* [RFC PATCH] net/core: fix skb handling on netif serves for both bridge and vlan
From: Xiaotian Feng @ 2011-03-03 10:55 UTC (permalink / raw)
To: netdev
Cc: linux-kernel, Xiaotian Feng, David S. Miller, Eric Dumazet,
Tom Herbert
Consider network topology as follows:
eth0 eth1
|_____|
|
bond0 --- br0
|
vlan0 --- br1
bond0 serves for both br0 and vlan0, if a vlan tagged packet was sent
to br1 through bond0, bridge handling code is seeing the packet on bond0
and handing it off to my "legacy" bridge before vlan_tx_tag_present
and vlan_hwaccel_do_receive even haven't a chance to look at it.
Moving the vlan_tx_tag_present before bridge/macvlan handling code could
cure this.
Signed-off-by: Xiaotian Feng <dfeng@redhat.com>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Eric Dumazet <eric.dumazet@gmail.com>
Cc: Tom Herbert <therbert@google.com>
---
net/core/dev.c | 20 ++++++++++----------
1 files changed, 10 insertions(+), 10 deletions(-)
diff --git a/net/core/dev.c b/net/core/dev.c
index 8ae6631..d2d12c2 100644
--- a/net/core/dev.c
+++ b/net/core/dev.c
@@ -3079,27 +3079,27 @@ static int __netif_receive_skb(struct sk_buff *skb)
ncls:
#endif
- /* Handle special case of bridge or macvlan */
- rx_handler = rcu_dereference(skb->dev->rx_handler);
- if (rx_handler) {
+ if (vlan_tx_tag_present(skb)) {
if (pt_prev) {
ret = deliver_skb(skb, pt_prev, orig_dev);
pt_prev = NULL;
}
- skb = rx_handler(skb);
- if (!skb)
+ if (vlan_hwaccel_do_receive(&skb)) {
+ ret = __netif_receive_skb(skb);
+ goto out;
+ } else if (unlikely(!skb))
goto out;
}
- if (vlan_tx_tag_present(skb)) {
+ /* Handle special case of bridge or macvlan */
+ rx_handler = rcu_dereference(skb->dev->rx_handler);
+ if (rx_handler) {
if (pt_prev) {
ret = deliver_skb(skb, pt_prev, orig_dev);
pt_prev = NULL;
}
- if (vlan_hwaccel_do_receive(&skb)) {
- ret = __netif_receive_skb(skb);
- goto out;
- } else if (unlikely(!skb))
+ skb = rx_handler(skb);
+ if (!skb)
goto out;
}
--
1.7.1
^ permalink raw reply related
* net-next-2.6: Phonet pull request
From: Rémi Denis-Courmont @ 2011-03-03 11:22 UTC (permalink / raw)
To: netdev
Hello,
As "promised", here is the cleaned up and unified Phonet pipe code.
The same kernel can now support both old (e.g. Nokia N900) and new
(e.g. ST-Ericsson U8500 SoC) modems.
Due to lack of hardware, I only tested the ST-E support in software.
But the earlier code was horribly broken anyway.
The following changes since commit dc6ed1df5a5f84e45e77e2acb6fd99b995414956:
dcbnl: add support for retrieving peer configuration - cee (2011-03-02 21:58:55 -0800)
are available in the git repository at:
git://git.remlab.net/linux-phonet.git master
Rémi Denis-Courmont (8):
Phonet: fix NULL-deref in previous patch series
Phonet: return an error when TX fails
Phonet: fix backlog callback return value
Phonet: factor common code to send control messages
Phonet: allocate sock from accept syscall rather than soft IRQ
Phonet: provide pipe socket option to retrieve the pipe identifier
Phonet: support active connection without pipe controller on modem
Phonet: kill the ST-Ericsson pipe controller Kconfig
Documentation/networking/phonet.txt | 67 ++---
include/linux/phonet.h | 4 +-
include/net/phonet/pep.h | 1 -
net/phonet/Kconfig | 12 -
net/phonet/af_phonet.c | 13 +-
net/phonet/pep.c | 682 ++++++++++++++---------------------
net/phonet/socket.c | 112 +++----
7 files changed, 352 insertions(+), 539 deletions(-)
--
Rémi Denis-Courmont
http://www.remlab.net/
^ permalink raw reply
* RE: r8169 disable ASPM patch
From: hayeswang @ 2011-03-03 11:35 UTC (permalink / raw)
To: 'Francois Romieu', 'David Miller'; +Cc: netdev, sgruszka
In-Reply-To: <20110303094518.GA13491@electric-eye.fr.zoreil.com>
Sorry for my late response.
> -----Original Message-----
> From: Francois Romieu [mailto:romieu@fr.zoreil.com]
> Sent: Thursday, March 03, 2011 5:45 PM
> To: David Miller
> Cc: netdev@vger.kernel.org; Hayeswang; sgruszka@redhat.com
> Subject: Re: r8169 disable ASPM patch
>
> (Hayes added to the Cc:)
>
> David Miller <davem@davemloft.net> :
> [...]
> > What is the status of Stanislaw Gruszka's "disable ASPM"
> r8169 patch?
> >
> > http://patchwork.ozlabs.org/patch/83961/
> >
> > Is it waiting for your testing ?
>
> No.
>
> I can test it against regression but I do not have the technical
> knowledge to tell if it is right or not.
>
> I have read (again) the PR as well as its sibling and there is little
> evidence for me that the patch is 1) needed 2) better than the global
> pcie_aspm switch. Either Realtek's in-house knowledge helps making a
> technical decision or it is a policy-guided choice.
>
> > Is it waiting for feedback from Hayes?
>
> Yes.
>
Our hardware engineer says that it has an issue indeed for 8111B (that is, rev
01) when the aspm is enabled. For the other chips, the aspm would affect the
performance. We suggest to disable for stablity.
> Something like :
> [ ] Ack. So far there is no way to handle ASPM reliably on the 816x [
> ] Not optimal but it is ok as an interim fix. For instance :
> - if some 816x (8168 ? 810{2, 3} ?) are known broken while others
> are not
> - it can be mitigated but the fix is not available yet
> - needs more tester / vendor work
> - etc.
> [ ] Nak. The adequate fix is ...
>
> > Is there going to be a MAINTAINERS patch that adds Hayes?
> If so what
> > is stalling that?
Realtek would continue handling the kernel. However, the person may be different
for different period. Therefore, I think keeping the current mode is better.
>
> Realtek nic team asked me privately (end 2010) in rather
> general terms if they could maintain the r8169 kernel driver.
> I have given some relevant pointers for kernel code
> maintenance. Realtek did its housework.
>
> I have stated publicly that I'll ack such a patch if Hayes
> wants to add himself. Hayes was Cced. Either Hayes sends a
> patch or a plain "please do it".
>
Best Regards,
Hayes
^ permalink raw reply
* Re: [PATCH] sctp: fix the fast retransmit limit
From: Neil Horman @ 2011-03-03 11:41 UTC (permalink / raw)
To: Wei Yongjun
Cc: David Miller, Vlad Yasevich, netdev@vger.kernel.org, lksctp,
Mingyuan Zhu
In-Reply-To: <4D6F5F83.705@cn.fujitsu.com>
On Thu, Mar 03, 2011 at 05:29:39PM +0800, Wei Yongjun wrote:
> If chunk is still lost after fast retransmit, SCTP stack will
> never allow the second fast retransmit of this chunk, even if
> the peer need we do this. This chunk will be retransmit until
> the rtx timeout. This limit is introduce by the following patch:
> sctp: reduce memory footprint of sctp_chunk structure
> (c226ef9b83694311327f3ab0036c6de9c22e9daf)
>
> This patch revert this limit and removed useless SCTP_DONT_FRTX.
>
> Signed-off-by: Wei Yongjun <yjwei@cn.fujitsu.com>
Acked-by: Neil Horman <nhorman@tuxdriver.com>
> ---
> include/net/sctp/structs.h | 1 -
> net/sctp/outqueue.c | 4 ++--
> 2 files changed, 2 insertions(+), 3 deletions(-)
>
> diff --git a/include/net/sctp/structs.h b/include/net/sctp/structs.h
> index cc9185c..82a0f84 100644
> --- a/include/net/sctp/structs.h
> +++ b/include/net/sctp/structs.h
> @@ -751,7 +751,6 @@ struct sctp_chunk {
>
> #define SCTP_CAN_FRTX 0x0
> #define SCTP_NEED_FRTX 0x1
> -#define SCTP_DONT_FRTX 0x2
> __u16 rtt_in_progress:1, /* This chunk used for RTT calc? */
> has_tsn:1, /* Does this chunk have a TSN yet? */
> has_ssn:1, /* Does this chunk have a SSN yet? */
> diff --git a/net/sctp/outqueue.c b/net/sctp/outqueue.c
> index 8c6d379..7ed5862 100644
> --- a/net/sctp/outqueue.c
> +++ b/net/sctp/outqueue.c
> @@ -657,7 +657,7 @@ redo:
> * after it is retransmitted.
> */
> if (chunk->fast_retransmit == SCTP_NEED_FRTX)
> - chunk->fast_retransmit = SCTP_DONT_FRTX;
> + chunk->fast_retransmit = SCTP_CAN_FRTX;
>
> q->empty = 0;
> break;
> @@ -679,7 +679,7 @@ redo:
> if (rtx_timeout || fast_rtx) {
> list_for_each_entry(chunk1, lqueue, transmitted_list) {
> if (chunk1->fast_retransmit == SCTP_NEED_FRTX)
> - chunk1->fast_retransmit = SCTP_DONT_FRTX;
> + chunk1->fast_retransmit = SCTP_CAN_FRTX;
> }
> }
>
> --
> 1.6.5.2
>
>
>
^ permalink raw reply
* Re: bonding...
From: Neil Horman @ 2011-03-03 11:45 UTC (permalink / raw)
To: David Miller; +Cc: fubar, netdev
In-Reply-To: <20110302.214910.112578818.davem@davemloft.net>
On Wed, Mar 02, 2011 at 09:49:10PM -0800, David Miller wrote:
>
> Hey, if someone could step up and help with bonding maintainence
> in some tangible way, I'd really appreciate it.
>
> Currently the situation is that many people work on bonding patches,
> and whilst I do try and wait for some ACKs to arrive, I am the person
> who has to sort out when changes are ready, decide to apply them, and
> poke for review when things fall through the cracks.
>
> Sometimes patches go for weeks without ACKs, and in that situation
> I have to either try to understand the changes myself, or wait
> potentially forever for someone with bonding knowledge to take a
> good look at the patch and properly review it.
>
> It was nearly 2 weeks before Oleg V. Ukhno's 802.3ad round-robin patch
> got looked at by anyone with bonding knowledge. And it only happened
> because I got tired of seeing his poor patch rot in patchwork
> and had to explicitly asked for review the other day.
>
> This is unacceptable, people are submitting multiple bonding patches
> every single day now. It needs a clueful bonding person looking at
> these submissions on a constant basis.
>
> This is a serious problem and is backlogging the netdev patch queue.
>
> So if someone would become an active bonding patch-accumulator, and
> send me sets of patches that are ready to apply, I would really
> appreciate it.
>
> Thanks.
I nominate gospo. If he doesn't want to, I can do it
Neil
> --
> To unsubscribe from this list: send the line "unsubscribe netdev" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at http://vger.kernel.org/majordomo-info.html
>
^ permalink raw reply
* Re: [PATCH 3/5] tg3: Enclose macro with complex values in parenthesis
From: Javier Martinez Canillas @ 2011-03-03 11:54 UTC (permalink / raw)
To: David Miller; +Cc: mcarlson, mchan, grant.likely, netdev, kernelnewbies
In-Reply-To: <20110302.205601.71129255.davem@davemloft.net>
>
> This seems unnecessary, what problem do the current macro definitions
> create?
>
Probably the change is unnecessary, the only value is that it silences
a checkpatch error:
ERROR: Macros with complex values should be enclosed in parenthesis
Sorry for the noise.
