, DMA_BIT_MASK(64));
if (!err) {
- err = pci_set_consistent_dma_mask(pdev, DMA_BIT_MASK(64));
+ err = dma_set_coherent_mask(&pdev->dev, DMA_BIT_MASK(64));
if (!err)
pci_using_dac = 1;
} else {
- err = pci_set_dma_mask(pdev, DMA_BIT_MASK(32));
+ err = dma_set_mask(&pdev->dev, DMA_BIT_MASK(32));
if (err) {
- err = pci_set_consistent_dma_mask(pdev,
- DMA_BIT_MASK(32));
+ err = dma_set_coherent_mask(&pdev->dev,
+ DMA_BIT_MASK(32));
if (err) {
dev_err(&pdev->dev, "No usable DMA "
"configuration, aborting\n");
^ permalink raw reply related
* [net-next-2.6 PATCH 1/7] e1000: use DMA API instead of PCI DMA functions
From: Jeff Kirsher @ 2010-04-27 23:08 UTC (permalink / raw)
To: davem; +Cc: netdev, gospo, Nicholas Nunley, Jeff Kirsher
From: Nick Nunley <nicholasx.d.nunley@intel.com>
Signed-off-by: Nicholas Nunley <nicholasx.d.nunley@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/e1000/e1000_ethtool.c | 46 +++++++++------
drivers/net/e1000/e1000_main.c | 110 +++++++++++++++++++++----------------
2 files changed, 88 insertions(+), 68 deletions(-)
diff --git a/drivers/net/e1000/e1000_ethtool.c b/drivers/net/e1000/e1000_ethtool.c
index c67e931..616b405 100644
--- a/drivers/net/e1000/e1000_ethtool.c
+++ b/drivers/net/e1000/e1000_ethtool.c
@@ -980,9 +980,10 @@ static void e1000_free_desc_rings(struct e1000_adapter *adapter)
if (txdr->desc && txdr->buffer_info) {
for (i = 0; i < txdr->count; i++) {
if (txdr->buffer_info[i].dma)
- pci_unmap_single(pdev, txdr->buffer_info[i].dma,
+ dma_unmap_single(&pdev->dev,
+ txdr->buffer_info[i].dma,
txdr->buffer_info[i].length,
- PCI_DMA_TODEVICE);
+ DMA_TO_DEVICE);
if (txdr->buffer_info[i].skb)
dev_kfree_skb(txdr->buffer_info[i].skb);
}
@@ -991,20 +992,23 @@ static void e1000_free_desc_rings(struct e1000_adapter *adapter)
if (rxdr->desc && rxdr->buffer_info) {
for (i = 0; i < rxdr->count; i++) {
if (rxdr->buffer_info[i].dma)
- pci_unmap_single(pdev, rxdr->buffer_info[i].dma,
+ dma_unmap_single(&pdev->dev,
+ rxdr->buffer_info[i].dma,
rxdr->buffer_info[i].length,
- PCI_DMA_FROMDEVICE);
+ DMA_FROM_DEVICE);
if (rxdr->buffer_info[i].skb)
dev_kfree_skb(rxdr->buffer_info[i].skb);
}
}
if (txdr->desc) {
- pci_free_consistent(pdev, txdr->size, txdr->desc, txdr->dma);
+ dma_free_coherent(&pdev->dev, txdr->size, txdr->desc,
+ txdr->dma);
txdr->desc = NULL;
}
if (rxdr->desc) {
- pci_free_consistent(pdev, rxdr->size, rxdr->desc, rxdr->dma);
+ dma_free_coherent(&pdev->dev, rxdr->size, rxdr->desc,
+ rxdr->dma);
rxdr->desc = NULL;
}
@@ -1039,7 +1043,8 @@ static int e1000_setup_desc_rings(struct e1000_adapter *adapter)
txdr->size = txdr->count * sizeof(struct e1000_tx_desc);
txdr->size = ALIGN(txdr->size, 4096);
- txdr->desc = pci_alloc_consistent(pdev, txdr->size, &txdr->dma);
+ txdr->desc = dma_alloc_coherent(&pdev->dev, txdr->size, &txdr->dma,
+ GFP_KERNEL);
if (!txdr->desc) {
ret_val = 2;
goto err_nomem;
