* Re: [PATCH] networking: fm10k: Fix build failure
From: David Miller @ 2014-10-10 5:01 UTC (permalink / raw)
To: bobby.prani
Cc: linux.nics, e1000-devel, bruce.w.allan, jesse.brandeburg,
linux-kernel, john.ronciak, netdev
In-Reply-To: <1412916929-21592-1-git-send-email-bobby.prani@gmail.com>
From: Pranith Kumar <bobby.prani@gmail.com>
Date: Fri, 10 Oct 2014 00:55:29 -0400
> The latest linus git tip (3.18-rc1) fails with the following build failure. Fix
> this by making PTP support explicit for fm10k driver.
>
> rivers/built-in.o: In function `fm10k_ptp_register':
> (.text+0x12e760): undefined reference to `ptp_clock_registER'
> drivers/built-in.o: In function `fm10k_ptp_unregister':
> (.text+0x12e7dc): undefined reference to `ptp_clock_unregister'
> Makefile:930: recipe for target 'vmlinux' failed
>
> Signed-off-by: Pranith Kumar <bobby.prani@gmail.com>
Please follow what other drivers do, which is to use "select" on this
Kconfig symbol.
Thanks.
------------------------------------------------------------------------------
Meet PCI DSS 3.0 Compliance Requirements with EventLog Analyzer
Achieve PCI DSS 3.0 Compliant Status with Out-of-the-box PCI DSS Reports
Are you Audit-Ready for PCI DSS 3.0 Compliance? Download White paper
Comply to PCI DSS 3.0 Requirement 10 and 11.5 with EventLog Analyzer
http://pubads.g.doubleclick.net/gampad/clk?id=154622311&iu=/4140/ostg.clktrk
_______________________________________________
E1000-devel mailing list
E1000-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/e1000-devel
To learn more about Intel® Ethernet, visit http://communities.intel.com/community/wired
^ permalink raw reply
* Re: [PATCH net-next 0/2] sunvnet: Packet processing in non-interrupt context.
From: Raghuram Kothakota @ 2014-10-10 4:56 UTC (permalink / raw)
To: David Miller; +Cc: sowmini.varadhan, netdev
In-Reply-To: <20141010.003628.646817757614430018.davem@davemloft.net>
On Oct 9, 2014, at 9:36 PM, David Miller <davem@davemloft.net> wrote:
> From: Raghuram Kothakota <Raghuram.Kothakota@oracle.com>
> Date: Thu, 9 Oct 2014 18:10:24 -0700
>
>> Lock less Tx and Rx implementation is very nice, but requires to the
>> code path to single threaded to achieve it.
>
> I would like to know how you believe the Linux LLTX "lockless TX"
> facility works.
>
Sorry, I used incorrect terminology in my email. My knowledge of LLTX
is limited and I am still learning. I was not referring to the LLTX, but about
the implementation of sunvnet transmit path and receive paths without locks. To me that
means only one thread of execution exists at a given time and I was
referring to it as single threadedness, which limits performance on SPARC CMT
processors today. Using methods to increase parallelism will help especially
when the traffic involves multiple connections, mainly from the point of view
of using multiple vCPUs to perform the processing where possible.
-Raghuram
> It's not truly lockless, it just turns off the generic networking TX
> path per-queue spinlock and puts the full burdon on the driver. It's
> just moving the locking from one place to another, not eliminating it.
> --
> To unsubscribe from this list: send the line "unsubscribe netdev" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply
* [PATCH] networking: fm10k: Fix build failure
From: Pranith Kumar @ 2014-10-10 4:55 UTC (permalink / raw)
To: Jeff Kirsher, Jesse Brandeburg, Bruce Allan, Carolyn Wyborny,
Don Skidmore, Greg Rose, Matthew Vick, John Ronciak,
Mitch Williams, Linux NICS, open list:INTEL ETHERNET DR...,
open list:NETWORKING DRIVERS, open list
The latest linus git tip (3.18-rc1) fails with the following build failure. Fix
this by making PTP support explicit for fm10k driver.
rivers/built-in.o: In function `fm10k_ptp_register':
(.text+0x12e760): undefined reference to `ptp_clock_registER'
drivers/built-in.o: In function `fm10k_ptp_unregister':
(.text+0x12e7dc): undefined reference to `ptp_clock_unregister'
Makefile:930: recipe for target 'vmlinux' failed
Signed-off-by: Pranith Kumar <bobby.prani@gmail.com>
---
drivers/net/ethernet/intel/Kconfig | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/net/ethernet/intel/Kconfig b/drivers/net/ethernet/intel/Kconfig
index 6a6d5ee..281d998 100644
--- a/drivers/net/ethernet/intel/Kconfig
+++ b/drivers/net/ethernet/intel/Kconfig
@@ -303,7 +303,7 @@ config I40EVF
config FM10K
tristate "Intel(R) FM10000 Ethernet Switch Host Interface Support"
default n
- depends on PCI_MSI
+ depends on PCI_MSI && PTP_1588_CLOCK
---help---
This driver supports Intel(R) FM10000 Ethernet Switch Host
Interface. For more information on how to identify your adapter,
--
1.9.1
------------------------------------------------------------------------------
Meet PCI DSS 3.0 Compliance Requirements with EventLog Analyzer
Achieve PCI DSS 3.0 Compliant Status with Out-of-the-box PCI DSS Reports
Are you Audit-Ready for PCI DSS 3.0 Compliance? Download White paper
Comply to PCI DSS 3.0 Requirement 10 and 11.5 with EventLog Analyzer
http://pubads.g.doubleclick.net/gampad/clk?id=154622311&iu=/4140/ostg.clktrk
_______________________________________________
E1000-devel mailing list
E1000-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/e1000-devel
To learn more about Intel® Ethernet, visit http://communities.intel.com/community/wired
^ permalink raw reply related
* Re: [PATCH] net: fs_enet: error: 'SCCE_ENET_TXF' undeclared
From: David Miller @ 2014-10-10 4:51 UTC (permalink / raw)
To: christophe.leroy
Cc: pantelis.antoniou, vbordug, linux-kernel, linuxppc-dev, netdev
In-Reply-To: <20141009145443.74AD61AB275@localhost.localdomain>
From: Christophe Leroy <christophe.leroy@c-s.fr>
Date: Thu, 9 Oct 2014 16:54:43 +0200 (CEST)
> [linux-devel:devel-hourly-2014100909 3763/3915] drivers/net/ethernet/freescale/fs_enet/mac-scc.c:119:32: error: 'SCCE_ENET_TXF' undeclared
>
> Due to patch d43a396 net: fs_enet: Add NAPI TX, it appears that some target
> compilations are broken.
> This is due to the fact that unlike the FEC, the SCC and FCC don't have a TXF
> event (complete Frame transmitted) but only TXB (buffer transmitted).
>
> Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
Applied, thanks.
^ permalink raw reply
* Re: [PATCH] flow-dissector: Fix alignment issue in __skb_flow_get_ports
From: David Miller @ 2014-10-10 4:47 UTC (permalink / raw)
To: alexander.duyck; +Cc: netdev
In-Reply-To: <20141010035840.21428.359.stgit@ahduyck-workstation.home>
From: alexander.duyck@gmail.com
Date: Thu, 09 Oct 2014 21:03:28 -0700
> From: Alexander Duyck <alexander.h.duyck@redhat.com>
>
> This patch addresses a kernel unaligned access bug seen on a sparc64 system
> with an igb adapter. Specifically the __skb_flow_get_ports was returning a
> be32 pointer which was then having the value directly returned.
>
> In order to keep the handling of the ports consistent with how we were
> handling the IPv4 and IPv6 addresses I have instead replaced the assignment
> with a memcpy to the flow key ports value. This way it should stay a
> memcpy on systems that cannot handle unaligned access, and will likely be
> converted to a 32b assignment on the systems that can support it.
>
> Reported-by: David S. Miller <davem@davemloft.net>
> Signed-off-by: Alexander Duyck <alexander.h.duyck@redhat.com>
Guess what the compiler will output for the memcpy()....
*(u32 *)dest = *(u32 *)src;
Using memcpy() is never a valid way to avoid misaligned loads and
stores.
If the types have a given alignment, the compiler can legitimately
expand the memcpy() inline with suitably sized loads and stores.
Please see my other reply in the original thread, we have to use
the hardware when we can to manage this situation, by configuring
it to output two garbage bytes before the packet contents when
DMA'ing into power-of-2 aligned blocks of memory.
Thanks.
^ permalink raw reply
* Re: eth_get_headlen() and unaligned accesses...
From: David Miller @ 2014-10-10 4:43 UTC (permalink / raw)
To: alexander.h.duyck; +Cc: netdev
In-Reply-To: <54374E09.1020907@redhat.com>
From: Alexander Duyck <alexander.h.duyck@redhat.com>
Date: Thu, 09 Oct 2014 20:10:01 -0700
> On 10/09/2014 05:12 PM, David Miller wrote:
>> So, we have a bit of a problem, this is on sparc64:
>>
>> [423475.740836] Kernel unaligned access at TPC[81d330]
>> __skb_flow_get_ports+0x70/0xe0
>> [423475.755756] CPU: 0 PID: 0 Comm: swapper/0 Not tainted 3.17.0+ #2
>> [423475.767854] Call Trace:
>> [423475.772877] [0000000000433288] kernel_unaligned_trap+0x368/0x5c0
>> [423475.785203] [000000000042a824] sun4v_do_mna+0x84/0xa0
>> [423475.795624] [0000000000406cd0] sun4v_mna+0x5c/0x68
>> [423475.805521] [000000000081d330] __skb_flow_get_ports+0x70/0xe0
>> [423475.817323] [000000000081d6ac] __skb_flow_dissect+0x1ac/0x460
>> [423475.829128] [0000000000843c98] eth_get_headlen+0x38/0xa0
>> [423475.840083] [0000000010064d54] igb_poll+0x8d4/0xf60 [igb]
>> [423475.851184] [00000000008243c8] net_rx_action+0xa8/0x1c0
>>
>> The chip DMA's to the beginning of a frag page and (unless timestamps
>> are enabled) that's where the ethernet header begins.
>>
>> So any larger than 16-bit access to the IP and later headers will be
>> unaligned.
>>
>> We have various ways we can deal with this based upon the capabilities
>> of the chips involved. Can we configure the IGB to put 2 "don't care"
>> bytes at the beginning of the packet?
>
> The problem is the igb part expects to be able to use 2K buffers which
> means it will always try to use the full half of a page.
Let me try to ask this again.
Can you configure the MAC to put two garbage bytes at the head of
the packet data as it feeds it into the DMA fifos on the IGB chip?
This is an essential (again: _essential_) feature for chips that
manage RX buffers as power-of-2 chunks of pages, as it is the only
cheap way to get the IP headers 32-bit aligned in those power-of-2 DMA
buffer blocks.
That would solve the whole problem.
^ permalink raw reply
* Re: [PATCH net-next 0/2] sunvnet: Packet processing in non-interrupt context.
From: David Miller @ 2014-10-10 4:36 UTC (permalink / raw)
To: Raghuram.Kothakota; +Cc: sowmini.varadhan, netdev
In-Reply-To: <EB9A1CD1-6E1D-409C-89CB-95CC9A5AFC1D@oracle.com>
From: Raghuram Kothakota <Raghuram.Kothakota@oracle.com>
Date: Thu, 9 Oct 2014 18:10:24 -0700
> Lock less Tx and Rx implementation is very nice, but requires to the
> code path to single threaded to achieve it.
I would like to know how you believe the Linux LLTX "lockless TX"
facility works.
It's not truly lockless, it just turns off the generic networking TX
path per-queue spinlock and puts the full burdon on the driver. It's
just moving the locking from one place to another, not eliminating it.
^ permalink raw reply
* Re: [PATCH][net-next] Documentation: replace __sk_run_filter with __bpf_prog_run
From: Alexei Starovoitov @ 2014-10-10 4:20 UTC (permalink / raw)
To: roy.qing.li, Daniel Borkmann; +Cc: netdev@vger.kernel.org
In-Reply-To: <1412912214-964-1-git-send-email-roy.qing.li@gmail.com>
On Thu, Oct 9, 2014 at 8:36 PM, <roy.qing.li@gmail.com> wrote:
> From: Li RongQing <roy.qing.li@gmail.com>
>
> __sk_run_filter has been renamed as __bpf_prog_run, so replace them in comments
>
> Signed-off-by: Li RongQing <roy.qing.li@gmail.com>
Acked-by: Alexei Starovoitov <ast@plumgrid.com>
> ---
> Documentation/networking/filter.txt | 4 ++--
> 1 file changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/Documentation/networking/filter.txt b/Documentation/networking/filter.txt
> index 5ce4d07..c22b797 100644
> --- a/Documentation/networking/filter.txt
> +++ b/Documentation/networking/filter.txt
> @@ -700,11 +700,11 @@ Some core changes of the new internal format:
> bpf_exit
>
> If f2 is JITed and the pointer stored to '_f2'. The calls f1 -> f2 -> f3 and
> - returns will be seamless. Without JIT, __sk_run_filter() interpreter needs to
> + returns will be seamless. Without JIT, __bpf_prog_run() interpreter needs to
> be used to call into f2.
>
> For practical reasons all eBPF programs have only one argument 'ctx' which is
> - already placed into R1 (e.g. on __sk_run_filter() startup) and the programs
> + already placed into R1 (e.g. on __bpf_prog_run() startup) and the programs
> can call kernel functions with up to 5 arguments. Calls with 6 or more arguments
> are currently not supported, but these restrictions can be lifted if necessary
> in the future.
> --
> 1.7.10.4
>
> --
> To unsubscribe from this list: send the line "unsubscribe netdev" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply
* Re: [PATCH][net-next] net: filter: replace sk_run_filter with __bpf_prog_run in comments
From: Alexei Starovoitov @ 2014-10-10 4:17 UTC (permalink / raw)
To: roy.qing.li; +Cc: netdev@vger.kernel.org
In-Reply-To: <1412912231-1004-1-git-send-email-roy.qing.li@gmail.com>
On Thu, Oct 9, 2014 at 8:37 PM, <roy.qing.li@gmail.com> wrote:
> From: Li RongQing <roy.qing.li@gmail.com>
>
> sk_run_filter has been renamed as __bpf_prog_run, so replace them in comments
>
> Signed-off-by: Li RongQing <roy.qing.li@gmail.com>
> ---
> net/core/filter.c | 8 ++++----
> 1 file changed, 4 insertions(+), 4 deletions(-)
>
> diff --git a/net/core/filter.c b/net/core/filter.c
> index fcd3f67..9324601 100644
> --- a/net/core/filter.c
> +++ b/net/core/filter.c
> @@ -51,9 +51,9 @@
> * @skb: buffer to filter
> *
> * Run the filter code and then cut skb->data to correct size returned by
> - * sk_run_filter. If pkt_len is 0 we toss packet. If skb->len is smaller
> + * __bpf_prog_run. If pkt_len is 0 we toss packet. If skb->len is smaller
> * than pkt_len we keep whole skb->data. This is the socket level
> - * wrapper to sk_run_filter. It returns 0 if the packet should
> + * wrapper to __bpf_prog_run. It returns 0 if the packet should
> * be accepted or -EPERM if the packet should be tossed.
not quite. I think in both of the above lines it's better to replace
sk_run_filter with SK_RUN_FILTER
> *
> */
> @@ -567,10 +567,10 @@ err:
> /* Security:
> *
> * A BPF program is able to use 16 cells of memory to store intermediate
> - * values (check u32 mem[BPF_MEMWORDS] in sk_run_filter()).
> + * values (check u32 mem[BPF_MEMWORDS] in __bpf_prog_run()).
this just wrong. Better to remove this line.
> *
> * As we dont want to clear mem[] array for each packet going through
> - * sk_run_filter(), we check that filter loaded by user never try to read
> + * __bpf_prog_run(), we check that filter loaded by user never try to read
here the change is fine.
^ permalink raw reply
* [PATCH] flow-dissector: Fix alignment issue in __skb_flow_get_ports
From: alexander.duyck @ 2014-10-10 4:03 UTC (permalink / raw)
To: netdev, davem
In-Reply-To: <20141009.201248.1210454965155680255.davem@davemloft.net>
From: Alexander Duyck <alexander.h.duyck@redhat.com>
This patch addresses a kernel unaligned access bug seen on a sparc64 system
with an igb adapter. Specifically the __skb_flow_get_ports was returning a
be32 pointer which was then having the value directly returned.
In order to keep the handling of the ports consistent with how we were
handling the IPv4 and IPv6 addresses I have instead replaced the assignment
with a memcpy to the flow key ports value. This way it should stay a
memcpy on systems that cannot handle unaligned access, and will likely be
converted to a 32b assignment on the systems that can support it.
Reported-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Alexander Duyck <alexander.h.duyck@redhat.com>
---
drivers/net/bonding/bond_main.c | 2 +-
include/net/flow_keys.h | 10 ++++++----
net/core/flow_dissector.c | 21 +++++++++++++--------
3 files changed, 20 insertions(+), 13 deletions(-)
diff --git a/drivers/net/bonding/bond_main.c b/drivers/net/bonding/bond_main.c
index c9ac06c..326eb3d 100644
--- a/drivers/net/bonding/bond_main.c
+++ b/drivers/net/bonding/bond_main.c
@@ -2993,7 +2993,7 @@ static bool bond_flow_dissect(struct bonding *bond, struct sk_buff *skb,
return false;
}
if (bond->params.xmit_policy == BOND_XMIT_POLICY_LAYER34 && proto >= 0)
- fk->ports = skb_flow_get_ports(skb, noff, proto);
+ skb_flow_get_ports(fk, skb, noff, proto);
return true;
}
diff --git a/include/net/flow_keys.h b/include/net/flow_keys.h
index 7ee2df0..66235f4 100644
--- a/include/net/flow_keys.h
+++ b/include/net/flow_keys.h
@@ -33,11 +33,13 @@ static inline bool skb_flow_dissect(const struct sk_buff *skb, struct flow_keys
{
return __skb_flow_dissect(skb, flow, NULL, 0, 0, 0);
}
-__be32 __skb_flow_get_ports(const struct sk_buff *skb, int thoff, u8 ip_proto,
- void *data, int hlen_proto);
-static inline __be32 skb_flow_get_ports(const struct sk_buff *skb, int thoff, u8 ip_proto)
+void __skb_flow_get_ports(struct flow_keys *flow, const struct sk_buff *skb,
+ int thoff, u8 ip_proto, void *data, int hlen_proto);
+static inline void skb_flow_get_ports(struct flow_keys *flow,
+ const struct sk_buff *skb, int thoff,
+ u8 ip_proto)
{
- return __skb_flow_get_ports(skb, thoff, ip_proto, NULL, 0);
+ __skb_flow_get_ports(flow, skb, thoff, ip_proto, NULL, 0);
}
u32 flow_hash_from_keys(struct flow_keys *keys);
unsigned int flow_get_hlen(const unsigned char *data, unsigned int max_len,
diff --git a/net/core/flow_dissector.c b/net/core/flow_dissector.c
index 8560dea..baf8fe3 100644
--- a/net/core/flow_dissector.c
+++ b/net/core/flow_dissector.c
@@ -28,6 +28,7 @@ static void iph_to_flow_copy_addrs(struct flow_keys *flow, const struct iphdr *i
/**
* __skb_flow_get_ports - extract the upper layer ports and return them
+ * @flow: Flow key to place port informatin into
* @skb: sk_buff to extract the ports from
* @thoff: transport header offset
* @ip_proto: protocol for which to get port offset
@@ -37,8 +38,8 @@ static void iph_to_flow_copy_addrs(struct flow_keys *flow, const struct iphdr *i
* The function will try to retrieve the ports at offset thoff + poff where poff
* is the protocol port offset returned from proto_ports_offset
*/
-__be32 __skb_flow_get_ports(const struct sk_buff *skb, int thoff, u8 ip_proto,
- void *data, int hlen)
+void __skb_flow_get_ports(struct flow_keys *flow, const struct sk_buff *skb,
+ int thoff, u8 ip_proto, void *data, int hlen)
{
int poff = proto_ports_offset(ip_proto);
@@ -48,15 +49,19 @@ __be32 __skb_flow_get_ports(const struct sk_buff *skb, int thoff, u8 ip_proto,
}
if (poff >= 0) {
- __be32 *ports, _ports;
+ __be32 *ports;
ports = __skb_header_pointer(skb, thoff + poff,
- sizeof(_ports), data, hlen, &_ports);
- if (ports)
- return *ports;
+ sizeof(flow->ports), data, hlen,
+ &flow->ports);
+ if (ports) {
+ if (&flow->ports != ports)
+ memcpy(&flow->ports, ports, sizeof(flow->ports));
+ return;
+ }
}
- return 0;
+ memset(&flow->ports, 0, sizeof(flow->ports));
}
EXPORT_SYMBOL(__skb_flow_get_ports);
@@ -231,7 +236,7 @@ ipv6:
flow->n_proto = proto;
flow->ip_proto = ip_proto;
- flow->ports = __skb_flow_get_ports(skb, nhoff, ip_proto, data, hlen);
+ __skb_flow_get_ports(flow, skb, nhoff, ip_proto, data, hlen);
flow->thoff = (u16) nhoff;
return true;
^ permalink raw reply related
* RE: [net-next PATCH v1 1/3] net: sched: af_packet support for direct ring access
From: Zhou, Danny @ 2014-10-10 3:49 UTC (permalink / raw)
To: David Miller
Cc: willemb@google.com, john.fastabend@gmail.com, dborkman@redhat.com,
fw@strlen.de, gerlitz.or@gmail.com, hannes@stressinduktion.org,
netdev@vger.kernel.org, Ronciak, John, amirv@mellanox.com,
eric.dumazet@gmail.com
In-Reply-To: <20141007.120534.1798634446901746809.davem@davemloft.net>
> -----Original Message-----
> From: David Miller [mailto:davem@davemloft.net]
> Sent: Wednesday, October 08, 2014 12:06 AM
> To: Zhou, Danny
> Cc: willemb@google.com; john.fastabend@gmail.com; dborkman@redhat.com; fw@strlen.de; gerlitz.or@gmail.com;
> hannes@stressinduktion.org; netdev@vger.kernel.org; Ronciak, John; amirv@mellanox.com; eric.dumazet@gmail.com
> Subject: Re: [net-next PATCH v1 1/3] net: sched: af_packet support for direct ring access
>
> From: "Zhou, Danny" <danny.zhou@intel.com>
> Date: Tue, 7 Oct 2014 15:21:15 +0000
>
> > Once qpairs split-off is done, the user space driver, as a slave
> > driver, will re-initialize those queues completely in user space by
> > using paddr(in the case of DPDK, vaddr of DPDK used huge pages are
> > translated to paddr) to fill in the packet descriptors. As of
> > security concern raised in previous discussion, the reason we
> > think(BTW, correct me if I am wrong) af_packet is most suitable is
> > because only user application with root permission is allowed to
> > successfully split-off queue pairs and mmap a small window of PCIe
> > I/O space to user space, so concern regarding "device can DMA
> > from/to any arbitrary physical memory." is not that big. As all user
> > space device drivers based on UIO mechanism has the same concern
> > issue, VFIO adds protection but it is based on IOMMU which is
> > specific to Intel silicons.
>
> Wait a second.
>
> If there is no memory protection performed I'm not merging this.
>
> I thought the user has to associate a fixed pool of memory to the
> queueus, the kernel attaches that memory, and then the user cannot
> modify the addresses _AT_ _ALL_.
>
> If the user can modify the addresses in the descriptors and make
> the chip crap on random memory, this is a non-starter.
>
> Sorry.
Fairly enough, we will manage to add memory protection in future versions.
Several options are under investigation.
^ permalink raw reply
* [PATCH][net-next] net: filter: replace sk_run_filter with __bpf_prog_run in comments
From: roy.qing.li @ 2014-10-10 3:37 UTC (permalink / raw)
To: netdev
From: Li RongQing <roy.qing.li@gmail.com>
sk_run_filter has been renamed as __bpf_prog_run, so replace them in comments
Signed-off-by: Li RongQing <roy.qing.li@gmail.com>
---
net/core/filter.c | 8 ++++----
1 file changed, 4 insertions(+), 4 deletions(-)
diff --git a/net/core/filter.c b/net/core/filter.c
index fcd3f67..9324601 100644
--- a/net/core/filter.c
+++ b/net/core/filter.c
@@ -51,9 +51,9 @@
* @skb: buffer to filter
*
* Run the filter code and then cut skb->data to correct size returned by
- * sk_run_filter. If pkt_len is 0 we toss packet. If skb->len is smaller
+ * __bpf_prog_run. If pkt_len is 0 we toss packet. If skb->len is smaller
* than pkt_len we keep whole skb->data. This is the socket level
- * wrapper to sk_run_filter. It returns 0 if the packet should
+ * wrapper to __bpf_prog_run. It returns 0 if the packet should
* be accepted or -EPERM if the packet should be tossed.
*
*/
@@ -567,10 +567,10 @@ err:
/* Security:
*
* A BPF program is able to use 16 cells of memory to store intermediate
- * values (check u32 mem[BPF_MEMWORDS] in sk_run_filter()).
+ * values (check u32 mem[BPF_MEMWORDS] in __bpf_prog_run()).
*
* As we dont want to clear mem[] array for each packet going through
- * sk_run_filter(), we check that filter loaded by user never try to read
+ * __bpf_prog_run(), we check that filter loaded by user never try to read
* a cell if not previously written, and we check all branches to be sure
* a malicious user doesn't try to abuse us.
*/
--
1.7.10.4
^ permalink raw reply related
* [PATCH][net-next] Documentation: replace __sk_run_filter with __bpf_prog_run
From: roy.qing.li @ 2014-10-10 3:36 UTC (permalink / raw)
To: netdev
From: Li RongQing <roy.qing.li@gmail.com>
__sk_run_filter has been renamed as __bpf_prog_run, so replace them in comments
Signed-off-by: Li RongQing <roy.qing.li@gmail.com>
---
Documentation/networking/filter.txt | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/Documentation/networking/filter.txt b/Documentation/networking/filter.txt
index 5ce4d07..c22b797 100644
--- a/Documentation/networking/filter.txt
+++ b/Documentation/networking/filter.txt
@@ -700,11 +700,11 @@ Some core changes of the new internal format:
bpf_exit
If f2 is JITed and the pointer stored to '_f2'. The calls f1 -> f2 -> f3 and
- returns will be seamless. Without JIT, __sk_run_filter() interpreter needs to
+ returns will be seamless. Without JIT, __bpf_prog_run() interpreter needs to
be used to call into f2.
For practical reasons all eBPF programs have only one argument 'ctx' which is
- already placed into R1 (e.g. on __sk_run_filter() startup) and the programs
+ already placed into R1 (e.g. on __bpf_prog_run() startup) and the programs
can call kernel functions with up to 5 arguments. Calls with 6 or more arguments
are currently not supported, but these restrictions can be lifted if necessary
in the future.
--
1.7.10.4
^ permalink raw reply related
* [PATCH net-next 2/2] macvlan: optimize the receive path
From: Jason Baron @ 2014-10-10 3:13 UTC (permalink / raw)
To: davem; +Cc: eric.dumazet, stephen, vyasevich, kaber, netdev
In-Reply-To: <cover.1412908929.git.jbaron@akamai.com>
The netif_rx() call on the fast path of macvlan_handle_frame() appears to
be there to ensure that we properly throttle incoming packets. However, it
would appear as though the proper throttling is already in place for all
possible ingress paths, and that the call is redundant. If packets are arriving
from the physical NIC, we've already throttled them by this point. Otherwise,
if they are coming via macvlan_queue_xmit(), it calls either
'dev_forward_skb()', which ends up calling netif_rx_internal(), or else in
the broadcast case, we are throttling via macvlan_broadcast_enqueue().
The test results below are from off the box to an lxc instance running macvlan.
Once the tranactions/sec stop increasing, the cpu idle time has gone to 0.
Results are from a quad core Intel E3-1270 V2@3.50GHz box with bnx2x 10G card.
for i in {10,100,200,300,400,500};
do super_netperf $i -H $ip -t TCP_RR; done
Average of 5 runs.
trans/sec trans/sec
(3.17-rc7-net-next) (3.17-rc7-net-next + this patch)
---------- ----------
208101 211534 (+1.6%)
839493 850162 (+1.3%)
845071 844053 (-.12%)
816330 819623 (+.4%)
778700 789938 (+1.4%)
735984 754408 (+2.5%)
Signed-off-by: Jason Baron <jbaron@akamai.com>
---
drivers/net/macvlan.c | 15 ++++++++++-----
1 file changed, 10 insertions(+), 5 deletions(-)
diff --git a/drivers/net/macvlan.c b/drivers/net/macvlan.c
index c7c58af..29b3bb4 100644
--- a/drivers/net/macvlan.c
+++ b/drivers/net/macvlan.c
@@ -407,7 +407,8 @@ static rx_handler_result_t macvlan_handle_frame(struct sk_buff **pskb)
const struct macvlan_dev *src;
struct net_device *dev;
unsigned int len = 0;
- int ret = NET_RX_DROP;
+ int ret;
+ rx_handler_result_t handle_res;
port = macvlan_port_get_rcu(skb->dev);
if (is_multicast_ether_addr(eth->h_dest)) {
@@ -423,6 +424,7 @@ static rx_handler_result_t macvlan_handle_frame(struct sk_buff **pskb)
vlan = src;
ret = macvlan_broadcast_one(skb, vlan, eth, 0) ?:
netif_rx(skb);
+ handle_res = RX_HANDLER_CONSUMED;
goto out;
}
@@ -448,17 +450,20 @@ static rx_handler_result_t macvlan_handle_frame(struct sk_buff **pskb)
}
len = skb->len + ETH_HLEN;
skb = skb_share_check(skb, GFP_ATOMIC);
- if (!skb)
+ if (!skb) {
+ ret = NET_RX_DROP;
+ handle_res = RX_HANDLER_CONSUMED;
goto out;
+ }
skb->dev = dev;
skb->pkt_type = PACKET_HOST;
- ret = netif_rx(skb);
-
+ ret = NET_RX_SUCCESS;
+ handle_res = RX_HANDLER_ANOTHER;
out:
macvlan_count_rx(vlan, len, ret == NET_RX_SUCCESS, false);
- return RX_HANDLER_CONSUMED;
+ return handle_res;
}
static int macvlan_queue_xmit(struct sk_buff *skb, struct net_device *dev)
--
1.8.2.rc2
^ permalink raw reply related
* [PATCH net-next 1/2] macvlan: pass 'bool' type to macvlan_count_rx()
From: Jason Baron @ 2014-10-10 3:13 UTC (permalink / raw)
To: davem; +Cc: eric.dumazet, stephen, vyasevich, kaber, netdev
In-Reply-To: <cover.1412908929.git.jbaron@akamai.com>
Pass last argument to macvlan_count_rx() as the correct bool type.
Signed-off-by: Jason Baron <jbaron@akamai.com>
---
drivers/net/macvlan.c | 6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)
diff --git a/drivers/net/macvlan.c b/drivers/net/macvlan.c
index 38b4fae..c7c58af 100644
--- a/drivers/net/macvlan.c
+++ b/drivers/net/macvlan.c
@@ -260,7 +260,7 @@ static void macvlan_broadcast(struct sk_buff *skb,
mode == MACVLAN_MODE_BRIDGE) ?:
netif_rx_ni(nskb);
macvlan_count_rx(vlan, skb->len + ETH_HLEN,
- err == NET_RX_SUCCESS, 1);
+ err == NET_RX_SUCCESS, true);
}
}
}
@@ -379,7 +379,7 @@ static void macvlan_forward_source_one(struct sk_buff *skb,
nskb->pkt_type = PACKET_HOST;
ret = netif_rx(nskb);
- macvlan_count_rx(vlan, len, ret == NET_RX_SUCCESS, 0);
+ macvlan_count_rx(vlan, len, ret == NET_RX_SUCCESS, false);
}
static void macvlan_forward_source(struct sk_buff *skb,
@@ -457,7 +457,7 @@ static rx_handler_result_t macvlan_handle_frame(struct sk_buff **pskb)
ret = netif_rx(skb);
out:
- macvlan_count_rx(vlan, len, ret == NET_RX_SUCCESS, 0);
+ macvlan_count_rx(vlan, len, ret == NET_RX_SUCCESS, false);
return RX_HANDLER_CONSUMED;
}
--
1.8.2.rc2
^ permalink raw reply related
* [PATCH net-next 0/2] macvlan: optimize receive path
From: Jason Baron @ 2014-10-10 3:13 UTC (permalink / raw)
To: davem; +Cc: eric.dumazet, stephen, vyasevich, kaber, netdev
Hi,
So after porting this optimization to net-next, I found that the netperf
results of TCP_RR regress right at the maximum peak of transactions/sec. That
is as I increase the number of threads via the first argument to super_netperf,
the number of transactions/sec keep increasing, peak, and then start
decreasing. It is right at the peak, that I see a small regression with this
patch (see results in patch 2/2).
Without the patch, the ksoftirqd threads are the top cpu consumers threads on
the system, since the extra 'netif_rx()', is queuing more softirq work, whereas
with the patch, the ksoftirqd threads are below all of the 'netserver' threads
in terms of their cpu usage. So there appears to be some interaction between how
softirqs are serviced at the peak here and this patch. I think the test results
are still supportive of this approach, but I wanted to be clear on my findings.
Thanks,
-Jason
Jason Baron (2):
macvlan: pass 'bool' type to macvlan_count_rx()
macvlan: optimize the receive path
drivers/net/macvlan.c | 21 +++++++++++++--------
1 file changed, 13 insertions(+), 8 deletions(-)
--
1.8.2.rc2
^ permalink raw reply
* Re: eth_get_headlen() and unaligned accesses...
From: Alexander Duyck @ 2014-10-10 3:10 UTC (permalink / raw)
To: David Miller, netdev
In-Reply-To: <20141009.201248.1210454965155680255.davem@davemloft.net>
On 10/09/2014 05:12 PM, David Miller wrote:
> So, we have a bit of a problem, this is on sparc64:
>
> [423475.740836] Kernel unaligned access at TPC[81d330] __skb_flow_get_ports+0x70/0xe0
> [423475.755756] CPU: 0 PID: 0 Comm: swapper/0 Not tainted 3.17.0+ #2
> [423475.767854] Call Trace:
> [423475.772877] [0000000000433288] kernel_unaligned_trap+0x368/0x5c0
> [423475.785203] [000000000042a824] sun4v_do_mna+0x84/0xa0
> [423475.795624] [0000000000406cd0] sun4v_mna+0x5c/0x68
> [423475.805521] [000000000081d330] __skb_flow_get_ports+0x70/0xe0
> [423475.817323] [000000000081d6ac] __skb_flow_dissect+0x1ac/0x460
> [423475.829128] [0000000000843c98] eth_get_headlen+0x38/0xa0
> [423475.840083] [0000000010064d54] igb_poll+0x8d4/0xf60 [igb]
> [423475.851184] [00000000008243c8] net_rx_action+0xa8/0x1c0
>
> The chip DMA's to the beginning of a frag page and (unless timestamps
> are enabled) that's where the ethernet header begins.
>
> So any larger than 16-bit access to the IP and later headers will be
> unaligned.
>
> We have various ways we can deal with this based upon the capabilities
> of the chips involved. Can we configure the IGB to put 2 "don't care"
> bytes at the beginning of the packet?
The problem is the igb part expects to be able to use 2K buffers which
means it will always try to use the full half of a page. I had
forgotten that the function this replaced had worked with unaligned
accesses as all of the fields I was pulling were only 16b in width. I
think I assumed that this function was already setup to handle that.
Actually the fix should be pretty simple. Just do what we already
appear to be doing for the iph_to_flow_copy_addrs. We can use memcpy to
copy the 4 bytes for the port data instead of doing the direct assignment.
I'll try to submit a patch, just need to see if I have a tree setup as
it has been a couple weeks.
Thanks,
Alex
^ permalink raw reply
* [PATCH v4 6/6] drivers: net: xgene: Add 10GbE ethtool support
From: Iyappan Subramanian @ 2014-10-10 1:32 UTC (permalink / raw)
To: davem, netdev, devicetree
Cc: linux-arm-kernel, patches, kchudgar, Iyappan Subramanian
In-Reply-To: <1412904727-23485-1-git-send-email-isubramanian@apm.com>
Signed-off-by: Iyappan Subramanian <isubramanian@apm.com>
Signed-off-by: Keyur Chudgar <kchudgar@apm.com>
---
.../net/ethernet/apm/xgene/xgene_enet_ethtool.c | 28 +++++++++++++++++-----
1 file changed, 22 insertions(+), 6 deletions(-)
diff --git a/drivers/net/ethernet/apm/xgene/xgene_enet_ethtool.c b/drivers/net/ethernet/apm/xgene/xgene_enet_ethtool.c
index 63f2aa5..c1c997b 100644
--- a/drivers/net/ethernet/apm/xgene/xgene_enet_ethtool.c
+++ b/drivers/net/ethernet/apm/xgene/xgene_enet_ethtool.c
@@ -59,10 +59,22 @@ static int xgene_get_settings(struct net_device *ndev, struct ethtool_cmd *cmd)
struct xgene_enet_pdata *pdata = netdev_priv(ndev);
struct phy_device *phydev = pdata->phy_dev;
- if (phydev == NULL)
- return -ENODEV;
+ if (pdata->phy_mode == PHY_INTERFACE_MODE_RGMII) {
+ if (phydev == NULL)
+ return -ENODEV;
- return phy_ethtool_gset(phydev, cmd);
+ return phy_ethtool_gset(phydev, cmd);
+ }
+
+ cmd->supported = SUPPORTED_10000baseT_Full | SUPPORTED_FIBRE;
+ cmd->advertising = cmd->supported;
+ ethtool_cmd_speed_set(cmd, SPEED_10000);
+ cmd->duplex = DUPLEX_FULL;
+ cmd->port = PORT_FIBRE;
+ cmd->transceiver = XCVR_EXTERNAL;
+ cmd->autoneg = AUTONEG_DISABLE;
+
+ return 0;
}
static int xgene_set_settings(struct net_device *ndev, struct ethtool_cmd *cmd)
@@ -70,10 +82,14 @@ static int xgene_set_settings(struct net_device *ndev, struct ethtool_cmd *cmd)
struct xgene_enet_pdata *pdata = netdev_priv(ndev);
struct phy_device *phydev = pdata->phy_dev;
- if (phydev == NULL)
- return -ENODEV;
+ if (pdata->phy_mode == PHY_INTERFACE_MODE_RGMII) {
+ if (phydev == NULL)
+ return -ENODEV;
+
+ return phy_ethtool_sset(phydev, cmd);
+ }
- return phy_ethtool_sset(phydev, cmd);
+ return -EINVAL;
}
static void xgene_get_strings(struct net_device *ndev, u32 stringset, u8 *data)
--
1.9.1
^ permalink raw reply related
* [PATCH v4 5/6] drivers: net: xgene: Add 10GbE support
From: Iyappan Subramanian @ 2014-10-10 1:32 UTC (permalink / raw)
To: davem, netdev, devicetree
Cc: linux-arm-kernel, patches, kchudgar, Iyappan Subramanian
In-Reply-To: <1412904727-23485-1-git-send-email-isubramanian@apm.com>
- Added 10GbE support
- Removed unused macros/variables
- Moved mac_init call to the end of hardware init
Signed-off-by: Iyappan Subramanian <isubramanian@apm.com>
Signed-off-by: Keyur Chudgar <kchudgar@apm.com>
---
drivers/net/ethernet/apm/xgene/Makefile | 3 +-
drivers/net/ethernet/apm/xgene/xgene_enet_hw.h | 14 +-
drivers/net/ethernet/apm/xgene/xgene_enet_main.c | 58 ++--
drivers/net/ethernet/apm/xgene/xgene_enet_main.h | 5 +-
drivers/net/ethernet/apm/xgene/xgene_enet_xgmac.c | 331 ++++++++++++++++++++++
drivers/net/ethernet/apm/xgene/xgene_enet_xgmac.h | 57 ++++
6 files changed, 438 insertions(+), 30 deletions(-)
create mode 100644 drivers/net/ethernet/apm/xgene/xgene_enet_xgmac.c
create mode 100644 drivers/net/ethernet/apm/xgene/xgene_enet_xgmac.h
diff --git a/drivers/net/ethernet/apm/xgene/Makefile b/drivers/net/ethernet/apm/xgene/Makefile
index c643e8a..589b352 100644
--- a/drivers/net/ethernet/apm/xgene/Makefile
+++ b/drivers/net/ethernet/apm/xgene/Makefile
@@ -2,5 +2,6 @@
# Makefile for APM X-Gene Ethernet Driver.
#
-xgene-enet-objs := xgene_enet_hw.o xgene_enet_main.o xgene_enet_ethtool.o
+xgene-enet-objs := xgene_enet_hw.o xgene_enet_xgmac.o \
+ xgene_enet_main.o xgene_enet_ethtool.o
obj-$(CONFIG_NET_XGENE) += xgene-enet.o
diff --git a/drivers/net/ethernet/apm/xgene/xgene_enet_hw.h b/drivers/net/ethernet/apm/xgene/xgene_enet_hw.h
index 084ac68..15ec426 100644
--- a/drivers/net/ethernet/apm/xgene/xgene_enet_hw.h
+++ b/drivers/net/ethernet/apm/xgene/xgene_enet_hw.h
@@ -42,6 +42,11 @@ static inline u32 xgene_get_bits(u32 val, u32 start, u32 end)
return (val & GENMASK(end, start)) >> start;
}
+enum xgene_enet_rm {
+ RM0,
+ RM3 = 3
+};
+
#define CSR_RING_ID 0x0008
#define OVERWRITE BIT(31)
#define IS_BUFFER_POOL BIT(20)
@@ -52,7 +57,6 @@ static inline u32 xgene_get_bits(u32 val, u32 start, u32 end)
#define CSR_RING_WR_BASE 0x0070
#define NUM_RING_CONFIG 5
#define BUFPOOL_MODE 3
-#define RM3 3
#define INC_DEC_CMD_ADDR 0x002c
#define UDP_HDR_SIZE 2
#define BUF_LEN_CODE_2K 0x5000
@@ -94,11 +98,9 @@ static inline u32 xgene_get_bits(u32 val, u32 start, u32 end)
#define BLOCK_ETH_CSR_OFFSET 0x2000
#define BLOCK_ETH_RING_IF_OFFSET 0x9000
-#define BLOCK_ETH_CLKRST_CSR_OFFSET 0xC000
#define BLOCK_ETH_DIAG_CSR_OFFSET 0xD000
#define BLOCK_ETH_MAC_OFFSET 0x0000
-#define BLOCK_ETH_STATS_OFFSET 0x0014
#define BLOCK_ETH_MAC_CSR_OFFSET 0x2800
#define MAC_ADDR_REG_OFFSET 0x00
@@ -107,12 +109,6 @@ static inline u32 xgene_get_bits(u32 val, u32 start, u32 end)
#define MAC_READ_REG_OFFSET 0x0c
#define MAC_COMMAND_DONE_REG_OFFSET 0x10
-#define STAT_ADDR_REG_OFFSET 0x00
-#define STAT_COMMAND_REG_OFFSET 0x04
-#define STAT_WRITE_REG_OFFSET 0x08
-#define STAT_READ_REG_OFFSET 0x0c
-#define STAT_COMMAND_DONE_REG_OFFSET 0x10
-
#define MII_MGMT_CONFIG_ADDR 0x20
#define MII_MGMT_COMMAND_ADDR 0x24
#define MII_MGMT_ADDRESS_ADDR 0x28
diff --git a/drivers/net/ethernet/apm/xgene/xgene_enet_main.c b/drivers/net/ethernet/apm/xgene/xgene_enet_main.c
index c432644..9b85239 100644
--- a/drivers/net/ethernet/apm/xgene/xgene_enet_main.c
+++ b/drivers/net/ethernet/apm/xgene/xgene_enet_main.c
@@ -21,6 +21,7 @@
#include "xgene_enet_main.h"
#include "xgene_enet_hw.h"
+#include "xgene_enet_xgmac.h"
static void xgene_enet_init_bufpool(struct xgene_enet_desc_ring *buf_pool)
{
@@ -390,7 +391,7 @@ static int xgene_enet_process_ring(struct xgene_enet_desc_ring *ring,
}
}
- return budget;
+ return count;
}
static int xgene_enet_napi(struct napi_struct *napi, const int budget)
@@ -456,8 +457,10 @@ static int xgene_enet_open(struct net_device *ndev)
return ret;
napi_enable(&pdata->rx_ring->napi);
- if (pdata->phy_dev)
+ if (pdata->phy_mode == PHY_INTERFACE_MODE_RGMII)
phy_start(pdata->phy_dev);
+ else
+ schedule_delayed_work(&pdata->link_work, PHY_POLL_LINK_OFF);
netif_start_queue(ndev);
@@ -471,8 +474,10 @@ static int xgene_enet_close(struct net_device *ndev)
netif_stop_queue(ndev);
- if (pdata->phy_dev)
+ if (pdata->phy_mode == PHY_INTERFACE_MODE_RGMII)
phy_stop(pdata->phy_dev);
+ else
+ cancel_delayed_work_sync(&pdata->link_work);
napi_disable(&pdata->rx_ring->napi);
xgene_enet_free_irq(ndev);
@@ -615,7 +620,6 @@ static struct xgene_enet_desc_ring *xgene_enet_create_desc_ring(
ring->cmd_base = pdata->ring_cmd_addr + (ring->num << 6);
ring->cmd = ring->cmd_base + INC_DEC_CMD_ADDR;
- pdata->rm = RM3;
ring = xgene_enet_setup_ring(ring);
netdev_dbg(ndev, "ring info: num=%d size=%d id=%d slots=%d\n",
ring->num, ring->size, ring->id, ring->slots);
@@ -805,8 +809,13 @@ static int xgene_enet_get_resources(struct xgene_enet_pdata *pdata)
pdata->phy_mode = of_get_phy_mode(pdev->dev.of_node);
if (pdata->phy_mode < 0) {
- dev_err(dev, "Incorrect phy-connection-type in DTS\n");
- return -EINVAL;
+ dev_err(dev, "Unable to get phy-connection-type\n");
+ return pdata->phy_mode;
+ }
+ if (pdata->phy_mode != PHY_INTERFACE_MODE_RGMII &&
+ pdata->phy_mode != PHY_INTERFACE_MODE_XGMII) {
+ dev_err(dev, "Incorrect phy-connection-type specified\n");
+ return -ENODEV;
}
pdata->clk = devm_clk_get(&pdev->dev, NULL);
@@ -821,12 +830,18 @@ static int xgene_enet_get_resources(struct xgene_enet_pdata *pdata)
pdata->eth_csr_addr = base_addr + BLOCK_ETH_CSR_OFFSET;
pdata->eth_ring_if_addr = base_addr + BLOCK_ETH_RING_IF_OFFSET;
pdata->eth_diag_csr_addr = base_addr + BLOCK_ETH_DIAG_CSR_OFFSET;
- pdata->mcx_mac_addr = base_addr + BLOCK_ETH_MAC_OFFSET;
- pdata->mcx_stats_addr = base_addr + BLOCK_ETH_STATS_OFFSET;
- pdata->mcx_mac_csr_addr = base_addr + BLOCK_ETH_MAC_CSR_OFFSET;
+ if (pdata->phy_mode == PHY_INTERFACE_MODE_RGMII) {
+ pdata->mcx_mac_addr = base_addr + BLOCK_ETH_MAC_OFFSET;
+ pdata->mcx_mac_csr_addr = base_addr + BLOCK_ETH_MAC_CSR_OFFSET;
+ pdata->rm = RM3;
+ } else {
+ pdata->mcx_mac_addr = base_addr + BLOCK_AXG_MAC_OFFSET;
+ pdata->mcx_mac_csr_addr = base_addr + BLOCK_AXG_MAC_CSR_OFFSET;
+ pdata->rm = RM0;
+ }
pdata->rx_buff_cnt = NUM_PKT_BUF;
- return ret;
+ return 0;
}
static int xgene_enet_init_hw(struct xgene_enet_pdata *pdata)
@@ -836,8 +851,7 @@ static int xgene_enet_init_hw(struct xgene_enet_pdata *pdata)
u16 dst_ring_num;
int ret;
- pdata->mac_ops->tx_disable(pdata);
- pdata->mac_ops->rx_disable(pdata);
+ pdata->port_ops->reset(pdata);
ret = xgene_enet_create_desc_rings(ndev);
if (ret) {
@@ -856,14 +870,23 @@ static int xgene_enet_init_hw(struct xgene_enet_pdata *pdata)
dst_ring_num = xgene_enet_dst_ring_num(pdata->rx_ring);
pdata->port_ops->cle_bypass(pdata, dst_ring_num, buf_pool->id);
+ pdata->mac_ops->init(pdata);
return ret;
}
static void xgene_enet_setup_ops(struct xgene_enet_pdata *pdata)
{
- pdata->mac_ops = &xgene_gmac_ops;
- pdata->port_ops = &xgene_gport_ops;
+ switch (pdata->phy_mode) {
+ case PHY_INTERFACE_MODE_RGMII:
+ pdata->mac_ops = &xgene_gmac_ops;
+ pdata->port_ops = &xgene_gport_ops;
+ break;
+ default:
+ pdata->mac_ops = &xgene_xgmac_ops;
+ pdata->port_ops = &xgene_xgport_ops;
+ break;
+ }
}
static int xgene_enet_probe(struct platform_device *pdev)
@@ -895,8 +918,6 @@ static int xgene_enet_probe(struct platform_device *pdev)
goto err;
xgene_enet_setup_ops(pdata);
- pdata->port_ops->reset(pdata);
- pdata->mac_ops->init(pdata);
ret = register_netdev(ndev);
if (ret) {
@@ -916,7 +937,10 @@ static int xgene_enet_probe(struct platform_device *pdev)
napi = &pdata->rx_ring->napi;
netif_napi_add(ndev, napi, xgene_enet_napi, NAPI_POLL_WEIGHT);
- ret = xgene_enet_mdio_config(pdata);
+ if (pdata->phy_mode == PHY_INTERFACE_MODE_RGMII)
+ ret = xgene_enet_mdio_config(pdata);
+ else
+ INIT_DELAYED_WORK(&pdata->link_work, xgene_enet_link_state);
return ret;
err:
diff --git a/drivers/net/ethernet/apm/xgene/xgene_enet_main.h b/drivers/net/ethernet/apm/xgene/xgene_enet_main.h
index ac180f9..86cf68b 100644
--- a/drivers/net/ethernet/apm/xgene/xgene_enet_main.h
+++ b/drivers/net/ethernet/apm/xgene/xgene_enet_main.h
@@ -105,18 +105,17 @@ struct xgene_enet_pdata {
void __iomem *eth_ring_if_addr;
void __iomem *eth_diag_csr_addr;
void __iomem *mcx_mac_addr;
- void __iomem *mcx_stats_addr;
void __iomem *mcx_mac_csr_addr;
void __iomem *base_addr;
void __iomem *ring_csr_addr;
void __iomem *ring_cmd_addr;
u32 phy_addr;
int phy_mode;
- u32 speed;
- u16 rm;
+ enum xgene_enet_rm rm;
struct rtnl_link_stats64 stats;
struct xgene_mac_ops *mac_ops;
struct xgene_port_ops *port_ops;
+ struct delayed_work link_work;
};
/* Set the specified value into a bit-field defined by its starting position
diff --git a/drivers/net/ethernet/apm/xgene/xgene_enet_xgmac.c b/drivers/net/ethernet/apm/xgene/xgene_enet_xgmac.c
new file mode 100644
index 0000000..cd64b9f
--- /dev/null
+++ b/drivers/net/ethernet/apm/xgene/xgene_enet_xgmac.c
@@ -0,0 +1,331 @@
+/* Applied Micro X-Gene SoC Ethernet Driver
+ *
+ * Copyright (c) 2014, Applied Micro Circuits Corporation
+ * Authors: Iyappan Subramanian <isubramanian@apm.com>
+ * Keyur Chudgar <kchudgar@apm.com>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License as published by the
+ * Free Software Foundation; either version 2 of the License, or (at your
+ * option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program. If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#include "xgene_enet_main.h"
+#include "xgene_enet_hw.h"
+#include "xgene_enet_xgmac.h"
+
+static void xgene_enet_wr_csr(struct xgene_enet_pdata *pdata,
+ u32 offset, u32 val)
+{
+ void __iomem *addr = pdata->eth_csr_addr + offset;
+
+ iowrite32(val, addr);
+}
+
+static void xgene_enet_wr_ring_if(struct xgene_enet_pdata *pdata,
+ u32 offset, u32 val)
+{
+ void __iomem *addr = pdata->eth_ring_if_addr + offset;
+
+ iowrite32(val, addr);
+}
+
+static void xgene_enet_wr_diag_csr(struct xgene_enet_pdata *pdata,
+ u32 offset, u32 val)
+{
+ void __iomem *addr = pdata->eth_diag_csr_addr + offset;
+
+ iowrite32(val, addr);
+}
+
+static bool xgene_enet_wr_indirect(void __iomem *addr, void __iomem *wr,
+ void __iomem *cmd, void __iomem *cmd_done,
+ u32 wr_addr, u32 wr_data)
+{
+ u32 done;
+ u8 wait = 10;
+
+ iowrite32(wr_addr, addr);
+ iowrite32(wr_data, wr);
+ iowrite32(XGENE_ENET_WR_CMD, cmd);
+
+ /* wait for write command to complete */
+ while (!(done = ioread32(cmd_done)) && wait--)
+ udelay(1);
+
+ if (!done)
+ return false;
+
+ iowrite32(0, cmd);
+
+ return true;
+}
+
+static void xgene_enet_wr_mac(struct xgene_enet_pdata *pdata,
+ u32 wr_addr, u32 wr_data)
+{
+ void __iomem *addr, *wr, *cmd, *cmd_done;
+
+ addr = pdata->mcx_mac_addr + MAC_ADDR_REG_OFFSET;
+ wr = pdata->mcx_mac_addr + MAC_WRITE_REG_OFFSET;
+ cmd = pdata->mcx_mac_addr + MAC_COMMAND_REG_OFFSET;
+ cmd_done = pdata->mcx_mac_addr + MAC_COMMAND_DONE_REG_OFFSET;
+
+ if (!xgene_enet_wr_indirect(addr, wr, cmd, cmd_done, wr_addr, wr_data))
+ netdev_err(pdata->ndev, "MCX mac write failed, addr: %04x\n",
+ wr_addr);
+}
+
+static void xgene_enet_rd_csr(struct xgene_enet_pdata *pdata,
+ u32 offset, u32 *val)
+{
+ void __iomem *addr = pdata->eth_csr_addr + offset;
+
+ *val = ioread32(addr);
+}
+
+static void xgene_enet_rd_diag_csr(struct xgene_enet_pdata *pdata,
+ u32 offset, u32 *val)
+{
+ void __iomem *addr = pdata->eth_diag_csr_addr + offset;
+
+ *val = ioread32(addr);
+}
+
+static bool xgene_enet_rd_indirect(void __iomem *addr, void __iomem *rd,
+ void __iomem *cmd, void __iomem *cmd_done,
+ u32 rd_addr, u32 *rd_data)
+{
+ u32 done;
+ u8 wait = 10;
+
+ iowrite32(rd_addr, addr);
+ iowrite32(XGENE_ENET_RD_CMD, cmd);
+
+ /* wait for read command to complete */
+ while (!(done = ioread32(cmd_done)) && wait--)
+ udelay(1);
+
+ if (!done)
+ return false;
+
+ *rd_data = ioread32(rd);
+ iowrite32(0, cmd);
+
+ return true;
+}
+
+static void xgene_enet_rd_mac(struct xgene_enet_pdata *pdata,
+ u32 rd_addr, u32 *rd_data)
+{
+ void __iomem *addr, *rd, *cmd, *cmd_done;
+
+ addr = pdata->mcx_mac_addr + MAC_ADDR_REG_OFFSET;
+ rd = pdata->mcx_mac_addr + MAC_READ_REG_OFFSET;
+ cmd = pdata->mcx_mac_addr + MAC_COMMAND_REG_OFFSET;
+ cmd_done = pdata->mcx_mac_addr + MAC_COMMAND_DONE_REG_OFFSET;
+
+ if (!xgene_enet_rd_indirect(addr, rd, cmd, cmd_done, rd_addr, rd_data))
+ netdev_err(pdata->ndev, "MCX mac read failed, addr: %04x\n",
+ rd_addr);
+}
+
+static int xgene_enet_ecc_init(struct xgene_enet_pdata *pdata)
+{
+ struct net_device *ndev = pdata->ndev;
+ u32 data;
+ u8 wait = 10;
+
+ xgene_enet_wr_diag_csr(pdata, ENET_CFG_MEM_RAM_SHUTDOWN_ADDR, 0x0);
+ do {
+ usleep_range(100, 110);
+ xgene_enet_rd_diag_csr(pdata, ENET_BLOCK_MEM_RDY_ADDR, &data);
+ } while ((data != 0xffffffff) && wait--);
+
+ if (data != 0xffffffff) {
+ netdev_err(ndev, "Failed to release memory from shutdown\n");
+ return -ENODEV;
+ }
+
+ return 0;
+}
+
+static void xgene_enet_config_ring_if_assoc(struct xgene_enet_pdata *pdata)
+{
+ xgene_enet_wr_ring_if(pdata, ENET_CFGSSQMIWQASSOC_ADDR, 0);
+ xgene_enet_wr_ring_if(pdata, ENET_CFGSSQMIFPQASSOC_ADDR, 0);
+ xgene_enet_wr_ring_if(pdata, ENET_CFGSSQMIQMLITEWQASSOC_ADDR, 0);
+ xgene_enet_wr_ring_if(pdata, ENET_CFGSSQMIQMLITEFPQASSOC_ADDR, 0);
+}
+
+static void xgene_xgmac_reset(struct xgene_enet_pdata *pdata)
+{
+ xgene_enet_wr_mac(pdata, AXGMAC_CONFIG_0, HSTMACRST);
+ xgene_enet_wr_mac(pdata, AXGMAC_CONFIG_0, 0);
+}
+
+static void xgene_xgmac_set_mac_addr(struct xgene_enet_pdata *pdata)
+{
+ u32 addr0, addr1;
+ u8 *dev_addr = pdata->ndev->dev_addr;
+
+ addr0 = (dev_addr[3] << 24) | (dev_addr[2] << 16) |
+ (dev_addr[1] << 8) | dev_addr[0];
+ addr1 = (dev_addr[5] << 24) | (dev_addr[4] << 16);
+
+ xgene_enet_wr_mac(pdata, HSTMACADR_LSW_ADDR, addr0);
+ xgene_enet_wr_mac(pdata, HSTMACADR_MSW_ADDR, addr1);
+}
+
+static u32 xgene_enet_link_status(struct xgene_enet_pdata *pdata)
+{
+ u32 data;
+
+ xgene_enet_rd_csr(pdata, XG_LINK_STATUS_ADDR, &data);
+
+ return data;
+}
+
+static void xgene_xgmac_init(struct xgene_enet_pdata *pdata)
+{
+ u32 data;
+
+ xgene_xgmac_reset(pdata);
+
+ xgene_enet_rd_mac(pdata, AXGMAC_CONFIG_1, &data);
+ data |= HSTPPEN;
+ data &= ~HSTLENCHK;
+ xgene_enet_wr_mac(pdata, AXGMAC_CONFIG_1, data);
+
+ xgene_enet_wr_mac(pdata, HSTMAXFRAME_LENGTH_ADDR, 0x06000600);
+ xgene_xgmac_set_mac_addr(pdata);
+
+ xgene_enet_rd_csr(pdata, XG_RSIF_CONFIG_REG_ADDR, &data);
+ data |= CFG_RSIF_FPBUFF_TIMEOUT_EN;
+ xgene_enet_wr_csr(pdata, XG_RSIF_CONFIG_REG_ADDR, data);
+
+ xgene_enet_wr_csr(pdata, XG_CFG_BYPASS_ADDR, RESUME_TX);
+ xgene_enet_wr_csr(pdata, XGENET_RX_DV_GATE_REG_0_ADDR, 0);
+ xgene_enet_rd_csr(pdata, XG_ENET_SPARE_CFG_REG_ADDR, &data);
+ data |= BIT(12);
+ xgene_enet_wr_csr(pdata, XG_ENET_SPARE_CFG_REG_ADDR, data);
+ xgene_enet_wr_csr(pdata, XG_ENET_SPARE_CFG_REG_1_ADDR, 0x82);
+}
+
+static void xgene_xgmac_rx_enable(struct xgene_enet_pdata *pdata)
+{
+ u32 data;
+
+ xgene_enet_rd_mac(pdata, AXGMAC_CONFIG_1, &data);
+ xgene_enet_wr_mac(pdata, AXGMAC_CONFIG_1, data | HSTRFEN);
+}
+
+static void xgene_xgmac_tx_enable(struct xgene_enet_pdata *pdata)
+{
+ u32 data;
+
+ xgene_enet_rd_mac(pdata, AXGMAC_CONFIG_1, &data);
+ xgene_enet_wr_mac(pdata, AXGMAC_CONFIG_1, data | HSTTFEN);
+}
+
+static void xgene_xgmac_rx_disable(struct xgene_enet_pdata *pdata)
+{
+ u32 data;
+
+ xgene_enet_rd_mac(pdata, AXGMAC_CONFIG_1, &data);
+ xgene_enet_wr_mac(pdata, AXGMAC_CONFIG_1, data & ~HSTRFEN);
+}
+
+static void xgene_xgmac_tx_disable(struct xgene_enet_pdata *pdata)
+{
+ u32 data;
+
+ xgene_enet_rd_mac(pdata, AXGMAC_CONFIG_1, &data);
+ xgene_enet_wr_mac(pdata, AXGMAC_CONFIG_1, data & ~HSTTFEN);
+}
+
+static void xgene_enet_reset(struct xgene_enet_pdata *pdata)
+{
+ clk_prepare_enable(pdata->clk);
+ clk_disable_unprepare(pdata->clk);
+ clk_prepare_enable(pdata->clk);
+
+ xgene_enet_ecc_init(pdata);
+ xgene_enet_config_ring_if_assoc(pdata);
+}
+
+static void xgene_enet_xgcle_bypass(struct xgene_enet_pdata *pdata,
+ u32 dst_ring_num, u16 bufpool_id)
+{
+ u32 cb, fpsel;
+
+ xgene_enet_rd_csr(pdata, XCLE_BYPASS_REG0_ADDR, &cb);
+ cb |= CFG_CLE_BYPASS_EN0;
+ CFG_CLE_IP_PROTOCOL0_SET(&cb, 3);
+ xgene_enet_wr_csr(pdata, XCLE_BYPASS_REG0_ADDR, cb);
+
+ fpsel = xgene_enet_ring_bufnum(bufpool_id) - 0x20;
+ xgene_enet_rd_csr(pdata, XCLE_BYPASS_REG1_ADDR, &cb);
+ CFG_CLE_DSTQID0_SET(&cb, dst_ring_num);
+ CFG_CLE_FPSEL0_SET(&cb, fpsel);
+ xgene_enet_wr_csr(pdata, XCLE_BYPASS_REG1_ADDR, cb);
+}
+
+static void xgene_enet_shutdown(struct xgene_enet_pdata *pdata)
+{
+ clk_disable_unprepare(pdata->clk);
+}
+
+void xgene_enet_link_state(struct work_struct *work)
+{
+ struct xgene_enet_pdata *pdata = container_of(to_delayed_work(work),
+ struct xgene_enet_pdata, link_work);
+ struct net_device *ndev = pdata->ndev;
+ u32 link_status, poll_interval;
+
+ link_status = xgene_enet_link_status(pdata);
+ if (link_status) {
+ if (!netif_carrier_ok(ndev)) {
+ netif_carrier_on(ndev);
+ xgene_xgmac_init(pdata);
+ xgene_xgmac_rx_enable(pdata);
+ xgene_xgmac_tx_enable(pdata);
+ netdev_info(ndev, "Link is Up - 10Gbps\n");
+ }
+ poll_interval = PHY_POLL_LINK_ON;
+ } else {
+ if (netif_carrier_ok(ndev)) {
+ xgene_xgmac_rx_disable(pdata);
+ xgene_xgmac_tx_disable(pdata);
+ netif_carrier_off(ndev);
+ netdev_info(ndev, "Link is Down\n");
+ }
+ poll_interval = PHY_POLL_LINK_OFF;
+ }
+
+ schedule_delayed_work(&pdata->link_work, poll_interval);
+}
+
+struct xgene_mac_ops xgene_xgmac_ops = {
+ .init = xgene_xgmac_init,
+ .reset = xgene_xgmac_reset,
+ .rx_enable = xgene_xgmac_rx_enable,
+ .tx_enable = xgene_xgmac_tx_enable,
+ .rx_disable = xgene_xgmac_rx_disable,
+ .tx_disable = xgene_xgmac_tx_disable,
+ .set_mac_addr = xgene_xgmac_set_mac_addr,
+};
+
+struct xgene_port_ops xgene_xgport_ops = {
+ .reset = xgene_enet_reset,
+ .cle_bypass = xgene_enet_xgcle_bypass,
+ .shutdown = xgene_enet_shutdown,
+};
diff --git a/drivers/net/ethernet/apm/xgene/xgene_enet_xgmac.h b/drivers/net/ethernet/apm/xgene/xgene_enet_xgmac.h
new file mode 100644
index 0000000..d2d59e7
--- /dev/null
+++ b/drivers/net/ethernet/apm/xgene/xgene_enet_xgmac.h
@@ -0,0 +1,57 @@
+/* Applied Micro X-Gene SoC Ethernet Driver
+ *
+ * Copyright (c) 2014, Applied Micro Circuits Corporation
+ * Authors: Iyappan Subramanian <isubramanian@apm.com>
+ * Keyur Chudgar <kchudgar@apm.com>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License as published by the
+ * Free Software Foundation; either version 2 of the License, or (at your
+ * option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program. If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#ifndef __XGENE_ENET_XGMAC_H__
+#define __XGENE_ENET_XGMAC_H__
+
+#define BLOCK_AXG_MAC_OFFSET 0x0800
+#define BLOCK_AXG_MAC_CSR_OFFSET 0x2000
+
+#define AXGMAC_CONFIG_0 0x0000
+#define AXGMAC_CONFIG_1 0x0004
+#define HSTMACRST BIT(31)
+#define HSTTCTLEN BIT(31)
+#define HSTTFEN BIT(30)
+#define HSTRCTLEN BIT(29)
+#define HSTRFEN BIT(28)
+#define HSTPPEN BIT(7)
+#define HSTDRPLT64 BIT(5)
+#define HSTLENCHK BIT(3)
+#define HSTMACADR_LSW_ADDR 0x0010
+#define HSTMACADR_MSW_ADDR 0x0014
+#define HSTMAXFRAME_LENGTH_ADDR 0x0020
+
+#define XG_RSIF_CONFIG_REG_ADDR 0x00a0
+#define XCLE_BYPASS_REG0_ADDR 0x0160
+#define XCLE_BYPASS_REG1_ADDR 0x0164
+#define XG_CFG_BYPASS_ADDR 0x0204
+#define XG_LINK_STATUS_ADDR 0x0228
+#define XG_ENET_SPARE_CFG_REG_ADDR 0x040c
+#define XG_ENET_SPARE_CFG_REG_1_ADDR 0x0410
+#define XGENET_RX_DV_GATE_REG_0_ADDR 0x0804
+
+#define PHY_POLL_LINK_ON (10 * HZ)
+#define PHY_POLL_LINK_OFF (PHY_POLL_LINK_ON / 5)
+
+void xgene_enet_link_state(struct work_struct *work);
+extern struct xgene_mac_ops xgene_xgmac_ops;
+extern struct xgene_port_ops xgene_xgport_ops;
+
+#endif /* __XGENE_ENET_XGMAC_H__ */
--
1.9.1
^ permalink raw reply related
* [PATCH v4 4/6] drivers: net: xgene: Preparing for adding 10GbE support
From: Iyappan Subramanian @ 2014-10-10 1:32 UTC (permalink / raw)
To: davem, netdev, devicetree
Cc: linux-arm-kernel, patches, kchudgar, Iyappan Subramanian
In-Reply-To: <1412904727-23485-1-git-send-email-isubramanian@apm.com>
- Rearranged code to pave the way for adding 10GbE support
- Added mac_ops structure containing function pointers for mac specific functions
- Added port_ops structure containing function pointers for port specific functions
Signed-off-by: Iyappan Subramanian <isubramanian@apm.com>
Signed-off-by: Keyur Chudgar <kchudgar@apm.com>
---
drivers/net/ethernet/apm/xgene/xgene_enet_hw.c | 44 ++++++++++++++++--------
drivers/net/ethernet/apm/xgene/xgene_enet_hw.h | 16 ++-------
drivers/net/ethernet/apm/xgene/xgene_enet_main.c | 40 +++++++++++++--------
drivers/net/ethernet/apm/xgene/xgene_enet_main.h | 19 ++++++++++
4 files changed, 78 insertions(+), 41 deletions(-)
diff --git a/drivers/net/ethernet/apm/xgene/xgene_enet_hw.c b/drivers/net/ethernet/apm/xgene/xgene_enet_hw.c
index 812d8d6..c8f3824 100644
--- a/drivers/net/ethernet/apm/xgene/xgene_enet_hw.c
+++ b/drivers/net/ethernet/apm/xgene/xgene_enet_hw.c
@@ -402,7 +402,7 @@ static int xgene_mii_phy_read(struct xgene_enet_pdata *pdata,
return data;
}
-void xgene_gmac_set_mac_addr(struct xgene_enet_pdata *pdata)
+static void xgene_gmac_set_mac_addr(struct xgene_enet_pdata *pdata)
{
u32 addr0, addr1;
u8 *dev_addr = pdata->ndev->dev_addr;
@@ -436,13 +436,13 @@ static int xgene_enet_ecc_init(struct xgene_enet_pdata *pdata)
return 0;
}
-void xgene_gmac_reset(struct xgene_enet_pdata *pdata)
+static void xgene_gmac_reset(struct xgene_enet_pdata *pdata)
{
xgene_enet_wr_mcx_mac(pdata, MAC_CONFIG_1_ADDR, SOFT_RESET1);
xgene_enet_wr_mcx_mac(pdata, MAC_CONFIG_1_ADDR, 0);
}
-void xgene_gmac_init(struct xgene_enet_pdata *pdata, int speed)
+static void xgene_gmac_init(struct xgene_enet_pdata *pdata)
{
u32 value, mc2;
u32 intf_ctl, rgmii;
@@ -456,7 +456,7 @@ void xgene_gmac_init(struct xgene_enet_pdata *pdata, int speed)
xgene_enet_rd_mcx_mac(pdata, INTERFACE_CONTROL_ADDR, &intf_ctl);
xgene_enet_rd_csr(pdata, RGMII_REG_0_ADDR, &rgmii);
- switch (speed) {
+ switch (pdata->phy_speed) {
case SPEED_10:
ENET_INTERFACE_MODE2_SET(&mc2, 1);
CFG_MACMODE_SET(&icm0, 0);
@@ -525,8 +525,8 @@ static void xgene_enet_config_ring_if_assoc(struct xgene_enet_pdata *pdata)
xgene_enet_wr_ring_if(pdata, ENET_CFGSSQMIQMLITEFPQASSOC_ADDR, val);
}
-void xgene_enet_cle_bypass(struct xgene_enet_pdata *pdata,
- u32 dst_ring_num, u16 bufpool_id)
+static void xgene_enet_cle_bypass(struct xgene_enet_pdata *pdata,
+ u32 dst_ring_num, u16 bufpool_id)
{
u32 cb;
u32 fpsel;
@@ -544,7 +544,7 @@ void xgene_enet_cle_bypass(struct xgene_enet_pdata *pdata,
xgene_enet_wr_csr(pdata, CLE_BYPASS_REG1_0_ADDR, cb);
}
-void xgene_gmac_rx_enable(struct xgene_enet_pdata *pdata)
+static void xgene_gmac_rx_enable(struct xgene_enet_pdata *pdata)
{
u32 data;
@@ -552,7 +552,7 @@ void xgene_gmac_rx_enable(struct xgene_enet_pdata *pdata)
xgene_enet_wr_mcx_mac(pdata, MAC_CONFIG_1_ADDR, data | RX_EN);
}
-void xgene_gmac_tx_enable(struct xgene_enet_pdata *pdata)
+static void xgene_gmac_tx_enable(struct xgene_enet_pdata *pdata)
{
u32 data;
@@ -560,7 +560,7 @@ void xgene_gmac_tx_enable(struct xgene_enet_pdata *pdata)
xgene_enet_wr_mcx_mac(pdata, MAC_CONFIG_1_ADDR, data | TX_EN);
}
-void xgene_gmac_rx_disable(struct xgene_enet_pdata *pdata)
+static void xgene_gmac_rx_disable(struct xgene_enet_pdata *pdata)
{
u32 data;
@@ -568,7 +568,7 @@ void xgene_gmac_rx_disable(struct xgene_enet_pdata *pdata)
xgene_enet_wr_mcx_mac(pdata, MAC_CONFIG_1_ADDR, data & ~RX_EN);
}
-void xgene_gmac_tx_disable(struct xgene_enet_pdata *pdata)
+static void xgene_gmac_tx_disable(struct xgene_enet_pdata *pdata)
{
u32 data;
@@ -576,7 +576,7 @@ void xgene_gmac_tx_disable(struct xgene_enet_pdata *pdata)
xgene_enet_wr_mcx_mac(pdata, MAC_CONFIG_1_ADDR, data & ~TX_EN);
}
-void xgene_enet_reset(struct xgene_enet_pdata *pdata)
+static void xgene_enet_reset(struct xgene_enet_pdata *pdata)
{
u32 val;
@@ -593,7 +593,7 @@ void xgene_enet_reset(struct xgene_enet_pdata *pdata)
xgene_enet_wr_mcx_mac(pdata, MII_MGMT_CONFIG_ADDR, val);
}
-void xgene_gport_shutdown(struct xgene_enet_pdata *pdata)
+static void xgene_gport_shutdown(struct xgene_enet_pdata *pdata)
{
clk_disable_unprepare(pdata->clk);
}
@@ -627,10 +627,10 @@ static void xgene_enet_adjust_link(struct net_device *ndev)
if (phydev->link) {
if (pdata->phy_speed != phydev->speed) {
- xgene_gmac_init(pdata, phydev->speed);
+ pdata->phy_speed = phydev->speed;
+ xgene_gmac_init(pdata);
xgene_gmac_rx_enable(pdata);
xgene_gmac_tx_enable(pdata);
- pdata->phy_speed = phydev->speed;
phy_print_status(phydev);
}
} else {
@@ -726,3 +726,19 @@ void xgene_enet_mdio_remove(struct xgene_enet_pdata *pdata)
mdiobus_free(pdata->mdio_bus);
pdata->mdio_bus = NULL;
}
+
+struct xgene_mac_ops xgene_gmac_ops = {
+ .init = xgene_gmac_init,
+ .reset = xgene_gmac_reset,
+ .rx_enable = xgene_gmac_rx_enable,
+ .tx_enable = xgene_gmac_tx_enable,
+ .rx_disable = xgene_gmac_rx_disable,
+ .tx_disable = xgene_gmac_tx_disable,
+ .set_mac_addr = xgene_gmac_set_mac_addr,
+};
+
+struct xgene_port_ops xgene_gport_ops = {
+ .reset = xgene_enet_reset,
+ .cle_bypass = xgene_enet_cle_bypass,
+ .shutdown = xgene_gport_shutdown,
+};
diff --git a/drivers/net/ethernet/apm/xgene/xgene_enet_hw.h b/drivers/net/ethernet/apm/xgene/xgene_enet_hw.h
index 371e7a5..084ac68 100644
--- a/drivers/net/ethernet/apm/xgene/xgene_enet_hw.h
+++ b/drivers/net/ethernet/apm/xgene/xgene_enet_hw.h
@@ -318,20 +318,10 @@ void xgene_enet_parse_error(struct xgene_enet_desc_ring *ring,
struct xgene_enet_pdata *pdata,
enum xgene_enet_err_code status);
-void xgene_enet_reset(struct xgene_enet_pdata *priv);
-void xgene_gmac_reset(struct xgene_enet_pdata *priv);
-void xgene_gmac_init(struct xgene_enet_pdata *priv, int speed);
-void xgene_gmac_tx_enable(struct xgene_enet_pdata *priv);
-void xgene_gmac_rx_enable(struct xgene_enet_pdata *priv);
-void xgene_gmac_tx_disable(struct xgene_enet_pdata *priv);
-void xgene_gmac_rx_disable(struct xgene_enet_pdata *priv);
-void xgene_gmac_set_mac_addr(struct xgene_enet_pdata *pdata);
-void xgene_enet_cle_bypass(struct xgene_enet_pdata *pdata,
- u32 dst_ring_num, u16 bufpool_id);
-void xgene_gport_shutdown(struct xgene_enet_pdata *priv);
-void xgene_gmac_get_tx_stats(struct xgene_enet_pdata *pdata);
-
int xgene_enet_mdio_config(struct xgene_enet_pdata *pdata);
void xgene_enet_mdio_remove(struct xgene_enet_pdata *pdata);
+extern struct xgene_mac_ops xgene_gmac_ops;
+extern struct xgene_port_ops xgene_gport_ops;
+
#endif /* __XGENE_ENET_HW_H__ */
diff --git a/drivers/net/ethernet/apm/xgene/xgene_enet_main.c b/drivers/net/ethernet/apm/xgene/xgene_enet_main.c
index e4222af..c432644 100644
--- a/drivers/net/ethernet/apm/xgene/xgene_enet_main.c
+++ b/drivers/net/ethernet/apm/xgene/xgene_enet_main.c
@@ -413,7 +413,7 @@ static void xgene_enet_timeout(struct net_device *ndev)
{
struct xgene_enet_pdata *pdata = netdev_priv(ndev);
- xgene_gmac_reset(pdata);
+ pdata->mac_ops->reset(pdata);
}
static int xgene_enet_register_irq(struct net_device *ndev)
@@ -445,10 +445,11 @@ static void xgene_enet_free_irq(struct net_device *ndev)
static int xgene_enet_open(struct net_device *ndev)
{
struct xgene_enet_pdata *pdata = netdev_priv(ndev);
+ struct xgene_mac_ops *mac_ops = pdata->mac_ops;
int ret;
- xgene_gmac_tx_enable(pdata);
- xgene_gmac_rx_enable(pdata);
+ mac_ops->tx_enable(pdata);
+ mac_ops->rx_enable(pdata);
ret = xgene_enet_register_irq(ndev);
if (ret)
@@ -466,6 +467,7 @@ static int xgene_enet_open(struct net_device *ndev)
static int xgene_enet_close(struct net_device *ndev)
{
struct xgene_enet_pdata *pdata = netdev_priv(ndev);
+ struct xgene_mac_ops *mac_ops = pdata->mac_ops;
netif_stop_queue(ndev);
@@ -476,8 +478,8 @@ static int xgene_enet_close(struct net_device *ndev)
xgene_enet_free_irq(ndev);
xgene_enet_process_ring(pdata->rx_ring, -1);
- xgene_gmac_tx_disable(pdata);
- xgene_gmac_rx_disable(pdata);
+ mac_ops->tx_disable(pdata);
+ mac_ops->rx_disable(pdata);
return 0;
}
@@ -724,7 +726,7 @@ static int xgene_enet_set_mac_address(struct net_device *ndev, void *addr)
ret = eth_mac_addr(ndev, addr);
if (ret)
return ret;
- xgene_gmac_set_mac_addr(pdata);
+ pdata->mac_ops->set_mac_addr(pdata);
return ret;
}
@@ -834,8 +836,8 @@ static int xgene_enet_init_hw(struct xgene_enet_pdata *pdata)
u16 dst_ring_num;
int ret;
- xgene_gmac_tx_disable(pdata);
- xgene_gmac_rx_disable(pdata);
+ pdata->mac_ops->tx_disable(pdata);
+ pdata->mac_ops->rx_disable(pdata);
ret = xgene_enet_create_desc_rings(ndev);
if (ret) {
@@ -853,11 +855,17 @@ static int xgene_enet_init_hw(struct xgene_enet_pdata *pdata)
}
dst_ring_num = xgene_enet_dst_ring_num(pdata->rx_ring);
- xgene_enet_cle_bypass(pdata, dst_ring_num, buf_pool->id);
+ pdata->port_ops->cle_bypass(pdata, dst_ring_num, buf_pool->id);
return ret;
}
+static void xgene_enet_setup_ops(struct xgene_enet_pdata *pdata)
+{
+ pdata->mac_ops = &xgene_gmac_ops;
+ pdata->port_ops = &xgene_gport_ops;
+}
+
static int xgene_enet_probe(struct platform_device *pdev)
{
struct net_device *ndev;
@@ -886,8 +894,9 @@ static int xgene_enet_probe(struct platform_device *pdev)
if (ret)
goto err;
- xgene_enet_reset(pdata);
- xgene_gmac_init(pdata, SPEED_1000);
+ xgene_enet_setup_ops(pdata);
+ pdata->port_ops->reset(pdata);
+ pdata->mac_ops->init(pdata);
ret = register_netdev(ndev);
if (ret) {
@@ -918,19 +927,21 @@ err:
static int xgene_enet_remove(struct platform_device *pdev)
{
struct xgene_enet_pdata *pdata;
+ struct xgene_mac_ops *mac_ops;
struct net_device *ndev;
pdata = platform_get_drvdata(pdev);
+ mac_ops = pdata->mac_ops;
ndev = pdata->ndev;
- xgene_gmac_rx_disable(pdata);
- xgene_gmac_tx_disable(pdata);
+ mac_ops->rx_disable(pdata);
+ mac_ops->tx_disable(pdata);
netif_napi_del(&pdata->rx_ring->napi);
xgene_enet_mdio_remove(pdata);
xgene_enet_delete_desc_rings(pdata);
unregister_netdev(ndev);
- xgene_gport_shutdown(pdata);
+ pdata->port_ops->shutdown(pdata);
free_netdev(ndev);
return 0;
@@ -956,5 +967,6 @@ module_platform_driver(xgene_enet_driver);
MODULE_DESCRIPTION("APM X-Gene SoC Ethernet driver");
MODULE_VERSION(XGENE_DRV_VERSION);
+MODULE_AUTHOR("Iyappan Subramanian <isubramanian@apm.com>");
MODULE_AUTHOR("Keyur Chudgar <kchudgar@apm.com>");
MODULE_LICENSE("GPL");
diff --git a/drivers/net/ethernet/apm/xgene/xgene_enet_main.h b/drivers/net/ethernet/apm/xgene/xgene_enet_main.h
index 0815866..ac180f9 100644
--- a/drivers/net/ethernet/apm/xgene/xgene_enet_main.h
+++ b/drivers/net/ethernet/apm/xgene/xgene_enet_main.h
@@ -68,6 +68,23 @@ struct xgene_enet_desc_ring {
};
};
+struct xgene_mac_ops {
+ void (*init)(struct xgene_enet_pdata *pdata);
+ void (*reset)(struct xgene_enet_pdata *pdata);
+ void (*tx_enable)(struct xgene_enet_pdata *pdata);
+ void (*rx_enable)(struct xgene_enet_pdata *pdata);
+ void (*tx_disable)(struct xgene_enet_pdata *pdata);
+ void (*rx_disable)(struct xgene_enet_pdata *pdata);
+ void (*set_mac_addr)(struct xgene_enet_pdata *pdata);
+};
+
+struct xgene_port_ops {
+ void (*reset)(struct xgene_enet_pdata *pdata);
+ void (*cle_bypass)(struct xgene_enet_pdata *pdata,
+ u32 dst_ring_num, u16 bufpool_id);
+ void (*shutdown)(struct xgene_enet_pdata *pdata);
+};
+
/* ethernet private data */
struct xgene_enet_pdata {
struct net_device *ndev;
@@ -98,6 +115,8 @@ struct xgene_enet_pdata {
u32 speed;
u16 rm;
struct rtnl_link_stats64 stats;
+ struct xgene_mac_ops *mac_ops;
+ struct xgene_port_ops *port_ops;
};
/* Set the specified value into a bit-field defined by its starting position
--
1.9.1
^ permalink raw reply related
* [PATCH v4 3/6] dtb: Add 10GbE node to APM X-Gene SoC device tree
From: Iyappan Subramanian @ 2014-10-10 1:32 UTC (permalink / raw)
To: davem, netdev, devicetree
Cc: linux-arm-kernel, patches, kchudgar, Iyappan Subramanian
In-Reply-To: <1412904727-23485-1-git-send-email-isubramanian@apm.com>
Added 10GbE interface and clock nodes.
Signed-off-by: Iyappan Subramanian <isubramanian@apm.com>
Signed-off-by: Keyur Chudgar <kchudgar@apm.com>
---
arch/arm64/boot/dts/apm-mustang.dts | 4 ++++
arch/arm64/boot/dts/apm-storm.dtsi | 29 +++++++++++++++++++++++++++--
2 files changed, 31 insertions(+), 2 deletions(-)
diff --git a/arch/arm64/boot/dts/apm-mustang.dts b/arch/arm64/boot/dts/apm-mustang.dts
index b2f5622..2ae782b 100644
--- a/arch/arm64/boot/dts/apm-mustang.dts
+++ b/arch/arm64/boot/dts/apm-mustang.dts
@@ -32,3 +32,7 @@
&menet {
status = "ok";
};
+
+&xgenet {
+ status = "ok";
+};
diff --git a/arch/arm64/boot/dts/apm-storm.dtsi b/arch/arm64/boot/dts/apm-storm.dtsi
index f391972..d16cc03 100644
--- a/arch/arm64/boot/dts/apm-storm.dtsi
+++ b/arch/arm64/boot/dts/apm-storm.dtsi
@@ -176,6 +176,16 @@
clock-output-names = "menetclk";
};
+ xge0clk: xge0clk@1f61c000 {
+ compatible = "apm,xgene-device-clock";
+ #clock-cells = <1>;
+ clocks = <&socplldiv2 0>;
+ reg = <0x0 0x1f61c000 0x0 0x1000>;
+ reg-names = "csr-reg";
+ csr-mask = <0x3>;
+ clock-output-names = "xge0clk";
+ };
+
sataphy1clk: sataphy1clk@1f21c000 {
compatible = "apm,xgene-device-clock";
#clock-cells = <1>;
@@ -420,7 +430,8 @@
interrupts = <0x0 0x3c 0x4>;
dma-coherent;
clocks = <&menetclk 0>;
- local-mac-address = [00 01 73 00 00 01];
+ /* mac address will be overwritten by the bootloader */
+ local-mac-address = [00 00 00 00 00 00];
phy-connection-type = "rgmii";
phy-handle = <&menetphy>;
mdio {
@@ -435,12 +446,26 @@
};
};
+ xgenet: ethernet@1f610000 {
+ compatible = "apm,xgene-enet";
+ status = "disabled";
+ reg = <0x0 0x1f610000 0x0 0xd100>,
+ <0x0 0x1f600000 0x0 0X400>,
+ <0x0 0x18000000 0x0 0X200>;
+ reg-names = "enet_csr", "ring_csr", "ring_cmd";
+ interrupts = <0x0 0x60 0x4>;
+ dma-coherent;
+ clocks = <&xge0clk 0>;
+ /* mac address will be overwritten by the bootloader */
+ local-mac-address = [00 00 00 00 00 00];
+ phy-connection-type = "xgmii";
+ };
+
rng: rng@10520000 {
compatible = "apm,xgene-rng";
reg = <0x0 0x10520000 0x0 0x100>;
interrupts = <0x0 0x41 0x4>;
clocks = <&rngpkaclk 0>;
};
-
};
};
--
1.9.1
^ permalink raw reply related
* [PATCH v4 1/6] MAINTAINERS: Update APM X-Gene section
From: Iyappan Subramanian @ 2014-10-10 1:32 UTC (permalink / raw)
To: davem-fT/PcQaiUtIeIZ0/mPfg9Q, netdev-u79uwXL29TY76Z2rM5mHXA,
devicetree-u79uwXL29TY76Z2rM5mHXA
Cc: linux-arm-kernel-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r,
patches-qTEPVZfXA3Y, kchudgar-qTEPVZfXA3Y, Iyappan Subramanian
In-Reply-To: <1412904727-23485-1-git-send-email-isubramanian-qTEPVZfXA3Y@public.gmane.org>
Updated APM X-Gene ethernet driver maintainers list.
Signed-off-by: Iyappan Subramanian <isubramanian-qTEPVZfXA3Y@public.gmane.org>
Signed-off-by: Keyur Chudgar <kchudgar-qTEPVZfXA3Y@public.gmane.org>
---
MAINTAINERS | 1 -
1 file changed, 1 deletion(-)
diff --git a/MAINTAINERS b/MAINTAINERS
index 9e315a4..1a85af8 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -734,7 +734,6 @@ F: net/appletalk/
APPLIED MICRO (APM) X-GENE SOC ETHERNET DRIVER
M: Iyappan Subramanian <isubramanian-qTEPVZfXA3Y@public.gmane.org>
M: Keyur Chudgar <kchudgar-qTEPVZfXA3Y@public.gmane.org>
-M: Ravi Patel <rapatel-qTEPVZfXA3Y@public.gmane.org>
S: Supported
F: drivers/net/ethernet/apm/xgene/
F: Documentation/devicetree/bindings/net/apm-xgene-enet.txt
--
1.9.1
--
To unsubscribe from this list: send the line "unsubscribe devicetree" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply related
* [PATCH v4 2/6] Documentation: dts: Update section header for APM X-Gene
From: Iyappan Subramanian @ 2014-10-10 1:32 UTC (permalink / raw)
To: davem, netdev, devicetree
Cc: linux-arm-kernel, patches, kchudgar, Iyappan Subramanian
In-Reply-To: <1412904727-23485-1-git-send-email-isubramanian@apm.com>
Signed-off-by: Iyappan Subramanian <isubramanian@apm.com>
Signed-off-by: Keyur Chudgar <kchudgar@apm.com>
---
Documentation/devicetree/bindings/net/apm-xgene-enet.txt | 4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)
diff --git a/Documentation/devicetree/bindings/net/apm-xgene-enet.txt b/Documentation/devicetree/bindings/net/apm-xgene-enet.txt
index ebcad25..cfcc527 100644
--- a/Documentation/devicetree/bindings/net/apm-xgene-enet.txt
+++ b/Documentation/devicetree/bindings/net/apm-xgene-enet.txt
@@ -3,7 +3,7 @@ APM X-Gene SoC Ethernet nodes
Ethernet nodes are defined to describe on-chip ethernet interfaces in
APM X-Gene SoC.
-Required properties:
+Required properties for all the ethernet interfaces:
- compatible: Should be "apm,xgene-enet"
- reg: Address and length of the register set for the device. It contains the
information of registers in the same order as described by reg-names
@@ -15,6 +15,8 @@ Required properties:
- clocks: Reference to the clock entry.
- local-mac-address: MAC address assigned to this device
- phy-connection-type: Interface type between ethernet device and PHY device
+
+Required properties for ethernet interfaces that have external PHY:
- phy-handle: Reference to a PHY node connected to this device
- mdio: Device tree subnode with the following required properties:
--
1.9.1
^ permalink raw reply related
* [PATCH v4 0/6] Add 10GbE support to APM X-Gene SoC ethernet driver
From: Iyappan Subramanian @ 2014-10-10 1:32 UTC (permalink / raw)
To: davem-fT/PcQaiUtIeIZ0/mPfg9Q, netdev-u79uwXL29TY76Z2rM5mHXA,
devicetree-u79uwXL29TY76Z2rM5mHXA
Cc: linux-arm-kernel-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r,
patches-qTEPVZfXA3Y, kchudgar-qTEPVZfXA3Y, Iyappan Subramanian
Adding 10GbE support to APM X-Gene SoC ethernet driver.
v4: Address comments from v3
* dtb: resolved merge conflict for the net tree
v3: Address comments from v2
* dtb: changed to use all-zeros for the mac address
v2: Address comments from v1
* created preparatory patch to review before adding new functionality
* dtb: updated to use tabs consistently
v1:
* Initial version
---
Iyappan Subramanian (6):
MAINTAINERS: Update APM X-Gene section
Documentation: dts: Update section header for APM X-Gene
dtb: Add 10GbE node to APM X-Gene SoC device tree
drivers: net: xgene: Preparing for adding 10GbE support
drivers: net: xgene: Add 10GbE support
drivers: net: xgene: Add 10GbE ethtool support
.../devicetree/bindings/net/apm-xgene-enet.txt | 4 +-
MAINTAINERS | 1 -
arch/arm64/boot/dts/apm-mustang.dts | 4 +
arch/arm64/boot/dts/apm-storm.dtsi | 29 +-
drivers/net/ethernet/apm/xgene/Makefile | 3 +-
.../net/ethernet/apm/xgene/xgene_enet_ethtool.c | 28 +-
drivers/net/ethernet/apm/xgene/xgene_enet_hw.c | 44 ++-
drivers/net/ethernet/apm/xgene/xgene_enet_hw.h | 30 +-
drivers/net/ethernet/apm/xgene/xgene_enet_main.c | 86 ++++--
drivers/net/ethernet/apm/xgene/xgene_enet_main.h | 24 +-
drivers/net/ethernet/apm/xgene/xgene_enet_xgmac.c | 331 +++++++++++++++++++++
drivers/net/ethernet/apm/xgene/xgene_enet_xgmac.h | 57 ++++
12 files changed, 566 insertions(+), 75 deletions(-)
create mode 100644 drivers/net/ethernet/apm/xgene/xgene_enet_xgmac.c
create mode 100644 drivers/net/ethernet/apm/xgene/xgene_enet_xgmac.h
--
1.9.1
--
To unsubscribe from this list: send the line "unsubscribe devicetree" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply
* Re: [PATCH net-next 0/2] sunvnet: Packet processing in non-interrupt context.
From: Raghuram Kothakota @ 2014-10-10 1:10 UTC (permalink / raw)
To: Sowmini Varadhan; +Cc: David Miller, netdev
In-Reply-To: <20141006193111.GE24721@oracle.com>
I would like to share my experience, on sparc we need more parallelism to
get the high performance that's just the nature of CMT processors(at least today).
Lock less Tx and Rx implementation is very nice, but requires to the code path to
single threaded to achieve it. This may limit the performance to the single thread
performance, that may be not very high for standard MTU packets. The large MTU packets
like 64K MTU may have some advantage due to less processing both in the stack
and the driver, but still single thread would limit the max performance. I would suggest
explore both lock less and high parallelism methods to see which one gives the
best performance.
-Raghuram
On Oct 6, 2014, at 12:31 PM, Sowmini Varadhan <sowmini.varadhan@oracle.com> wrote:
> On (10/06/14 15:25), David Miller wrote:
>>
>>> But we still need to hold the vio lock around the ldc_write
>>> (and also around dring write) in vnet_start_xmit, right?
>>
>> You might be able to avoid it, you're fully serialized by the TX queue
>> lock.
>
> yes, I was just noticing that. The only place where I believe I need
> to hold the vio spin-lock is to sync with the dr->cons checks
> (the "should I send a start_cons LDC message?" check in vnet_start_xmit()
> vs the vnet_ack() updates).
>
> But isn't it better in general to declare NETIF_F_LLTX and have finer lock
> granularity in the driver?
>
> --Sowmini
>
> --
> To unsubscribe from this list: send the line "unsubscribe netdev" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply
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