--
-----------------------------------------
Javier Martínez Canillas
(+34) 682 39 81 69
PhD Student in High Performance Computing
Computer Architecture and Operating System Department (CAOS)
Universitat Autònoma de Barcelona
Barcelona, Spain
^ permalink raw reply
* Re: [PATCH 1/5] tg3: use usleep_range not msleep for small sleeps
From: Javier Martinez Canillas @ 2011-03-03 12:14 UTC (permalink / raw)
To: David Miller; +Cc: mcarlson, mchan, grant.likely, netdev, kernelnewbies
In-Reply-To: <20110302.205536.226759949.davem@davemloft.net>
>
>>
>> Signed-off-by: Javier Martinez Canillas <martinez.javier@gmail.com>
>
> Why?
>
As I understand msleep() will stop for a minimum of two jiffies and
that resolution is too low in systems with small HZ values (big
jiffies).
Using msleep() (with small values) in this systems will sleep for more
time than the caller expects.
I used usleep_range() because it is a finer precision implementations of msleep.
Sorry if I'm wrong or the sleep precision is not a concern in this
driver and the change doesn't add any value.
Thanks a lot.
--
-----------------------------------------
Javier Martínez Canillas
(+34) 682 39 81 69
PhD Student in High Performance Computing
Computer Architecture and Operating System Department (CAOS)
Universitat Autònoma de Barcelona
Barcelona, Spain
^ permalink raw reply
* Re: [BUG] VPN broken in net-next
From: Julian Anastasov @ 2011-03-03 12:41 UTC (permalink / raw)
To: David Miller; +Cc: shemminger, netdev
In-Reply-To: <20110302.165421.258093056.davem@davemloft.net>
Hello,
On Wed, 2 Mar 2011, David Miller wrote:
>> I really need to know what addresses interfaces have the time of the
>> __ip_dev_find() call which, if I'm not mistaken, is before the VPN is
>> up.
>
> Looking at pptp_connect(), it seems to be trying to do a route lookup
> using the source address, before it registers the PPP channel and that
> even gets brought up.
>
> I suspect that was working by luck previously, and it was getting the
> default route out perhaps another interface. If that is the case,
> removing the explicit "fl4_src" source address specification in the lookup
> flow pptp_connect() uses should fix the problem.
May be the problem is in inet_hash_insert(), it should
hash ifa_local, not ifa_address. May be they are equal for
the common case but not for peer addresses. In devinet_ioctl()
we can see they are equal initially:
ifa->ifa_address = ifa->ifa_local = sin->sin_addr.s_addr;
but later SIOCSIFDSTADDR can change ifa_address which
is destination address from the same prefix:
ifa->ifa_address = sin->sin_addr.s_addr;
fib_add_ifaddr() does more things but not for this
case with /32 mask, so the problem should be in using
ifa_address instead of ifa_local.
Regards
--
Julian Anastasov <ja@ssi.bg>
^ permalink raw reply
* Re: [RFC LOL OMG] pfifo_lat: qdisc that limits dequeueing based on estimated link latency
From: Eric Dumazet @ 2011-03-03 12:51 UTC (permalink / raw)
To: John W. Linville; +Cc: netdev, bloat-devel
In-Reply-To: <1299102850-2883-1-git-send-email-linville@tuxdriver.com>
Le mercredi 02 mars 2011 à 16:54 -0500, John W. Linville a écrit :
> This is a qdisc based on the existing pfifo_fast code. The difference
> is that this qdisc limits the dequeue rate based on estimates of how
> many packets can be in-flight at a given time while maintaining a target
> link latency.
>
> This work is based on the eBDP documented in Section IV of "Buffer
> Sizing for 802.11 Based Networks" by Tianji Li, et al.
>
> http://www.hamilton.ie/tianji_li/buffersizing.pdf
>
> This implementation timestamps an skb as it dequeues it, then
> computes the service time when the frame is freed by the driver.
> An exponentially weighted moving average of per fragment service times
> is used to restrict queueing delays in hopes of achieving a target
> fragment transmission latency. The skb->deconstructor mechanism is
> abused in order to obtain packet service time estimates.
>
> Signed-off-by: John W. Linville <linville@tuxdriver.com>
> ---
> I took a whack at reimplementing my eBDP patch at the qdisc level.
> Unfortunately, it doesn't seem to work very well and I'm at a loss
> as to why... :-( Comments welcome -- maybe I'm doing something really
> stupid in the math and just can't see it.
>
> The skb->deconstructor abuse includes adding a union member in the skb
> to record the qdisc->handle on the way out so that it can be used for
> accounting in the deconstructor -- thanks to Neil Horman for the
> suggestion!
>
> The reason I think this is an idea worth exploring is that existing
> qdisc code doesn't seem to account for the fact that the devices could
> be doing a lot of queueing behind them. Even Jussi's recent
> sch_fifo_ewma post doesn't seem to take into account how long the device
> holds-on to packets, which limits his ability to fight latency.
>
> Anyway, all comments appreciated!
>
>
Well, many issues in your patch.
skb destructor cannot be used like that (think about locking, and
various context where drivers actually free skbs (from interrupt, from
softirq, or even _before_ sending data on wire).
qdisc_lookup(skb->dev, skb->qdhandle) for example is only safe if run
with RTNL held. Its not meant to be used in fast path at all, but
management code only.
Being able to have a feedback on when a skb is freed (with a
notification of being delivered or dropped) is a recurring idea, so we
might design a stackable infrastructure.
^ permalink raw reply
* [net-next-2.6 00/24][pull request] Intel Wired LAN Driver Updates
From: Jeff Kirsher @ 2011-03-03 13:03 UTC (permalink / raw)
To: davem; +Cc: Jeff Kirsher, netdev, gospo, bphilips
The following series contains cleanups and fixes for igb, ixgbevf
and mainly ixgbe.
The following are changes since commit dc6ed1df5a5f84e45e77e2acb6fd99b995414956:
dcbnl: add support for retrieving peer configuration - cee
and are available in the git repository at:
master.kernel.org:/pub/scm/linux/kernel/git/jkirsher/net-next-2.6 master
Alexander Duyck (1):
ixgbe: balance free_irq calls with request_irq calls
Carolyn Wyborny (2):
igb: Fix reg pattern test in ethtool for i350 devices
igb: Fix strncpy calls to be safe per source code review tools
Don Skidmore (3):
ixgbe: cleanup X540 PHY reset function pointer
ixgbe: add function pointer for semaphore function
ixgbe: cleanup copyright string for 2011
Emil Tantilov (15):
ixgbe: cleanup wake on LAN defines
ixgbe: cleanup logic related to HW semaphores
ixgbe: set media type for 82599 T3 LOM
ixgbe: Add ability to double reset on failure to clear master enable
ixgbe: cleanup code in ixgbe_identify_sfp_module_generic
ixgbe: Check link wants report current link state
ixgbe: add polling test to end of PHY reset
ixgbe: Fill out PCIe speed and width enums with values
ixgbe: Bounds checking for set_rar, clear_rar, set_vmdq, clear_vmdq
ixgbe: rework ixgbe MTA handling to not drop packets
ixgbe: Drop unused code for setting up unicast addresses
ixgbe: Specific check for 100 Full link speed
ixgbe: Numerous whitespace / formatting cleanups
ixgbe: store permanent address before initializing Rx addresses
ixgbe: cleanup handling of I2C interface to PHY
Greg Rose (2):
ixgbevf: Fix Compiler Warnings
ixgbe: X540 Cleanup
Stefan Assmann (1):
igb: warn if max_vfs limit is exceeded
drivers/net/igb/igb_ethtool.c | 13 +-
drivers/net/igb/igb_main.c | 7 +-
drivers/net/ixgbe/ixgbe.h | 2 +-
drivers/net/ixgbe/ixgbe_82598.c | 63 ++++--
drivers/net/ixgbe/ixgbe_82599.c | 104 ++++++---
drivers/net/ixgbe/ixgbe_common.c | 438 ++++++++++++++---------------------
drivers/net/ixgbe/ixgbe_common.h | 4 +-
drivers/net/ixgbe/ixgbe_dcb.c | 2 +-
drivers/net/ixgbe/ixgbe_dcb.h | 2 +-
drivers/net/ixgbe/ixgbe_dcb_82598.c | 2 +-
drivers/net/ixgbe/ixgbe_dcb_82598.h | 2 +-
drivers/net/ixgbe/ixgbe_dcb_82599.c | 2 +-
drivers/net/ixgbe/ixgbe_dcb_82599.h | 2 +-
drivers/net/ixgbe/ixgbe_dcb_nl.c | 2 +-
drivers/net/ixgbe/ixgbe_ethtool.c | 2 +-
drivers/net/ixgbe/ixgbe_fcoe.c | 2 +-
drivers/net/ixgbe/ixgbe_fcoe.h | 2 +-
drivers/net/ixgbe/ixgbe_main.c | 12 +-
drivers/net/ixgbe/ixgbe_mbx.c | 2 +-
drivers/net/ixgbe/ixgbe_mbx.h | 2 +-
drivers/net/ixgbe/ixgbe_phy.c | 238 ++++++++++++++-----
drivers/net/ixgbe/ixgbe_phy.h | 2 +-
drivers/net/ixgbe/ixgbe_sriov.c | 2 +-
drivers/net/ixgbe/ixgbe_sriov.h | 2 +-
drivers/net/ixgbe/ixgbe_type.h | 40 ++--
drivers/net/ixgbe/ixgbe_x540.c | 34 ++-
drivers/net/ixgbevf/ixgbevf_main.c | 9 +-
27 files changed, 547 insertions(+), 447 deletions(-)
--
1.7.4
^ permalink raw reply
* [net-next-2.6 01/24] igb: warn if max_vfs limit is exceeded
From: Jeff Kirsher @ 2011-03-03 13:03 UTC (permalink / raw)
To: davem; +Cc: Stefan Assmann, netdev, gospo, bphilips, Jeff Kirsher
In-Reply-To: <1299157429-15878-1-git-send-email-jeffrey.t.kirsher@intel.com>
From: Stefan Assmann <sassmann@redhat.com>
Currently there's no warning printed when max_vfs > 7 is specified with
igb and the maximum of 7 is silently enforced. This patch prints a
warning and informs the user of the actions taken.
Signed-off-by: Stefan Assmann <sassmann@redhat.com>
Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/igb/igb_main.c | 7 ++++++-
1 files changed, 6 insertions(+), 1 deletions(-)
diff --git a/drivers/net/igb/igb_main.c b/drivers/net/igb/igb_main.c
index 579dbba..eef380a 100644
--- a/drivers/net/igb/igb_main.c
+++ b/drivers/net/igb/igb_main.c
@@ -2291,7 +2291,12 @@ static int __devinit igb_sw_init(struct igb_adapter *adapter)
switch (hw->mac.type) {
case e1000_82576:
case e1000_i350:
- adapter->vfs_allocated_count = (max_vfs > 7) ? 7 : max_vfs;
+ if (max_vfs > 7) {
+ dev_warn(&pdev->dev,
+ "Maximum of 7 VFs per PF, using max\n");
+ adapter->vfs_allocated_count = 7;
+ } else
+ adapter->vfs_allocated_count = max_vfs;
break;
default:
break;
--
1.7.4
^ permalink raw reply related
* [net-next-2.6 03/24] igb: Fix strncpy calls to be safe per source code review tools
From: Jeff Kirsher @ 2011-03-03 13:03 UTC (permalink / raw)
To: davem; +Cc: Carolyn Wyborny, netdev, gospo, bphilips, Jeff Kirsher
In-Reply-To: <1299157429-15878-1-git-send-email-jeffrey.t.kirsher@intel.com>
From: Carolyn Wyborny <carolyn.wyborny@intel.com>
This fix changes the remaining calls to strncpy that have not yet
been changed to use the "sizeof(buf) - 1" syntax rather than just
a number for buffer size.