@@ -1070,8 +1075,8 @@ static int e1000_setup_desc_rings(struct e1000_adapter *adapter)
txdr->buffer_info[i].skb = skb;
txdr->buffer_info[i].length = skb->len;
txdr->buffer_info[i].dma =
- pci_map_single(pdev, skb->data, skb->len,
- PCI_DMA_TODEVICE);
+ dma_map_single(&pdev->dev, skb->data, skb->len,
+ DMA_TO_DEVICE);
tx_desc->buffer_addr = cpu_to_le64(txdr->buffer_info[i].dma);
tx_desc->lower.data = cpu_to_le32(skb->len);
tx_desc->lower.data |= cpu_to_le32(E1000_TXD_CMD_EOP |
@@ -1093,7 +1098,8 @@ static int e1000_setup_desc_rings(struct e1000_adapter *adapter)
}
rxdr->size = rxdr->count * sizeof(struct e1000_rx_desc);
- rxdr->desc = pci_alloc_consistent(pdev, rxdr->size, &rxdr->dma);
+ rxdr->desc = dma_alloc_coherent(&pdev->dev, rxdr->size, &rxdr->dma,
+ GFP_KERNEL);
if (!rxdr->desc) {
ret_val = 5;
goto err_nomem;
@@ -1126,8 +1132,8 @@ static int e1000_setup_desc_rings(struct e1000_adapter *adapter)
rxdr->buffer_info[i].skb = skb;
rxdr->buffer_info[i].length = E1000_RXBUFFER_2048;
rxdr->buffer_info[i].dma =
- pci_map_single(pdev, skb->data, E1000_RXBUFFER_2048,
- PCI_DMA_FROMDEVICE);
+ dma_map_single(&pdev->dev, skb->data,
+ E1000_RXBUFFER_2048, DMA_FROM_DEVICE);
rx_desc->buffer_addr = cpu_to_le64(rxdr->buffer_info[i].dma);
memset(skb->data, 0x00, skb->len);
}
@@ -1444,10 +1450,10 @@ static int e1000_run_loopback_test(struct e1000_adapter *adapter)
for (i = 0; i < 64; i++) { /* send the packets */
e1000_create_lbtest_frame(txdr->buffer_info[i].skb,
1024);
- pci_dma_sync_single_for_device(pdev,
- txdr->buffer_info[k].dma,
- txdr->buffer_info[k].length,
- PCI_DMA_TODEVICE);
+ dma_sync_single_for_device(&pdev->dev,
+ txdr->buffer_info[k].dma,
+ txdr->buffer_info[k].length,
+ DMA_TO_DEVICE);
if (unlikely(++k == txdr->count)) k = 0;
}
ew32(TDT, k);
@@ -1455,10 +1461,10 @@ static int e1000_run_loopback_test(struct e1000_adapter *adapter)
time = jiffies; /* set the start time for the receive */
good_cnt = 0;
do { /* receive the sent packets */
- pci_dma_sync_single_for_cpu(pdev,
- rxdr->buffer_info[l].dma,
- rxdr->buffer_info[l].length,
- PCI_DMA_FROMDEVICE);
+ dma_sync_single_for_cpu(&pdev->dev,
+ rxdr->buffer_info[l].dma,
+ rxdr->buffer_info[l].length,
+ DMA_FROM_DEVICE);
ret_val = e1000_check_lbtest_frame(
rxdr->buffer_info[l].skb,
diff --git a/drivers/net/e1000/e1000_main.c b/drivers/net/e1000/e1000_main.c
index 974a02d..0a56e7d 100644
--- a/drivers/net/e1000/e1000_main.c
+++ b/drivers/net/e1000/e1000_main.c
@@ -823,13 +823,14 @@ static int __devinit e1000_probe(struct pci_dev *pdev,
if (err)
return err;
- if (!pci_set_dma_mask(pdev, DMA_BIT_MASK(64)) &&
- !pci_set_consistent_dma_mask(pdev, DMA_BIT_MASK(64))) {
+ if (!dma_set_mask(&pdev->dev, DMA_BIT_MASK(64)) &&
+ !dma_set_coherent_mask(&pdev->dev, DMA_BIT_MASK(64))) {
pci_using_dac = 1;
} else {
- err = pci_set_dma_mask(pdev, DMA_BIT_MASK(32));
+ err = dma_set_mask(&pdev->dev, DMA_BIT_MASK(32));
