* Re: Netlink mmap tx security?
From: Daniel Borkmann @ 2014-10-16 6:45 UTC (permalink / raw)
To: David Miller; +Cc: luto, torvalds, kaber, netdev, tgraf
In-Reply-To: <20141014.220908.123550384430402000.davem@davemloft.net>
On 10/15/2014 04:09 AM, David Miller wrote:
> From: Andy Lutomirski <luto@amacapital.net>
> Date: Tue, 14 Oct 2014 19:03:11 -0700
>
>> On Tue, Oct 14, 2014 at 7:01 PM, David Miller <davem@davemloft.net> wrote:
>>> I really think this means I'll have to remove all of the netlink
>>> mmap() support in order to prevent from breaking applications. :(
>>>
>>> The other option is to keep NETLINK_TX_RING, but copy the data into
>>> a kernel side buffer before acting upon it.
>>
>> Option 3, which sucks but maybe not that badly: change the value of
>> NETLINK_RX_RING. (Practically: add NETLINK_RX_RING2 or something like
>> that.)
>
> That would work as well.
>
> There are pros and cons to all of these approaches.
>
> I was thinking that if we do the "TX mmap --> copy to kernel buffer"
> approach, then if in the future we find a way to make it work
> reliably, we can avoid the copy. And frankly performance wise it's no
> worse than what happens via normal sendmsg() calls.
>
> And all applications using NETLINK_RX_RING keep working and keep
> getting the performance boost.
That would be better, yes. This would avoid having such a TPACKET_V*
API chaos we have in packet sockets if this could be fixed for netlink
eventually.
^ permalink raw reply
* Re: tcpdump's capture filter: "vlan" doesn't match
From: Daniel Borkmann @ 2014-10-16 6:10 UTC (permalink / raw)
To: Lukas Tribus
Cc: netdev@vger.kernel.org, John Fastabend, Michał Mirosław,
Jiri Pirko, Ben Hutchings, Atzm Watanabe, Patrick McHardy,
Jesse Gross
In-Reply-To: <DUB123-W23D5ECAD4869F9A799E1F6EDAA0@phx.gbl>
On 10/16/2014 12:58 AM, Lukas Tribus wrote:
> Hi,
>
>
>
> since 2.6.39 (including -rc1), tcpdump "vlan" capture filters don't match
> anymore. All 2.6.38 and older kernels are fine.
>
>
> I reproduced this specifically on a r8169 NIC on 2.6.39-rc1, but I found
> this problem initially on bnx2 and e1000e nics.
>
>
> Howto reproduce: just tcpdump with a "not vlan", "vlan" or "vlan <vlanid>"
> capture filter on a passive eth interface (dot1q/vlan/ip config not necessary).
>
> Actual behavior is that a "vlan [vlanid]" capture filter doesn't match the
> (tagged) packet, and a "not vlan" capture filter matches everything.
>
>
> Disabling rx-vlan-offloading via
> ethtool -K eth0 rxvlan off
>
> doesn't change anything.
>
>
> Here we are filtering for "not vlan" and we can see that the matched frame
> is vlan tagged:
>
> # tcpdump -Uenc1 not vlan
> tcpdump: verbose output suppressed, use -v or -vv for full protocol decode
> listening on eth0, link-type EN10MB (Ethernet), capture size 65535 bytes
> 22:03:39.077584 70:ca:9b:01:23:34> 00:18:f8:01:23:34, \
> *ethertype 802.1Q (0x8100), length 70: vlan 7, p 0*, ethertype IPv4, \
> 192.168.47.9.443> 192.168.32.30.39436: Flags [.], ack 255248912, \
> [...]
> 1 packet captured
> 169 packets received by filter
> 0 packets dropped by kernel
> 59 packets dropped by interface
> #
>
>
>
>
> As suggested here [1], we can pipe everything through another tcpdump
> instance:
> tcpdump -Uw - | tcpdump -en -r - vlan <vlanid>
>
>
> But that is not something that works for my specific use-case (dedicated
> sniffer box, dedicated interface connected to a Cisco SPAN/mirror port,
> un/single/double-tagged packets, remotely accessible via remote-pcap [2]).
>
>
> The sniffer should also be able to:
> - maintain the frame as-is, including dot1q, dot1p (preferably
> without artificial recreation of header fields/values and including CFI/DEI)
> - "direct" capture filter based on vlan (not through multiple userspace
> instances)
>
> Kernel <= 2.6.38 perfectly satisfies those requirements.
>
>
> Isn't disabling rx-vlan-offloading supposed to remedy those problems?
There were some discussions on this in the past e.g. [1]. We have
SKF_AD_VLAN_TAG and SKF_AD_VLAN_TAG_PRESENT for the BPF filter on
this, but libpcap is currently not making use of any of them.
[1] http://thread.gmane.org/gmane.linux.network/247947
> Thanks,
>
> Lukas
>
>
>
>
> [1] https://bugzilla.redhat.com/show_bug.cgi?id=498981
> [2] https://github.com/frgtn/rpcapd-linux
>
> --
> 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: Netlink mmap tx security?
From: Thomas Graf @ 2014-10-16 5:52 UTC (permalink / raw)
To: David Miller; +Cc: dborkman, luto, torvalds, kaber, netdev
In-Reply-To: <20141015.195737.1429281929513331763.davem@davemloft.net>
On 10/15/14 at 07:57pm, David Miller wrote:
> From: Daniel Borkmann <dborkman@redhat.com>
> Date: Thu, 16 Oct 2014 01:45:22 +0200
>
> > On 10/15/2014 04:01 AM, David Miller wrote:
> >> From: Andy Lutomirski <luto@amacapital.net>
> >> Date: Tue, 14 Oct 2014 15:16:46 -0700
> >>
> >>> It's at least remotely possible that there's something that assumes
> >>> that assumes that the availability of NETLINK_RX_RING implies
> >>> NETLINK_TX_RING, which would be unfortunate.
> >>
> >> I already found one such case, nlmon :-/
> >
> > Hmm, can you elaborate? I currently don't think that nlmon cares
> > actually.
>
> nlmon cares, openvswitch cares, etc:
>
> http://openvswitch.org/pipermail/dev/2013-December/034496.html
(Fortunately) the OVS patch has not been merged yet because the number
of Netlink sockets created per vport in the current architecture
currently make it a non scalable approach.
I think introdcing a NETLINK_RX_RING2 and having NETLINK_RX_RING fail
is not a bad way out of this.
^ permalink raw reply
* RE: [PATCH] net: use hardware buffer pool to allocate skb
From: Jiafei.Pan @ 2014-10-16 5:15 UTC (permalink / raw)
To: Eric Dumazet
Cc: David Miller, jkosina@suse.cz, netdev@vger.kernel.org,
LeoLi@freescale.com, linux-doc@vger.kernel.org,
Jiafei.Pan@freescale.com
In-Reply-To: <1413432912.28798.7.camel@edumazet-glaptop2.roam.corp.google.com>
> -----Original Message-----
> From: Eric Dumazet [mailto:eric.dumazet@gmail.com]
> Sent: Thursday, October 16, 2014 12:15 PM
> To: Pan Jiafei-B37022
> Cc: David Miller; jkosina@suse.cz; netdev@vger.kernel.org; Li Yang-Leo-R58472;
> linux-doc@vger.kernel.org
> Subject: Re: [PATCH] net: use hardware buffer pool to allocate skb
>
> On Thu, 2014-10-16 at 02:17 +0000, Jiafei.Pan@freescale.com wrote:
>
> > Thanks for your comments and suggestion. In my case, I want to build skb
> > from hardware block specified memory, I only can see two ways, one is modified
> > net card driver replace common skb allocation function with my specially
> > functions, another way is to hack common skb allocation function in which
> > redirect to my specially functions. My patch is just for the second way.
> > Except these two ways, would you please give me some advice for some other
> > ways for my case? Thanks
>
> I suggest you read drivers/net/ethernet numerous examples.
>
> No need to change anything in net/* or include/*, really.
>
> For a start, look at drivers/net/ethernet/intel/igb/igb_main.c
>
> Mentioning 'hack' in your mails simply should hint you are doing
> something very wrong.
>
> What makes you think your hardware is so special ?
>
In fact, I am developing a bridge driver, it can bridge between any other the
third party net card and my own net card. My target is to let any other the
third party net card can directly use my own net card specified buffer, then
there will be no memory copy in the whole bridge process.
By the way, I don’t see any similar between igb_main.c and my case. And also
My bridge also can’t implemented with "skb frag" in order to aim at zero memory
copy.
^ permalink raw reply
* Re: [Patch net 3/3] ipv4: clean up cookie_v4_check()
From: Eric Dumazet @ 2014-10-16 4:17 UTC (permalink / raw)
To: Cong Wang; +Cc: netdev, davem, Krzysztof Kolasa, Eric Dumazet, Cong Wang
In-Reply-To: <1413408802-21052-3-git-send-email-xiyou.wangcong@gmail.com>
On Wed, 2014-10-15 at 14:33 -0700, Cong Wang wrote:
> From: Cong Wang <cwang@twopensource.com>
>
> We can retrieve opt from skb, no need to pass it as a parameter.
> And opt should always be non-NULL, no need to check.
>
> Cc: Krzysztof Kolasa <kkolasa@winsoft.pl>
> Cc: Eric Dumazet <edumazet@google.com>
> Tested-by: Krzysztof Kolasa <kkolasa@winsoft.pl>
> Signed-off-by: Cong Wang <cwang@twopensource.com>
> Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com>
> ---
Acked-by: Eric Dumazet <edumazet@google.com>
^ permalink raw reply
* Re: [Patch net 2/3] ipv4: share tcp_v4_save_options() with cookie_v4_check()
From: Eric Dumazet @ 2014-10-16 4:16 UTC (permalink / raw)
To: Cong Wang; +Cc: netdev, davem, Krzysztof Kolasa, Eric Dumazet, Cong Wang
In-Reply-To: <1413408802-21052-2-git-send-email-xiyou.wangcong@gmail.com>
On Wed, 2014-10-15 at 14:33 -0700, Cong Wang wrote:
> From: Cong Wang <cwang@twopensource.com>
>
> cookie_v4_check() allocates ip_options_rcu in the same way
> with tcp_v4_save_options(), we can just make it a helper function.
>
> Cc: Krzysztof Kolasa <kkolasa@winsoft.pl>
> Cc: Eric Dumazet <edumazet@google.com>
> Signed-off-by: Cong Wang <cwang@twopensource.com>
> Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com>
> ---
Acked-by: Eric Dumazet <edumazet@google.com>
^ permalink raw reply
* Re: [Patch net 1/3] ipv4: call __ip_options_echo() in cookie_v4_check()
From: Eric Dumazet @ 2014-10-16 4:15 UTC (permalink / raw)
To: Cong Wang; +Cc: netdev, davem, Krzysztof Kolasa, Eric Dumazet, Cong Wang
In-Reply-To: <1413408802-21052-1-git-send-email-xiyou.wangcong@gmail.com>
On Wed, 2014-10-15 at 14:33 -0700, Cong Wang wrote:
> From: Cong Wang <cwang@twopensource.com>
>
> commit 971f10eca186cab238c49da ("tcp: better TCP_SKB_CB layout to reduce cache line misses")
> missed that cookie_v4_check() still calls ip_options_echo() which uses
> IPCB(). It should use TCPCB() at TCP layer, so call __ip_options_echo()
> instead.
>
> Fixes: commit 971f10eca186cab238c49da ("tcp: better TCP_SKB_CB layout to reduce cache line misses")
> Cc: Krzysztof Kolasa <kkolasa@winsoft.pl>
> Cc: Eric Dumazet <edumazet@google.com>
> Reported-by: Krzysztof Kolasa <kkolasa@winsoft.pl>
> Tested-by: Krzysztof Kolasa <kkolasa@winsoft.pl>
> Signed-off-by: Cong Wang <cwang@twopensource.com>
> Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com>
> ---
> net/ipv4/syncookies.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/net/ipv4/syncookies.c b/net/ipv4/syncookies.c
> index 0431a8f..7e7401c 100644
> --- a/net/ipv4/syncookies.c
> +++ b/net/ipv4/syncookies.c
> @@ -321,7 +321,7 @@ struct sock *cookie_v4_check(struct sock *sk, struct sk_buff *skb,
> int opt_size = sizeof(struct ip_options_rcu) + opt->optlen;
>
> ireq->opt = kmalloc(opt_size, GFP_ATOMIC);
> - if (ireq->opt != NULL && ip_options_echo(&ireq->opt->opt, skb)) {
> + if (ireq->opt != NULL && __ip_options_echo(&ireq->opt->opt, skb, opt)) {
> kfree(ireq->opt);
> ireq->opt = NULL;
> }
Signed-off-by: Eric Dumazet <edumazet@google.com>
^ permalink raw reply
* Re: [PATCH] net: use hardware buffer pool to allocate skb
From: Eric Dumazet @ 2014-10-16 4:15 UTC (permalink / raw)
To: Jiafei.Pan@freescale.com
Cc: David Miller, jkosina@suse.cz, netdev@vger.kernel.org,
LeoLi@freescale.com, linux-doc@vger.kernel.org
In-Reply-To: <524626e093684abeba65839d26e94262@BLUPR03MB517.namprd03.prod.outlook.com>
On Thu, 2014-10-16 at 02:17 +0000, Jiafei.Pan@freescale.com wrote:
> Thanks for your comments and suggestion. In my case, I want to build skb
> from hardware block specified memory, I only can see two ways, one is modified
> net card driver replace common skb allocation function with my specially
> functions, another way is to hack common skb allocation function in which
> redirect to my specially functions. My patch is just for the second way.