Signed-off-by: Carolyn Wyborny <carolyn.wyborny@intel.com>
Tested-by: Aaron Brown <aaron.f.brown@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/igb/igb_ethtool.c | 11 +++++++----
1 files changed, 7 insertions(+), 4 deletions(-)
diff --git a/drivers/net/igb/igb_ethtool.c b/drivers/net/igb/igb_ethtool.c
index 16bbd49..61f7849 100644
--- a/drivers/net/igb/igb_ethtool.c
+++ b/drivers/net/igb/igb_ethtool.c
@@ -727,8 +727,9 @@ static void igb_get_drvinfo(struct net_device *netdev,
char firmware_version[32];
u16 eeprom_data;
- strncpy(drvinfo->driver, igb_driver_name, 32);
- strncpy(drvinfo->version, igb_driver_version, 32);
+ strncpy(drvinfo->driver, igb_driver_name, sizeof(drvinfo->driver) - 1);
+ strncpy(drvinfo->version, igb_driver_version,
+ sizeof(drvinfo->version) - 1);
/* EEPROM image version # is reported as firmware version # for
* 82575 controllers */
@@ -738,8 +739,10 @@ static void igb_get_drvinfo(struct net_device *netdev,
(eeprom_data & 0x0FF0) >> 4,
eeprom_data & 0x000F);
- strncpy(drvinfo->fw_version, firmware_version, 32);
- strncpy(drvinfo->bus_info, pci_name(adapter->pdev), 32);
+ strncpy(drvinfo->fw_version, firmware_version,
+ sizeof(drvinfo->fw_version) - 1);
+ strncpy(drvinfo->bus_info, pci_name(adapter->pdev),
+ sizeof(drvinfo->bus_info) - 1);
drvinfo->n_stats = IGB_STATS_LEN;
drvinfo->testinfo_len = IGB_TEST_LEN;
drvinfo->regdump_len = igb_get_regs_len(netdev);
--
1.7.4
^ permalink raw reply related
* [net-next-2.6 02/24] igb: Fix reg pattern test in ethtool for i350 devices
From: Jeff Kirsher @ 2011-03-03 13:03 UTC (permalink / raw)
To: davem; +Cc: Carolyn Wyborny, netdev, gospo, bphilips, Jeff Kirsher
In-Reply-To: <1299157429-15878-1-git-send-email-jeffrey.t.kirsher@intel.com>
From: Carolyn Wyborny <carolyn.wyborny@intel.com>
This fixes the reg_pattern_test so that the test does not fail
on i350 parts.
Signed-off-by: Carolyn Wyborny <carolyn.wyborny@intel.com>
Tested-by: Jeff Pieper <jeffrey.e.pieper@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/igb/igb_ethtool.c | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/drivers/net/igb/igb_ethtool.c b/drivers/net/igb/igb_ethtool.c
index a70e16b..16bbd49 100644
--- a/drivers/net/igb/igb_ethtool.c
+++ b/drivers/net/igb/igb_ethtool.c
@@ -1070,7 +1070,7 @@ static bool reg_pattern_test(struct igb_adapter *adapter, u64 *data,
{0x5A5A5A5A, 0xA5A5A5A5, 0x00000000, 0xFFFFFFFF};
for (pat = 0; pat < ARRAY_SIZE(_test); pat++) {
wr32(reg, (_test[pat] & write));
- val = rd32(reg);
+ val = rd32(reg) & mask;
if (val != (_test[pat] & write & mask)) {
dev_err(&adapter->pdev->dev, "pattern test reg %04X "
"failed: got 0x%08X expected 0x%08X\n",
--
1.7.4
^ permalink raw reply related
* [net-next-2.6 04/24] ixgbevf: Fix Compiler Warnings
From: Jeff Kirsher @ 2011-03-03 13:03 UTC (permalink / raw)
To: davem; +Cc: Greg Rose, netdev, gospo, bphilips, Jeff Kirsher
In-Reply-To: <1299157429-15878-1-git-send-email-jeffrey.t.kirsher@intel.com>
From: Greg Rose <gregory.v.rose@intel.com>
Fix Compiler warnings of variables that are initialized but not used.
Signed-off-by: Greg Rose <gregory.v.rose@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/ixgbevf/ixgbevf_main.c | 9 ++-------
1 files changed, 2 insertions(+), 7 deletions(-)
diff --git a/drivers/net/ixgbevf/ixgbevf_main.c b/drivers/net/ixgbevf/ixgbevf_main.c
index 1e735a1..8276881 100644
--- a/drivers/net/ixgbevf/ixgbevf_main.c
+++ b/drivers/net/ixgbevf/ixgbevf_main.c
@@ -178,8 +178,6 @@ static inline bool ixgbevf_check_tx_hang(struct ixgbevf_adapter *adapter,
tx_ring->tx_buffer_info[eop].time_stamp &&
time_after(jiffies, tx_ring->tx_buffer_info[eop].time_stamp + HZ)) {
/* detected Tx unit hang */
- union ixgbe_adv_tx_desc *tx_desc;
- tx_desc = IXGBE_TX_DESC_ADV(*tx_ring, eop);
printk(KERN_ERR "Detected Tx Unit Hang\n"
" Tx Queue <%d>\n"
" TDH, TDT <%x>, <%x>\n"
@@ -334,7 +332,6 @@ static void ixgbevf_receive_skb(struct ixgbevf_q_vector *q_vector,
struct ixgbevf_adapter *adapter = q_vector->adapter;
bool is_vlan = (status & IXGBE_RXD_STAT_VP);
u16 tag = le16_to_cpu(rx_desc->wb.upper.vlan);
- int ret;
if (!(adapter->flags & IXGBE_FLAG_IN_NETPOLL)) {
if (adapter->vlgrp && is_vlan)
@@ -345,9 +342,9 @@ static void ixgbevf_receive_skb(struct ixgbevf_q_vector *q_vector,
napi_gro_receive(&q_vector->napi, skb);
} else {
if (adapter->vlgrp && is_vlan)
- ret = vlan_hwaccel_rx(skb, adapter->vlgrp, tag);
+ vlan_hwaccel_rx(skb, adapter->vlgrp, tag);
else
- ret = netif_rx(skb);
+ netif_rx(skb);
}
}
@@ -3287,8 +3284,6 @@ static const struct net_device_ops ixgbe_netdev_ops = {
static void ixgbevf_assign_netdev_ops(struct net_device *dev)
{
- struct ixgbevf_adapter *adapter;
- adapter = netdev_priv(dev);
dev->netdev_ops = &ixgbe_netdev_ops;
ixgbevf_set_ethtool_ops(dev);
dev->watchdog_timeo = 5 * HZ;
--
1.7.4
^ permalink raw reply related
* [net-next-2.6 05/24] ixgbe: cleanup wake on LAN defines
From: Jeff Kirsher @ 2011-03-03 13:03 UTC (permalink / raw)
To: davem; +Cc: Emil Tantilov, netdev, gospo, bphilips, Jeff Kirsher
In-Reply-To: <1299157429-15878-1-git-send-email-jeffrey.t.kirsher@intel.com>
From: Emil Tantilov <emil.s.tantilov@intel.com>
This change just cleans up a few defines in ixgbe_type.h related to wake on LAN.
Signed-off-by: Emil Tantilov <emil.s.tantilov@intel.com>
Tested-by: Stephen Ko <stephen.s.ko@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/ixgbe/ixgbe_type.h | 4 ++--
1 files changed, 2 insertions(+), 2 deletions(-)
diff --git a/drivers/net/ixgbe/ixgbe_type.h b/drivers/net/ixgbe/ixgbe_type.h
index ab65d13..18780d0 100644
--- a/drivers/net/ixgbe/ixgbe_type.h
+++ b/drivers/net/ixgbe/ixgbe_type.h
@@ -91,7 +91,7 @@
/* General Receive Control */
#define IXGBE_GRC_MNG 0x00000001 /* Manageability Enable */
-#define IXGBE_GRC_APME 0x00000002 /* Advanced Power Management Enable */
+#define IXGBE_GRC_APME 0x00000002 /* APM enabled in EEPROM */
#define IXGBE_VPDDIAG0 0x10204
#define IXGBE_VPDDIAG1 0x10208
@@ -342,7 +342,7 @@
/* Wake Up Control */
#define IXGBE_WUC_PME_EN 0x00000002 /* PME Enable */
#define IXGBE_WUC_PME_STATUS 0x00000004 /* PME Status */
-#define IXGBE_WUC_ADVD3WUC 0x00000010 /* D3Cold wake up cap. enable*/
+#define IXGBE_WUC_WKEN 0x00000010 /* Enable PE_WAKE_N pin assertion */
/* Wake Up Filter Control */
#define IXGBE_WUFC_LNKC 0x00000001 /* Link Status Change Wakeup Enable */
--
1.7.4
^ permalink raw reply related
* [net-next-2.6 07/24] ixgbe: balance free_irq calls with request_irq calls
From: Jeff Kirsher @ 2011-03-03 13:03 UTC (permalink / raw)
To: davem; +Cc: Alexander Duyck, netdev, gospo, bphilips, Jeff Kirsher
In-Reply-To: <1299157429-15878-1-git-send-email-jeffrey.t.kirsher@intel.com>
From: Alexander Duyck <alexander.h.duyck@intel.com>
We were incorrectly freeing IRQs that we had not requested. This change
corrects that by making certain we only free q_vectors that we have
requested IRQs for.
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Tested-by: Ross Brattain <ross.b.brattain@intel.com>
Tested-by: Stephen Ko <stephen.s.ko@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/ixgbe/ixgbe_main.c | 5 +++++
1 files changed, 5 insertions(+), 0 deletions(-)
diff --git a/drivers/net/ixgbe/ixgbe_main.c b/drivers/net/ixgbe/ixgbe_main.c
index f0d0c5a..588661b 100644
--- a/drivers/net/ixgbe/ixgbe_main.c
+++ b/drivers/net/ixgbe/ixgbe_main.c
@@ -2597,6 +2597,11 @@ static void ixgbe_free_irq(struct ixgbe_adapter *adapter)
i--;
for (; i >= 0; i--) {
+ /* free only the irqs that were actually requested */
+ if (!adapter->q_vector[i]->rxr_count &&
+ !adapter->q_vector[i]->txr_count)
+ continue;
+
free_irq(adapter->msix_entries[i].vector,
adapter->q_vector[i]);
}
--
1.7.4
^ permalink raw reply related
* [net-next-2.6 06/24] ixgbe: cleanup logic related to HW semaphores
From: Jeff Kirsher @ 2011-03-03 13:03 UTC (permalink / raw)
To: davem; +Cc: Emil Tantilov, netdev, gospo, bphilips, Jeff Kirsher
In-Reply-To: <1299157429-15878-1-git-send-email-jeffrey.t.kirsher@intel.com>
From: Emil Tantilov <emil.s.tantilov@intel.com>
This change cleans up much of the logic related to the hardware semaphores
on the adapters. There were a number of issues with timings that needed to
be addressed.