if (err) {
- err = pci_set_consistent_dma_mask(pdev, DMA_BIT_MASK(32));
+ err = dma_set_coherent_mask(&pdev->dev,
+ DMA_BIT_MASK(32));
if (err) {
E1000_ERR("No usable DMA configuration, "
"aborting\n");
@@ -1395,7 +1396,8 @@ static int e1000_setup_tx_resources(struct e1000_adapter *adapter,
txdr->size = txdr->count * sizeof(struct e1000_tx_desc);
txdr->size = ALIGN(txdr->size, 4096);
- txdr->desc = pci_alloc_consistent(pdev, txdr->size, &txdr->dma);
+ txdr->desc = dma_alloc_coherent(&pdev->dev, txdr->size, &txdr->dma,
+ GFP_KERNEL);
if (!txdr->desc) {
setup_tx_desc_die:
vfree(txdr->buffer_info);
@@ -1411,18 +1413,21 @@ setup_tx_desc_die:
DPRINTK(TX_ERR, ERR, "txdr align check failed: %u bytes "
"at %p\n", txdr->size, txdr->desc);
/* Try again, without freeing the previous */
- txdr->desc = pci_alloc_consistent(pdev, txdr->size, &txdr->dma);
+ txdr->desc = dma_alloc_coherent(&pdev->dev, txdr->size,
+ &txdr->dma, GFP_KERNEL);
/* Failed allocation, critical failure */
if (!txdr->desc) {
- pci_free_consistent(pdev, txdr->size, olddesc, olddma);
+ dma_free_coherent(&pdev->dev, txdr->size, olddesc,
+ olddma);
goto setup_tx_desc_die;
}
if (!e1000_check_64k_bound(adapter, txdr->desc, txdr->size)) {
/* give up */
- pci_free_consistent(pdev, txdr->size, txdr->desc,
- txdr->dma);
- pci_free_consistent(pdev, txdr->size, olddesc, olddma);
+ dma_free_coherent(&pdev->dev, txdr->size, txdr->desc,
+ txdr->dma);
+ dma_free_coherent(&pdev->dev, txdr->size, olddesc,
+ olddma);
DPRINTK(PROBE, ERR,
"Unable to allocate aligned memory "
"for the transmit descriptor ring\n");
@@ -1430,7 +1435,8 @@ setup_tx_desc_die:
return -ENOMEM;
} else {
/* Free old allocation, new allocation was successful */
- pci_free_consistent(pdev, txdr->size, olddesc, olddma);
+ dma_free_coherent(&pdev->dev, txdr->size, olddesc,
+ olddma);
}
}
memset(txdr->desc, 0, txdr->size);
@@ -1590,7 +1596,8 @@ static int e1000_setup_rx_resources(struct e1000_adapter *adapter,
rxdr->size = rxdr->count * desc_len;
rxdr->size = ALIGN(rxdr->size, 4096);
- rxdr->desc = pci_alloc_consistent(pdev, rxdr->size, &rxdr->dma);
+ rxdr->desc = dma_alloc_coherent(&pdev->dev, rxdr->size, &rxdr->dma,
+ GFP_KERNEL);
if (!rxdr->desc) {
DPRINTK(PROBE, ERR,
@@ -1607,10 +1614,12 @@ setup_rx_desc_die:
DPRINTK(RX_ERR, ERR, "rxdr align check failed: %u bytes "
"at %p\n", rxdr->size, rxdr->desc);
/* Try again, without freeing the previous */
- rxdr->desc = pci_alloc_consistent(pdev, rxdr->size, &rxdr->dma);
+ rxdr->desc = dma_alloc_coherent(&pdev->dev, rxdr->size,
+ &rxdr->dma, GFP_KERNEL);
/* Failed allocation, critical failure */
if (!rxdr->desc) {
- pci_free_consistent(pdev, rxdr->size, olddesc, olddma);
+ dma_free_coherent(&pdev->dev, rxdr->size, olddesc,
+ olddma);
DPRINTK(PROBE, ERR,
"Unable to allocate memory "
"for the receive descriptor ring\n");
@@ -1619,16 +1628,18 @@ setup_rx_desc_die:
if (!e1000_check_64k_bound(adapter, rxdr->desc, rxdr->size)) {
/* give up */
- pci_free_consistent(pdev, rxdr->size, rxdr->desc,
- rxdr->dma);
- pci_free_consistent(pdev, rxdr->size, olddesc, olddma);
+ dma_free_coherent(&pdev->dev, rxdr->size, rxdr->desc,
+ rxdr->dma);
+ dma_free_coherent(&pdev->dev, rxdr->size, olddesc,
+ olddma);
DPRINTK(PROBE, ERR,
"Unable to allocate aligned memory "
"for the receive descriptor ring\n");
goto setup_rx_desc_die;
} else {
/* Free old allocation, new allocation was successful */
- pci_free_consistent(pdev, rxdr->size, olddesc, olddma);
+ dma_free_coherent(&pdev->dev, rxdr->size, olddesc,
+ olddma);
}
}
memset(rxdr->desc, 0, rxdr->size);
@@ -1804,7 +1815,8 @@ static void e1000_free_tx_resources(struct e1000_adapter *adapter,
vfree(tx_ring->buffer_info);
tx_ring->buffer_info = NULL;
- pci_free_consistent(pdev, tx_ring->size, tx_ring->desc, tx_ring->dma);
+ dma_free_coherent(&pdev->dev, tx_ring->size, tx_ring->desc,
+ tx_ring->dma);
tx_ring->desc = NULL;
}
@@ -1829,12 +1841,12 @@ static void e1000_unmap_and_free_tx_resource(struct e1000_adapter *adapter,
{
if (buffer_info->dma) {
if (buffer_info->mapped_as_page)
- pci_unmap_page(adapter->pdev, buffer_info->dma,
- buffer_info->length, PCI_DMA_TODEVICE);
+ dma_unmap_page(&adapter->pdev->dev, buffer_info->dma,
+ buffer_info->length, DMA_TO_DEVICE);
else
- pci_unmap_single(adapter->pdev, buffer_info->dma,
+ dma_unmap_single(&adapter->pdev->dev, buffer_info->dma,
buffer_info->length,
- PCI_DMA_TODEVICE);
+ DMA_TO_DEVICE);
buffer_info->dma = 0;
}
if (buffer_info->skb) {
@@ -1912,7 +1924,8 @@ static void e1000_free_rx_resources(struct e1000_adapter *adapter,
vfree(rx_ring->buffer_info);
rx_ring->buffer_info = NULL;
- pci_free_consistent(pdev, rx_ring->size, rx_ring->desc, rx_ring->dma);
+ dma_free_coherent(&pdev->dev, rx_ring->size, rx_ring->desc,
+ rx_ring->dma);
rx_ring->desc = NULL;
}
@@ -1952,14 +1965,14 @@ static void e1000_clean_rx_ring(struct e1000_adapter *adapter,
buffer_info = &rx_ring->buffer_info[i];
if (buffer_info->dma &&
adapter->clean_rx == e1000_clean_rx_irq) {
- pci_unmap_single(pdev, buffer_info->dma,
+ dma_unmap_single(&pdev->dev, buffer_info->dma,
buffer_info->length,
- PCI_DMA_FROMDEVICE);
+ DMA_FROM_DEVICE);
} else if (buffer_info->dma &&
adapter->clean_rx == e1000_clean_jumbo_rx_irq) {
- pci_unmap_page(pdev, buffer_info->dma,
- buffer_info->length,
- PCI_DMA_FROMDEVICE);
+ dma_unmap_page(&pdev->dev, buffer_info->dma,
+ buffer_info->length,
+ DMA_FROM_DEVICE);
}
buffer_info->dma = 0;
@@ -2714,9 +2727,10 @@ static int e1000_tx_map(struct e1000_adapter *adapter,
/* set time_stamp *before* dma to help avoid a possible race */
buffer_info->time_stamp = jiffies;
buffer_info->mapped_as_page = false;
- buffer_info->dma = pci_map_single(pdev, skb->data + offset,
- size, PCI_DMA_TODEVICE);
- if (pci_dma_mapping_error(pdev, buffer_info->dma))
+ buffer_info->dma = dma_map_single(&pdev->dev,
+ skb->data + offset,
+ size, DMA_TO_DEVICE);
+ if (dma_mapping_error(&pdev->dev, buffer_info->dma))
goto dma_error;
buffer_info->next_to_watch = i;
@@ -2760,10 +2774,10 @@ static int e1000_tx_map(struct e1000_adapter *adapter,
buffer_info->length = size;
buffer_info->time_stamp = jiffies;
buffer_info->mapped_as_page = true;
- buffer_info->dma = pci_map_page(pdev, frag->page,
+ buffer_info->dma = dma_map_page(&pdev->dev, frag->page,
offset, size,
- PCI_DMA_TODEVICE);
- if (pci_dma_mapping_error(pdev, buffer_info->dma))
+ DMA_TO_DEVICE);
+ if (dma_mapping_error(&pdev->dev, buffer_info->dma))
goto dma_error;
buffer_info->next_to_watch = i;
@@ -3634,8 +3648,8 @@ static bool e1000_clean_jumbo_rx_irq(struct e1000_adapter *adapter,
cleaned = true;
cleaned_count++;
- pci_unmap_page(pdev, buffer_info->dma, buffer_info->length,
- PCI_DMA_FROMDEVICE);
+ dma_unmap_page(&pdev->dev, buffer_info->dma,
+ buffer_info->length, DMA_FROM_DEVICE);
buffer_info->dma = 0;
length = le16_to_cpu(rx_desc->length);
@@ -3817,8 +3831,8 @@ static bool e1000_clean_rx_irq(struct e1000_adapter *adapter,
cleaned = true;
cleaned_count++;
- pci_unmap_single(pdev, buffer_info->dma, buffer_info->length,
- PCI_DMA_FROMDEVICE);
+ dma_unmap_single(&pdev->dev, buffer_info->dma,
+ buffer_info->length, DMA_FROM_DEVICE);
buffer_info->dma = 0;
length = le16_to_cpu(rx_desc->length);
@@ -3998,11 +4012,11 @@ check_page:
}
if (!buffer_info->dma) {
- buffer_info->dma = pci_map_page(pdev,
+ buffer_info->dma = dma_map_page(&pdev->dev,
buffer_info->page, 0,
- buffer_info->length,
- PCI_DMA_FROMDEVICE);
- if (pci_dma_mapping_error(pdev, buffer_info->dma)) {
+ buffer_info->length,
+ DMA_FROM_DEVICE);
+ if (dma_mapping_error(&pdev->dev, buffer_info->dma)) {
put_page(buffer_info->page);
dev_kfree_skb(skb);
buffer_info->page = NULL;
@@ -4098,11 +4112,11 @@ static void e1000_alloc_rx_buffers(struct e1000_adapter *adapter,
buffer_info->skb = skb;
buffer_info->length = adapter->rx_buffer_len;
map_skb:
- buffer_info->dma = pci_map_single(pdev,
+ buffer_info->dma = dma_map_single(&pdev->dev,
skb->data,
buffer_info->length,
- PCI_DMA_FROMDEVICE);
- if (pci_dma_mapping_error(pdev, buffer_info->dma)) {
+ DMA_FROM_DEVICE);
+ if (dma_mapping_error(&pdev->dev, buffer_info->dma)) {
dev_kfree_skb(skb);
buffer_info->skb = NULL;
buffer_info->dma = 0;
@@ -4126,9 +4140,9 @@ map_skb:
dev_kfree_skb(skb);
buffer_info->skb = NULL;
- pci_unmap_single(pdev, buffer_info->dma,
+ dma_unmap_single(&pdev->dev, buffer_info->dma,
adapter->rx_buffer_len,
- PCI_DMA_FROMDEVICE);
+ DMA_FROM_DEVICE);
buffer_info->dma = 0;
adapter->alloc_rx_buff_failed++;
^ permalink raw reply related
page: next (older) | prev (newer) | latest
- recent:[subjects (threaded)|topics (new)|topics (active)]
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox