* Re: [PATCH v4 20/20] kbuild: add a new kselftest_install make target to install selftests
From: Shuah Khan @ 2015-01-15 23:41 UTC (permalink / raw)
To: Michael Ellerman
Cc: mmarek-AlSwsSmVLrQ, masami.hiramatsu.pt-FCd8Q96Dh0JBDgjK7y7TUQ,
gregkh-hQyY1W1yCW8ekmWlsbkhG0B+6BGkLq7r,
akpm-de/tnXTf+JLsfHDXvbKv3WD2FQJk+8+b,
rostedt-nx8X9YLhiw1AfugRpC6u6w, mingo-H+wXaHxf7aLQT0dZR+AlfA,
davem-fT/PcQaiUtIeIZ0/mPfg9Q, keescook-F7+t8E8rja9g9hUCZPvPmw,
tranmanphong-Re5JQEeQqe8AvxtiuMwx3w, cov-sgV2jX0FEOL9JmXXK+q4OQ,
dh.herrmann-Re5JQEeQqe8AvxtiuMwx3w, hughd-hpIqsD4AKlfQT0dZR+AlfA,
bobby.prani-Re5JQEeQqe8AvxtiuMwx3w,
serge.hallyn-GeWIH/nMZzLQT0dZR+AlfA,
ebiederm-aS9lmoZGLiVWk0Htik3J/w, tim.bird-/MT0OVThwyLZJqsBc5GL+g,
josh-iaAMLnmF4UmaiuxdJuQwMA, koct9i-Re5JQEeQqe8AvxtiuMwx3w,
linux-kbuild-u79uwXL29TY76Z2rM5mHXA,
linux-kernel-u79uwXL29TY76Z2rM5mHXA,
linux-api-u79uwXL29TY76Z2rM5mHXA, netdev-u79uwXL29TY76Z2rM5mHXA
In-Reply-To: <1421362698.23332.6.camel-Gsx/Oe8HsFggBc27wqDAHg@public.gmane.org>
On 01/15/2015 03:58 PM, Michael Ellerman wrote:
> On Wed, 2015-01-14 at 09:32 -0700, Shuah Khan wrote:
>> On 01/06/2015 12:43 PM, Shuah Khan wrote:
>>> Add a new make target to install to install kernel selftests.
>>> This new target will build and install selftests. kselftest
>>> target now depends on kselftest_install and runs the generated
>>> kselftest script to reduce duplicate work and for common look
>>> and feel when running tests.
>>>
>>> make kselftest_target:
>>> -- exports kselftest INSTALL_KSFT_PATH
>>> default $(INSTALL_MOD_PATH)/lib/kselftest/$(KERNELRELEASE)
>>> -- exports INSTALL_KSFT_PATH
>>> -- runs selftests make install target
>>>
>>> Signed-off-by: Shuah Khan <shuahkh-JPH+aEBZ4P+UEJcrhfAQsw@public.gmane.org>
>>
>> Hi Marek,
>>
>> Could you please Ack this patch, if this version looks good,
>> so I can take this through ksefltest tree.
>
> Hi Shuah,
>
> I posted what I think is a better approach last week. It's less code and less
> repetition, and installs the majority of tests this series missed - ie. the
> powerpc ones.
>
> Please at least respond to it before you put this series into next.
>
> https://lkml.org/lkml/2015/1/9/45
>
Michael,
I didn't get a chance to review your patches yet. I plan to
review and give you feedback by end of this week. This patch
is needed in any case for either approach.
thanks,
-- Shuah
--
Shuah Khan
Sr. Linux Kernel Developer
Open Source Innovation Group
Samsung Research America (Silicon Valley)
shuahkh-JPH+aEBZ4P+UEJcrhfAQsw@public.gmane.org | (970) 217-8978
^ permalink raw reply
* Re: [PATCH] i40e: don't enable and init FCOE by default when do PF reset
From: Dev, Vasu @ 2015-01-15 23:45 UTC (permalink / raw)
To: ethan zhao
Cc: e1000-devel@lists.sourceforge.net, brian.maly@oracle.com,
Allan, Bruce W, Brandeburg, Jesse, Parikh, Neerav, Linux NICS,
Ronciak, John, netdev@vger.kernel.org, Ethan Zhao,
linux-kernel@vger.kernel.org
> -----Original Message-----
> From: ethan zhao [mailto:ethan.zhao@oracle.com]
> Sent: Tuesday, January 13, 2015 6:41 PM
> To: Dev, Vasu
> Cc: Ethan Zhao; Ronciak, John; Kirsher, Jeffrey T; Brandeburg, Jesse; Allan,
> Bruce W; Wyborny, Carolyn; Skidmore, Donald C; Rose, Gregory V; Vick,
> Matthew; Williams, Mitch A; Parikh, Neerav; Linux NICS; e1000-
> devel@lists.sourceforge.net; netdev@vger.kernel.org; linux-
> kernel@vger.kernel.org; brian.maly@oracle.com
> Subject: Re: [PATCH] i40e: don't enable and init FCOE by default when do PF
> reset
>
> Vasu,
>
> On 2015/1/14 3:38, Dev, Vasu wrote:
> >> -----Original Message-----
> >>>>> diff --git a/drivers/net/ethernet/intel/i40e/i40e_main.c
> >>>>> b/drivers/net/ethernet/intel/i40e/i40e_main.c
> >>>>> index a5f2660..a2572cc 100644
> >>>>> --- a/drivers/net/ethernet/intel/i40e/i40e_main.c
> >>>>> +++ b/drivers/net/ethernet/intel/i40e/i40e_main.c
> >>>>> @@ -6180,9 +6180,12 @@ static void i40e_reset_and_rebuild(struct
> >>>>> i40e_pf *pf, bool reinit)
> >>>>> }
> >>>>> #endif /* CONFIG_I40E_DCB */
> >>>>> #ifdef I40E_FCOE
> >>>>> - ret = i40e_init_pf_fcoe(pf);
> >>>>> - if (ret)
> >>>>> - dev_info(&pf->pdev->dev, "init_pf_fcoe failed: %d\n", ret);
> >>>>> + if (pf->flags & I40E_FLAG_FCOE_ENABLED) {
> >>>>> + ret = i40e_init_pf_fcoe(pf);
> >>> Calling i40e_init_pf_fcoe() here conflicts with its
> >> I40E_FLAG_FCOE_ENABLED pre-condition since
> I40E_FLAG_FCOE_ENABLED is
> >> set by very same i40e_init_pf_fcoe(), in turn i40e_init_pf_fcoe()
> >> will never get called.
> >>
> >> I don't think so, here ,i40e_reset_and_rebuild() is not the only
> >> and the first place that i40e_init_pf_fcoe() is called, see
> >> i40e_probe(), that is the first chance.
> >>
> >> i40e_probe()
> >> -->i40e_sw_init()
> >> -->i40e_init_pf_fcoe()
> >>
> >> And the I40E_FLAG_FCOE_ENABLED is possible be set by
> >> i40e_fcoe_enable() or i40e_fcoe_disable() interface before the reset
> >> action is to be done.
> >>
> > It is set by i40e_init_pf_fcoe() and you are right that the modified call flow
> by your patch won't impact setting of I40E_FLAG_FCOE_ENABLED anyway
> which could have prevented calling i40e_init_pf_fcoe() as I described above,
> so this is not an issue with the patch.
> >
> >> BTW, the reason I post this patch is that we hit a bug, after setup
> >> vlan, the PF is enabled to FCOE.
> >>
> > Then that BUG would still remain un-fixed and calling i40e_init_pf_fcoe()
> under I40E_FLAG_FCOE_ENABLED flag really won't affect call flow to fix
> anything. I mean I40E_FLAG_FCOE_ENABLED condition will be true with "pf-
> >hw.func_caps.fcoe == true" and otherwise calling i40e_init_pf_fcoe() simply
> returns back early on after checking "pf->hw.func_caps.fcoe == false", so
> how that bug is fixed here by added I40E_FLAG_FCOE_ENABLED condition ?
> What is the bug ?
> The func_caps.fcoe is assigned by following call path, under our test
> environment,
>
> i40e_probe()
> ->i40e_get_capabilities()
> ->i40e_aq_discover_capabilities()
> ->i40e_parse_discover_capabilities()
>
> Or
>
> i40e_reset_and_rebuild()
> ->i40e_get_capabilities()
> ->i40e_aq_discover_capabilities()
> ->i40e_parse_discover_capabilities()
>
> Under our test environment, the "pf->hw.func_caps.fcoe" is true. so if
> i40e_reset_and_rebuild() is called for VLAN setup, ethtool diagnostic test.
> And then i40e_init_pf_fcoe() is to be called,
>
> While if (!pf->hw.func_caps.fcoe) wouldn't return,
>
I said it would return with "pf->hw.func_caps.fcoe == false" in my last response, more details below.
> So pf->flags is set to I40E_FLAG_FCOE_ENABLED.
>
> With my patch, i40e_init_pf_fcoe() is only called after
> I40E_FLAG_FCOE_ENABLED is set before reset.
>
> Enable FCOE in i40e_probe() or not is another issue.
>
Nope since both cases we should do i40e_init_pf_fcoe() or don't based on fcoe cap true or false.
I don't have much to add as I described before with the your patch that "calling i40e_init_pf_fcoe() under I40E_FLAG_FCOE_ENABLED flag really won't affect call flow to fix anything. I mean I40E_FLAG_FCOE_ENABLED condition will be true with "pf->hw.func_caps.fcoe == true" and otherwise calling i40e_init_pf_fcoe() simply returns back early on after checking "pf->hw.func_caps.fcoe == false".
May be I'm missing something, I guess next either go with CONFIG_I40E_FCOE disable as I suggested before and now it in upstream kernel or we can have further off list discussion to fix the issue you are trying to fix with the patch.
Thanks,
Vasu
> Thanks,
> Ethan
>
>
> >
> >>> Jeff Kirsher should be getting out a patch queued by me which adds
> >> I40E_FCoE Kbuild option, in that FCoE is disabled by default and
> >> user could enable FCoE only if needed, that patch would do same of
> >> skipping
> >> i40e_init_pf_fcoe() whether FCoE capability in device enabled or not
> >> in default config.
> >> The following patch will not fix the above issue -- configuration of
> >> PF will be changed via reset.
> >> How about the FCOE is configured and disabled by i40e_fcoe_disable()
> >> , then reset happens ?
> >>
> > May be but if the BUG is due to FCoE being enabled then having it disabled
> in config will avoid the bug for non FCoE config option and once bug is
> understood then that has to be fixed for FCoE enabled config also as I asked
> above.
> >
> > Thanks Ethan for detailed response.
> > Vasu
> >
> >>> From patchwork Wed Oct 2 23:26:08 2013
> >>> Content-Type: text/plain; charset="utf-8"
> >>> MIME-Version: 1.0
> >>> Content-Transfer-Encoding: 7bit
> >>> Subject: [net] i40e: adds FCoE configure option
> >>> Date: Thu, 03 Oct 2013 07:26:08 -0000
> >>> From: Vasu Dev <vasu.dev@intel.com>
> >>> X-Patchwork-Id: 11797
> >>>
> >>> Adds FCoE config option I40E_FCOE, so that FCoE can be enabled as
> >>> needed but otherwise have it disabled by default.
> >>>
> >>> This also eliminate multiple FCoE config checks, instead now just
> >>> one config check for CONFIG_I40E_FCOE.
> >>>
> >>> The I40E FCoE was added with 3.17 kernel and therefore this patch
> >>> shall be applied to stable 3.17 kernel also.
> >>>
> >>> CC: <stable@vger.kernel.org>
> >>> Signed-off-by: Vasu Dev <vasu.dev@intel.com>
> >>> Tested-by: Jim Young <jamesx.m.young@intel.com>
> >>>
> >>> ---
> >>> drivers/net/ethernet/intel/Kconfig | 11 +++++++++++
> >>> drivers/net/ethernet/intel/i40e/Makefile | 2 +-
> >>> drivers/net/ethernet/intel/i40e/i40e_osdep.h | 4 ++--
> >>> 3 files changed, 14 insertions(+), 3 deletions(-)
> >>>
> >>> diff --git a/drivers/net/ethernet/intel/Kconfig
> >>> b/drivers/net/ethernet/intel/Kconfig
> >>> index 5b8300a..4d61ef5 100644
> >>> --- a/drivers/net/ethernet/intel/Kconfig
> >>> +++ b/drivers/net/ethernet/intel/Kconfig
> >>> @@ -281,6 +281,17 @@ config I40E_DCB
> >>>
> >>> If unsure, say N.
> >>>
> >>> +config I40E_FCOE
> >>> + bool "Fibre Channel over Ethernet (FCoE)"
> >>> + default n
> >>> + depends on I40E && DCB && FCOE
> >>> + ---help---
> >>> + Say Y here if you want to use Fibre Channel over Ethernet (FCoE)
> >>> + in the driver. This will create new netdev for exclusive FCoE
> >>> + use with XL710 FCoE offloads enabled.
> >>> +
> >>> + If unsure, say N.
> >>> +
> >>> config I40EVF
> >>> tristate "Intel(R) XL710 X710 Virtual Function Ethernet support"
> >>> depends on PCI_MSI
> >>> diff --git a/drivers/net/ethernet/intel/i40e/Makefile
> >>> b/drivers/net/ethernet/intel/i40e/Makefile
> >>> index 4b94ddb..c405819 100644
> >>> --- a/drivers/net/ethernet/intel/i40e/Makefile
> >>> +++ b/drivers/net/ethernet/intel/i40e/Makefile
> >>> @@ -44,4 +44,4 @@ i40e-objs := i40e_main.o \
> >>> i40e_virtchnl_pf.o
> >>>
> >>> i40e-$(CONFIG_I40E_DCB) += i40e_dcb.o i40e_dcb_nl.o
> >>> -i40e-$(CONFIG_FCOE:m=y) += i40e_fcoe.o
> >>> +i40e-$(CONFIG_I40E_FCOE) += i40e_fcoe.o
> >>> diff --git a/drivers/net/ethernet/intel/i40e/i40e_osdep.h
> >>> b/drivers/net/ethernet/intel/i40e/i40e_osdep.h
> >>> index 045b5c4..ad802dd 100644
> >>> --- a/drivers/net/ethernet/intel/i40e/i40e_osdep.h
> >>> +++ b/drivers/net/ethernet/intel/i40e/i40e_osdep.h
> >>> @@ -78,7 +78,7 @@ do { \
> >>> } while (0)
> >>>
> >>> typedef enum i40e_status_code i40e_status; -#if
> >>> defined(CONFIG_FCOE)
> >>> || defined(CONFIG_FCOE_MODULE)
> >>> +#ifdef CONFIG_I40E_FCOE
> >>> #define I40E_FCOE
> >>> -#endif /* CONFIG_FCOE or CONFIG_FCOE_MODULE */
> >>> +#endif
> >>> #endif /* _I40E_OSDEP_H_ */
> >>>
> >>>>> + if (ret)
> >>>>> + dev_info(&pf->pdev->dev,
> >>>>> + "init_pf_fcoe failed: %d\n", ret);
> >>>>> + }
> >>>>>
> >>>>> #endif
> >>>>> /* do basic switch setup */
> >>>>> --
> >>>>> 1.8.3.1
> >> Thanks,
> >> Ethan
------------------------------------------------------------------------------
New Year. New Location. New Benefits. New Data Center in Ashburn, VA.
GigeNET is offering a free month of service with a new server in Ashburn.
Choose from 2 high performing configs, both with 100TB of bandwidth.
Higher redundancy.Lower latency.Increased capacity.Completely compliant.
http://p.sf.net/sfu/gigenet
_______________________________________________
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 v2 0/2] cxgb4/cxgb4i : Update & use ipv6 handling api
From: David Miller @ 2015-01-16 0:00 UTC (permalink / raw)
To: anish; +Cc: netdev, hariprasad, kxie, deepak.s
In-Reply-To: <1421277455-20158-1-git-send-email-anish@chelsio.com>
From: Anish Bhatt <anish@chelsio.com>
Date: Wed, 14 Jan 2015 15:17:33 -0800
> This patch series consolidates and updates the ipv6 api, as well as exports
> it for use by upper level drivers dependent on cxgb4
>
> v2: Fix formatting issues in clip_tbl.c
Series applied, thanks.
^ permalink raw reply
* [PATCH net-next v8 0/4] net: Add Keystone NetCP ethernet driver support
From: Murali Karicheri @ 2015-01-16 0:10 UTC (permalink / raw)
To: davem-fT/PcQaiUtIeIZ0/mPfg9Q, devicetree-u79uwXL29TY76Z2rM5mHXA,
linux-kernel-u79uwXL29TY76Z2rM5mHXA,
netdev-u79uwXL29TY76Z2rM5mHXA
Cc: Murali Karicheri
The Network Coprocessor (NetCP) is a hardware accelerator that processes
Ethernet packets. NetCP has a gigabit Ethernet (GbE) subsystem with a ethernet
switch sub-module to send and receive packets. NetCP also includes a packet
accelerator (PA) module to perform packet classification operations such as
header matching, and packet modification operations such as checksum
generation. NetCP can also optionally include a Security Accelerator(SA)
capable of performing IPSec operations on ingress/egress packets.
Keystone SoC's also have a 10 Gigabit Ethernet Subsystem (XGbE) which
includes a 3-port Ethernet switch sub-module capable of 10Gb/s and
1Gb/s rates per Ethernet port.
Both GBE and XGBE network processors supported using common driver. It
is also designed to handle future variants of NetCP.
version history
---------------
v7->v8
- Reworked comments against v7, related to checker warning.
- Patch 2/4 that has all of the driver code in v7 is now split into 3
patches based on functionality so that we have 3 smaller patches
review instead of a big patch.
- Patch for MAINTAINER is merged to 2/4 along with netcp core driver
- Separate patch (3/4) for 1G and (4/4) for 10G
- Removed big endian support for initial version (will add it later)
v6->v7
- Fixed some minor documentation error and also modified the netcp driver
to fix the set* functions to include correct le/be macros.
v5->v6
- updated version after incorporating comments [6] from David Miller,
David Laight & Geert Uytterhoeven on v5. I would like get this in
for v3.19 merge window if the latest version is acceptable.
v4->v5
- Sorry to spin v5 quickly but I missed few check-patch warnings which
were pointed by Joe Perches(thanks). I folded his changes [5] along with
few more check-patch warning fixes. I would like get this in for v3.18
merge window if David is happy with this version.
v3->v4
- Couple of fixes in in error path as pointed [4] out by David. Rest of
the patches are unchanged from v3.
v2->v3
- Update v3 after incorporating Jamal and David Miller's comment/suggestion
from earlier versions [1] [2]. After per the discussion here [3], the
controversial custom exports have been dropped now. And for future
future offload support additions, we will plug into generic frameworks
as an when they are available.
Murali Karicheri (2):
Documentation: dt: net: Add binding doc for Keystone NetCP ethernet
driver
net: netcp: Add Keystone NetCP core ethernet driver
Wingman Kwok (2):
net: netcp: Add Keystone NetCP GbE driver
net: netcp: Enhance GBE driver to support 10G Ethernet
Cc: David Miller <davem-fT/PcQaiUtIeIZ0/mPfg9Q@public.gmane.org>
Cc: Rob Herring <robh+dt-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org>
Cc: Grant Likely <grant.likely-QSEj5FYQhm4dnm+yROfE0A@public.gmane.org>
Cc: Santosh Shilimkar <santosh.shilimkar-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org>
Cc: Pawel Moll <pawel.moll-5wv7dgnIgG8@public.gmane.org>
Cc: Mark Rutland <mark.rutland-5wv7dgnIgG8@public.gmane.org>
Cc: Ian Campbell <ijc+devicetree-KcIKpvwj1kUDXYZnReoRVg@public.gmane.org>
Cc: Kumar Gala <galak-sgV2jX0FEOL9JmXXK+q4OQ@public.gmane.org>
.../devicetree/bindings/net/keystone-netcp.txt | 197 ++
MAINTAINERS | 7 +
drivers/net/ethernet/ti/Kconfig | 11 +
drivers/net/ethernet/ti/Makefile | 4 +
drivers/net/ethernet/ti/netcp.h | 229 +++
drivers/net/ethernet/ti/netcp_core.c | 2141 +++++++++++++++++++
drivers/net/ethernet/ti/netcp_ethss.c | 2156 ++++++++++++++++++++
drivers/net/ethernet/ti/netcp_sgmii.c | 131 ++
drivers/net/ethernet/ti/netcp_xgbepcsr.c | 501 +++++
9 files changed, 5377 insertions(+)
create mode 100644 Documentation/devicetree/bindings/net/keystone-netcp.txt
create mode 100644 drivers/net/ethernet/ti/netcp.h
create mode 100644 drivers/net/ethernet/ti/netcp_core.c
create mode 100644 drivers/net/ethernet/ti/netcp_ethss.c
create mode 100644 drivers/net/ethernet/ti/netcp_sgmii.c
create mode 100644 drivers/net/ethernet/ti/netcp_xgbepcsr.c
[1] https://lkml.org/lkml/2014/4/22/805
[2] https://lkml.org/lkml/2014/8/15/218
[3] https://lkml.org/lkml/2014/9/11/691
[4] https://lkml.org/lkml/2014/9/15/403
[5] https://lkml.org/lkml/2014/9/25/332
[6] https://lkml.org/lkml/2014/9/29/639
--
1.7.9.5
--
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
* [PATCH net-next v8 1/4] Documentation: dt: net: Add binding doc for Keystone NetCP ethernet driver
From: Murali Karicheri @ 2015-01-16 0:12 UTC (permalink / raw)
To: davem, devicetree, linux-kernel, netdev
Cc: Murali Karicheri, Rob Herring, Grant Likely, Pawel Moll,
Mark Rutland, Ian Campbell, Kumar Gala, Santosh Shilimkar
The network coprocessor (NetCP) is a hardware accelerator that processes
Ethernet packets. NetCP has a gigabit Ethernet (GbE) subsystem with a ethernet
switch sub-module to send and receive packets. NetCP also includes a packet
accelerator (PA) module to perform packet classification operations such as
header matching, and packet modification operations such as checksum
generation. NetCP can also optionally include a Security Accelerator(SA)
capable of performing IPSec operations on ingress/egress packets.
Keystone SoC's also have a 10 Gigabit Ethernet Subsystem (XGbE) which
includes a 3-port Ethernet switch sub-module capable of 10Gb/s and
1Gb/s rates per Ethernet port.
NetCP Subsystem device tree layout looks something like below:
-----------------------------
NetCP subsystem(10G or 1G)
-----------------------------
|
|-> NetCP Devices -> |
| |-> GBE/XGBE Switch
| |
| |-> Packet Accelerator
| |
| |-> Security Accelerator
|
|
|
|-> NetCP Interfaces -> |
|-> Ethernet Port 0
|
|-> Ethernet Port 1
|
|-> Ethernet Port 2
|
|-> Ethernet Port 3
Common driver supports GBE as well XGBE network processors.
Cc: Rob Herring <robh+dt@kernel.org>
Cc: Grant Likely <grant.likely@linaro.org>
Cc: Pawel Moll <pawel.moll@arm.com>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Ian Campbell <ijc+devicetree@hellion.org.uk>
Cc: Kumar Gala <galak@codeaurora.org>
Cc: David Miller <davem@davemloft.net>
Cc: Santosh Shilimkar <santosh.shilimkar@kernel.org>
Signed-off-by: Murali Karicheri <m-karicheri2@ti.com>
---
.../devicetree/bindings/net/keystone-netcp.txt | 197 ++++++++++++++++++++
1 file changed, 197 insertions(+)
create mode 100644 Documentation/devicetree/bindings/net/keystone-netcp.txt
diff --git a/Documentation/devicetree/bindings/net/keystone-netcp.txt b/Documentation/devicetree/bindings/net/keystone-netcp.txt
new file mode 100644
index 0000000..f9c0771
--- /dev/null
+++ b/Documentation/devicetree/bindings/net/keystone-netcp.txt
@@ -0,0 +1,197 @@
+This document describes the device tree bindings associated with the
+keystone network coprocessor(NetCP) driver support.
+
+The network coprocessor (NetCP) is a hardware accelerator that processes
+Ethernet packets. NetCP has a gigabit Ethernet (GbE) subsytem with a ethernet
+switch sub-module to send and receive packets. NetCP also includes a packet
+accelerator (PA) module to perform packet classification operations such as
+header matching, and packet modification operations such as checksum
+generation. NetCP can also optionally include a Security Accelerator (SA)
+capable of performing IPSec operations on ingress/egress packets.
+
+Keystone II SoC's also have a 10 Gigabit Ethernet Subsystem (XGbE) which
+includes a 3-port Ethernet switch sub-module capable of 10Gb/s and 1Gb/s rates
+per Ethernet port.
+
+Keystone NetCP driver has a plug-in module architecture where each of the NetCP
+sub-modules exist as a loadable kernel module which plug in to the netcp core.
+These sub-modules are represented as "netcp-devices" in the dts bindings. It is
+mandatory to have the ethernet switch sub-module for the ethernet interface to
+be operational. Any other sub-module like the PA is optional.
+
+NetCP Ethernet SubSystem Layout:
+
+-----------------------------
+ NetCP subsystem(10G or 1G)
+-----------------------------
+ |
+ |-> NetCP Devices -> |
+ | |-> GBE/XGBE Switch
+ | |
+ | |-> Packet Accelerator
+ | |
+ | |-> Security Accelerator
+ |
+ |
+ |
+ |-> NetCP Interfaces -> |
+ |-> Ethernet Port 0
+ |
+ |-> Ethernet Port 1
+ |
+ |-> Ethernet Port 2
+ |
+ |-> Ethernet Port 3
+
+
+NetCP subsystem properties:
+Required properties:
+- compatible: Should be "ti,netcp-1.0"
+- clocks: phandle to the reference clocks for the subsystem.
+- dma-id: Navigator packet dma instance id.
+
+Optional properties:
+- reg: register location and the size for the following register
+ regions in the specified order.
+ - Efuse MAC address register
+- dma-coherent: Present if dma operations are coherent
+- big-endian: Keystone devices can be operated in a mode where the DSP is in
+ the big endian mode. In such cases enable this option. This
+ option should also be enabled if the ARM is operated in
+ big endian mode with the DSP in little endian.
+
+NetCP device properties: Device specification for NetCP sub-modules.
+1Gb/10Gb (gbe/xgbe) ethernet switch sub-module specifications.
+Required properties:
+- label: Must be "netcp-gbe" for 1Gb & "netcp-xgbe" for 10Gb.
+- reg: register location and the size for the following register
+ regions in the specified order.
+ - subsystem registers
+ - serdes registers
+- tx-channel: the navigator packet dma channel name for tx.
+- tx-queue: the navigator queue number associated with the tx dma channel.
+- interfaces: specification for each of the switch port to be registered as a
+ network interface in the stack.
+-- slave-port: Switch port number, 0 based numbering.
+-- link-interface: type of link interface, supported options are
+ - mac<->mac auto negotiate mode: 0
+ - mac<->phy mode: 1
+ - mac<->mac forced mode: 2
+ - mac<->fiber mode: 3
+ - mac<->phy mode with no mdio: 4
+ - 10Gb mac<->phy mode : 10
+ - 10Gb mac<->mac forced mode : 11
+----phy-handle: phandle to PHY device
+
+Optional properties:
+- enable-ale: NetCP driver keeps the address learning feature in the ethernet
+ switch module disabled. This attribute is to enable the address
+ learning.
+- secondary-slave-ports: specification for each of the switch port not be
+ registered as a network interface. NetCP driver
+ will only initialize these ports and attach PHY
+ driver to them if needed.
+
+NetCP interface properties: Interface specification for NetCP sub-modules.
+Required properties:
+- rx-channel: the navigator packet dma channel name for rx.
+- rx-queue: the navigator queue number associated with rx dma channel.
+- rx-pool: specifies the number of descriptors to be used & the region-id
+ for creating the rx descriptor pool.
+- tx-pool: specifies the number of descriptors to be used & the region-id
+ for creating the tx descriptor pool.
+- rx-queue-depth: number of descriptors in each of the free descriptor
+ queue (FDQ) for the pktdma Rx flow. There can be at
+ present a maximum of 4 queues per Rx flow.
+- rx-buffer-size: the buffer size for each of the Rx flow FDQ.
+- tx-completion-queue: the navigator queue number where the descriptors are
+ recycled after Tx DMA completion.
+
+Optional properties:
+- efuse-mac: If this is 1, then the MAC address for the interface is
+ obtained from the device efuse mac address register
+- local-mac-address: the driver is designed to use the of_get_mac_address api
+ only if efuse-mac is 0. When efuse-mac is 0, the MAC
+ address is obtained from local-mac-address. If this
+ attribute is not present, then the driver will use a
+ random MAC address.
+- "netcp-device label": phandle to the device specification for each of NetCP
+ sub-module attached to this interface.
+
+Example binding:
+
+netcp: netcp@2090000 {
+ reg = <0x2620110 0x8>;
+ reg-names = "efuse";
+ compatible = "ti,netcp-1.0";
+ #address-cells = <1>;
+ #size-cells = <1>;
+ ranges;
+
+ clocks = <&papllclk>, <&clkcpgmac>, <&chipclk12>;
+ dma-coherent;
+ /* big-endian; */
+ dma-id = <0>;
+
+ netcp-devices {
+ #address-cells = <1>;
+ #size-cells = <1>;
+ ranges;
+ gbe@0x2090000 {
+ label = "netcp-gbe";
+ reg = <0x2090000 0xf00>;
+ /* enable-ale; */
+ tx-queue = <648>;
+ tx-channel = <8>;
+
+ interfaces {
+ gbe0: interface-0 {
+ slave-port = <0>;
+ link-interface = <4>;
+ };
+ gbe1: interface-1 {
+ slave-port = <1>;
+ link-interface = <4>;
+ };
+ };
+
+ secondary-slave-ports {
+ port-2 {
+ slave-port = <2>;
+ link-interface = <2>;
+ };
+ port-3 {
+ slave-port = <3>;
+ link-interface = <2>;
+ };
+ };
+ };
+ };
+
+ netcp-interfaces {
+ interface-0 {
+ rx-channel = <22>;
+ rx-pool = <1024 12>;
+ tx-pool = <1024 12>;
+ rx-queue-depth = <128 128 0 0>;
+ rx-buffer-size = <1518 4096 0 0>;
+ rx-queue = <8704>;
+ tx-completion-queue = <8706>;
+ efuse-mac = <1>;
+ netcp-gbe = <&gbe0>;
+
+ };
+ interface-1 {
+ rx-channel = <23>;
+ rx-pool = <1024 12>;
+ tx-pool = <1024 12>;
+ rx-queue-depth = <128 128 0 0>;
+ rx-buffer-size = <1518 4096 0 0>;
+ rx-queue = <8705>;
+ tx-completion-queue = <8707>;
+ efuse-mac = <0>;
+ local-mac-address = [02 18 31 7e 3e 6f];
+ netcp-gbe = <&gbe1>;
+ };
+ };
+};
--
1.7.9.5
^ permalink raw reply related
* [PATCH net-next v8 2/4] net: netcp: Add Keystone NetCP core ethernet driver
From: Murali Karicheri @ 2015-01-16 0:12 UTC (permalink / raw)
To: davem-fT/PcQaiUtIeIZ0/mPfg9Q, devicetree-u79uwXL29TY76Z2rM5mHXA,
linux-kernel-u79uwXL29TY76Z2rM5mHXA,
netdev-u79uwXL29TY76Z2rM5mHXA
Cc: Murali Karicheri, Wingman Kwok
In-Reply-To: <1421367172-19793-1-git-send-email-m-karicheri2-l0cyMroinI0@public.gmane.org>
The network coprocessor (NetCP) is a hardware accelerator available in
Keystone SoCs that processes Ethernet packets. NetCP consists of following
hardware components
1 Gigabit Ethernet (GbE) subsystem with a Ethernet switch sub-module to
send and receive packets.
2 Packet Accelerator (PA) module to perform packet classification
operations such as header matching, and packet modification operations
such as checksum generation.
3 Security Accelerator(SA) capable of performing IPSec operations on
ingress/egress packets.
4 An optional 10 Gigabit Ethernet Subsystem (XGbE) which includes a
3-port Ethernet switch sub-module capable of 10Gb/s and 1Gb/s rates
per Ethernet port.
5 Packet DMA and Queue Management Subsystem (QMSS) to enqueue and dequeue
packets and DMA the packets between memory and NetCP hardware components
described above.
NetCP core driver make use of the Keystone Navigator driver API to allocate
DMA channel for the Ethenet device and to handle packet queue/de-queue,
Please refer API's in include/linux/soc/ti/knav_dma.h and
drivers/soc/ti/knav_qmss.h for details.
NetCP driver consists of NetCP core driver and at a minimum Gigabit
Ethernet (GBE) module (1) driver to implement the Network device function.
Other modules (2,3) can be optionally added to achieve supported hardware
acceleration function. The initial version of the driver include NetCP
core driver and GBE driver modules.
Please refer Documentation/devicetree/bindings/net/keystone-netcp.txt
for design of the driver.
Cc: David Miller <davem-fT/PcQaiUtIeIZ0/mPfg9Q@public.gmane.org>
Cc: Rob Herring <robh+dt-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org>
Cc: Grant Likely <grant.likely-QSEj5FYQhm4dnm+yROfE0A@public.gmane.org>
Cc: Santosh Shilimkar <santosh.shilimkar-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org>
Cc: Pawel Moll <pawel.moll-5wv7dgnIgG8@public.gmane.org>
Cc: Mark Rutland <mark.rutland-5wv7dgnIgG8@public.gmane.org>
Cc: Ian Campbell <ijc+devicetree-KcIKpvwj1kUDXYZnReoRVg@public.gmane.org>
Cc: Kumar Gala <galak-sgV2jX0FEOL9JmXXK+q4OQ@public.gmane.org>
Signed-off-by: Murali Karicheri <m-karicheri2-l0cyMroinI0@public.gmane.org>
Signed-off-by: Wingman Kwok <w-kwok2-l0cyMroinI0@public.gmane.org>
---
MAINTAINERS | 7 +
drivers/net/ethernet/ti/Kconfig | 11 +
drivers/net/ethernet/ti/Makefile | 3 +
drivers/net/ethernet/ti/netcp.h | 229 ++++
drivers/net/ethernet/ti/netcp_core.c | 2141 ++++++++++++++++++++++++++++++++++
5 files changed, 2391 insertions(+)
create mode 100644 drivers/net/ethernet/ti/netcp.h
create mode 100644 drivers/net/ethernet/ti/netcp_core.c
diff --git a/MAINTAINERS b/MAINTAINERS
index 3309895..0cddbf0 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -9591,6 +9591,13 @@ F: drivers/power/lp8788-charger.c
F: drivers/regulator/lp8788-*.c
F: include/linux/mfd/lp8788*.h
+TI NETCP ETHERNET DRIVER
+M: Wingman Kwok <w-kwok2-l0cyMroinI0@public.gmane.org>
+M: Murali Karicheri <m-karicheri2-l0cyMroinI0@public.gmane.org>
+L: netdev-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
+S: Maintained
+F: drivers/net/ethernet/ti/netcp*
+
TI TWL4030 SERIES SOC CODEC DRIVER
M: Peter Ujfalusi <peter.ujfalusi-l0cyMroinI0@public.gmane.org>
L: alsa-devel-K7yf7f+aM1XWsZ/bQMPhNw@public.gmane.org (moderated for non-subscribers)
diff --git a/drivers/net/ethernet/ti/Kconfig b/drivers/net/ethernet/ti/Kconfig
index 605dd90..e11bcfa 100644
--- a/drivers/net/ethernet/ti/Kconfig
+++ b/drivers/net/ethernet/ti/Kconfig
@@ -73,12 +73,23 @@ config TI_CPSW
config TI_CPTS
boolean "TI Common Platform Time Sync (CPTS) Support"
depends on TI_CPSW
+ depends on TI_CPSW || TI_KEYSTONE_NET
select PTP_1588_CLOCK
---help---
This driver supports the Common Platform Time Sync unit of
the CPSW Ethernet Switch. The unit can time stamp PTP UDP/IPv4
and Layer 2 packets, and the driver offers a PTP Hardware Clock.
+config TI_KEYSTONE_NETCP
+ tristate "TI Keystone NETCP Ethernet subsystem Support"
+ depends on OF
+ depends on KEYSTONE_NAVIGATOR_DMA && KEYSTONE_NAVIGATOR_QMSS
+ ---help---
+ This driver supports TI's Keystone NETCP Ethernet subsystem.
+
+ To compile this driver as a module, choose M here: the module
+ will be called keystone_netcp.
+
config TLAN
tristate "TI ThunderLAN support"
depends on (PCI || EISA)
diff --git a/drivers/net/ethernet/ti/Makefile b/drivers/net/ethernet/ti/Makefile
index 9cfaab8..4e8a8e4 100644
--- a/drivers/net/ethernet/ti/Makefile
+++ b/drivers/net/ethernet/ti/Makefile
@@ -10,3 +10,6 @@ obj-$(CONFIG_TI_DAVINCI_CPDMA) += davinci_cpdma.o
obj-$(CONFIG_TI_CPSW_PHY_SEL) += cpsw-phy-sel.o
obj-$(CONFIG_TI_CPSW) += ti_cpsw.o
ti_cpsw-y := cpsw_ale.o cpsw.o cpts.o
+
+obj-$(CONFIG_TI_KEYSTONE_NETCP) += keystone_netcp.o
+keystone_netcp-y := netcp_core.o
diff --git a/drivers/net/ethernet/ti/netcp.h b/drivers/net/ethernet/ti/netcp.h
new file mode 100644
index 0000000..906e9bc
--- /dev/null
+++ b/drivers/net/ethernet/ti/netcp.h
@@ -0,0 +1,229 @@
+/*
+ * NetCP driver local header
+ *
+ * Copyright (C) 2014 Texas Instruments Incorporated
+ * Authors: Sandeep Nair <sandeep_n-l0cyMroinI0@public.gmane.org>
+ * Sandeep Paulraj <s-paulraj-l0cyMroinI0@public.gmane.org>
+ * Cyril Chemparathy <cyril-l0cyMroinI0@public.gmane.org>
+ * Santosh Shilimkar <santosh.shilimkar-l0cyMroinI0@public.gmane.org>
+ * Wingman Kwok <w-kwok2-l0cyMroinI0@public.gmane.org>
+ * Murali Karicheri <m-karicheri2-l0cyMroinI0@public.gmane.org>
+ *
+ * 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 version 2.
+ *
+ * This program is distributed "as is" WITHOUT ANY WARRANTY of any
+ * kind, whether express or implied; without even the implied warranty
+ * of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ */
+#ifndef __NETCP_H__
+#define __NETCP_H__
+
+#include <linux/netdevice.h>
+#include <linux/soc/ti/knav_dma.h>
+
+/* Maximum Ethernet frame size supported by Keystone switch */
+#define NETCP_MAX_FRAME_SIZE 9504
+
+#define SGMII_LINK_MAC_MAC_AUTONEG 0
+#define SGMII_LINK_MAC_PHY 1
+#define SGMII_LINK_MAC_MAC_FORCED 2
+#define SGMII_LINK_MAC_FIBER 3
+#define SGMII_LINK_MAC_PHY_NO_MDIO 4
+#define XGMII_LINK_MAC_PHY 10
+#define XGMII_LINK_MAC_MAC_FORCED 11
+
+struct netcp_device;
+
+struct netcp_tx_pipe {
+ struct netcp_device *netcp_device;
+ void *dma_queue;
+ unsigned int dma_queue_id;
+ u8 dma_psflags;
+ void *dma_channel;
+ const char *dma_chan_name;
+};
+
+#define ADDR_NEW BIT(0)
+#define ADDR_VALID BIT(1)
+
+enum netcp_addr_type {
+ ADDR_ANY,
+ ADDR_DEV,
+ ADDR_UCAST,
+ ADDR_MCAST,
+ ADDR_BCAST
+};
+
+struct netcp_addr {
+ struct netcp_intf *netcp;
+ unsigned char addr[ETH_ALEN];
+ enum netcp_addr_type type;
+ unsigned int flags;
+ struct list_head node;
+};
+
+struct netcp_intf {
+ struct device *dev;
+ struct device *ndev_dev;
+ struct net_device *ndev;
+ bool big_endian;
+ unsigned int tx_compl_qid;
+ void *tx_pool;
+ struct list_head txhook_list_head;
+ unsigned int tx_pause_threshold;
+ void *tx_compl_q;
+
+ unsigned int tx_resume_threshold;
+ void *rx_queue;
+ void *rx_pool;
+ struct list_head rxhook_list_head;
+ unsigned int rx_queue_id;
+ void *rx_fdq[KNAV_DMA_FDQ_PER_CHAN];
+ u32 rx_buffer_sizes[KNAV_DMA_FDQ_PER_CHAN];
+ struct napi_struct rx_napi;
+ struct napi_struct tx_napi;
+
+ void *rx_channel;
+ const char *dma_chan_name;
+ u32 rx_pool_size;
+ u32 rx_pool_region_id;
+ u32 tx_pool_size;
+ u32 tx_pool_region_id;
+ struct list_head module_head;
+ struct list_head interface_list;
+ struct list_head addr_list;
+ bool netdev_registered;
+ bool primary_module_attached;
+
+ /* Lock used for protecting Rx/Tx hook list management */
+ spinlock_t lock;
+ struct netcp_device *netcp_device;
+ struct device_node *node_interface;
+
+ /* DMA configuration data */
+ u32 msg_enable;
+ u32 rx_queue_depths[KNAV_DMA_FDQ_PER_CHAN];
+};
+
+#define NETCP_PSDATA_LEN KNAV_DMA_NUM_PS_WORDS
+struct netcp_packet {
+ struct sk_buff *skb;
+ u32 *epib;
+ u32 *psdata;
+ unsigned int psdata_len;
+ struct netcp_intf *netcp;
+ struct netcp_tx_pipe *tx_pipe;
+ bool rxtstamp_complete;
+ void *ts_context;
+
+ int (*txtstamp_complete)(void *ctx, struct netcp_packet *pkt);
+};
+
+static inline u32 *netcp_push_psdata(struct netcp_packet *p_info,
+ unsigned int bytes)
+{
+ u32 *buf;
+ unsigned int words;
+
+ if ((bytes & 0x03) != 0)
+ return NULL;
+ words = bytes >> 2;
+
+ if ((p_info->psdata_len + words) > NETCP_PSDATA_LEN)
+ return NULL;
+
+ p_info->psdata_len += words;
+ buf = &p_info->psdata[NETCP_PSDATA_LEN - p_info->psdata_len];
+ return buf;
+}
+
+static inline int netcp_align_psdata(struct netcp_packet *p_info,
+ unsigned int byte_align)
+{
+ int padding;
+
+ switch (byte_align) {
+ case 0:
+ padding = -EINVAL;
+ break;
+ case 1:
+ case 2:
+ case 4:
+ padding = 0;
+ break;
+ case 8:
+ padding = (p_info->psdata_len << 2) % 8;
+ break;
+ case 16:
+ padding = (p_info->psdata_len << 2) % 16;
+ break;
+ default:
+ padding = (p_info->psdata_len << 2) % byte_align;
+ break;
+ }
+ return padding;
+}
+
+struct netcp_module {
+ const char *name;
+ struct module *owner;
+ bool primary;
+
+ /* probe/remove: called once per NETCP instance */
+ int (*probe)(struct netcp_device *netcp_device,
+ struct device *device, struct device_node *node,
+ void **inst_priv);
+ int (*remove)(struct netcp_device *netcp_device, void *inst_priv);
+
+ /* attach/release: called once per network interface */
+ int (*attach)(void *inst_priv, struct net_device *ndev,
+ struct device_node *node, void **intf_priv);
+ int (*release)(void *intf_priv);
+ int (*open)(void *intf_priv, struct net_device *ndev);
+ int (*close)(void *intf_priv, struct net_device *ndev);
+ int (*add_addr)(void *intf_priv, struct netcp_addr *naddr);
+ int (*del_addr)(void *intf_priv, struct netcp_addr *naddr);
+ int (*add_vid)(void *intf_priv, int vid);
+ int (*del_vid)(void *intf_priv, int vid);
+ int (*ioctl)(void *intf_priv, struct ifreq *req, int cmd);
+
+ /* used internally */
+ struct list_head module_list;
+ struct list_head interface_list;
+};
+
+int netcp_register_module(struct netcp_module *module);
+void netcp_unregister_module(struct netcp_module *module);
+void *netcp_module_get_intf_data(struct netcp_module *module,
+ struct netcp_intf *intf);
+
+int netcp_txpipe_init(struct netcp_tx_pipe *tx_pipe,
+ struct netcp_device *netcp_device,
+ const char *dma_chan_name, unsigned int dma_queue_id);
+int netcp_txpipe_open(struct netcp_tx_pipe *tx_pipe);
+int netcp_txpipe_close(struct netcp_tx_pipe *tx_pipe);
+
+typedef int netcp_hook_rtn(int order, void *data, struct netcp_packet *packet);
+int netcp_register_txhook(struct netcp_intf *netcp_priv, int order,
+ netcp_hook_rtn *hook_rtn, void *hook_data);
+int netcp_unregister_txhook(struct netcp_intf *netcp_priv, int order,
+ netcp_hook_rtn *hook_rtn, void *hook_data);
+int netcp_register_rxhook(struct netcp_intf *netcp_priv, int order,
+ netcp_hook_rtn *hook_rtn, void *hook_data);
+int netcp_unregister_rxhook(struct netcp_intf *netcp_priv, int order,
+ netcp_hook_rtn *hook_rtn, void *hook_data);
+void *netcp_device_find_module(struct netcp_device *netcp_device,
+ const char *name);
+
+/* SGMII functions */
+int netcp_sgmii_reset(void __iomem *sgmii_ofs, int port);
+int netcp_sgmii_get_port_link(void __iomem *sgmii_ofs, int port);
+int netcp_sgmii_config(void __iomem *sgmii_ofs, int port, u32 interface);
+
+/* XGBE SERDES init functions */
+int netcp_xgbe_serdes_init(void __iomem *serdes_regs, void __iomem *xgbe_regs);
+
+#endif /* __NETCP_H__ */
diff --git a/drivers/net/ethernet/ti/netcp_core.c b/drivers/net/ethernet/ti/netcp_core.c
new file mode 100644
index 0000000..ba3002e
--- /dev/null
+++ b/drivers/net/ethernet/ti/netcp_core.c
@@ -0,0 +1,2141 @@
+/*
+ * Keystone NetCP Core driver
+ *
+ * Copyright (C) 2014 Texas Instruments Incorporated
+ * Authors: Sandeep Nair <sandeep_n-l0cyMroinI0@public.gmane.org>
+ * Sandeep Paulraj <s-paulraj-l0cyMroinI0@public.gmane.org>
+ * Cyril Chemparathy <cyril-l0cyMroinI0@public.gmane.org>
+ * Santosh Shilimkar <santosh.shilimkar-l0cyMroinI0@public.gmane.org>
+ * Murali Karicheri <m-karicheri2-l0cyMroinI0@public.gmane.org>
+ * Wingman Kwok <w-kwok2-l0cyMroinI0@public.gmane.org>
+ *
+ * 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 version 2.
+ *
+ * This program is distributed "as is" WITHOUT ANY WARRANTY of any
+ * kind, whether express or implied; without even the implied warranty
+ * of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ */
+
+#include <linux/io.h>
+#include <linux/module.h>
+#include <linux/of_net.h>
+#include <linux/of_address.h>
+#include <linux/if_vlan.h>
+#include <linux/pm_runtime.h>
+#include <linux/platform_device.h>
+#include <linux/soc/ti/knav_qmss.h>
+#include <linux/soc/ti/knav_dma.h>
+
+#include "netcp.h"
+
+#define NETCP_SOP_OFFSET (NET_IP_ALIGN + NET_SKB_PAD)
+#define NETCP_NAPI_WEIGHT 64
+#define NETCP_TX_TIMEOUT (5 * HZ)
+#define NETCP_MIN_PACKET_SIZE ETH_ZLEN
+#define NETCP_MAX_MCAST_ADDR 16
+
+#define NETCP_EFUSE_REG_INDEX 0
+
+#define NETCP_MOD_PROBE_SKIPPED 1
+#define NETCP_MOD_PROBE_FAILED 2
+
+#define NETCP_DEBUG (NETIF_MSG_HW | NETIF_MSG_WOL | \
+ NETIF_MSG_DRV | NETIF_MSG_LINK | \
+ NETIF_MSG_IFUP | NETIF_MSG_INTR | \
+ NETIF_MSG_PROBE | NETIF_MSG_TIMER | \
+ NETIF_MSG_IFDOWN | NETIF_MSG_RX_ERR | \
+ NETIF_MSG_TX_ERR | NETIF_MSG_TX_DONE | \
+ NETIF_MSG_PKTDATA | NETIF_MSG_TX_QUEUED | \
+ NETIF_MSG_RX_STATUS)
+
+#define knav_queue_get_id(q) knav_queue_device_control(q, \
+ KNAV_QUEUE_GET_ID, (unsigned long)NULL)
+
+#define knav_queue_enable_notify(q) knav_queue_device_control(q, \
+ KNAV_QUEUE_ENABLE_NOTIFY, \
+ (unsigned long)NULL)
+
+#define knav_queue_disable_notify(q) knav_queue_device_control(q, \
+ KNAV_QUEUE_DISABLE_NOTIFY, \
+ (unsigned long)NULL)
+
+#define knav_queue_get_count(q) knav_queue_device_control(q, \
+ KNAV_QUEUE_GET_COUNT, (unsigned long)NULL)
+
+#define for_each_netcp_module(module) \
+ list_for_each_entry(module, &netcp_modules, module_list)
+
+#define for_each_netcp_device_module(netcp_device, inst_modpriv) \
+ list_for_each_entry(inst_modpriv, \
+ &((netcp_device)->modpriv_head), inst_list)
+
+#define for_each_module(netcp, intf_modpriv) \
+ list_for_each_entry(intf_modpriv, &netcp->module_head, intf_list)
+
+/* Module management structures */
+struct netcp_device {
+ struct list_head device_list;
+ struct list_head interface_head;
+ struct list_head modpriv_head;
+ struct device *device;
+};
+
+struct netcp_inst_modpriv {
+ struct netcp_device *netcp_device;
+ struct netcp_module *netcp_module;
+ struct list_head inst_list;
+ void *module_priv;
+};
+
+struct netcp_intf_modpriv {
+ struct netcp_intf *netcp_priv;
+ struct netcp_module *netcp_module;
+ struct list_head intf_list;
+ void *module_priv;
+};
+
+static LIST_HEAD(netcp_devices);
+static LIST_HEAD(netcp_modules);
+static DEFINE_MUTEX(netcp_modules_lock);
+
+static int netcp_debug_level = -1;
+module_param(netcp_debug_level, int, 0);
+MODULE_PARM_DESC(netcp_debug_level, "Netcp debug level (NETIF_MSG bits) (0=none,...,16=all)");
+
+/* Helper functions - Get/Set */
+static void get_pkt_info(u32 *buff, u32 *buff_len, u32 *ndesc,
+ struct knav_dma_desc *desc)
+{
+ *buff_len = desc->buff_len;
+ *buff = desc->buff;
+ *ndesc = desc->next_desc;
+}
+
+static void get_pad_info(u32 *pad0, u32 *pad1, struct knav_dma_desc *desc)
+{
+ *pad0 = desc->pad[0];
+ *pad1 = desc->pad[1];
+}
+
+static void get_org_pkt_info(u32 *buff, u32 *buff_len,
+ struct knav_dma_desc *desc)
+{
+ *buff = desc->orig_buff;
+ *buff_len = desc->orig_len;
+}
+
+static void get_words(u32 *words, int num_words, u32 *desc)
+{
+ int i;
+
+ for (i = 0; i < num_words; i++)
+ words[i] = desc[i];
+}
+
+static void set_pkt_info(u32 buff, u32 buff_len, u32 ndesc,
+ struct knav_dma_desc *desc)
+{
+ desc->buff_len = buff_len;
+ desc->buff = buff;
+ desc->next_desc = ndesc;
+}
+
+static void set_desc_info(u32 desc_info, u32 pkt_info,
+ struct knav_dma_desc *desc)
+{
+ desc->desc_info = desc_info;
+ desc->packet_info = pkt_info;
+}
+
+static void set_pad_info(u32 pad0, u32 pad1, struct knav_dma_desc *desc)
+{
+ desc->pad[0] = pad0;
+ desc->pad[1] = pad1;
+}
+
+static void set_org_pkt_info(u32 buff, u32 buff_len,
+ struct knav_dma_desc *desc)
+{
+ desc->orig_buff = buff;
+ desc->orig_len = buff_len;
+}
+
+static void set_words(u32 *words, int num_words, u32 *desc)
+{
+ int i;
+
+ for (i = 0; i < num_words; i++)
+ desc[i] = words[i];
+}
+
+/* Read the e-fuse value as 32 bit values to be endian independent */
+static int emac_arch_get_mac_addr(char *x, void __iomem *efuse_mac)
+{
+ unsigned int addr0, addr1;
+
+ addr1 = readl(efuse_mac + 4);
+ addr0 = readl(efuse_mac);
+
+ x[0] = (addr1 & 0x0000ff00) >> 8;
+ x[1] = addr1 & 0x000000ff;
+ x[2] = (addr0 & 0xff000000) >> 24;
+ x[3] = (addr0 & 0x00ff0000) >> 16;
+ x[4] = (addr0 & 0x0000ff00) >> 8;
+ x[5] = addr0 & 0x000000ff;
+
+ return 0;
+}
+
+static const char *netcp_node_name(struct device_node *node)
+{
+ const char *name;
+
+ if (of_property_read_string(node, "label", &name) < 0)
+ name = node->name;
+ if (!name)
+ name = "unknown";
+ return name;
+}
+
+/* Module management routines */
+static int netcp_register_interface(struct netcp_intf *netcp)
+{
+ int ret;
+
+ ret = register_netdev(netcp->ndev);
+ if (!ret)
+ netcp->netdev_registered = true;
+ return ret;
+}
+
+static int netcp_module_probe(struct netcp_device *netcp_device,
+ struct netcp_module *module)
+{
+ struct device *dev = netcp_device->device;
+ struct device_node *devices, *interface, *node = dev->of_node;
+ struct device_node *child;
+ struct netcp_inst_modpriv *inst_modpriv;
+ struct netcp_intf *netcp_intf;
+ struct netcp_module *tmp;
+ bool primary_module_registered = false;
+ int ret;
+
+ /* Find this module in the sub-tree for this device */
+ devices = of_get_child_by_name(node, "netcp-devices");
+ if (!devices) {
+ dev_err(dev, "could not find netcp-devices node\n");
+ return NETCP_MOD_PROBE_SKIPPED;
+ }
+
+ for_each_available_child_of_node(devices, child) {
+ const char *name = netcp_node_name(child);
+
+ if (!strcasecmp(module->name, name))
+ break;
+ }
+
+ of_node_put(devices);
+ /* If module not used for this device, skip it */
+ if (!child) {
+ dev_warn(dev, "module(%s) not used for device\n", module->name);
+ return NETCP_MOD_PROBE_SKIPPED;
+ }
+
+ inst_modpriv = devm_kzalloc(dev, sizeof(*inst_modpriv), GFP_KERNEL);
+ if (!inst_modpriv) {
+ of_node_put(child);
+ return -ENOMEM;
+ }
+
+ inst_modpriv->netcp_device = netcp_device;
+ inst_modpriv->netcp_module = module;
+ list_add_tail(&inst_modpriv->inst_list, &netcp_device->modpriv_head);
+
+ ret = module->probe(netcp_device, dev, child,
+ &inst_modpriv->module_priv);
+ of_node_put(child);
+ if (ret) {
+ dev_err(dev, "Probe of module(%s) failed with %d\n",
+ module->name, ret);
+ list_del(&inst_modpriv->inst_list);
+ devm_kfree(dev, inst_modpriv);
+ return NETCP_MOD_PROBE_FAILED;
+ }
+
+ /* Attach modules only if the primary module is probed */
+ for_each_netcp_module(tmp) {
+ if (tmp->primary)
+ primary_module_registered = true;
+ }
+
+ if (!primary_module_registered)
+ return 0;
+
+ /* Attach module to interfaces */
+ list_for_each_entry(netcp_intf, &netcp_device->interface_head,
+ interface_list) {
+ struct netcp_intf_modpriv *intf_modpriv;
+
+ /* If interface not registered then register now */
+ if (!netcp_intf->netdev_registered)
+ ret = netcp_register_interface(netcp_intf);
+
+ if (ret)
+ return -ENODEV;
+
+ intf_modpriv = devm_kzalloc(dev, sizeof(*intf_modpriv),
+ GFP_KERNEL);
+ if (!intf_modpriv)
+ return -ENOMEM;
+
+ interface = of_parse_phandle(netcp_intf->node_interface,
+ module->name, 0);
+
+ intf_modpriv->netcp_priv = netcp_intf;
+ intf_modpriv->netcp_module = module;
+ list_add_tail(&intf_modpriv->intf_list,
+ &netcp_intf->module_head);
+
+ ret = module->attach(inst_modpriv->module_priv,
+ netcp_intf->ndev, interface,
+ &intf_modpriv->module_priv);
+ of_node_put(interface);
+ if (ret) {
+ dev_dbg(dev, "Attach of module %s declined with %d\n",
+ module->name, ret);
+ list_del(&intf_modpriv->intf_list);
+ devm_kfree(dev, intf_modpriv);
+ continue;
+ }
+ }
+ return 0;
+}
+
+int netcp_register_module(struct netcp_module *module)
+{
+ struct netcp_device *netcp_device;
+ struct netcp_module *tmp;
+ int ret;
+
+ if (!module->name) {
+ WARN(1, "error registering netcp module: no name\n");
+ return -EINVAL;
+ }
+
+ if (!module->probe) {
+ WARN(1, "error registering netcp module: no probe\n");
+ return -EINVAL;
+ }
+
+ mutex_lock(&netcp_modules_lock);
+
+ for_each_netcp_module(tmp) {
+ if (!strcasecmp(tmp->name, module->name)) {
+ mutex_unlock(&netcp_modules_lock);
+ return -EEXIST;
+ }
+ }
+ list_add_tail(&module->module_list, &netcp_modules);
+
+ list_for_each_entry(netcp_device, &netcp_devices, device_list) {
+ ret = netcp_module_probe(netcp_device, module);
+ if (ret < 0)
+ goto fail;
+ }
+
+ mutex_unlock(&netcp_modules_lock);
+ return 0;
+
+fail:
+ mutex_unlock(&netcp_modules_lock);
+ netcp_unregister_module(module);
+ return ret;
+}
+
+static void netcp_release_module(struct netcp_device *netcp_device,
+ struct netcp_module *module)
+{
+ struct netcp_inst_modpriv *inst_modpriv, *inst_tmp;
+ struct netcp_intf *netcp_intf, *netcp_tmp;
+ struct device *dev = netcp_device->device;
+
+ /* Release the module from each interface */
+ list_for_each_entry_safe(netcp_intf, netcp_tmp,
+ &netcp_device->interface_head,
+ interface_list) {
+ struct netcp_intf_modpriv *intf_modpriv, *intf_tmp;
+
+ list_for_each_entry_safe(intf_modpriv, intf_tmp,
+ &netcp_intf->module_head,
+ intf_list) {
+ if (intf_modpriv->netcp_module == module) {
+ module->release(intf_modpriv->module_priv);
+ list_del(&intf_modpriv->intf_list);
+ devm_kfree(dev, intf_modpriv);
+ break;
+ }
+ }
+ }
+
+ /* Remove the module from each instance */
+ list_for_each_entry_safe(inst_modpriv, inst_tmp,
+ &netcp_device->modpriv_head, inst_list) {
+ if (inst_modpriv->netcp_module == module) {
+ module->remove(netcp_device,
+ inst_modpriv->module_priv);
+ list_del(&inst_modpriv->inst_list);
+ devm_kfree(dev, inst_modpriv);
+ break;
+ }
+ }
+}
+
+void netcp_unregister_module(struct netcp_module *module)
+{
+ struct netcp_device *netcp_device;
+ struct netcp_module *module_tmp;
+
+ mutex_lock(&netcp_modules_lock);
+
+ list_for_each_entry(netcp_device, &netcp_devices, device_list) {
+ netcp_release_module(netcp_device, module);
+ }
+
+ /* Remove the module from the module list */
+ for_each_netcp_module(module_tmp) {
+ if (module == module_tmp) {
+ list_del(&module->module_list);
+ break;
+ }
+ }
+
+ mutex_unlock(&netcp_modules_lock);
+}
+
+void *netcp_module_get_intf_data(struct netcp_module *module,
+ struct netcp_intf *intf)
+{
+ struct netcp_intf_modpriv *intf_modpriv;
+
+ list_for_each_entry(intf_modpriv, &intf->module_head, intf_list)
+ if (intf_modpriv->netcp_module == module)
+ return intf_modpriv->module_priv;
+ return NULL;
+}
+
+/* Module TX and RX Hook management */
+struct netcp_hook_list {
+ struct list_head list;
+ netcp_hook_rtn *hook_rtn;
+ void *hook_data;
+ int order;
+};
+
+int netcp_register_txhook(struct netcp_intf *netcp_priv, int order,
+ netcp_hook_rtn *hook_rtn, void *hook_data)
+{
+ struct netcp_hook_list *entry;
+ struct netcp_hook_list *next;
+ unsigned long flags;
+
+ entry = devm_kzalloc(netcp_priv->dev, sizeof(*entry), GFP_KERNEL);
+ if (!entry)
+ return -ENOMEM;
+
+ entry->hook_rtn = hook_rtn;
+ entry->hook_data = hook_data;
+ entry->order = order;
+
+ spin_lock_irqsave(&netcp_priv->lock, flags);
+ list_for_each_entry(next, &netcp_priv->txhook_list_head, list) {
+ if (next->order > order)
+ break;
+ }
+ __list_add(&entry->list, next->list.prev, &next->list);
+ spin_unlock_irqrestore(&netcp_priv->lock, flags);
+
+ return 0;
+}
+
+int netcp_unregister_txhook(struct netcp_intf *netcp_priv, int order,
+ netcp_hook_rtn *hook_rtn, void *hook_data)
+{
+ struct netcp_hook_list *next, *n;
+ unsigned long flags;
+
+ spin_lock_irqsave(&netcp_priv->lock, flags);
+ list_for_each_entry_safe(next, n, &netcp_priv->txhook_list_head, list) {
+ if ((next->order == order) &&
+ (next->hook_rtn == hook_rtn) &&
+ (next->hook_data == hook_data)) {
+ list_del(&next->list);
+ spin_unlock_irqrestore(&netcp_priv->lock, flags);
+ devm_kfree(netcp_priv->dev, next);
+ return 0;
+ }
+ }
+ spin_unlock_irqrestore(&netcp_priv->lock, flags);
+ return -ENOENT;
+}
+
+int netcp_register_rxhook(struct netcp_intf *netcp_priv, int order,
+ netcp_hook_rtn *hook_rtn, void *hook_data)
+{
+ struct netcp_hook_list *entry;
+ struct netcp_hook_list *next;
+ unsigned long flags;
+
+ entry = devm_kzalloc(netcp_priv->dev, sizeof(*entry), GFP_KERNEL);
+ if (!entry)
+ return -ENOMEM;
+
+ entry->hook_rtn = hook_rtn;
+ entry->hook_data = hook_data;
+ entry->order = order;
+
+ spin_lock_irqsave(&netcp_priv->lock, flags);
+ list_for_each_entry(next, &netcp_priv->rxhook_list_head, list) {
+ if (next->order > order)
+ break;
+ }
+ __list_add(&entry->list, next->list.prev, &next->list);
+ spin_unlock_irqrestore(&netcp_priv->lock, flags);
+
+ return 0;
+}
+
+int netcp_unregister_rxhook(struct netcp_intf *netcp_priv, int order,
+ netcp_hook_rtn *hook_rtn, void *hook_data)
+{
+ struct netcp_hook_list *next, *n;
+ unsigned long flags;
+
+ spin_lock_irqsave(&netcp_priv->lock, flags);
+ list_for_each_entry_safe(next, n, &netcp_priv->rxhook_list_head, list) {
+ if ((next->order == order) &&
+ (next->hook_rtn == hook_rtn) &&
+ (next->hook_data == hook_data)) {
+ list_del(&next->list);
+ spin_unlock_irqrestore(&netcp_priv->lock, flags);
+ devm_kfree(netcp_priv->dev, next);
+ return 0;
+ }
+ }
+ spin_unlock_irqrestore(&netcp_priv->lock, flags);
+
+ return -ENOENT;
+}
+
+static void netcp_frag_free(bool is_frag, void *ptr)
+{
+ if (is_frag)
+ put_page(virt_to_head_page(ptr));
+ else
+ kfree(ptr);
+}
+
+static void netcp_free_rx_desc_chain(struct netcp_intf *netcp,
+ struct knav_dma_desc *desc)
+{
+ struct knav_dma_desc *ndesc;
+ dma_addr_t dma_desc, dma_buf;
+ unsigned int buf_len, dma_sz = sizeof(*ndesc);
+ void *buf_ptr;
+ u32 tmp;
+
+ get_words(&dma_desc, 1, &desc->next_desc);
+
+ while (dma_desc) {
+ ndesc = knav_pool_desc_unmap(netcp->rx_pool, dma_desc, dma_sz);
+ if (unlikely(!ndesc)) {
+ dev_err(netcp->ndev_dev, "failed to unmap Rx desc\n");
+ break;
+ }
+ get_pkt_info(&dma_buf, &tmp, &dma_desc, ndesc);
+ get_pad_info((u32 *)&buf_ptr, &tmp, ndesc);
+ dma_unmap_page(netcp->dev, dma_buf, PAGE_SIZE, DMA_FROM_DEVICE);
+ __free_page(buf_ptr);
+ knav_pool_desc_put(netcp->rx_pool, desc);
+ }
+
+ get_pad_info((u32 *)&buf_ptr, &buf_len, desc);
+ if (buf_ptr)
+ netcp_frag_free(buf_len <= PAGE_SIZE, buf_ptr);
+ knav_pool_desc_put(netcp->rx_pool, desc);
+}
+
+static void netcp_empty_rx_queue(struct netcp_intf *netcp)
+{
+ struct knav_dma_desc *desc;
+ unsigned int dma_sz;
+ dma_addr_t dma;
+
+ for (; ;) {
+ dma = knav_queue_pop(netcp->rx_queue, &dma_sz);
+ if (!dma)
+ break;
+
+ desc = knav_pool_desc_unmap(netcp->rx_pool, dma, dma_sz);
+ if (unlikely(!desc)) {
+ dev_err(netcp->ndev_dev, "%s: failed to unmap Rx desc\n",
+ __func__);
+ netcp->ndev->stats.rx_errors++;
+ continue;
+ }
+ netcp_free_rx_desc_chain(netcp, desc);
+ netcp->ndev->stats.rx_dropped++;
+ }
+}
+
+static int netcp_process_one_rx_packet(struct netcp_intf *netcp)
+{
+ unsigned int dma_sz, buf_len, org_buf_len;
+ struct knav_dma_desc *desc, *ndesc;
+ unsigned int pkt_sz = 0, accum_sz;
+ struct netcp_hook_list *rx_hook;
+ dma_addr_t dma_desc, dma_buff;
+ struct netcp_packet p_info;
+ struct sk_buff *skb;
+ void *org_buf_ptr;
+ u32 tmp;
+
+ dma_desc = knav_queue_pop(netcp->rx_queue, &dma_sz);
+ if (!dma_desc)
+ return -1;
+
+ desc = knav_pool_desc_unmap(netcp->rx_pool, dma_desc, dma_sz);
+ if (unlikely(!desc)) {
+ dev_err(netcp->ndev_dev, "failed to unmap Rx desc\n");
+ return 0;
+ }
+
+ get_pkt_info(&dma_buff, &buf_len, &dma_desc, desc);
+ get_pad_info((u32 *)&org_buf_ptr, &org_buf_len, desc);
+
+ if (unlikely(!org_buf_ptr)) {
+ dev_err(netcp->ndev_dev, "NULL bufptr in desc\n");
+ goto free_desc;
+ }
+
+ pkt_sz &= KNAV_DMA_DESC_PKT_LEN_MASK;
+ accum_sz = buf_len;
+ dma_unmap_single(netcp->dev, dma_buff, buf_len, DMA_FROM_DEVICE);
+
+ /* Build a new sk_buff for the primary buffer */
+ skb = build_skb(org_buf_ptr, org_buf_len);
+ if (unlikely(!skb)) {
+ dev_err(netcp->ndev_dev, "build_skb() failed\n");
+ goto free_desc;
+ }
+
+ /* update data, tail and len */
+ skb_reserve(skb, NETCP_SOP_OFFSET);
+ __skb_put(skb, buf_len);
+
+ /* Fill in the page fragment list */
+ while (dma_desc) {
+ struct page *page;
+
+ ndesc = knav_pool_desc_unmap(netcp->rx_pool, dma_desc, dma_sz);
+ if (unlikely(!ndesc)) {
+ dev_err(netcp->ndev_dev, "failed to unmap Rx desc\n");
+ goto free_desc;
+ }
+
+ get_pkt_info(&dma_buff, &buf_len, &dma_desc, ndesc);
+ get_pad_info((u32 *)&page, &tmp, ndesc);
+
+ if (likely(dma_buff && buf_len && page)) {
+ dma_unmap_page(netcp->dev, dma_buff, PAGE_SIZE,
+ DMA_FROM_DEVICE);
+ } else {
+ dev_err(netcp->ndev_dev, "Bad Rx desc dma_buff(%p), len(%d), page(%p)\n",
+ (void *)dma_buff, buf_len, page);
+ goto free_desc;
+ }
+
+ skb_add_rx_frag(skb, skb_shinfo(skb)->nr_frags, page,
+ offset_in_page(dma_buff), buf_len, PAGE_SIZE);
+ accum_sz += buf_len;
+
+ /* Free the descriptor */
+ knav_pool_desc_put(netcp->rx_pool, ndesc);
+ }
+
+ /* Free the primary descriptor */
+ knav_pool_desc_put(netcp->rx_pool, desc);
+
+ /* check for packet len and warn */
+ if (unlikely(pkt_sz != accum_sz))
+ dev_dbg(netcp->ndev_dev, "mismatch in packet size(%d) & sum of fragments(%d)\n",
+ pkt_sz, accum_sz);
+
+ /* Remove ethernet FCS from the packet */
+ __pskb_trim(skb, skb->len - ETH_FCS_LEN);
+
+ /* Call each of the RX hooks */
+ p_info.skb = skb;
+ p_info.rxtstamp_complete = false;
+ list_for_each_entry(rx_hook, &netcp->rxhook_list_head, list) {
+ int ret;
+
+ ret = rx_hook->hook_rtn(rx_hook->order, rx_hook->hook_data,
+ &p_info);
+ if (unlikely(ret)) {
+ dev_err(netcp->ndev_dev, "RX hook %d failed: %d\n",
+ rx_hook->order, ret);
+ netcp->ndev->stats.rx_errors++;
+ dev_kfree_skb(skb);
+ return 0;
+ }
+ }
+
+ netcp->ndev->last_rx = jiffies;
+ netcp->ndev->stats.rx_packets++;
+ netcp->ndev->stats.rx_bytes += skb->len;
+
+ /* push skb up the stack */
+ skb->protocol = eth_type_trans(skb, netcp->ndev);
+ netif_receive_skb(skb);
+ return 0;
+
+free_desc:
+ netcp_free_rx_desc_chain(netcp, desc);
+ netcp->ndev->stats.rx_errors++;
+ return 0;
+}
+
+static int netcp_process_rx_packets(struct netcp_intf *netcp,
+ unsigned int budget)
+{
+ int i;
+
+ for (i = 0; (i < budget) && !netcp_process_one_rx_packet(netcp); i++)
+ ;
+ return i;
+}
+
+/* Release descriptors and attached buffers from Rx FDQ */
+static void netcp_free_rx_buf(struct netcp_intf *netcp, int fdq)
+{
+ struct knav_dma_desc *desc;
+ unsigned int buf_len, dma_sz;
+ dma_addr_t dma;
+ void *buf_ptr;
+ u32 tmp;
+
+ /* Allocate descriptor */
+ while ((dma = knav_queue_pop(netcp->rx_fdq[fdq], &dma_sz))) {
+ desc = knav_pool_desc_unmap(netcp->rx_pool, dma, dma_sz);
+ if (unlikely(!desc)) {
+ dev_err(netcp->ndev_dev, "failed to unmap Rx desc\n");
+ continue;
+ }
+
+ get_org_pkt_info(&dma, &buf_len, desc);
+ get_pad_info((u32 *)&buf_ptr, &tmp, desc);
+
+ if (unlikely(!dma)) {
+ dev_err(netcp->ndev_dev, "NULL orig_buff in desc\n");
+ knav_pool_desc_put(netcp->rx_pool, desc);
+ continue;
+ }
+
+ if (unlikely(!buf_ptr)) {
+ dev_err(netcp->ndev_dev, "NULL bufptr in desc\n");
+ knav_pool_desc_put(netcp->rx_pool, desc);
+ continue;
+ }
+
+ if (fdq == 0) {
+ dma_unmap_single(netcp->dev, dma, buf_len,
+ DMA_FROM_DEVICE);
+ netcp_frag_free((buf_len <= PAGE_SIZE), buf_ptr);
+ } else {
+ dma_unmap_page(netcp->dev, dma, buf_len,
+ DMA_FROM_DEVICE);
+ __free_page(buf_ptr);
+ }
+
+ knav_pool_desc_put(netcp->rx_pool, desc);
+ }
+}
+
+static void netcp_rxpool_free(struct netcp_intf *netcp)
+{
+ int i;
+
+ for (i = 0; i < KNAV_DMA_FDQ_PER_CHAN &&
+ !IS_ERR_OR_NULL(netcp->rx_fdq[i]); i++)
+ netcp_free_rx_buf(netcp, i);
+
+ if (knav_pool_count(netcp->rx_pool) != netcp->rx_pool_size)
+ dev_err(netcp->ndev_dev, "Lost Rx (%d) descriptors\n",
+ netcp->rx_pool_size - knav_pool_count(netcp->rx_pool));
+
+ knav_pool_destroy(netcp->rx_pool);
+ netcp->rx_pool = NULL;
+}
+
+static void netcp_allocate_rx_buf(struct netcp_intf *netcp, int fdq)
+{
+ struct knav_dma_desc *hwdesc;
+ unsigned int buf_len, dma_sz;
+ u32 desc_info, pkt_info;
+ struct page *page;
+ dma_addr_t dma;
+ void *bufptr;
+ u32 pad[2];
+
+ /* Allocate descriptor */
+ hwdesc = knav_pool_desc_get(netcp->rx_pool);
+ if (IS_ERR_OR_NULL(hwdesc)) {
+ dev_dbg(netcp->ndev_dev, "out of rx pool desc\n");
+ return;
+ }
+
+ if (likely(fdq == 0)) {
+ unsigned int primary_buf_len;
+ /* Allocate a primary receive queue entry */
+ buf_len = netcp->rx_buffer_sizes[0] + NETCP_SOP_OFFSET;
+ primary_buf_len = SKB_DATA_ALIGN(buf_len) +
+ SKB_DATA_ALIGN(sizeof(struct skb_shared_info));
+
+ if (primary_buf_len <= PAGE_SIZE) {
+ bufptr = netdev_alloc_frag(primary_buf_len);
+ pad[1] = primary_buf_len;
+ } else {
+ bufptr = kmalloc(primary_buf_len, GFP_ATOMIC |
+ GFP_DMA32 | __GFP_COLD);
+ pad[1] = 0;
+ }
+
+ if (unlikely(!bufptr)) {
+ dev_warn_ratelimited(netcp->ndev_dev, "Primary RX buffer alloc failed\n");
+ goto fail;
+ }
+ dma = dma_map_single(netcp->dev, bufptr, buf_len,
+ DMA_TO_DEVICE);
+ pad[0] = (u32)bufptr;
+
+ } else {
+ /* Allocate a secondary receive queue entry */
+ page = alloc_page(GFP_ATOMIC | GFP_DMA32 | __GFP_COLD);
+ if (unlikely(!page)) {
+ dev_warn_ratelimited(netcp->ndev_dev, "Secondary page alloc failed\n");
+ goto fail;
+ }
+ buf_len = PAGE_SIZE;
+ dma = dma_map_page(netcp->dev, page, 0, buf_len, DMA_TO_DEVICE);
+ pad[0] = (u32)page;
+ pad[1] = 0;
+ }
+
+ desc_info = KNAV_DMA_DESC_PS_INFO_IN_DESC;
+ desc_info |= buf_len & KNAV_DMA_DESC_PKT_LEN_MASK;
+ pkt_info = KNAV_DMA_DESC_HAS_EPIB;
+ pkt_info |= KNAV_DMA_NUM_PS_WORDS << KNAV_DMA_DESC_PSLEN_SHIFT;
+ pkt_info |= (netcp->rx_queue_id & KNAV_DMA_DESC_RETQ_MASK) <<
+ KNAV_DMA_DESC_RETQ_SHIFT;
+ set_org_pkt_info(dma, buf_len, hwdesc);
+ set_pad_info(pad[0], pad[1], hwdesc);
+ set_desc_info(desc_info, pkt_info, hwdesc);
+
+ /* Push to FDQs */
+ knav_pool_desc_map(netcp->rx_pool, hwdesc, sizeof(*hwdesc), &dma,
+ &dma_sz);
+ knav_queue_push(netcp->rx_fdq[fdq], dma, sizeof(*hwdesc), 0);
+ return;
+
+fail:
+ knav_pool_desc_put(netcp->rx_pool, hwdesc);
+}
+
+/* Refill Rx FDQ with descriptors & attached buffers */
+static void netcp_rxpool_refill(struct netcp_intf *netcp)
+{
+ u32 fdq_deficit[KNAV_DMA_FDQ_PER_CHAN] = {0};
+ int i;
+
+ /* Calculate the FDQ deficit and refill */
+ for (i = 0; i < KNAV_DMA_FDQ_PER_CHAN && netcp->rx_fdq[i]; i++) {
+ fdq_deficit[i] = netcp->rx_queue_depths[i] -
+ knav_queue_get_count(netcp->rx_fdq[i]);
+
+ while (fdq_deficit[i]--)
+ netcp_allocate_rx_buf(netcp, i);
+ } /* end for fdqs */
+}
+
+/* NAPI poll */
+static int netcp_rx_poll(struct napi_struct *napi, int budget)
+{
+ struct netcp_intf *netcp = container_of(napi, struct netcp_intf,
+ rx_napi);
+ unsigned int packets;
+
+ packets = netcp_process_rx_packets(netcp, budget);
+
+ if (packets < budget) {
+ napi_complete(&netcp->rx_napi);
+ knav_queue_enable_notify(netcp->rx_queue);
+ }
+
+ netcp_rxpool_refill(netcp);
+ return packets;
+}
+
+static void netcp_rx_notify(void *arg)
+{
+ struct netcp_intf *netcp = arg;
+
+ knav_queue_disable_notify(netcp->rx_queue);
+ napi_schedule(&netcp->rx_napi);
+}
+
+static void netcp_free_tx_desc_chain(struct netcp_intf *netcp,
+ struct knav_dma_desc *desc,
+ unsigned int desc_sz)
+{
+ struct knav_dma_desc *ndesc = desc;
+ dma_addr_t dma_desc, dma_buf;
+ unsigned int buf_len;
+
+ while (ndesc) {
+ get_pkt_info(&dma_buf, &buf_len, &dma_desc, ndesc);
+
+ if (dma_buf && buf_len)
+ dma_unmap_single(netcp->dev, dma_buf, buf_len,
+ DMA_TO_DEVICE);
+ else
+ dev_warn(netcp->ndev_dev, "bad Tx desc buf(%p), len(%d)\n",
+ (void *)dma_buf, buf_len);
+
+ knav_pool_desc_put(netcp->tx_pool, ndesc);
+ ndesc = NULL;
+ if (dma_desc) {
+ ndesc = knav_pool_desc_unmap(netcp->tx_pool, dma_desc,
+ desc_sz);
+ if (!ndesc)
+ dev_err(netcp->ndev_dev, "failed to unmap Tx desc\n");
+ }
+ }
+}
+
+static int netcp_process_tx_compl_packets(struct netcp_intf *netcp,
+ unsigned int budget)
+{
+ struct knav_dma_desc *desc;
+ struct sk_buff *skb;
+ unsigned int dma_sz;
+ dma_addr_t dma;
+ int pkts = 0;
+ u32 tmp;
+
+ while (budget--) {
+ dma = knav_queue_pop(netcp->tx_compl_q, &dma_sz);
+ if (!dma)
+ break;
+ desc = knav_pool_desc_unmap(netcp->tx_pool, dma, dma_sz);
+ if (unlikely(!desc)) {
+ dev_err(netcp->ndev_dev, "failed to unmap Tx desc\n");
+ netcp->ndev->stats.tx_errors++;
+ continue;
+ }
+
+ get_pad_info((u32 *)&skb, &tmp, desc);
+ netcp_free_tx_desc_chain(netcp, desc, dma_sz);
+ if (!skb) {
+ dev_err(netcp->ndev_dev, "No skb in Tx desc\n");
+ netcp->ndev->stats.tx_errors++;
+ continue;
+ }
+
+ if (netif_subqueue_stopped(netcp->ndev, skb) &&
+ netif_running(netcp->ndev) &&
+ (knav_pool_count(netcp->tx_pool) >
+ netcp->tx_resume_threshold)) {
+ u16 subqueue = skb_get_queue_mapping(skb);
+
+ netif_wake_subqueue(netcp->ndev, subqueue);
+ }
+
+ netcp->ndev->stats.tx_packets++;
+ netcp->ndev->stats.tx_bytes += skb->len;
+ dev_kfree_skb(skb);
+ pkts++;
+ }
+ return pkts;
+}
+
+static int netcp_tx_poll(struct napi_struct *napi, int budget)
+{
+ int packets;
+ struct netcp_intf *netcp = container_of(napi, struct netcp_intf,
+ tx_napi);
+
+ packets = netcp_process_tx_compl_packets(netcp, budget);
+ if (packets < budget) {
+ napi_complete(&netcp->tx_napi);
+ knav_queue_enable_notify(netcp->tx_compl_q);
+ }
+
+ return packets;
+}
+
+static void netcp_tx_notify(void *arg)
+{
+ struct netcp_intf *netcp = arg;
+
+ knav_queue_disable_notify(netcp->tx_compl_q);
+ napi_schedule(&netcp->tx_napi);
+}
+
+static struct knav_dma_desc*
+netcp_tx_map_skb(struct sk_buff *skb, struct netcp_intf *netcp)
+{
+ struct knav_dma_desc *desc, *ndesc, *pdesc;
+ unsigned int pkt_len = skb_headlen(skb);
+ struct device *dev = netcp->dev;
+ dma_addr_t dma_addr;
+ unsigned int dma_sz;
+ int i;
+
+ /* Map the linear buffer */
+ dma_addr = dma_map_single(dev, skb->data, pkt_len, DMA_TO_DEVICE);
+ if (unlikely(!dma_addr)) {
+ dev_err(netcp->ndev_dev, "Failed to map skb buffer\n");
+ return NULL;
+ }
+
+ desc = knav_pool_desc_get(netcp->tx_pool);
+ if (unlikely(IS_ERR_OR_NULL(desc))) {
+ dev_err(netcp->ndev_dev, "out of TX desc\n");
+ dma_unmap_single(dev, dma_addr, pkt_len, DMA_TO_DEVICE);
+ return NULL;
+ }
+
+ set_pkt_info(dma_addr, pkt_len, 0, desc);
+ if (skb_is_nonlinear(skb)) {
+ prefetchw(skb_shinfo(skb));
+ } else {
+ desc->next_desc = 0;
+ goto upd_pkt_len;
+ }
+
+ pdesc = desc;
+
+ /* Handle the case where skb is fragmented in pages */
+ for (i = 0; i < skb_shinfo(skb)->nr_frags; i++) {
+ skb_frag_t *frag = &skb_shinfo(skb)->frags[i];
+ struct page *page = skb_frag_page(frag);
+ u32 page_offset = frag->page_offset;
+ u32 buf_len = skb_frag_size(frag);
+ dma_addr_t desc_dma;
+ u32 pkt_info;
+
+ dma_addr = dma_map_page(dev, page, page_offset, buf_len,
+ DMA_TO_DEVICE);
+ if (unlikely(!dma_addr)) {
+ dev_err(netcp->ndev_dev, "Failed to map skb page\n");
+ goto free_descs;
+ }
+
+ ndesc = knav_pool_desc_get(netcp->tx_pool);
+ if (unlikely(IS_ERR_OR_NULL(ndesc))) {
+ dev_err(netcp->ndev_dev, "out of TX desc for frags\n");
+ dma_unmap_page(dev, dma_addr, buf_len, DMA_TO_DEVICE);
+ goto free_descs;
+ }
+
+ desc_dma = knav_pool_desc_virt_to_dma(netcp->tx_pool,
+ (void *)ndesc);
+ pkt_info =
+ (netcp->tx_compl_qid & KNAV_DMA_DESC_RETQ_MASK) <<
+ KNAV_DMA_DESC_RETQ_SHIFT;
+ set_pkt_info(dma_addr, buf_len, 0, ndesc);
+ set_words(&desc_dma, 1, &pdesc->next_desc);
+ pkt_len += buf_len;
+ if (pdesc != desc)
+ knav_pool_desc_map(netcp->tx_pool, pdesc,
+ sizeof(*pdesc), &desc_dma, &dma_sz);
+ pdesc = ndesc;
+ }
+ if (pdesc != desc)
+ knav_pool_desc_map(netcp->tx_pool, pdesc, sizeof(*pdesc),
+ &dma_addr, &dma_sz);
+
+ /* frag list based linkage is not supported for now. */
+ if (skb_shinfo(skb)->frag_list) {
+ dev_err_ratelimited(netcp->ndev_dev, "NETIF_F_FRAGLIST not supported\n");
+ goto free_descs;
+ }
+
+upd_pkt_len:
+ WARN_ON(pkt_len != skb->len);
+
+ pkt_len &= KNAV_DMA_DESC_PKT_LEN_MASK;
+ set_words(&pkt_len, 1, &desc->desc_info);
+ return desc;
+
+free_descs:
+ netcp_free_tx_desc_chain(netcp, desc, sizeof(*desc));
+ return NULL;
+}
+
+static int netcp_tx_submit_skb(struct netcp_intf *netcp,
+ struct sk_buff *skb,
+ struct knav_dma_desc *desc)
+{
+ struct netcp_tx_pipe *tx_pipe = NULL;
+ struct netcp_hook_list *tx_hook;
+ struct netcp_packet p_info;
+ u32 packet_info = 0;
+ unsigned int dma_sz;
+ dma_addr_t dma;
+ int ret = 0;
+
+ p_info.netcp = netcp;
+ p_info.skb = skb;
+ p_info.tx_pipe = NULL;
+ p_info.psdata_len = 0;
+ p_info.ts_context = NULL;
+ p_info.txtstamp_complete = NULL;
+ p_info.epib = desc->epib;
+ p_info.psdata = desc->psdata;
+ memset(p_info.epib, 0, KNAV_DMA_NUM_EPIB_WORDS * sizeof(u32));
+
+ /* Find out where to inject the packet for transmission */
+ list_for_each_entry(tx_hook, &netcp->txhook_list_head, list) {
+ ret = tx_hook->hook_rtn(tx_hook->order, tx_hook->hook_data,
+ &p_info);
+ if (unlikely(ret != 0)) {
+ dev_err(netcp->ndev_dev, "TX hook %d rejected the packet with reason(%d)\n",
+ tx_hook->order, ret);
+ ret = (ret < 0) ? ret : NETDEV_TX_OK;
+ goto out;
+ }
+ }
+
+ /* Make sure some TX hook claimed the packet */
+ tx_pipe = p_info.tx_pipe;
+ if (!tx_pipe) {
+ dev_err(netcp->ndev_dev, "No TX hook claimed the packet!\n");
+ ret = -ENXIO;
+ goto out;
+ }
+
+ /* update descriptor */
+ if (p_info.psdata_len) {
+ u32 *psdata = p_info.psdata;
+
+ memmove(p_info.psdata, p_info.psdata + p_info.psdata_len,
+ p_info.psdata_len);
+ set_words(psdata, p_info.psdata_len, psdata);
+ packet_info |=
+ (p_info.psdata_len & KNAV_DMA_DESC_PSLEN_MASK) <<
+ KNAV_DMA_DESC_PSLEN_SHIFT;
+ }
+
+ packet_info |= KNAV_DMA_DESC_HAS_EPIB |
+ ((netcp->tx_compl_qid & KNAV_DMA_DESC_RETQ_MASK) <<
+ KNAV_DMA_DESC_RETQ_SHIFT) |
+ ((tx_pipe->dma_psflags & KNAV_DMA_DESC_PSFLAG_MASK) <<
+ KNAV_DMA_DESC_PSFLAG_SHIFT);
+
+ set_words(&packet_info, 1, &desc->packet_info);
+ set_words((u32 *)&skb, 1, &desc->pad[0]);
+
+ /* submit packet descriptor */
+ ret = knav_pool_desc_map(netcp->tx_pool, desc, sizeof(*desc), &dma,
+ &dma_sz);
+ if (unlikely(ret)) {
+ dev_err(netcp->ndev_dev, "%s() failed to map desc\n", __func__);
+ ret = -ENOMEM;
+ goto out;
+ }
+ skb_tx_timestamp(skb);
+ knav_queue_push(tx_pipe->dma_queue, dma, dma_sz, 0);
+
+out:
+ return ret;
+}
+
+/* Submit the packet */
+static int netcp_ndo_start_xmit(struct sk_buff *skb, struct net_device *ndev)
+{
+ struct netcp_intf *netcp = netdev_priv(ndev);
+ int subqueue = skb_get_queue_mapping(skb);
+ struct knav_dma_desc *desc;
+ int desc_count, ret = 0;
+
+ if (unlikely(skb->len <= 0)) {
+ dev_kfree_skb(skb);
+ return NETDEV_TX_OK;
+ }
+
+ if (unlikely(skb->len < NETCP_MIN_PACKET_SIZE)) {
+ ret = skb_padto(skb, NETCP_MIN_PACKET_SIZE);
+ if (ret < 0) {
+ /* If we get here, the skb has already been dropped */
+ dev_warn(netcp->ndev_dev, "padding failed (%d), packet dropped\n",
+ ret);
+ ndev->stats.tx_dropped++;
+ return ret;
+ }
+ skb->len = NETCP_MIN_PACKET_SIZE;
+ }
+
+ desc = netcp_tx_map_skb(skb, netcp);
+ if (unlikely(!desc)) {
+ netif_stop_subqueue(ndev, subqueue);
+ ret = -ENOBUFS;
+ goto drop;
+ }
+
+ ret = netcp_tx_submit_skb(netcp, skb, desc);
+ if (ret)
+ goto drop;
+
+ ndev->trans_start = jiffies;
+
+ /* Check Tx pool count & stop subqueue if needed */
+ desc_count = knav_pool_count(netcp->tx_pool);
+ if (desc_count < netcp->tx_pause_threshold) {
+ dev_dbg(netcp->ndev_dev, "pausing tx, count(%d)\n", desc_count);
+ netif_stop_subqueue(ndev, subqueue);
+ }
+ return NETDEV_TX_OK;
+
+drop:
+ ndev->stats.tx_dropped++;
+ if (desc)
+ netcp_free_tx_desc_chain(netcp, desc, sizeof(*desc));
+ dev_kfree_skb(skb);
+ return ret;
+}
+
+int netcp_txpipe_close(struct netcp_tx_pipe *tx_pipe)
+{
+ if (tx_pipe->dma_channel) {
+ knav_dma_close_channel(tx_pipe->dma_channel);
+ tx_pipe->dma_channel = NULL;
+ }
+ return 0;
+}
+
+int netcp_txpipe_open(struct netcp_tx_pipe *tx_pipe)
+{
+ struct device *dev = tx_pipe->netcp_device->device;
+ struct knav_dma_cfg config;
+ int ret = 0;
+ u8 name[16];
+
+ memset(&config, 0, sizeof(config));
+ config.direction = DMA_MEM_TO_DEV;
+ config.u.tx.filt_einfo = false;
+ config.u.tx.filt_pswords = false;
+ config.u.tx.priority = DMA_PRIO_MED_L;
+
+ tx_pipe->dma_channel = knav_dma_open_channel(dev,
+ tx_pipe->dma_chan_name, &config);
+ if (IS_ERR_OR_NULL(tx_pipe->dma_channel)) {
+ dev_err(dev, "failed opening tx chan(%s)\n",
+ tx_pipe->dma_chan_name);
+ goto err;
+ }
+
+ snprintf(name, sizeof(name), "tx-pipe-%s", dev_name(dev));
+ tx_pipe->dma_queue = knav_queue_open(name, tx_pipe->dma_queue_id,
+ KNAV_QUEUE_SHARED);
+ if (IS_ERR(tx_pipe->dma_queue)) {
+ dev_err(dev, "Could not open DMA queue for channel \"%s\": %d\n",
+ name, ret);
+ ret = PTR_ERR(tx_pipe->dma_queue);
+ goto err;
+ }
+
+ dev_dbg(dev, "opened tx pipe %s\n", name);
+ return 0;
+
+err:
+ if (!IS_ERR_OR_NULL(tx_pipe->dma_channel))
+ knav_dma_close_channel(tx_pipe->dma_channel);
+ tx_pipe->dma_channel = NULL;
+ return ret;
+}
+
+int netcp_txpipe_init(struct netcp_tx_pipe *tx_pipe,
+ struct netcp_device *netcp_device,
+ const char *dma_chan_name, unsigned int dma_queue_id)
+{
+ memset(tx_pipe, 0, sizeof(*tx_pipe));
+ tx_pipe->netcp_device = netcp_device;
+ tx_pipe->dma_chan_name = dma_chan_name;
+ tx_pipe->dma_queue_id = dma_queue_id;
+ return 0;
+}
+
+static struct netcp_addr *netcp_addr_find(struct netcp_intf *netcp,
+ const u8 *addr,
+ enum netcp_addr_type type)
+{
+ struct netcp_addr *naddr;
+
+ list_for_each_entry(naddr, &netcp->addr_list, node) {
+ if (naddr->type != type)
+ continue;
+ if (addr && memcmp(addr, naddr->addr, ETH_ALEN))
+ continue;
+ return naddr;
+ }
+
+ return NULL;
+}
+
+static struct netcp_addr *netcp_addr_add(struct netcp_intf *netcp,
+ const u8 *addr,
+ enum netcp_addr_type type)
+{
+ struct netcp_addr *naddr;
+
+ naddr = devm_kmalloc(netcp->dev, sizeof(*naddr), GFP_ATOMIC);
+ if (!naddr)
+ return NULL;
+
+ naddr->type = type;
+ naddr->flags = 0;
+ naddr->netcp = netcp;
+ if (addr)
+ ether_addr_copy(naddr->addr, addr);
+ else
+ memset(naddr->addr, 0, ETH_ALEN);
+ list_add_tail(&naddr->node, &netcp->addr_list);
+
+ return naddr;
+}
+
+static void netcp_addr_del(struct netcp_intf *netcp, struct netcp_addr *naddr)
+{
+ list_del(&naddr->node);
+ devm_kfree(netcp->dev, naddr);
+}
+
+static void netcp_addr_clear_mark(struct netcp_intf *netcp)
+{
+ struct netcp_addr *naddr;
+
+ list_for_each_entry(naddr, &netcp->addr_list, node)
+ naddr->flags = 0;
+}
+
+static void netcp_addr_add_mark(struct netcp_intf *netcp, const u8 *addr,
+ enum netcp_addr_type type)
+{
+ struct netcp_addr *naddr;
+
+ naddr = netcp_addr_find(netcp, addr, type);
+ if (naddr) {
+ naddr->flags |= ADDR_VALID;
+ return;
+ }
+
+ naddr = netcp_addr_add(netcp, addr, type);
+ if (!WARN_ON(!naddr))
+ naddr->flags |= ADDR_NEW;
+}
+
+static void netcp_addr_sweep_del(struct netcp_intf *netcp)
+{
+ struct netcp_addr *naddr, *tmp;
+ struct netcp_intf_modpriv *priv;
+ struct netcp_module *module;
+ int error;
+
+ list_for_each_entry_safe(naddr, tmp, &netcp->addr_list, node) {
+ if (naddr->flags & (ADDR_VALID | ADDR_NEW))
+ continue;
+ dev_dbg(netcp->ndev_dev, "deleting address %pM, type %x\n",
+ naddr->addr, naddr->type);
+ mutex_lock(&netcp_modules_lock);
+ for_each_module(netcp, priv) {
+ module = priv->netcp_module;
+ if (!module->del_addr)
+ continue;
+ error = module->del_addr(priv->module_priv,
+ naddr);
+ WARN_ON(error);
+ }
+ mutex_unlock(&netcp_modules_lock);
+ netcp_addr_del(netcp, naddr);
+ }
+}
+
+static void netcp_addr_sweep_add(struct netcp_intf *netcp)
+{
+ struct netcp_addr *naddr, *tmp;
+ struct netcp_intf_modpriv *priv;
+ struct netcp_module *module;
+ int error;
+
+ list_for_each_entry_safe(naddr, tmp, &netcp->addr_list, node) {
+ if (!(naddr->flags & ADDR_NEW))
+ continue;
+ dev_dbg(netcp->ndev_dev, "adding address %pM, type %x\n",
+ naddr->addr, naddr->type);
+ mutex_lock(&netcp_modules_lock);
+ for_each_module(netcp, priv) {
+ module = priv->netcp_module;
+ if (!module->add_addr)
+ continue;
+ error = module->add_addr(priv->module_priv, naddr);
+ WARN_ON(error);
+ }
+ mutex_unlock(&netcp_modules_lock);
+ }
+}
+
+static void netcp_set_rx_mode(struct net_device *ndev)
+{
+ struct netcp_intf *netcp = netdev_priv(ndev);
+ struct netdev_hw_addr *ndev_addr;
+ bool promisc;
+
+ promisc = (ndev->flags & IFF_PROMISC ||
+ ndev->flags & IFF_ALLMULTI ||
+ netdev_mc_count(ndev) > NETCP_MAX_MCAST_ADDR);
+
+ /* first clear all marks */
+ netcp_addr_clear_mark(netcp);
+
+ /* next add new entries, mark existing ones */
+ netcp_addr_add_mark(netcp, ndev->broadcast, ADDR_BCAST);
+ for_each_dev_addr(ndev, ndev_addr)
+ netcp_addr_add_mark(netcp, ndev_addr->addr, ADDR_DEV);
+ netdev_for_each_uc_addr(ndev_addr, ndev)
+ netcp_addr_add_mark(netcp, ndev_addr->addr, ADDR_UCAST);
+ netdev_for_each_mc_addr(ndev_addr, ndev)
+ netcp_addr_add_mark(netcp, ndev_addr->addr, ADDR_MCAST);
+
+ if (promisc)
+ netcp_addr_add_mark(netcp, NULL, ADDR_ANY);
+
+ /* finally sweep and callout into modules */
+ netcp_addr_sweep_del(netcp);
+ netcp_addr_sweep_add(netcp);
+}
+
+static void netcp_free_navigator_resources(struct netcp_intf *netcp)
+{
+ int i;
+
+ if (netcp->rx_channel) {
+ knav_dma_close_channel(netcp->rx_channel);
+ netcp->rx_channel = NULL;
+ }
+
+ if (!IS_ERR_OR_NULL(netcp->rx_pool))
+ netcp_rxpool_free(netcp);
+
+ if (!IS_ERR_OR_NULL(netcp->rx_queue)) {
+ knav_queue_close(netcp->rx_queue);
+ netcp->rx_queue = NULL;
+ }
+
+ for (i = 0; i < KNAV_DMA_FDQ_PER_CHAN &&
+ !IS_ERR_OR_NULL(netcp->rx_fdq[i]) ; ++i) {
+ knav_queue_close(netcp->rx_fdq[i]);
+ netcp->rx_fdq[i] = NULL;
+ }
+
+ if (!IS_ERR_OR_NULL(netcp->tx_compl_q)) {
+ knav_queue_close(netcp->tx_compl_q);
+ netcp->tx_compl_q = NULL;
+ }
+
+ if (!IS_ERR_OR_NULL(netcp->tx_pool)) {
+ knav_pool_destroy(netcp->tx_pool);
+ netcp->tx_pool = NULL;
+ }
+}
+
+static int netcp_setup_navigator_resources(struct net_device *ndev)
+{
+ struct netcp_intf *netcp = netdev_priv(ndev);
+ struct knav_queue_notify_config notify_cfg;
+ struct knav_dma_cfg config;
+ u32 last_fdq = 0;
+ u8 name[16];
+ int ret;
+ int i;
+
+ /* Create Rx/Tx descriptor pools */
+ snprintf(name, sizeof(name), "rx-pool-%s", ndev->name);
+ netcp->rx_pool = knav_pool_create(name, netcp->rx_pool_size,
+ netcp->rx_pool_region_id);
+ if (IS_ERR_OR_NULL(netcp->rx_pool)) {
+ dev_err(netcp->ndev_dev, "Couldn't create rx pool\n");
+ ret = PTR_ERR(netcp->rx_pool);
+ goto fail;
+ }
+
+ snprintf(name, sizeof(name), "tx-pool-%s", ndev->name);
+ netcp->tx_pool = knav_pool_create(name, netcp->tx_pool_size,
+ netcp->tx_pool_region_id);
+ if (IS_ERR_OR_NULL(netcp->tx_pool)) {
+ dev_err(netcp->ndev_dev, "Couldn't create tx pool\n");
+ ret = PTR_ERR(netcp->tx_pool);
+ goto fail;
+ }
+
+ /* open Tx completion queue */
+ snprintf(name, sizeof(name), "tx-compl-%s", ndev->name);
+ netcp->tx_compl_q = knav_queue_open(name, netcp->tx_compl_qid, 0);
+ if (IS_ERR_OR_NULL(netcp->tx_compl_q)) {
+ ret = PTR_ERR(netcp->tx_compl_q);
+ goto fail;
+ }
+ netcp->tx_compl_qid = knav_queue_get_id(netcp->tx_compl_q);
+
+ /* Set notification for Tx completion */
+ notify_cfg.fn = netcp_tx_notify;
+ notify_cfg.fn_arg = netcp;
+ ret = knav_queue_device_control(netcp->tx_compl_q,
+ KNAV_QUEUE_SET_NOTIFIER,
+ (unsigned long)¬ify_cfg);
+ if (ret)
+ goto fail;
+
+ knav_queue_disable_notify(netcp->tx_compl_q);
+
+ /* open Rx completion queue */
+ snprintf(name, sizeof(name), "rx-compl-%s", ndev->name);
+ netcp->rx_queue = knav_queue_open(name, netcp->rx_queue_id, 0);
+ if (IS_ERR_OR_NULL(netcp->rx_queue)) {
+ ret = PTR_ERR(netcp->rx_queue);
+ goto fail;
+ }
+ netcp->rx_queue_id = knav_queue_get_id(netcp->rx_queue);
+
+ /* Set notification for Rx completion */
+ notify_cfg.fn = netcp_rx_notify;
+ notify_cfg.fn_arg = netcp;
+ ret = knav_queue_device_control(netcp->rx_queue,
+ KNAV_QUEUE_SET_NOTIFIER,
+ (unsigned long)¬ify_cfg);
+ if (ret)
+ goto fail;
+
+ knav_queue_disable_notify(netcp->rx_queue);
+
+ /* open Rx FDQs */
+ for (i = 0; i < KNAV_DMA_FDQ_PER_CHAN &&
+ netcp->rx_queue_depths[i] && netcp->rx_buffer_sizes[i]; ++i) {
+ snprintf(name, sizeof(name), "rx-fdq-%s-%d", ndev->name, i);
+ netcp->rx_fdq[i] = knav_queue_open(name, KNAV_QUEUE_GP, 0);
+ if (IS_ERR_OR_NULL(netcp->rx_fdq[i])) {
+ ret = PTR_ERR(netcp->rx_fdq[i]);
+ goto fail;
+ }
+ }
+
+ memset(&config, 0, sizeof(config));
+ config.direction = DMA_DEV_TO_MEM;
+ config.u.rx.einfo_present = true;
+ config.u.rx.psinfo_present = true;
+ config.u.rx.err_mode = DMA_DROP;
+ config.u.rx.desc_type = DMA_DESC_HOST;
+ config.u.rx.psinfo_at_sop = false;
+ config.u.rx.sop_offset = NETCP_SOP_OFFSET;
+ config.u.rx.dst_q = netcp->rx_queue_id;
+ config.u.rx.thresh = DMA_THRESH_NONE;
+
+ for (i = 0; i < KNAV_DMA_FDQ_PER_CHAN; ++i) {
+ if (netcp->rx_fdq[i])
+ last_fdq = knav_queue_get_id(netcp->rx_fdq[i]);
+ config.u.rx.fdq[i] = last_fdq;
+ }
+
+ netcp->rx_channel = knav_dma_open_channel(netcp->netcp_device->device,
+ netcp->dma_chan_name, &config);
+ if (IS_ERR_OR_NULL(netcp->rx_channel)) {
+ dev_err(netcp->ndev_dev, "failed opening rx chan(%s\n",
+ netcp->dma_chan_name);
+ goto fail;
+ }
+
+ dev_dbg(netcp->ndev_dev, "opened RX channel: %p\n", netcp->rx_channel);
+ return 0;
+
+fail:
+ netcp_free_navigator_resources(netcp);
+ return ret;
+}
+
+/* Open the device */
+static int netcp_ndo_open(struct net_device *ndev)
+{
+ struct netcp_intf *netcp = netdev_priv(ndev);
+ struct netcp_intf_modpriv *intf_modpriv;
+ struct netcp_module *module;
+ int ret;
+
+ netif_carrier_off(ndev);
+ ret = netcp_setup_navigator_resources(ndev);
+ if (ret) {
+ dev_err(netcp->ndev_dev, "Failed to setup navigator resources\n");
+ goto fail;
+ }
+
+ mutex_lock(&netcp_modules_lock);
+ for_each_module(netcp, intf_modpriv) {
+ module = intf_modpriv->netcp_module;
+ if (module->open) {
+ ret = module->open(intf_modpriv->module_priv, ndev);
+ if (ret != 0) {
+ dev_err(netcp->ndev_dev, "module open failed\n");
+ goto fail_open;
+ }
+ }
+ }
+ mutex_unlock(&netcp_modules_lock);
+
+ netcp_rxpool_refill(netcp);
+ napi_enable(&netcp->rx_napi);
+ napi_enable(&netcp->tx_napi);
+ knav_queue_enable_notify(netcp->tx_compl_q);
+ knav_queue_enable_notify(netcp->rx_queue);
+ netif_tx_wake_all_queues(ndev);
+ dev_dbg(netcp->ndev_dev, "netcp device %s opened\n", ndev->name);
+ return 0;
+
+fail_open:
+ for_each_module(netcp, intf_modpriv) {
+ module = intf_modpriv->netcp_module;
+ if (module->close)
+ module->close(intf_modpriv->module_priv, ndev);
+ }
+ mutex_unlock(&netcp_modules_lock);
+
+fail:
+ netcp_free_navigator_resources(netcp);
+ return ret;
+}
+
+/* Close the device */
+static int netcp_ndo_stop(struct net_device *ndev)
+{
+ struct netcp_intf *netcp = netdev_priv(ndev);
+ struct netcp_intf_modpriv *intf_modpriv;
+ struct netcp_module *module;
+ int err = 0;
+
+ netif_tx_stop_all_queues(ndev);
+ netif_carrier_off(ndev);
+ netcp_addr_clear_mark(netcp);
+ netcp_addr_sweep_del(netcp);
+ knav_queue_disable_notify(netcp->rx_queue);
+ knav_queue_disable_notify(netcp->tx_compl_q);
+ napi_disable(&netcp->rx_napi);
+ napi_disable(&netcp->tx_napi);
+
+ mutex_lock(&netcp_modules_lock);
+ for_each_module(netcp, intf_modpriv) {
+ module = intf_modpriv->netcp_module;
+ if (module->close) {
+ err = module->close(intf_modpriv->module_priv, ndev);
+ if (err != 0)
+ dev_err(netcp->ndev_dev, "Close failed\n");
+ }
+ }
+ mutex_unlock(&netcp_modules_lock);
+
+ /* Recycle Rx descriptors from completion queue */
+ netcp_empty_rx_queue(netcp);
+
+ /* Recycle Tx descriptors from completion queue */
+ netcp_process_tx_compl_packets(netcp, netcp->tx_pool_size);
+
+ if (knav_pool_count(netcp->tx_pool) != netcp->tx_pool_size)
+ dev_err(netcp->ndev_dev, "Lost (%d) Tx descs\n",
+ netcp->tx_pool_size - knav_pool_count(netcp->tx_pool));
+
+ netcp_free_navigator_resources(netcp);
+ dev_dbg(netcp->ndev_dev, "netcp device %s stopped\n", ndev->name);
+ return 0;
+}
+
+static int netcp_ndo_ioctl(struct net_device *ndev,
+ struct ifreq *req, int cmd)
+{
+ struct netcp_intf *netcp = netdev_priv(ndev);
+ struct netcp_intf_modpriv *intf_modpriv;
+ struct netcp_module *module;
+ int ret = -1, err = -EOPNOTSUPP;
+
+ if (!netif_running(ndev))
+ return -EINVAL;
+
+ mutex_lock(&netcp_modules_lock);
+ for_each_module(netcp, intf_modpriv) {
+ module = intf_modpriv->netcp_module;
+ if (!module->ioctl)
+ continue;
+
+ err = module->ioctl(intf_modpriv->module_priv, req, cmd);
+ if ((err < 0) && (err != -EOPNOTSUPP)) {
+ ret = err;
+ goto out;
+ }
+ if (err == 0)
+ ret = err;
+ }
+
+out:
+ mutex_unlock(&netcp_modules_lock);
+ return (ret == 0) ? 0 : err;
+}
+
+static int netcp_ndo_change_mtu(struct net_device *ndev, int new_mtu)
+{
+ struct netcp_intf *netcp = netdev_priv(ndev);
+
+ /* MTU < 68 is an error for IPv4 traffic */
+ if ((new_mtu < 68) ||
+ (new_mtu > (NETCP_MAX_FRAME_SIZE - ETH_HLEN - ETH_FCS_LEN))) {
+ dev_err(netcp->ndev_dev, "Invalid mtu size = %d\n", new_mtu);
+ return -EINVAL;
+ }
+
+ ndev->mtu = new_mtu;
+ return 0;
+}
+
+static void netcp_ndo_tx_timeout(struct net_device *ndev)
+{
+ struct netcp_intf *netcp = netdev_priv(ndev);
+ unsigned int descs = knav_pool_count(netcp->tx_pool);
+
+ dev_err(netcp->ndev_dev, "transmit timed out tx descs(%d)\n", descs);
+ netcp_process_tx_compl_packets(netcp, netcp->tx_pool_size);
+ ndev->trans_start = jiffies;
+ netif_tx_wake_all_queues(ndev);
+}
+
+static int netcp_rx_add_vid(struct net_device *ndev, __be16 proto, u16 vid)
+{
+ struct netcp_intf *netcp = netdev_priv(ndev);
+ struct netcp_intf_modpriv *intf_modpriv;
+ struct netcp_module *module;
+ int err = 0;
+
+ dev_dbg(netcp->ndev_dev, "adding rx vlan id: %d\n", vid);
+
+ mutex_lock(&netcp_modules_lock);
+ for_each_module(netcp, intf_modpriv) {
+ module = intf_modpriv->netcp_module;
+ if ((module->add_vid) && (vid != 0)) {
+ err = module->add_vid(intf_modpriv->module_priv, vid);
+ if (err != 0) {
+ dev_err(netcp->ndev_dev, "Could not add vlan id = %d\n",
+ vid);
+ break;
+ }
+ }
+ }
+ mutex_unlock(&netcp_modules_lock);
+ return err;
+}
+
+static int netcp_rx_kill_vid(struct net_device *ndev, __be16 proto, u16 vid)
+{
+ struct netcp_intf *netcp = netdev_priv(ndev);
+ struct netcp_intf_modpriv *intf_modpriv;
+ struct netcp_module *module;
+ int err = 0;
+
+ dev_dbg(netcp->ndev_dev, "removing rx vlan id: %d\n", vid);
+
+ mutex_lock(&netcp_modules_lock);
+ for_each_module(netcp, intf_modpriv) {
+ module = intf_modpriv->netcp_module;
+ if (module->del_vid) {
+ err = module->del_vid(intf_modpriv->module_priv, vid);
+ if (err != 0) {
+ dev_err(netcp->ndev_dev, "Could not delete vlan id = %d\n",
+ vid);
+ break;
+ }
+ }
+ }
+ mutex_unlock(&netcp_modules_lock);
+ return err;
+}
+
+static u16 netcp_select_queue(struct net_device *dev, struct sk_buff *skb,
+ void *accel_priv,
+ select_queue_fallback_t fallback)
+{
+ return 0;
+}
+
+static int netcp_setup_tc(struct net_device *dev, u8 num_tc)
+{
+ int i;
+
+ /* setup tc must be called under rtnl lock */
+ ASSERT_RTNL();
+
+ /* Sanity-check the number of traffic classes requested */
+ if ((dev->real_num_tx_queues <= 1) ||
+ (dev->real_num_tx_queues < num_tc))
+ return -EINVAL;
+
+ /* Configure traffic class to queue mappings */
+ if (num_tc) {
+ netdev_set_num_tc(dev, num_tc);
+ for (i = 0; i < num_tc; i++)
+ netdev_set_tc_queue(dev, i, 1, i);
+ } else {
+ netdev_reset_tc(dev);
+ }
+
+ return 0;
+}
+
+static const struct net_device_ops netcp_netdev_ops = {
+ .ndo_open = netcp_ndo_open,
+ .ndo_stop = netcp_ndo_stop,
+ .ndo_start_xmit = netcp_ndo_start_xmit,
+ .ndo_set_rx_mode = netcp_set_rx_mode,
+ .ndo_do_ioctl = netcp_ndo_ioctl,
+ .ndo_change_mtu = netcp_ndo_change_mtu,
+ .ndo_set_mac_address = eth_mac_addr,
+ .ndo_validate_addr = eth_validate_addr,
+ .ndo_vlan_rx_add_vid = netcp_rx_add_vid,
+ .ndo_vlan_rx_kill_vid = netcp_rx_kill_vid,
+ .ndo_tx_timeout = netcp_ndo_tx_timeout,
+ .ndo_select_queue = netcp_select_queue,
+ .ndo_setup_tc = netcp_setup_tc,
+};
+
+static int netcp_create_interface(struct netcp_device *netcp_device,
+ struct device_node *node_interface)
+{
+ struct device *dev = netcp_device->device;
+ struct device_node *node = dev->of_node;
+ struct netcp_intf *netcp;
+ struct net_device *ndev;
+ resource_size_t size;
+ struct resource res;
+ void __iomem *efuse = NULL;
+ u32 efuse_mac = 0;
+ const void *mac_addr;
+ u8 efuse_mac_addr[6];
+ u32 temp[2];
+ int ret = 0;
+
+ ndev = alloc_etherdev_mqs(sizeof(*netcp), 1, 1);
+ if (!ndev) {
+ dev_err(dev, "Error allocating netdev\n");
+ return -ENOMEM;
+ }
+
+ ndev->features |= NETIF_F_SG;
+ ndev->features |= NETIF_F_HW_VLAN_CTAG_FILTER;
+ ndev->hw_features = ndev->features;
+ ndev->vlan_features |= NETIF_F_SG;
+
+ netcp = netdev_priv(ndev);
+ spin_lock_init(&netcp->lock);
+ INIT_LIST_HEAD(&netcp->module_head);
+ INIT_LIST_HEAD(&netcp->txhook_list_head);
+ INIT_LIST_HEAD(&netcp->rxhook_list_head);
+ INIT_LIST_HEAD(&netcp->addr_list);
+ netcp->netcp_device = netcp_device;
+ netcp->dev = netcp_device->device;
+ netcp->ndev = ndev;
+ netcp->ndev_dev = &ndev->dev;
+ netcp->msg_enable = netif_msg_init(netcp_debug_level, NETCP_DEBUG);
+ netcp->tx_pause_threshold = MAX_SKB_FRAGS;
+ netcp->tx_resume_threshold = netcp->tx_pause_threshold;
+ netcp->node_interface = node_interface;
+
+ ret = of_property_read_u32(node_interface, "efuse-mac", &efuse_mac);
+ if (efuse_mac) {
+ if (of_address_to_resource(node, NETCP_EFUSE_REG_INDEX, &res)) {
+ dev_err(dev, "could not find efuse-mac reg resource\n");
+ ret = -ENODEV;
+ goto quit;
+ }
+ size = resource_size(&res);
+
+ if (!devm_request_mem_region(dev, res.start, size,
+ dev_name(dev))) {
+ dev_err(dev, "could not reserve resource\n");
+ ret = -ENOMEM;
+ goto quit;
+ }
+
+ efuse = devm_ioremap_nocache(dev, res.start, size);
+ if (!efuse) {
+ dev_err(dev, "could not map resource\n");
+ devm_release_mem_region(dev, res.start, size);
+ ret = -ENOMEM;
+ goto quit;
+ }
+
+ emac_arch_get_mac_addr(efuse_mac_addr, efuse);
+ if (is_valid_ether_addr(efuse_mac_addr))
+ ether_addr_copy(ndev->dev_addr, efuse_mac_addr);
+ else
+ random_ether_addr(ndev->dev_addr);
+
+ devm_iounmap(dev, efuse);
+ devm_release_mem_region(dev, res.start, size);
+ } else {
+ mac_addr = of_get_mac_address(node_interface);
+ if (mac_addr)
+ ether_addr_copy(ndev->dev_addr, mac_addr);
+ else
+ random_ether_addr(ndev->dev_addr);
+ }
+
+ ret = of_property_read_string(node_interface, "rx-channel",
+ &netcp->dma_chan_name);
+ if (ret < 0) {
+ dev_err(dev, "missing \"rx-channel\" parameter\n");
+ ret = -ENODEV;
+ goto quit;
+ }
+
+ ret = of_property_read_u32(node_interface, "rx-queue",
+ &netcp->rx_queue_id);
+ if (ret < 0) {
+ dev_warn(dev, "missing \"rx-queue\" parameter\n");
+ netcp->rx_queue_id = KNAV_QUEUE_QPEND;
+ }
+
+ ret = of_property_read_u32_array(node_interface, "rx-queue-depth",
+ netcp->rx_queue_depths,
+ KNAV_DMA_FDQ_PER_CHAN);
+ if (ret < 0) {
+ dev_err(dev, "missing \"rx-queue-depth\" parameter\n");
+ netcp->rx_queue_depths[0] = 128;
+ }
+
+ ret = of_property_read_u32_array(node_interface, "rx-buffer-size",
+ netcp->rx_buffer_sizes,
+ KNAV_DMA_FDQ_PER_CHAN);
+ if (ret) {
+ dev_err(dev, "missing \"rx-buffer-size\" parameter\n");
+ netcp->rx_buffer_sizes[0] = 1536;
+ }
+
+ ret = of_property_read_u32_array(node_interface, "rx-pool", temp, 2);
+ if (ret < 0) {
+ dev_err(dev, "missing \"rx-pool\" parameter\n");
+ ret = -ENODEV;
+ goto quit;
+ }
+ netcp->rx_pool_size = temp[0];
+ netcp->rx_pool_region_id = temp[1];
+
+ ret = of_property_read_u32_array(node_interface, "tx-pool", temp, 2);
+ if (ret < 0) {
+ dev_err(dev, "missing \"tx-pool\" parameter\n");
+ ret = -ENODEV;
+ goto quit;
+ }
+ netcp->tx_pool_size = temp[0];
+ netcp->tx_pool_region_id = temp[1];
+
+ if (netcp->tx_pool_size < MAX_SKB_FRAGS) {
+ dev_err(dev, "tx-pool size too small, must be atleast(%ld)\n",
+ MAX_SKB_FRAGS);
+ ret = -ENODEV;
+ goto quit;
+ }
+
+ ret = of_property_read_u32(node_interface, "tx-completion-queue",
+ &netcp->tx_compl_qid);
+ if (ret < 0) {
+ dev_warn(dev, "missing \"tx-completion-queue\" parameter\n");
+ netcp->tx_compl_qid = KNAV_QUEUE_QPEND;
+ }
+
+ /* NAPI register */
+ netif_napi_add(ndev, &netcp->rx_napi, netcp_rx_poll, NETCP_NAPI_WEIGHT);
+ netif_napi_add(ndev, &netcp->tx_napi, netcp_tx_poll, NETCP_NAPI_WEIGHT);
+
+ /* Register the network device */
+ ndev->dev_id = 0;
+ ndev->watchdog_timeo = NETCP_TX_TIMEOUT;
+ ndev->netdev_ops = &netcp_netdev_ops;
+ SET_NETDEV_DEV(ndev, dev);
+
+ list_add_tail(&netcp->interface_list, &netcp_device->interface_head);
+ return 0;
+
+quit:
+ free_netdev(ndev);
+ return ret;
+}
+
+static void netcp_delete_interface(struct netcp_device *netcp_device,
+ struct net_device *ndev)
+{
+ struct netcp_intf_modpriv *intf_modpriv, *tmp;
+ struct netcp_intf *netcp = netdev_priv(ndev);
+ struct netcp_module *module;
+
+ dev_dbg(netcp_device->device, "Removing interface \"%s\"\n",
+ ndev->name);
+
+ /* Notify each of the modules that the interface is going away */
+ list_for_each_entry_safe(intf_modpriv, tmp, &netcp->module_head,
+ intf_list) {
+ module = intf_modpriv->netcp_module;
+ dev_dbg(netcp_device->device, "Releasing module \"%s\"\n",
+ module->name);
+ if (module->release)
+ module->release(intf_modpriv->module_priv);
+ list_del(&intf_modpriv->intf_list);
+ kfree(intf_modpriv);
+ }
+ WARN(!list_empty(&netcp->module_head), "%s interface module list is not empty!\n",
+ ndev->name);
+
+ list_del(&netcp->interface_list);
+
+ of_node_put(netcp->node_interface);
+ unregister_netdev(ndev);
+ netif_napi_del(&netcp->rx_napi);
+ free_netdev(ndev);
+}
+
+static int netcp_probe(struct platform_device *pdev)
+{
+ struct device_node *node = pdev->dev.of_node;
+ struct netcp_intf *netcp_intf, *netcp_tmp;
+ struct device_node *child, *interfaces;
+ struct netcp_device *netcp_device;
+ struct device *dev = &pdev->dev;
+ struct netcp_module *module;
+ int ret;
+
+ if (!node) {
+ dev_err(dev, "could not find device info\n");
+ return -ENODEV;
+ }
+
+ /* Allocate a new NETCP device instance */
+ netcp_device = devm_kzalloc(dev, sizeof(*netcp_device), GFP_KERNEL);
+ if (!netcp_device)
+ return -ENOMEM;
+
+ pm_runtime_enable(&pdev->dev);
+ ret = pm_runtime_get_sync(&pdev->dev);
+ if (ret < 0) {
+ dev_err(dev, "Failed to enable NETCP power-domain\n");
+ pm_runtime_disable(&pdev->dev);
+ return ret;
+ }
+
+ /* Initialize the NETCP device instance */
+ INIT_LIST_HEAD(&netcp_device->interface_head);
+ INIT_LIST_HEAD(&netcp_device->modpriv_head);
+ netcp_device->device = dev;
+ platform_set_drvdata(pdev, netcp_device);
+
+ /* create interfaces */
+ interfaces = of_get_child_by_name(node, "netcp-interfaces");
+ if (!interfaces) {
+ dev_err(dev, "could not find netcp-interfaces node\n");
+ ret = -ENODEV;
+ goto probe_quit;
+ }
+
+ for_each_available_child_of_node(interfaces, child) {
+ ret = netcp_create_interface(netcp_device, child);
+ if (ret) {
+ dev_err(dev, "could not create interface(%s)\n",
+ child->name);
+ goto probe_quit_interface;
+ }
+ }
+
+ /* Add the device instance to the list */
+ list_add_tail(&netcp_device->device_list, &netcp_devices);
+
+ /* Probe & attach any modules already registered */
+ mutex_lock(&netcp_modules_lock);
+ for_each_netcp_module(module) {
+ ret = netcp_module_probe(netcp_device, module);
+ if (ret < 0)
+ dev_err(dev, "module(%s) probe failed\n", module->name);
+ }
+ mutex_unlock(&netcp_modules_lock);
+ return 0;
+
+probe_quit_interface:
+ list_for_each_entry_safe(netcp_intf, netcp_tmp,
+ &netcp_device->interface_head,
+ interface_list) {
+ netcp_delete_interface(netcp_device, netcp_intf->ndev);
+ }
+
+probe_quit:
+ pm_runtime_put_sync(&pdev->dev);
+ pm_runtime_disable(&pdev->dev);
+ platform_set_drvdata(pdev, NULL);
+ return ret;
+}
+
+static int netcp_remove(struct platform_device *pdev)
+{
+ struct netcp_device *netcp_device = platform_get_drvdata(pdev);
+ struct netcp_inst_modpriv *inst_modpriv, *tmp;
+ struct netcp_module *module;
+
+ list_for_each_entry_safe(inst_modpriv, tmp, &netcp_device->modpriv_head,
+ inst_list) {
+ module = inst_modpriv->netcp_module;
+ dev_dbg(&pdev->dev, "Removing module \"%s\"\n", module->name);
+ module->remove(netcp_device, inst_modpriv->module_priv);
+ list_del(&inst_modpriv->inst_list);
+ kfree(inst_modpriv);
+ }
+ WARN(!list_empty(&netcp_device->interface_head), "%s interface list not empty!\n",
+ pdev->name);
+
+ devm_kfree(&pdev->dev, netcp_device);
+ pm_runtime_put_sync(&pdev->dev);
+ pm_runtime_disable(&pdev->dev);
+ platform_set_drvdata(pdev, NULL);
+ return 0;
+}
+
+static struct of_device_id of_match[] = {
+ { .compatible = "ti,netcp-1.0", },
+ {},
+};
+MODULE_DEVICE_TABLE(of, of_match);
+
+static struct platform_driver netcp_driver = {
+ .driver = {
+ .name = "netcp-1.0",
+ .owner = THIS_MODULE,
+ .of_match_table = of_match,
+ },
+ .probe = netcp_probe,
+ .remove = netcp_remove,
+};
+module_platform_driver(netcp_driver);
+
+MODULE_LICENSE("GPL v2");
+MODULE_DESCRIPTION("TI NETCP driver for Keystone SOCs");
+MODULE_AUTHOR("Sandeep Nair <sandeep_n-l0cyMroinI0@public.gmane.org");
--
1.7.9.5
--
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 net-next v8 3/4] net: netcp: Add Keystone NetCP GbE driver
From: Murali Karicheri @ 2015-01-16 0:12 UTC (permalink / raw)
To: davem, devicetree, linux-kernel, netdev; +Cc: Wingman Kwok, Murali Karicheri
In-Reply-To: <1421367172-19793-1-git-send-email-m-karicheri2@ti.com>
From: Wingman Kwok <w-kwok2@ti.com>
This patch add support for 1G Ethernet driver based on Keystone
NetCP hardware. The gigabit Ethernet (GbE) switch subsystem is one of the main
components of the network coprocessor (NETCP) peripheral. The purpose of the
gigabit Ethernet switch subsystem in the NETCP is to provide an interface to
transfer data between the host device and another connected device in
compliance with the Ethernet protocol. GbE consists of 5 port Ethernet Switch
module, 4 Serial Gigabit Media Independent Interface (SGMII) modules, MDIO
module and SerDes.
Driver for 5 port GbE switch and SGMII module is added in this patch. These
hardware modules along with netcp core driver provides Network driver functions
for 1G Ethernet.
Detailed hardware spec is available at
http://www.ti.com/lit/ug/sprugv9d/sprugv9d.pdf
Cc: David Miller <davem@davemloft.net>
Cc: Rob Herring <robh+dt@kernel.org>
Cc: Grant Likely <grant.likely@linaro.org>
Cc: Santosh Shilimkar <santosh.shilimkar@kernel.org>
Cc: Pawel Moll <pawel.moll@arm.com>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Ian Campbell <ijc+devicetree@hellion.org.uk>
Cc: Kumar Gala <galak@codeaurora.org>
Signed-off-by: Wingman Kwok <w-kwok2@ti.com>
Signed-off-by: Murali Karicheri <m-karicheri2@ti.com>
---
drivers/net/ethernet/ti/Makefile | 3 +-
drivers/net/ethernet/ti/netcp_ethss.c | 1714 +++++++++++++++++++++++++++++++++
drivers/net/ethernet/ti/netcp_sgmii.c | 131 +++
3 files changed, 1847 insertions(+), 1 deletion(-)
create mode 100644 drivers/net/ethernet/ti/netcp_ethss.c
create mode 100644 drivers/net/ethernet/ti/netcp_sgmii.c
diff --git a/drivers/net/ethernet/ti/Makefile b/drivers/net/ethernet/ti/Makefile
index 4e8a8e4..6d4dde8 100644
--- a/drivers/net/ethernet/ti/Makefile
+++ b/drivers/net/ethernet/ti/Makefile
@@ -12,4 +12,5 @@ obj-$(CONFIG_TI_CPSW) += ti_cpsw.o
ti_cpsw-y := cpsw_ale.o cpsw.o cpts.o
obj-$(CONFIG_TI_KEYSTONE_NETCP) += keystone_netcp.o
-keystone_netcp-y := netcp_core.o
+keystone_netcp-y := netcp_core.o netcp_ethss.o netcp_sgmii.o \
+ cpsw_ale.o cpts.o
diff --git a/drivers/net/ethernet/ti/netcp_ethss.c b/drivers/net/ethernet/ti/netcp_ethss.c
new file mode 100644
index 0000000..58f00c5
--- /dev/null
+++ b/drivers/net/ethernet/ti/netcp_ethss.c
@@ -0,0 +1,1714 @@
+/*
+ * Keystone GBE subsystem code
+ *
+ * Copyright (C) 2014 Texas Instruments Incorporated
+ * Authors: Sandeep Nair <sandeep_n@ti.com>
+ * Sandeep Paulraj <s-paulraj@ti.com>
+ * Cyril Chemparathy <cyril@ti.com>
+ * Santosh Shilimkar <santosh.shilimkar@ti.com>
+ * Wingman Kwok <w-kwok2@ti.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 version 2.
+ *
+ * This program is distributed "as is" WITHOUT ANY WARRANTY of any
+ * kind, whether express or implied; without even the implied warranty
+ * of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ */
+
+#include <linux/io.h>
+#include <linux/of_mdio.h>
+#include <linux/of_address.h>
+#include <linux/if_vlan.h>
+#include <linux/ethtool.h>
+
+#include "cpsw_ale.h"
+#include "netcp.h"
+
+#define NETCP_DRIVER_NAME "TI KeyStone Ethernet Driver"
+#define NETCP_DRIVER_VERSION "v1.0"
+
+#define GBE_IDENT(reg) ((reg >> 16) & 0xffff)
+#define GBE_MAJOR_VERSION(reg) (reg >> 8 & 0x7)
+#define GBE_MINOR_VERSION(reg) (reg & 0xff)
+#define GBE_RTL_VERSION(reg) ((reg >> 11) & 0x1f)
+
+/* 1G Ethernet SS defines */
+#define GBE_MODULE_NAME "netcp-gbe"
+#define GBE_SS_VERSION_14 0x4ed21104
+
+#define GBE13_SGMII_MODULE_OFFSET 0x100
+#define GBE13_SGMII34_MODULE_OFFSET 0x400
+#define GBE13_SWITCH_MODULE_OFFSET 0x800
+#define GBE13_HOST_PORT_OFFSET 0x834
+#define GBE13_SLAVE_PORT_OFFSET 0x860
+#define GBE13_EMAC_OFFSET 0x900
+#define GBE13_SLAVE_PORT2_OFFSET 0xa00
+#define GBE13_HW_STATS_OFFSET 0xb00
+#define GBE13_ALE_OFFSET 0xe00
+#define GBE13_HOST_PORT_NUM 0
+#define GBE13_NUM_SLAVES 4
+#define GBE13_NUM_ALE_PORTS (GBE13_NUM_SLAVES + 1)
+#define GBE13_NUM_ALE_ENTRIES 1024
+
+#define GBE_TIMER_INTERVAL (HZ / 2)
+
+/* Soft reset register values */
+#define SOFT_RESET_MASK BIT(0)
+#define SOFT_RESET BIT(0)
+#define DEVICE_EMACSL_RESET_POLL_COUNT 100
+#define GMACSL_RET_WARN_RESET_INCOMPLETE -2
+
+#define MACSL_RX_ENABLE_CSF BIT(23)
+#define MACSL_ENABLE_EXT_CTL BIT(18)
+#define MACSL_GIG_MODE BIT(7)
+#define MACSL_GMII_ENABLE BIT(5)
+#define MACSL_FULLDUPLEX BIT(0)
+
+#define GBE_CTL_P0_ENABLE BIT(2)
+#define GBE_REG_VAL_STAT_ENABLE_ALL 0xff
+#define GBE_STATS_CD_SEL BIT(28)
+
+#define GBE_PORT_MASK(x) (BIT(x) - 1)
+#define GBE_MASK_NO_PORTS 0
+
+#define GBE_DEF_1G_MAC_CONTROL \
+ (MACSL_GIG_MODE | MACSL_GMII_ENABLE | \
+ MACSL_ENABLE_EXT_CTL | MACSL_RX_ENABLE_CSF)
+
+#define GBE_STATSA_MODULE 0
+#define GBE_STATSB_MODULE 1
+#define GBE_STATSC_MODULE 2
+#define GBE_STATSD_MODULE 3
+
+#define MAX_SLAVES GBE13_NUM_SLAVES
+/* s: 0-based slave_port */
+#define SGMII_BASE(s) \
+ (((s) < 2) ? gbe_dev->sgmii_port_regs : gbe_dev->sgmii_port34_regs)
+
+#define GBE_TX_QUEUE 648
+#define GBE_TXHOOK_ORDER 0
+#define GBE_DEFAULT_ALE_AGEOUT 30
+#define NETCP_LINK_STATE_INVALID -1
+
+#define GBE_SET_REG_OFS(p, rb, rn) p->rb##_ofs.rn = \
+ offsetof(struct gbe##_##rb, rn)
+#define GBE_REG_ADDR(p, rb, rn) (p->rb + p->rb##_ofs.rn)
+
+struct gbe_ss_regs {
+ u32 id_ver;
+ u32 synce_count;
+ u32 synce_mux;
+};
+
+struct gbe_ss_regs_ofs {
+ u16 id_ver;
+ u16 control;
+};
+
+struct gbe_switch_regs {
+ u32 id_ver;
+ u32 control;
+ u32 soft_reset;
+ u32 stat_port_en;
+ u32 ptype;
+ u32 soft_idle;
+ u32 thru_rate;
+ u32 gap_thresh;
+ u32 tx_start_wds;
+ u32 flow_control;
+};
+
+struct gbe_switch_regs_ofs {
+ u16 id_ver;
+ u16 control;
+ u16 soft_reset;
+ u16 emcontrol;
+ u16 stat_port_en;
+ u16 ptype;
+ u16 flow_control;
+};
+
+struct gbe_port_regs {
+ u32 max_blks;
+ u32 blk_cnt;
+ u32 port_vlan;
+ u32 tx_pri_map;
+ u32 sa_lo;
+ u32 sa_hi;
+ u32 ts_ctl;
+ u32 ts_seq_ltype;
+ u32 ts_vlan;
+ u32 ts_ctl_ltype2;
+ u32 ts_ctl2;
+};
+
+struct gbe_port_regs_ofs {
+ u16 port_vlan;
+ u16 tx_pri_map;
+ u16 sa_lo;
+ u16 sa_hi;
+ u16 ts_ctl;
+ u16 ts_seq_ltype;
+ u16 ts_vlan;
+ u16 ts_ctl_ltype2;
+ u16 ts_ctl2;
+};
+
+struct gbe_host_port_regs {
+ u32 src_id;
+ u32 port_vlan;
+ u32 rx_pri_map;
+ u32 rx_maxlen;
+};
+
+struct gbe_host_port_regs_ofs {
+ u16 port_vlan;
+ u16 tx_pri_map;
+ u16 rx_maxlen;
+};
+
+struct gbe_emac_regs {
+ u32 id_ver;
+ u32 mac_control;
+ u32 mac_status;
+ u32 soft_reset;
+ u32 rx_maxlen;
+ u32 __reserved_0;
+ u32 rx_pause;
+ u32 tx_pause;
+ u32 __reserved_1;
+ u32 rx_pri_map;
+ u32 rsvd[6];
+};
+
+struct gbe_emac_regs_ofs {
+ u16 mac_control;
+ u16 soft_reset;
+ u16 rx_maxlen;
+};
+
+struct gbe_hw_stats {
+ u32 rx_good_frames;
+ u32 rx_broadcast_frames;
+ u32 rx_multicast_frames;
+ u32 rx_pause_frames;
+ u32 rx_crc_errors;
+ u32 rx_align_code_errors;
+ u32 rx_oversized_frames;
+ u32 rx_jabber_frames;
+ u32 rx_undersized_frames;
+ u32 rx_fragments;
+ u32 __pad_0[2];
+ u32 rx_bytes;
+ u32 tx_good_frames;
+ u32 tx_broadcast_frames;
+ u32 tx_multicast_frames;
+ u32 tx_pause_frames;
+ u32 tx_deferred_frames;
+ u32 tx_collision_frames;
+ u32 tx_single_coll_frames;
+ u32 tx_mult_coll_frames;
+ u32 tx_excessive_collisions;
+ u32 tx_late_collisions;
+ u32 tx_underrun;
+ u32 tx_carrier_sense_errors;
+ u32 tx_bytes;
+ u32 tx_64byte_frames;
+ u32 tx_65_to_127byte_frames;
+ u32 tx_128_to_255byte_frames;
+ u32 tx_256_to_511byte_frames;
+ u32 tx_512_to_1023byte_frames;
+ u32 tx_1024byte_frames;
+ u32 net_bytes;
+ u32 rx_sof_overruns;
+ u32 rx_mof_overruns;
+ u32 rx_dma_overruns;
+};
+
+#define GBE13_NUM_HW_STAT_ENTRIES (sizeof(struct gbe_hw_stats)/sizeof(u32))
+#define GBE13_NUM_HW_STATS_MOD 2
+#define GBE_MAX_HW_STAT_MODS 3
+#define GBE_HW_STATS_REG_MAP_SZ 0x100
+
+struct gbe_slave {
+ void __iomem *port_regs;
+ void __iomem *emac_regs;
+ struct gbe_port_regs_ofs port_regs_ofs;
+ struct gbe_emac_regs_ofs emac_regs_ofs;
+ int slave_num; /* 0 based logical number */
+ int port_num; /* actual port number */
+ atomic_t link_state;
+ bool open;
+ struct phy_device *phy;
+ u32 link_interface;
+ u32 mac_control;
+ u8 phy_port_t;
+ struct device_node *phy_node;
+ struct list_head slave_list;
+};
+
+struct gbe_priv {
+ struct device *dev;
+ struct netcp_device *netcp_device;
+ struct timer_list timer;
+ u32 num_slaves;
+ u32 ale_entries;
+ u32 ale_ports;
+ bool enable_ale;
+ struct netcp_tx_pipe tx_pipe;
+
+ int host_port;
+ u32 rx_packet_max;
+ u32 ss_version;
+
+ void __iomem *ss_regs;
+ void __iomem *switch_regs;
+ void __iomem *host_port_regs;
+ void __iomem *ale_reg;
+ void __iomem *sgmii_port_regs;
+ void __iomem *sgmii_port34_regs;
+ void __iomem *xgbe_serdes_regs;
+ void __iomem *hw_stats_regs[GBE_MAX_HW_STAT_MODS];
+
+ struct gbe_ss_regs_ofs ss_regs_ofs;
+ struct gbe_switch_regs_ofs switch_regs_ofs;
+ struct gbe_host_port_regs_ofs host_port_regs_ofs;
+
+ struct cpsw_ale *ale;
+ unsigned int tx_queue_id;
+ const char *dma_chan_name;
+
+ struct list_head gbe_intf_head;
+ struct list_head secondary_slaves;
+ struct net_device *dummy_ndev;
+
+ u64 *hw_stats;
+ const struct netcp_ethtool_stat *et_stats;
+ int num_et_stats;
+ /* Lock for updating the hwstats */
+ spinlock_t hw_stats_lock;
+};
+
+struct gbe_intf {
+ struct net_device *ndev;
+ struct device *dev;
+ struct gbe_priv *gbe_dev;
+ struct netcp_tx_pipe tx_pipe;
+ struct gbe_slave *slave;
+ struct list_head gbe_intf_list;
+ unsigned long active_vlans[BITS_TO_LONGS(VLAN_N_VID)];
+};
+
+static struct netcp_module gbe_module;
+
+/* Statistic management */
+struct netcp_ethtool_stat {
+ char desc[ETH_GSTRING_LEN];
+ int type;
+ u32 size;
+ int offset;
+};
+
+#define GBE_STATSA_INFO(field) "GBE_A:"#field, GBE_STATSA_MODULE,\
+ FIELD_SIZEOF(struct gbe_hw_stats, field), \
+ offsetof(struct gbe_hw_stats, field)
+
+#define GBE_STATSB_INFO(field) "GBE_B:"#field, GBE_STATSB_MODULE,\
+ FIELD_SIZEOF(struct gbe_hw_stats, field), \
+ offsetof(struct gbe_hw_stats, field)
+
+#define GBE_STATSC_INFO(field) "GBE_C:"#field, GBE_STATSC_MODULE,\
+ FIELD_SIZEOF(struct gbe_hw_stats, field), \
+ offsetof(struct gbe_hw_stats, field)
+
+#define GBE_STATSD_INFO(field) "GBE_D:"#field, GBE_STATSD_MODULE,\
+ FIELD_SIZEOF(struct gbe_hw_stats, field), \
+ offsetof(struct gbe_hw_stats, field)
+
+static const struct netcp_ethtool_stat gbe13_et_stats[] = {
+ /* GBE module A */
+ {GBE_STATSA_INFO(rx_good_frames)},
+ {GBE_STATSA_INFO(rx_broadcast_frames)},
+ {GBE_STATSA_INFO(rx_multicast_frames)},
+ {GBE_STATSA_INFO(rx_pause_frames)},
+ {GBE_STATSA_INFO(rx_crc_errors)},
+ {GBE_STATSA_INFO(rx_align_code_errors)},
+ {GBE_STATSA_INFO(rx_oversized_frames)},
+ {GBE_STATSA_INFO(rx_jabber_frames)},
+ {GBE_STATSA_INFO(rx_undersized_frames)},
+ {GBE_STATSA_INFO(rx_fragments)},
+ {GBE_STATSA_INFO(rx_bytes)},
+ {GBE_STATSA_INFO(tx_good_frames)},
+ {GBE_STATSA_INFO(tx_broadcast_frames)},
+ {GBE_STATSA_INFO(tx_multicast_frames)},
+ {GBE_STATSA_INFO(tx_pause_frames)},
+ {GBE_STATSA_INFO(tx_deferred_frames)},
+ {GBE_STATSA_INFO(tx_collision_frames)},
+ {GBE_STATSA_INFO(tx_single_coll_frames)},
+ {GBE_STATSA_INFO(tx_mult_coll_frames)},
+ {GBE_STATSA_INFO(tx_excessive_collisions)},
+ {GBE_STATSA_INFO(tx_late_collisions)},
+ {GBE_STATSA_INFO(tx_underrun)},
+ {GBE_STATSA_INFO(tx_carrier_sense_errors)},
+ {GBE_STATSA_INFO(tx_bytes)},
+ {GBE_STATSA_INFO(tx_64byte_frames)},
+ {GBE_STATSA_INFO(tx_65_to_127byte_frames)},
+ {GBE_STATSA_INFO(tx_128_to_255byte_frames)},
+ {GBE_STATSA_INFO(tx_256_to_511byte_frames)},
+ {GBE_STATSA_INFO(tx_512_to_1023byte_frames)},
+ {GBE_STATSA_INFO(tx_1024byte_frames)},
+ {GBE_STATSA_INFO(net_bytes)},
+ {GBE_STATSA_INFO(rx_sof_overruns)},
+ {GBE_STATSA_INFO(rx_mof_overruns)},
+ {GBE_STATSA_INFO(rx_dma_overruns)},
+ /* GBE module B */
+ {GBE_STATSB_INFO(rx_good_frames)},
+ {GBE_STATSB_INFO(rx_broadcast_frames)},
+ {GBE_STATSB_INFO(rx_multicast_frames)},
+ {GBE_STATSB_INFO(rx_pause_frames)},
+ {GBE_STATSB_INFO(rx_crc_errors)},
+ {GBE_STATSB_INFO(rx_align_code_errors)},
+ {GBE_STATSB_INFO(rx_oversized_frames)},
+ {GBE_STATSB_INFO(rx_jabber_frames)},
+ {GBE_STATSB_INFO(rx_undersized_frames)},
+ {GBE_STATSB_INFO(rx_fragments)},
+ {GBE_STATSB_INFO(rx_bytes)},
+ {GBE_STATSB_INFO(tx_good_frames)},
+ {GBE_STATSB_INFO(tx_broadcast_frames)},
+ {GBE_STATSB_INFO(tx_multicast_frames)},
+ {GBE_STATSB_INFO(tx_pause_frames)},
+ {GBE_STATSB_INFO(tx_deferred_frames)},
+ {GBE_STATSB_INFO(tx_collision_frames)},
+ {GBE_STATSB_INFO(tx_single_coll_frames)},
+ {GBE_STATSB_INFO(tx_mult_coll_frames)},
+ {GBE_STATSB_INFO(tx_excessive_collisions)},
+ {GBE_STATSB_INFO(tx_late_collisions)},
+ {GBE_STATSB_INFO(tx_underrun)},
+ {GBE_STATSB_INFO(tx_carrier_sense_errors)},
+ {GBE_STATSB_INFO(tx_bytes)},
+ {GBE_STATSB_INFO(tx_64byte_frames)},
+ {GBE_STATSB_INFO(tx_65_to_127byte_frames)},
+ {GBE_STATSB_INFO(tx_128_to_255byte_frames)},
+ {GBE_STATSB_INFO(tx_256_to_511byte_frames)},
+ {GBE_STATSB_INFO(tx_512_to_1023byte_frames)},
+ {GBE_STATSB_INFO(tx_1024byte_frames)},
+ {GBE_STATSB_INFO(net_bytes)},
+ {GBE_STATSB_INFO(rx_sof_overruns)},
+ {GBE_STATSB_INFO(rx_mof_overruns)},
+ {GBE_STATSB_INFO(rx_dma_overruns)},
+ /* GBE module C */
+ {GBE_STATSC_INFO(rx_good_frames)},
+ {GBE_STATSC_INFO(rx_broadcast_frames)},
+ {GBE_STATSC_INFO(rx_multicast_frames)},
+ {GBE_STATSC_INFO(rx_pause_frames)},
+ {GBE_STATSC_INFO(rx_crc_errors)},
+ {GBE_STATSC_INFO(rx_align_code_errors)},
+ {GBE_STATSC_INFO(rx_oversized_frames)},
+ {GBE_STATSC_INFO(rx_jabber_frames)},
+ {GBE_STATSC_INFO(rx_undersized_frames)},
+ {GBE_STATSC_INFO(rx_fragments)},
+ {GBE_STATSC_INFO(rx_bytes)},
+ {GBE_STATSC_INFO(tx_good_frames)},
+ {GBE_STATSC_INFO(tx_broadcast_frames)},
+ {GBE_STATSC_INFO(tx_multicast_frames)},
+ {GBE_STATSC_INFO(tx_pause_frames)},
+ {GBE_STATSC_INFO(tx_deferred_frames)},
+ {GBE_STATSC_INFO(tx_collision_frames)},
+ {GBE_STATSC_INFO(tx_single_coll_frames)},
+ {GBE_STATSC_INFO(tx_mult_coll_frames)},
+ {GBE_STATSC_INFO(tx_excessive_collisions)},
+ {GBE_STATSC_INFO(tx_late_collisions)},
+ {GBE_STATSC_INFO(tx_underrun)},
+ {GBE_STATSC_INFO(tx_carrier_sense_errors)},
+ {GBE_STATSC_INFO(tx_bytes)},
+ {GBE_STATSC_INFO(tx_64byte_frames)},
+ {GBE_STATSC_INFO(tx_65_to_127byte_frames)},
+ {GBE_STATSC_INFO(tx_128_to_255byte_frames)},
+ {GBE_STATSC_INFO(tx_256_to_511byte_frames)},
+ {GBE_STATSC_INFO(tx_512_to_1023byte_frames)},
+ {GBE_STATSC_INFO(tx_1024byte_frames)},
+ {GBE_STATSC_INFO(net_bytes)},
+ {GBE_STATSC_INFO(rx_sof_overruns)},
+ {GBE_STATSC_INFO(rx_mof_overruns)},
+ {GBE_STATSC_INFO(rx_dma_overruns)},
+ /* GBE module D */
+ {GBE_STATSD_INFO(rx_good_frames)},
+ {GBE_STATSD_INFO(rx_broadcast_frames)},
+ {GBE_STATSD_INFO(rx_multicast_frames)},
+ {GBE_STATSD_INFO(rx_pause_frames)},
+ {GBE_STATSD_INFO(rx_crc_errors)},
+ {GBE_STATSD_INFO(rx_align_code_errors)},
+ {GBE_STATSD_INFO(rx_oversized_frames)},
+ {GBE_STATSD_INFO(rx_jabber_frames)},
+ {GBE_STATSD_INFO(rx_undersized_frames)},
+ {GBE_STATSD_INFO(rx_fragments)},
+ {GBE_STATSD_INFO(rx_bytes)},
+ {GBE_STATSD_INFO(tx_good_frames)},
+ {GBE_STATSD_INFO(tx_broadcast_frames)},
+ {GBE_STATSD_INFO(tx_multicast_frames)},
+ {GBE_STATSD_INFO(tx_pause_frames)},
+ {GBE_STATSD_INFO(tx_deferred_frames)},
+ {GBE_STATSD_INFO(tx_collision_frames)},
+ {GBE_STATSD_INFO(tx_single_coll_frames)},
+ {GBE_STATSD_INFO(tx_mult_coll_frames)},
+ {GBE_STATSD_INFO(tx_excessive_collisions)},
+ {GBE_STATSD_INFO(tx_late_collisions)},
+ {GBE_STATSD_INFO(tx_underrun)},
+ {GBE_STATSD_INFO(tx_carrier_sense_errors)},
+ {GBE_STATSD_INFO(tx_bytes)},
+ {GBE_STATSD_INFO(tx_64byte_frames)},
+ {GBE_STATSD_INFO(tx_65_to_127byte_frames)},
+ {GBE_STATSD_INFO(tx_128_to_255byte_frames)},
+ {GBE_STATSD_INFO(tx_256_to_511byte_frames)},
+ {GBE_STATSD_INFO(tx_512_to_1023byte_frames)},
+ {GBE_STATSD_INFO(tx_1024byte_frames)},
+ {GBE_STATSD_INFO(net_bytes)},
+ {GBE_STATSD_INFO(rx_sof_overruns)},
+ {GBE_STATSD_INFO(rx_mof_overruns)},
+ {GBE_STATSD_INFO(rx_dma_overruns)},
+};
+
+#define for_each_intf(i, priv) \
+ list_for_each_entry((i), &(priv)->gbe_intf_head, gbe_intf_list)
+
+#define for_each_sec_slave(slave, priv) \
+ list_for_each_entry((slave), &(priv)->secondary_slaves, slave_list)
+
+#define first_sec_slave(priv) \
+ list_first_entry(&priv->secondary_slaves, \
+ struct gbe_slave, slave_list)
+
+static void keystone_get_drvinfo(struct net_device *ndev,
+ struct ethtool_drvinfo *info)
+{
+ strncpy(info->driver, NETCP_DRIVER_NAME, sizeof(info->driver));
+ strncpy(info->version, NETCP_DRIVER_VERSION, sizeof(info->version));
+}
+
+static u32 keystone_get_msglevel(struct net_device *ndev)
+{
+ struct netcp_intf *netcp = netdev_priv(ndev);
+
+ return netcp->msg_enable;
+}
+
+static void keystone_set_msglevel(struct net_device *ndev, u32 value)
+{
+ struct netcp_intf *netcp = netdev_priv(ndev);
+
+ netcp->msg_enable = value;
+}
+
+static void keystone_get_stat_strings(struct net_device *ndev,
+ uint32_t stringset, uint8_t *data)
+{
+ struct netcp_intf *netcp = netdev_priv(ndev);
+ struct gbe_intf *gbe_intf;
+ struct gbe_priv *gbe_dev;
+ int i;
+
+ gbe_intf = netcp_module_get_intf_data(&gbe_module, netcp);
+ if (!gbe_intf)
+ return;
+ gbe_dev = gbe_intf->gbe_dev;
+
+ switch (stringset) {
+ case ETH_SS_STATS:
+ for (i = 0; i < gbe_dev->num_et_stats; i++) {
+ memcpy(data, gbe_dev->et_stats[i].desc,
+ ETH_GSTRING_LEN);
+ data += ETH_GSTRING_LEN;
+ }
+ break;
+ case ETH_SS_TEST:
+ break;
+ }
+}
+
+static int keystone_get_sset_count(struct net_device *ndev, int stringset)
+{
+ struct netcp_intf *netcp = netdev_priv(ndev);
+ struct gbe_intf *gbe_intf;
+ struct gbe_priv *gbe_dev;
+
+ gbe_intf = netcp_module_get_intf_data(&gbe_module, netcp);
+ if (!gbe_intf)
+ return -EINVAL;
+ gbe_dev = gbe_intf->gbe_dev;
+
+ switch (stringset) {
+ case ETH_SS_TEST:
+ return 0;
+ case ETH_SS_STATS:
+ return gbe_dev->num_et_stats;
+ default:
+ return -EINVAL;
+ }
+}
+
+static void gbe_update_stats(struct gbe_priv *gbe_dev, uint64_t *data)
+{
+ void __iomem *base = NULL;
+ u32 __iomem *p;
+ u32 tmp = 0;
+ int i;
+
+ for (i = 0; i < gbe_dev->num_et_stats; i++) {
+ base = gbe_dev->hw_stats_regs[gbe_dev->et_stats[i].type];
+ p = base + gbe_dev->et_stats[i].offset;
+ tmp = readl(p);
+ gbe_dev->hw_stats[i] = gbe_dev->hw_stats[i] + tmp;
+ if (data)
+ data[i] = gbe_dev->hw_stats[i];
+ /* write-to-decrement:
+ * new register value = old register value - write value
+ */
+ writel(tmp, p);
+ }
+}
+
+static void gbe_update_stats_ver14(struct gbe_priv *gbe_dev, uint64_t *data)
+{
+ void __iomem *gbe_statsa = gbe_dev->hw_stats_regs[0];
+ void __iomem *gbe_statsb = gbe_dev->hw_stats_regs[1];
+ u64 *hw_stats = &gbe_dev->hw_stats[0];
+ void __iomem *base = NULL;
+ u32 __iomem *p;
+ u32 tmp = 0, val, pair_size = (gbe_dev->num_et_stats / 2);
+ int i, j, pair;
+
+ for (pair = 0; pair < 2; pair++) {
+ val = readl(GBE_REG_ADDR(gbe_dev, switch_regs, stat_port_en));
+
+ if (pair == 0)
+ val &= ~GBE_STATS_CD_SEL;
+ else
+ val |= GBE_STATS_CD_SEL;
+
+ /* make the stat modules visible */
+ writel(val, GBE_REG_ADDR(gbe_dev, switch_regs, stat_port_en));
+
+ for (i = 0; i < pair_size; i++) {
+ j = pair * pair_size + i;
+ switch (gbe_dev->et_stats[j].type) {
+ case GBE_STATSA_MODULE:
+ case GBE_STATSC_MODULE:
+ base = gbe_statsa;
+ break;
+ case GBE_STATSB_MODULE:
+ case GBE_STATSD_MODULE:
+ base = gbe_statsb;
+ break;
+ }
+
+ p = base + gbe_dev->et_stats[j].offset;
+ tmp = readl(p);
+ hw_stats[j] += tmp;
+ if (data)
+ data[j] = hw_stats[j];
+ /* write-to-decrement:
+ * new register value = old register value - write value
+ */
+ writel(tmp, p);
+ }
+ }
+}
+
+static void keystone_get_ethtool_stats(struct net_device *ndev,
+ struct ethtool_stats *stats,
+ uint64_t *data)
+{
+ struct netcp_intf *netcp = netdev_priv(ndev);
+ struct gbe_intf *gbe_intf;
+ struct gbe_priv *gbe_dev;
+
+ gbe_intf = netcp_module_get_intf_data(&gbe_module, netcp);
+ if (!gbe_intf)
+ return;
+
+ gbe_dev = gbe_intf->gbe_dev;
+ spin_lock_bh(&gbe_dev->hw_stats_lock);
+ gbe_update_stats_ver14(gbe_dev, data);
+ spin_unlock_bh(&gbe_dev->hw_stats_lock);
+}
+
+static int keystone_get_settings(struct net_device *ndev,
+ struct ethtool_cmd *cmd)
+{
+ struct netcp_intf *netcp = netdev_priv(ndev);
+ struct phy_device *phy = ndev->phydev;
+ struct gbe_intf *gbe_intf;
+ int ret;
+
+ if (!phy)
+ return -EINVAL;
+
+ gbe_intf = netcp_module_get_intf_data(&gbe_module, netcp);
+ if (!gbe_intf)
+ return -EINVAL;
+
+ if (!gbe_intf->slave)
+ return -EINVAL;
+
+ ret = phy_ethtool_gset(phy, cmd);
+ if (!ret)
+ cmd->port = gbe_intf->slave->phy_port_t;
+
+ return ret;
+}
+
+static int keystone_set_settings(struct net_device *ndev,
+ struct ethtool_cmd *cmd)
+{
+ struct netcp_intf *netcp = netdev_priv(ndev);
+ struct phy_device *phy = ndev->phydev;
+ struct gbe_intf *gbe_intf;
+ u32 features = cmd->advertising & cmd->supported;
+
+ if (!phy)
+ return -EINVAL;
+
+ gbe_intf = netcp_module_get_intf_data(&gbe_module, netcp);
+ if (!gbe_intf)
+ return -EINVAL;
+
+ if (!gbe_intf->slave)
+ return -EINVAL;
+
+ if (cmd->port != gbe_intf->slave->phy_port_t) {
+ if ((cmd->port == PORT_TP) && !(features & ADVERTISED_TP))
+ return -EINVAL;
+
+ if ((cmd->port == PORT_AUI) && !(features & ADVERTISED_AUI))
+ return -EINVAL;
+
+ if ((cmd->port == PORT_BNC) && !(features & ADVERTISED_BNC))
+ return -EINVAL;
+
+ if ((cmd->port == PORT_MII) && !(features & ADVERTISED_MII))
+ return -EINVAL;
+
+ if ((cmd->port == PORT_FIBRE) && !(features & ADVERTISED_FIBRE))
+ return -EINVAL;
+ }
+
+ gbe_intf->slave->phy_port_t = cmd->port;
+ return phy_ethtool_sset(phy, cmd);
+}
+
+static const struct ethtool_ops keystone_ethtool_ops = {
+ .get_drvinfo = keystone_get_drvinfo,
+ .get_link = ethtool_op_get_link,
+ .get_msglevel = keystone_get_msglevel,
+ .set_msglevel = keystone_set_msglevel,
+ .get_strings = keystone_get_stat_strings,
+ .get_sset_count = keystone_get_sset_count,
+ .get_ethtool_stats = keystone_get_ethtool_stats,
+ .get_settings = keystone_get_settings,
+ .set_settings = keystone_set_settings,
+};
+
+#define mac_hi(mac) (((mac)[0] << 0) | ((mac)[1] << 8) | \
+ ((mac)[2] << 16) | ((mac)[3] << 24))
+#define mac_lo(mac) (((mac)[4] << 0) | ((mac)[5] << 8))
+
+static void gbe_set_slave_mac(struct gbe_slave *slave,
+ struct gbe_intf *gbe_intf)
+{
+ struct net_device *ndev = gbe_intf->ndev;
+
+ writel(mac_hi(ndev->dev_addr), GBE_REG_ADDR(slave, port_regs, sa_hi));
+ writel(mac_lo(ndev->dev_addr), GBE_REG_ADDR(slave, port_regs, sa_lo));
+}
+
+static int gbe_get_slave_port(struct gbe_priv *priv, u32 slave_num)
+{
+ if (priv->host_port == 0)
+ return slave_num + 1;
+
+ return slave_num;
+}
+
+static void netcp_ethss_link_state_action(struct gbe_priv *gbe_dev,
+ struct net_device *ndev,
+ struct gbe_slave *slave,
+ int up)
+{
+ struct phy_device *phy = slave->phy;
+ u32 mac_control = 0;
+
+ if (up) {
+ mac_control = slave->mac_control;
+ if (phy && (phy->speed == SPEED_1000))
+ mac_control |= MACSL_GIG_MODE;
+
+ writel(mac_control, GBE_REG_ADDR(slave, emac_regs,
+ mac_control));
+
+ cpsw_ale_control_set(gbe_dev->ale, slave->port_num,
+ ALE_PORT_STATE,
+ ALE_PORT_STATE_FORWARD);
+
+ if (ndev && slave->open)
+ netif_carrier_on(ndev);
+ } else {
+ writel(mac_control, GBE_REG_ADDR(slave, emac_regs,
+ mac_control));
+ cpsw_ale_control_set(gbe_dev->ale, slave->port_num,
+ ALE_PORT_STATE,
+ ALE_PORT_STATE_DISABLE);
+ if (ndev)
+ netif_carrier_off(ndev);
+ }
+
+ if (phy)
+ phy_print_status(phy);
+}
+
+static bool gbe_phy_link_status(struct gbe_slave *slave)
+{
+ return !slave->phy || slave->phy->link;
+}
+
+static void netcp_ethss_update_link_state(struct gbe_priv *gbe_dev,
+ struct gbe_slave *slave,
+ struct net_device *ndev)
+{
+ int sp = slave->slave_num;
+ int phy_link_state, sgmii_link_state = 1, link_state;
+
+ if (!slave->open)
+ return;
+
+ sgmii_link_state = netcp_sgmii_get_port_link(SGMII_BASE(sp), sp);
+ phy_link_state = gbe_phy_link_status(slave);
+ link_state = phy_link_state & sgmii_link_state;
+
+ if (atomic_xchg(&slave->link_state, link_state) != link_state)
+ netcp_ethss_link_state_action(gbe_dev, ndev, slave,
+ link_state);
+}
+
+static void gbe_adjust_link(struct net_device *ndev)
+{
+ struct netcp_intf *netcp = netdev_priv(ndev);
+ struct gbe_intf *gbe_intf;
+
+ gbe_intf = netcp_module_get_intf_data(&gbe_module, netcp);
+ if (!gbe_intf)
+ return;
+
+ netcp_ethss_update_link_state(gbe_intf->gbe_dev, gbe_intf->slave,
+ ndev);
+}
+
+static void gbe_adjust_link_sec_slaves(struct net_device *ndev)
+{
+ struct gbe_priv *gbe_dev = netdev_priv(ndev);
+ struct gbe_slave *slave;
+
+ for_each_sec_slave(slave, gbe_dev)
+ netcp_ethss_update_link_state(gbe_dev, slave, NULL);
+}
+
+/* Reset EMAC
+ * Soft reset is set and polled until clear, or until a timeout occurs
+ */
+static int gbe_port_reset(struct gbe_slave *slave)
+{
+ u32 i, v;
+
+ /* Set the soft reset bit */
+ writel(SOFT_RESET, GBE_REG_ADDR(slave, emac_regs, soft_reset));
+
+ /* Wait for the bit to clear */
+ for (i = 0; i < DEVICE_EMACSL_RESET_POLL_COUNT; i++) {
+ v = readl(GBE_REG_ADDR(slave, emac_regs, soft_reset));
+ if ((v & SOFT_RESET_MASK) != SOFT_RESET)
+ return 0;
+ }
+
+ /* Timeout on the reset */
+ return GMACSL_RET_WARN_RESET_INCOMPLETE;
+}
+
+/* Configure EMAC */
+static void gbe_port_config(struct gbe_priv *gbe_dev, struct gbe_slave *slave,
+ int max_rx_len)
+{
+ if (max_rx_len > NETCP_MAX_FRAME_SIZE)
+ max_rx_len = NETCP_MAX_FRAME_SIZE;
+
+ writel(max_rx_len, GBE_REG_ADDR(slave, emac_regs, rx_maxlen));
+ writel(slave->mac_control, GBE_REG_ADDR(slave, emac_regs, mac_control));
+}
+
+static void gbe_slave_stop(struct gbe_intf *intf)
+{
+ struct gbe_priv *gbe_dev = intf->gbe_dev;
+ struct gbe_slave *slave = intf->slave;
+
+ gbe_port_reset(slave);
+ /* Disable forwarding */
+ cpsw_ale_control_set(gbe_dev->ale, slave->port_num,
+ ALE_PORT_STATE, ALE_PORT_STATE_DISABLE);
+ cpsw_ale_del_mcast(gbe_dev->ale, intf->ndev->broadcast,
+ 1 << slave->port_num, 0, 0);
+
+ if (!slave->phy)
+ return;
+
+ phy_stop(slave->phy);
+ phy_disconnect(slave->phy);
+ slave->phy = NULL;
+}
+
+static void gbe_sgmii_config(struct gbe_priv *priv, struct gbe_slave *slave)
+{
+ void __iomem *sgmii_port_regs;
+
+ sgmii_port_regs = priv->sgmii_port_regs;
+ if ((priv->ss_version == GBE_SS_VERSION_14) && (slave->slave_num >= 2))
+ sgmii_port_regs = priv->sgmii_port34_regs;
+
+ netcp_sgmii_reset(sgmii_port_regs, slave->slave_num);
+ netcp_sgmii_config(sgmii_port_regs, slave->slave_num,
+ slave->link_interface);
+}
+
+static int gbe_slave_open(struct gbe_intf *gbe_intf)
+{
+ struct gbe_priv *priv = gbe_intf->gbe_dev;
+ struct gbe_slave *slave = gbe_intf->slave;
+ phy_interface_t phy_mode;
+ bool has_phy = false;
+
+ void (*hndlr)(struct net_device *) = gbe_adjust_link;
+
+ gbe_sgmii_config(priv, slave);
+ gbe_port_reset(slave);
+ gbe_port_config(priv, slave, priv->rx_packet_max);
+ gbe_set_slave_mac(slave, gbe_intf);
+ /* enable forwarding */
+ cpsw_ale_control_set(priv->ale, slave->port_num,
+ ALE_PORT_STATE, ALE_PORT_STATE_FORWARD);
+ cpsw_ale_add_mcast(priv->ale, gbe_intf->ndev->broadcast,
+ 1 << slave->port_num, 0, 0, ALE_MCAST_FWD_2);
+
+ if (slave->link_interface == SGMII_LINK_MAC_PHY) {
+ has_phy = true;
+ phy_mode = PHY_INTERFACE_MODE_SGMII;
+ slave->phy_port_t = PORT_MII;
+ } else if (slave->link_interface == XGMII_LINK_MAC_PHY) {
+ has_phy = true;
+ phy_mode = PHY_INTERFACE_MODE_NA;
+ slave->phy_port_t = PORT_FIBRE;
+ }
+
+ if (has_phy) {
+ slave->phy = of_phy_connect(gbe_intf->ndev,
+ slave->phy_node,
+ hndlr, 0,
+ phy_mode);
+ if (!slave->phy) {
+ dev_err(priv->dev, "phy not found on slave %d\n",
+ slave->slave_num);
+ return -ENODEV;
+ }
+ dev_dbg(priv->dev, "phy found: id is: 0x%s\n",
+ dev_name(&slave->phy->dev));
+ phy_start(slave->phy);
+ phy_read_status(slave->phy);
+ }
+ return 0;
+}
+
+static void gbe_init_host_port(struct gbe_priv *priv)
+{
+ int bypass_en = 1;
+ /* Max length register */
+ writel(NETCP_MAX_FRAME_SIZE, GBE_REG_ADDR(priv, host_port_regs,
+ rx_maxlen));
+
+ cpsw_ale_start(priv->ale);
+
+ if (priv->enable_ale)
+ bypass_en = 0;
+
+ cpsw_ale_control_set(priv->ale, 0, ALE_BYPASS, bypass_en);
+
+ cpsw_ale_control_set(priv->ale, 0, ALE_NO_PORT_VLAN, 1);
+
+ cpsw_ale_control_set(priv->ale, priv->host_port,
+ ALE_PORT_STATE, ALE_PORT_STATE_FORWARD);
+
+ cpsw_ale_control_set(priv->ale, 0,
+ ALE_PORT_UNKNOWN_VLAN_MEMBER,
+ GBE_PORT_MASK(priv->ale_ports));
+
+ cpsw_ale_control_set(priv->ale, 0,
+ ALE_PORT_UNKNOWN_MCAST_FLOOD,
+ GBE_PORT_MASK(priv->ale_ports - 1));
+
+ cpsw_ale_control_set(priv->ale, 0,
+ ALE_PORT_UNKNOWN_REG_MCAST_FLOOD,
+ GBE_PORT_MASK(priv->ale_ports));
+
+ cpsw_ale_control_set(priv->ale, 0,
+ ALE_PORT_UNTAGGED_EGRESS,
+ GBE_PORT_MASK(priv->ale_ports));
+}
+
+static void gbe_add_mcast_addr(struct gbe_intf *gbe_intf, u8 *addr)
+{
+ struct gbe_priv *gbe_dev = gbe_intf->gbe_dev;
+ u16 vlan_id;
+
+ cpsw_ale_add_mcast(gbe_dev->ale, addr,
+ GBE_PORT_MASK(gbe_dev->ale_ports), 0, 0,
+ ALE_MCAST_FWD_2);
+ for_each_set_bit(vlan_id, gbe_intf->active_vlans, VLAN_N_VID) {
+ cpsw_ale_add_mcast(gbe_dev->ale, addr,
+ GBE_PORT_MASK(gbe_dev->ale_ports),
+ ALE_VLAN, vlan_id, ALE_MCAST_FWD_2);
+ }
+}
+
+static void gbe_add_ucast_addr(struct gbe_intf *gbe_intf, u8 *addr)
+{
+ struct gbe_priv *gbe_dev = gbe_intf->gbe_dev;
+ u16 vlan_id;
+
+ cpsw_ale_add_ucast(gbe_dev->ale, addr, gbe_dev->host_port, 0, 0);
+
+ for_each_set_bit(vlan_id, gbe_intf->active_vlans, VLAN_N_VID)
+ cpsw_ale_add_ucast(gbe_dev->ale, addr, gbe_dev->host_port,
+ ALE_VLAN, vlan_id);
+}
+
+static void gbe_del_mcast_addr(struct gbe_intf *gbe_intf, u8 *addr)
+{
+ struct gbe_priv *gbe_dev = gbe_intf->gbe_dev;
+ u16 vlan_id;
+
+ cpsw_ale_del_mcast(gbe_dev->ale, addr, 0, 0, 0);
+
+ for_each_set_bit(vlan_id, gbe_intf->active_vlans, VLAN_N_VID) {
+ cpsw_ale_del_mcast(gbe_dev->ale, addr, 0, ALE_VLAN, vlan_id);
+ }
+}
+
+static void gbe_del_ucast_addr(struct gbe_intf *gbe_intf, u8 *addr)
+{
+ struct gbe_priv *gbe_dev = gbe_intf->gbe_dev;
+ u16 vlan_id;
+
+ cpsw_ale_del_ucast(gbe_dev->ale, addr, gbe_dev->host_port, 0, 0);
+
+ for_each_set_bit(vlan_id, gbe_intf->active_vlans, VLAN_N_VID) {
+ cpsw_ale_del_ucast(gbe_dev->ale, addr, gbe_dev->host_port,
+ ALE_VLAN, vlan_id);
+ }
+}
+
+static int gbe_add_addr(void *intf_priv, struct netcp_addr *naddr)
+{
+ struct gbe_intf *gbe_intf = intf_priv;
+ struct gbe_priv *gbe_dev = gbe_intf->gbe_dev;
+
+ dev_dbg(gbe_dev->dev, "ethss adding address %pM, type %d\n",
+ naddr->addr, naddr->type);
+
+ switch (naddr->type) {
+ case ADDR_MCAST:
+ case ADDR_BCAST:
+ gbe_add_mcast_addr(gbe_intf, naddr->addr);
+ break;
+ case ADDR_UCAST:
+ case ADDR_DEV:
+ gbe_add_ucast_addr(gbe_intf, naddr->addr);
+ break;
+ case ADDR_ANY:
+ /* nothing to do for promiscuous */
+ default:
+ break;
+ }
+
+ return 0;
+}
+
+static int gbe_del_addr(void *intf_priv, struct netcp_addr *naddr)
+{
+ struct gbe_intf *gbe_intf = intf_priv;
+ struct gbe_priv *gbe_dev = gbe_intf->gbe_dev;
+
+ dev_dbg(gbe_dev->dev, "ethss deleting address %pM, type %d\n",
+ naddr->addr, naddr->type);
+
+ switch (naddr->type) {
+ case ADDR_MCAST:
+ case ADDR_BCAST:
+ gbe_del_mcast_addr(gbe_intf, naddr->addr);
+ break;
+ case ADDR_UCAST:
+ case ADDR_DEV:
+ gbe_del_ucast_addr(gbe_intf, naddr->addr);
+ break;
+ case ADDR_ANY:
+ /* nothing to do for promiscuous */
+ default:
+ break;
+ }
+
+ return 0;
+}
+
+static int gbe_add_vid(void *intf_priv, int vid)
+{
+ struct gbe_intf *gbe_intf = intf_priv;
+ struct gbe_priv *gbe_dev = gbe_intf->gbe_dev;
+
+ set_bit(vid, gbe_intf->active_vlans);
+
+ cpsw_ale_add_vlan(gbe_dev->ale, vid,
+ GBE_PORT_MASK(gbe_dev->ale_ports),
+ GBE_MASK_NO_PORTS,
+ GBE_PORT_MASK(gbe_dev->ale_ports),
+ GBE_PORT_MASK(gbe_dev->ale_ports - 1));
+
+ return 0;
+}
+
+static int gbe_del_vid(void *intf_priv, int vid)
+{
+ struct gbe_intf *gbe_intf = intf_priv;
+ struct gbe_priv *gbe_dev = gbe_intf->gbe_dev;
+
+ cpsw_ale_del_vlan(gbe_dev->ale, vid, 0);
+ clear_bit(vid, gbe_intf->active_vlans);
+ return 0;
+}
+
+static int gbe_ioctl(void *intf_priv, struct ifreq *req, int cmd)
+{
+ struct gbe_intf *gbe_intf = intf_priv;
+ struct phy_device *phy = gbe_intf->slave->phy;
+ int ret = -EOPNOTSUPP;
+
+ if (phy)
+ ret = phy_mii_ioctl(phy, req, cmd);
+
+ return ret;
+}
+
+static void netcp_ethss_timer(unsigned long arg)
+{
+ struct gbe_priv *gbe_dev = (struct gbe_priv *)arg;
+ struct gbe_intf *gbe_intf;
+ struct gbe_slave *slave;
+
+ /* Check & update SGMII link state of interfaces */
+ for_each_intf(gbe_intf, gbe_dev) {
+ if (!gbe_intf->slave->open)
+ continue;
+ netcp_ethss_update_link_state(gbe_dev, gbe_intf->slave,
+ gbe_intf->ndev);
+ }
+
+ /* Check & update SGMII link state of secondary ports */
+ for_each_sec_slave(slave, gbe_dev) {
+ netcp_ethss_update_link_state(gbe_dev, slave, NULL);
+ }
+
+ spin_lock_bh(&gbe_dev->hw_stats_lock);
+
+ if (gbe_dev->ss_version == GBE_SS_VERSION_14)
+ gbe_update_stats_ver14(gbe_dev, NULL);
+ else
+ gbe_update_stats(gbe_dev, NULL);
+
+ spin_unlock_bh(&gbe_dev->hw_stats_lock);
+
+ gbe_dev->timer.expires = jiffies + GBE_TIMER_INTERVAL;
+ add_timer(&gbe_dev->timer);
+}
+
+static int gbe_tx_hook(int order, void *data, struct netcp_packet *p_info)
+{
+ struct gbe_intf *gbe_intf = data;
+
+ p_info->tx_pipe = &gbe_intf->tx_pipe;
+ return 0;
+}
+
+static int gbe_open(void *intf_priv, struct net_device *ndev)
+{
+ struct gbe_intf *gbe_intf = intf_priv;
+ struct gbe_priv *gbe_dev = gbe_intf->gbe_dev;
+ struct netcp_intf *netcp = netdev_priv(ndev);
+ struct gbe_slave *slave = gbe_intf->slave;
+ int port_num = slave->port_num;
+ u32 reg;
+ int ret;
+
+ reg = readl(GBE_REG_ADDR(gbe_dev, switch_regs, id_ver));
+ dev_dbg(gbe_dev->dev, "initializing gbe version %d.%d (%d) GBE identification value 0x%x\n",
+ GBE_MAJOR_VERSION(reg), GBE_MINOR_VERSION(reg),
+ GBE_RTL_VERSION(reg), GBE_IDENT(reg));
+
+ if (gbe_dev->enable_ale)
+ gbe_intf->tx_pipe.dma_psflags = 0;
+ else
+ gbe_intf->tx_pipe.dma_psflags = port_num;
+
+ dev_dbg(gbe_dev->dev, "opened TX channel %s: %p with psflags %d\n",
+ gbe_intf->tx_pipe.dma_chan_name,
+ gbe_intf->tx_pipe.dma_channel,
+ gbe_intf->tx_pipe.dma_psflags);
+
+ gbe_slave_stop(gbe_intf);
+
+ /* disable priority elevation and enable statistics on all ports */
+ writel(0, GBE_REG_ADDR(gbe_dev, switch_regs, ptype));
+
+ /* Control register */
+ writel(GBE_CTL_P0_ENABLE, GBE_REG_ADDR(gbe_dev, switch_regs, control));
+
+ /* All statistics enabled and STAT AB visible by default */
+ writel(GBE_REG_VAL_STAT_ENABLE_ALL, GBE_REG_ADDR(gbe_dev, switch_regs,
+ stat_port_en));
+
+ ret = gbe_slave_open(gbe_intf);
+ if (ret)
+ goto fail;
+
+ netcp_register_txhook(netcp, GBE_TXHOOK_ORDER, gbe_tx_hook,
+ gbe_intf);
+
+ slave->open = true;
+ netcp_ethss_update_link_state(gbe_dev, slave, ndev);
+ return 0;
+
+fail:
+ gbe_slave_stop(gbe_intf);
+ return ret;
+}
+
+static int gbe_close(void *intf_priv, struct net_device *ndev)
+{
+ struct gbe_intf *gbe_intf = intf_priv;
+ struct netcp_intf *netcp = netdev_priv(ndev);
+
+ gbe_slave_stop(gbe_intf);
+ netcp_unregister_txhook(netcp, GBE_TXHOOK_ORDER, gbe_tx_hook,
+ gbe_intf);
+
+ gbe_intf->slave->open = false;
+ atomic_set(&gbe_intf->slave->link_state, NETCP_LINK_STATE_INVALID);
+ return 0;
+}
+
+static int init_slave(struct gbe_priv *gbe_dev, struct gbe_slave *slave,
+ struct device_node *node)
+{
+ int port_reg_num;
+ u32 port_reg_ofs, emac_reg_ofs;
+
+ if (of_property_read_u32(node, "slave-port", &slave->slave_num)) {
+ dev_err(gbe_dev->dev, "missing slave-port parameter\n");
+ return -EINVAL;
+ }
+
+ if (of_property_read_u32(node, "link-interface",
+ &slave->link_interface)) {
+ dev_warn(gbe_dev->dev,
+ "missing link-interface value defaulting to 1G mac-phy link\n");
+ slave->link_interface = SGMII_LINK_MAC_PHY;
+ }
+
+ slave->open = false;
+ slave->phy_node = of_parse_phandle(node, "phy-handle", 0);
+ slave->port_num = gbe_get_slave_port(gbe_dev, slave->slave_num);
+
+ slave->mac_control = GBE_DEF_1G_MAC_CONTROL;
+
+ /* Emac regs memmap are contiguous but port regs are not */
+ port_reg_num = slave->slave_num;
+ if (gbe_dev->ss_version == GBE_SS_VERSION_14) {
+ if (slave->slave_num > 1) {
+ port_reg_ofs = GBE13_SLAVE_PORT2_OFFSET;
+ port_reg_num -= 2;
+ } else {
+ port_reg_ofs = GBE13_SLAVE_PORT_OFFSET;
+ }
+ } else {
+ dev_err(gbe_dev->dev, "unknown ethss(0x%x)\n",
+ gbe_dev->ss_version);
+ return -EINVAL;
+ }
+
+ if (gbe_dev->ss_version == GBE_SS_VERSION_14)
+ emac_reg_ofs = GBE13_EMAC_OFFSET;
+
+ slave->port_regs = gbe_dev->ss_regs + port_reg_ofs +
+ (0x30 * port_reg_num);
+ slave->emac_regs = gbe_dev->ss_regs + emac_reg_ofs +
+ (0x40 * slave->slave_num);
+
+ if (gbe_dev->ss_version == GBE_SS_VERSION_14) {
+ /* Initialize slave port register offsets */
+ GBE_SET_REG_OFS(slave, port_regs, port_vlan);
+ GBE_SET_REG_OFS(slave, port_regs, tx_pri_map);
+ GBE_SET_REG_OFS(slave, port_regs, sa_lo);
+ GBE_SET_REG_OFS(slave, port_regs, sa_hi);
+ GBE_SET_REG_OFS(slave, port_regs, ts_ctl);
+ GBE_SET_REG_OFS(slave, port_regs, ts_seq_ltype);
+ GBE_SET_REG_OFS(slave, port_regs, ts_vlan);
+ GBE_SET_REG_OFS(slave, port_regs, ts_ctl_ltype2);
+ GBE_SET_REG_OFS(slave, port_regs, ts_ctl2);
+
+ /* Initialize EMAC register offsets */
+ GBE_SET_REG_OFS(slave, emac_regs, mac_control);
+ GBE_SET_REG_OFS(slave, emac_regs, soft_reset);
+ GBE_SET_REG_OFS(slave, emac_regs, rx_maxlen);
+
+ } else {
+ dev_err(gbe_dev->dev, "unknown ethss(0x%x)\n",
+ gbe_dev->ss_version);
+ return -EINVAL;
+ }
+
+ atomic_set(&slave->link_state, NETCP_LINK_STATE_INVALID);
+ return 0;
+}
+
+static void init_secondary_ports(struct gbe_priv *gbe_dev,
+ struct device_node *node)
+{
+ struct device *dev = gbe_dev->dev;
+ phy_interface_t phy_mode;
+ struct gbe_priv **priv;
+ struct device_node *port;
+ struct gbe_slave *slave;
+ bool mac_phy_link = false;
+
+ for_each_child_of_node(node, port) {
+ slave = devm_kzalloc(dev, sizeof(*slave), GFP_KERNEL);
+ if (!slave) {
+ dev_err(dev,
+ "memomry alloc failed for secondary port(%s), skipping...\n",
+ port->name);
+ continue;
+ }
+
+ if (init_slave(gbe_dev, slave, port)) {
+ dev_err(dev,
+ "Failed to initialize secondary port(%s), skipping...\n",
+ port->name);
+ devm_kfree(dev, slave);
+ continue;
+ }
+
+ gbe_sgmii_config(gbe_dev, slave);
+ gbe_port_reset(slave);
+ gbe_port_config(gbe_dev, slave, gbe_dev->rx_packet_max);
+ list_add_tail(&slave->slave_list, &gbe_dev->secondary_slaves);
+ gbe_dev->num_slaves++;
+ if (slave->link_interface == SGMII_LINK_MAC_PHY)
+ mac_phy_link = true;
+
+ slave->open = true;
+ }
+
+ /* of_phy_connect() is needed only for MAC-PHY interface */
+ if (!mac_phy_link)
+ return;
+
+ /* Allocate dummy netdev device for attaching to phy device */
+ gbe_dev->dummy_ndev = alloc_netdev(sizeof(gbe_dev), "dummy",
+ NET_NAME_UNKNOWN, ether_setup);
+ if (!gbe_dev->dummy_ndev) {
+ dev_err(dev,
+ "Failed to allocate dummy netdev for secondary ports, skipping phy_connect()...\n");
+ return;
+ }
+ priv = netdev_priv(gbe_dev->dummy_ndev);
+ *priv = gbe_dev;
+
+ if (slave->link_interface == SGMII_LINK_MAC_PHY) {
+ phy_mode = PHY_INTERFACE_MODE_SGMII;
+ slave->phy_port_t = PORT_MII;
+ } else {
+ phy_mode = PHY_INTERFACE_MODE_NA;
+ slave->phy_port_t = PORT_FIBRE;
+ }
+
+ for_each_sec_slave(slave, gbe_dev) {
+ if (slave->link_interface != SGMII_LINK_MAC_PHY)
+ continue;
+ slave->phy =
+ of_phy_connect(gbe_dev->dummy_ndev,
+ slave->phy_node,
+ gbe_adjust_link_sec_slaves,
+ 0, phy_mode);
+ if (!slave->phy) {
+ dev_err(dev, "phy not found for slave %d\n",
+ slave->slave_num);
+ slave->phy = NULL;
+ } else {
+ dev_dbg(dev, "phy found: id is: 0x%s\n",
+ dev_name(&slave->phy->dev));
+ phy_start(slave->phy);
+ phy_read_status(slave->phy);
+ }
+ }
+}
+
+static void free_secondary_ports(struct gbe_priv *gbe_dev)
+{
+ struct gbe_slave *slave;
+
+ for (;;) {
+ slave = first_sec_slave(gbe_dev);
+ if (!slave)
+ break;
+ if (slave->phy)
+ phy_disconnect(slave->phy);
+ list_del(&slave->slave_list);
+ }
+ if (gbe_dev->dummy_ndev)
+ free_netdev(gbe_dev->dummy_ndev);
+}
+
+static int get_gbe_resource_version(struct gbe_priv *gbe_dev,
+ struct device_node *node)
+{
+ struct resource res;
+ void __iomem *regs;
+ int ret;
+
+ ret = of_address_to_resource(node, 0, &res);
+ if (ret) {
+ dev_err(gbe_dev->dev, "Can't translate of node(%s) address\n",
+ node->name);
+ return ret;
+ }
+
+ regs = devm_ioremap_resource(gbe_dev->dev, &res);
+ if (IS_ERR(regs)) {
+ dev_err(gbe_dev->dev, "Failed to map gbe register base\n");
+ return PTR_ERR(regs);
+ }
+ gbe_dev->ss_regs = regs;
+ gbe_dev->ss_version = readl(gbe_dev->ss_regs);
+ return 0;
+}
+
+static int set_gbe_ethss14_priv(struct gbe_priv *gbe_dev,
+ struct device_node *node)
+{
+ void __iomem *regs;
+ int i;
+
+ gbe_dev->hw_stats = devm_kzalloc(gbe_dev->dev,
+ GBE13_NUM_HW_STAT_ENTRIES *
+ GBE13_NUM_SLAVES * sizeof(u64),
+ GFP_KERNEL);
+ if (!gbe_dev->hw_stats) {
+ dev_err(gbe_dev->dev, "hw_stats memory allocation failed\n");
+ return -ENOMEM;
+ }
+
+ regs = gbe_dev->ss_regs;
+ gbe_dev->sgmii_port_regs = regs + GBE13_SGMII_MODULE_OFFSET;
+ gbe_dev->sgmii_port34_regs = regs + GBE13_SGMII34_MODULE_OFFSET;
+ gbe_dev->switch_regs = regs + GBE13_SWITCH_MODULE_OFFSET;
+ gbe_dev->host_port_regs = regs + GBE13_HOST_PORT_OFFSET;
+
+ for (i = 0; i < GBE13_NUM_HW_STATS_MOD; i++)
+ gbe_dev->hw_stats_regs[i] = regs + GBE13_HW_STATS_OFFSET +
+ (GBE_HW_STATS_REG_MAP_SZ * i);
+
+ gbe_dev->ale_reg = regs + GBE13_ALE_OFFSET;
+ gbe_dev->ale_ports = GBE13_NUM_ALE_PORTS;
+ gbe_dev->host_port = GBE13_HOST_PORT_NUM;
+ gbe_dev->ale_entries = GBE13_NUM_ALE_ENTRIES;
+ gbe_dev->et_stats = gbe13_et_stats;
+ gbe_dev->num_et_stats = ARRAY_SIZE(gbe13_et_stats);
+
+ /* Subsystem registers */
+ GBE_SET_REG_OFS(gbe_dev, ss_regs, id_ver);
+
+ /* Switch module registers */
+ GBE_SET_REG_OFS(gbe_dev, switch_regs, id_ver);
+ GBE_SET_REG_OFS(gbe_dev, switch_regs, control);
+ GBE_SET_REG_OFS(gbe_dev, switch_regs, soft_reset);
+ GBE_SET_REG_OFS(gbe_dev, switch_regs, stat_port_en);
+ GBE_SET_REG_OFS(gbe_dev, switch_regs, ptype);
+ GBE_SET_REG_OFS(gbe_dev, switch_regs, flow_control);
+
+ /* Host port registers */
+ GBE_SET_REG_OFS(gbe_dev, host_port_regs, port_vlan);
+ GBE_SET_REG_OFS(gbe_dev, host_port_regs, rx_maxlen);
+ return 0;
+}
+
+static int gbe_probe(struct netcp_device *netcp_device, struct device *dev,
+ struct device_node *node, void **inst_priv)
+{
+ struct device_node *interfaces, *interface;
+ struct device_node *secondary_ports;
+ struct cpsw_ale_params ale_params;
+ struct gbe_priv *gbe_dev;
+ u32 slave_num;
+ int ret = 0;
+
+ if (!node) {
+ dev_err(dev, "device tree info unavailable\n");
+ return -ENODEV;
+ }
+
+ gbe_dev = devm_kzalloc(dev, sizeof(struct gbe_priv), GFP_KERNEL);
+ if (!gbe_dev)
+ return -ENOMEM;
+
+ gbe_dev->dev = dev;
+ gbe_dev->netcp_device = netcp_device;
+ gbe_dev->rx_packet_max = NETCP_MAX_FRAME_SIZE;
+
+ /* init the hw stats lock */
+ spin_lock_init(&gbe_dev->hw_stats_lock);
+
+ if (of_find_property(node, "enable-ale", NULL)) {
+ gbe_dev->enable_ale = true;
+ dev_info(dev, "ALE enabled\n");
+ } else {
+ gbe_dev->enable_ale = false;
+ dev_dbg(dev, "ALE bypass enabled*\n");
+ }
+
+ ret = of_property_read_u32(node, "tx-queue",
+ &gbe_dev->tx_queue_id);
+ if (ret < 0) {
+ dev_err(dev, "missing tx_queue parameter\n");
+ gbe_dev->tx_queue_id = GBE_TX_QUEUE;
+ }
+
+ ret = of_property_read_string(node, "tx-channel",
+ &gbe_dev->dma_chan_name);
+ if (ret < 0) {
+ dev_err(dev, "missing \"tx-channel\" parameter\n");
+ ret = -ENODEV;
+ goto quit;
+ }
+
+ if (!strcmp(node->name, "gbe")) {
+ ret = get_gbe_resource_version(gbe_dev, node);
+ if (ret)
+ goto quit;
+
+ ret = set_gbe_ethss14_priv(gbe_dev, node);
+ if (ret)
+ goto quit;
+ } else {
+ dev_err(dev, "unknown GBE node(%s)\n", node->name);
+ ret = -ENODEV;
+ goto quit;
+ }
+
+ interfaces = of_get_child_by_name(node, "interfaces");
+ if (!interfaces)
+ dev_err(dev, "could not find interfaces\n");
+
+ ret = netcp_txpipe_init(&gbe_dev->tx_pipe, netcp_device,
+ gbe_dev->dma_chan_name, gbe_dev->tx_queue_id);
+ if (ret)
+ goto quit;
+
+ ret = netcp_txpipe_open(&gbe_dev->tx_pipe);
+ if (ret)
+ goto quit;
+
+ /* Create network interfaces */
+ INIT_LIST_HEAD(&gbe_dev->gbe_intf_head);
+ for_each_child_of_node(interfaces, interface) {
+ ret = of_property_read_u32(interface, "slave-port", &slave_num);
+ if (ret) {
+ dev_err(dev, "missing slave-port parameter, skipping interface configuration for %s\n",
+ interface->name);
+ continue;
+ }
+ gbe_dev->num_slaves++;
+ }
+
+ if (!gbe_dev->num_slaves)
+ dev_warn(dev, "No network interface configured\n");
+
+ /* Initialize Secondary slave ports */
+ secondary_ports = of_get_child_by_name(node, "secondary-slave-ports");
+ INIT_LIST_HEAD(&gbe_dev->secondary_slaves);
+ if (secondary_ports)
+ init_secondary_ports(gbe_dev, secondary_ports);
+ of_node_put(secondary_ports);
+
+ if (!gbe_dev->num_slaves) {
+ dev_err(dev, "No network interface or secondary ports configured\n");
+ ret = -ENODEV;
+ goto quit;
+ }
+
+ memset(&ale_params, 0, sizeof(ale_params));
+ ale_params.dev = gbe_dev->dev;
+ ale_params.ale_regs = gbe_dev->ale_reg;
+ ale_params.ale_ageout = GBE_DEFAULT_ALE_AGEOUT;
+ ale_params.ale_entries = gbe_dev->ale_entries;
+ ale_params.ale_ports = gbe_dev->ale_ports;
+
+ gbe_dev->ale = cpsw_ale_create(&ale_params);
+ if (!gbe_dev->ale) {
+ dev_err(gbe_dev->dev, "error initializing ale engine\n");
+ ret = -ENODEV;
+ goto quit;
+ } else {
+ dev_dbg(gbe_dev->dev, "Created a gbe ale engine\n");
+ }
+
+ /* initialize host port */
+ gbe_init_host_port(gbe_dev);
+
+ init_timer(&gbe_dev->timer);
+ gbe_dev->timer.data = (unsigned long)gbe_dev;
+ gbe_dev->timer.function = netcp_ethss_timer;
+ gbe_dev->timer.expires = jiffies + GBE_TIMER_INTERVAL;
+ add_timer(&gbe_dev->timer);
+ *inst_priv = gbe_dev;
+ return 0;
+
+quit:
+ if (gbe_dev->hw_stats)
+ devm_kfree(dev, gbe_dev->hw_stats);
+ if (gbe_dev->ale)
+ cpsw_ale_destroy(gbe_dev->ale);
+ if (gbe_dev->ss_regs)
+ devm_iounmap(dev, gbe_dev->ss_regs);
+ if (interfaces)
+ of_node_put(interfaces);
+ devm_kfree(dev, gbe_dev);
+ return ret;
+}
+
+static int gbe_attach(void *inst_priv, struct net_device *ndev,
+ struct device_node *node, void **intf_priv)
+{
+ struct gbe_priv *gbe_dev = inst_priv;
+ struct gbe_intf *gbe_intf;
+ int ret;
+
+ if (!node) {
+ dev_err(gbe_dev->dev, "interface node not available\n");
+ return -ENODEV;
+ }
+
+ gbe_intf = devm_kzalloc(gbe_dev->dev, sizeof(*gbe_intf), GFP_KERNEL);
+ if (!gbe_intf)
+ return -ENOMEM;
+
+ gbe_intf->ndev = ndev;
+ gbe_intf->dev = gbe_dev->dev;
+ gbe_intf->gbe_dev = gbe_dev;
+
+ gbe_intf->slave = devm_kzalloc(gbe_dev->dev,
+ sizeof(*gbe_intf->slave),
+ GFP_KERNEL);
+ if (!gbe_intf->slave) {
+ ret = -ENOMEM;
+ goto fail;
+ }
+
+ if (init_slave(gbe_dev, gbe_intf->slave, node)) {
+ ret = -ENODEV;
+ goto fail;
+ }
+
+ gbe_intf->tx_pipe = gbe_dev->tx_pipe;
+ ndev->ethtool_ops = &keystone_ethtool_ops;
+ list_add_tail(&gbe_intf->gbe_intf_list, &gbe_dev->gbe_intf_head);
+ *intf_priv = gbe_intf;
+ return 0;
+
+fail:
+ if (gbe_intf->slave)
+ devm_kfree(gbe_dev->dev, gbe_intf->slave);
+ if (gbe_intf)
+ devm_kfree(gbe_dev->dev, gbe_intf);
+ return ret;
+}
+
+static int gbe_release(void *intf_priv)
+{
+ struct gbe_intf *gbe_intf = intf_priv;
+
+ gbe_intf->ndev->ethtool_ops = NULL;
+ list_del(&gbe_intf->gbe_intf_list);
+ devm_kfree(gbe_intf->dev, gbe_intf->slave);
+ devm_kfree(gbe_intf->dev, gbe_intf);
+ return 0;
+}
+
+static int gbe_remove(struct netcp_device *netcp_device, void *inst_priv)
+{
+ struct gbe_priv *gbe_dev = inst_priv;
+
+ del_timer_sync(&gbe_dev->timer);
+ cpsw_ale_stop(gbe_dev->ale);
+ cpsw_ale_destroy(gbe_dev->ale);
+ netcp_txpipe_close(&gbe_dev->tx_pipe);
+ free_secondary_ports(gbe_dev);
+
+ if (!list_empty(&gbe_dev->gbe_intf_head))
+ dev_alert(gbe_dev->dev, "unreleased ethss interfaces present\n");
+
+ devm_kfree(gbe_dev->dev, gbe_dev->hw_stats);
+ devm_iounmap(gbe_dev->dev, gbe_dev->ss_regs);
+ memset(gbe_dev, 0x00, sizeof(*gbe_dev));
+ devm_kfree(gbe_dev->dev, gbe_dev);
+ return 0;
+}
+
+static struct netcp_module gbe_module = {
+ .name = GBE_MODULE_NAME,
+ .owner = THIS_MODULE,
+ .primary = true,
+ .probe = gbe_probe,
+ .open = gbe_open,
+ .close = gbe_close,
+ .remove = gbe_remove,
+ .attach = gbe_attach,
+ .release = gbe_release,
+ .add_addr = gbe_add_addr,
+ .del_addr = gbe_del_addr,
+ .add_vid = gbe_add_vid,
+ .del_vid = gbe_del_vid,
+ .ioctl = gbe_ioctl,
+};
+
+static int __init keystone_gbe_init(void)
+{
+ int ret;
+
+ ret = netcp_register_module(&gbe_module);
+ if (ret)
+ return ret;
+
+ return 0;
+}
+module_init(keystone_gbe_init);
+
+static void __exit keystone_gbe_exit(void)
+{
+ netcp_unregister_module(&gbe_module);
+}
+module_exit(keystone_gbe_exit);
diff --git a/drivers/net/ethernet/ti/netcp_sgmii.c b/drivers/net/ethernet/ti/netcp_sgmii.c
new file mode 100644
index 0000000..dbeb142
--- /dev/null
+++ b/drivers/net/ethernet/ti/netcp_sgmii.c
@@ -0,0 +1,131 @@
+/*
+ * SGMI module initialisation
+ *
+ * Copyright (C) 2014 Texas Instruments Incorporated
+ * Authors: Sandeep Nair <sandeep_n@ti.com>
+ * Sandeep Paulraj <s-paulraj@ti.com>
+ * Wingman Kwok <w-kwok2@ti.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 version 2.
+ *
+ * This program is distributed "as is" WITHOUT ANY WARRANTY of any
+ * kind, whether express or implied; without even the implied warranty
+ * of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ */
+
+#include "netcp.h"
+
+#define SGMII_REG_STATUS_LOCK BIT(4)
+#define SGMII_REG_STATUS_LINK BIT(0)
+#define SGMII_REG_STATUS_AUTONEG BIT(2)
+#define SGMII_REG_CONTROL_AUTONEG BIT(0)
+
+#define SGMII23_OFFSET(x) ((x - 2) * 0x100)
+#define SGMII_OFFSET(x) ((x <= 1) ? (x * 0x100) : (SGMII23_OFFSET(x)))
+
+/* SGMII registers */
+#define SGMII_SRESET_REG(x) (SGMII_OFFSET(x) + 0x004)
+#define SGMII_CTL_REG(x) (SGMII_OFFSET(x) + 0x010)
+#define SGMII_STATUS_REG(x) (SGMII_OFFSET(x) + 0x014)
+#define SGMII_MRADV_REG(x) (SGMII_OFFSET(x) + 0x018)
+
+static void sgmii_write_reg(void __iomem *base, int reg, u32 val)
+{
+ writel(val, base + reg);
+}
+
+static u32 sgmii_read_reg(void __iomem *base, int reg)
+{
+ return readl(base + reg);
+}
+
+static void sgmii_write_reg_bit(void __iomem *base, int reg, u32 val)
+{
+ writel((readl(base + reg) | val), base + reg);
+}
+
+/* port is 0 based */
+int netcp_sgmii_reset(void __iomem *sgmii_ofs, int port)
+{
+ /* Soft reset */
+ sgmii_write_reg_bit(sgmii_ofs, SGMII_SRESET_REG(port), 0x1);
+ while (sgmii_read_reg(sgmii_ofs, SGMII_SRESET_REG(port)) != 0x0)
+ ;
+ return 0;
+}
+
+int netcp_sgmii_get_port_link(void __iomem *sgmii_ofs, int port)
+{
+ u32 status = 0, link = 0;
+
+ status = sgmii_read_reg(sgmii_ofs, SGMII_STATUS_REG(port));
+ if ((status & SGMII_REG_STATUS_LINK) != 0)
+ link = 1;
+ return link;
+}
+
+int netcp_sgmii_config(void __iomem *sgmii_ofs, int port, u32 interface)
+{
+ unsigned int i, status, mask;
+ u32 mr_adv_ability;
+ u32 control;
+
+ switch (interface) {
+ case SGMII_LINK_MAC_MAC_AUTONEG:
+ mr_adv_ability = 0x9801;
+ control = 0x21;
+ break;
+
+ case SGMII_LINK_MAC_PHY:
+ case SGMII_LINK_MAC_PHY_NO_MDIO:
+ mr_adv_ability = 1;
+ control = 1;
+ break;
+
+ case SGMII_LINK_MAC_MAC_FORCED:
+ mr_adv_ability = 0x9801;
+ control = 0x20;
+ break;
+
+ case SGMII_LINK_MAC_FIBER:
+ mr_adv_ability = 0x20;
+ control = 0x1;
+ break;
+
+ default:
+ WARN_ONCE(1, "Invalid sgmii interface: %d\n", interface);
+ return -EINVAL;
+ }
+
+ sgmii_write_reg(sgmii_ofs, SGMII_CTL_REG(port), 0);
+
+ /* Wait for the SerDes pll to lock */
+ for (i = 0; i < 1000; i++) {
+ usleep_range(1000, 2000);
+ status = sgmii_read_reg(sgmii_ofs, SGMII_STATUS_REG(port));
+ if ((status & SGMII_REG_STATUS_LOCK) != 0)
+ break;
+ }
+
+ if ((status & SGMII_REG_STATUS_LOCK) == 0)
+ pr_err("serdes PLL not locked\n");
+
+ sgmii_write_reg(sgmii_ofs, SGMII_MRADV_REG(port), mr_adv_ability);
+ sgmii_write_reg(sgmii_ofs, SGMII_CTL_REG(port), control);
+
+ mask = SGMII_REG_STATUS_LINK;
+ if (control & SGMII_REG_CONTROL_AUTONEG)
+ mask |= SGMII_REG_STATUS_AUTONEG;
+
+ for (i = 0; i < 1000; i++) {
+ usleep_range(200, 500);
+ status = sgmii_read_reg(sgmii_ofs, SGMII_STATUS_REG(port));
+ if ((status & mask) == mask)
+ break;
+ }
+
+ return 0;
+}
--
1.7.9.5
^ permalink raw reply related
* [PATCH net-next v8 4/4] net: netcp: Enhance GBE driver to support 10G Ethernet
From: Murali Karicheri @ 2015-01-16 0:12 UTC (permalink / raw)
To: davem, devicetree, linux-kernel, netdev; +Cc: Wingman Kwok, Murali Karicheri
In-Reply-To: <1421367172-19793-1-git-send-email-m-karicheri2@ti.com>
From: Wingman Kwok <w-kwok2@ti.com>
This patch enhances the NetCP gbe driver to support 10GbE subsystem
available in Keystone NetCP. The 3-port 10GbE switch sub-module contains
the following components:- 10GbE Switch, MDIO Module, 2 PCS-R Modules
(10GBase-R) and 2 SGMII modules (10/100/1000Base-T). The GBE driver
together with netcp core driver provides support for 10G Ethernet
on Keystone SoCs.
10GbE hardware spec is available at
http://www.ti.com/general/docs/lit/getliterature.tsp?baseLiteratureNumber=spruhj5&fileType=pdf
Cc: David Miller <davem@davemloft.net>
Cc: Rob Herring <robh+dt@kernel.org>
Cc: Grant Likely <grant.likely@linaro.org>
Cc: Santosh Shilimkar <santosh.shilimkar@kernel.org>
Cc: Pawel Moll <pawel.moll@arm.com>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Ian Campbell <ijc+devicetree@hellion.org.uk>
Cc: Kumar Gala <galak@codeaurora.org>
Signed-off-by: Wingman Kwok <w-kwok2@ti.com>
Signed-off-by: Murali Karicheri <m-karicheri2@ti.com>
---
drivers/net/ethernet/ti/Makefile | 2 +-
drivers/net/ethernet/ti/netcp_ethss.c | 470 +++++++++++++++++++++++++++-
drivers/net/ethernet/ti/netcp_xgbepcsr.c | 501 ++++++++++++++++++++++++++++++
3 files changed, 958 insertions(+), 15 deletions(-)
create mode 100644 drivers/net/ethernet/ti/netcp_xgbepcsr.c
diff --git a/drivers/net/ethernet/ti/Makefile b/drivers/net/ethernet/ti/Makefile
index 6d4dde8..465d03d 100644
--- a/drivers/net/ethernet/ti/Makefile
+++ b/drivers/net/ethernet/ti/Makefile
@@ -13,4 +13,4 @@ ti_cpsw-y := cpsw_ale.o cpsw.o cpts.o
obj-$(CONFIG_TI_KEYSTONE_NETCP) += keystone_netcp.o
keystone_netcp-y := netcp_core.o netcp_ethss.o netcp_sgmii.o \
- cpsw_ale.o cpts.o
+ netcp_xgbepcsr.o cpsw_ale.o cpts.o
diff --git a/drivers/net/ethernet/ti/netcp_ethss.c b/drivers/net/ethernet/ti/netcp_ethss.c
index 58f00c5..fa1041a 100644
--- a/drivers/net/ethernet/ti/netcp_ethss.c
+++ b/drivers/net/ethernet/ti/netcp_ethss.c
@@ -1,5 +1,5 @@
/*
- * Keystone GBE subsystem code
+ * Keystone GBE and XGBE subsystem code
*
* Copyright (C) 2014 Texas Instruments Incorporated
* Authors: Sandeep Nair <sandeep_n@ti.com>
@@ -53,6 +53,23 @@
#define GBE13_NUM_ALE_PORTS (GBE13_NUM_SLAVES + 1)
#define GBE13_NUM_ALE_ENTRIES 1024
+/* 10G Ethernet SS defines */
+#define XGBE_MODULE_NAME "netcp-xgbe"
+#define XGBE_SS_VERSION_10 0x4ee42100
+
+#define XGBE_SERDES_REG_INDEX 1
+#define XGBE10_SGMII_MODULE_OFFSET 0x100
+#define XGBE10_SWITCH_MODULE_OFFSET 0x1000
+#define XGBE10_HOST_PORT_OFFSET 0x1034
+#define XGBE10_SLAVE_PORT_OFFSET 0x1064
+#define XGBE10_EMAC_OFFSET 0x1400
+#define XGBE10_ALE_OFFSET 0x1700
+#define XGBE10_HW_STATS_OFFSET 0x1800
+#define XGBE10_HOST_PORT_NUM 0
+#define XGBE10_NUM_SLAVES 2
+#define XGBE10_NUM_ALE_PORTS (XGBE10_NUM_SLAVES + 1)
+#define XGBE10_NUM_ALE_ENTRIES 1024
+
#define GBE_TIMER_INTERVAL (HZ / 2)
/* Soft reset register values */
@@ -63,12 +80,15 @@
#define MACSL_RX_ENABLE_CSF BIT(23)
#define MACSL_ENABLE_EXT_CTL BIT(18)
+#define MACSL_XGMII_ENABLE BIT(13)
+#define MACSL_XGIG_MODE BIT(8)
#define MACSL_GIG_MODE BIT(7)
#define MACSL_GMII_ENABLE BIT(5)
#define MACSL_FULLDUPLEX BIT(0)
#define GBE_CTL_P0_ENABLE BIT(2)
#define GBE_REG_VAL_STAT_ENABLE_ALL 0xff
+#define XGBE_REG_VAL_STAT_ENABLE_ALL 0xf
#define GBE_STATS_CD_SEL BIT(28)
#define GBE_PORT_MASK(x) (BIT(x) - 1)
@@ -78,11 +98,19 @@
(MACSL_GIG_MODE | MACSL_GMII_ENABLE | \
MACSL_ENABLE_EXT_CTL | MACSL_RX_ENABLE_CSF)
+#define GBE_DEF_10G_MAC_CONTROL \
+ (MACSL_XGIG_MODE | MACSL_XGMII_ENABLE | \
+ MACSL_ENABLE_EXT_CTL | MACSL_RX_ENABLE_CSF)
+
#define GBE_STATSA_MODULE 0
#define GBE_STATSB_MODULE 1
#define GBE_STATSC_MODULE 2
#define GBE_STATSD_MODULE 3
+#define XGBE_STATS0_MODULE 0
+#define XGBE_STATS1_MODULE 1
+#define XGBE_STATS2_MODULE 2
+
#define MAX_SLAVES GBE13_NUM_SLAVES
/* s: 0-based slave_port */
#define SGMII_BASE(s) \
@@ -91,12 +119,144 @@
#define GBE_TX_QUEUE 648
#define GBE_TXHOOK_ORDER 0
#define GBE_DEFAULT_ALE_AGEOUT 30
+#define SLAVE_LINK_IS_XGMII(s) ((s)->link_interface >= XGMII_LINK_MAC_PHY)
#define NETCP_LINK_STATE_INVALID -1
#define GBE_SET_REG_OFS(p, rb, rn) p->rb##_ofs.rn = \
offsetof(struct gbe##_##rb, rn)
+#define XGBE_SET_REG_OFS(p, rb, rn) p->rb##_ofs.rn = \
+ offsetof(struct xgbe##_##rb, rn)
#define GBE_REG_ADDR(p, rb, rn) (p->rb + p->rb##_ofs.rn)
+struct xgbe_ss_regs {
+ u32 id_ver;
+ u32 synce_count;
+ u32 synce_mux;
+ u32 control;
+};
+
+struct xgbe_switch_regs {
+ u32 id_ver;
+ u32 control;
+ u32 emcontrol;
+ u32 stat_port_en;
+ u32 ptype;
+ u32 soft_idle;
+ u32 thru_rate;
+ u32 gap_thresh;
+ u32 tx_start_wds;
+ u32 flow_control;
+ u32 cppi_thresh;
+};
+
+struct xgbe_port_regs {
+ u32 blk_cnt;
+ u32 port_vlan;
+ u32 tx_pri_map;
+ u32 sa_lo;
+ u32 sa_hi;
+ u32 ts_ctl;
+ u32 ts_seq_ltype;
+ u32 ts_vlan;
+ u32 ts_ctl_ltype2;
+ u32 ts_ctl2;
+ u32 control;
+};
+
+struct xgbe_host_port_regs {
+ u32 blk_cnt;
+ u32 port_vlan;
+ u32 tx_pri_map;
+ u32 src_id;
+ u32 rx_pri_map;
+ u32 rx_maxlen;
+};
+
+struct xgbe_emac_regs {
+ u32 id_ver;
+ u32 mac_control;
+ u32 mac_status;
+ u32 soft_reset;
+ u32 rx_maxlen;
+ u32 __reserved_0;
+ u32 rx_pause;
+ u32 tx_pause;
+ u32 em_control;
+ u32 __reserved_1;
+ u32 tx_gap;
+ u32 rsvd[4];
+};
+
+struct xgbe_host_hw_stats {
+ u32 rx_good_frames;
+ u32 rx_broadcast_frames;
+ u32 rx_multicast_frames;
+ u32 __rsvd_0[3];
+ u32 rx_oversized_frames;
+ u32 __rsvd_1;
+ u32 rx_undersized_frames;
+ u32 __rsvd_2;
+ u32 overrun_type4;
+ u32 overrun_type5;
+ u32 rx_bytes;
+ u32 tx_good_frames;
+ u32 tx_broadcast_frames;
+ u32 tx_multicast_frames;
+ u32 __rsvd_3[9];
+ u32 tx_bytes;
+ u32 tx_64byte_frames;
+ u32 tx_65_to_127byte_frames;
+ u32 tx_128_to_255byte_frames;
+ u32 tx_256_to_511byte_frames;
+ u32 tx_512_to_1023byte_frames;
+ u32 tx_1024byte_frames;
+ u32 net_bytes;
+ u32 rx_sof_overruns;
+ u32 rx_mof_overruns;
+ u32 rx_dma_overruns;
+};
+
+struct xgbe_hw_stats {
+ u32 rx_good_frames;
+ u32 rx_broadcast_frames;
+ u32 rx_multicast_frames;
+ u32 rx_pause_frames;
+ u32 rx_crc_errors;
+ u32 rx_align_code_errors;
+ u32 rx_oversized_frames;
+ u32 rx_jabber_frames;
+ u32 rx_undersized_frames;
+ u32 rx_fragments;
+ u32 overrun_type4;
+ u32 overrun_type5;
+ u32 rx_bytes;
+ u32 tx_good_frames;
+ u32 tx_broadcast_frames;
+ u32 tx_multicast_frames;
+ u32 tx_pause_frames;
+ u32 tx_deferred_frames;
+ u32 tx_collision_frames;
+ u32 tx_single_coll_frames;
+ u32 tx_mult_coll_frames;
+ u32 tx_excessive_collisions;
+ u32 tx_late_collisions;
+ u32 tx_underrun;
+ u32 tx_carrier_sense_errors;
+ u32 tx_bytes;
+ u32 tx_64byte_frames;
+ u32 tx_65_to_127byte_frames;
+ u32 tx_128_to_255byte_frames;
+ u32 tx_256_to_511byte_frames;
+ u32 tx_512_to_1023byte_frames;
+ u32 tx_1024byte_frames;
+ u32 net_bytes;
+ u32 rx_sof_overruns;
+ u32 rx_mof_overruns;
+ u32 rx_dma_overruns;
+};
+
+#define XGBE10_NUM_STAT_ENTRIES (sizeof(struct xgbe_hw_stats)/sizeof(u32))
+
struct gbe_ss_regs {
u32 id_ver;
u32 synce_count;
@@ -230,6 +390,7 @@ struct gbe_hw_stats {
#define GBE13_NUM_HW_STAT_ENTRIES (sizeof(struct gbe_hw_stats)/sizeof(u32))
#define GBE13_NUM_HW_STATS_MOD 2
+#define XGBE10_NUM_HW_STATS_MOD 3
#define GBE_MAX_HW_STAT_MODS 3
#define GBE_HW_STATS_REG_MAP_SZ 0x100
@@ -303,6 +464,7 @@ struct gbe_intf {
};
static struct netcp_module gbe_module;
+static struct netcp_module xgbe_module;
/* Statistic management */
struct netcp_ethtool_stat {
@@ -471,6 +633,118 @@ static const struct netcp_ethtool_stat gbe13_et_stats[] = {
{GBE_STATSD_INFO(rx_dma_overruns)},
};
+#define XGBE_STATS0_INFO(field) "GBE_0:"#field, XGBE_STATS0_MODULE, \
+ FIELD_SIZEOF(struct xgbe_hw_stats, field), \
+ offsetof(struct xgbe_hw_stats, field)
+
+#define XGBE_STATS1_INFO(field) "GBE_1:"#field, XGBE_STATS1_MODULE, \
+ FIELD_SIZEOF(struct xgbe_hw_stats, field), \
+ offsetof(struct xgbe_hw_stats, field)
+
+#define XGBE_STATS2_INFO(field) "GBE_2:"#field, XGBE_STATS2_MODULE, \
+ FIELD_SIZEOF(struct xgbe_hw_stats, field), \
+ offsetof(struct xgbe_hw_stats, field)
+
+static const struct netcp_ethtool_stat xgbe10_et_stats[] = {
+ /* GBE module 0 */
+ {XGBE_STATS0_INFO(rx_good_frames)},
+ {XGBE_STATS0_INFO(rx_broadcast_frames)},
+ {XGBE_STATS0_INFO(rx_multicast_frames)},
+ {XGBE_STATS0_INFO(rx_oversized_frames)},
+ {XGBE_STATS0_INFO(rx_undersized_frames)},
+ {XGBE_STATS0_INFO(overrun_type4)},
+ {XGBE_STATS0_INFO(overrun_type5)},
+ {XGBE_STATS0_INFO(rx_bytes)},
+ {XGBE_STATS0_INFO(tx_good_frames)},
+ {XGBE_STATS0_INFO(tx_broadcast_frames)},
+ {XGBE_STATS0_INFO(tx_multicast_frames)},
+ {XGBE_STATS0_INFO(tx_bytes)},
+ {XGBE_STATS0_INFO(tx_64byte_frames)},
+ {XGBE_STATS0_INFO(tx_65_to_127byte_frames)},
+ {XGBE_STATS0_INFO(tx_128_to_255byte_frames)},
+ {XGBE_STATS0_INFO(tx_256_to_511byte_frames)},
+ {XGBE_STATS0_INFO(tx_512_to_1023byte_frames)},
+ {XGBE_STATS0_INFO(tx_1024byte_frames)},
+ {XGBE_STATS0_INFO(net_bytes)},
+ {XGBE_STATS0_INFO(rx_sof_overruns)},
+ {XGBE_STATS0_INFO(rx_mof_overruns)},
+ {XGBE_STATS0_INFO(rx_dma_overruns)},
+ /* XGBE module 1 */
+ {XGBE_STATS1_INFO(rx_good_frames)},
+ {XGBE_STATS1_INFO(rx_broadcast_frames)},
+ {XGBE_STATS1_INFO(rx_multicast_frames)},
+ {XGBE_STATS1_INFO(rx_pause_frames)},
+ {XGBE_STATS1_INFO(rx_crc_errors)},
+ {XGBE_STATS1_INFO(rx_align_code_errors)},
+ {XGBE_STATS1_INFO(rx_oversized_frames)},
+ {XGBE_STATS1_INFO(rx_jabber_frames)},
+ {XGBE_STATS1_INFO(rx_undersized_frames)},
+ {XGBE_STATS1_INFO(rx_fragments)},
+ {XGBE_STATS1_INFO(overrun_type4)},
+ {XGBE_STATS1_INFO(overrun_type5)},
+ {XGBE_STATS1_INFO(rx_bytes)},
+ {XGBE_STATS1_INFO(tx_good_frames)},
+ {XGBE_STATS1_INFO(tx_broadcast_frames)},
+ {XGBE_STATS1_INFO(tx_multicast_frames)},
+ {XGBE_STATS1_INFO(tx_pause_frames)},
+ {XGBE_STATS1_INFO(tx_deferred_frames)},
+ {XGBE_STATS1_INFO(tx_collision_frames)},
+ {XGBE_STATS1_INFO(tx_single_coll_frames)},
+ {XGBE_STATS1_INFO(tx_mult_coll_frames)},
+ {XGBE_STATS1_INFO(tx_excessive_collisions)},
+ {XGBE_STATS1_INFO(tx_late_collisions)},
+ {XGBE_STATS1_INFO(tx_underrun)},
+ {XGBE_STATS1_INFO(tx_carrier_sense_errors)},
+ {XGBE_STATS1_INFO(tx_bytes)},
+ {XGBE_STATS1_INFO(tx_64byte_frames)},
+ {XGBE_STATS1_INFO(tx_65_to_127byte_frames)},
+ {XGBE_STATS1_INFO(tx_128_to_255byte_frames)},
+ {XGBE_STATS1_INFO(tx_256_to_511byte_frames)},
+ {XGBE_STATS1_INFO(tx_512_to_1023byte_frames)},
+ {XGBE_STATS1_INFO(tx_1024byte_frames)},
+ {XGBE_STATS1_INFO(net_bytes)},
+ {XGBE_STATS1_INFO(rx_sof_overruns)},
+ {XGBE_STATS1_INFO(rx_mof_overruns)},
+ {XGBE_STATS1_INFO(rx_dma_overruns)},
+ /* XGBE module 2 */
+ {XGBE_STATS2_INFO(rx_good_frames)},
+ {XGBE_STATS2_INFO(rx_broadcast_frames)},
+ {XGBE_STATS2_INFO(rx_multicast_frames)},
+ {XGBE_STATS2_INFO(rx_pause_frames)},
+ {XGBE_STATS2_INFO(rx_crc_errors)},
+ {XGBE_STATS2_INFO(rx_align_code_errors)},
+ {XGBE_STATS2_INFO(rx_oversized_frames)},
+ {XGBE_STATS2_INFO(rx_jabber_frames)},
+ {XGBE_STATS2_INFO(rx_undersized_frames)},
+ {XGBE_STATS2_INFO(rx_fragments)},
+ {XGBE_STATS2_INFO(overrun_type4)},
+ {XGBE_STATS2_INFO(overrun_type5)},
+ {XGBE_STATS2_INFO(rx_bytes)},
+ {XGBE_STATS2_INFO(tx_good_frames)},
+ {XGBE_STATS2_INFO(tx_broadcast_frames)},
+ {XGBE_STATS2_INFO(tx_multicast_frames)},
+ {XGBE_STATS2_INFO(tx_pause_frames)},
+ {XGBE_STATS2_INFO(tx_deferred_frames)},
+ {XGBE_STATS2_INFO(tx_collision_frames)},
+ {XGBE_STATS2_INFO(tx_single_coll_frames)},
+ {XGBE_STATS2_INFO(tx_mult_coll_frames)},
+ {XGBE_STATS2_INFO(tx_excessive_collisions)},
+ {XGBE_STATS2_INFO(tx_late_collisions)},
+ {XGBE_STATS2_INFO(tx_underrun)},
+ {XGBE_STATS2_INFO(tx_carrier_sense_errors)},
+ {XGBE_STATS2_INFO(tx_bytes)},
+ {XGBE_STATS2_INFO(tx_64byte_frames)},
+ {XGBE_STATS2_INFO(tx_65_to_127byte_frames)},
+ {XGBE_STATS2_INFO(tx_128_to_255byte_frames)},
+ {XGBE_STATS2_INFO(tx_256_to_511byte_frames)},
+ {XGBE_STATS2_INFO(tx_512_to_1023byte_frames)},
+ {XGBE_STATS2_INFO(tx_1024byte_frames)},
+ {XGBE_STATS2_INFO(net_bytes)},
+ {XGBE_STATS2_INFO(rx_sof_overruns)},
+ {XGBE_STATS2_INFO(rx_mof_overruns)},
+ {XGBE_STATS2_INFO(rx_dma_overruns)},
+};
+
#define for_each_intf(i, priv) \
list_for_each_entry((i), &(priv)->gbe_intf_head, gbe_intf_list)
@@ -631,7 +905,10 @@ static void keystone_get_ethtool_stats(struct net_device *ndev,
gbe_dev = gbe_intf->gbe_dev;
spin_lock_bh(&gbe_dev->hw_stats_lock);
- gbe_update_stats_ver14(gbe_dev, data);
+ if (gbe_dev->ss_version == GBE_SS_VERSION_14)
+ gbe_update_stats_ver14(gbe_dev, data);
+ else
+ gbe_update_stats(gbe_dev, data);
spin_unlock_bh(&gbe_dev->hw_stats_lock);
}
@@ -742,8 +1019,13 @@ static void netcp_ethss_link_state_action(struct gbe_priv *gbe_dev,
if (up) {
mac_control = slave->mac_control;
- if (phy && (phy->speed == SPEED_1000))
+ if (phy && (phy->speed == SPEED_1000)) {
mac_control |= MACSL_GIG_MODE;
+ mac_control &= ~MACSL_XGIG_MODE;
+ } else if (phy && (phy->speed == SPEED_10000)) {
+ mac_control |= MACSL_XGIG_MODE;
+ mac_control &= ~MACSL_GIG_MODE;
+ }
writel(mac_control, GBE_REG_ADDR(slave, emac_regs,
mac_control));
@@ -783,7 +1065,9 @@ static void netcp_ethss_update_link_state(struct gbe_priv *gbe_dev,
if (!slave->open)
return;
- sgmii_link_state = netcp_sgmii_get_port_link(SGMII_BASE(sp), sp);
+ if (!SLAVE_LINK_IS_XGMII(slave))
+ sgmii_link_state = netcp_sgmii_get_port_link(SGMII_BASE(sp),
+ sp);
phy_link_state = gbe_phy_link_status(slave);
link_state = phy_link_state & sgmii_link_state;
@@ -792,6 +1076,19 @@ static void netcp_ethss_update_link_state(struct gbe_priv *gbe_dev,
link_state);
}
+static void xgbe_adjust_link(struct net_device *ndev)
+{
+ struct netcp_intf *netcp = netdev_priv(ndev);
+ struct gbe_intf *gbe_intf;
+
+ gbe_intf = netcp_module_get_intf_data(&xgbe_module, netcp);
+ if (!gbe_intf)
+ return;
+
+ netcp_ethss_update_link_state(gbe_intf->gbe_dev, gbe_intf->slave,
+ ndev);
+}
+
static void gbe_adjust_link(struct net_device *ndev)
{
struct netcp_intf *netcp = netdev_priv(ndev);
@@ -839,9 +1136,19 @@ static int gbe_port_reset(struct gbe_slave *slave)
static void gbe_port_config(struct gbe_priv *gbe_dev, struct gbe_slave *slave,
int max_rx_len)
{
+ u32 xgmii_mode;
+
if (max_rx_len > NETCP_MAX_FRAME_SIZE)
max_rx_len = NETCP_MAX_FRAME_SIZE;
+ /* Enable correct MII mode at SS level */
+ if ((gbe_dev->ss_version == XGBE_SS_VERSION_10) &&
+ (slave->link_interface >= XGMII_LINK_MAC_PHY)) {
+ xgmii_mode = readl(GBE_REG_ADDR(gbe_dev, ss_regs, control));
+ xgmii_mode |= (1 << slave->slave_num);
+ writel(xgmii_mode, GBE_REG_ADDR(gbe_dev, ss_regs, control));
+ }
+
writel(max_rx_len, GBE_REG_ADDR(slave, emac_regs, rx_maxlen));
writel(slave->mac_control, GBE_REG_ADDR(slave, emac_regs, mac_control));
}
@@ -874,9 +1181,11 @@ static void gbe_sgmii_config(struct gbe_priv *priv, struct gbe_slave *slave)
if ((priv->ss_version == GBE_SS_VERSION_14) && (slave->slave_num >= 2))
sgmii_port_regs = priv->sgmii_port34_regs;
- netcp_sgmii_reset(sgmii_port_regs, slave->slave_num);
- netcp_sgmii_config(sgmii_port_regs, slave->slave_num,
- slave->link_interface);
+ if (!SLAVE_LINK_IS_XGMII(slave)) {
+ netcp_sgmii_reset(sgmii_port_regs, slave->slave_num);
+ netcp_sgmii_config(sgmii_port_regs, slave->slave_num,
+ slave->link_interface);
+ }
}
static int gbe_slave_open(struct gbe_intf *gbe_intf)
@@ -909,6 +1218,9 @@ static int gbe_slave_open(struct gbe_intf *gbe_intf)
}
if (has_phy) {
+ if (priv->ss_version == XGBE_SS_VERSION_10)
+ hndlr = xgbe_adjust_link;
+
slave->phy = of_phy_connect(gbe_intf->ndev,
slave->phy_node,
hndlr, 0,
@@ -1233,7 +1545,10 @@ static int init_slave(struct gbe_priv *gbe_dev, struct gbe_slave *slave,
slave->phy_node = of_parse_phandle(node, "phy-handle", 0);
slave->port_num = gbe_get_slave_port(gbe_dev, slave->slave_num);
- slave->mac_control = GBE_DEF_1G_MAC_CONTROL;
+ if (slave->link_interface >= XGMII_LINK_MAC_PHY)
+ slave->mac_control = GBE_DEF_10G_MAC_CONTROL;
+ else
+ slave->mac_control = GBE_DEF_1G_MAC_CONTROL;
/* Emac regs memmap are contiguous but port regs are not */
port_reg_num = slave->slave_num;
@@ -1244,6 +1559,8 @@ static int init_slave(struct gbe_priv *gbe_dev, struct gbe_slave *slave,
} else {
port_reg_ofs = GBE13_SLAVE_PORT_OFFSET;
}
+ } else if (gbe_dev->ss_version == XGBE_SS_VERSION_10) {
+ port_reg_ofs = XGBE10_SLAVE_PORT_OFFSET;
} else {
dev_err(gbe_dev->dev, "unknown ethss(0x%x)\n",
gbe_dev->ss_version);
@@ -1252,6 +1569,8 @@ static int init_slave(struct gbe_priv *gbe_dev, struct gbe_slave *slave,
if (gbe_dev->ss_version == GBE_SS_VERSION_14)
emac_reg_ofs = GBE13_EMAC_OFFSET;
+ else if (gbe_dev->ss_version == XGBE_SS_VERSION_10)
+ emac_reg_ofs = XGBE10_EMAC_OFFSET;
slave->port_regs = gbe_dev->ss_regs + port_reg_ofs +
(0x30 * port_reg_num);
@@ -1275,10 +1594,22 @@ static int init_slave(struct gbe_priv *gbe_dev, struct gbe_slave *slave,
GBE_SET_REG_OFS(slave, emac_regs, soft_reset);
GBE_SET_REG_OFS(slave, emac_regs, rx_maxlen);
- } else {
- dev_err(gbe_dev->dev, "unknown ethss(0x%x)\n",
- gbe_dev->ss_version);
- return -EINVAL;
+ } else if (gbe_dev->ss_version == XGBE_SS_VERSION_10) {
+ /* Initialize slave port register offsets */
+ XGBE_SET_REG_OFS(slave, port_regs, port_vlan);
+ XGBE_SET_REG_OFS(slave, port_regs, tx_pri_map);
+ XGBE_SET_REG_OFS(slave, port_regs, sa_lo);
+ XGBE_SET_REG_OFS(slave, port_regs, sa_hi);
+ XGBE_SET_REG_OFS(slave, port_regs, ts_ctl);
+ XGBE_SET_REG_OFS(slave, port_regs, ts_seq_ltype);
+ XGBE_SET_REG_OFS(slave, port_regs, ts_vlan);
+ XGBE_SET_REG_OFS(slave, port_regs, ts_ctl_ltype2);
+ XGBE_SET_REG_OFS(slave, port_regs, ts_ctl2);
+
+ /* Initialize EMAC register offsets */
+ XGBE_SET_REG_OFS(slave, emac_regs, mac_control);
+ XGBE_SET_REG_OFS(slave, emac_regs, soft_reset);
+ XGBE_SET_REG_OFS(slave, emac_regs, rx_maxlen);
}
atomic_set(&slave->link_state, NETCP_LINK_STATE_INVALID);
@@ -1317,7 +1648,8 @@ static void init_secondary_ports(struct gbe_priv *gbe_dev,
gbe_port_config(gbe_dev, slave, gbe_dev->rx_packet_max);
list_add_tail(&slave->slave_list, &gbe_dev->secondary_slaves);
gbe_dev->num_slaves++;
- if (slave->link_interface == SGMII_LINK_MAC_PHY)
+ if ((slave->link_interface == SGMII_LINK_MAC_PHY) ||
+ (slave->link_interface == XGMII_LINK_MAC_PHY))
mac_phy_link = true;
slave->open = true;
@@ -1347,7 +1679,8 @@ static void init_secondary_ports(struct gbe_priv *gbe_dev,
}
for_each_sec_slave(slave, gbe_dev) {
- if (slave->link_interface != SGMII_LINK_MAC_PHY)
+ if ((slave->link_interface != SGMII_LINK_MAC_PHY) &&
+ (slave->link_interface != XGMII_LINK_MAC_PHY))
continue;
slave->phy =
of_phy_connect(gbe_dev->dummy_ndev,
@@ -1383,6 +1716,85 @@ static void free_secondary_ports(struct gbe_priv *gbe_dev)
free_netdev(gbe_dev->dummy_ndev);
}
+static int set_xgbe_ethss10_priv(struct gbe_priv *gbe_dev,
+ struct device_node *node)
+{
+ struct resource res;
+ void __iomem *regs;
+ int ret, i;
+
+ ret = of_address_to_resource(node, 0, &res);
+ if (ret) {
+ dev_err(gbe_dev->dev, "Can't translate of node(%s) address for xgbe subsystem regs\n",
+ node->name);
+ return ret;
+ }
+
+ regs = devm_ioremap_resource(gbe_dev->dev, &res);
+ if (IS_ERR(regs)) {
+ dev_err(gbe_dev->dev, "Failed to map xgbe register base\n");
+ return PTR_ERR(regs);
+ }
+ gbe_dev->ss_regs = regs;
+
+ ret = of_address_to_resource(node, XGBE_SERDES_REG_INDEX, &res);
+ if (ret) {
+ dev_err(gbe_dev->dev, "Can't translate of node(%s) address for xgbe serdes regs\n",
+ node->name);
+ return ret;
+ }
+
+ regs = devm_ioremap_resource(gbe_dev->dev, &res);
+ if (IS_ERR(regs)) {
+ dev_err(gbe_dev->dev, "Failed to map xgbe serdes register base\n");
+ return PTR_ERR(regs);
+ }
+ gbe_dev->xgbe_serdes_regs = regs;
+
+ gbe_dev->hw_stats = devm_kzalloc(gbe_dev->dev,
+ XGBE10_NUM_STAT_ENTRIES *
+ (XGBE10_NUM_SLAVES + 1) * sizeof(u64),
+ GFP_KERNEL);
+ if (!gbe_dev->hw_stats) {
+ dev_err(gbe_dev->dev, "hw_stats memory allocation failed\n");
+ return -ENOMEM;
+ }
+
+ gbe_dev->ss_version = XGBE_SS_VERSION_10;
+ gbe_dev->sgmii_port_regs = gbe_dev->ss_regs +
+ XGBE10_SGMII_MODULE_OFFSET;
+ gbe_dev->switch_regs = gbe_dev->ss_regs + XGBE10_SWITCH_MODULE_OFFSET;
+ gbe_dev->host_port_regs = gbe_dev->ss_regs + XGBE10_HOST_PORT_OFFSET;
+
+ for (i = 0; i < XGBE10_NUM_HW_STATS_MOD; i++)
+ gbe_dev->hw_stats_regs[i] = gbe_dev->ss_regs +
+ XGBE10_HW_STATS_OFFSET + (GBE_HW_STATS_REG_MAP_SZ * i);
+
+ gbe_dev->ale_reg = gbe_dev->ss_regs + XGBE10_ALE_OFFSET;
+ gbe_dev->ale_ports = XGBE10_NUM_ALE_PORTS;
+ gbe_dev->host_port = XGBE10_HOST_PORT_NUM;
+ gbe_dev->ale_entries = XGBE10_NUM_ALE_ENTRIES;
+ gbe_dev->et_stats = xgbe10_et_stats;
+ gbe_dev->num_et_stats = ARRAY_SIZE(xgbe10_et_stats);
+
+ /* Subsystem registers */
+ XGBE_SET_REG_OFS(gbe_dev, ss_regs, id_ver);
+ XGBE_SET_REG_OFS(gbe_dev, ss_regs, control);
+
+ /* Switch module registers */
+ XGBE_SET_REG_OFS(gbe_dev, switch_regs, id_ver);
+ XGBE_SET_REG_OFS(gbe_dev, switch_regs, control);
+ XGBE_SET_REG_OFS(gbe_dev, switch_regs, ptype);
+ XGBE_SET_REG_OFS(gbe_dev, switch_regs, stat_port_en);
+ XGBE_SET_REG_OFS(gbe_dev, switch_regs, flow_control);
+
+ /* Host port registers */
+ XGBE_SET_REG_OFS(gbe_dev, host_port_regs, port_vlan);
+ XGBE_SET_REG_OFS(gbe_dev, host_port_regs, tx_pri_map);
+ XGBE_SET_REG_OFS(gbe_dev, host_port_regs, rx_maxlen);
+ return 0;
+}
+
static int get_gbe_resource_version(struct gbe_priv *gbe_dev,
struct device_node *node)
{
@@ -1513,6 +1925,14 @@ static int gbe_probe(struct netcp_device *netcp_device, struct device *dev,
ret = set_gbe_ethss14_priv(gbe_dev, node);
if (ret)
goto quit;
+ } else if (!strcmp(node->name, "xgbe")) {
+ ret = set_xgbe_ethss10_priv(gbe_dev, node);
+ if (ret)
+ goto quit;
+ ret = netcp_xgbe_serdes_init(gbe_dev->xgbe_serdes_regs,
+ gbe_dev->ss_regs);
+ if (ret)
+ goto quit;
} else {
dev_err(dev, "unknown GBE node(%s)\n", node->name);
ret = -ENODEV;
@@ -1695,6 +2115,23 @@ static struct netcp_module gbe_module = {
.ioctl = gbe_ioctl,
};
+static struct netcp_module xgbe_module = {
+ .name = XGBE_MODULE_NAME,
+ .owner = THIS_MODULE,
+ .primary = true,
+ .probe = gbe_probe,
+ .open = gbe_open,
+ .close = gbe_close,
+ .remove = gbe_remove,
+ .attach = gbe_attach,
+ .release = gbe_release,
+ .add_addr = gbe_add_addr,
+ .del_addr = gbe_del_addr,
+ .add_vid = gbe_add_vid,
+ .del_vid = gbe_del_vid,
+ .ioctl = gbe_ioctl,
+};
+
static int __init keystone_gbe_init(void)
{
int ret;
@@ -1703,6 +2140,10 @@ static int __init keystone_gbe_init(void)
if (ret)
return ret;
+ ret = netcp_register_module(&xgbe_module);
+ if (ret)
+ return ret;
+
return 0;
}
module_init(keystone_gbe_init);
@@ -1710,5 +2151,6 @@ module_init(keystone_gbe_init);
static void __exit keystone_gbe_exit(void)
{
netcp_unregister_module(&gbe_module);
+ netcp_unregister_module(&xgbe_module);
}
module_exit(keystone_gbe_exit);
diff --git a/drivers/net/ethernet/ti/netcp_xgbepcsr.c b/drivers/net/ethernet/ti/netcp_xgbepcsr.c
new file mode 100644
index 0000000..33571ac
--- /dev/null
+++ b/drivers/net/ethernet/ti/netcp_xgbepcsr.c
@@ -0,0 +1,501 @@
+/*
+ * XGE PCSR module initialisation
+ *
+ * Copyright (C) 2014 Texas Instruments Incorporated
+ * Authors: Sandeep Nair <sandeep_n@ti.com>
+ * WingMan Kwok <w-kwok2@ti.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 version 2.
+ *
+ * This program is distributed "as is" WITHOUT ANY WARRANTY of any
+ * kind, whether express or implied; without even the implied warranty
+ * of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ */
+#include "netcp.h"
+
+/* XGBE registers */
+#define XGBE_CTRL_OFFSET 0x0c
+#define XGBE_SGMII_1_OFFSET 0x0114
+#define XGBE_SGMII_2_OFFSET 0x0214
+
+/* PCS-R registers */
+#define PCSR_CPU_CTRL_OFFSET 0x1fd0
+#define POR_EN BIT(29)
+
+#define reg_rmw(addr, value, mask) \
+ writel(((readl(addr) & (~(mask))) | \
+ (value & (mask))), (addr))
+
+/* bit mask of width w at offset s */
+#define MASK_WID_SH(w, s) (((1 << w) - 1) << s)
+
+/* shift value v to offset s */
+#define VAL_SH(v, s) (v << s)
+
+#define PHY_A(serdes) 0
+
+struct serdes_cfg {
+ u32 ofs;
+ u32 val;
+ u32 mask;
+};
+
+static struct serdes_cfg cfg_phyb_1p25g_156p25mhz_cmu0[] = {
+ {0x0000, 0x00800002, 0x00ff00ff},
+ {0x0014, 0x00003838, 0x0000ffff},
+ {0x0060, 0x1c44e438, 0xffffffff},
+ {0x0064, 0x00c18400, 0x00ffffff},
+ {0x0068, 0x17078200, 0xffffff00},
+ {0x006c, 0x00000014, 0x000000ff},
+ {0x0078, 0x0000c000, 0x0000ff00},
+ {0x0000, 0x00000003, 0x000000ff},
+};
+
+static struct serdes_cfg cfg_phyb_10p3125g_156p25mhz_cmu1[] = {
+ {0x0c00, 0x00030002, 0x00ff00ff},
+ {0x0c14, 0x00005252, 0x0000ffff},
+ {0x0c28, 0x80000000, 0xff000000},
+ {0x0c2c, 0x000000f6, 0x000000ff},
+ {0x0c3c, 0x04000405, 0xff00ffff},
+ {0x0c40, 0xc0800000, 0xffff0000},
+ {0x0c44, 0x5a202062, 0xffffffff},
+ {0x0c48, 0x40040424, 0xffffffff},
+ {0x0c4c, 0x00004002, 0x0000ffff},
+ {0x0c50, 0x19001c00, 0xff00ff00},
+ {0x0c54, 0x00002100, 0x0000ff00},
+ {0x0c58, 0x00000060, 0x000000ff},
+ {0x0c60, 0x80131e7c, 0xffffffff},
+ {0x0c64, 0x8400cb02, 0xff00ffff},
+ {0x0c68, 0x17078200, 0xffffff00},
+ {0x0c6c, 0x00000016, 0x000000ff},
+ {0x0c74, 0x00000400, 0x0000ff00},
+ {0x0c78, 0x0000c000, 0x0000ff00},
+ {0x0c00, 0x00000003, 0x000000ff},
+};
+
+static struct serdes_cfg cfg_phyb_10p3125g_16bit_lane[] = {
+ {0x0204, 0x00000080, 0x000000ff},
+ {0x0208, 0x0000920d, 0x0000ffff},
+ {0x0204, 0xfc000000, 0xff000000},
+ {0x0208, 0x00009104, 0x0000ffff},
+ {0x0210, 0x1a000000, 0xff000000},
+ {0x0214, 0x00006b58, 0x00ffffff},
+ {0x0218, 0x75800084, 0xffff00ff},
+ {0x022c, 0x00300000, 0x00ff0000},
+ {0x0230, 0x00003800, 0x0000ff00},
+ {0x024c, 0x008f0000, 0x00ff0000},
+ {0x0250, 0x30000000, 0xff000000},
+ {0x0260, 0x00000002, 0x000000ff},
+ {0x0264, 0x00000057, 0x000000ff},
+ {0x0268, 0x00575700, 0x00ffff00},
+ {0x0278, 0xff000000, 0xff000000},
+ {0x0280, 0x00500050, 0x00ff00ff},
+ {0x0284, 0x00001f15, 0x0000ffff},
+ {0x028c, 0x00006f00, 0x0000ff00},
+ {0x0294, 0x00000000, 0xffffff00},
+ {0x0298, 0x00002640, 0xff00ffff},
+ {0x029c, 0x00000003, 0x000000ff},
+ {0x02a4, 0x00000f13, 0x0000ffff},
+ {0x02a8, 0x0001b600, 0x00ffff00},
+ {0x0380, 0x00000030, 0x000000ff},
+ {0x03c0, 0x00000200, 0x0000ff00},
+ {0x03cc, 0x00000018, 0x000000ff},
+ {0x03cc, 0x00000000, 0x000000ff},
+};
+
+static struct serdes_cfg cfg_phyb_10p3125g_comlane[] = {
+ {0x0a00, 0x00000800, 0x0000ff00},
+ {0x0a84, 0x00000000, 0x000000ff},
+ {0x0a8c, 0x00130000, 0x00ff0000},
+ {0x0a90, 0x77a00000, 0xffff0000},
+ {0x0a94, 0x00007777, 0x0000ffff},
+ {0x0b08, 0x000f0000, 0xffff0000},
+ {0x0b0c, 0x000f0000, 0x00ffffff},
+ {0x0b10, 0xbe000000, 0xff000000},
+ {0x0b14, 0x000000ff, 0x000000ff},
+ {0x0b18, 0x00000014, 0x000000ff},
+ {0x0b5c, 0x981b0000, 0xffff0000},
+ {0x0b64, 0x00001100, 0x0000ff00},
+ {0x0b78, 0x00000c00, 0x0000ff00},
+ {0x0abc, 0xff000000, 0xff000000},
+ {0x0ac0, 0x0000008b, 0x000000ff},
+};
+
+static struct serdes_cfg cfg_cm_c1_c2[] = {
+ {0x0208, 0x00000000, 0x00000f00},
+ {0x0208, 0x00000000, 0x0000001f},
+ {0x0204, 0x00000000, 0x00040000},
+ {0x0208, 0x000000a0, 0x000000e0},
+};
+
+static void netcp_xgbe_serdes_cmu_init(void __iomem *serdes_regs)
+{
+ int i;
+
+ /* cmu0 setup */
+ for (i = 0; i < ARRAY_SIZE(cfg_phyb_1p25g_156p25mhz_cmu0); i++) {
+ reg_rmw(serdes_regs + cfg_phyb_1p25g_156p25mhz_cmu0[i].ofs,
+ cfg_phyb_1p25g_156p25mhz_cmu0[i].val,
+ cfg_phyb_1p25g_156p25mhz_cmu0[i].mask);
+ }
+
+ /* cmu1 setup */
+ for (i = 0; i < ARRAY_SIZE(cfg_phyb_10p3125g_156p25mhz_cmu1); i++) {
+ reg_rmw(serdes_regs + cfg_phyb_10p3125g_156p25mhz_cmu1[i].ofs,
+ cfg_phyb_10p3125g_156p25mhz_cmu1[i].val,
+ cfg_phyb_10p3125g_156p25mhz_cmu1[i].mask);
+ }
+}
+
+/* lane is 0 based */
+static void netcp_xgbe_serdes_lane_config(
+ void __iomem *serdes_regs, int lane)
+{
+ int i;
+
+ /* lane setup */
+ for (i = 0; i < ARRAY_SIZE(cfg_phyb_10p3125g_16bit_lane); i++) {
+ reg_rmw(serdes_regs +
+ cfg_phyb_10p3125g_16bit_lane[i].ofs +
+ (0x200 * lane),
+ cfg_phyb_10p3125g_16bit_lane[i].val,
+ cfg_phyb_10p3125g_16bit_lane[i].mask);
+ }
+
+ /* disable auto negotiation*/
+ reg_rmw(serdes_regs + (0x200 * lane) + 0x0380,
+ 0x00000000, 0x00000010);
+
+ /* disable link training */
+ reg_rmw(serdes_regs + (0x200 * lane) + 0x03c0,
+ 0x00000000, 0x00000200);
+}
+
+static void netcp_xgbe_serdes_com_enable(void __iomem *serdes_regs)
+{
+ int i;
+
+ for (i = 0; i < ARRAY_SIZE(cfg_phyb_10p3125g_comlane); i++) {
+ reg_rmw(serdes_regs + cfg_phyb_10p3125g_comlane[i].ofs,
+ cfg_phyb_10p3125g_comlane[i].val,
+ cfg_phyb_10p3125g_comlane[i].mask);
+ }
+}
+
+static void netcp_xgbe_serdes_lane_enable(
+ void __iomem *serdes_regs, int lane)
+{
+ /* Set Lane Control Rate */
+ writel(0xe0e9e038, serdes_regs + 0x1fe0 + (4 * lane));
+}
+
+static void netcp_xgbe_serdes_phyb_rst_clr(void __iomem *serdes_regs)
+{
+ reg_rmw(serdes_regs + 0x0a00, 0x0000001f, 0x000000ff);
+}
+
+static void netcp_xgbe_serdes_pll_disable(void __iomem *serdes_regs)
+{
+ writel(0x88000000, serdes_regs + 0x1ff4);
+}
+
+static void netcp_xgbe_serdes_pll_enable(void __iomem *serdes_regs)
+{
+ netcp_xgbe_serdes_phyb_rst_clr(serdes_regs);
+ writel(0xee000000, serdes_regs + 0x1ff4);
+}
+
+static int netcp_xgbe_wait_pll_locked(void __iomem *sw_regs)
+{
+ unsigned long timeout;
+ int ret = 0;
+ u32 val_1, val_0;
+
+ timeout = jiffies + msecs_to_jiffies(500);
+ do {
+ val_0 = (readl(sw_regs + XGBE_SGMII_1_OFFSET) & BIT(4));
+ val_1 = (readl(sw_regs + XGBE_SGMII_2_OFFSET) & BIT(4));
+
+ if (val_1 && val_0)
+ return 0;
+
+ if (time_after(jiffies, timeout)) {
+ ret = -ETIMEDOUT;
+ break;
+ }
+
+ cpu_relax();
+ } while (true);
+
+ pr_err("XGBE serdes not locked: time out.\n");
+ return ret;
+}
+
+static void netcp_xgbe_serdes_enable_xgmii_port(void __iomem *sw_regs)
+{
+ writel(0x03, sw_regs + XGBE_CTRL_OFFSET);
+}
+
+static u32 netcp_xgbe_serdes_read_tbus_val(void __iomem *serdes_regs)
+{
+ u32 tmp;
+
+ if (PHY_A(serdes_regs)) {
+ tmp = (readl(serdes_regs + 0x0ec) >> 24) & 0x0ff;
+ tmp |= ((readl(serdes_regs + 0x0fc) >> 16) & 0x00f00);
+ } else {
+ tmp = (readl(serdes_regs + 0x0f8) >> 16) & 0x0fff;
+ }
+
+ return tmp;
+}
+
+static void netcp_xgbe_serdes_write_tbus_addr(void __iomem *serdes_regs,
+ int select, int ofs)
+{
+ if (PHY_A(serdes_regs)) {
+ reg_rmw(serdes_regs + 0x0008, ((select << 5) + ofs) << 24,
+ ~0x00ffffff);
+ return;
+ }
+
+ /* For 2 lane Phy-B, lane0 is actually lane1 */
+ switch (select) {
+ case 1:
+ select = 2;
+ break;
+ case 2:
+ select = 3;
+ break;
+ default:
+ return;
+ }
+
+ reg_rmw(serdes_regs + 0x00fc, ((select << 8) + ofs) << 16, ~0xf800ffff);
+}
+
+static u32 netcp_xgbe_serdes_read_select_tbus(void __iomem *serdes_regs,
+ int select, int ofs)
+{
+ /* Set tbus address */
+ netcp_xgbe_serdes_write_tbus_addr(serdes_regs, select, ofs);
+ /* Get TBUS Value */
+ return netcp_xgbe_serdes_read_tbus_val(serdes_regs);
+}
+
+static void netcp_xgbe_serdes_reset_cdr(void __iomem *serdes_regs,
+ void __iomem *sig_detect_reg, int lane)
+{
+ u32 tmp, dlpf, tbus;
+
+ /*Get the DLPF values */
+ tmp = netcp_xgbe_serdes_read_select_tbus(
+ serdes_regs, lane + 1, 5);
+
+ dlpf = tmp >> 2;
+
+ if (dlpf < 400 || dlpf > 700) {
+ reg_rmw(sig_detect_reg, VAL_SH(2, 1), MASK_WID_SH(2, 1));
+ mdelay(1);
+ reg_rmw(sig_detect_reg, VAL_SH(0, 1), MASK_WID_SH(2, 1));
+ } else {
+ tbus = netcp_xgbe_serdes_read_select_tbus(serdes_regs, lane +
+ 1, 0xe);
+
+ pr_debug("XGBE: CDR centered, DLPF: %4d,%d,%d.\n",
+ tmp >> 2, tmp & 3, (tbus >> 2) & 3);
+ }
+}
+
+/* Call every 100 ms */
+static int netcp_xgbe_check_link_status(void __iomem *serdes_regs,
+ void __iomem *sw_regs, u32 lanes,
+ u32 *current_state, u32 *lane_down)
+{
+ void __iomem *pcsr_base = sw_regs + 0x0600;
+ void __iomem *sig_detect_reg;
+ u32 pcsr_rx_stat, blk_lock, blk_errs;
+ int loss, i, status = 1;
+
+ for (i = 0; i < lanes; i++) {
+ /* Get the Loss bit */
+ loss = readl(serdes_regs + 0x1fc0 + 0x20 + (i * 0x04)) & 0x1;
+
+ /* Get Block Errors and Block Lock bits */
+ pcsr_rx_stat = readl(pcsr_base + 0x0c + (i * 0x80));
+ blk_lock = (pcsr_rx_stat >> 30) & 0x1;
+ blk_errs = (pcsr_rx_stat >> 16) & 0x0ff;
+
+ /* Get Signal Detect Overlay Address */
+ sig_detect_reg = serdes_regs + (i * 0x200) + 0x200 + 0x04;
+
+ /* If Block errors maxed out, attempt recovery! */
+ if (blk_errs == 0x0ff)
+ blk_lock = 0;
+
+ switch (current_state[i]) {
+ case 0:
+ /* if good link lock the signal detect ON! */
+ if (!loss && blk_lock) {
+ pr_debug("XGBE PCSR Linked Lane: %d\n", i);
+ reg_rmw(sig_detect_reg, VAL_SH(3, 1),
+ MASK_WID_SH(2, 1));
+ current_state[i] = 1;
+ } else if (!blk_lock) {
+ /* if no lock, then reset CDR */
+ pr_debug("XGBE PCSR Recover Lane: %d\n", i);
+ netcp_xgbe_serdes_reset_cdr(serdes_regs,
+ sig_detect_reg, i);
+ }
+ break;
+
+ case 1:
+ if (!blk_lock) {
+ /* Link Lost? */
+ lane_down[i] = 1;
+ current_state[i] = 2;
+ }
+ break;
+
+ case 2:
+ if (blk_lock)
+ /* Nope just noise */
+ current_state[i] = 1;
+ else {
+ /* Lost the block lock, reset CDR if it is
+ * not centered and go back to sync state
+ */
+ netcp_xgbe_serdes_reset_cdr(serdes_regs,
+ sig_detect_reg, i);
+ current_state[i] = 0;
+ }
+ break;
+
+ default:
+ pr_err("XGBE: unknown current_state[%d] %d\n",
+ i, current_state[i]);
+ break;
+ }
+
+ if (blk_errs > 0) {
+ /* Reset the Error counts! */
+ reg_rmw(pcsr_base + 0x08 + (i * 0x80), VAL_SH(0x19, 0),
+ MASK_WID_SH(8, 0));
+
+ reg_rmw(pcsr_base + 0x08 + (i * 0x80), VAL_SH(0x00, 0),
+ MASK_WID_SH(8, 0));
+ }
+
+ status &= (current_state[i] == 1);
+ }
+
+ return status;
+}
+
+static int netcp_xgbe_serdes_check_lane(void __iomem *serdes_regs,
+ void __iomem *sw_regs)
+{
+ u32 current_state[2] = {0, 0};
+ int retries = 0, link_up;
+ u32 lane_down[2];
+
+ do {
+ lane_down[0] = 0;
+ lane_down[1] = 0;
+
+ link_up = netcp_xgbe_check_link_status(serdes_regs, sw_regs, 2,
+ current_state,
+ lane_down);
+
+ /* if we did not get link up then wait 100ms before calling
+ * it again
+ */
+ if (link_up)
+ break;
+
+ if (lane_down[0])
+ pr_debug("XGBE: detected link down on lane 0\n");
+
+ if (lane_down[1])
+ pr_debug("XGBE: detected link down on lane 1\n");
+
+ if (++retries > 1) {
+ pr_debug("XGBE: timeout waiting for serdes link up\n");
+ return -ETIMEDOUT;
+ }
+ mdelay(100);
+ } while (!link_up);
+
+ pr_debug("XGBE: PCSR link is up\n");
+ return 0;
+}
+
+static void netcp_xgbe_serdes_setup_cm_c1_c2(void __iomem *serdes_regs,
+ int lane, int cm, int c1, int c2)
+{
+ int i;
+
+ for (i = 0; i < ARRAY_SIZE(cfg_cm_c1_c2); i++) {
+ reg_rmw(serdes_regs + cfg_cm_c1_c2[i].ofs + (0x200 * lane),
+ cfg_cm_c1_c2[i].val,
+ cfg_cm_c1_c2[i].mask);
+ }
+}
+
+static void netcp_xgbe_reset_serdes(void __iomem *serdes_regs)
+{
+ /* Toggle the POR_EN bit in CONFIG.CPU_CTRL */
+ /* enable POR_EN bit */
+ reg_rmw(serdes_regs + PCSR_CPU_CTRL_OFFSET, POR_EN, POR_EN);
+ usleep_range(10, 100);
+
+ /* disable POR_EN bit */
+ reg_rmw(serdes_regs + PCSR_CPU_CTRL_OFFSET, 0, POR_EN);
+ usleep_range(10, 100);
+}
+
+static int netcp_xgbe_serdes_config(void __iomem *serdes_regs,
+ void __iomem *sw_regs)
+{
+ u32 ret, i;
+
+ netcp_xgbe_serdes_pll_disable(serdes_regs);
+ netcp_xgbe_serdes_cmu_init(serdes_regs);
+
+ for (i = 0; i < 2; i++)
+ netcp_xgbe_serdes_lane_config(serdes_regs, i);
+
+ netcp_xgbe_serdes_com_enable(serdes_regs);
+ /* This is EVM + RTM-BOC specific */
+ for (i = 0; i < 2; i++)
+ netcp_xgbe_serdes_setup_cm_c1_c2(serdes_regs, i, 0, 0, 5);
+
+ netcp_xgbe_serdes_pll_enable(serdes_regs);
+ for (i = 0; i < 2; i++)
+ netcp_xgbe_serdes_lane_enable(serdes_regs, i);
+
+ /* SB PLL Status Poll */
+ ret = netcp_xgbe_wait_pll_locked(sw_regs);
+ if (ret)
+ return ret;
+
+ netcp_xgbe_serdes_enable_xgmii_port(sw_regs);
+ netcp_xgbe_serdes_check_lane(serdes_regs, sw_regs);
+ return ret;
+}
+
+int netcp_xgbe_serdes_init(void __iomem *serdes_regs, void __iomem *xgbe_regs)
+{
+ u32 val;
+
+ /* read COMLANE bits 4:0 */
+ val = readl(serdes_regs + 0xa00);
+ if (val & 0x1f) {
+ pr_debug("XGBE: serdes already in operation - reset\n");
+ netcp_xgbe_reset_serdes(serdes_regs);
+ }
+ return netcp_xgbe_serdes_config(serdes_regs, xgbe_regs);
+}
--
1.7.9.5
^ permalink raw reply related
* Re: [PATCH] net: sxgbe: Fix NULL dereferece when using DT
From: David Miller @ 2015-01-16 0:14 UTC (permalink / raw)
To: kgene; +Cc: netdev, davem, ks.giri, bh74.an
In-Reply-To: <1421286107-9514-1-git-send-email-kgene@kernel.org>
From: Kukjin Kim <kgene@kernel.org>
Date: Thu, 15 Jan 2015 10:41:47 +0900
> From: Girish K.S <ks.giri@samsung.com>
>
> When the MAC address is provided in the device tree file, the
> condition is true and kernel crashes due to NULL dereference.
>
> Signed-off-by: Girish K.S <ks.giri@samsung.com>
> Signed-off-by: Byungho An <bh74.an@samsung.com>
> Signed-off-by: Kukjin Kim <kgene@kernel.org>
Applied.
^ permalink raw reply
* Re: [PATCH] [PATCH] net: sxgbe: Fix waring for double kfree()
From: David Miller @ 2015-01-16 0:14 UTC (permalink / raw)
To: kgene; +Cc: netdev, davem, dan.carpenter, bh74.an
In-Reply-To: <1421286191-9550-1-git-send-email-kgene@kernel.org>
From: Kukjin Kim <kgene@kernel.org>
Date: Thu, 15 Jan 2015 10:43:11 +0900
> From: Byungho An <bh74.an@samsung.com>
>
> This patch fixes double kfree() calls at init_rx_ring() because
> it causes static checker warning.
>
> Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
> Signed-off-by: Byungho An <bh74.an@samsung.com>
> Signed-off-by: Kukjin Kim <kgene@kernel.org>
Applied.
^ permalink raw reply
* Re: [PATCH net] net: rps: fix cpu unplug
From: Eric Dumazet @ 2015-01-16 0:14 UTC (permalink / raw)
To: subashab; +Cc: Prasad Sodagudi, netdev, Tom Herbert
In-Reply-To: <9d793ea2f58c16ec1ceca0124eed4d67.squirrel@www.codeaurora.org>
On Thu, 2015-01-15 at 22:29 +0000, subashab@codeaurora.org wrote:
> Thanks for the patch. I shall try it out and provide feedback soon.
> But we think the race condition issue is different. The crash was observed
> in the process_queue.
>
> On the event of a CPU hotplug, the NAPI poll_list is copied over from the
> offline CPU to the CPU on which dev_cpu_callback() was called. These
> operations happens in dev_cpu_callback() in the context of the notifier
> chain from hotplug framework. Also in the same hotplug notifier context
> (dev_cpu_callback) the input_pkt_queue and process_queue of the offline
> CPU are dequeued and sent up the network stack and this is where I think
> the race/problem is.
>
> Context1: The online CPU starts processing the poll_list from
> net_rx_action since a
> softIRQ was raised in dev_cpu_callback(). process_backlog() draining the
> process queue
>
> Context2: hotplug notifier dev_cpu_callback() draining the queues and
> calling netif_rx().
>
> from dev_cpu_callback()
> /* Process offline CPU's input_pkt_queue */
> while ((skb = __skb_dequeue(&oldsd->process_queue))) {
> netif_rx(skb);
> input_queue_head_incr(oldsd);
> }
> while ((skb = __skb_dequeue(&oldsd->input_pkt_queue))) {
> netif_rx(skb);
> input_queue_head_incr(oldsd);
> }
>
> Is this de-queuing(the above code snippet from dev_cpu_callback())
> actually necessary since the poll_list should already have the backlog
> napi struct of the old CPU? In this case when process_backlog()
> actually runs it should drain these two queues of older CPU.
> Let me know your thoughts.
input_pkt_queue and process_queue have nothing to do with NAPI
poll_list : They store skbs.
dev_cpu_callback() is called when the cpu we are offlining is no longer
running. No interrupts either serviced by this offline cpu.
You have the absolute guarantee No one is manipulating process_queue at
the same time than you.
It looks like you found another issue, not related to RPS, but due to
the fact that commit 264524d5e5195f6e
("net: cpu offline cause napi stall") did not exclude the percpu
backlog.
process_backlog() MUST be called by the owner cpu. Otherwise we would
need to add locking everywhere, as you did, and this is simply insane.
I'll send a V2
^ permalink raw reply
* Re: pull-request: mac80211-next 2015-01-15
From: David Miller @ 2015-01-16 0:28 UTC (permalink / raw)
To: johannes; +Cc: netdev, linux-wireless
In-Reply-To: <1421325902.1962.8.camel@sipsolutions.net>
From: Johannes Berg <johannes@sipsolutions.net>
Date: Thu, 15 Jan 2015 13:45:02 +0100
> My first pull request for -next, sorry I dragged my feet on this.
>
> I merged mac80211.git to avoid some conflicts, but the content was in
> net-next already anyway.
>
> Let me know if there's any issue or any format changes you'd like to make.
I had to resolve a minor merge conflict, please take a look to make sure
I got it right.
Pulled, thanks.
^ permalink raw reply
* Re: pull-request: mac80211 2015-01-15
From: David Miller @ 2015-01-16 0:29 UTC (permalink / raw)
To: johannes-cdvu00un1VgdHxzADdlk8Q
Cc: netdev-u79uwXL29TY76Z2rM5mHXA,
linux-wireless-u79uwXL29TY76Z2rM5mHXA
In-Reply-To: <1421326080.1962.10.camel-cdvu00un1VgdHxzADdlk8Q@public.gmane.org>
From: Johannes Berg <johannes-cdvu00un1VgdHxzADdlk8Q@public.gmane.org>
Date: Thu, 15 Jan 2015 13:48:00 +0100
> Hope it's not too late for 3.19 - we found two more issues, pull request
> below.
Pulled, thanks Johannes.
--
To unsubscribe from this list: send the line "unsubscribe linux-wireless" 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: smc91x: Add Atari EtherNAT support
From: David Miller @ 2015-01-16 0:33 UTC (permalink / raw)
To: geert; +Cc: nico, netdev, linux-m68k, schmitzmic, schmitz
In-Reply-To: <1421327175-7613-1-git-send-email-geert@linux-m68k.org>
From: Geert Uytterhoeven <geert@linux-m68k.org>
Date: Thu, 15 Jan 2015 14:06:15 +0100
> From: Michael Schmitz <schmitzmic@gmail.com>
>
> Add Atari specific code to the smc91x Ethernet driver. This code is used
> on the EtherNAT adapter card for the Atari Falcon extension port.
>
> Signed-off-by: Michael Schmitz <schmitz@debian.org>
> Tested-by: Christian Steigies <cts@debian.org>
> [geert: Sort Kconfig entries, split in hard and soft dependencies]
> Signed-off-by: Geert Uytterhoeven <geert@linux-m68k.org>
Applied to net-next, thanks.
^ permalink raw reply
* Re: [PATCH net] net/mlx4: Don't disable vxlan offloads under DMFS-A0 optimized steering
From: David Miller @ 2015-01-16 0:35 UTC (permalink / raw)
To: ogerlitz; +Cc: netdev, matanb, amirv
In-Reply-To: <1421328534-14216-1-git-send-email-ogerlitz@mellanox.com>
From: Or Gerlitz <ogerlitz@mellanox.com>
Date: Thu, 15 Jan 2015 15:28:54 +0200
> Except for VXLAN steering rules, all offloads should work as they were
> under plain DMFS mode. Fix that by enabling all the offloads under
> DMFS-A0 mode, except for VXLAN steering rules.
>
> Fixes: d57febe1a478 "net/mlx4: Add A0 hybrid steering"
> Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com>
Applied, thanks.
^ permalink raw reply
* Re: [PATCH net-next] bridge: use MDBA_SET_ENTRY_MAX for maxtype in nlmsg_parse()
From: David Miller @ 2015-01-16 0:37 UTC (permalink / raw)
To: nicolas.dichtel; +Cc: netdev
In-Reply-To: <1421335752-25204-1-git-send-email-nicolas.dichtel@6wind.com>
From: Nicolas Dichtel <nicolas.dichtel@6wind.com>
Date: Thu, 15 Jan 2015 16:29:12 +0100
> This is just a cleanup, because in the current code MDBA_SET_ENTRY_MAX ==
> MDBA_SET_ENTRY.
>
> Signed-off-by: Nicolas Dichtel <nicolas.dichtel@6wind.com>
Applied, thanks.
^ permalink raw reply
* Re: pull-request: can 2015-01-15
From: David Miller @ 2015-01-16 0:39 UTC (permalink / raw)
To: mkl; +Cc: netdev, linux-can, kernel
In-Reply-To: <1421338283-12417-1-git-send-email-mkl@pengutronix.de>
From: Marc Kleine-Budde <mkl@pengutronix.de>
Date: Thu, 15 Jan 2015 17:11:15 +0100
> this is a pull request of 8 patches.
>
> Ahmed S. Darwish contributes 4 fixes for the kvaser_usb driver. The two patches
> by Oliver Hartkopp mark the m_can driver as non-ISO, as the CANFD standard was
> updated. Roger Quadros's patch for the c_can driver fixes the register access
> during RAMINIT. And one patch by my, which updates the MAINTAINERS file, as we
> moved the git repos to the kernel.org infrastructure.
Pulled, thank you Marc.
^ permalink raw reply
* [PATCH v2 net] net: rps: fix cpu unplug
From: Eric Dumazet @ 2015-01-16 1:04 UTC (permalink / raw)
To: subashab; +Cc: Prasad Sodagudi, netdev, Tom Herbert
In-Reply-To: <1421358383.11734.97.camel@edumazet-glaptop2.roam.corp.google.com>
From: Eric Dumazet <edumazet@google.com>
softnet_data.input_pkt_queue is protected by a spinlock that
we must hold when transferring packets from victim queue to an active
one. This is because other cpus could still be trying to enqueue packets
into victim queue.
A second problem is that when we transfert the NAPI poll_list from
victim to current cpu, we absolutely need to special case the percpu
backlog, because we do not want to add complex locking to protect
process_queue : Only owner cpu is allowed to manipulate it, unless cpu
is offline.
Based on initial patch from Prasad Sodagudi & Subash Abhinov
Kasiviswanathan.
This version is better because we do not slow down packet processing,
only make migration safer.
Reported-by: Prasad Sodagudi <psodagud@codeaurora.org>
Reported-by: Subash Abhinov Kasiviswanathan <subashab@codeaurora.org>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Tom Herbert <therbert@google.com>
---
net/core/dev.c | 20 +++++++++++++++-----
1 file changed, 15 insertions(+), 5 deletions(-)
diff --git a/net/core/dev.c b/net/core/dev.c
index 1e325adc4367..7f028d441e98 100644
--- a/net/core/dev.c
+++ b/net/core/dev.c
@@ -7075,10 +7075,20 @@ static int dev_cpu_callback(struct notifier_block *nfb,
oldsd->output_queue = NULL;
oldsd->output_queue_tailp = &oldsd->output_queue;
}
- /* Append NAPI poll list from offline CPU. */
- if (!list_empty(&oldsd->poll_list)) {
- list_splice_init(&oldsd->poll_list, &sd->poll_list);
- raise_softirq_irqoff(NET_RX_SOFTIRQ);
+ /* Append NAPI poll list from offline CPU, with one exception :
+ * process_backlog() must be called by cpu owning percpu backlog.
+ * We properly handle process_queue & input_pkt_queue later.
+ */
+ while (!list_empty(&oldsd->poll_list)) {
+ struct napi_struct *napi = list_first_entry(&oldsd->poll_list,
+ struct napi_struct,
+ poll_list);
+
+ list_del_init(&napi->poll_list);
+ if (napi->poll == process_backlog)
+ napi->state = 0;
+ else
+ ____napi_schedule(sd, napi);
}
raise_softirq_irqoff(NET_TX_SOFTIRQ);
@@ -7089,7 +7099,7 @@ static int dev_cpu_callback(struct notifier_block *nfb,
netif_rx_internal(skb);
input_queue_head_incr(oldsd);
}
- while ((skb = __skb_dequeue(&oldsd->input_pkt_queue))) {
+ while ((skb = skb_dequeue(&oldsd->input_pkt_queue))) {
netif_rx_internal(skb);
input_queue_head_incr(oldsd);
}
^ permalink raw reply related
* Re: [PATCH v2 net] net: rps: fix cpu unplug
From: Eric Dumazet @ 2015-01-16 1:21 UTC (permalink / raw)
To: subashab; +Cc: Prasad Sodagudi, netdev, Tom Herbert
In-Reply-To: <1421370262.11734.111.camel@edumazet-glaptop2.roam.corp.google.com>
On Thu, 2015-01-15 at 17:04 -0800, Eric Dumazet wrote:
> From: Eric Dumazet <edumazet@google.com>
>
Small note : this patch was based on net-next but applies correctly to
net tree
Hunk #1 succeeded at 7072 (offset -3 lines).
Hunk #2 succeeded at 7096 (offset -3 lines).
^ permalink raw reply
* Re: [patch-net-next v2 3/3] net: ethernet: cpsw: don't requests IRQs we don't use
From: Felipe Balbi @ 2015-01-16 1:28 UTC (permalink / raw)
To: David Miller; +Cc: balbi, mugunthanvnm, tony, linux-omap, netdev
In-Reply-To: <20150115.181615.498992970722065060.davem@davemloft.net>
[-- Attachment #1: Type: text/plain, Size: 1184 bytes --]
On Thu, Jan 15, 2015 at 06:16:15PM -0500, David Miller wrote:
> From: Felipe Balbi <balbi@ti.com>
> Date: Thu, 15 Jan 2015 09:20:53 -0600
>
> > On Thu, Jan 15, 2015 at 01:19:16PM +0530, Mugunthan V N wrote:
> >> On Wednesday 14 January 2015 10:28 PM, Felipe Balbi wrote:
> >> > CPSW never uses RX_THRESHOLD or MISC interrupts. In
> >> > fact, they are always kept masked in their appropriate
> >> > IRQ Enable register.
> >> >
> >> > Instead of allocating an IRQ that never fires, it's best
> >> > to remove that code altogether and let future patches
> >> > implement it if anybody needs those.
> >> >
> >> > Signed-off-by: Felipe Balbi <balbi@ti.com>
> >>
> >> Instead of introducing dummy ISR in previous patch and then removing in
> >> this patch, both can be squashed into a single patch.
> >
> > sure they can. I decided to split to ease review and to make sure only
> > one thing happens in a single patch.
>
> Indeed, I agree that adding something as a placeholder that just gets
> immediately removed should be avoided unless it is extremely difficult
> to do so.
what does this mean ? you prefer both patches to be combined ?
--
balbi
[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 819 bytes --]
^ permalink raw reply
* Re: [PATCH] i40e: don't enable and init FCOE by default when do PF reset
From: ethan zhao @ 2015-01-16 1:48 UTC (permalink / raw)
To: Dev, Vasu
Cc: e1000-devel@lists.sourceforge.net, brian.maly@oracle.com,
Allan, Bruce W, Brandeburg, Jesse, Parikh, Neerav, Linux NICS,
Ronciak, John, netdev@vger.kernel.org, Ethan Zhao,
linux-kernel@vger.kernel.org
In-Reply-To: <933BEC2E04D6A5458F4B0239FB547F9A34CCCD35@fmsmsx118.amr.corp.intel.com>
Vasu,
OK, disable FCOE as default configuration as a temporary step to
make it work.
Thanks,
Ethan
On 2015/1/16 7:45, Dev, Vasu wrote:
>> -----Original Message-----
>> From: ethan zhao [mailto:ethan.zhao@oracle.com]
>> Sent: Tuesday, January 13, 2015 6:41 PM
>> To: Dev, Vasu
>> Cc: Ethan Zhao; Ronciak, John; Kirsher, Jeffrey T; Brandeburg, Jesse; Allan,
>> Bruce W; Wyborny, Carolyn; Skidmore, Donald C; Rose, Gregory V; Vick,
>> Matthew; Williams, Mitch A; Parikh, Neerav; Linux NICS; e1000-
>> devel@lists.sourceforge.net; netdev@vger.kernel.org; linux-
>> kernel@vger.kernel.org; brian.maly@oracle.com
>> Subject: Re: [PATCH] i40e: don't enable and init FCOE by default when do PF
>> reset
>>
>> Vasu,
>>
>> On 2015/1/14 3:38, Dev, Vasu wrote:
>>>> -----Original Message-----
>>>>>>> diff --git a/drivers/net/ethernet/intel/i40e/i40e_main.c
>>>>>>> b/drivers/net/ethernet/intel/i40e/i40e_main.c
>>>>>>> index a5f2660..a2572cc 100644
>>>>>>> --- a/drivers/net/ethernet/intel/i40e/i40e_main.c
>>>>>>> +++ b/drivers/net/ethernet/intel/i40e/i40e_main.c
>>>>>>> @@ -6180,9 +6180,12 @@ static void i40e_reset_and_rebuild(struct
>>>>>>> i40e_pf *pf, bool reinit)
>>>>>>> }
>>>>>>> #endif /* CONFIG_I40E_DCB */
>>>>>>> #ifdef I40E_FCOE
>>>>>>> - ret = i40e_init_pf_fcoe(pf);
>>>>>>> - if (ret)
>>>>>>> - dev_info(&pf->pdev->dev, "init_pf_fcoe failed: %d\n", ret);
>>>>>>> + if (pf->flags & I40E_FLAG_FCOE_ENABLED) {
>>>>>>> + ret = i40e_init_pf_fcoe(pf);
>>>>> Calling i40e_init_pf_fcoe() here conflicts with its
>>>> I40E_FLAG_FCOE_ENABLED pre-condition since
>> I40E_FLAG_FCOE_ENABLED is
>>>> set by very same i40e_init_pf_fcoe(), in turn i40e_init_pf_fcoe()
>>>> will never get called.
>>>>
>>>> I don't think so, here ,i40e_reset_and_rebuild() is not the only
>>>> and the first place that i40e_init_pf_fcoe() is called, see
>>>> i40e_probe(), that is the first chance.
>>>>
>>>> i40e_probe()
>>>> -->i40e_sw_init()
>>>> -->i40e_init_pf_fcoe()
>>>>
>>>> And the I40E_FLAG_FCOE_ENABLED is possible be set by
>>>> i40e_fcoe_enable() or i40e_fcoe_disable() interface before the reset
>>>> action is to be done.
>>>>
>>> It is set by i40e_init_pf_fcoe() and you are right that the modified call flow
>> by your patch won't impact setting of I40E_FLAG_FCOE_ENABLED anyway
>> which could have prevented calling i40e_init_pf_fcoe() as I described above,
>> so this is not an issue with the patch.
>>>> BTW, the reason I post this patch is that we hit a bug, after setup
>>>> vlan, the PF is enabled to FCOE.
>>>>
>>> Then that BUG would still remain un-fixed and calling i40e_init_pf_fcoe()
>> under I40E_FLAG_FCOE_ENABLED flag really won't affect call flow to fix
>> anything. I mean I40E_FLAG_FCOE_ENABLED condition will be true with "pf-
>>> hw.func_caps.fcoe == true" and otherwise calling i40e_init_pf_fcoe() simply
>> returns back early on after checking "pf->hw.func_caps.fcoe == false", so
>> how that bug is fixed here by added I40E_FLAG_FCOE_ENABLED condition ?
>> What is the bug ?
>> The func_caps.fcoe is assigned by following call path, under our test
>> environment,
>>
>> i40e_probe()
>> ->i40e_get_capabilities()
>> ->i40e_aq_discover_capabilities()
>> ->i40e_parse_discover_capabilities()
>>
>> Or
>>
>> i40e_reset_and_rebuild()
>> ->i40e_get_capabilities()
>> ->i40e_aq_discover_capabilities()
>> ->i40e_parse_discover_capabilities()
>>
>> Under our test environment, the "pf->hw.func_caps.fcoe" is true. so if
>> i40e_reset_and_rebuild() is called for VLAN setup, ethtool diagnostic test.
>> And then i40e_init_pf_fcoe() is to be called,
>>
>> While if (!pf->hw.func_caps.fcoe) wouldn't return,
>>
> I said it would return with "pf->hw.func_caps.fcoe == false" in my last response, more details below.
>
>> So pf->flags is set to I40E_FLAG_FCOE_ENABLED.
>>
>> With my patch, i40e_init_pf_fcoe() is only called after
>> I40E_FLAG_FCOE_ENABLED is set before reset.
>>
>> Enable FCOE in i40e_probe() or not is another issue.
>>
> Nope since both cases we should do i40e_init_pf_fcoe() or don't based on fcoe cap true or false.
>
> I don't have much to add as I described before with the your patch that "calling i40e_init_pf_fcoe() under I40E_FLAG_FCOE_ENABLED flag really won't affect call flow to fix anything. I mean I40E_FLAG_FCOE_ENABLED condition will be true with "pf->hw.func_caps.fcoe == true" and otherwise calling i40e_init_pf_fcoe() simply returns back early on after checking "pf->hw.func_caps.fcoe == false".
>
> May be I'm missing something, I guess next either go with CONFIG_I40E_FCOE disable as I suggested before and now it in upstream kernel or we can have further off list discussion to fix the issue you are trying to fix with the patch.
>
> Thanks,
> Vasu
>
>> Thanks,
>> Ethan
>>
>>
>>>>> Jeff Kirsher should be getting out a patch queued by me which adds
>>>> I40E_FCoE Kbuild option, in that FCoE is disabled by default and
>>>> user could enable FCoE only if needed, that patch would do same of
>>>> skipping
>>>> i40e_init_pf_fcoe() whether FCoE capability in device enabled or not
>>>> in default config.
>>>> The following patch will not fix the above issue -- configuration of
>>>> PF will be changed via reset.
>>>> How about the FCOE is configured and disabled by i40e_fcoe_disable()
>>>> , then reset happens ?
>>>>
>>> May be but if the BUG is due to FCoE being enabled then having it disabled
>> in config will avoid the bug for non FCoE config option and once bug is
>> understood then that has to be fixed for FCoE enabled config also as I asked
>> above.
>>> Thanks Ethan for detailed response.
>>> Vasu
>>>
>>>>> From patchwork Wed Oct 2 23:26:08 2013
>>>>> Content-Type: text/plain; charset="utf-8"
>>>>> MIME-Version: 1.0
>>>>> Content-Transfer-Encoding: 7bit
>>>>> Subject: [net] i40e: adds FCoE configure option
>>>>> Date: Thu, 03 Oct 2013 07:26:08 -0000
>>>>> From: Vasu Dev <vasu.dev@intel.com>
>>>>> X-Patchwork-Id: 11797
>>>>>
>>>>> Adds FCoE config option I40E_FCOE, so that FCoE can be enabled as
>>>>> needed but otherwise have it disabled by default.
>>>>>
>>>>> This also eliminate multiple FCoE config checks, instead now just
>>>>> one config check for CONFIG_I40E_FCOE.
>>>>>
>>>>> The I40E FCoE was added with 3.17 kernel and therefore this patch
>>>>> shall be applied to stable 3.17 kernel also.
>>>>>
>>>>> CC: <stable@vger.kernel.org>
>>>>> Signed-off-by: Vasu Dev <vasu.dev@intel.com>
>>>>> Tested-by: Jim Young <jamesx.m.young@intel.com>
>>>>>
>>>>> ---
>>>>> drivers/net/ethernet/intel/Kconfig | 11 +++++++++++
>>>>> drivers/net/ethernet/intel/i40e/Makefile | 2 +-
>>>>> drivers/net/ethernet/intel/i40e/i40e_osdep.h | 4 ++--
>>>>> 3 files changed, 14 insertions(+), 3 deletions(-)
>>>>>
>>>>> diff --git a/drivers/net/ethernet/intel/Kconfig
>>>>> b/drivers/net/ethernet/intel/Kconfig
>>>>> index 5b8300a..4d61ef5 100644
>>>>> --- a/drivers/net/ethernet/intel/Kconfig
>>>>> +++ b/drivers/net/ethernet/intel/Kconfig
>>>>> @@ -281,6 +281,17 @@ config I40E_DCB
>>>>>
>>>>> If unsure, say N.
>>>>>
>>>>> +config I40E_FCOE
>>>>> + bool "Fibre Channel over Ethernet (FCoE)"
>>>>> + default n
>>>>> + depends on I40E && DCB && FCOE
>>>>> + ---help---
>>>>> + Say Y here if you want to use Fibre Channel over Ethernet (FCoE)
>>>>> + in the driver. This will create new netdev for exclusive FCoE
>>>>> + use with XL710 FCoE offloads enabled.
>>>>> +
>>>>> + If unsure, say N.
>>>>> +
>>>>> config I40EVF
>>>>> tristate "Intel(R) XL710 X710 Virtual Function Ethernet support"
>>>>> depends on PCI_MSI
>>>>> diff --git a/drivers/net/ethernet/intel/i40e/Makefile
>>>>> b/drivers/net/ethernet/intel/i40e/Makefile
>>>>> index 4b94ddb..c405819 100644
>>>>> --- a/drivers/net/ethernet/intel/i40e/Makefile
>>>>> +++ b/drivers/net/ethernet/intel/i40e/Makefile
>>>>> @@ -44,4 +44,4 @@ i40e-objs := i40e_main.o \
>>>>> i40e_virtchnl_pf.o
>>>>>
>>>>> i40e-$(CONFIG_I40E_DCB) += i40e_dcb.o i40e_dcb_nl.o
>>>>> -i40e-$(CONFIG_FCOE:m=y) += i40e_fcoe.o
>>>>> +i40e-$(CONFIG_I40E_FCOE) += i40e_fcoe.o
>>>>> diff --git a/drivers/net/ethernet/intel/i40e/i40e_osdep.h
>>>>> b/drivers/net/ethernet/intel/i40e/i40e_osdep.h
>>>>> index 045b5c4..ad802dd 100644
>>>>> --- a/drivers/net/ethernet/intel/i40e/i40e_osdep.h
>>>>> +++ b/drivers/net/ethernet/intel/i40e/i40e_osdep.h
>>>>> @@ -78,7 +78,7 @@ do { \
>>>>> } while (0)
>>>>>
>>>>> typedef enum i40e_status_code i40e_status; -#if
>>>>> defined(CONFIG_FCOE)
>>>>> || defined(CONFIG_FCOE_MODULE)
>>>>> +#ifdef CONFIG_I40E_FCOE
>>>>> #define I40E_FCOE
>>>>> -#endif /* CONFIG_FCOE or CONFIG_FCOE_MODULE */
>>>>> +#endif
>>>>> #endif /* _I40E_OSDEP_H_ */
>>>>>
>>>>>>> + if (ret)
>>>>>>> + dev_info(&pf->pdev->dev,
>>>>>>> + "init_pf_fcoe failed: %d\n", ret);
>>>>>>> + }
>>>>>>>
>>>>>>> #endif
>>>>>>> /* do basic switch setup */
>>>>>>> --
>>>>>>> 1.8.3.1
>>>> Thanks,
>>>> Ethan
------------------------------------------------------------------------------
New Year. New Location. New Benefits. New Data Center in Ashburn, VA.
GigeNET is offering a free month of service with a new server in Ashburn.
Choose from 2 high performing configs, both with 100TB of bandwidth.
Higher redundancy.Lower latency.Increased capacity.Completely compliant.
http://p.sf.net/sfu/gigenet
_______________________________________________
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] ethernet: atheros: Add nss-gmac driver
From: Stephen Hemminger @ 2015-01-16 1:56 UTC (permalink / raw)
To: Stephen Wang
Cc: jcliburn, grant.likely, robh+dt, linux-kernel, netdev, devicetree
In-Reply-To: <1420754626-30121-1-git-send-email-wstephen@codeaurora.org>
On Thu, 8 Jan 2015 14:03:46 -0800
Stephen Wang <wstephen@codeaurora.org> wrote:
> + uint32_t rx_bytes; /**< Number of RX bytes */
> + uint32_t rx_packets; /**< Number of RX packets */
32 bit packet stats are a problem with lots of traffic.
^ permalink raw reply
* Re: [PATCH] ethernet: atheros: Add nss-gmac driver
From: Stephen Hemminger @ 2015-01-16 1:58 UTC (permalink / raw)
To: Stephen Wang
Cc: jcliburn, grant.likely, robh+dt, linux-kernel, netdev, devicetree
In-Reply-To: <1420754626-30121-1-git-send-email-wstephen@codeaurora.org>
On Thu, 8 Jan 2015 14:03:46 -0800
Stephen Wang <wstephen@codeaurora.org> wrote:
> + struct nss_gmac_data_plane_ops *data_plane_ops;
> + /* ops to send messages to nss-drv */
This should be a const pointer.
^ permalink raw reply
* Re: [PATCH] ethernet: atheros: Add nss-gmac driver
From: Stephen Hemminger @ 2015-01-16 2:00 UTC (permalink / raw)
To: Stephen Wang
Cc: jcliburn-Re5JQEeQqe8AvxtiuMwx3w,
grant.likely-QSEj5FYQhm4dnm+yROfE0A,
robh+dt-DgEjT+Ai2ygdnm+yROfE0A,
linux-kernel-u79uwXL29TY76Z2rM5mHXA,
netdev-u79uwXL29TY76Z2rM5mHXA, devicetree-u79uwXL29TY76Z2rM5mHXA
In-Reply-To: <1420754626-30121-1-git-send-email-wstephen-sgV2jX0FEOL9JmXXK+q4OQ@public.gmane.org>
On Thu, 8 Jan 2015 14:03:46 -0800
Stephen Wang <wstephen-sgV2jX0FEOL9JmXXK+q4OQ@public.gmane.org> wrote:
> +static int32_t nss_gmac_setup_tx_desc_queue(struct nss_gmac_dev *gmacdev,
> + struct device *dev,
> + uint32_t no_of_desc,
> + uint32_t desc_mode)
> +{
> + int32_t i;
> + struct dma_desc *first_desc = NULL;
> + dma_addr_t dma_addr;
> +
> + gmacdev->tx_desc_count = 0;
> +
> + BUG_ON(desc_mode != RINGMODE);
> + BUG_ON((no_of_desc & (no_of_desc - 1)) != 0);
> +
> + netdev_dbg(gmacdev->netdev, "Total size of memory required for Tx Descriptors in Ring Mode = 0x%08x"
> + , (uint32_t) ((sizeof(struct dma_desc) * no_of_desc)));
> +
> + first_desc = dma_alloc_coherent(dev, sizeof(struct dma_desc) * no_of_desc
> + , &dma_addr, GFP_KERNEL);
> + if (first_desc == NULL) {
In a lot of places you first initialize a variable then assign it 5 lines
later to it's initial value. Get rid of the first initialization.
--
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
* [PATCH net-next v4] rhashtable: Fix race in rhashtable_destroy() and use regular work_struct
From: Ying Xue @ 2015-01-16 2:23 UTC (permalink / raw)
To: tgraf; +Cc: davem, sergei.shtylyov, netdev
When we put our declared work task in the global workqueue with
schedule_delayed_work(), its delay parameter is always zero.
Therefore, we should define a regular work in rhashtable structure
instead of a delayed work.
By the way, we add a condition to check whether resizing functions
are NULL before cancelling the work, avoiding to cancel an
uninitialized work.
Lastly, while we wait for all work items we submitted before to run
to completion with cancel_delayed_work(), ht->mutex has been taken in
rhashtable_destroy(). Moreover, cancel_delayed_work() doesn't return
until all work items are accomplished, and when work items are
scheduled, the work's function - rht_deferred_worker() will be called.
However, as rht_deferred_worker() also needs to acquire the lock,
deadlock might happen at the moment as the lock is already held before.
So if the cancel work function is moved out of the lock covered scope,
this will avoid the deadlock.
Fixes: 97defe1 ("rhashtable: Per bucket locks & deferred expansion/shrinking")
Signed-off-by: Ying Xue <ying.xue@windriver.com>
Cc: Thomas Graf <tgraf@suug.ch>
Signed-off-by: Ying Xue <ying.xue@windriver.com>
---
Change log:
v4:
- Add an empty line in rhashtable_destroy()
- Revise commit description
v3:
- Revise patch header description from Thomas's suggestion
v2:
- Move cancel_work_sync() out of ht->mutex lock scope from Thomas's
comments
include/linux/rhashtable.h | 2 +-
lib/rhashtable.c | 12 ++++++------
2 files changed, 7 insertions(+), 7 deletions(-)
diff --git a/include/linux/rhashtable.h b/include/linux/rhashtable.h
index 9570832..a2562ed 100644
--- a/include/linux/rhashtable.h
+++ b/include/linux/rhashtable.h
@@ -119,7 +119,7 @@ struct rhashtable {
atomic_t nelems;
atomic_t shift;
struct rhashtable_params p;
- struct delayed_work run_work;
+ struct work_struct run_work;
struct mutex mutex;
bool being_destroyed;
};
diff --git a/lib/rhashtable.c b/lib/rhashtable.c
index aca6998..84a78e3 100644
--- a/lib/rhashtable.c
+++ b/lib/rhashtable.c
@@ -485,7 +485,7 @@ static void rht_deferred_worker(struct work_struct *work)
struct rhashtable *ht;
struct bucket_table *tbl;
- ht = container_of(work, struct rhashtable, run_work.work);
+ ht = container_of(work, struct rhashtable, run_work);
mutex_lock(&ht->mutex);
tbl = rht_dereference(ht->tbl, ht);
@@ -507,7 +507,7 @@ static void rhashtable_wakeup_worker(struct rhashtable *ht)
if (tbl == new_tbl &&
((ht->p.grow_decision && ht->p.grow_decision(ht, size)) ||
(ht->p.shrink_decision && ht->p.shrink_decision(ht, size))))
- schedule_delayed_work(&ht->run_work, 0);
+ schedule_work(&ht->run_work);
}
static void __rhashtable_insert(struct rhashtable *ht, struct rhash_head *obj,
@@ -903,7 +903,7 @@ int rhashtable_init(struct rhashtable *ht, struct rhashtable_params *params)
get_random_bytes(&ht->p.hash_rnd, sizeof(ht->p.hash_rnd));
if (ht->p.grow_decision || ht->p.shrink_decision)
- INIT_DEFERRABLE_WORK(&ht->run_work, rht_deferred_worker);
+ INIT_WORK(&ht->run_work, rht_deferred_worker);
return 0;
}
@@ -921,11 +921,11 @@ void rhashtable_destroy(struct rhashtable *ht)
{
ht->being_destroyed = true;
- mutex_lock(&ht->mutex);
+ if (ht->p.grow_decision || ht->p.shrink_decision)
+ cancel_work_sync(&ht->run_work);
- cancel_delayed_work(&ht->run_work);
+ mutex_lock(&ht->mutex);
bucket_table_free(rht_dereference(ht->tbl, ht));
-
mutex_unlock(&ht->mutex);
}
EXPORT_SYMBOL_GPL(rhashtable_destroy);
--
1.7.9.5
^ 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