Signed-off-by: Emil Tantilov <emil.s.tantilov@intel.com>
Tested-by: Stephen Ko <stephen.s.ko@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/ixgbe/ixgbe_common.c | 42 +++++++++++++++++++++----------------
1 files changed, 24 insertions(+), 18 deletions(-)
diff --git a/drivers/net/ixgbe/ixgbe_common.c b/drivers/net/ixgbe/ixgbe_common.c
index ebbda7d..345c32e 100644
--- a/drivers/net/ixgbe/ixgbe_common.c
+++ b/drivers/net/ixgbe/ixgbe_common.c
@@ -603,7 +603,6 @@ s32 ixgbe_write_eeprom_generic(struct ixgbe_hw *hw, u16 offset, u16 data)
ixgbe_shift_out_eeprom_bits(hw, data, 16);
ixgbe_standby_eeprom(hw);
- msleep(hw->eeprom.semaphore_delay);
/* Done with writing - release the EEPROM */
ixgbe_release_eeprom(hw);
}
@@ -747,7 +746,7 @@ s32 ixgbe_poll_eerd_eewr_done(struct ixgbe_hw *hw, u32 ee_reg)
static s32 ixgbe_acquire_eeprom(struct ixgbe_hw *hw)
{
s32 status = 0;
- u32 eec = 0;
+ u32 eec;
u32 i;
if (ixgbe_acquire_swfw_sync(hw, IXGBE_GSSR_EEP_SM) != 0)
@@ -776,15 +775,15 @@ static s32 ixgbe_acquire_eeprom(struct ixgbe_hw *hw)
ixgbe_release_swfw_sync(hw, IXGBE_GSSR_EEP_SM);
status = IXGBE_ERR_EEPROM;
}
- }
- /* Setup EEPROM for Read/Write */
- if (status == 0) {
- /* Clear CS and SK */
- eec &= ~(IXGBE_EEC_CS | IXGBE_EEC_SK);
- IXGBE_WRITE_REG(hw, IXGBE_EEC, eec);
- IXGBE_WRITE_FLUSH(hw);
- udelay(1);
+ /* Setup EEPROM for Read/Write */
+ if (status == 0) {
+ /* Clear CS and SK */
+ eec &= ~(IXGBE_EEC_CS | IXGBE_EEC_SK);
+ IXGBE_WRITE_REG(hw, IXGBE_EEC, eec);
+ IXGBE_WRITE_FLUSH(hw);
+ udelay(1);
+ }
}
return status;
}
@@ -798,13 +797,10 @@ static s32 ixgbe_acquire_eeprom(struct ixgbe_hw *hw)
static s32 ixgbe_get_eeprom_semaphore(struct ixgbe_hw *hw)
{
s32 status = IXGBE_ERR_EEPROM;
- u32 timeout;
+ u32 timeout = 2000;
u32 i;
u32 swsm;
- /* Set timeout value based on size of EEPROM */
- timeout = hw->eeprom.word_size + 1;
-
/* Get SMBI software semaphore between device drivers first */
for (i = 0; i < timeout; i++) {
/*
@@ -816,7 +812,7 @@ static s32 ixgbe_get_eeprom_semaphore(struct ixgbe_hw *hw)
status = 0;
break;
}
- msleep(1);
+ udelay(50);
}
/* Now get the semaphore between SW/FW through the SWESMBI bit */
@@ -844,11 +840,14 @@ static s32 ixgbe_get_eeprom_semaphore(struct ixgbe_hw *hw)
* was not granted because we don't have access to the EEPROM
*/
if (i >= timeout) {
- hw_dbg(hw, "Driver can't access the Eeprom - Semaphore "
+ hw_dbg(hw, "SWESMBI Software EEPROM semaphore "
"not granted.\n");
ixgbe_release_eeprom_semaphore(hw);
status = IXGBE_ERR_EEPROM;
}
+ } else {
+ hw_dbg(hw, "Software semaphore SMBI between device drivers "
+ "not granted.\n");
}
return status;
@@ -1081,10 +1080,13 @@ static void ixgbe_release_eeprom(struct ixgbe_hw *hw)
IXGBE_WRITE_REG(hw, IXGBE_EEC, eec);
ixgbe_release_swfw_sync(hw, IXGBE_GSSR_EEP_SM);
+
+ /* Delay before attempt to obtain semaphore again to allow FW access */
+ msleep(hw->eeprom.semaphore_delay);
}
/**
- * ixgbe_calc_eeprom_checksum - Calculates and returns the checksum
+ * ixgbe_calc_eeprom_checksum_generic - Calculates and returns the checksum
* @hw: pointer to hardware structure
**/
u16 ixgbe_calc_eeprom_checksum_generic(struct ixgbe_hw *hw)
@@ -2206,6 +2208,10 @@ s32 ixgbe_acquire_swfw_sync(struct ixgbe_hw *hw, u16 mask)
s32 timeout = 200;
while (timeout) {
+ /*
+ * SW EEPROM semaphore bit is used for access to all
+ * SW_FW_SYNC/GSSR bits (not just EEPROM)
+ */
if (ixgbe_get_eeprom_semaphore(hw))
return IXGBE_ERR_SWFW_SYNC;
@@ -2223,7 +2229,7 @@ s32 ixgbe_acquire_swfw_sync(struct ixgbe_hw *hw, u16 mask)
}
if (!timeout) {
- hw_dbg(hw, "Driver can't access resource, GSSR timeout.\n");
+ hw_dbg(hw, "Driver can't access resource, SW_FW_SYNC timeout.\n");
return IXGBE_ERR_SWFW_SYNC;
}
--
1.7.4
^ permalink raw reply related
* [net-next-2.6 08/24] ixgbe: set media type for 82599 T3 LOM
From: Jeff Kirsher @ 2011-03-03 13:03 UTC (permalink / raw)
To: davem; +Cc: Emil Tantilov, netdev, gospo, bphilips, Jeff Kirsher
In-Reply-To: <1299157429-15878-1-git-send-email-jeffrey.t.kirsher@intel.com>
From: Emil Tantilov <emil.s.tantilov@intel.com>
The media type was not being set for the 82599 T3 LAN on motherboard. This
change corrects that.
Signed-off-by: Emil Tantilov <emil.s.tantilov@intel.com>
Tested-by: Stephen Ko <stephen.s.ko@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/ixgbe/ixgbe_82599.c | 50 +++++++++++++++++++++++++++++++-------
drivers/net/ixgbe/ixgbe_main.c | 2 +-
drivers/net/ixgbe/ixgbe_type.h | 1 +
3 files changed, 42 insertions(+), 11 deletions(-)
diff --git a/drivers/net/ixgbe/ixgbe_82599.c b/drivers/net/ixgbe/ixgbe_82599.c
index a21f581..b45a491 100644
--- a/drivers/net/ixgbe/ixgbe_82599.c
+++ b/drivers/net/ixgbe/ixgbe_82599.c
@@ -329,11 +329,14 @@ static enum ixgbe_media_type ixgbe_get_media_type_82599(struct ixgbe_hw *hw)
enum ixgbe_media_type media_type;
/* Detect if there is a copper PHY attached. */
- if (hw->phy.type == ixgbe_phy_cu_unknown ||
- hw->phy.type == ixgbe_phy_tn ||
- hw->phy.type == ixgbe_phy_aq) {
+ switch (hw->phy.type) {
+ case ixgbe_phy_cu_unknown:
+ case ixgbe_phy_tn:
+ case ixgbe_phy_aq:
media_type = ixgbe_media_type_copper;
goto out;
+ default:
+ break;
}
switch (hw->device_id) {
@@ -354,6 +357,9 @@ static enum ixgbe_media_type ixgbe_get_media_type_82599(struct ixgbe_hw *hw)
case IXGBE_DEV_ID_82599_CX4:
media_type = ixgbe_media_type_cx4;
break;
+ case IXGBE_DEV_ID_82599_T3_LOM:
+ media_type = ixgbe_media_type_copper;
+ break;
default:
media_type = ixgbe_media_type_unknown;
break;
@@ -1733,13 +1739,34 @@ static s32 ixgbe_start_hw_82599(struct ixgbe_hw *hw)
* @hw: pointer to hardware structure
*
* Determines the physical layer module found on the current adapter.
+ * If PHY already detected, maintains current PHY type in hw struct,
+ * otherwise executes the PHY detection routine.
**/
-static s32 ixgbe_identify_phy_82599(struct ixgbe_hw *hw)
+s32 ixgbe_identify_phy_82599(struct ixgbe_hw *hw)
{
s32 status = IXGBE_ERR_PHY_ADDR_INVALID;
+
+ /* Detect PHY if not unknown - returns success if already detected. */
status = ixgbe_identify_phy_generic(hw);
- if (status != 0)
- status = ixgbe_identify_sfp_module_generic(hw);
+ if (status != 0) {
+ /* 82599 10GBASE-T requires an external PHY */
+ if (hw->mac.ops.get_media_type(hw) == ixgbe_media_type_copper)
+ goto out;
+ else
+ status = ixgbe_identify_sfp_module_generic(hw);
+ }
+
+ /* Set PHY type none if no PHY detected */
+ if (hw->phy.type == ixgbe_phy_unknown) {
+ hw->phy.type = ixgbe_phy_none;
+ status = 0;
+ }
+
+ /* Return error if SFP module has been detected but is not supported */
+ if (hw->phy.type == ixgbe_phy_sfp_unsupported)
+ status = IXGBE_ERR_SFP_NOT_SUPPORTED;
+
+out:
return status;
}
@@ -1763,11 +1790,12 @@ static u32 ixgbe_get_supported_physical_layer_82599(struct ixgbe_hw *hw)
hw->phy.ops.identify(hw);
- if (hw->phy.type == ixgbe_phy_tn ||
- hw->phy.type == ixgbe_phy_aq ||
- hw->phy.type == ixgbe_phy_cu_unknown) {
+ switch (hw->phy.type) {
+ case ixgbe_phy_tn:
+ case ixgbe_phy_aq:
+ case ixgbe_phy_cu_unknown:
hw->phy.ops.read_reg(hw, MDIO_PMA_EXTABLE, MDIO_MMD_PMAPMD,
- &ext_ability);
+ &ext_ability);
if (ext_ability & MDIO_PMA_EXTABLE_10GBT)
physical_layer |= IXGBE_PHYSICAL_LAYER_10GBASE_T;
if (ext_ability & MDIO_PMA_EXTABLE_1000BT)
@@ -1775,6 +1803,8 @@ static u32 ixgbe_get_supported_physical_layer_82599(struct ixgbe_hw *hw)
if (ext_ability & MDIO_PMA_EXTABLE_100BTX)
physical_layer |= IXGBE_PHYSICAL_LAYER_100BASE_TX;
goto out;
+ default:
+ break;
}
switch (autoc & IXGBE_AUTOC_LMS_MASK) {
diff --git a/drivers/net/ixgbe/ixgbe_main.c b/drivers/net/ixgbe/ixgbe_main.c
index 588661b..9877719 100644
--- a/drivers/net/ixgbe/ixgbe_main.c
+++ b/drivers/net/ixgbe/ixgbe_main.c
@@ -3889,7 +3889,7 @@ static int ixgbe_up_complete(struct ixgbe_adapter *adapter)
* If we're not hot-pluggable SFP+, we just need to configure link
* and bring it up.
*/
- if (hw->phy.type == ixgbe_phy_unknown)
+ if (hw->phy.type == ixgbe_phy_none)
schedule_work(&adapter->sfp_config_module_task);
/* enable transmits */
diff --git a/drivers/net/ixgbe/ixgbe_type.h b/drivers/net/ixgbe/ixgbe_type.h
index 18780d0..af5ad40 100644
--- a/drivers/net/ixgbe/ixgbe_type.h
+++ b/drivers/net/ixgbe/ixgbe_type.h
@@ -2242,6 +2242,7 @@ enum ixgbe_mac_type {
enum ixgbe_phy_type {
ixgbe_phy_unknown = 0,
+ ixgbe_phy_none,
ixgbe_phy_tn,
ixgbe_phy_aq,
ixgbe_phy_cu_unknown,
--
1.7.4
^ permalink raw reply related
* [net-next-2.6 09/24] ixgbe: Add ability to double reset on failure to clear master enable
From: Jeff Kirsher @ 2011-03-03 13:03 UTC (permalink / raw)
To: davem; +Cc: Emil Tantilov, netdev, gospo, bphilips, Jeff Kirsher
In-Reply-To: <1299157429-15878-1-git-send-email-jeffrey.t.kirsher@intel.com>
From: Emil Tantilov <emil.s.tantilov@intel.com>
Double resets are required for recovery from certain error conditions.
Between resets, it is necessary to stall to allow time for any pending HW
events to complete. We use 1usec since that is what is needed for
ixgbe_disable_pcie_master(). The second reset then clears out any effects
of those events.
Signed-off-by: Emil Tantilov <emil.s.tantilov@intel.com>
Tested-by: Stephen Ko <stephen.s.ko@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/ixgbe/ixgbe_82598.c | 21 ++++++++++++-----
drivers/net/ixgbe/ixgbe_82599.c | 20 ++++++++++++----
drivers/net/ixgbe/ixgbe_common.c | 47 +++++++++++++++++++++++++++++++++-----
drivers/net/ixgbe/ixgbe_type.h | 4 +++
drivers/net/ixgbe/ixgbe_x540.c | 20 ++++++++++++----
5 files changed, 90 insertions(+), 22 deletions(-)
diff --git a/drivers/net/ixgbe/ixgbe_82598.c b/drivers/net/ixgbe/ixgbe_82598.c
index d0f1d9d..291b1e6 100644
--- a/drivers/net/ixgbe/ixgbe_82598.c
+++ b/drivers/net/ixgbe/ixgbe_82598.c
@@ -770,7 +770,6 @@ static s32 ixgbe_reset_hw_82598(struct ixgbe_hw *hw)
else if (phy_status == IXGBE_ERR_SFP_NOT_PRESENT)
goto no_phy_reset;
-
hw->phy.ops.reset(hw);
}
@@ -779,12 +778,9 @@ no_phy_reset:
* Prevent the PCI-E bus from from hanging by disabling PCI-E master
* access and verify no pending requests before reset
*/
- status = ixgbe_disable_pcie_master(hw);
- if (status != 0) {
- status = IXGBE_ERR_MASTER_REQUESTS_PENDING;
- hw_dbg(hw, "PCI-E Master disable polling has failed.\n");
- }
+ ixgbe_disable_pcie_master(hw);
+mac_reset_top:
/*
* Issue global reset to the MAC. This needs to be a SW reset.
* If link reset is used, it might reset the MAC when mng is using it
@@ -805,6 +801,19 @@ no_phy_reset:
hw_dbg(hw, "Reset polling failed to complete.\n");
}
+ /*
+ * Double resets are required for recovery from certain error
+ * conditions. Between resets, it is necessary to stall to allow time
+ * for any pending HW events to complete. We use 1usec since that is
+ * what is needed for ixgbe_disable_pcie_master(). The second reset
+ * then clears out any effects of those events.
+ */
+ if (hw->mac.flags & IXGBE_FLAGS_DOUBLE_RESET_REQUIRED) {
+ hw->mac.flags &= ~IXGBE_FLAGS_DOUBLE_RESET_REQUIRED;
+ udelay(1);
+ goto mac_reset_top;
+ }
+
msleep(50);
gheccr = IXGBE_READ_REG(hw, IXGBE_GHECCR);
diff --git a/drivers/net/ixgbe/ixgbe_82599.c b/drivers/net/ixgbe/ixgbe_82599.c
index b45a491..126a06f 100644
--- a/drivers/net/ixgbe/ixgbe_82599.c
+++ b/drivers/net/ixgbe/ixgbe_82599.c
@@ -904,12 +904,9 @@ static s32 ixgbe_reset_hw_82599(struct ixgbe_hw *hw)
* Prevent the PCI-E bus from from hanging by disabling PCI-E master
* access and verify no pending requests before reset
*/
- status = ixgbe_disable_pcie_master(hw);
- if (status != 0) {
- status = IXGBE_ERR_MASTER_REQUESTS_PENDING;
- hw_dbg(hw, "PCI-E Master disable polling has failed.\n");
- }
+ ixgbe_disable_pcie_master(hw);
+mac_reset_top:
/*
* Issue global reset to the MAC. This needs to be a SW reset.
* If link reset is used, it might reset the MAC when mng is using it
@@ -930,6 +927,19 @@ static s32 ixgbe_reset_hw_82599(struct ixgbe_hw *hw)
hw_dbg(hw, "Reset polling failed to complete.\n");
}
+ /*
+ * Double resets are required for recovery from certain error
+ * conditions. Between resets, it is necessary to stall to allow time
+ * for any pending HW events to complete. We use 1usec since that is
+ * what is needed for ixgbe_disable_pcie_master(). The second reset
+ * then clears out any effects of those events.
+ */
+ if (hw->mac.flags & IXGBE_FLAGS_DOUBLE_RESET_REQUIRED) {
+ hw->mac.flags &= ~IXGBE_FLAGS_DOUBLE_RESET_REQUIRED;
+ udelay(1);
+ goto mac_reset_top;
+ }
+
msleep(50);
/*
diff --git a/drivers/net/ixgbe/ixgbe_common.c b/drivers/net/ixgbe/ixgbe_common.c
index 345c32e..6d87c74 100644
--- a/drivers/net/ixgbe/ixgbe_common.c
+++ b/drivers/net/ixgbe/ixgbe_common.c
@@ -454,8 +454,7 @@ s32 ixgbe_stop_adapter_generic(struct ixgbe_hw *hw)
* Prevent the PCI-E bus from from hanging by disabling PCI-E master
* access and verify no pending requests
*/
- if (ixgbe_disable_pcie_master(hw) != 0)
- hw_dbg(hw, "PCI-E Master disable polling has failed.\n");
+ ixgbe_disable_pcie_master(hw);
return 0;
}
@@ -2161,10 +2160,16 @@ out:
**/
s32 ixgbe_disable_pcie_master(struct ixgbe_hw *hw)
{
+ struct ixgbe_adapter *adapter = hw->back;
u32 i;
u32 reg_val;
u32 number_of_queues;
- s32 status = IXGBE_ERR_MASTER_REQUESTS_PENDING;
+ s32 status = 0;
+ u16 dev_status = 0;
+
+ /* Just jump out if bus mastering is already disabled */
+ if (!(IXGBE_READ_REG(hw, IXGBE_STATUS) & IXGBE_STATUS_GIO))
+ goto out;
/* Disable the receive unit by stopping each queue */
number_of_queues = hw->mac.max_rx_queues;
@@ -2181,13 +2186,43 @@ s32 ixgbe_disable_pcie_master(struct ixgbe_hw *hw)
IXGBE_WRITE_REG(hw, IXGBE_CTRL, reg_val);
for (i = 0; i < IXGBE_PCI_MASTER_DISABLE_TIMEOUT; i++) {
- if (!(IXGBE_READ_REG(hw, IXGBE_STATUS) & IXGBE_STATUS_GIO)) {
- status = 0;
+ if (!(IXGBE_READ_REG(hw, IXGBE_STATUS) & IXGBE_STATUS_GIO))
+ goto check_device_status;
+ udelay(100);
+ }
+
+ hw_dbg(hw, "GIO Master Disable bit didn't clear - requesting resets\n");
+ status = IXGBE_ERR_MASTER_REQUESTS_PENDING;
+
+ /*
+ * Before proceeding, make sure that the PCIe block does not have
+ * transactions pending.
+ */
+check_device_status:
+ for (i = 0; i < IXGBE_PCI_MASTER_DISABLE_TIMEOUT; i++) {
+ pci_read_config_word(adapter->pdev, IXGBE_PCI_DEVICE_STATUS,
+ &dev_status);
+ if (!(dev_status & IXGBE_PCI_DEVICE_STATUS_TRANSACTION_PENDING))
break;
- }
udelay(100);
}
+ if (i == IXGBE_PCI_MASTER_DISABLE_TIMEOUT)
+ hw_dbg(hw, "PCIe transaction pending bit also did not clear.\n");
+ else
+ goto out;
+
+ /*
+ * Two consecutive resets are required via CTRL.RST per datasheet
+ * 5.2.5.3.2 Master Disable. We set a flag to inform the reset routine
+ * of this need. The first reset prevents new master requests from
+ * being issued by our device. We then must wait 1usec for any
+ * remaining completions from the PCIe bus to trickle in, and then reset
+ * again to clear out any effects they may have had on our device.
+ */
+ hw->mac.flags |= IXGBE_FLAGS_DOUBLE_RESET_REQUIRED;
+
+out:
return status;
}
diff --git a/drivers/net/ixgbe/ixgbe_type.h b/drivers/net/ixgbe/ixgbe_type.h
index af5ad40..5ede03c 100644
--- a/drivers/net/ixgbe/ixgbe_type.h
+++ b/drivers/net/ixgbe/ixgbe_type.h
@@ -1614,6 +1614,8 @@
#define IXGBE_ALT_SAN_MAC_ADDR_CAPS_ALTWWN 0x1 /* Alt. WWN base exists */
/* PCI Bus Info */
+#define IXGBE_PCI_DEVICE_STATUS 0xAA
+#define IXGBE_PCI_DEVICE_STATUS_TRANSACTION_PENDING 0x0020
#define IXGBE_PCI_LINK_STATUS 0xB2
#define IXGBE_PCI_DEVICE_CONTROL2 0xC8
#define IXGBE_PCI_LINK_WIDTH 0x3F0
@@ -2557,6 +2559,7 @@ struct ixgbe_eeprom_info {
u16 address_bits;
};
+#define IXGBE_FLAGS_DOUBLE_RESET_REQUIRED 0x01
struct ixgbe_mac_info {
struct ixgbe_mac_operations ops;
enum ixgbe_mac_type type;
@@ -2579,6 +2582,7 @@ struct ixgbe_mac_info {
u32 orig_autoc2;
bool orig_link_settings_stored;
bool autotry_restart;
+ u8 flags;
};
struct ixgbe_phy_info {
diff --git a/drivers/net/ixgbe/ixgbe_x540.c b/drivers/net/ixgbe/ixgbe_x540.c
index f2518b0..a6f06d5 100644
--- a/drivers/net/ixgbe/ixgbe_x540.c
+++ b/drivers/net/ixgbe/ixgbe_x540.c
@@ -110,12 +110,9 @@ static s32 ixgbe_reset_hw_X540(struct ixgbe_hw *hw)
* Prevent the PCI-E bus from from hanging by disabling PCI-E master
* access and verify no pending requests before reset
*/
- status = ixgbe_disable_pcie_master(hw);
- if (status != 0) {
- status = IXGBE_ERR_MASTER_REQUESTS_PENDING;
- hw_dbg(hw, "PCI-E Master disable polling has failed.\n");
- }
+ ixgbe_disable_pcie_master(hw);
+mac_reset_top:
/*
* Issue global reset to the MAC. Needs to be SW reset if link is up.
* If link reset is used when link is up, it might reset the PHY when
@@ -148,6 +145,19 @@ static s32 ixgbe_reset_hw_X540(struct ixgbe_hw *hw)
hw_dbg(hw, "Reset polling failed to complete.\n");
}
+ /*
+ * Double resets are required for recovery from certain error
+ * conditions. Between resets, it is necessary to stall to allow time
+ * for any pending HW events to complete. We use 1usec since that is
+ * what is needed for ixgbe_disable_pcie_master(). The second reset
+ * then clears out any effects of those events.
+ */
+ if (hw->mac.flags & IXGBE_FLAGS_DOUBLE_RESET_REQUIRED) {
+ hw->mac.flags &= ~IXGBE_FLAGS_DOUBLE_RESET_REQUIRED;
+ udelay(1);
+ goto mac_reset_top;
+ }
+
/* Clear PF Reset Done bit so PF/VF Mail Ops can work */
ctrl_ext = IXGBE_READ_REG(hw, IXGBE_CTRL_EXT);
ctrl_ext |= IXGBE_CTRL_EXT_PFRSTD;
--
1.7.4
^ permalink raw reply related
* [net-next-2.6 12/24] ixgbe: add polling test to end of PHY reset
From: Jeff Kirsher @ 2011-03-03 13:03 UTC (permalink / raw)
To: davem; +Cc: Emil Tantilov, netdev, gospo, bphilips, Jeff Kirsher
In-Reply-To: <1299157429-15878-1-git-send-email-jeffrey.t.kirsher@intel.com>
From: Emil Tantilov <emil.s.tantilov@intel.com>
Some PHYs require that we poll the reset bit and wait for it to clear
before continuing initialization. As such we should add this check to the
end of the ixgbe_reset_phy_generic routine.
Signed-off-by: Emil Tantilov <emil.s.tantilov@intel.com>
Tested-by: Stephen Ko <stephen.s.ko@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/ixgbe/ixgbe_phy.c | 40 +++++++++++++++++++++++++++++++++++++---
1 files changed, 37 insertions(+), 3 deletions(-)
diff --git a/drivers/net/ixgbe/ixgbe_phy.c b/drivers/net/ixgbe/ixgbe_phy.c
index f8a60ca..ebd6e44 100644
--- a/drivers/net/ixgbe/ixgbe_phy.c
+++ b/drivers/net/ixgbe/ixgbe_phy.c
@@ -138,17 +138,51 @@ static enum ixgbe_phy_type ixgbe_get_phy_type_from_id(u32 phy_id)
**/
s32 ixgbe_reset_phy_generic(struct ixgbe_hw *hw)
{
+ u32 i;
+ u16 ctrl = 0;
+ s32 status = 0;
+
+ if (hw->phy.type == ixgbe_phy_unknown)
+ status = ixgbe_identify_phy_generic(hw);
+
+ if (status != 0 || hw->phy.type == ixgbe_phy_none)
+ goto out;
+
/* Don't reset PHY if it's shut down due to overtemp. */
if (!hw->phy.reset_if_overtemp &&
(IXGBE_ERR_OVERTEMP == hw->phy.ops.check_overtemp(hw)))
- return 0;
+ goto out;
/*
* Perform soft PHY reset to the PHY_XS.
* This will cause a soft reset to the PHY
*/
- return hw->phy.ops.write_reg(hw, MDIO_CTRL1, MDIO_MMD_PHYXS,
- MDIO_CTRL1_RESET);
+ hw->phy.ops.write_reg(hw, MDIO_CTRL1,
+ MDIO_MMD_PHYXS,
+ MDIO_CTRL1_RESET);
+
+ /*
+ * Poll for reset bit to self-clear indicating reset is complete.
+ * Some PHYs could take up to 3 seconds to complete and need about
+ * 1.7 usec delay after the reset is complete.
+ */
+ for (i = 0; i < 30; i++) {
+ msleep(100);
+ hw->phy.ops.read_reg(hw, MDIO_CTRL1,
+ MDIO_MMD_PHYXS, &ctrl);
+ if (!(ctrl & MDIO_CTRL1_RESET)) {
+ udelay(2);
+ break;
+ }
+ }
+
+ if (ctrl & MDIO_CTRL1_RESET) {
+ status = IXGBE_ERR_RESET_FAILED;
+ hw_dbg(hw, "PHY reset polling failed to complete.\n");
+ }
+
+out:
+ return status;
}
/**
--
1.7.4
^ permalink raw reply related
* [net-next-2.6 10/24] ixgbe: cleanup code in ixgbe_identify_sfp_module_generic
From: Jeff Kirsher @ 2011-03-03 13:03 UTC (permalink / raw)
To: davem; +Cc: Emil Tantilov, netdev, gospo, bphilips, Jeff Kirsher
In-Reply-To: <1299157429-15878-1-git-send-email-jeffrey.t.kirsher@intel.com>
From: Emil Tantilov <emil.s.tantilov@intel.com>
This change cleans up several issues in ixgbe_identify_sfp_module_generic
including whitespace, redundant code, I2C EEPROM reads without exception
handling, and an if/elseif/else without braces.
Signed-off-by: Emil Tantilov <emil.s.tantilov@intel.com>
Tested-by: Stephen Ko <stephen.s.ko@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/ixgbe/ixgbe_phy.c | 149 ++++++++++++++++++++++++++---------------
1 files changed, 94 insertions(+), 55 deletions(-)
diff --git a/drivers/net/ixgbe/ixgbe_phy.c b/drivers/net/ixgbe/ixgbe_phy.c
index 8f7123e..f8a60ca 100644
--- a/drivers/net/ixgbe/ixgbe_phy.c
+++ b/drivers/net/ixgbe/ixgbe_phy.c
@@ -556,11 +556,10 @@ out:
}
/**
- * ixgbe_identify_sfp_module_generic - Identifies SFP module and assigns
- * the PHY type.
+ * ixgbe_identify_sfp_module_generic - Identifies SFP modules
* @hw: pointer to hardware structure
*
- * Searches for and indentifies the SFP module. Assings appropriate PHY type.
+ * Searches for and identifies the SFP module and assigns appropriate PHY type.
**/
s32 ixgbe_identify_sfp_module_generic(struct ixgbe_hw *hw)
{
@@ -581,41 +580,62 @@ s32 ixgbe_identify_sfp_module_generic(struct ixgbe_hw *hw)
goto out;
}
- status = hw->phy.ops.read_i2c_eeprom(hw, IXGBE_SFF_IDENTIFIER,
+ status = hw->phy.ops.read_i2c_eeprom(hw,
+ IXGBE_SFF_IDENTIFIER,
&identifier);
- if (status == IXGBE_ERR_SFP_NOT_PRESENT || status == IXGBE_ERR_I2C) {
- status = IXGBE_ERR_SFP_NOT_PRESENT;
- hw->phy.sfp_type = ixgbe_sfp_type_not_present;
- if (hw->phy.type != ixgbe_phy_nl) {
- hw->phy.id = 0;
- hw->phy.type = ixgbe_phy_unknown;
- }
- goto out;
- }
+ if (status == IXGBE_ERR_SWFW_SYNC ||
+ status == IXGBE_ERR_I2C ||
+ status == IXGBE_ERR_SFP_NOT_PRESENT)
+ goto err_read_i2c_eeprom;
- if (identifier == IXGBE_SFF_IDENTIFIER_SFP) {
- hw->phy.ops.read_i2c_eeprom(hw, IXGBE_SFF_1GBE_COMP_CODES,
- &comp_codes_1g);
- hw->phy.ops.read_i2c_eeprom(hw, IXGBE_SFF_10GBE_COMP_CODES,
- &comp_codes_10g);
- hw->phy.ops.read_i2c_eeprom(hw, IXGBE_SFF_CABLE_TECHNOLOGY,
- &cable_tech);
-
- /* ID Module
- * =========
- * 0 SFP_DA_CU
- * 1 SFP_SR
- * 2 SFP_LR
- * 3 SFP_DA_CORE0 - 82599-specific
- * 4 SFP_DA_CORE1 - 82599-specific
- * 5 SFP_SR/LR_CORE0 - 82599-specific
- * 6 SFP_SR/LR_CORE1 - 82599-specific
- * 7 SFP_act_lmt_DA_CORE0 - 82599-specific
- * 8 SFP_act_lmt_DA_CORE1 - 82599-specific
- * 9 SFP_1g_cu_CORE0 - 82599-specific
- * 10 SFP_1g_cu_CORE1 - 82599-specific
- */
+ /* LAN ID is needed for sfp_type determination */
+ hw->mac.ops.set_lan_id(hw);
+
+ if (identifier != IXGBE_SFF_IDENTIFIER_SFP) {
+ hw->phy.type = ixgbe_phy_sfp_unsupported;
+ status = IXGBE_ERR_SFP_NOT_SUPPORTED;
+ } else {
+ status = hw->phy.ops.read_i2c_eeprom(hw,
+ IXGBE_SFF_1GBE_COMP_CODES,
+ &comp_codes_1g);
+
+ if (status == IXGBE_ERR_SWFW_SYNC ||
+ status == IXGBE_ERR_I2C ||
+ status == IXGBE_ERR_SFP_NOT_PRESENT)
+ goto err_read_i2c_eeprom;
+
+ status = hw->phy.ops.read_i2c_eeprom(hw,
+ IXGBE_SFF_10GBE_COMP_CODES,
+ &comp_codes_10g);
+
+ if (status == IXGBE_ERR_SWFW_SYNC ||
+ status == IXGBE_ERR_I2C ||
+ status == IXGBE_ERR_SFP_NOT_PRESENT)
+ goto err_read_i2c_eeprom;
+ status = hw->phy.ops.read_i2c_eeprom(hw,
+ IXGBE_SFF_CABLE_TECHNOLOGY,
+ &cable_tech);
+
+ if (status == IXGBE_ERR_SWFW_SYNC ||
+ status == IXGBE_ERR_I2C ||
+ status == IXGBE_ERR_SFP_NOT_PRESENT)
+ goto err_read_i2c_eeprom;
+
+ /* ID Module
+ * =========
+ * 0 SFP_DA_CU
+ * 1 SFP_SR
+ * 2 SFP_LR
+ * 3 SFP_DA_CORE0 - 82599-specific
+ * 4 SFP_DA_CORE1 - 82599-specific
+ * 5 SFP_SR/LR_CORE0 - 82599-specific
+ * 6 SFP_SR/LR_CORE1 - 82599-specific
+ * 7 SFP_act_lmt_DA_CORE0 - 82599-specific
+ * 8 SFP_act_lmt_DA_CORE1 - 82599-specific
+ * 9 SFP_1g_cu_CORE0 - 82599-specific
+ * 10 SFP_1g_cu_CORE1 - 82599-specific
+ */
if (hw->mac.type == ixgbe_mac_82598EB) {
if (cable_tech & IXGBE_SFF_DA_PASSIVE_CABLE)
hw->phy.sfp_type = ixgbe_sfp_type_da_cu;
@@ -647,31 +667,27 @@ s32 ixgbe_identify_sfp_module_generic(struct ixgbe_hw *hw)
ixgbe_sfp_type_da_act_lmt_core1;
} else {
hw->phy.sfp_type =
- ixgbe_sfp_type_unknown;
+ ixgbe_sfp_type_unknown;
}
- } else if (comp_codes_10g & IXGBE_SFF_10GBASESR_CAPABLE)
- if (hw->bus.lan_id == 0)
- hw->phy.sfp_type =
- ixgbe_sfp_type_srlr_core0;
- else
- hw->phy.sfp_type =
- ixgbe_sfp_type_srlr_core1;
- else if (comp_codes_10g & IXGBE_SFF_10GBASELR_CAPABLE)
+ } else if (comp_codes_10g &
+ (IXGBE_SFF_10GBASESR_CAPABLE |
+ IXGBE_SFF_10GBASELR_CAPABLE)) {
if (hw->bus.lan_id == 0)
hw->phy.sfp_type =
ixgbe_sfp_type_srlr_core0;
else
hw->phy.sfp_type =
ixgbe_sfp_type_srlr_core1;
- else if (comp_codes_1g & IXGBE_SFF_1GBASET_CAPABLE)
+ } else if (comp_codes_1g & IXGBE_SFF_1GBASET_CAPABLE) {
if (hw->bus.lan_id == 0)
hw->phy.sfp_type =
ixgbe_sfp_type_1g_cu_core0;
else
hw->phy.sfp_type =
ixgbe_sfp_type_1g_cu_core1;
- else
+ } else {
hw->phy.sfp_type = ixgbe_sfp_type_unknown;
+ }
}
if (hw->phy.sfp_type != stored_sfp_type)
@@ -688,16 +704,33 @@ s32 ixgbe_identify_sfp_module_generic(struct ixgbe_hw *hw)
/* Determine PHY vendor */
if (hw->phy.type != ixgbe_phy_nl) {
hw->phy.id = identifier;
- hw->phy.ops.read_i2c_eeprom(hw,
+ status = hw->phy.ops.read_i2c_eeprom(hw,
IXGBE_SFF_VENDOR_OUI_BYTE0,
&oui_bytes[0]);
- hw->phy.ops.read_i2c_eeprom(hw,
+
+ if (status == IXGBE_ERR_SWFW_SYNC ||
+ status == IXGBE_ERR_I2C ||
+ status == IXGBE_ERR_SFP_NOT_PRESENT)
+ goto err_read_i2c_eeprom;
+
+ status = hw->phy.ops.read_i2c_eeprom(hw,
IXGBE_SFF_VENDOR_OUI_BYTE1,
&oui_bytes[1]);
- hw->phy.ops.read_i2c_eeprom(hw,
+
+ if (status == IXGBE_ERR_SWFW_SYNC ||
+ status == IXGBE_ERR_I2C ||
+ status == IXGBE_ERR_SFP_NOT_PRESENT)
+ goto err_read_i2c_eeprom;
+
+ status = hw->phy.ops.read_i2c_eeprom(hw,
IXGBE_SFF_VENDOR_OUI_BYTE2,
&oui_bytes[2]);
+ if (status == IXGBE_ERR_SWFW_SYNC ||
+ status == IXGBE_ERR_I2C ||
+ status == IXGBE_ERR_SFP_NOT_PRESENT)
+ goto err_read_i2c_eeprom;
+
vendor_oui =
((oui_bytes[0] << IXGBE_SFF_VENDOR_OUI_BYTE0_SHIFT) |
(oui_bytes[1] << IXGBE_SFF_VENDOR_OUI_BYTE1_SHIFT) |
@@ -707,7 +740,7 @@ s32 ixgbe_identify_sfp_module_generic(struct ixgbe_hw *hw)
case IXGBE_SFF_VENDOR_OUI_TYCO:
if (cable_tech & IXGBE_SFF_DA_PASSIVE_CABLE)
hw->phy.type =
- ixgbe_phy_sfp_passive_tyco;
+ ixgbe_phy_sfp_passive_tyco;
break;
case IXGBE_SFF_VENDOR_OUI_FTL:
if (cable_tech & IXGBE_SFF_DA_ACTIVE_CABLE)
@@ -724,7 +757,7 @@ s32 ixgbe_identify_sfp_module_generic(struct ixgbe_hw *hw)
default:
if (cable_tech & IXGBE_SFF_DA_PASSIVE_CABLE)
hw->phy.type =
- ixgbe_phy_sfp_passive_unknown;
+ ixgbe_phy_sfp_passive_unknown;
else if (cable_tech & IXGBE_SFF_DA_ACTIVE_CABLE)
hw->phy.type =
ixgbe_phy_sfp_active_unknown;
@@ -734,7 +767,7 @@ s32 ixgbe_identify_sfp_module_generic(struct ixgbe_hw *hw)
}
}
- /* All passive DA cables are supported */
+ /* Allow any DA cable vendor */
if (cable_tech & (IXGBE_SFF_DA_PASSIVE_CABLE |
IXGBE_SFF_DA_ACTIVE_CABLE)) {
status = 0;
@@ -776,12 +809,18 @@ s32 ixgbe_identify_sfp_module_generic(struct ixgbe_hw *hw)
out:
return status;
+
+err_read_i2c_eeprom:
+ hw->phy.sfp_type = ixgbe_sfp_type_not_present;
+ if (hw->phy.type != ixgbe_phy_nl) {
+ hw->phy.id = 0;
+ hw->phy.type = ixgbe_phy_unknown;
+ }
+ return IXGBE_ERR_SFP_NOT_PRESENT;
}
/**
- * ixgbe_get_sfp_init_sequence_offsets - Checks the MAC's EEPROM to see
- * if it supports a given SFP+ module type, if so it returns the offsets to the
- * phy init sequence block.
+ * ixgbe_get_sfp_init_sequence_offsets - Provides offset of PHY init sequence
* @hw: pointer to hardware structure
* @list_offset: offset to the SFP ID list
* @data_offset: offset to the SFP data block
--
1.7.4
^ permalink raw reply related
* [net-next-2.6 11/24] ixgbe: Check link wants report current link state
From: Jeff Kirsher @ 2011-03-03 13:03 UTC (permalink / raw)
To: davem; +Cc: Emil Tantilov, netdev, gospo, bphilips, Jeff Kirsher
In-Reply-To: <1299157429-15878-1-git-send-email-jeffrey.t.kirsher@intel.com>
From: Emil Tantilov <emil.s.tantilov@intel.com>
Currently check link reports the link state as down, if at any time
the link had previously gone down since the last time the LINKS
register was read. This does not accurately reflect the function of
the check link call, which should be to return the CURRENT link
state. Code now reads the LINKS registers twice, once to clear the
previous and again to get the current value.
Signed-off-by: Emil Tantilov <emil.s.tantilov@intel.com>
Tested-by: Stephen Ko <stephen.s.ko@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/ixgbe/ixgbe_common.c | 11 ++++++++++-
1 files changed, 10 insertions(+), 1 deletions(-)
diff --git a/drivers/net/ixgbe/ixgbe_common.c b/drivers/net/ixgbe/ixgbe_common.c
index 6d87c74..7e3bb55 100644
--- a/drivers/net/ixgbe/ixgbe_common.c
+++ b/drivers/net/ixgbe/ixgbe_common.c
@@ -2769,10 +2769,19 @@ s32 ixgbe_clear_vfta_generic(struct ixgbe_hw *hw)
s32 ixgbe_check_mac_link_generic(struct ixgbe_hw *hw, ixgbe_link_speed *speed,
bool *link_up, bool link_up_wait_to_complete)
{
- u32 links_reg;
+ u32 links_reg, links_orig;
u32 i;
+ /* clear the old state */
+ links_orig = IXGBE_READ_REG(hw, IXGBE_LINKS);
+
links_reg = IXGBE_READ_REG(hw, IXGBE_LINKS);
+
+ if (links_orig != links_reg) {
+ hw_dbg(hw, "LINKS changed from %08X to %08X\n",
+ links_orig, links_reg);
+ }
+
if (link_up_wait_to_complete) {
for (i = 0; i < IXGBE_LINK_UP_TIME; i++) {
if (links_reg & IXGBE_LINKS_UP) {
--
1.7.4
^ permalink raw reply related
* [net-next-2.6 13/24] ixgbe: Fill out PCIe speed and width enums with values
From: Jeff Kirsher @ 2011-03-03 13:03 UTC (permalink / raw)
To: davem; +Cc: Emil Tantilov, netdev, gospo, bphilips, Jeff Kirsher
In-Reply-To: <1299157429-15878-1-git-send-email-jeffrey.t.kirsher@intel.com>
From: Emil Tantilov <emil.s.tantilov@intel.com>
This patch fills in the values for bus speed and width of the
ixgbe_bus_speed and ixgbe_bus_width enums.
Signed-off-by: Emil Tantilov <emil.s.tantilov@intel.com>
Tested-by: Stephen Ko <stephen.s.ko@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/ixgbe/ixgbe_type.h | 22 +++++++++++-----------
1 files changed, 11 insertions(+), 11 deletions(-)
diff --git a/drivers/net/ixgbe/ixgbe_type.h b/drivers/net/ixgbe/ixgbe_type.h
index 5ede03c..a737b13 100644
--- a/drivers/net/ixgbe/ixgbe_type.h
+++ b/drivers/net/ixgbe/ixgbe_type.h
@@ -2333,25 +2333,25 @@ enum ixgbe_bus_type {
/* PCI bus speeds */
enum ixgbe_bus_speed {
ixgbe_bus_speed_unknown = 0,
- ixgbe_bus_speed_33,
- ixgbe_bus_speed_66,
- ixgbe_bus_speed_100,
- ixgbe_bus_speed_120,
- ixgbe_bus_speed_133,
- ixgbe_bus_speed_2500,
- ixgbe_bus_speed_5000,
+ ixgbe_bus_speed_33 = 33,
+ ixgbe_bus_speed_66 = 66,
+ ixgbe_bus_speed_100 = 100,
+ ixgbe_bus_speed_120 = 120,
+ ixgbe_bus_speed_133 = 133,
+ ixgbe_bus_speed_2500 = 2500,
+ ixgbe_bus_speed_5000 = 5000,
ixgbe_bus_speed_reserved
};
/* PCI bus widths */
enum ixgbe_bus_width {
ixgbe_bus_width_unknown = 0,
- ixgbe_bus_width_pcie_x1,
- ixgbe_bus_width_pcie_x2,
+ ixgbe_bus_width_pcie_x1 = 1,
+ ixgbe_bus_width_pcie_x2 = 2,
ixgbe_bus_width_pcie_x4 = 4,
ixgbe_bus_width_pcie_x8 = 8,
- ixgbe_bus_width_32,
- ixgbe_bus_width_64,
+ ixgbe_bus_width_32 = 32,
+ ixgbe_bus_width_64 = 64,
ixgbe_bus_width_reserved
};
--
1.7.4
^ permalink raw reply related
* [net-next-2.6 14/24] ixgbe: Bounds checking for set_rar, clear_rar, set_vmdq, clear_vmdq
From: Jeff Kirsher @ 2011-03-03 13:03 UTC (permalink / raw)
To: davem; +Cc: Emil Tantilov, netdev, gospo, bphilips, Jeff Kirsher
In-Reply-To: <1299157429-15878-1-git-send-email-jeffrey.t.kirsher@intel.com>
From: Emil Tantilov <emil.s.tantilov@intel.com>
This change makes it so that out of bounds requests to these calls will
now return IXGBE_ERR_INVALID_ARGUMENT instead of returning 0.
Signed-off-by: Emil Tantilov <emil.s.tantilov@intel.com>
Tested-by: Stephen Ko <stephen.s.ko@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/ixgbe/ixgbe_82598.c | 24 ++++--
drivers/net/ixgbe/ixgbe_common.c | 153 +++++++++++++++++++-------------------
drivers/net/ixgbe/ixgbe_type.h | 1 -
3 files changed, 95 insertions(+), 83 deletions(-)
diff --git a/drivers/net/ixgbe/ixgbe_82598.c b/drivers/net/ixgbe/ixgbe_82598.c
index 291b1e6..bff7011 100644
--- a/drivers/net/ixgbe/ixgbe_82598.c
+++ b/drivers/net/ixgbe/ixgbe_82598.c
@@ -858,6 +858,13 @@ reset_hw_out:
static s32 ixgbe_set_vmdq_82598(struct ixgbe_hw *hw, u32 rar, u32 vmdq)
{
u32 rar_high;
+ u32 rar_entries = hw->mac.num_rar_entries;
+
+ /* Make sure we are using a valid rar index range */
+ if (rar >= rar_entries) {
+ hw_dbg(hw, "RAR index %d is out of range.\n", rar);
+ return IXGBE_ERR_INVALID_ARGUMENT;
+ }
rar_high = IXGBE_READ_REG(hw, IXGBE_RAH(rar));
rar_high &= ~IXGBE_RAH_VIND_MASK;
@@ -877,14 +884,17 @@ static s32 ixgbe_clear_vmdq_82598(struct ixgbe_hw *hw, u32 rar, u32 vmdq)
u32 rar_high;
u32 rar_entries = hw->mac.num_rar_entries;
- if (rar < rar_entries) {
- rar_high = IXGBE_READ_REG(hw, IXGBE_RAH(rar));
- if (rar_high & IXGBE_RAH_VIND_MASK) {
- rar_high &= ~IXGBE_RAH_VIND_MASK;
- IXGBE_WRITE_REG(hw, IXGBE_RAH(rar), rar_high);
- }
- } else {
+
+ /* Make sure we are using a valid rar index range */
+ if (rar >= rar_entries) {
hw_dbg(hw, "RAR index %d is out of range.\n", rar);
+ return IXGBE_ERR_INVALID_ARGUMENT;
+ }
+
+ rar_high = IXGBE_READ_REG(hw, IXGBE_RAH(rar));
+ if (rar_high & IXGBE_RAH_VIND_MASK) {
+ rar_high &= ~IXGBE_RAH_VIND_MASK;
+ IXGBE_WRITE_REG(hw, IXGBE_RAH(rar), rar_high);
}
return 0;
diff --git a/drivers/net/ixgbe/ixgbe_common.c b/drivers/net/ixgbe/ixgbe_common.c
index 7e3bb55..882a350 100644
--- a/drivers/net/ixgbe/ixgbe_common.c
+++ b/drivers/net/ixgbe/ixgbe_common.c
@@ -1239,37 +1239,37 @@ s32 ixgbe_set_rar_generic(struct ixgbe_hw *hw, u32 index, u8 *addr, u32 vmdq,
u32 rar_low, rar_high;
u32 rar_entries = hw->mac.num_rar_entries;
+ /* Make sure we are using a valid rar index range */
+ if (index >= rar_entries) {
+ hw_dbg(hw, "RAR index %d is out of range.\n", index);
+ return IXGBE_ERR_INVALID_ARGUMENT;
+ }
+
/* setup VMDq pool selection before this RAR gets enabled */
hw->mac.ops.set_vmdq(hw, index, vmdq);
- /* Make sure we are using a valid rar index range */
- if (index < rar_entries) {
- /*
- * HW expects these in little endian so we reverse the byte
- * order from network order (big endian) to little endian
- */
- rar_low = ((u32)addr[0] |
- ((u32)addr[1] << 8) |
- ((u32)addr[2] << 16) |
- ((u32)addr[3] << 24));
- /*
- * Some parts put the VMDq setting in the extra RAH bits,
- * so save everything except the lower 16 bits that hold part
- * of the address and the address valid bit.
- */
- rar_high = IXGBE_READ_REG(hw, IXGBE_RAH(index));
- rar_high &= ~(0x0000FFFF | IXGBE_RAH_AV);
- rar_high |= ((u32)addr[4] | ((u32)addr[5] << 8));
+ /*
+ * HW expects these in little endian so we reverse the byte
+ * order from network order (big endian) to little endian
+ */
+ rar_low = ((u32)addr[0] |
+ ((u32)addr[1] << 8) |
+ ((u32)addr[2] << 16) |
+ ((u32)addr[3] << 24));
+ /*
+ * Some parts put the VMDq setting in the extra RAH bits,
+ * so save everything except the lower 16 bits that hold part
+ * of the address and the address valid bit.
+ */
+ rar_high = IXGBE_READ_REG(hw, IXGBE_RAH(index));
+ rar_high &= ~(0x0000FFFF | IXGBE_RAH_AV);
+ rar_high |= ((u32)addr[4] | ((u32)addr[5] << 8));
- if (enable_addr != 0)
- rar_high |= IXGBE_RAH_AV;
+ if (enable_addr != 0)
+ rar_high |= IXGBE_RAH_AV;
- IXGBE_WRITE_REG(hw, IXGBE_RAL(index), rar_low);
- IXGBE_WRITE_REG(hw, IXGBE_RAH(index), rar_high);
- } else {
- hw_dbg(hw, "RAR index %d is out of range.\n", index);
- return IXGBE_ERR_RAR_INDEX;
- }
+ IXGBE_WRITE_REG(hw, IXGBE_RAL(index), rar_low);
+ IXGBE_WRITE_REG(hw, IXGBE_RAH(index), rar_high);
return 0;
}
@@ -1287,22 +1287,22 @@ s32 ixgbe_clear_rar_generic(struct ixgbe_hw *hw, u32 index)
u32 rar_entries = hw->mac.num_rar_entries;
/* Make sure we are using a valid rar index range */
- if (index < rar_entries) {
- /*
- * Some parts put the VMDq setting in the extra RAH bits,
- * so save everything except the lower 16 bits that hold part
- * of the address and the address valid bit.
- */
- rar_high = IXGBE_READ_REG(hw, IXGBE_RAH(index));
- rar_high &= ~(0x0000FFFF | IXGBE_RAH_AV);
-
- IXGBE_WRITE_REG(hw, IXGBE_RAL(index), 0);
- IXGBE_WRITE_REG(hw, IXGBE_RAH(index), rar_high);
- } else {
+ if (index >= rar_entries) {
hw_dbg(hw, "RAR index %d is out of range.\n", index);
- return IXGBE_ERR_RAR_INDEX;
+ return IXGBE_ERR_INVALID_ARGUMENT;
}
+ /*
+ * Some parts put the VMDq setting in the extra RAH bits,
+ * so save everything except the lower 16 bits that hold part
+ * of the address and the address valid bit.
+ */
+ rar_high = IXGBE_READ_REG(hw, IXGBE_RAH(index));
+ rar_high &= ~(0x0000FFFF | IXGBE_RAH_AV);
+
+ IXGBE_WRITE_REG(hw, IXGBE_RAL(index), 0);
+ IXGBE_WRITE_REG(hw, IXGBE_RAH(index), rar_high);
+
/* clear VMDq pool/queue selection for this RAR */
hw->mac.ops.clear_vmdq(hw, index, IXGBE_CLEAR_VMDQ_ALL);
@@ -2468,37 +2468,38 @@ s32 ixgbe_clear_vmdq_generic(struct ixgbe_hw *hw, u32 rar, u32 vmdq)
u32 mpsar_lo, mpsar_hi;
u32 rar_entries = hw->mac.num_rar_entries;
- if (rar < rar_entries) {
- mpsar_lo = IXGBE_READ_REG(hw, IXGBE_MPSAR_LO(rar));
- mpsar_hi = IXGBE_READ_REG(hw, IXGBE_MPSAR_HI(rar));
+ /* Make sure we are using a valid rar index range */
+ if (rar >= rar_entries) {
+ hw_dbg(hw, "RAR index %d is out of range.\n", rar);
+ return IXGBE_ERR_INVALID_ARGUMENT;
+ }
- if (!mpsar_lo && !mpsar_hi)
- goto done;
+ mpsar_lo = IXGBE_READ_REG(hw, IXGBE_MPSAR_LO(rar));
+ mpsar_hi = IXGBE_READ_REG(hw, IXGBE_MPSAR_HI(rar));
- if (vmdq == IXGBE_CLEAR_VMDQ_ALL) {
- if (mpsar_lo) {
- IXGBE_WRITE_REG(hw, IXGBE_MPSAR_LO(rar), 0);
- mpsar_lo = 0;
- }
- if (mpsar_hi) {
- IXGBE_WRITE_REG(hw, IXGBE_MPSAR_HI(rar), 0);
- mpsar_hi = 0;
- }
- } else if (vmdq < 32) {
- mpsar_lo &= ~(1 << vmdq);
- IXGBE_WRITE_REG(hw, IXGBE_MPSAR_LO(rar), mpsar_lo);
- } else {
- mpsar_hi &= ~(1 << (vmdq - 32));
- IXGBE_WRITE_REG(hw, IXGBE_MPSAR_HI(rar), mpsar_hi);
- }
+ if (!mpsar_lo && !mpsar_hi)
+ goto done;
- /* was that the last pool using this rar? */
- if (mpsar_lo == 0 && mpsar_hi == 0 && rar != 0)
- hw->mac.ops.clear_rar(hw, rar);
+ if (vmdq == IXGBE_CLEAR_VMDQ_ALL) {
+ if (mpsar_lo) {
+ IXGBE_WRITE_REG(hw, IXGBE_MPSAR_LO(rar), 0);
+ mpsar_lo = 0;
+ }
+ if (mpsar_hi) {
+ IXGBE_WRITE_REG(hw, IXGBE_MPSAR_HI(rar), 0);
+ mpsar_hi = 0;
+ }
+ } else if (vmdq < 32) {
+ mpsar_lo &= ~(1 << vmdq);
+ IXGBE_WRITE_REG(hw, IXGBE_MPSAR_LO(rar), mpsar_lo);
} else {
- hw_dbg(hw, "RAR index %d is out of range.\n", rar);
+ mpsar_hi &= ~(1 << (vmdq - 32));
+ IXGBE_WRITE_REG(hw, IXGBE_MPSAR_HI(rar), mpsar_hi);
}
+ /* was that the last pool using this rar? */
+ if (mpsar_lo == 0 && mpsar_hi == 0 && rar != 0)
+ hw->mac.ops.clear_rar(hw, rar);
done:
return 0;
}
@@ -2514,18 +2515,20 @@ s32 ixgbe_set_vmdq_generic(struct ixgbe_hw *hw, u32 rar, u32 vmdq)
u32 mpsar;
u32 rar_entries = hw->mac.num_rar_entries;
- if (rar < rar_entries) {
- if (vmdq < 32) {
- mpsar = IXGBE_READ_REG(hw, IXGBE_MPSAR_LO(rar));
- mpsar |= 1 << vmdq;
- IXGBE_WRITE_REG(hw, IXGBE_MPSAR_LO(rar), mpsar);
- } else {
- mpsar = IXGBE_READ_REG(hw, IXGBE_MPSAR_HI(rar));
- mpsar |= 1 << (vmdq - 32);
- IXGBE_WRITE_REG(hw, IXGBE_MPSAR_HI(rar), mpsar);
- }
- } else {
+ /* Make sure we are using a valid rar index range */
+ if (rar >= rar_entries) {
hw_dbg(hw, "RAR index %d is out of range.\n", rar);
+ return IXGBE_ERR_INVALID_ARGUMENT;
+ }
+
+ if (vmdq < 32) {
+ mpsar = IXGBE_READ_REG(hw, IXGBE_MPSAR_LO(rar));
+ mpsar |= 1 << vmdq;
+ IXGBE_WRITE_REG(hw, IXGBE_MPSAR_LO(rar), mpsar);
+ } else {
+ mpsar = IXGBE_READ_REG(hw, IXGBE_MPSAR_HI(rar));
+ mpsar |= 1 << (vmdq - 32);
+ IXGBE_WRITE_REG(hw, IXGBE_MPSAR_HI(rar), mpsar);
}
return 0;
}
diff --git a/drivers/net/ixgbe/ixgbe_type.h b/drivers/net/ixgbe/ixgbe_type.h
index a737b13..b1ae185 100644
--- a/drivers/net/ixgbe/ixgbe_type.h
+++ b/drivers/net/ixgbe/ixgbe_type.h
@@ -2689,7 +2689,6 @@ struct ixgbe_info {
#define IXGBE_ERR_EEPROM_VERSION -24
#define IXGBE_ERR_NO_SPACE -25
#define IXGBE_ERR_OVERTEMP -26
-#define IXGBE_ERR_RAR_INDEX -27
#define IXGBE_ERR_SFP_SETUP_NOT_COMPLETE -30
#define IXGBE_ERR_PBA_SECTION -31
#define IXGBE_ERR_INVALID_ARGUMENT -32
--
1.7.4
^ permalink raw reply related
* [net-next-2.6 15/24] ixgbe: cleanup X540 PHY reset function pointer
From: Jeff Kirsher @ 2011-03-03 13:03 UTC (permalink / raw)
To: davem; +Cc: Don Skidmore, netdev, gospo, bphilips, Jeff Kirsher
In-Reply-To: <1299157429-15878-1-git-send-email-jeffrey.t.kirsher@intel.com>
From: Don Skidmore <donald.c.skidmore@intel.com>
The X540 PHY reset pointer isn't currently used which is a good thing as it
wouldn't work as implemented. On top of that the X540 firmware is written
with the assumption that is does not need to be reset for proper
initialization so it's not needed. I'm just assigning the pointer at NULL
as the current implementation is rather misleading.
Signed-off-by: Don Skidmore <donald.c.skidmore@intel.com>
Tested-by: Stephen Ko <stephen.s.ko@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/ixgbe/ixgbe_x540.c | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/drivers/net/ixgbe/ixgbe_x540.c b/drivers/net/ixgbe/ixgbe_x540.c
index a6f06d5..3229398 100644
--- a/drivers/net/ixgbe/ixgbe_x540.c
+++ b/drivers/net/ixgbe/ixgbe_x540.c
@@ -712,7 +712,7 @@ static struct ixgbe_phy_operations phy_ops_X540 = {
.identify = &ixgbe_identify_phy_generic,
.identify_sfp = &ixgbe_identify_sfp_module_generic,
.init = NULL,
- .reset = &ixgbe_reset_phy_generic,
+ .reset = NULL,
.read_reg = &ixgbe_read_phy_reg_generic,
.write_reg = &ixgbe_write_phy_reg_generic,
.setup_link = &ixgbe_setup_phy_link_generic,
--
1.7.4
^ 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