> Except these two ways, would you please give me some advice for some other
> ways for my case? Thanks
I suggest you read drivers/net/ethernet numerous examples.
No need to change anything in net/* or include/*, really.
For a start, look at drivers/net/ethernet/intel/igb/igb_main.c
Mentioning 'hack' in your mails simply should hint you are doing
something very wrong.
What makes you think your hardware is so special ?
^ permalink raw reply
* Re: Netlink mmap tx security?
From: David Miller @ 2014-10-16 3:34 UTC (permalink / raw)
To: luto; +Cc: dborkman, torvalds, kaber, netdev, tgraf
In-Reply-To: <CALCETrUW6yX8B=ANyxYqhwPbKkosbuBGXO8omTD6rGA1A1XhvA@mail.gmail.com>
From: Andy Lutomirski <luto@amacapital.net>
Date: Wed, 15 Oct 2014 16:58:39 -0700
> On Wed, Oct 15, 2014 at 4:57 PM, David Miller <davem@davemloft.net> wrote:
>> From: Daniel Borkmann <dborkman@redhat.com>
>> Date: Thu, 16 Oct 2014 01:45:22 +0200
>>
>>> On 10/15/2014 04:01 AM, David Miller wrote:
>>>> From: Andy Lutomirski <luto@amacapital.net>
>>>> Date: Tue, 14 Oct 2014 15:16:46 -0700
>>>>
>>>>> It's at least remotely possible that there's something that assumes
>>>>> that assumes that the availability of NETLINK_RX_RING implies
>>>>> NETLINK_TX_RING, which would be unfortunate.
>>>>
>>>> I already found one such case, nlmon :-/
>>>
>>> Hmm, can you elaborate? I currently don't think that nlmon cares
>>> actually.
>>
>> nlmon cares, openvswitch cares, etc:
>
> It'll still work, just slower, right?
>
> + if (setsockopt(sock->fd, SOL_NETLINK, NETLINK_RX_RING, &req,
> sizeof(req)) < 0
> + || setsockopt(sock->fd, SOL_NETLINK, NETLINK_TX_RING, &req,
> sizeof(req)) < 0) {
> + VLOG_INFO("mmap netlink is not supported");
> + return 0;
> + }
So NETLINK_RX_RING succeeds but NETLINK_TX_RING fails, so now the
recvmsg() path will take the RX ring code path because this code
doesn't clean up and undo the setsockopt().
This is the common coding paradigm I've seen in all NETLINK_*_RING
uses.
^ permalink raw reply
* Re: [PATCH] vxlan: using pskb_may_pull as early as possible
From: David Miller @ 2014-10-16 3:33 UTC (permalink / raw)
To: roy.qing.li; +Cc: netdev, xiyou.wangcong
In-Reply-To: <1413422238-2562-1-git-send-email-roy.qing.li@gmail.com>
From: roy.qing.li@gmail.com
Date: Thu, 16 Oct 2014 09:17:18 +0800
> From: Li RongQing <roy.qing.li@gmail.com>
>
> pskb_may_pull should be used to check if skb->data has enough space,
> skb->len can not ensure that.
>
> Cc: Cong Wang <xiyou.wangcong@gmail.com>
> Signed-off-by: Li RongQing <roy.qing.li@gmail.com>
Also applied and queued up for -stable,thanks.
^ permalink raw reply
* Re: [PATCH] vxlan: fix a use after free in vxlan_encap_bypass
From: David Miller @ 2014-10-16 3:31 UTC (permalink / raw)
To: roy.qing.li; +Cc: netdev
In-Reply-To: <1413420581-12989-1-git-send-email-roy.qing.li@gmail.com>
From: roy.qing.li@gmail.com
Date: Thu, 16 Oct 2014 08:49:41 +0800
> From: Li RongQing <roy.qing.li@gmail.com>
>
> when netif_rx() is done, the netif_rx handled skb maybe be freed,
> and should not be used.
>
> Signed-off-by: Li RongQing <roy.qing.li@gmail.com>
Looks good, applied, and queued up for -stable thanks!
^ permalink raw reply
* Re: [PATCH 1/1 net-next] openvswitch: use vport instead of p
From: David Miller @ 2014-10-16 3:26 UTC (permalink / raw)
To: fabf; +Cc: linux-kernel, pshelar, dev, netdev
In-Reply-To: <1413399821-8558-1-git-send-email-fabf@skynet.be>
From: Fabian Frederick <fabf@skynet.be>
Date: Wed, 15 Oct 2014 21:03:41 +0200
> All functions used struct vport *vport except
> ovs_vport_find_upcall_portid.
>
> This fixes 1 kerneldoc warning
>
> Signed-off-by: Fabian Frederick <fabf@skynet.be>
Applied.
^ permalink raw reply
* Re: [PATCH 1/1 net-next] openvswitch: kerneldoc warning fix
From: David Miller @ 2014-10-16 3:25 UTC (permalink / raw)
To: fabf; +Cc: linux-kernel, pshelar, dev, netdev
In-Reply-To: <1413399798-8514-1-git-send-email-fabf@skynet.be>
From: Fabian Frederick <fabf@skynet.be>
Date: Wed, 15 Oct 2014 21:03:18 +0200
> s/sock/gs
>
> Signed-off-by: Fabian Frederick <fabf@skynet.be>
Applied.
^ permalink raw reply
* Re: [PATCH net-next] cxgb4 : Improve handling of DCB negotiation or loss thereof
From: David Miller @ 2014-10-16 3:24 UTC (permalink / raw)
To: anish; +Cc: netdev, hariprasad, leedom
In-Reply-To: <1413425295-22824-1-git-send-email-anish@chelsio.com>
From: Anish Bhatt <anish@chelsio.com>
Date: Wed, 15 Oct 2014 19:08:15 -0700
> Clear out any DCB apps we might have added to kernel table when we lose DCB
> sync (or IEEE equivalent event). IEEE allows individual components to work
> independently, so improve check for IEEE completion by specifying individual
> components.
>
> Signed-off-by: Anish Bhatt <anish@chelsio.com>
net-next is not open at this time, please resubmit this patch
when it is open again.
Thank you.
^ permalink raw reply
* Re: [PATCH net-next 0/4] Add support for few debugfs entries
From: David Miller @ 2014-10-16 3:24 UTC (permalink / raw)
To: hariprasad; +Cc: netdev, leedom, kumaras, nirranjan, santosh, anish
In-Reply-To: <1413422524-14054-1-git-send-email-hariprasad@chelsio.com>
From: Hariprasad Shenai <hariprasad@chelsio.com>
Date: Thu, 16 Oct 2014 06:52:00 +0530
> This patch series adds a new file for debugfs and support for devlog, cim_la,
> cim_la_qcfg and mps_tcam debugfs entries.
>
> The patches series is created against 'net-next' tree.
> And includes patches on cxgb4 driver.
>
> We have included all the maintainers of respective drivers. Kindly review the
> change and let us know in case of any review comments.
net-next is not open at this time, please resubmit this series
when it is open again.
Thank you.
^ permalink raw reply
* RE: [PATCH] net: use hardware buffer pool to allocate skb
From: Jiafei.Pan @ 2014-10-16 2:30 UTC (permalink / raw)
To: Stephen Hemminger
Cc: davem@davemloft.net, jkosina@suse.cz, netdev@vger.kernel.org,
LeoLi@freescale.com, linux-doc@vger.kernel.org,
Jiafei.Pan@freescale.com
In-Reply-To: <20141015113323.5321b2f7@uryu.home.lan>
> -----Original Message-----
> From: Stephen Hemminger [mailto:stephen@networkplumber.org]
> Sent: Wednesday, October 15, 2014 5:33 PM
> To: Pan Jiafei-B37022
> Cc: davem@davemloft.net; jkosina@suse.cz; netdev@vger.kernel.org; Li Yang-Leo-
> R58472; linux-doc@vger.kernel.org
> Subject: Re: [PATCH] net: use hardware buffer pool to allocate skb
>
> Since an skb can sit forever in an application queue, you have created
> an easy way to livelock the system when enough skb's are waiting to be
> read.
I think there is no possible to livelock the system, because in my patch
The function __netdev_alloc_skb will try to allocate hardware block buffer
Firstly if dev->alloc_hw_skb is set, but it will continue allocate normal
skb buffer if the hardware block buffer allocation fails.
^ permalink raw reply
* RE: [PATCH] net: use hardware buffer pool to allocate skb
From: Jiafei.Pan @ 2014-10-16 2:17 UTC (permalink / raw)
To: Eric Dumazet
Cc: David Miller, jkosina@suse.cz, netdev@vger.kernel.org,
LeoLi@freescale.com, linux-doc@vger.kernel.org,
Jiafei.Pan@freescale.com
In-Reply-To: <1413364533.12304.44.camel@edumazet-glaptop2.roam.corp.google.com>
> -----Original Message-----
> From: Eric Dumazet [mailto:eric.dumazet@gmail.com]
> Sent: Wednesday, October 15, 2014 5:16 PM
> To: Pan Jiafei-B37022
> Cc: David Miller; jkosina@suse.cz; netdev@vger.kernel.org; Li Yang-Leo-R58472;
> linux-doc@vger.kernel.org
> Subject: Re: [PATCH] net: use hardware buffer pool to allocate skb
>
> On Wed, 2014-10-15 at 05:34 +0000, Jiafei.Pan@freescale.com wrote:
>
> > Yes, for this matter, in order to do this we should modify the Ethernet
> > drivers. For example, driver A want to driver B, C, D.. to support driver
> > A's Hardware block access functions, so we have to modify driver B, C, D...
> > It will be so complex for this matter.
> >
> > But by using my patch, I just modify a Ethernet device (I don't care
> > Which driver it is used) flag in driver A in order to implement this
> > Ethernet device using hardware block access functions provided by
> > Driver A.
>
> We care a lot of all the bugs added by your patches. You have little
> idea of how many of them were added. We do not want to spend days of
> work explaining everything or fixing all the details for you.
>
> Carefully read net/core/skbuff.c, net/core/dev.c, GRO layer, you'll see
> how many spots you missed.
>
> You cannot control how skbs are cooked before reaching your driver
> ndo_start_xmit(). You are not going to add hooks in UDP , TCP, or other
> drivers RX path. This would be absolutely insane.
>
> Trying to control how skb are cooked in RX path is absolutely something
> drivers do, using page frags that are read-only by all the stack.
>
> Fix your driver to use existing infra, your suggestion is not going to
> be accepted.
>
I think my patch can connect some hardware buffer block with the third party
net card drivers. this should be a general requirement in order to get
a better performance. Yes, maybe some defect in my patch, but any comments
and suggestions for this target is welcome and thanks.
^ permalink raw reply
* RE: [PATCH] net: use hardware buffer pool to allocate skb
From: Jiafei.Pan @ 2014-10-16 2:17 UTC (permalink / raw)
To: Eric Dumazet
Cc: David Miller, jkosina@suse.cz, netdev@vger.kernel.org,
LeoLi@freescale.com, linux-doc@vger.kernel.org,
Jiafei.Pan@freescale.com
In-Reply-To: <1413364533.12304.44.camel@edumazet-glaptop2.roam.corp.google.com>
> -----Original Message-----
> From: Eric Dumazet [mailto:eric.dumazet@gmail.com]
> Sent: Wednesday, October 15, 2014 5:16 PM
> To: Pan Jiafei-B37022
> Cc: David Miller; jkosina@suse.cz; netdev@vger.kernel.org; Li Yang-Leo-R58472;
> linux-doc@vger.kernel.org
> Subject: Re: [PATCH] net: use hardware buffer pool to allocate skb
>
> On Wed, 2014-10-15 at 05:34 +0000, Jiafei.Pan@freescale.com wrote:
>
> > Yes, for this matter, in order to do this we should modify the Ethernet
> > drivers. For example, driver A want to driver B, C, D.. to support driver
> > A's Hardware block access functions, so we have to modify driver B, C, D...
> > It will be so complex for this matter.
> >
> > But by using my patch, I just modify a Ethernet device (I don't care
> > Which driver it is used) flag in driver A in order to implement this
> > Ethernet device using hardware block access functions provided by
> > Driver A.
>
> We care a lot of all the bugs added by your patches. You have little
> idea of how many of them were added. We do not want to spend days of
> work explaining everything or fixing all the details for you.
>
> Carefully read net/core/skbuff.c, net/core/dev.c, GRO layer, you'll see
> how many spots you missed.
>
> You cannot control how skbs are cooked before reaching your driver
> ndo_start_xmit(). You are not going to add hooks in UDP , TCP, or other
> drivers RX path. This would be absolutely insane.
>
Thanks for your comments and suggestion. In my case, I want to build skb
from hardware block specified memory, I only can see two ways, one is modified
net card driver replace common skb allocation function with my specially
functions, another way is to hack common skb allocation function in which
redirect to my specially functions. My patch is just for the second way.
Except these two ways, would you please give me some advice for some other
ways for my case? Thanks.
> Trying to control how skb are cooked in RX path is absolutely something
> drivers do, using page frags that are read-only by all the stack.
>
> Fix your driver to use existing infra, your suggestion is not going to
> be accepted.
>
^ permalink raw reply
* [PATCH net-next] cxgb4 : Improve handling of DCB negotiation or loss thereof
From: Anish Bhatt @ 2014-10-16 2:08 UTC (permalink / raw)
To: netdev; +Cc: davem, hariprasad, leedom, Anish Bhatt
Clear out any DCB apps we might have added to kernel table when we lose DCB
sync (or IEEE equivalent event). IEEE allows individual components to work
independently, so improve check for IEEE completion by specifying individual
components.
Signed-off-by: Anish Bhatt <anish@chelsio.com>
---
drivers/net/ethernet/chelsio/cxgb4/cxgb4_dcb.c | 48 ++++++++++++++++++++++++--
1 file changed, 45 insertions(+), 3 deletions(-)
diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_dcb.c b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_dcb.c
index 8edf0f5bd679..ee819fd12bd2 100644
--- a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_dcb.c
+++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_dcb.c
@@ -60,6 +60,42 @@ void cxgb4_dcb_version_init(struct net_device *dev)
dcb->dcb_version = FW_PORT_DCB_VER_AUTO;
}
+static void cxgb4_dcb_cleanup_apps(struct net_device *dev)
+{
+ struct port_info *pi = netdev2pinfo(dev);
+ struct adapter *adap = pi->adapter;
+ struct port_dcb_info *dcb = &pi->dcb;
+ struct dcb_app app;
+ int i, err;
+
+ /* zero priority implies remove */
+ app.priority = 0;
+
+ for (i = 0; i < CXGB4_MAX_DCBX_APP_SUPPORTED; i++) {
+ /* Check if app list is exhausted */
+ if (!dcb->app_priority[i].protocolid)
+ break;
+
+ app.protocol = dcb->app_priority[i].protocolid;
+
+ if (dcb->dcb_version == FW_PORT_DCB_VER_IEEE) {
+ app.selector = dcb->app_priority[i].sel_field + 1;
+ err = dcb_ieee_setapp(dev, &app);
+ } else {
+ app.selector = !!(dcb->app_priority[i].sel_field);
+ err = dcb_setapp(dev, &app);
+ }
+
+ if (err) {
+ dev_err(adap->pdev_dev,
+ "Failed DCB Clear %s Application Priority: sel=%d, prot=%d, , err=%d\n",
+ dcb_ver_array[dcb->dcb_version], app.selector,
+ app.protocol, -err);
+ break;
+ }
+ }
+}
+
/* Finite State machine for Data Center Bridging.
*/
void cxgb4_dcb_state_fsm(struct net_device *dev,
@@ -145,6 +181,7 @@ void cxgb4_dcb_state_fsm(struct net_device *dev,
* state. We need to reset back to a ground state
* of incomplete.
*/
+ cxgb4_dcb_cleanup_apps(dev);
cxgb4_dcb_state_init(dev);
dcb->state = CXGB4_DCB_STATE_FW_INCOMPLETE;
dcb->supported = CXGB4_DCBX_FW_SUPPORT;
@@ -833,11 +870,16 @@ static int cxgb4_setapp(struct net_device *dev, u8 app_idtype, u16 app_id,
/* Return whether IEEE Data Center Bridging has been negotiated.
*/
-static inline int cxgb4_ieee_negotiation_complete(struct net_device *dev)
+static inline int
+cxgb4_ieee_negotiation_complete(struct net_device *dev,
+ enum cxgb4_dcb_fw_msgs dcb_subtype)
{
struct port_info *pi = netdev2pinfo(dev);
struct port_dcb_info *dcb = &pi->dcb;
+ if (dcb_subtype && !(dcb->msgs & dcb_subtype))
+ return 0;
+
return (dcb->state == CXGB4_DCB_STATE_FW_ALLSYNCED &&
(dcb->supported & DCB_CAP_DCBX_VER_IEEE));
}
@@ -850,7 +892,7 @@ static int cxgb4_ieee_getapp(struct net_device *dev, struct dcb_app *app)
{
int prio;
- if (!cxgb4_ieee_negotiation_complete(dev))
+ if (!cxgb4_ieee_negotiation_complete(dev, CXGB4_DCB_FW_APP_ID))
return -EINVAL;
if (!(app->selector && app->protocol))
return -EINVAL;
@@ -872,7 +914,7 @@ static int cxgb4_ieee_setapp(struct net_device *dev, struct dcb_app *app)
{
int ret;
- if (!cxgb4_ieee_negotiation_complete(dev))
+ if (!cxgb4_ieee_negotiation_complete(dev, CXGB4_DCB_FW_APP_ID))
return -EINVAL;
if (!(app->selector && app->protocol))
return -EINVAL;
--
2.1.2
^ permalink raw reply related
* [PATCH] vxlan: using pskb_may_pull as early as possible
From: roy.qing.li @ 2014-10-16 1:17 UTC (permalink / raw)
To: netdev; +Cc: xiyou.wangcong
From: Li RongQing <roy.qing.li@gmail.com>
pskb_may_pull should be used to check if skb->data has enough space,
skb->len can not ensure that.
Cc: Cong Wang <xiyou.wangcong@gmail.com>
Signed-off-by: Li RongQing <roy.qing.li@gmail.com>
---
drivers/net/vxlan.c | 6 ++----
1 file changed, 2 insertions(+), 4 deletions(-)
diff --git a/drivers/net/vxlan.c b/drivers/net/vxlan.c
index faf1bd1..77ab844 100644
--- a/drivers/net/vxlan.c
+++ b/drivers/net/vxlan.c
@@ -1437,9 +1437,6 @@ static int neigh_reduce(struct net_device *dev, struct sk_buff *skb)
if (!in6_dev)
goto out;
- if (!pskb_may_pull(skb, skb->len))
- goto out;
-
iphdr = ipv6_hdr(skb);
saddr = &iphdr->saddr;
daddr = &iphdr->daddr;
@@ -1880,7 +1877,8 @@ static netdev_tx_t vxlan_xmit(struct sk_buff *skb, struct net_device *dev)
return arp_reduce(dev, skb);
#if IS_ENABLED(CONFIG_IPV6)
else if (ntohs(eth->h_proto) == ETH_P_IPV6 &&
- skb->len >= sizeof(struct ipv6hdr) + sizeof(struct nd_msg) &&
+ pskb_may_pull(skb, sizeof(struct ipv6hdr)
+ + sizeof(struct nd_msg)) &&
ipv6_hdr(skb)->nexthdr == IPPROTO_ICMPV6) {
struct nd_msg *msg;
--
1.7.10.4
^ permalink raw reply related
* [PATCH net-next 4/4] cxgb4: Add support for mps_tcam debugfs
From: Hariprasad Shenai @ 2014-10-16 1:22 UTC (permalink / raw)
To: netdev; +Cc: davem, leedom, kumaras, nirranjan, santosh, anish,
Hariprasad Shenai
In-Reply-To: <1413422524-14054-1-git-send-email-hariprasad@chelsio.com>
Debug log to get the MPS TCAM table
Signed-off-by: Hariprasad Shenai <hariprasad@chelsio.com>
---
drivers/net/ethernet/chelsio/cxgb4/cxgb4_debugfs.c | 130 ++++++++++++++++++++
drivers/net/ethernet/chelsio/cxgb4/t4_regs.h | 16 +++
2 files changed, 146 insertions(+), 0 deletions(-)
diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_debugfs.c b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_debugfs.c
index e8b84dc..7d114ed 100644
--- a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_debugfs.c
+++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_debugfs.c
@@ -435,6 +435,135 @@ static const struct file_operations devlog_fops = {
.release = seq_release_private
};
+static inline void tcamxy2valmask(u64 x, u64 y, u8 *addr, u64 *mask)
+{
+ *mask = x | y;
+ y = (__force u64)cpu_to_be64(y);
+ memcpy(addr, (char *)&y + 2, ETH_ALEN);
+}
+
+static int mps_tcam_show(struct seq_file *seq, void *v)
+{
+ if (v == SEQ_START_TOKEN)
+ seq_puts(seq, "Idx Ethernet address Mask Vld Ports PF"
+ " VF Replication P0 P1 P2 P3 ML\n");
+ else {
+ u64 mask;
+ u8 addr[ETH_ALEN];
+ struct adapter *adap = seq->private;
+ unsigned int idx = (uintptr_t)v - 2;
+ u64 tcamy = t4_read_reg64(adap, MPS_CLS_TCAM_Y_L(idx));
+ u64 tcamx = t4_read_reg64(adap, MPS_CLS_TCAM_X_L(idx));
+ u32 cls_lo = t4_read_reg(adap, MPS_CLS_SRAM_L(idx));
+ u32 cls_hi = t4_read_reg(adap, MPS_CLS_SRAM_H(idx));
+ u32 rplc[4] = {0, 0, 0, 0};
+
+ if (tcamx & tcamy) {
+ seq_printf(seq, "%3u -\n", idx);
+ goto out;
+ }
+
+ if (cls_lo & REPLICATE) {
+ struct fw_ldst_cmd ldst_cmd;
+ int ret;
+
+ memset(&ldst_cmd, 0, sizeof(ldst_cmd));
+ ldst_cmd.op_to_addrspace =
+ htonl(FW_CMD_OP(FW_LDST_CMD) |
+ FW_CMD_REQUEST |
+ FW_CMD_READ |
+ FW_LDST_CMD_ADDRSPACE(
+ FW_LDST_ADDRSPC_MPS));
+ ldst_cmd.cycles_to_len16 = htonl(FW_LEN16(ldst_cmd));
+ ldst_cmd.u.mps.fid_ctl =
+ htons(FW_LDST_CMD_FID(FW_LDST_MPS_RPLC) |
+ FW_LDST_CMD_CTL(idx));
+ ret = t4_wr_mbox(adap, adap->mbox, &ldst_cmd,
+ sizeof(ldst_cmd), &ldst_cmd);
+ if (ret)
+ dev_warn(adap->pdev_dev, "Can't read MPS "
+ "replication map for idx %d: %d\n",
+ idx, -ret);
+ else {
+ rplc[0] = ntohl(ldst_cmd.u.mps.rplc31_0);
+ rplc[1] = ntohl(ldst_cmd.u.mps.rplc63_32);
+ rplc[2] = ntohl(ldst_cmd.u.mps.rplc95_64);
+ rplc[3] = ntohl(ldst_cmd.u.mps.rplc127_96);
+ }
+ }
+
+ tcamxy2valmask(tcamx, tcamy, addr, &mask);
+ seq_printf(seq, "%3u %02x:%02x:%02x:%02x:%02x:%02x %012llx"
+ "%3c %#x%4u%4d",
+ idx, addr[0], addr[1], addr[2], addr[3], addr[4],
+ addr[5], (unsigned long long)mask,
+ (cls_lo & SRAM_VLD) ? 'Y' : 'N', PORTMAP_GET(cls_hi),
+ PF_GET(cls_lo),
+ (cls_lo & VF_VALID) ? VF_GET(cls_lo) : -1);
+ if (cls_lo & REPLICATE)
+ seq_printf(seq, " %08x %08x %08x %08x",
+ rplc[3], rplc[2], rplc[1], rplc[0]);
+ else
+ seq_printf(seq, "%36c", ' ');
+ seq_printf(seq, "%4u%3u%3u%3u %#x\n",
+ SRAM_PRIO0_GET(cls_lo), SRAM_PRIO1_GET(cls_lo),
+ SRAM_PRIO2_GET(cls_lo), SRAM_PRIO3_GET(cls_lo),
+ (cls_lo >> MULTILISTEN0_SHIFT) & 0xf);
+ }
+out: return 0;
+}
+
+static inline void *mps_tcam_get_idx(struct seq_file *seq, loff_t pos)
+{
+ struct adapter *adap = seq->private;
+ int max_mac_addr = is_t4(adap->params.chip) ?
+ NUM_MPS_CLS_SRAM_L_INSTANCES :
+ NUM_MPS_T5_CLS_SRAM_L_INSTANCES;
+ return ((pos <= max_mac_addr) ? (void *)(uintptr_t)(pos + 1) : NULL);
+}
+
+static void *mps_tcam_start(struct seq_file *seq, loff_t *pos)
+{
+ return *pos ? mps_tcam_get_idx(seq, *pos) : SEQ_START_TOKEN;
+}
+
+static void *mps_tcam_next(struct seq_file *seq, void *v, loff_t *pos)
+{
+ ++*pos;
+ return mps_tcam_get_idx(seq, *pos);
+}
+
+static void mps_tcam_stop(struct seq_file *seq, void *v)
+{
+}
+
+static const struct seq_operations mps_tcam_seq_ops = {
+ .start = mps_tcam_start,
+ .next = mps_tcam_next,
+ .stop = mps_tcam_stop,
+ .show = mps_tcam_show
+};
+
+static int mps_tcam_open(struct inode *inode, struct file *file)
+{
+ int res = seq_open(file, &mps_tcam_seq_ops);
+
+ if (!res) {
+ struct seq_file *seq = file->private_data;
+
+ seq->private = inode->i_private;
+ }
+ return res;
+}
+
+static const struct file_operations mps_tcam_debugfs_fops = {
+ .owner = THIS_MODULE,
+ .open = mps_tcam_open,
+ .read = seq_read,
+ .llseek = seq_lseek,
+ .release = seq_release,
+};
+
static ssize_t mem_read(struct file *file, char __user *buf, size_t count,
loff_t *ppos)
{
@@ -517,6 +646,7 @@ int t4_setup_debugfs(struct adapter *adap)
{ "cim_qcfg", &cim_qcfg_fops, S_IRUSR, 0 },
{ "devlog", &devlog_fops, S_IRUSR, 0 },
{ "l2t", &t4_l2t_fops, S_IRUSR, 0},
+ { "mps_tcam", &mps_tcam_debugfs_fops, S_IRUSR, 0 },
};
add_debugfs_files(adap,
diff --git a/drivers/net/ethernet/chelsio/cxgb4/t4_regs.h b/drivers/net/ethernet/chelsio/cxgb4/t4_regs.h
index 562cd70..6c4369e 100644
--- a/drivers/net/ethernet/chelsio/cxgb4/t4_regs.h
+++ b/drivers/net/ethernet/chelsio/cxgb4/t4_regs.h
@@ -1054,6 +1054,22 @@
#define MPS_RX_PERR_INT_CAUSE 0x11074
+#define MPS_CLS_TCAM_Y_L(idx) (0xf000 + (idx) * 16)
+#define MPS_CLS_TCAM_X_L(idx) (0xf008 + (idx) * 16)
+#define MPS_CLS_SRAM_L(idx) (0xe000 + (idx) * 8)
+#define MPS_CLS_SRAM_H(idx) (0xe004 + (idx) * 8)
+#define REPLICATE (1U << 11)
+#define SRAM_VLD (1U << 12)
+#define SRAM_PRIO0_GET(x) (((x) >> 13) & 0x7U)
+#define SRAM_PRIO1_GET(x) (((x) >> 16) & 0x7U)
+#define SRAM_PRIO2_GET(x) (((x) >> 19) & 0x7U)
+#define SRAM_PRIO3_GET(x) (((x) >> 22) & 0x7U)
+#define PORTMAP_GET(x) (((x) >> 0) & 0xfU)
+#define PF_GET(x) (((x) >> 8) & 0x7U)
+#define VF_GET(x) (((x) >> 0) & 0x7fU)
+#define VF_VALID (1U << 7)
+#define MULTILISTEN0_SHIFT 25
+
#define CPL_INTR_CAUSE 0x19054
#define CIM_OP_MAP_PERR 0x00000020U
#define CIM_OVFL_ERROR 0x00000010U
--
1.7.1
^ permalink raw reply related
* [PATCH net-next 3/4] cxgb4: Add support for cim_la_qcfg entry in debugfs
From: Hariprasad Shenai @ 2014-10-16 1:22 UTC (permalink / raw)
To: netdev; +Cc: davem, leedom, kumaras, nirranjan, santosh, anish,
Hariprasad Shenai
In-Reply-To: <1413422524-14054-1-git-send-email-hariprasad@chelsio.com>
Adds debug log to get cim la queue config
Signed-off-by: Hariprasad Shenai <hariprasad@chelsio.com>
---
drivers/net/ethernet/chelsio/cxgb4/cxgb4.h | 1 +
drivers/net/ethernet/chelsio/cxgb4/cxgb4_debugfs.c | 70 ++++++++++++++++++++
drivers/net/ethernet/chelsio/cxgb4/t4_hw.c | 35 ++++++++++
drivers/net/ethernet/chelsio/cxgb4/t4_hw.h | 3 +
drivers/net/ethernet/chelsio/cxgb4/t4_regs.h | 22 ++++++
5 files changed, 131 insertions(+), 0 deletions(-)
diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4.h b/drivers/net/ethernet/chelsio/cxgb4/cxgb4.h
index d8a62cc..ab3f925 100644
--- a/drivers/net/ethernet/chelsio/cxgb4/cxgb4.h
+++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4.h
@@ -1020,6 +1020,7 @@ int t4_cim_read(struct adapter *adap, unsigned int addr, unsigned int n,
int t4_cim_write(struct adapter *adap, unsigned int addr, unsigned int n,
const unsigned int *valp);
int t4_cim_read_la(struct adapter *adap, u32 *la_buf, unsigned int *wrptr);
+void t4_read_cimq_cfg(struct adapter *adap, u16 *base, u16 *size, u16 *thres);
const char *t4_get_port_type_description(enum fw_port_type port_type);
void t4_get_port_stats(struct adapter *adap, int idx, struct port_stats *p);
void t4_read_mtu_tbl(struct adapter *adap, u16 *mtus, u8 *mtu_log);
diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_debugfs.c b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_debugfs.c
index 04b490e..e8b84dc 100644
--- a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_debugfs.c
+++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_debugfs.c
@@ -168,6 +168,75 @@ static const struct file_operations cim_la_fops = {
.release = seq_release_private
};
+static int cim_qcfg_show(struct seq_file *seq, void *v)
+{
+ static const char * const qname[] = {
+ "TP0", "TP1", "ULP", "SGE0", "SGE1", "NC-SI",
+ "ULP0", "ULP1", "ULP2", "ULP3", "SGE", "NC-SI",
+ "SGE0-RX", "SGE1-RX"
+ };
+
+ int i;
+ struct adapter *adap = seq->private;
+ u16 base[CIM_NUM_IBQ + CIM_NUM_OBQ_T5];
+ u16 size[CIM_NUM_IBQ + CIM_NUM_OBQ_T5];
+ u32 stat[(4 * (CIM_NUM_IBQ + CIM_NUM_OBQ_T5))];
+ u16 thres[CIM_NUM_IBQ];
+ u32 obq_wr_t4[2 * CIM_NUM_OBQ], *wr;
+ u32 obq_wr_t5[2 * CIM_NUM_OBQ_T5];
+ u32 *p = stat;
+ int cim_num_obq = is_t4(adap->params.chip) ?
+ CIM_NUM_OBQ : CIM_NUM_OBQ_T5;
+
+ i = t4_cim_read(adap, is_t4(adap->params.chip) ? UP_IBQ_0_RDADDR :
+ UP_IBQ_0_SHADOW_RDADDR,
+ ARRAY_SIZE(stat), stat);
+ if (!i) {
+ if (is_t4(adap->params.chip)) {
+ i = t4_cim_read(adap, UP_OBQ_0_REALADDR,
+ ARRAY_SIZE(obq_wr_t4), obq_wr_t4);
+ wr = obq_wr_t4;
+ } else {
+ i = t4_cim_read(adap, UP_OBQ_0_SHADOW_REALADDR,
+ ARRAY_SIZE(obq_wr_t5), obq_wr_t5);
+ wr = obq_wr_t5;
+ }
+ }
+ if (i)
+ return i;
+
+ t4_read_cimq_cfg(adap, base, size, thres);
+
+ seq_printf(seq,
+ " Queue Base Size Thres RdPtr WrPtr SOP EOP Avail\n");
+ for (i = 0; i < CIM_NUM_IBQ; i++, p += 4)
+ seq_printf(seq, "%7s %5x %5u %5u %6x %4x %4u %4u %5u\n",
+ qname[i], base[i], size[i], thres[i],
+ IBQRDADDR_GET(p[0]), IBQWRADDR_GET(p[1]),
+ QUESOPCNT_GET(p[3]), QUEEOPCNT_GET(p[3]),
+ QUEREMFLITS_GET(p[2]) * 16);
+ for ( ; i < CIM_NUM_IBQ + cim_num_obq; i++, p += 4, wr += 2)
+ seq_printf(seq, "%7s %5x %5u %12x %4x %4u %4u %5u\n",
+ qname[i], base[i], size[i],
+ QUERDADDR_GET(p[0]) & 0x3fff, wr[0] - base[i],
+ QUESOPCNT_GET(p[3]), QUEEOPCNT_GET(p[3]),
+ QUEREMFLITS_GET(p[2]) * 16);
+ return 0;
+}
+
+static int cim_qcfg_open(struct inode *inode, struct file *file)
+{
+ return single_open(file, cim_qcfg_show, inode->i_private);
+}
+
+static const struct file_operations cim_qcfg_fops = {
+ .owner = THIS_MODULE,
+ .open = cim_qcfg_open,
+ .read = seq_read,
+ .llseek = seq_lseek,
+ .release = single_release,
+};
+
/* Firmware Device Log dump. */
static const char * const devlog_level_strings[] = {
[FW_DEVLOG_LEVEL_EMERG] = "EMERG",
@@ -445,6 +514,7 @@ int t4_setup_debugfs(struct adapter *adap)
static struct t4_debugfs_entry t4_debugfs_files[] = {
{ "cim_la", &cim_la_fops, S_IRUSR, 0 },
+ { "cim_qcfg", &cim_qcfg_fops, S_IRUSR, 0 },
{ "devlog", &devlog_fops, S_IRUSR, 0 },
{ "l2t", &t4_l2t_fops, S_IRUSR, 0},
};
diff --git a/drivers/net/ethernet/chelsio/cxgb4/t4_hw.c b/drivers/net/ethernet/chelsio/cxgb4/t4_hw.c
index 81e1ef2..900f275 100644
--- a/drivers/net/ethernet/chelsio/cxgb4/t4_hw.c
+++ b/drivers/net/ethernet/chelsio/cxgb4/t4_hw.c
@@ -4128,6 +4128,41 @@ int t4_port_init(struct adapter *adap, int mbox, int pf, int vf)
}
/**
+ * t4_read_cimq_cfg - read CIM queue configuration
+ * @adap: the adapter
+ * @base: holds the queue base addresses in bytes
+ * @size: holds the queue sizes in bytes
+ * @thres: holds the queue full thresholds in bytes
+ *
+ * Returns the current configuration of the CIM queues, starting with
+ * the IBQs, then the OBQs.
+ */
+void t4_read_cimq_cfg(struct adapter *adap, u16 *base, u16 *size, u16 *thres)
+{
+ unsigned int i, v;
+ int cim_num_obq = is_t4(adap->params.chip) ?
+ CIM_NUM_OBQ : CIM_NUM_OBQ_T5;
+
+ for (i = 0; i < CIM_NUM_IBQ; i++) {
+ t4_write_reg(adap, CIM_QUEUE_CONFIG_REF, IBQSELECT |
+ QUENUMSELECT(i));
+ v = t4_read_reg(adap, CIM_QUEUE_CONFIG_CTRL);
+ /* value is in 256-byte units */
+ *base++ = CIMQBASE_GET(v) * 256;
+ *size++ = CIMQSIZE_GET(v) * 256;
+ *thres++ = QUEFULLTHRSH_GET(v) * 8; /* 8-byte unit */
+ }
+ for (i = 0; i < cim_num_obq; i++) {
+ t4_write_reg(adap, CIM_QUEUE_CONFIG_REF, OBQSELECT |
+ QUENUMSELECT(i));
+ v = t4_read_reg(adap, CIM_QUEUE_CONFIG_CTRL);
+ /* value is in 256-byte units */
+ *base++ = CIMQBASE_GET(v) * 256;
+ *size++ = CIMQSIZE_GET(v) * 256;
+ }
+}
+
+/**
* t4_cim_read - read a block from CIM internal address space
* @adap: the adapter
* @addr: the start address within the CIM address space
diff --git a/drivers/net/ethernet/chelsio/cxgb4/t4_hw.h b/drivers/net/ethernet/chelsio/cxgb4/t4_hw.h
index bcc925b..f6b82da 100644
--- a/drivers/net/ethernet/chelsio/cxgb4/t4_hw.h
+++ b/drivers/net/ethernet/chelsio/cxgb4/t4_hw.h
@@ -56,6 +56,9 @@ enum {
};
enum {
+ CIM_NUM_IBQ = 6, /* # of CIM IBQs */
+ CIM_NUM_OBQ = 6, /* # of CIM OBQs */
+ CIM_NUM_OBQ_T5 = 8, /* # of CIM OBQs for T5 adapter */
CIMLA_SIZE = 2048, /* # of 32-bit words in CIM LA */
};
diff --git a/drivers/net/ethernet/chelsio/cxgb4/t4_regs.h b/drivers/net/ethernet/chelsio/cxgb4/t4_regs.h
index 4f6d6fa..562cd70 100644
--- a/drivers/net/ethernet/chelsio/cxgb4/t4_regs.h
+++ b/drivers/net/ethernet/chelsio/cxgb4/t4_regs.h
@@ -623,6 +623,28 @@
#define ILLTRANSINT 0x00000002U
#define RSVDSPACEINT 0x00000001U
+#define CIM_QUEUE_CONFIG_REF 0x7b48
+#define CIM_QUEUE_CONFIG_CTRL 0x7b4c
+
+#define CIMQBASE_GET(x) (((x) >> 16) & 0x3fU)
+#define CIMQSIZE_GET(x) (((x) >> 24) & 0x3fU)
+#define QUEFULLTHRSH_GET(x) (((x) >> 0) & 0x1ffU)
+
+#define UP_IBQ_0_RDADDR 0x10
+#define UP_IBQ_0_SHADOW_RDADDR 0x280
+#define UP_OBQ_0_REALADDR 0x104
+#define UP_OBQ_0_SHADOW_REALADDR 0x394
+
+#define IBQRDADDR_GET(x) (((x) >> 0) & 0x1fffU)
+#define IBQWRADDR_GET(x) (((x) >> 0) & 0x1fffU)
+#define QUERDADDR_GET(x) (((x) >> 0) & 0x7fffU)
+#define QUESOPCNT_GET(x) (((x) >> 0) & 0xfffU)
+#define QUEEOPCNT_GET(x) (((x) >> 16) & 0xfffU)
+#define QUEREMFLITS_GET(x) (((x) >> 0) & 0x7ffU)
+#define OBQSELECT (1U << 4)
+#define IBQSELECT (1U << 3)
+#define QUENUMSELECT(x) ((x) << 0)
+
#define TP_OUT_CONFIG 0x7d04
#define VLANEXTENABLE_MASK 0x0000f000U
#define VLANEXTENABLE_SHIFT 12
--
1.7.1
^ permalink raw reply related
* [PATCH net-next 2/4] cxgb4: Add support for cim_la entry in debugfs
From: Hariprasad Shenai @ 2014-10-16 1:22 UTC (permalink / raw)
To: netdev; +Cc: davem, leedom, kumaras, nirranjan, santosh, anish,
Hariprasad Shenai
In-Reply-To: <1413422524-14054-1-git-send-email-hariprasad@chelsio.com>
The CIM LA captures the embedded processor’s internal state. Optionally, it can
also trace the flow of data in and out of the embedded processor. Therefore, the
CIM LA output contains detailed information of what code the embedded processor
executed prior to the CIM LA capture.
Signed-off-by: Hariprasad Shenai <hariprasad@chelsio.com>
---
drivers/net/ethernet/chelsio/cxgb4/cxgb4.h | 7 +
drivers/net/ethernet/chelsio/cxgb4/cxgb4_debugfs.c | 129 +++++++++++++++++++-
drivers/net/ethernet/chelsio/cxgb4/cxgb4_debugfs.h | 12 ++
drivers/net/ethernet/chelsio/cxgb4/t4_hw.c | 120 ++++++++++++++++++
drivers/net/ethernet/chelsio/cxgb4/t4_hw.h | 4 +
drivers/net/ethernet/chelsio/cxgb4/t4_regs.h | 14 ++
6 files changed, 284 insertions(+), 2 deletions(-)
diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4.h b/drivers/net/ethernet/chelsio/cxgb4/cxgb4.h
index 081b7ef..d8a62cc 100644
--- a/drivers/net/ethernet/chelsio/cxgb4/cxgb4.h
+++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4.h
@@ -297,6 +297,8 @@ struct adapter_params {
struct devlog_params devlog;
enum pcie_memwin drv_memwin;
+ unsigned int cim_la_size;
+
unsigned int sf_size; /* serial flash size in bytes */
unsigned int sf_nsec; /* # of flash sectors */
unsigned int sf_fw_start; /* start of FW image in flash */
@@ -1013,6 +1015,11 @@ int t4_mc_read(struct adapter *adap, int idx, u32 addr, __be32 *data,
u64 *parity);
int t4_edc_read(struct adapter *adap, int idx, u32 addr, __be32 *data,
u64 *parity);
+int t4_cim_read(struct adapter *adap, unsigned int addr, unsigned int n,
+ unsigned int *valp);
+int t4_cim_write(struct adapter *adap, unsigned int addr, unsigned int n,
+ const unsigned int *valp);
+int t4_cim_read_la(struct adapter *adap, u32 *la_buf, unsigned int *wrptr);
const char *t4_get_port_type_description(enum fw_port_type port_type);
void t4_get_port_stats(struct adapter *adap, int idx, struct port_stats *p);
void t4_read_mtu_tbl(struct adapter *adap, u16 *mtus, u8 *mtu_log);
diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_debugfs.c b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_debugfs.c
index 0704af5..04b490e 100644
--- a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_debugfs.c
+++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_debugfs.c
@@ -43,8 +43,132 @@
#include "cxgb4_debugfs.h"
#include "l2t.h"
-/* Firmware Device Log dump.
- */
+/* generic seq_file support for showing a table of size rows x width. */
+static void *seq_tab_get_idx(struct seq_tab *tb, loff_t pos)
+{
+ pos -= tb->skip_first;
+ return pos >= tb->rows ? NULL : &tb->data[pos * tb->width];
+}
+
+static void *seq_tab_start(struct seq_file *seq, loff_t *pos)
+{
+ struct seq_tab *tb = seq->private;
+
+ if (tb->skip_first && *pos == 0)
+ return SEQ_START_TOKEN;
+
+ return seq_tab_get_idx(tb, *pos);
+}
+
+static void *seq_tab_next(struct seq_file *seq, void *v, loff_t *pos)
+{
+ v = seq_tab_get_idx(seq->private, *pos + 1);
+ if (v)
+ ++*pos;
+ return v;
+}
+
+static void seq_tab_stop(struct seq_file *seq, void *v)
+{
+}
+
+static int seq_tab_show(struct seq_file *seq, void *v)
+{
+ const struct seq_tab *tb = seq->private;
+
+ return tb->show(seq, v, ((char *)v - tb->data) / tb->width);
+}
+
+static const struct seq_operations seq_tab_ops = {
+ .start = seq_tab_start,
+ .next = seq_tab_next,
+ .stop = seq_tab_stop,
+ .show = seq_tab_show
+};
+
+struct seq_tab *seq_open_tab(struct file *f, unsigned int rows,
+ unsigned int width, unsigned int have_header,
+ int (*show)(struct seq_file *seq, void *v, int i))
+{
+ struct seq_tab *p;
+
+ p = __seq_open_private(f, &seq_tab_ops, sizeof(*p) + rows * width);
+ if (p) {
+ p->show = show;
+ p->rows = rows;
+ p->width = width;
+ p->skip_first = have_header != 0;
+ }
+ return p;
+}
+
+static int cim_la_show(struct seq_file *seq, void *v, int idx)
+{
+ if (v == SEQ_START_TOKEN)
+ seq_puts(seq, "Status Data PC LS0Stat LS0Addr "
+ " LS0Data\n");
+ else {
+ const u32 *p = v;
+
+ seq_printf(seq,
+ " %02x %x%07x %x%07x %08x %08x %08x%08x%08x%08x\n",
+ (p[0] >> 4) & 0xff, p[0] & 0xf, p[1] >> 4,
+ p[1] & 0xf, p[2] >> 4, p[2] & 0xf, p[3], p[4], p[5],
+ p[6], p[7]);
+ }
+ return 0;
+}
+
+static int cim_la_show_3in1(struct seq_file *seq, void *v, int idx)
+{
+ if (v == SEQ_START_TOKEN) {
+ seq_puts(seq, "Status Data PC\n");
+ } else {
+ const u32 *p = v;
+
+ seq_printf(seq, " %02x %08x %08x\n", p[5] & 0xff, p[6],
+ p[7]);
+ seq_printf(seq, " %02x %02x%06x %02x%06x\n",
+ (p[3] >> 8) & 0xff, p[3] & 0xff, p[4] >> 8,
+ p[4] & 0xff, p[5] >> 8);
+ seq_printf(seq, " %02x %x%07x %x%07x\n", (p[0] >> 4) & 0xff,
+ p[0] & 0xf, p[1] >> 4, p[1] & 0xf, p[2] >> 4);
+ }
+ return 0;
+}
+
+static int cim_la_open(struct inode *inode, struct file *file)
+{
+ int ret;
+ unsigned int cfg;
+ struct seq_tab *p;
+ struct adapter *adap = inode->i_private;
+
+ ret = t4_cim_read(adap, UP_UP_DBG_LA_CFG, 1, &cfg);
+ if (ret)
+ return ret;
+
+ p = seq_open_tab(file, adap->params.cim_la_size / 8, 8 * sizeof(u32), 1,
+ cfg & UPDBGLACAPTPCONLY ?
+ cim_la_show_3in1 : cim_la_show);
+ if (!p)
+ return -ENOMEM;
+
+ ret = t4_cim_read_la(adap, (u32 *)p->data, NULL);
+ if (ret)
+ seq_release_private(inode, file);
+ return ret;
+}
+
+static const struct file_operations cim_la_fops = {
+ .owner = THIS_MODULE,
+ .open = cim_la_open,
+ .read = seq_read,
+ .llseek = seq_lseek,
+ .release = seq_release_private
+};
+
+/* Firmware Device Log dump. */
static const char * const devlog_level_strings[] = {
[FW_DEVLOG_LEVEL_EMERG] = "EMERG",
[FW_DEVLOG_LEVEL_CRIT] = "CRIT",
@@ -320,6 +444,7 @@ int t4_setup_debugfs(struct adapter *adap)
u32 size;
static struct t4_debugfs_entry t4_debugfs_files[] = {
+ { "cim_la", &cim_la_fops, S_IRUSR, 0 },
{ "devlog", &devlog_fops, S_IRUSR, 0 },
{ "l2t", &t4_l2t_fops, S_IRUSR, 0},
};
diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_debugfs.h b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_debugfs.h
index a3d8867..70fcbc9 100644
--- a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_debugfs.h
+++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_debugfs.h
@@ -44,6 +44,18 @@ struct t4_debugfs_entry {
unsigned char data;
};
+struct seq_tab {
+ int (*show)(struct seq_file *seq, void *v, int idx);
+ unsigned int rows; /* # of entries */
+ unsigned char width; /* size in bytes of each entry */
+ unsigned char skip_first; /* whether the first line is a header */
+ char data[0]; /* the table data */
+};
+
+struct seq_tab *seq_open_tab(struct file *f, unsigned int rows,
+ unsigned int width, unsigned int have_header,
+ int (*show)(struct seq_file *seq, void *v, int i));
+
int t4_setup_debugfs(struct adapter *adap);
void add_debugfs_files(struct adapter *adap,
struct t4_debugfs_entry *files,
diff --git a/drivers/net/ethernet/chelsio/cxgb4/t4_hw.c b/drivers/net/ethernet/chelsio/cxgb4/t4_hw.c
index 1fff149..81e1ef2 100644
--- a/drivers/net/ethernet/chelsio/cxgb4/t4_hw.c
+++ b/drivers/net/ethernet/chelsio/cxgb4/t4_hw.c
@@ -3953,6 +3953,7 @@ int t4_prep_adapter(struct adapter *adapter)
return -EINVAL;
}
+ adapter->params.cim_la_size = CIMLA_SIZE;
init_cong_ctrl(adapter->params.a_wnd, adapter->params.b_wnd);
/*
@@ -4125,3 +4126,122 @@ int t4_port_init(struct adapter *adap, int mbox, int pf, int vf)
}
return 0;
}
+
+/**
+ * t4_cim_read - read a block from CIM internal address space
+ * @adap: the adapter
+ * @addr: the start address within the CIM address space
+ * @n: number of words to read
+ * @valp: where to store the result
+ *
+ * Reads a block of 4-byte words from the CIM intenal address space.
+ */
+int t4_cim_read(struct adapter *adap, unsigned int addr, unsigned int n,
+ unsigned int *valp)
+{
+ int ret = 0;
+
+ if (t4_read_reg(adap, CIM_HOST_ACC_CTRL) & HOSTBUSY)
+ return -EBUSY;
+
+ for ( ; !ret && n--; addr += 4) {
+ t4_write_reg(adap, CIM_HOST_ACC_CTRL, addr);
+ ret = t4_wait_op_done(adap, CIM_HOST_ACC_CTRL, HOSTBUSY,
+ 0, 5, 2);
+ if (!ret)
+ *valp++ = t4_read_reg(adap, CIM_HOST_ACC_DATA);
+ }
+ return ret;
+}
+
+/**
+ * t4_cim_write - write a block into CIM internal address space
+ * @adap: the adapter
+ * @addr: the start address within the CIM address space
+ * @n: number of words to write
+ * @valp: set of values to write
+ *
+ * Writes a block of 4-byte words into the CIM intenal address space.
+ */
+int t4_cim_write(struct adapter *adap, unsigned int addr, unsigned int n,
+ const unsigned int *valp)
+{
+ int ret = 0;
+
+ if (t4_read_reg(adap, CIM_HOST_ACC_CTRL) & HOSTBUSY)
+ return -EBUSY;
+
+ for ( ; !ret && n--; addr += 4) {
+ t4_write_reg(adap, CIM_HOST_ACC_DATA, *valp++);
+ t4_write_reg(adap, CIM_HOST_ACC_CTRL, addr | HOSTWRITE);
+ ret = t4_wait_op_done(adap, CIM_HOST_ACC_CTRL, HOSTBUSY,
+ 0, 5, 2);
+ }
+ return ret;
+}
+
+static int t4_cim_write1(struct adapter *adap, unsigned int addr,
+ unsigned int val)
+{
+ return t4_cim_write(adap, addr, 1, &val);
+}
+
+/**
+ * t4_cim_read_la - read CIM LA capture buffer
+ * @adap: the adapter
+ * @la_buf: where to store the LA data
+ * @wrptr: the HW write pointer within the capture buffer
+ *
+ * Reads the contents of the CIM LA buffer with the most recent entry at
+ * the end of the returned data and with the entry at @wrptr first.
+ * We try to leave the LA in the running state we find it in.
+ */
+int t4_cim_read_la(struct adapter *adap, u32 *la_buf, unsigned int *wrptr)
+{
+ int i, ret;
+ unsigned int cfg, val, idx;
+
+ ret = t4_cim_read(adap, UP_UP_DBG_LA_CFG, 1, &cfg);
+ if (ret)
+ return ret;
+
+ if (cfg & UPDBGLAEN) { /* LA is running, freeze it */
+ ret = t4_cim_write1(adap, UP_UP_DBG_LA_CFG, 0);
+ if (ret)
+ return ret;
+ }
+
+ ret = t4_cim_read(adap, UP_UP_DBG_LA_CFG, 1, &val);
+ if (ret)
+ goto restart;
+
+ idx = UPDBGLAWRPTR_GET(val);
+ if (wrptr)
+ *wrptr = idx;
+
+ for (i = 0; i < adap->params.cim_la_size; i++) {
+ ret = t4_cim_write1(adap, UP_UP_DBG_LA_CFG,
+ UPDBGLARDPTR(idx) | UPDBGLARDEN);
+ if (ret)
+ break;
+ ret = t4_cim_read(adap, UP_UP_DBG_LA_CFG, 1, &val);
+ if (ret)
+ break;
+ if (val & UPDBGLARDEN) {
+ ret = -ETIMEDOUT;
+ break;
+ }
+ ret = t4_cim_read(adap, UP_UP_DBG_LA_DATA, 1, &la_buf[i]);
+ if (ret)
+ break;
+ idx = (idx + 1) & UPDBGLARDPTR_MASK;
+ }
+restart:
+ if (cfg & UPDBGLAEN) {
+ int r = t4_cim_write1(adap, UP_UP_DBG_LA_CFG,
+ cfg & ~UPDBGLARDEN);
+ if (!ret)
+ ret = r;
+ }
+ return ret;
+}
diff --git a/drivers/net/ethernet/chelsio/cxgb4/t4_hw.h b/drivers/net/ethernet/chelsio/cxgb4/t4_hw.h
index 5e5eee6..bcc925b 100644
--- a/drivers/net/ethernet/chelsio/cxgb4/t4_hw.h
+++ b/drivers/net/ethernet/chelsio/cxgb4/t4_hw.h
@@ -56,6 +56,10 @@ enum {
};
enum {
+ CIMLA_SIZE = 2048, /* # of 32-bit words in CIM LA */
+};
+
+enum {
SF_PAGE_SIZE = 256, /* serial flash page size */
SF_SEC_SIZE = 64 * 1024, /* serial flash sector size */
};
diff --git a/drivers/net/ethernet/chelsio/cxgb4/t4_regs.h b/drivers/net/ethernet/chelsio/cxgb4/t4_regs.h
index a1024db..4f6d6fa 100644
--- a/drivers/net/ethernet/chelsio/cxgb4/t4_regs.h
+++ b/drivers/net/ethernet/chelsio/cxgb4/t4_regs.h
@@ -1331,4 +1331,18 @@
#define S_FT_VNID_ID_VLD 16
#define V_FT_VNID_ID_VLD(x) ((x) << S_FT_VNID_ID_VLD)
+/* registers for module CIM */
+#define CIM_HOST_ACC_CTRL 0x7b50
+#define CIM_HOST_ACC_DATA 0x7b54
+#define UP_UP_DBG_LA_CFG 0x140
+#define HOSTBUSY (1U << 17)
+#define HOSTWRITE (1U << 16)
+#define UPDBGLAEN (1U << 0)
+#define UPDBGLARDEN (1U << 1)
+#define UP_UP_DBG_LA_DATA 0x144
+#define UPDBGLARDPTR_MASK 0xfffU
+#define UPDBGLARDPTR(x) ((x) << 2)
+#define UPDBGLAWRPTR_GET(x) (((x) >> 16) & 0xfffU)
+#define UPDBGLACAPTPCONLY (1U << 30)
+
#endif /* __T4_REGS_H */
--
1.7.1
^ permalink raw reply related
* [PATCH net-next 1/4] cxgb4: Add cxgb4_debugfs.c and devlog support
From: Hariprasad Shenai @ 2014-10-16 1:22 UTC (permalink / raw)
To: netdev; +Cc: davem, leedom, kumaras, nirranjan, santosh, anish,
Hariprasad Shenai
In-Reply-To: <1413422524-14054-1-git-send-email-hariprasad@chelsio.com>
All the debugfs stuff is moved to cxgb4_debugfs.c.
Also add support for device log entry in debugfs
Signed-off-by: Hariprasad Shenai <hariprasad@chelsio.com>
---
drivers/net/ethernet/chelsio/cxgb4/Makefile | 2 +-
drivers/net/ethernet/chelsio/cxgb4/cxgb4.h | 9 +
drivers/net/ethernet/chelsio/cxgb4/cxgb4_debugfs.c | 358 ++++++++++++++++++++
drivers/net/ethernet/chelsio/cxgb4/cxgb4_debugfs.h | 52 +++
drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c | 121 ++-----
drivers/net/ethernet/chelsio/cxgb4/t4_hw.h | 12 +
drivers/net/ethernet/chelsio/cxgb4/t4fw_api.h | 72 ++++
7 files changed, 532 insertions(+), 94 deletions(-)
create mode 100644 drivers/net/ethernet/chelsio/cxgb4/cxgb4_debugfs.c
create mode 100644 drivers/net/ethernet/chelsio/cxgb4/cxgb4_debugfs.h
diff --git a/drivers/net/ethernet/chelsio/cxgb4/Makefile b/drivers/net/ethernet/chelsio/cxgb4/Makefile
index 1df65c9..de34d4b 100644
--- a/drivers/net/ethernet/chelsio/cxgb4/Makefile
+++ b/drivers/net/ethernet/chelsio/cxgb4/Makefile
@@ -4,5 +4,5 @@
obj-$(CONFIG_CHELSIO_T4) += cxgb4.o
-cxgb4-objs := cxgb4_main.o l2t.o t4_hw.o sge.o
+cxgb4-objs := cxgb4_main.o l2t.o t4_hw.o sge.o cxgb4_debugfs.o
cxgb4-$(CONFIG_CHELSIO_T4_DCB) += cxgb4_dcb.o
diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4.h b/drivers/net/ethernet/chelsio/cxgb4/cxgb4.h
index 410ed58..081b7ef 100644
--- a/drivers/net/ethernet/chelsio/cxgb4/cxgb4.h
+++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4.h
@@ -284,10 +284,18 @@ enum chip_type {
T5_LAST_REV = T5_A1,
};
+struct devlog_params {
+ u32 memtype; /* which memory (EDC0, EDC1, MC) */
+ u32 start; /* start of log in firmware memory */
+ u32 size; /* size of log */
+};
+
struct adapter_params {
struct tp_params tp;
struct vpd_params vpd;
struct pci_params pci;
+ struct devlog_params devlog;
+ enum pcie_memwin drv_memwin;
unsigned int sf_size; /* serial flash size in bytes */
unsigned int sf_nsec; /* # of flash sectors */
@@ -1083,4 +1091,5 @@ void t4_db_dropped(struct adapter *adapter);
int t4_fwaddrspace_write(struct adapter *adap, unsigned int mbox,
u32 addr, u32 val);
void t4_sge_decode_idma_state(struct adapter *adapter, int state);
+void t4_free_mem(void *addr);
#endif /* __CXGB4_H__ */
diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_debugfs.c b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_debugfs.c
new file mode 100644
index 0000000..0704af5
--- /dev/null
+++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_debugfs.c
@@ -0,0 +1,358 @@
+/*
+ * This file is part of the Chelsio T4 Ethernet driver for Linux.
+ *
+ * Copyright (c) 2003-2014 Chelsio Communications, Inc. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses. You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ * Redistribution and use in source and binary forms, with or
+ * without modification, are permitted provided that the following
+ * conditions are met:
+ *
+ * - Redistributions of source code must retain the above
+ * copyright notice, this list of conditions and the following
+ * disclaimer.
+ *
+ * - Redistributions in binary form must reproduce the above
+ * copyright notice, this list of conditions and the following
+ * disclaimer in the documentation and/or other materials
+ * provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+
+#include <linux/seq_file.h>
+#include <linux/debugfs.h>
+#include <linux/string_helpers.h>
+#include <linux/sort.h>
+
+#include "cxgb4.h"
+#include "t4_regs.h"
+#include "t4fw_api.h"
+#include "cxgb4_debugfs.h"
+#include "l2t.h"
+
+/* Firmware Device Log dump.
+ */
+static const char * const devlog_level_strings[] = {
+ [FW_DEVLOG_LEVEL_EMERG] = "EMERG",
+ [FW_DEVLOG_LEVEL_CRIT] = "CRIT",
+ [FW_DEVLOG_LEVEL_ERR] = "ERR",
+ [FW_DEVLOG_LEVEL_NOTICE] = "NOTICE",
+ [FW_DEVLOG_LEVEL_INFO] = "INFO",
+ [FW_DEVLOG_LEVEL_DEBUG] = "DEBUG"
+};
+
+static const char * const devlog_facility_strings[] = {
+ [FW_DEVLOG_FACILITY_CORE] = "CORE",
+ [FW_DEVLOG_FACILITY_SCHED] = "SCHED",
+ [FW_DEVLOG_FACILITY_TIMER] = "TIMER",
+ [FW_DEVLOG_FACILITY_RES] = "RES",
+ [FW_DEVLOG_FACILITY_HW] = "HW",
+ [FW_DEVLOG_FACILITY_FLR] = "FLR",
+ [FW_DEVLOG_FACILITY_DMAQ] = "DMAQ",
+ [FW_DEVLOG_FACILITY_PHY] = "PHY",
+ [FW_DEVLOG_FACILITY_MAC] = "MAC",
+ [FW_DEVLOG_FACILITY_PORT] = "PORT",
+ [FW_DEVLOG_FACILITY_VI] = "VI",
+ [FW_DEVLOG_FACILITY_FILTER] = "FILTER",
+ [FW_DEVLOG_FACILITY_ACL] = "ACL",
+ [FW_DEVLOG_FACILITY_TM] = "TM",
+ [FW_DEVLOG_FACILITY_QFC] = "QFC",
+ [FW_DEVLOG_FACILITY_DCB] = "DCB",
+ [FW_DEVLOG_FACILITY_ETH] = "ETH",
+ [FW_DEVLOG_FACILITY_OFLD] = "OFLD",
+ [FW_DEVLOG_FACILITY_RI] = "RI",
+ [FW_DEVLOG_FACILITY_ISCSI] = "ISCSI",
+ [FW_DEVLOG_FACILITY_FCOE] = "FCOE",
+ [FW_DEVLOG_FACILITY_FOISCSI] = "FOISCSI",
+ [FW_DEVLOG_FACILITY_FOFCOE] = "FOFCOE"
+};
+
+/* Information gathered by Device Log Open routine for the display routine.
+ */
+struct devlog_info {
+ unsigned int nentries; /* number of entries in log[] */
+ unsigned int first; /* first [temporal] entry in log[] */
+ struct fw_devlog_e log[0]; /* Firmware Device Log */
+};
+
+/* Dump a Firmaware Device Log entry.
+ */
+static int devlog_show(struct seq_file *seq, void *v)
+{
+ if (v == SEQ_START_TOKEN)
+ seq_printf(seq, "%10s %15s %8s %8s %s\n",
+ "Seq#", "Tstamp", "Level", "Facility", "Message");
+ else {
+ struct devlog_info *dinfo = seq->private;
+ int fidx = (uintptr_t)v - 2;
+ unsigned long index;
+ struct fw_devlog_e *e;
+
+ /* Get a pointer to the log entry to display. Skip unused log
+ * entries.
+ */
+ index = dinfo->first + fidx;
+ if (index >= dinfo->nentries)
+ index -= dinfo->nentries;
+ e = &dinfo->log[index];
+ if (e->timestamp == 0)
+ return 0;
+
+ /* Print the message. This depends on the firmware using
+ * exactly the same formating strings as the kernel so we may
+ * eventually have to put a format interpreter in here ...
+ */
+ seq_printf(seq, "%10d %15llu %8s %8s ",
+ e->seqno, e->timestamp,
+ (e->level < ARRAY_SIZE(devlog_level_strings)
+ ? devlog_level_strings[e->level]
+ : "UNKNOWN"),
+ (e->facility < ARRAY_SIZE(devlog_facility_strings)
+ ? devlog_facility_strings[e->facility]
+ : "UNKNOWN"));
+ seq_printf(seq, e->fmt, e->params[0], e->params[1],
+ e->params[2], e->params[3], e->params[4],
+ e->params[5], e->params[6], e->params[7]);
+ }
+
+ return 0;
+}
+
+/* Sequential File Operations for Device Log.
+ */
+static inline void *devlog_get_idx(struct devlog_info *dinfo, loff_t pos)
+{
+ if (pos > dinfo->nentries)
+ return NULL;
+
+ return (void *)(uintptr_t)(pos + 1);
+}
+
+static void *devlog_start(struct seq_file *seq, loff_t *pos)
+{
+ struct devlog_info *dinfo = seq->private;
+
+ return (*pos
+ ? devlog_get_idx(dinfo, *pos)
+ : SEQ_START_TOKEN);
+}
+
+static void *devlog_next(struct seq_file *seq, void *v, loff_t *pos)
+{
+ struct devlog_info *dinfo = seq->private;
+
+ (*pos)++;
+ return devlog_get_idx(dinfo, *pos);
+}
+
+static void devlog_stop(struct seq_file *seq, void *v)
+{
+}
+
+static const struct seq_operations devlog_seq_ops = {
+ .start = devlog_start,
+ .next = devlog_next,
+ .stop = devlog_stop,
+ .show = devlog_show
+};
+
+/* Set up for reading the firmware's device log. We read the entire log here
+ * and then display it incrementally in devlog_show().
+ */
+static int devlog_open(struct inode *inode, struct file *file)
+{
+ struct adapter *adap = inode->i_private;
+ struct devlog_params *dparams = &adap->params.devlog;
+ struct devlog_info *dinfo;
+ unsigned int index;
+ u32 fseqno;
+ int ret;
+
+ /* If we don't know where the log is we can't do anything.
+ */
+ if (dparams->start == 0)
+ return -ENXIO;
+
+ /* Allocate the space to read in the firmware's device log and set up
+ * for the iterated call to our display function.
+ */
+ dinfo = __seq_open_private(file, &devlog_seq_ops,
+ sizeof(*dinfo) + dparams->size);
+ if (dinfo == NULL)
+ return -ENOMEM;
+
+ /* Record the basic log buffer information and read in the raw log.
+ */
+ dinfo->nentries = (dparams->size / sizeof(struct fw_devlog_e));
+ dinfo->first = 0;
+ spin_lock(&adap->win0_lock);
+ ret = t4_memory_rw(adap, adap->params.drv_memwin, dparams->memtype,
+ dparams->start, dparams->size, (__be32 *)dinfo->log,
+ T4_MEMORY_READ);
+ spin_unlock(&adap->win0_lock);
+ if (ret) {
+ seq_release_private(inode, file);
+ return ret;
+ }
+
+ /* Translate log multi-byte integral elements into host native format
+ * and determine where the first entry in the log is.
+ */
+ for (fseqno = ~((u32)0), index = 0; index < dinfo->nentries; index++) {
+ struct fw_devlog_e *e = &dinfo->log[index];
+ int i;
+ __u32 seqno;
+
+ if (e->timestamp == 0)
+ continue;
+
+ e->timestamp = (__force __be64)be64_to_cpu(e->timestamp);
+ seqno = be32_to_cpu(e->seqno);
+ for (i = 0; i < 8; i++)
+ e->params[i] =
+ (__force __be32)be32_to_cpu(e->params[i]);
+
+ if (seqno < fseqno) {
+ fseqno = seqno;
+ dinfo->first = index;
+ }
+ }
+
+ return 0;
+}
+
+static const struct file_operations devlog_fops = {
+ .owner = THIS_MODULE,
+ .open = devlog_open,
+ .read = seq_read,
+ .llseek = seq_lseek,
+ .release = seq_release_private
+};
+
+static ssize_t mem_read(struct file *file, char __user *buf, size_t count,
+ loff_t *ppos)
+{
+ loff_t pos = *ppos;
+ loff_t avail = file_inode(file)->i_size;
+ unsigned int mem = (uintptr_t)file->private_data & 3;
+ struct adapter *adap = file->private_data - mem;
+ __be32 *data;
+ int ret;
+
+ if (pos < 0)
+ return -EINVAL;
+ if (pos >= avail)
+ return 0;
+ if (count > avail - pos)
+ count = avail - pos;
+
+ data = t4_alloc_mem(count);
+ if (!data)
+ return -ENOMEM;
+
+ spin_lock(&adap->win0_lock);
+ ret = t4_memory_rw(adap, 0, mem, pos, count, data, T4_MEMORY_READ);
+ spin_unlock(&adap->win0_lock);
+ if (ret) {
+ t4_free_mem(data);
+ return ret;
+ }
+ ret = copy_to_user(buf, data, count);
+
+ t4_free_mem(data);
+ if (ret)
+ return -EFAULT;
+
+ *ppos = pos + count;
+ return count;
+}
+
+static const struct file_operations mem_debugfs_fops = {
+ .owner = THIS_MODULE,
+ .open = simple_open,
+ .read = mem_read,
+ .llseek = default_llseek,
+};
+
+static void add_debugfs_mem(struct adapter *adap, const char *name,
+ unsigned int idx, unsigned int size_mb)
+{
+ struct dentry *de;
+
+ de = debugfs_create_file(name, S_IRUSR, adap->debugfs_root,
+ (void *)adap + idx, &mem_debugfs_fops);
+ if (de && de->d_inode)
+ de->d_inode->i_size = size_mb << 20;
+}
+
+/* Add an array of Debug FS files.
+ */
+void add_debugfs_files(struct adapter *adap,
+ struct t4_debugfs_entry *files,
+ unsigned int nfiles)
+{
+ int i;
+
+ /* debugfs support is best effort */
+ for (i = 0; i < nfiles; i++)
+ debugfs_create_file(files[i].name, files[i].mode,
+ adap->debugfs_root,
+ (void *)adap + files[i].data,
+ files[i].ops);
+}
+
+int t4_setup_debugfs(struct adapter *adap)
+{
+ int i;
+ u32 size;
+
+ static struct t4_debugfs_entry t4_debugfs_files[] = {
+ { "devlog", &devlog_fops, S_IRUSR, 0 },
+ { "l2t", &t4_l2t_fops, S_IRUSR, 0},
+ };
+
+ add_debugfs_files(adap,
+ t4_debugfs_files,
+ ARRAY_SIZE(t4_debugfs_files));
+
+ i = t4_read_reg(adap, MA_TARGET_MEM_ENABLE);
+ if (i & EDRAM0_ENABLE) {
+ size = t4_read_reg(adap, MA_EDRAM0_BAR);
+ add_debugfs_mem(adap, "edc0", MEM_EDC0, EDRAM_SIZE_GET(size));
+ }
+ if (i & EDRAM1_ENABLE) {
+ size = t4_read_reg(adap, MA_EDRAM1_BAR);
+ add_debugfs_mem(adap, "edc1", MEM_EDC1, EDRAM_SIZE_GET(size));
+ }
+ if (is_t4(adap->params.chip)) {
+ size = t4_read_reg(adap, MA_EXT_MEMORY_BAR);
+ if (i & EXT_MEM_ENABLE)
+ add_debugfs_mem(adap, "mc", MEM_MC,
+ EXT_MEM_SIZE_GET(size));
+ } else {
+ if (i & EXT_MEM_ENABLE) {
+ size = t4_read_reg(adap, MA_EXT_MEMORY_BAR);
+ add_debugfs_mem(adap, "mc0", MEM_MC0,
+ EXT_MEM_SIZE_GET(size));
+ }
+ if (i & EXT_MEM1_ENABLE) {
+ size = t4_read_reg(adap, MA_EXT_MEMORY1_BAR);
+ add_debugfs_mem(adap, "mc1", MEM_MC1,
+ EXT_MEM_SIZE_GET(size));
+ }
+ }
+ return 0;
+}
diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_debugfs.h b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_debugfs.h
new file mode 100644
index 0000000..a3d8867
--- /dev/null
+++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_debugfs.h
@@ -0,0 +1,52 @@
+/*
+ * This file is part of the Chelsio T4 Ethernet driver for Linux.
+ *
+ * Copyright (c) 2003-2014 Chelsio Communications, Inc. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses. You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ * Redistribution and use in source and binary forms, with or
+ * without modification, are permitted provided that the following
+ * conditions are met:
+ *
+ * - Redistributions of source code must retain the above
+ * copyright notice, this list of conditions and the following
+ * disclaimer.
+ *
+ * - Redistributions in binary form must reproduce the above
+ * copyright notice, this list of conditions and the following
+ * disclaimer in the documentation and/or other materials
+ * provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+
+#ifndef __CXGB4_DEBUGFS_H
+#define __CXGB4_DEBUGFS_H
+
+#include <linux/export.h>
+
+struct t4_debugfs_entry {
+ const char *name;
+ const struct file_operations *ops;
+ mode_t mode;
+ unsigned char data;
+};
+
+int t4_setup_debugfs(struct adapter *adap);
+void add_debugfs_files(struct adapter *adap,
+ struct t4_debugfs_entry *files,
+ unsigned int nfiles);
+
+#endif
diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c
index 5b38e95..645defa 100644
--- a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c
+++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c
@@ -68,6 +68,7 @@
#include "t4_msg.h"
#include "t4fw_api.h"
#include "cxgb4_dcb.h"
+#include "cxgb4_debugfs.h"
#include "l2t.h"
#include <../drivers/net/bonding/bonding.h>
@@ -1283,7 +1284,7 @@ void *t4_alloc_mem(size_t size)
/*
* Free memory allocated through alloc_mem().
*/
-static void t4_free_mem(void *addr)
+void t4_free_mem(void *addr)
{
if (is_vmalloc_addr(addr))
vfree(addr);
@@ -3113,103 +3114,12 @@ static const struct ethtool_ops cxgb_ethtool_ops = {
.flash_device = set_flash,
};
-/*
- * debugfs support
- */
-static ssize_t mem_read(struct file *file, char __user *buf, size_t count,
- loff_t *ppos)
-{
- loff_t pos = *ppos;
- loff_t avail = file_inode(file)->i_size;
- unsigned int mem = (uintptr_t)file->private_data & 3;
- struct adapter *adap = file->private_data - mem;
- __be32 *data;
- int ret;
-
- if (pos < 0)
- return -EINVAL;
- if (pos >= avail)
- return 0;
- if (count > avail - pos)
- count = avail - pos;
-
- data = t4_alloc_mem(count);
- if (!data)
- return -ENOMEM;
-
- spin_lock(&adap->win0_lock);
- ret = t4_memory_rw(adap, 0, mem, pos, count, data, T4_MEMORY_READ);
- spin_unlock(&adap->win0_lock);
- if (ret) {
- t4_free_mem(data);
- return ret;
- }
- ret = copy_to_user(buf, data, count);
-
- t4_free_mem(data);
- if (ret)
- return -EFAULT;
-
- *ppos = pos + count;
- return count;
-}
-
-static const struct file_operations mem_debugfs_fops = {
- .owner = THIS_MODULE,
- .open = simple_open,
- .read = mem_read,
- .llseek = default_llseek,
-};
-
-static void add_debugfs_mem(struct adapter *adap, const char *name,
- unsigned int idx, unsigned int size_mb)
-{
- struct dentry *de;
-
- de = debugfs_create_file(name, S_IRUSR, adap->debugfs_root,
- (void *)adap + idx, &mem_debugfs_fops);
- if (de && de->d_inode)
- de->d_inode->i_size = size_mb << 20;
-}
-
static int setup_debugfs(struct adapter *adap)
{
- int i;
- u32 size;
-
if (IS_ERR_OR_NULL(adap->debugfs_root))
return -1;
- i = t4_read_reg(adap, MA_TARGET_MEM_ENABLE);
- if (i & EDRAM0_ENABLE) {
- size = t4_read_reg(adap, MA_EDRAM0_BAR);
- add_debugfs_mem(adap, "edc0", MEM_EDC0, EDRAM_SIZE_GET(size));
- }
- if (i & EDRAM1_ENABLE) {
- size = t4_read_reg(adap, MA_EDRAM1_BAR);
- add_debugfs_mem(adap, "edc1", MEM_EDC1, EDRAM_SIZE_GET(size));
- }
- if (is_t4(adap->params.chip)) {
- size = t4_read_reg(adap, MA_EXT_MEMORY_BAR);
- if (i & EXT_MEM_ENABLE)
- add_debugfs_mem(adap, "mc", MEM_MC,
- EXT_MEM_SIZE_GET(size));
- } else {
- if (i & EXT_MEM_ENABLE) {
- size = t4_read_reg(adap, MA_EXT_MEMORY_BAR);
- add_debugfs_mem(adap, "mc0", MEM_MC0,
- EXT_MEM_SIZE_GET(size));
- }
- if (i & EXT_MEM1_ENABLE) {
- size = t4_read_reg(adap, MA_EXT_MEMORY1_BAR);
- add_debugfs_mem(adap, "mc1", MEM_MC1,
- EXT_MEM_SIZE_GET(size));
- }
- }
- if (adap->l2t)
- debugfs_create_file("l2t", S_IRUSR, adap->debugfs_root, adap,
- &t4_l2t_fops);
- return 0;
+ return t4_setup_debugfs(adap);
}
/*
@@ -5656,6 +5566,8 @@ static int adap_init0(struct adapter *adap)
enum dev_state state;
u32 params[7], val[7];
struct fw_caps_config_cmd caps_cmd;
+ struct fw_devlog_cmd devlog_cmd;
+ u32 devlog_meminfo;
int reset = 1;
/*
@@ -5744,6 +5656,29 @@ static int adap_init0(struct adapter *adap)
if (ret < 0)
goto bye;
+ /* Read firmware device log parameters. We really need to find a way
+ * to get these parameters initialized with some default values (which
+ * are likely to be correct) for the case where we either don't
+ * attache to the firmware or it's crashed when we probe the adapter.
+ * That way we'll still be able to perform early firmware startup
+ * debugging ... If the request to get the Firmware's Device Log
+ * parameters fails, we'll live so we don't make that a fatal error.
+ */
+ memset(&devlog_cmd, 0, sizeof(devlog_cmd));
+ devlog_cmd.op_to_write = htonl(FW_CMD_OP(FW_DEVLOG_CMD) |
+ FW_CMD_REQUEST | FW_CMD_READ);
+ devlog_cmd.retval_len16 = htonl(FW_LEN16(devlog_cmd));
+ ret = t4_wr_mbox(adap, adap->mbox, &devlog_cmd, sizeof(devlog_cmd),
+ &devlog_cmd);
+ if (ret == 0) {
+ devlog_meminfo =
+ ntohl(devlog_cmd.memtype_devlog_memaddr16_devlog);
+ adap->params.devlog.memtype =
+ FW_DEVLOG_CMD_MEMTYPE_DEVLOG_GET(devlog_meminfo);
+ adap->params.devlog.start =
+ FW_DEVLOG_CMD_MEMADDR16_DEVLOG_GET(devlog_meminfo) << 4;
+ adap->params.devlog.size = ntohl(devlog_cmd.memsize_devlog);
+ }
/*
* Find out what ports are available to us. Note that we need to do
* this before calling adap_init0_no_config() since it needs nports
diff --git a/drivers/net/ethernet/chelsio/cxgb4/t4_hw.h b/drivers/net/ethernet/chelsio/cxgb4/t4_hw.h
index c19a90e..5e5eee6 100644
--- a/drivers/net/ethernet/chelsio/cxgb4/t4_hw.h
+++ b/drivers/net/ethernet/chelsio/cxgb4/t4_hw.h
@@ -110,6 +110,18 @@ enum {
SGE_INGPADBOUNDARY_SHIFT = 5,/* ingress queue pad boundary */
};
+/* PCI-e memory window access */
+enum pcie_memwin {
+ MEMWIN_NIC = 0,
+ MEMWIN_RSVD1 = 1,
+ MEMWIN_RSVD2 = 2,
+ MEMWIN_RDMA = 3,
+ MEMWIN_RSVD4 = 4,
+ MEMWIN_FOISCSI = 5,
+ MEMWIN_CSIOSTOR = 6,
+ MEMWIN_RSVD7 = 7,
+};
+
struct sge_qstat { /* data written to SGE queue status entries */
__be32 qid;
__be16 cidx;
diff --git a/drivers/net/ethernet/chelsio/cxgb4/t4fw_api.h b/drivers/net/ethernet/chelsio/cxgb4/t4fw_api.h
index 3409756..e9ceb3a 100644
--- a/drivers/net/ethernet/chelsio/cxgb4/t4fw_api.h
+++ b/drivers/net/ethernet/chelsio/cxgb4/t4fw_api.h
@@ -618,6 +618,7 @@ enum fw_cmd_opcodes {
FW_RSS_IND_TBL_CMD = 0x20,
FW_RSS_GLB_CONFIG_CMD = 0x22,
FW_RSS_VI_CONFIG_CMD = 0x23,
+ FW_DEVLOG_CMD = 0x25,
FW_CLIP_CMD = 0x28,
FW_LASTC2E_CMD = 0x40,
FW_ERROR_CMD = 0x80,
@@ -2279,4 +2280,75 @@ enum fw_hdr_flags {
FW_HDR_FLAGS_RESET_HALT = 0x00000001,
};
+/* length of the formatting string */
+#define FW_DEVLOG_FMT_LEN 192
+
+/* maximum number of the formatting string parameters */
+#define FW_DEVLOG_FMT_PARAMS_NUM 8
+
+/* priority levels */
+enum fw_devlog_level {
+ FW_DEVLOG_LEVEL_EMERG = 0x0,
+ FW_DEVLOG_LEVEL_CRIT = 0x1,
+ FW_DEVLOG_LEVEL_ERR = 0x2,
+ FW_DEVLOG_LEVEL_NOTICE = 0x3,
+ FW_DEVLOG_LEVEL_INFO = 0x4,
+ FW_DEVLOG_LEVEL_DEBUG = 0x5,
+ FW_DEVLOG_LEVEL_MAX = 0x5,
+};
+
+/* facilities that may send a log message */
+enum fw_devlog_facility {
+ FW_DEVLOG_FACILITY_CORE = 0x00,
+ FW_DEVLOG_FACILITY_CF = 0x01,
+ FW_DEVLOG_FACILITY_SCHED = 0x02,
+ FW_DEVLOG_FACILITY_TIMER = 0x04,
+ FW_DEVLOG_FACILITY_RES = 0x06,
+ FW_DEVLOG_FACILITY_HW = 0x08,
+ FW_DEVLOG_FACILITY_FLR = 0x10,
+ FW_DEVLOG_FACILITY_DMAQ = 0x12,
+ FW_DEVLOG_FACILITY_PHY = 0x14,
+ FW_DEVLOG_FACILITY_MAC = 0x16,
+ FW_DEVLOG_FACILITY_PORT = 0x18,
+ FW_DEVLOG_FACILITY_VI = 0x1A,
+ FW_DEVLOG_FACILITY_FILTER = 0x1C,
+ FW_DEVLOG_FACILITY_ACL = 0x1E,
+ FW_DEVLOG_FACILITY_TM = 0x20,
+ FW_DEVLOG_FACILITY_QFC = 0x22,
+ FW_DEVLOG_FACILITY_DCB = 0x24,
+ FW_DEVLOG_FACILITY_ETH = 0x26,
+ FW_DEVLOG_FACILITY_OFLD = 0x28,
+ FW_DEVLOG_FACILITY_RI = 0x2A,
+ FW_DEVLOG_FACILITY_ISCSI = 0x2C,
+ FW_DEVLOG_FACILITY_FCOE = 0x2E,
+ FW_DEVLOG_FACILITY_FOISCSI = 0x30,
+ FW_DEVLOG_FACILITY_FOFCOE = 0x32,
+ FW_DEVLOG_FACILITY_MAX = 0x32,
+};
+
+/* log message format */
+struct fw_devlog_e {
+ __be64 timestamp;
+ __be32 seqno;
+ __be16 reserved1;
+ __u8 level;
+ __u8 facility;
+ __u8 fmt[FW_DEVLOG_FMT_LEN];
+ __be32 params[FW_DEVLOG_FMT_PARAMS_NUM];
+ __be32 reserved3[4];
+};
+
+struct fw_devlog_cmd {
+ __be32 op_to_write;
+ __be32 retval_len16;
+ __u8 level;
+ __u8 r2[7];
+ __be32 memtype_devlog_memaddr16_devlog;
+ __be32 memsize_devlog;
+ __be32 r3[2];
+};
+
+#define FW_DEVLOG_CMD_MEMTYPE_DEVLOG_GET(x) (((x) >> 28) & 0xf)
+#define FW_DEVLOG_CMD_MEMADDR16_DEVLOG_GET(x) (((x) >> 0) & 0xfffffff)
+
#endif /* _T4FW_INTERFACE_H_ */
--
1.7.1
^ permalink raw reply related
* [PATCH net-next 0/4] Add support for few debugfs entries
From: Hariprasad Shenai @ 2014-10-16 1:22 UTC (permalink / raw)
To: netdev; +Cc: davem, leedom, kumaras, nirranjan, santosh, anish,
Hariprasad Shenai
Hi,
This patch series adds a new file for debugfs and support for devlog, cim_la,
cim_la_qcfg and mps_tcam debugfs entries.
The patches series is created against 'net-next' tree.
And includes patches on cxgb4 driver.
We have included all the maintainers of respective drivers. Kindly review the
change and let us know in case of any review comments.
Thanks
Hariprasad Shenai (4):
cxgb4: Add cxgb4_debugfs.c and devlog support
cxgb4: Add support for cim_la entry in debugfs
cxgb4: Add support for cim_la_qcfg entry in debugfs
cxgb4: Add support for mps_tcam debugfs
drivers/net/ethernet/chelsio/cxgb4/Makefile | 2 +-
drivers/net/ethernet/chelsio/cxgb4/cxgb4.h | 17 +
drivers/net/ethernet/chelsio/cxgb4/cxgb4_debugfs.c | 683 ++++++++++++++++++++
drivers/net/ethernet/chelsio/cxgb4/cxgb4_debugfs.h | 64 ++
drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c | 121 +---
drivers/net/ethernet/chelsio/cxgb4/t4_hw.c | 155 +++++
drivers/net/ethernet/chelsio/cxgb4/t4_hw.h | 19 +
drivers/net/ethernet/chelsio/cxgb4/t4_regs.h | 52 ++
drivers/net/ethernet/chelsio/cxgb4/t4fw_api.h | 72 ++
9 files changed, 1091 insertions(+), 94 deletions(-)
create mode 100644 drivers/net/ethernet/chelsio/cxgb4/cxgb4_debugfs.c
create mode 100644 drivers/net/ethernet/chelsio/cxgb4/cxgb4_debugfs.h
^ permalink raw reply
page: next (older) | prev (newer) | latest
- recent:[subjects (threaded)|topics (new)|topics (active)]
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox