* Re: [PATCH NET-2.6 1/1] qlcnic: limit skb frags for non tso packet
From: Amit Salecha @ 2011-04-13 15:59 UTC (permalink / raw)
To: Greg KH
Cc: netdev@vger.kernel.org, Anirban Chakraborty, David Miller,
Ameen Rahman, stable@kernel.org
In-Reply-To: <20110413140307.GA7426@kroah.com>
> On Wed, Apr 13, 2011 at 12:56:56AM -0500, Amit Salecha wrote:
> > > On Tue, Apr 12, 2011 at 09:01:13PM -0500, Amit Salecha wrote:
> > > >
> > > > This message and any attached documents contain information from
> > > QLogic Corporation or its wholly-owned subsidiaries that may be
> > > confidential. If you are not the intended recipient, you may not
> read,
> > > copy, distribute, or use this information. If you have received
> this
> > > transmission in error, please notify the sender immediately by
> reply e-
> > > mail and then delete this message.
> > >
> > > I have received this transmission in error.
> > >
> > > Please remove this from your footer, otherwise we can not accept
> any
> > > emails sent from you as actually being allowed to contribute to the
> > > kernel properly :(
> > >
> > I have send version two of this patch, that doesn't have this footer.
> > Please discard this one.
> >
> > This message and any attached documents contain information from
> QLogic Corporation or its wholly-owned subsidiaries that may be
> confidential. If you are not the intended recipient, you may not read,
> copy, distribute, or use this information. If you have received this
> transmission in error, please notify the sender immediately by reply e-
> mail and then delete this message.
>
> Your footer is still present, please fix this.
Footer will present in my reply to this email. But footer should not be there in patches sent by me.
Can you verify patch version 2 again ? Here http://patchwork.ozlabs.org/patch/90938/ I don't see any footer.
If you see footer with patch version 2, please send me that.
-Amit
This message and any attached documents contain information from QLogic Corporation or its wholly-owned subsidiaries that may be confidential. If you are not the intended recipient, you may not read, copy, distribute, or use this information. If you have received this transmission in error, please notify the sender immediately by reply e-mail and then delete this message.
_______________________________________________
stable mailing list
stable@linux.kernel.org
http://linux.kernel.org/mailman/listinfo/stable
^ permalink raw reply
* Re: [PATCH NET-2.6 1/1] qlcnic: limit skb frags for non tso packet
From: Greg KH @ 2011-04-13 16:15 UTC (permalink / raw)
To: Amit Salecha
Cc: netdev@vger.kernel.org, Anirban Chakraborty, David Miller,
Ameen Rahman, stable@kernel.org
In-Reply-To: <99737F4847ED0A48AECC9F4A1974A4B80FD138407E@MNEXMB2.qlogic.org>
On Wed, Apr 13, 2011 at 10:59:56AM -0500, Amit Salecha wrote:
> > On Wed, Apr 13, 2011 at 12:56:56AM -0500, Amit Salecha wrote:
> > > > On Tue, Apr 12, 2011 at 09:01:13PM -0500, Amit Salecha wrote:
> > > > >
> > > > > This message and any attached documents contain information from
> > > > QLogic Corporation or its wholly-owned subsidiaries that may be
> > > > confidential. If you are not the intended recipient, you may not
> > read,
> > > > copy, distribute, or use this information. If you have received
> > this
> > > > transmission in error, please notify the sender immediately by
> > reply e-
> > > > mail and then delete this message.
> > > >
> > > > I have received this transmission in error.
> > > >
> > > > Please remove this from your footer, otherwise we can not accept
> > any
> > > > emails sent from you as actually being allowed to contribute to the
> > > > kernel properly :(
> > > >
> > > I have send version two of this patch, that doesn't have this footer.
> > > Please discard this one.
> > >
> > > This message and any attached documents contain information from
> > QLogic Corporation or its wholly-owned subsidiaries that may be
> > confidential. If you are not the intended recipient, you may not read,
> > copy, distribute, or use this information. If you have received this
> > transmission in error, please notify the sender immediately by reply e-
> > mail and then delete this message.
> >
> > Your footer is still present, please fix this.
>
>
> Footer will present in my reply to this email. But footer should not be there in patches sent by me.
> Can you verify patch version 2 again ? Here http://patchwork.ozlabs.org/patch/90938/ I don't see any footer.
> If you see footer with patch version 2, please send me that.
Your footer was not in your patch, correct. But it was in this email.
And that's the issue, you can't have that footer on emails you send to a
public list where you are going to be collaborating on a public project,
otherwise no one can use anything you say.
Now if you only think that people will just accept your patches, without
being able to have you participate in development and maintance of those
patches (which is required to be done through email), you are mistaken.
So please fix your email issue, otherwise it is not going to work.
Note, other people at qualcomm have fixed this, so you are not alone.
thanks,
greg k-h
_______________________________________________
stable mailing list
stable@linux.kernel.org
http://linux.kernel.org/mailman/listinfo/stable
^ permalink raw reply
* Re: [PATCH 2/4] [RFC rev2] virtio-net changes
From: Krishna Kumar2 @ 2011-04-13 16:20 UTC (permalink / raw)
To: Rusty Russell
Cc: anthony, arnd, avi, davem, eric.dumazet, horms, kvm, mst, netdev
In-Reply-To: <878vvf0wkd.fsf@rustcorp.com.au>
Hi Rusty,
Thanks for your feedback. I agree with all the changes, and will
make it and resubmit next.
thanks,
- KK
Rusty Russell <rusty@rustcorp.com.au> wrote on 04/13/2011 06:58:02 AM:
> Rusty Russell <rusty@rustcorp.com.au>
> 04/13/2011 06:58 AM
>
> To
>
> Krishna Kumar2/India/IBM@IBMIN, davem@davemloft.net, mst@redhat.com
>
> cc
>
> eric.dumazet@gmail.com, arnd@arndb.de, netdev@vger.kernel.org,
> horms@verge.net.au, avi@redhat.com, anthony@codemonkey.ws,
> kvm@vger.kernel.org, Krishna Kumar2/India/IBM@IBMIN
>
> Subject
>
> Re: [PATCH 2/4] [RFC rev2] virtio-net changes
>
> On Tue, 05 Apr 2011 20:38:52 +0530, Krishna Kumar <krkumar2@in.ibm.com>
wrote:
> > Implement mq virtio-net driver.
> >
> > Though struct virtio_net_config changes, it works with the old
> > qemu since the last element is not accessed unless qemu sets
> > VIRTIO_NET_F_MULTIQUEUE.
> >
> > Signed-off-by: Krishna Kumar <krkumar2@in.ibm.com>
>
> Hi Krishna!
>
> This change looks fairly solid, but I'd prefer it split into a few
> stages for clarity.
>
> The first patch should extract out the struct send_queue and struct
> receive_queue, even though there's still only one. The second patch
> can then introduce VIRTIO_NET_F_MULTIQUEUE.
>
> You could split into more parts if that makes sense, but I'd prefer to
> see the mechanical changes separate from the feature addition.
>
> > -struct virtnet_info {
> > - struct virtio_device *vdev;
> > - struct virtqueue *rvq, *svq, *cvq;
> > - struct net_device *dev;
> > +/* Internal representation of a send virtqueue */
> > +struct send_queue {
> > + /* Virtqueue associated with this send _queue */
> > + struct virtqueue *svq;
>
> You can simply call this vq now it's inside 'send_queue'.
>
> > +
> > + /* TX: fragments + linear part + virtio header */
> > + struct scatterlist tx_sg[MAX_SKB_FRAGS + 2];
>
> Similarly, this can just be sg.
>
> > +static void free_receive_bufs(struct virtnet_info *vi)
> > +{
> > + int i;
> > +
> > + for (i = 0; i < vi->numtxqs; i++) {
> > + BUG_ON(vi->rq[i] == NULL);
> > + while (vi->rq[i]->pages)
> > + __free_pages(get_a_page(vi->rq[i], GFP_KERNEL), 0);
> > + }
> > +}
>
> You can skip the BUG_ON(), since the next line will have the same effect.
>
> > +/* Free memory allocated for send and receive queues */
> > +static void free_rq_sq(struct virtnet_info *vi)
> > +{
> > + int i;
> > +
> > + if (vi->rq) {
> > + for (i = 0; i < vi->numtxqs; i++)
> > + kfree(vi->rq[i]);
> > + kfree(vi->rq);
> > + }
> > +
> > + if (vi->sq) {
> > + for (i = 0; i < vi->numtxqs; i++)
> > + kfree(vi->sq[i]);
> > + kfree(vi->sq);
> > + }
>
> This looks weird, even though it's correct.
>
> I think we need a better name than numtxqs and shorter than
> num_queue_pairs. Let's just use num_queues; sure, there are both tx and
> rq queues, but I still think it's pretty clear.
>
> > + for (i = 0; i < vi->numtxqs; i++) {
> > + struct virtqueue *svq = vi->sq[i]->svq;
> > +
> > + while (1) {
> > + buf = virtqueue_detach_unused_buf(svq);
> > + if (!buf)
> > + break;
> > + dev_kfree_skb(buf);
> > + }
> > + }
>
> I know this isn't your code, but it's ugly :)
>
> while ((buf = virtqueue_detach_unused_buf(svq)) != NULL)
> dev_kfree_skb(buf);
>
> > + for (i = 0; i < vi->numtxqs; i++) {
> > + struct virtqueue *rvq = vi->rq[i]->rvq;
> > +
> > + while (1) {
> > + buf = virtqueue_detach_unused_buf(rvq);
> > + if (!buf)
> > + break;
>
> Here too...
>
> > +#define MAX_DEVICE_NAME 16
>
> This isn't a good idea, see below.
>
> > +static int initialize_vqs(struct virtnet_info *vi, int numtxqs)
> > +{
> > + vq_callback_t **callbacks;
> > + struct virtqueue **vqs;
> > + int i, err = -ENOMEM;
> > + int totalvqs;
> > + char **names;
>
> This whole routine is really messy. How about doing find_vqs first,
> then have routines like setup_rxq(), setup_txq() and setup_controlq()
> would make this neater:
>
> static int setup_rxq(struct send_queue *sq, char *name);
>
> Also, use kasprintf() instead of kmalloc & sprintf.
>
> > +#if 1
> > + /* Allocate/initialize parameters for recv/send virtqueues */
>
> Why is this #if 1'd?
>
> I do prefer the #else method of doing two loops, myself (but use
> kasprintf).
>
> Cheers,
> Rusty.
^ permalink raw reply
* Re: [PATCH 0/4] [RFC rev2] Implement multiqueue (RX & TX) virtio-net
From: Krishna Kumar2 @ 2011-04-13 16:22 UTC (permalink / raw)
To: Avi Kivity
Cc: anthony, arnd, davem, eric.dumazet, horms, kvm, mst, netdev,
rusty
In-Reply-To: <4DA5904B.9090006@redhat.com>
Avi Kivity <avi@redhat.com> wrote on 04/13/2011 05:30:11 PM:
Hi Avi,
> > --------
> > 1. Reduce vectors for find_vqs().
> > 2. Make vhost changes minimal. For now, I have restricted the number of
> > vhost threads to 4. This can be either made unrestricted; or if the
> > userspace vhost works, it can be removed altogether.
> >
> > Please review and provide feedback. I am travelling a bit in the next
> > few days but will respond at the earliest.
>
> Do you have an update to the virtio-pci spec for this?
Not yet, will keep it in my TODO list.
thanks,
- KK
^ permalink raw reply
* Re: [PATCHv4 NEXT 1/1] net: ethtool support to configure number of channels
From: David Miller @ 2011-04-13 17:40 UTC (permalink / raw)
To: bhutchings
Cc: amit.salecha, netdev, ameen.rahman, sucheta.chakraborty,
anirban.chakraborty
In-Reply-To: <1302696197.5282.612.camel@localhost>
From: Ben Hutchings <bhutchings@solarflare.com>
Date: Wed, 13 Apr 2011 13:03:17 +0100
> It's marked as 'deferred'; not sure what that means:
>
> http://patchwork.ozlabs.org/patch/90166/
It means it's out of my TODO list until something happens.
Specifically Ben I'm waiting for you to provide some review and
feedback on this patch since in the past you were the one saying
what needed to be changed in this patch.
I can't tell if you're requests have been met unless you actually
provide an ACK, or more feedback.
^ permalink raw reply
* Re: [Bugme-new] [Bug 32832] New: shutdown(2) does not fully shut down socket any more
From: David Miller @ 2011-04-13 17:43 UTC (permalink / raw)
To: dbaluta; +Cc: eric.dumazet, akpm, netdev, bugzilla-daemon, bugme-daemon, kees
In-Reply-To: <BANLkTi=1wmQYRgHmO5SbPBiSwzKwVz4_ag@mail.gmail.com>
From: Daniel Baluta <dbaluta@ixiacom.com>
Date: Wed, 13 Apr 2011 14:57:18 +0300
> Cyril's use case looks suspect. I don't think that this is a good
> reason for reverting this commit.
I complete disagree.
Something that worked perfectly fine, probably for years, we broke.
We simply cannot do that, especially since we do not have a reasonable
alternative at this time.
Adding SO_REUSEPORT is a long range option, and not something that
will provide a fix for users right now.
So please don't even pretend to suggest that we shouldn't fix this
with a revert unless a simple, obvious, kernel fix presents itself.
^ permalink raw reply
* Re: [net-next-2.6] if_vlan.h: fix compile errors
From: David Miller @ 2011-04-13 17:46 UTC (permalink / raw)
To: jeffrey.t.kirsher; +Cc: jpirko, netdev
In-Reply-To: <BANLkTimn2GfWs3tmoh+GRX6vX6tH9FpEyg@mail.gmail.com>
From: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Date: Wed, 13 Apr 2011 04:03:57 -0700
> On Wed, Apr 13, 2011 at 03:57, Jiri Pirko <jpirko@redhat.com> wrote:
>> Thanks Jeff, this fix was already submitted by DaveM
>
> Ah, ok. I did not see anything in netdev patchworks and I did a
> recent pull of Dave's net-next tree so I knew I had the latest.
> Good to know it was caught, thanks Jiri.
Sorry, forgot to push before going to bed :-/
It's there now.
^ permalink raw reply
* Re: [PATCH] r8169: Be verbose when unable to load fw patch
From: François Romieu @ 2011-04-13 17:48 UTC (permalink / raw)
To: David Miller; +Cc: bp, romieu, linux-kernel, borislav.petkov, netdev
In-Reply-To: <20110412.162036.241924043.davem@davemloft.net>
On Tue, Apr 12, 2011 at 04:20:36PM -0700, David Miller wrote:
[...]
> Seems reasonable, Francois?
Yes, it makes sense.
The patch goes way beyond 80 cols but I'll do a round of
cleanups later anyway.
--
Ueimor
^ permalink raw reply
* Re: r8169 doesn't report link state correctly.
From: François Romieu @ 2011-04-13 17:48 UTC (permalink / raw)
To: Ben Greear; +Cc: netdev
In-Reply-To: <4DA35FEF.1030009@candelatech.com>
On Mon, Apr 11, 2011 at 01:09:19PM -0700, Ben Greear wrote:
> I notice that in kernel 2.6.38-wl, the realtek 8169 NIC doesn't
> report link down when in fact there is no cable connected. Instead,
> it shows 10Mbps half-duplex. I have no idea if this previously worked
> or not.
Thanks for the report. I'll try it tomorrow or friday.
--
Ueimor
^ permalink raw reply
* Re: [PATCH v2] net: r8169: convert to hw_features
From: François Romieu @ 2011-04-13 17:47 UTC (permalink / raw)
To: hayeswang
Cc: 'Michał Mirosław', netdev,
'David Dillow'
In-Reply-To: <BAFB047F568645FE911024B882E5C022@realtek.com.tw>
On Tue, Apr 12, 2011 at 10:10:20AM +0800, hayeswang wrote:
[...]
> Yes, all 8168 have the same Tx descriptors layout except for 8168B series.
Thanks Hayes.
I have started playing with the patch below on top of davem-next but
something in it is probably broken as enabling TSO kills the traffic
from several 10s of Mbytes per second to a few kbytes. I'll dig
further.
Subject: [PATCH] r8169: TSO fixes.
- the MSS value is actually contained in a 11 bits wide (0x7ff) field.
The extra bit in the former MSSMask did encompass the TSO command
bit ("LargeSend") as well (0xfff). Oops.
- the Tx descriptor layout is not the same through the whole chipset
family. The 8169 documentation, the 8168c documentation and Realtek's
drivers (8.020.00, 1.019.00, 6.014.00) highlight two layouts:
1. 8169, 8168 up to 8168b (included) and 8101
2. {8102e, 8168c} and beyond
- notwithstanding the "first descriptor" and "last descriptor" bits, the
same Tx descriptor content is enforced when a packet consists of several
descriptors. The chipsets are documented to require it.
Signed-off-by: Francois Romieu <romieu@fr.zoreil.com>
Cc: Hayes Wang <hayeswang@realtek.com>
---
drivers/net/r8169.c | 207 ++++++++++++++++++++++++++++++++++-----------------
1 files changed, 137 insertions(+), 70 deletions(-)
diff --git a/drivers/net/r8169.c b/drivers/net/r8169.c
index 058524f..61196a1 100644
--- a/drivers/net/r8169.c
+++ b/drivers/net/r8169.c
@@ -134,47 +134,52 @@ enum mac_version {
RTL_GIGA_MAC_VER_33 = 0x21, // 8168E
};
-#define _R(NAME,MAC,MASK) \
- { .name = NAME, .mac_version = MAC, .RxConfigMask = MASK }
+enum rtl_tx_desc_version {
+ RTL_TD_0 = 0,
+ RTL_TD_1 = 1,
+};
+
+#define _R(NAME,MAC,TD) \
+ { .name = NAME, .mac_version = MAC, .txd_version = TD }
static const struct {
const char *name;
u8 mac_version;
- u32 RxConfigMask; /* Clears the bits supported by this chip */
+ enum rtl_tx_desc_version txd_version;
} rtl_chip_info[] = {
- _R("RTL8169", RTL_GIGA_MAC_VER_01, 0xff7e1880), // 8169
- _R("RTL8169s", RTL_GIGA_MAC_VER_02, 0xff7e1880), // 8169S
- _R("RTL8110s", RTL_GIGA_MAC_VER_03, 0xff7e1880), // 8110S
- _R("RTL8169sb/8110sb", RTL_GIGA_MAC_VER_04, 0xff7e1880), // 8169SB
- _R("RTL8169sc/8110sc", RTL_GIGA_MAC_VER_05, 0xff7e1880), // 8110SCd
- _R("RTL8169sc/8110sc", RTL_GIGA_MAC_VER_06, 0xff7e1880), // 8110SCe
- _R("RTL8102e", RTL_GIGA_MAC_VER_07, 0xff7e1880), // PCI-E
- _R("RTL8102e", RTL_GIGA_MAC_VER_08, 0xff7e1880), // PCI-E
- _R("RTL8102e", RTL_GIGA_MAC_VER_09, 0xff7e1880), // PCI-E
- _R("RTL8101e", RTL_GIGA_MAC_VER_10, 0xff7e1880), // PCI-E
- _R("RTL8168b/8111b", RTL_GIGA_MAC_VER_11, 0xff7e1880), // PCI-E
- _R("RTL8168b/8111b", RTL_GIGA_MAC_VER_12, 0xff7e1880), // PCI-E
- _R("RTL8101e", RTL_GIGA_MAC_VER_13, 0xff7e1880), // PCI-E 8139
- _R("RTL8100e", RTL_GIGA_MAC_VER_14, 0xff7e1880), // PCI-E 8139
- _R("RTL8100e", RTL_GIGA_MAC_VER_15, 0xff7e1880), // PCI-E 8139
- _R("RTL8168b/8111b", RTL_GIGA_MAC_VER_17, 0xff7e1880), // PCI-E
- _R("RTL8101e", RTL_GIGA_MAC_VER_16, 0xff7e1880), // PCI-E
- _R("RTL8168cp/8111cp", RTL_GIGA_MAC_VER_18, 0xff7e1880), // PCI-E
- _R("RTL8168c/8111c", RTL_GIGA_MAC_VER_19, 0xff7e1880), // PCI-E
- _R("RTL8168c/8111c", RTL_GIGA_MAC_VER_20, 0xff7e1880), // PCI-E
- _R("RTL8168c/8111c", RTL_GIGA_MAC_VER_21, 0xff7e1880), // PCI-E
- _R("RTL8168c/8111c", RTL_GIGA_MAC_VER_22, 0xff7e1880), // PCI-E
- _R("RTL8168cp/8111cp", RTL_GIGA_MAC_VER_23, 0xff7e1880), // PCI-E
- _R("RTL8168cp/8111cp", RTL_GIGA_MAC_VER_24, 0xff7e1880), // PCI-E
- _R("RTL8168d/8111d", RTL_GIGA_MAC_VER_25, 0xff7e1880), // PCI-E
- _R("RTL8168d/8111d", RTL_GIGA_MAC_VER_26, 0xff7e1880), // PCI-E
- _R("RTL8168dp/8111dp", RTL_GIGA_MAC_VER_27, 0xff7e1880), // PCI-E
- _R("RTL8168dp/8111dp", RTL_GIGA_MAC_VER_28, 0xff7e1880), // PCI-E
- _R("RTL8105e", RTL_GIGA_MAC_VER_29, 0xff7e1880), // PCI-E
- _R("RTL8105e", RTL_GIGA_MAC_VER_30, 0xff7e1880), // PCI-E
- _R("RTL8168dp/8111dp", RTL_GIGA_MAC_VER_31, 0xff7e1880), // PCI-E
- _R("RTL8168e/8111e", RTL_GIGA_MAC_VER_32, 0xff7e1880), // PCI-E
- _R("RTL8168e/8111e", RTL_GIGA_MAC_VER_33, 0xff7e1880) // PCI-E
+ _R("RTL8169", RTL_GIGA_MAC_VER_01, RTL_TD_0), // 8169
+ _R("RTL8169s", RTL_GIGA_MAC_VER_02, RTL_TD_0), // 8169S
+ _R("RTL8110s", RTL_GIGA_MAC_VER_03, RTL_TD_0), // 8110S
+ _R("RTL8169sb/8110sb", RTL_GIGA_MAC_VER_04, RTL_TD_0), // 8169SB
+ _R("RTL8169sc/8110sc", RTL_GIGA_MAC_VER_05, RTL_TD_0), // 8110SCd
+ _R("RTL8169sc/8110sc", RTL_GIGA_MAC_VER_06, RTL_TD_0), // 8110SCe
+ _R("RTL8102e", RTL_GIGA_MAC_VER_07, RTL_TD_1), // PCI-E
+ _R("RTL8102e", RTL_GIGA_MAC_VER_08, RTL_TD_1), // PCI-E
+ _R("RTL8102e", RTL_GIGA_MAC_VER_09, RTL_TD_1), // PCI-E
+ _R("RTL8101e", RTL_GIGA_MAC_VER_10, RTL_TD_0), // PCI-E
+ _R("RTL8168b/8111b", RTL_GIGA_MAC_VER_11, RTL_TD_0), // PCI-E
+ _R("RTL8168b/8111b", RTL_GIGA_MAC_VER_12, RTL_TD_0), // PCI-E
+ _R("RTL8101e", RTL_GIGA_MAC_VER_13, RTL_TD_0), // PCI-E 8139
+ _R("RTL8100e", RTL_GIGA_MAC_VER_14, RTL_TD_0), // PCI-E 8139
+ _R("RTL8100e", RTL_GIGA_MAC_VER_15, RTL_TD_0), // PCI-E 8139
+ _R("RTL8168b/8111b", RTL_GIGA_MAC_VER_17, RTL_TD_0), // PCI-E
+ _R("RTL8101e", RTL_GIGA_MAC_VER_16, RTL_TD_0), // PCI-E
+ _R("RTL8168cp/8111cp", RTL_GIGA_MAC_VER_18, RTL_TD_1), // PCI-E
+ _R("RTL8168c/8111c", RTL_GIGA_MAC_VER_19, RTL_TD_1), // PCI-E
+ _R("RTL8168c/8111c", RTL_GIGA_MAC_VER_20, RTL_TD_1), // PCI-E
+ _R("RTL8168c/8111c", RTL_GIGA_MAC_VER_21, RTL_TD_1), // PCI-E
+ _R("RTL8168c/8111c", RTL_GIGA_MAC_VER_22, RTL_TD_1), // PCI-E
+ _R("RTL8168cp/8111cp", RTL_GIGA_MAC_VER_23, RTL_TD_1), // PCI-E
+ _R("RTL8168cp/8111cp", RTL_GIGA_MAC_VER_24, RTL_TD_1), // PCI-E
+ _R("RTL8168d/8111d", RTL_GIGA_MAC_VER_25, RTL_TD_1), // PCI-E
+ _R("RTL8168d/8111d", RTL_GIGA_MAC_VER_26, RTL_TD_1), // PCI-E
+ _R("RTL8168dp/8111dp", RTL_GIGA_MAC_VER_27, RTL_TD_1), // PCI-E
+ _R("RTL8168dp/8111dp", RTL_GIGA_MAC_VER_28, RTL_TD_1), // PCI-E
+ _R("RTL8105e", RTL_GIGA_MAC_VER_29, RTL_TD_1), // PCI-E
+ _R("RTL8105e", RTL_GIGA_MAC_VER_30, RTL_TD_1), // PCI-E
+ _R("RTL8168dp/8111dp", RTL_GIGA_MAC_VER_31, RTL_TD_1), // PCI-E
+ _R("RTL8168e/8111e", RTL_GIGA_MAC_VER_32, RTL_TD_1), // PCI-E
+ _R("RTL8168e/8111e", RTL_GIGA_MAC_VER_33, RTL_TD_1) // PCI-E
};
#undef _R
@@ -230,6 +235,9 @@ enum rtl_registers {
IntrStatus = 0x3e,
TxConfig = 0x40,
RxConfig = 0x44,
+
+#define RTL_RX_CONFIG_MASK 0xff7e1880u
+
RxMissed = 0x4c,
Cfg9346 = 0x50,
Config0 = 0x51,
@@ -452,21 +460,69 @@ enum rtl_register_content {
CounterDump = 0x8,
};
-enum desc_status_bit {
+enum rtl_desc_bit {
+ /* First doubleword. */
DescOwn = (1 << 31), /* Descriptor is owned by NIC */
RingEnd = (1 << 30), /* End of descriptor ring */
FirstFrag = (1 << 29), /* First segment of a packet */
LastFrag = (1 << 28), /* Final segment of a packet */
+};
+
+/* Generic case. */
+enum rtl_tx_desc_bit {
+ /* First doubleword. */
+ TD_LSO = (1 << 27), /* Large Send Offload */
+#define TD_MSS_MAX 0x07ffu /* MSS value */
- /* Tx private */
- LargeSend = (1 << 27), /* TCP Large Send Offload (TSO) */
- MSSShift = 16, /* MSS value position */
- MSSMask = 0xfff, /* MSS value + LargeSend bit: 12 bits */
- IPCS = (1 << 18), /* Calculate IP checksum */
- UDPCS = (1 << 17), /* Calculate UDP/IP checksum */
- TCPCS = (1 << 16), /* Calculate TCP/IP checksum */
- TxVlanTag = (1 << 17), /* Add VLAN tag */
+ /* Second doubleword. */
+ TxVlanTag = (1 << 17), /* Add VLAN tag */
+};
+
+/* 8169, 8168b and 810x except 8102e. */
+enum rtl_tx_desc_bit_0 {
+ /* First doubleword. */
+#define TD0_MSS_SHIFT 16 /* MSS position (11 bits) */
+ TD0_TCP_CS = (1 << 16), /* Calculate TCP/IP checksum */
+ TD0_UDP_CS = (1 << 17), /* Calculate UDP/IP checksum */
+ TD0_IP_CS = (1 << 18), /* Calculate IP checksum */
+};
+
+/* 8102e, 8168c and beyond. */
+enum rtl_tx_desc_bit_1 {
+ /* Second doubleword. */
+#define TD1_MSS_SHIFT 18 /* MSS position (11 bits) */
+ TD1_IP_CS = (1 << 29), /* Calculate IP checksum */
+ TD1_TCP_CS = (1 << 30), /* Calculate TCP/IP checksum */
+ TD1_UDP_CS = (1 << 31), /* Calculate UDP/IP checksum */
+};
+static const struct rtl_tx_desc_info {
+ struct {
+ u32 udp;
+ u32 tcp;
+ } checksum;
+ u16 mss_shift;
+ u16 opts_offset;
+} tx_desc_info [] = {
+ [RTL_TD_0] = {
+ .checksum = {
+ .udp = TD0_IP_CS | TD0_UDP_CS,
+ .tcp = TD0_IP_CS | TD0_TCP_CS
+ },
+ .mss_shift = TD0_MSS_SHIFT,
+ .opts_offset = 0
+ },
+ [RTL_TD_1] = {
+ .checksum = {
+ .udp = TD1_IP_CS | TD1_UDP_CS,
+ .tcp = TD1_IP_CS | TD1_TCP_CS
+ },
+ .mss_shift = TD1_MSS_SHIFT,
+ .opts_offset = 1
+ }
+};
+
+enum rtl_rx_desc_bit {
/* Rx private */
PID1 = (1 << 18), /* Protocol ID bit 1/2 */
PID0 = (1 << 17), /* Protocol ID bit 2/2 */
@@ -531,8 +587,8 @@ struct rtl8169_private {
struct napi_struct napi;
spinlock_t lock; /* spin lock flag */
u32 msg_enable;
- int chipset;
- int mac_version;
+ u16 txd_version;
+ u16 mac_version;
u32 cur_rx; /* Index into the Rx descriptor buffer of next Rx pkt. */
u32 cur_tx; /* Index into the Tx descriptor buffer of next Rx pkt. */
u32 dirty_rx;
@@ -1288,7 +1344,7 @@ static int rtl8169_set_settings(struct net_device *dev, struct ethtool_cmd *cmd)
static u32 rtl8169_fix_features(struct net_device *dev, u32 features)
{
- if (dev->mtu > MSSMask)
+ if (dev->mtu > TD_MSS_MAX)
features &= ~NETIF_F_ALL_TSO;
return features;
@@ -3194,7 +3250,7 @@ rtl8169_init_one(struct pci_dev *pdev, const struct pci_device_id *ent)
struct mii_if_info *mii;
struct net_device *dev;
void __iomem *ioaddr;
- unsigned int i;
+ int chipset, i;
int rc;
if (netif_msg_drv(&debug)) {
@@ -3336,7 +3392,8 @@ rtl8169_init_one(struct pci_dev *pdev, const struct pci_device_id *ent)
"driver bug, MAC version not found in rtl_chip_info\n");
goto err_out_msi_4;
}
- tp->chipset = i;
+ chipset = i;
+ tp->txd_version = rtl_chip_info[chipset].txd_version;
RTL_W8(Cfg9346, Cfg9346_Unlock);
RTL_W8(Config1, RTL_R8(Config1) | PMEnable);
@@ -3413,8 +3470,7 @@ rtl8169_init_one(struct pci_dev *pdev, const struct pci_device_id *ent)
pci_set_drvdata(pdev, dev);
netif_info(tp, probe, dev, "%s at 0x%lx, %pM, XID %08x IRQ %d\n",
- rtl_chip_info[tp->chipset].name,
- dev->base_addr, dev->dev_addr,
+ rtl_chip_info[chipset].name, dev->base_addr, dev->dev_addr,
(u32)(RTL_R32(TxConfig) & 0x9cf0f8ff), dev->irq);
if ((tp->mac_version == RTL_GIGA_MAC_VER_27) ||
@@ -3572,7 +3628,7 @@ static void rtl_set_rx_tx_config_registers(struct rtl8169_private *tp)
void __iomem *ioaddr = tp->mmio_addr;
u32 cfg = rtl8169_rx_config;
- cfg |= (RTL_R32(RxConfig) & rtl_chip_info[tp->chipset].RxConfigMask);
+ cfg |= (RTL_R32(RxConfig) & RTL_RX_CONFIG_MASK);
RTL_W32(RxConfig, cfg);
/* Set DMA burst size and Interframe Gap Time */
@@ -4564,7 +4620,7 @@ static void rtl8169_tx_timeout(struct net_device *dev)
}
static int rtl8169_xmit_frags(struct rtl8169_private *tp, struct sk_buff *skb,
- u32 opts1)
+ u32 *opts)
{
struct skb_shared_info *info = skb_shinfo(skb);
unsigned int cur_frag, entry;
@@ -4592,9 +4648,11 @@ static int rtl8169_xmit_frags(struct rtl8169_private *tp, struct sk_buff *skb,
}
/* anti gcc 2.95.3 bugware (sic) */
- status = opts1 | len | (RingEnd * !((entry + 1) % NUM_TX_DESC));
+ status = opts[0] | len |
+ (RingEnd * !((entry + 1) % NUM_TX_DESC));
txd->opts1 = cpu_to_le32(status);
+ txd->opts2 = cpu_to_le32(opts[1]);
txd->addr = cpu_to_le64(mapping);
tp->tx_skb[entry].len = len;
@@ -4612,23 +4670,28 @@ err_out:
return -EIO;
}
-static inline u32 rtl8169_tso_csum(struct sk_buff *skb, struct net_device *dev)
+static inline void rtl8169_tso_csum(struct rtl8169_private *tp,
+ struct sk_buff *skb, u32 *opts)
{
+ const struct rtl_tx_desc_info *info = tx_desc_info + tp->txd_version;
u32 mss = skb_shinfo(skb)->gso_size;
+ int offset = info->opts_offset;
- if (mss)
- return LargeSend | ((mss & MSSMask) << MSSShift);
+ if (mss) {
+ opts[0] |= TD_LSO;
+ opts[offset] |= min(mss, TD_MSS_MAX) << info->mss_shift;
+ }
if (skb->ip_summed == CHECKSUM_PARTIAL) {
const struct iphdr *ip = ip_hdr(skb);
if (ip->protocol == IPPROTO_TCP)
- return IPCS | TCPCS;
+ opts[offset] |= info->checksum.tcp;
else if (ip->protocol == IPPROTO_UDP)
- return IPCS | UDPCS;
- WARN_ON(1); /* we need a WARN() */
+ opts[offset] |= info->checksum.udp;
+ else
+ WARN_ON(1); /* we need a WARN() */
}
- return 0;
}
static netdev_tx_t rtl8169_start_xmit(struct sk_buff *skb,
@@ -4641,7 +4704,7 @@ static netdev_tx_t rtl8169_start_xmit(struct sk_buff *skb,
struct device *d = &tp->pci_dev->dev;
dma_addr_t mapping;
u32 status, len;
- u32 opts1;
+ u32 opts[2];
int frags;
if (unlikely(TX_BUFFS_AVAIL(tp) < skb_shinfo(skb)->nr_frags)) {
@@ -4662,24 +4725,28 @@ static netdev_tx_t rtl8169_start_xmit(struct sk_buff *skb,
tp->tx_skb[entry].len = len;
txd->addr = cpu_to_le64(mapping);
- txd->opts2 = cpu_to_le32(rtl8169_tx_vlan_tag(tp, skb));
- opts1 = DescOwn | rtl8169_tso_csum(skb, dev);
+ opts[1] = cpu_to_le32(rtl8169_tx_vlan_tag(tp, skb));
+ opts[0] = DescOwn;
- frags = rtl8169_xmit_frags(tp, skb, opts1);
+ rtl8169_tso_csum(tp, skb, opts);
+
+ frags = rtl8169_xmit_frags(tp, skb, opts);
if (frags < 0)
goto err_dma_1;
else if (frags)
- opts1 |= FirstFrag;
+ opts[0] |= FirstFrag;
else {
- opts1 |= FirstFrag | LastFrag;
+ opts[0] |= FirstFrag | LastFrag;
tp->tx_skb[entry].skb = skb;
}
+ txd->opts2 = cpu_to_le32(opts[1]);
+
wmb();
/* anti gcc 2.95.3 bugware (sic) */
- status = opts1 | len | (RingEnd * !((entry + 1) % NUM_TX_DESC));
+ status = opts[0] | len | (RingEnd * !((entry + 1) % NUM_TX_DESC));
txd->opts1 = cpu_to_le32(status);
tp->cur_tx += frags + 1;
@@ -5167,7 +5234,7 @@ static void rtl_set_rx_mode(struct net_device *dev)
spin_lock_irqsave(&tp->lock, flags);
tmp = rtl8169_rx_config | rx_mode |
- (RTL_R32(RxConfig) & rtl_chip_info[tp->chipset].RxConfigMask);
+ (RTL_R32(RxConfig) & RTL_RX_CONFIG_MASK);
if (tp->mac_version > RTL_GIGA_MAC_VER_06) {
u32 data = mc_filter[0];
--
1.7.4.2
^ permalink raw reply related
* Re: [PATCH] r8169: Be verbose when unable to load fw patch
From: David Miller @ 2011-04-13 17:57 UTC (permalink / raw)
To: romieu; +Cc: bp, linux-kernel, borislav.petkov, netdev
In-Reply-To: <20110413174852.GC17579@electric-eye.fr.zoreil.com>
From: François Romieu <romieu@fr.zoreil.com>
Date: Wed, 13 Apr 2011 19:48:52 +0200
> On Tue, Apr 12, 2011 at 04:20:36PM -0700, David Miller wrote:
> [...]
>> Seems reasonable, Francois?
>
> Yes, it makes sense.
>
> The patch goes way beyond 80 cols but I'll do a round of
> cleanups later anyway.
Ok. So will you submit this yourself later or should I just
apply this myself directly to net-2.6?
I'm fine either way.
Thanks!
^ permalink raw reply
* Re: Bonding/LACP on RTL8169sc/8110sc (R1869)
From: François Romieu @ 2011-04-13 17:49 UTC (permalink / raw)
To: Jay Vosburgh; +Cc: Jonathan Thibault, netdev
In-Reply-To: <13118.1302648465@death>
On Tue, Apr 12, 2011 at 03:47:45PM -0700, Jay Vosburgh wrote:
> Jonathan Thibault <jonathan@navigue.com> wrote:
[2.6.32.old]
> That's a fairly old kernel; one relatively recent fix that comes
> to mind is:
The r8169 driver has undergone some multicast fixes since 2.6.32 as well.
If I remember correctly, the old 8169 - not the 8168 - was hit.
Upgrading is really suggested.
--
Ueimor
^ permalink raw reply
* Re: [Bugme-new] [Bug 33042] New: Marvell 88E1145 phy configured incorrectly in fiber mode
From: Andy Fleming @ 2011-04-13 18:01 UTC (permalink / raw)
To: Alex Dubov
Cc: Andrew Morton, David Daney, netdev, bugzilla-daemon, bugme-daemon,
Grant Likely, Andy Fleming
In-Reply-To: <392488.26736.qm@web37606.mail.mud.yahoo.com>
On Mon, Apr 11, 2011 at 10:45 PM, Alex Dubov <oakad@yahoo.com> wrote:
>>
>> How does your u-boot configure the part? Does it
>> write any of the
>> configuration registers, or is it just the default
>> configuration set via
>> the strapping pins?
>
> U-boot configures this phy just like any other phy - by running a set of
> register assignments from phy_info_M88E1145.
>
> Unfortunately, I don't have a datasheet for this phy and kernel does
> quite a few things differently, so simply copying stuff from u-boot
> does not work well (in kernel, phy initialization is broken into 3
> functions, if I'm not mistaken).
I've just rewritten the U-Boot code for PHY management, so I'd be
interested in hearing if this breaks your board. But what's
interesting to me is that, in order for U-Boot to report that the link
is a "fiber" link, something had to set the TSEC_FIBER flag, and only
one PHY in the public source did. This implies to me that your board
isn't supported by mainline U-Boot, and suggests that someone may have
modified the 88e1145 driver. Otherwise, I don't see any fiber-related
differences between the U-Boot 1145 driver, and the Linux one.
>
>
>>
>> In any event, you will probably have to read the
>> configuration before
>> the drivers/net/phy/marvel.c changes them. Then
>> compare that to what
>> the driver is trying to set. Then you will either
>> have to override the
>> configuration with the device tree "marvell,reg-init"
>> property, or if
>> you are not using the device tree, add a 88e1145 specific
>> flag that you
>> set when calling phy_connect().
Reading the configuration from U-Boot is straightforward. use the
"mii" command to read the registers. But don't forget to set register
22 (16 - mii command only reads hex) to 1, and read all of the
registers that way, too.
You will either need to add some code to detect when the PHY is using
fiber, and change the phydev->port to PORT_FIBRE, or you will need to
add a board-level "fixup" to change the port to PORT_FIBRE on your
board.
Then the PHY driver should use that information to do the right configuration.
Andy
^ permalink raw reply
* Re: [PATCHv4 NEXT 1/1] net: ethtool support to configure number of channels
From: Ben Hutchings @ 2011-04-13 18:22 UTC (permalink / raw)
To: Amit Kumar Salecha
Cc: davem, netdev, ameen.rahman, sucheta.chakraborty,
anirban.chakraborty
In-Reply-To: <1302177522-17815-1-git-send-email-amit.salecha@qlogic.com>
On Thu, 2011-04-07 at 04:58 -0700, Amit Kumar Salecha wrote:
> Ethtool support to configure RX, TX and other channels. combined field
> in struct ethtool_channels to reflect set of channel (RX, TX or other).
> Other channel can be link interrupts, SR-IOV coordination etc.
>
> ETHTOOL_GCHANNELS will report max and current number of RX channels,
> max and current number of TX channels, max and current number of other channel
> or max and current number of combined channel.
>
> Number of channel can be modify upto max number of channel through
> ETHTOOL_SCHANNELS command.
>
> Ben Hutchings:
> o define 'combined' and 'other' types. Most multiqueue drivers pair up RX and TX
> queues so that most channels combine RX and TX work.
> o Please could you use a kernel-doc comment to describe the structure.
>
> Signed-off-by: Amit Kumar Salecha <amit.salecha@qlogic.com>
Reviewed-by: Ben Hutchings <bhutchings@solarflare.com>
Sorry for the delay.
Ben.
--
Ben Hutchings, Senior Software Engineer, Solarflare
Not speaking for my employer; that's the marketing department's job.
They asked us to note that Solarflare product names are trademarked.
^ permalink raw reply
* Re: panic in tg3 driver
From: Stephen Clark @ 2011-04-13 18:23 UTC (permalink / raw)
To: Matt Carlson; +Cc: Linux Kernel Network Developers, Michael Chan
In-Reply-To: <20110125022532.GA19884@mcarlson.broadcom.com>
On 01/24/2011 09:25 PM, Matt Carlson wrote:
> On Mon, Jan 24, 2011 at 04:59:22PM -0800, Matt Carlson wrote:
>
>> On Sun, Jan 16, 2011 at 10:11:50AM -0800, Stephen Clark wrote:
>>
>>> On 01/13/2011 08:12 AM, Stephen Clark wrote:
>>>
>>>> On 01/11/2011 10:06 PM, Matt Carlson wrote:
>>>>
>>>>> lspci -vvv -xxx -s 81:00.0
>>>>>
>>>>
>>>>
>>>> Further information - I found these messages in /var/log/messages. It
>>>> looks
>>>> like after it switched to INTx mode interrupts for other devices were
>>>> hosed.
>>>>
>>>> Jan 12 08:37:49 localhost kernel: tg3 0000:81:00.0: eth2: No interrupt
>>>> was gener
>>>> ated using MSI. Switching to INTx mode. Please report this failure to
>>>> the PCI ma
>>>> intainer and include system chipset information
>>>> Jan 12 08:37:49 localhost kernel: ADDRCONF(NETDEV_UP): eth2: link is
>>>> not ready
>>>> Jan 12 08:38:50 localhost kernel: ata2: lost interrupt (Status 0x50)
>>>> Jan 12 08:38:50 localhost kernel: ata2.01: exception Emask 0x0 SAct
>>>> 0x0 SErr 0x0
>>>> action 0x6 frozen
>>>> Jan 12 08:38:50 localhost kernel: ata2.01: failed command: WRITE DMA
>>>> Jan 12 08:38:50 localhost kernel: ata2.01: cmd
>>>> ca/00:08:e0:bc:51/00:00:00:00:00/f0 tag 0 dma 4096 out
>>>> Jan 12 08:38:50 localhost kernel: res
>>>> 40/00:01:00:4f:c2/00:00:00:00:00/b0 Emask 0x4 (timeout)
>>>> Jan 12 08:38:50 localhost kernel: ata2.01: status: { DRDY }
>>>> Jan 12 08:38:50 localhost kernel: ata2: soft resetting link
>>>> Jan 12 08:38:50 localhost kernel: do_IRQ: 0.64 No irq handler for
>>>> vector (irq -1)
>>>> Jan 12 08:38:50 localhost kernel: ata2.01: configured for UDMA/33
>>>> Jan 12 08:38:54 localhost pppd[1983]: No response to 3 echo-requests
>>>> Jan 12 08:39:55 localhost pppoe[1988]: Inactivity timeout... something
>>>> wicked happened on session 3363
>>>>
>>> Just checking to make sure you have everything you need?
>>>
>> Sorry for the delay Stephen.
>>
>> It looks to me like interrupts aren't being setup correctly on this
>> system. I tested MSI and INTx interrupt modes locally and they both
>> work. I'm guessing one of two things could be happening:
>>
>> 1) The 2nd parameter of the low-level ISR (tg3_interrupt_tagged()) is
>> not correct. The ISR tries to tell the hardware the interrupt is
>> acknowledged, but the message goes unheard. (This might also explain
>> why other devices are also afflicted.)
>>
>> 2) Something is blocking the delivery of the interrupt to the tg3 driver
>> altogether.
>>
>> In both cases, the hardware persistently nags the host to ack the
>> interrupt, hence the interrupt storm.
>>
> Just curious, is the problem still there if you add pci=nomsi to the
> kernel command line?
>
>
Sorry I have been tied up.
With kernel 2.6.32-44.1.el6.i686 and pci=nomsi on the kernel command
line it seems to work great.
[root@Z1010 ~]# ping -f 3.3.3.2
PING 3.3.3.2 (3.3.3.2) 56(84) bytes of data.
.^
--- 3.3.3.2 ping statistics ---
20562 packets transmitted, 20562 received, 0% packet loss, time 4408ms
rtt min/avg/max/mdev = 0.141/0.163/1.021/0.034 ms, ipg/ewma 0.214/0.161 ms
> --
> 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
>
>
--
"They that give up essential liberty to obtain temporary safety,
deserve neither liberty nor safety." (Ben Franklin)
"The course of history shows that as a government grows, liberty
decreases." (Thomas Jefferson)
^ permalink raw reply
* Re: Bonding/LACP on RTL8169sc/8110sc (R1869)
From: Ben Hutchings @ 2011-04-13 18:28 UTC (permalink / raw)
To: François Romieu; +Cc: Jay Vosburgh, Jonathan Thibault, netdev
In-Reply-To: <20110413174943.GD17579@electric-eye.fr.zoreil.com>
On Wed, 2011-04-13 at 19:49 +0200, François Romieu wrote:
> On Tue, Apr 12, 2011 at 03:47:45PM -0700, Jay Vosburgh wrote:
> > Jonathan Thibault <jonathan@navigue.com> wrote:
> [2.6.32.old]
> > That's a fairly old kernel; one relatively recent fix that comes
> > to mind is:
>
> The r8169 driver has undergone some multicast fixes since 2.6.32 as well.
>
> If I remember correctly, the old 8169 - not the 8168 - was hit.
>
> Upgrading is really suggested.
If you mean the changes to register writes (commits
78f1cd02457252e1ffbc6caa44a17424a45286b8 and
908ba2bfd22253f26fa910cd855e4ccffb1467d0), those went into 2.6.32.y.
Ben.
--
Ben Hutchings, Senior Software Engineer, Solarflare
Not speaking for my employer; that's the marketing department's job.
They asked us to note that Solarflare product names are trademarked.
^ permalink raw reply
* Re: [PATCHv4 NEXT 1/1] net: ethtool support to configure number of channels
From: David Miller @ 2011-04-13 18:31 UTC (permalink / raw)
To: bhutchings
Cc: amit.salecha, netdev, ameen.rahman, sucheta.chakraborty,
anirban.chakraborty
In-Reply-To: <1302718921.2873.1.camel@bwh-desktop>
From: Ben Hutchings <bhutchings@solarflare.com>
Date: Wed, 13 Apr 2011 19:22:01 +0100
> On Thu, 2011-04-07 at 04:58 -0700, Amit Kumar Salecha wrote:
>> Ethtool support to configure RX, TX and other channels. combined field
>> in struct ethtool_channels to reflect set of channel (RX, TX or other).
>> Other channel can be link interrupts, SR-IOV coordination etc.
>>
>> ETHTOOL_GCHANNELS will report max and current number of RX channels,
>> max and current number of TX channels, max and current number of other channel
>> or max and current number of combined channel.
>>
>> Number of channel can be modify upto max number of channel through
>> ETHTOOL_SCHANNELS command.
>>
>> Ben Hutchings:
>> o define 'combined' and 'other' types. Most multiqueue drivers pair up RX and TX
>> queues so that most channels combine RX and TX work.
>> o Please could you use a kernel-doc comment to describe the structure.
>>
>> Signed-off-by: Amit Kumar Salecha <amit.salecha@qlogic.com>
> Reviewed-by: Ben Hutchings <bhutchings@solarflare.com>
Applied, thanks everyone.
^ permalink raw reply
* Re: [PATCH] net: can: mscan: fix build breakage in mpc5xxx_can
From: David Miller @ 2011-04-13 18:33 UTC (permalink / raw)
To: mkl; +Cc: agust, netdev, grant.likely
In-Reply-To: <4DA57F55.7080304@pengutronix.de>
From: Marc Kleine-Budde <mkl@pengutronix.de>
Date: Wed, 13 Apr 2011 12:47:49 +0200
> On 04/13/2011 11:49 AM, Anatolij Gustschin wrote:
>> Commit 74888760d40b3ac9054f9c5fa07b566c0676ba2d
>> "dt/net: Eliminate users of of_platform_{,un}register_driver"
>> broke building mscan driver. Fix it.
>>
>> Signed-off-by: Anatolij Gustschin <agust@denx.de>
>> Cc: Grant Likely <grant.likely@secretlab.ca>
>
> Acked-by: Marc Kleine-Budde <mkl@pengutronix.de>
I'll apply this to net-2.6, thanks everyone.
^ permalink raw reply
* Re: [Bugme-new] [Bug 32832] New: shutdown(2) does not fully shut down socket any more
From: Stephen Hemminger @ 2011-04-13 18:47 UTC (permalink / raw)
To: David Miller
Cc: dbaluta, eric.dumazet, akpm, netdev, bugzilla-daemon,
bugme-daemon, kees
In-Reply-To: <20110413.104317.71110417.davem@davemloft.net>
On Wed, 13 Apr 2011 10:43:17 -0700 (PDT)
David Miller <davem@davemloft.net> wrote:
> From: Daniel Baluta <dbaluta@ixiacom.com>
> Date: Wed, 13 Apr 2011 14:57:18 +0300
>
> > Cyril's use case looks suspect. I don't think that this is a good
> > reason for reverting this commit.
>
> I complete disagree.
>
> Something that worked perfectly fine, probably for years, we broke.
>
> We simply cannot do that, especially since we do not have a reasonable
> alternative at this time.
>
> Adding SO_REUSEPORT is a long range option, and not something that
> will provide a fix for users right now.
>
> So please don't even pretend to suggest that we shouldn't fix this
> with a revert unless a simple, obvious, kernel fix presents itself.
Just to echo what Dave said.
Even though the semantics of this is not documented in some standard,
applications have been built on Linux expecting a certain behavior.
If you want to change what happens in this case, you have to have a
really good reason (like crash, security hole, or standards violation).
^ permalink raw reply
* [PATCH] net: allow shifted access in smsc911x V2
From: mathieu.poirier @ 2011-04-13 18:50 UTC (permalink / raw)
To: netdev, mathieu.poirier; +Cc: lee.jones, patches, rubini, linus.walleij
From: Mathieu J. Poirier <mathieu.poirier@linaro.org>
This is a revised patch that permits a shifted access to the
LAN9221 registers. More specifically:
It adds a shift parameter in the platform_data.
It introduces an ops in smsc911x_data.
A choice of access function to use at run-time.
Four new shifted access function.
Signed-off-by: Mathieu Poirier <mathieu.poirier@linaro.org>
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>
---
drivers/net/smsc911x.c | 156 +++++++++++++++++++++++++++++++++++++++++++--
include/linux/smsc911x.h | 1 +
2 files changed, 150 insertions(+), 7 deletions(-)
diff --git a/drivers/net/smsc911x.c b/drivers/net/smsc911x.c
index d70bde9..2d65fa4 100644
--- a/drivers/net/smsc911x.c
+++ b/drivers/net/smsc911x.c
@@ -69,6 +69,17 @@ static int debug = 3;
module_param(debug, int, 0);
MODULE_PARM_DESC(debug, "Debug level (0=none,...,16=all)");
+struct smsc911x_data;
+
+struct smsc911x_ops {
+ u32 (*reg_read)(struct smsc911x_data *pdata, u32 reg);
+ void (*reg_write)(struct smsc911x_data *pdata, u32 reg, u32 val);
+ void (*rx_readfifo)(struct smsc911x_data *pdata,
+ unsigned int *buf, unsigned int wordcount);
+ void (*tx_writefifo)(struct smsc911x_data *pdata,
+ unsigned int *buf, unsigned int wordcount);
+};
+
struct smsc911x_data {
void __iomem *ioaddr;
@@ -116,8 +127,14 @@ struct smsc911x_data {
unsigned int clear_bits_mask;
unsigned int hashhi;
unsigned int hashlo;
+
+ /* register access functions */
+ const struct smsc911x_ops *ops;
};
+/* Easy access to information */
+#define __smsc_shift(pdata, reg) ((reg) << ((pdata)->config.shift))
+
static inline u32 __smsc911x_reg_read(struct smsc911x_data *pdata, u32 reg)
{
if (pdata->config.flags & SMSC911X_USE_32BIT)
@@ -131,13 +148,29 @@ static inline u32 __smsc911x_reg_read(struct smsc911x_data *pdata, u32 reg)
return 0;
}
+static inline u32
+__smsc911x_reg_read_shift(struct smsc911x_data *pdata, u32 reg)
+{
+ if (pdata->config.flags & SMSC911X_USE_32BIT)
+ return readl(pdata->ioaddr + __smsc_shift(pdata, reg));
+
+ if (pdata->config.flags & SMSC911X_USE_16BIT)
+ return (readw(pdata->ioaddr +
+ __smsc_shift(pdata, reg)) & 0xFFFF) |
+ ((readw(pdata->ioaddr +
+ __smsc_shift(pdata, reg + 2)) & 0xFFFF) << 16);
+
+ BUG();
+ return 0;
+}
+
static inline u32 smsc911x_reg_read(struct smsc911x_data *pdata, u32 reg)
{
u32 data;
unsigned long flags;
spin_lock_irqsave(&pdata->dev_lock, flags);
- data = __smsc911x_reg_read(pdata, reg);
+ data = pdata->ops->reg_read(pdata, reg);
spin_unlock_irqrestore(&pdata->dev_lock, flags);
return data;
@@ -160,13 +193,32 @@ static inline void __smsc911x_reg_write(struct smsc911x_data *pdata, u32 reg,
BUG();
}
+static inline void
+__smsc911x_reg_write_shift(struct smsc911x_data *pdata, u32 reg, u32 val)
+{
+ if (pdata->config.flags & SMSC911X_USE_32BIT) {
+ writel(val, pdata->ioaddr + __smsc_shift(pdata, reg));
+ return;
+ }
+
+ if (pdata->config.flags & SMSC911X_USE_16BIT) {
+ writew(val & 0xFFFF,
+ pdata->ioaddr + __smsc_shift(pdata, reg));
+ writew((val >> 16) & 0xFFFF,
+ pdata->ioaddr + __smsc_shift(pdata, reg + 2));
+ return;
+ }
+
+ BUG();
+}
+
static inline void smsc911x_reg_write(struct smsc911x_data *pdata, u32 reg,
u32 val)
{
unsigned long flags;
spin_lock_irqsave(&pdata->dev_lock, flags);
- __smsc911x_reg_write(pdata, reg, val);
+ pdata->ops->reg_write(pdata, reg, val);
spin_unlock_irqrestore(&pdata->dev_lock, flags);
}
@@ -202,6 +254,40 @@ out:
spin_unlock_irqrestore(&pdata->dev_lock, flags);
}
+/* Writes a packet to the TX_DATA_FIFO - shifted version */
+static inline void
+smsc911x_tx_writefifo_shift(struct smsc911x_data *pdata, unsigned int *buf,
+ unsigned int wordcount)
+{
+ unsigned long flags;
+
+ spin_lock_irqsave(&pdata->dev_lock, flags);
+
+ if (pdata->config.flags & SMSC911X_SWAP_FIFO) {
+ while (wordcount--)
+ __smsc911x_reg_write_shift(pdata, TX_DATA_FIFO,
+ swab32(*buf++));
+ goto out;
+ }
+
+ if (pdata->config.flags & SMSC911X_USE_32BIT) {
+ writesl(pdata->ioaddr + __smsc_shift(pdata,
+ TX_DATA_FIFO), buf, wordcount);
+ goto out;
+ }
+
+ if (pdata->config.flags & SMSC911X_USE_16BIT) {
+ while (wordcount--)
+ __smsc911x_reg_write_shift(pdata,
+ TX_DATA_FIFO, *buf++);
+ goto out;
+ }
+
+ BUG();
+out:
+ spin_unlock_irqrestore(&pdata->dev_lock, flags);
+}
+
/* Reads a packet out of the RX_DATA_FIFO */
static inline void
smsc911x_rx_readfifo(struct smsc911x_data *pdata, unsigned int *buf,
@@ -234,6 +320,40 @@ out:
spin_unlock_irqrestore(&pdata->dev_lock, flags);
}
+/* Reads a packet out of the RX_DATA_FIFO - shifted version */
+static inline void
+smsc911x_rx_readfifo_shift(struct smsc911x_data *pdata, unsigned int *buf,
+ unsigned int wordcount)
+{
+ unsigned long flags;
+
+ spin_lock_irqsave(&pdata->dev_lock, flags);
+
+ if (pdata->config.flags & SMSC911X_SWAP_FIFO) {
+ while (wordcount--)
+ *buf++ = swab32(__smsc911x_reg_read_shift(pdata,
+ RX_DATA_FIFO));
+ goto out;
+ }
+
+ if (pdata->config.flags & SMSC911X_USE_32BIT) {
+ readsl(pdata->ioaddr + __smsc_shift(pdata,
+ RX_DATA_FIFO), buf, wordcount);
+ goto out;
+ }
+
+ if (pdata->config.flags & SMSC911X_USE_16BIT) {
+ while (wordcount--)
+ *buf++ = __smsc911x_reg_read_shift(pdata,
+ RX_DATA_FIFO);
+ goto out;
+ }
+
+ BUG();
+out:
+ spin_unlock_irqrestore(&pdata->dev_lock, flags);
+}
+
/* waits for MAC not busy, with timeout. Only called by smsc911x_mac_read
* and smsc911x_mac_write, so assumes mac_lock is held */
static int smsc911x_mac_complete(struct smsc911x_data *pdata)
@@ -499,7 +619,7 @@ static int smsc911x_phy_check_loopbackpkt(struct smsc911x_data *pdata)
wrsz += (u32)((ulong)pdata->loopback_tx_pkt & 0x3);
wrsz >>= 2;
- smsc911x_tx_writefifo(pdata, (unsigned int *)bufp, wrsz);
+ pdata->ops->tx_writefifo(pdata, (unsigned int *)bufp, wrsz);
/* Wait till transmit is done */
i = 60;
@@ -543,7 +663,7 @@ static int smsc911x_phy_check_loopbackpkt(struct smsc911x_data *pdata)
rdsz += (u32)((ulong)pdata->loopback_rx_pkt & 0x3);
rdsz >>= 2;
- smsc911x_rx_readfifo(pdata, (unsigned int *)bufp, rdsz);
+ pdata->ops->rx_readfifo(pdata, (unsigned int *)bufp, rdsz);
if (pktlength != (MIN_PACKET_SIZE + 4)) {
SMSC_WARNING(HW, "Unexpected packet size "
@@ -1046,8 +1166,8 @@ static int smsc911x_poll(struct napi_struct *napi, int budget)
/* Align IP on 16B boundary */
skb_reserve(skb, NET_IP_ALIGN);
skb_put(skb, pktlength - 4);
- smsc911x_rx_readfifo(pdata, (unsigned int *)skb->head,
- pktwords);
+ pdata->ops->rx_readfifo(pdata,
+ (unsigned int *)skb->head, pktwords);
skb->protocol = eth_type_trans(skb, dev);
skb_checksum_none_assert(skb);
netif_receive_skb(skb);
@@ -1350,7 +1470,7 @@ static int smsc911x_hard_start_xmit(struct sk_buff *skb, struct net_device *dev)
wrsz += (u32)((ulong)skb->data & 0x3);
wrsz >>= 2;
- smsc911x_tx_writefifo(pdata, (unsigned int *)bufp, wrsz);
+ pdata->ops->tx_writefifo(pdata, (unsigned int *)bufp, wrsz);
freespace -= (skb->len + 32);
dev_kfree_skb(skb);
@@ -1951,6 +2071,22 @@ static int __devexit smsc911x_drv_remove(struct platform_device *pdev)
return 0;
}
+/* standard register acces */
+static const struct smsc911x_ops standard_smsc911x_ops = {
+ .reg_read = __smsc911x_reg_read,
+ .reg_write = __smsc911x_reg_write,
+ .rx_readfifo = smsc911x_rx_readfifo,
+ .tx_writefifo = smsc911x_tx_writefifo,
+};
+
+/* shifted register access */
+static const struct smsc911x_ops shifted_smsc911x_ops = {
+ .reg_read = __smsc911x_reg_read_shift,
+ .reg_write = __smsc911x_reg_write_shift,
+ .rx_readfifo = smsc911x_rx_readfifo_shift,
+ .tx_writefifo = smsc911x_tx_writefifo_shift,
+};
+
static int __devinit smsc911x_drv_probe(struct platform_device *pdev)
{
struct net_device *dev;
@@ -2023,6 +2159,12 @@ static int __devinit smsc911x_drv_probe(struct platform_device *pdev)
goto out_free_netdev_2;
}
+ /* assume standard, non-shifted, access to HW registers */
+ pdata->ops = &standard_smsc911x_ops;
+ /* apply the right access if shifting is needed */
+ if (config->shift)
+ pdata->ops = &shifted_smsc911x_ops;
+
retval = smsc911x_init(dev);
if (retval < 0)
goto out_unmap_io_3;
diff --git a/include/linux/smsc911x.h b/include/linux/smsc911x.h
index 7144e8a..4dde70e 100644
--- a/include/linux/smsc911x.h
+++ b/include/linux/smsc911x.h
@@ -29,6 +29,7 @@ struct smsc911x_platform_config {
unsigned int irq_polarity;
unsigned int irq_type;
unsigned int flags;
+ unsigned int shift;
phy_interface_t phy_interface;
unsigned char mac[6];
};
--
1.7.1
^ permalink raw reply related
* Re: [PATCH] macb: Add rx overrun counter
From: David Miller @ 2011-04-13 18:50 UTC (permalink / raw)
To: alexander.stein; +Cc: nicolas.ferre, netdev
In-Reply-To: <1302707004-15818-1-git-send-email-alexander.stein@systec-electronic.com>
From: Alexander Stein <alexander.stein@systec-electronic.com>
Date: Wed, 13 Apr 2011 17:03:24 +0200
> Signed-off-by: Alexander Stein <alexander.stein@systec-electronic.com>
Applied.
^ permalink raw reply
* Re: [PATCH] stmmac: review Wol and enable the Unicast support
From: David Miller @ 2011-04-13 18:51 UTC (permalink / raw)
To: peppe.cavallaro; +Cc: netdev
In-Reply-To: <1302679635-7035-1-git-send-email-peppe.cavallaro@st.com>
From: Giuseppe CAVALLARO <peppe.cavallaro@st.com>
Date: Wed, 13 Apr 2011 09:27:15 +0200
> Signed-off-by: Giuseppe Cavallaro <peppe.cavallaro@st.com>
Applied, thanks.
^ permalink raw reply
* Re: [PATCHv2 NET-2.6 1/1] qlcnic: limit skb frags for non tso packet
From: David Miller @ 2011-04-13 18:53 UTC (permalink / raw)
To: amit.salecha; +Cc: netdev, ameen.rahman, anirban.chakraborty, stable
In-Reply-To: <1302663955-31849-2-git-send-email-amit.salecha@qlogic.com>
From: amit.salecha@qlogic.com
Date: Tue, 12 Apr 2011 20:05:55 -0700
> From: Amit Kumar Salecha <amit.salecha@qlogic.com>
>
> Machines are getting deadlock in four node cluster environment.
> All nodes are accessing (find /gfs2 -depth -print|cpio -ocv > /dev/null)
> 200 GB storage on a GFS2 filesystem.
> This result in memory fragmentation and driver receives 18 frags for
> 1448 byte packets.
> For non tso packet, fw drops the tx request, if it has >14 frags.
>
> Fixing it by pulling extra frags.
>
> Cc: stable@kernel.org
> Signed-off-by: Amit Kumar Salecha <amit.salecha@qlogic.com>
Applied.
But like Greg said, please kill that footer notice on your outgoing
emails when engaging in any discussion on a public mailing list.
Thanks.
^ permalink raw reply
* [PATCH] Add Qualcomm Gobi 2000/3000 driver.
From: Elly Jones @ 2011-04-13 19:00 UTC (permalink / raw)
To: netdev; +Cc: dcbw, mjg59, jglasgow, trond
From: Elizabeth Jones <ellyjones@google.com>
This is a rewrite and unification of the Qualcomm Gobi 2000 and Gobi 3000
drivers, available at:
https://www.codeaurora.org/patches/quic/gobi/VT773.Gobi2000Drivers_03022011.tar.gz
https://www.codeaurora.org/patches/quic/gobi/Gobi3000Drivers1040_03022011.tar.gz
Both devices need firmware loaded in order to be useful.
BUG=chromium-os:5521
TEST=suite_Cellular
Signed-off-by: Elizabeth Jones <ellyjones@chromium.org>
Signed-off-by: Jason Glasgow <jglasgow@chromium.org>
Signed-off-by: Olof Johansson <olofj@chromium.org>
Review URL: http://codereview.chromium.org/6539018
---
drivers/net/usb/Kconfig | 6 +
drivers/net/usb/Makefile | 1 +
drivers/net/usb/gobi/Makefile | 2 +
drivers/net/usb/gobi/README | 30 +
drivers/net/usb/gobi/qcusbnet.c | 717 +++++++++++++++++
drivers/net/usb/gobi/qcusbnet.h | 28 +
drivers/net/usb/gobi/qmi.c | 358 +++++++++
drivers/net/usb/gobi/qmi.h | 67 ++
drivers/net/usb/gobi/qmidevice.c | 1562 ++++++++++++++++++++++++++++++++++++++
drivers/net/usb/gobi/qmidevice.h | 35 +
drivers/net/usb/gobi/structs.h | 96 +++
11 files changed, 2902 insertions(+), 0 deletions(-)
create mode 100644 drivers/net/usb/gobi/Makefile
create mode 100644 drivers/net/usb/gobi/README
create mode 100644 drivers/net/usb/gobi/qcusbnet.c
create mode 100644 drivers/net/usb/gobi/qcusbnet.h
create mode 100644 drivers/net/usb/gobi/qmi.c
create mode 100644 drivers/net/usb/gobi/qmi.h
create mode 100644 drivers/net/usb/gobi/qmidevice.c
create mode 100644 drivers/net/usb/gobi/qmidevice.h
create mode 100644 drivers/net/usb/gobi/structs.h
diff --git a/drivers/net/usb/Kconfig b/drivers/net/usb/Kconfig
index 3ec22c3..3f43266 100644
--- a/drivers/net/usb/Kconfig
+++ b/drivers/net/usb/Kconfig
@@ -448,4 +448,10 @@ config USB_VL600
http://ubuntuforums.org/showpost.php?p=10589647&postcount=17
+config USB_NET_GOBI
+ tristate "Qualcomm Gobi"
+ depends on USB_USBNET
+ help
+ Qualcomm Gobi 2k/3k support.
+
endmenu
diff --git a/drivers/net/usb/Makefile b/drivers/net/usb/Makefile
index c7ec8a5..dd1f24c 100644
--- a/drivers/net/usb/Makefile
+++ b/drivers/net/usb/Makefile
@@ -28,4 +28,5 @@ obj-$(CONFIG_USB_SIERRA_NET) += sierra_net.o
obj-$(CONFIG_USB_NET_CX82310_ETH) += cx82310_eth.o
obj-$(CONFIG_USB_NET_CDC_NCM) += cdc_ncm.o
obj-$(CONFIG_USB_VL600) += lg-vl600.o
+obj-$(CONFIG_USB_NET_GOBI) += gobi/
diff --git a/drivers/net/usb/gobi/Makefile b/drivers/net/usb/gobi/Makefile
new file mode 100644
index 0000000..a6635f6
--- /dev/null
+++ b/drivers/net/usb/gobi/Makefile
@@ -0,0 +1,2 @@
+obj-$(CONFIG_USB_NET_GOBI) += gobi.o
+gobi-objs += qcusbnet.o qmidevice.o qmi.o
diff --git a/drivers/net/usb/gobi/README b/drivers/net/usb/gobi/README
new file mode 100644
index 0000000..113c523
--- /dev/null
+++ b/drivers/net/usb/gobi/README
@@ -0,0 +1,30 @@
+Qualcomm Gobi 2000 driver
+-------------------------
+
+This directory contains a driver for the Qualcomm Gobi 2000 CDMA/GSM modem. The
+device speaks a protocol called QMI with its host; this driver's responsibility
+is to multiplex transactions over QMI connections. Note that the relatively
+simple encoding defined in qmi.h and qmi.c is only the encapsulating protocol;
+the device speaks a more complex set of protocols embodied in the closed-source
+Gobi SDK which are necessary to use advanced features of the device.
+Furthermore, firmware (also proprietary) is needed to make the device useful. An
+open-source firmware loader exists: http://www.codon.org.uk/~mjg59/gobi_loader/.
+
+Interfaces:
+This driver presents two separate interfaces to userspace:
+- usbN, an ordinary usb network interface
+- /dev/qcqmiN, a channel to speak to the device using QMI messages
+The latter is for use by the Gobi SDK.
+
+History:
+This driver's original incarnation (of which this is a rewrite) was released
+under GPLv2 by Qualcomm on their Code Aurora site:
+https://www.codeaurora.org/gitweb/quic/la/?p=kernel/msm.git;a=commit;h=b5135a880f8942f990e8c2e383f7f876beacc55b
+
+Issues:
+Help and commentary on any of the issues below would be appreciated.
+- The locking in devqmi_close() is not threadsafe: we're walking a list of tasks
+ without holding a lock on the list. The original driver did this and I have
+ left it intact because I do not fully understand what's going on here.
+- The driver seems to deadlock very occasionally (one in ~6000 reboots during
+ stress testing). I've had no luck tracking this problem down.
diff --git a/drivers/net/usb/gobi/qcusbnet.c b/drivers/net/usb/gobi/qcusbnet.c
new file mode 100644
index 0000000..954066d
--- /dev/null
+++ b/drivers/net/usb/gobi/qcusbnet.c
@@ -0,0 +1,717 @@
+/* qcusbnet.c - gobi network device
+ * Copyright (c) 2011, Code Aurora Forum. All rights reserved.
+
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 and
+ * only version 2 as published by the Free Software Foundation.
+
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
+ * 02110-1301, USA.
+ */
+
+#include "structs.h"
+#include "qmidevice.h"
+#include "qmi.h"
+#include "qcusbnet.h"
+
+#include <linux/ctype.h>
+
+#define DRIVER_VERSION "1.0.110+google"
+#define DRIVER_AUTHOR "Qualcomm Innovation Center"
+#define DRIVER_DESC "gobi"
+
+static LIST_HEAD(qcusbnet_list);
+static DEFINE_MUTEX(qcusbnet_lock);
+
+int qcusbnet_debug;
+static struct class *devclass;
+
+static void free_dev(struct kref *ref)
+{
+ struct qcusbnet *dev = container_of(ref, struct qcusbnet, refcount);
+ list_del(&dev->node);
+ kfree(dev);
+}
+
+void qcusbnet_put(struct qcusbnet *dev)
+{
+ mutex_lock(&qcusbnet_lock);
+ kref_put(&dev->refcount, free_dev);
+ mutex_unlock(&qcusbnet_lock);
+}
+
+struct qcusbnet *qcusbnet_get(struct qcusbnet *key)
+{
+ /* Given a putative qcusbnet struct, return either the struct itself
+ * (with a ref taken) if the struct is still visible, or NULL if it's
+ * not. This prevents object-visibility races where someone is looking
+ * up an object as the last ref gets dropped; dropping the last ref and
+ * removing the object from the list are atomic with respect to getting
+ * a new ref. */
+ struct qcusbnet *entry;
+ mutex_lock(&qcusbnet_lock);
+ list_for_each_entry(entry, &qcusbnet_list, node) {
+ if (entry == key) {
+ kref_get(&entry->refcount);
+ mutex_unlock(&qcusbnet_lock);
+ return entry;
+ }
+ }
+ mutex_unlock(&qcusbnet_lock);
+ return NULL;
+}
+
+int qc_suspend(struct usb_interface *iface, pm_message_t event)
+{
+ struct usbnet *usbnet;
+ struct qcusbnet *dev;
+
+ if (!iface)
+ return -ENOMEM;
+
+ usbnet = usb_get_intfdata(iface);
+
+ if (!usbnet || !usbnet->net) {
+ DBG("failed to get netdevice\n");
+ return -ENXIO;
+ }
+
+ dev = (struct qcusbnet *)usbnet->data[0];
+ if (!dev) {
+ DBG("failed to get QMIDevice\n");
+ return -ENXIO;
+ }
+
+ if (!(event.event & PM_EVENT_AUTO)) {
+ DBG("device suspended to power level %d\n",
+ event.event);
+ qc_setdown(dev, DOWN_DRIVER_SUSPENDED);
+ } else {
+ DBG("device autosuspend\n");
+ }
+
+ if (event.event & PM_EVENT_SUSPEND) {
+ qc_stopread(dev);
+ usbnet->udev->reset_resume = 0;
+ iface->dev.power.power_state.event = event.event;
+ } else {
+ usbnet->udev->reset_resume = 1;
+ }
+
+ return usbnet_suspend(iface, event);
+}
+
+static int qc_resume(struct usb_interface *iface)
+{
+ struct usbnet *usbnet;
+ struct qcusbnet *dev;
+ int ret;
+ int oldstate;
+
+ if (iface == 0)
+ return -ENOMEM;
+
+ usbnet = usb_get_intfdata(iface);
+
+ if (!usbnet || !usbnet->net) {
+ DBG("failed to get netdevice\n");
+ return -ENXIO;
+ }
+
+ dev = (struct qcusbnet *)usbnet->data[0];
+ if (!dev) {
+ DBG("failed to get QMIDevice\n");
+ return -ENXIO;
+ }
+
+ oldstate = iface->dev.power.power_state.event;
+ iface->dev.power.power_state.event = PM_EVENT_ON;
+ DBG("resuming from power mode %d\n", oldstate);
+
+ if (oldstate & PM_EVENT_SUSPEND) {
+ qc_cleardown(dev, DOWN_DRIVER_SUSPENDED);
+
+ ret = usbnet_resume(iface);
+ if (ret) {
+ DBG("usbnet_resume error %d\n", ret);
+ return ret;
+ }
+
+ ret = qc_startread(dev);
+ if (ret) {
+ DBG("qc_startread error %d\n", ret);
+ return ret;
+ }
+
+ complete(&dev->worker.work);
+ } else {
+ DBG("nothing to resume\n");
+ return 0;
+ }
+
+ return ret;
+}
+
+static int qcnet_bind(struct usbnet *usbnet, struct usb_interface *iface)
+{
+ int numends;
+ int i;
+ struct usb_host_endpoint *endpoint = NULL;
+ struct usb_host_endpoint *in = NULL;
+ struct usb_host_endpoint *out = NULL;
+
+ if (iface->num_altsetting != 1) {
+ DBG("invalid num_altsetting %u\n", iface->num_altsetting);
+ return -EINVAL;
+ }
+
+ if (iface->cur_altsetting->desc.bInterfaceNumber != 0
+ && iface->cur_altsetting->desc.bInterfaceNumber != 5) {
+ DBG("invalid interface %d\n",
+ iface->cur_altsetting->desc.bInterfaceNumber);
+ return -EINVAL;
+ }
+
+ numends = iface->cur_altsetting->desc.bNumEndpoints;
+ for (i = 0; i < numends; i++) {
+ endpoint = iface->cur_altsetting->endpoint + i;
+ if (!endpoint) {
+ DBG("invalid endpoint %u\n", i);
+ return -EINVAL;
+ }
+
+ if (usb_endpoint_dir_in(&endpoint->desc)
+ && !usb_endpoint_xfer_int(&endpoint->desc)) {
+ in = endpoint;
+ } else if (!usb_endpoint_dir_out(&endpoint->desc)) {
+ out = endpoint;
+ }
+ }
+
+ if (!in || !out) {
+ DBG("invalid endpoints\n");
+ return -EINVAL;
+ }
+
+ if (usb_set_interface(usbnet->udev,
+ iface->cur_altsetting->desc.bInterfaceNumber, 0)) {
+ DBG("unable to set interface\n");
+ return -EINVAL;
+ }
+
+ usbnet->in = usb_rcvbulkpipe(usbnet->udev, in->desc.bEndpointAddress & USB_ENDPOINT_NUMBER_MASK);
+ usbnet->out = usb_sndbulkpipe(usbnet->udev, out->desc.bEndpointAddress & USB_ENDPOINT_NUMBER_MASK);
+
+ DBG("in %x, out %x\n",
+ in->desc.bEndpointAddress,
+ out->desc.bEndpointAddress);
+
+ return 0;
+}
+
+static void qcnet_unbind(struct usbnet *usbnet, struct usb_interface *iface)
+{
+ struct qcusbnet *dev = (struct qcusbnet *)usbnet->data[0];
+
+ netif_carrier_off(usbnet->net);
+ qc_deregister(dev);
+
+ kfree(usbnet->net->netdev_ops);
+ usbnet->net->netdev_ops = NULL;
+ /* drop the list's ref */
+ qcusbnet_put(dev);
+}
+
+static void qcnet_urbhook(struct urb *urb)
+{
+ unsigned long flags;
+ struct worker *worker = urb->context;
+ if (!worker) {
+ DBG("bad context\n");
+ return;
+ }
+
+ if (urb->status) {
+ DBG("urb finished with error %d\n", urb->status);
+ }
+
+ spin_lock_irqsave(&worker->active_lock, flags);
+ worker->active = ERR_PTR(-EAGAIN);
+ spin_unlock_irqrestore(&worker->active_lock, flags);
+ /* XXX-fix race against qcnet_stop()? */
+ complete(&worker->work);
+ usb_free_urb(urb);
+}
+
+static void qcnet_txtimeout(struct net_device *netdev)
+{
+ struct list_head *node, *tmp;
+ struct qcusbnet *dev;
+ struct worker *worker;
+ struct urbreq *req;
+ unsigned long activeflags, listflags;
+ struct usbnet *usbnet = netdev_priv(netdev);
+
+ if (!usbnet || !usbnet->net) {
+ DBG("failed to get usbnet device\n");
+ return;
+ }
+
+ dev = (struct qcusbnet *)usbnet->data[0];
+ if (!dev) {
+ DBG("failed to get QMIDevice\n");
+ return;
+ }
+ worker = &dev->worker;
+
+ DBG("\n");
+
+ spin_lock_irqsave(&worker->active_lock, activeflags);
+ if (worker->active)
+ usb_kill_urb(worker->active);
+ spin_unlock_irqrestore(&worker->active_lock, activeflags);
+
+ spin_lock_irqsave(&worker->urbs_lock, listflags);
+ list_for_each_safe(node, tmp, &worker->urbs) {
+ req = list_entry(node, struct urbreq, node);
+ usb_free_urb(req->urb);
+ list_del(&req->node);
+ kfree(req);
+ }
+ spin_unlock_irqrestore(&worker->urbs_lock, listflags);
+
+ complete(&worker->work);
+}
+
+static int qcnet_worker(void *arg)
+{
+ struct list_head *node, *tmp;
+ unsigned long activeflags, listflags;
+ struct urbreq *req;
+ int status;
+ struct usb_device *usbdev;
+ struct worker *worker = arg;
+ if (!worker) {
+ DBG("passed null pointer\n");
+ return -EINVAL;
+ }
+
+ usbdev = interface_to_usbdev(worker->iface);
+
+ DBG("traffic thread started\n");
+
+ while (!kthread_should_stop()) {
+ wait_for_completion_interruptible(&worker->work);
+
+ if (kthread_should_stop()) {
+ spin_lock_irqsave(&worker->active_lock, activeflags);
+ if (worker->active) {
+ usb_kill_urb(worker->active);
+ }
+ spin_unlock_irqrestore(&worker->active_lock, activeflags);
+
+ spin_lock_irqsave(&worker->urbs_lock, listflags);
+ list_for_each_safe(node, tmp, &worker->urbs) {
+ req = list_entry(node, struct urbreq, node);
+ usb_free_urb(req->urb);
+ list_del(&req->node);
+ kfree(req);
+ }
+ spin_unlock_irqrestore(&worker->urbs_lock, listflags);
+
+ break;
+ }
+
+ spin_lock_irqsave(&worker->active_lock, activeflags);
+ if (IS_ERR(worker->active) && PTR_ERR(worker->active) == -EAGAIN) {
+ worker->active = NULL;
+ spin_unlock_irqrestore(&worker->active_lock, activeflags);
+ usb_autopm_put_interface(worker->iface);
+ spin_lock_irqsave(&worker->active_lock, activeflags);
+ }
+
+ if (worker->active) {
+ spin_unlock_irqrestore(&worker->active_lock, activeflags);
+ continue;
+ }
+
+ spin_lock_irqsave(&worker->urbs_lock, listflags);
+ if (list_empty(&worker->urbs)) {
+ spin_unlock_irqrestore(&worker->urbs_lock, listflags);
+ spin_unlock_irqrestore(&worker->active_lock, activeflags);
+ continue;
+ }
+
+ req = list_first_entry(&worker->urbs, struct urbreq, node);
+ list_del(&req->node);
+ spin_unlock_irqrestore(&worker->urbs_lock, listflags);
+
+ worker->active = req->urb;
+ spin_unlock_irqrestore(&worker->active_lock, activeflags);
+
+ status = usb_autopm_get_interface(worker->iface);
+ if (status < 0) {
+ DBG("unable to autoresume interface: %d\n", status);
+ if (status == -EPERM) {
+ qc_suspend(worker->iface, PMSG_SUSPEND);
+ }
+
+ spin_lock_irqsave(&worker->urbs_lock, listflags);
+ list_add(&req->node, &worker->urbs);
+ spin_unlock_irqrestore(&worker->urbs_lock, listflags);
+
+ spin_lock_irqsave(&worker->active_lock, activeflags);
+ worker->active = NULL;
+ spin_unlock_irqrestore(&worker->active_lock, activeflags);
+
+ continue;
+ }
+
+ status = usb_submit_urb(worker->active, GFP_KERNEL);
+ if (status < 0) {
+ DBG("Failed to submit URB: %d. Packet dropped\n", status);
+ spin_lock_irqsave(&worker->active_lock, activeflags);
+ usb_free_urb(worker->active);
+ worker->active = NULL;
+ spin_unlock_irqrestore(&worker->active_lock, activeflags);
+ usb_autopm_put_interface(worker->iface);
+ complete(&worker->work);
+ }
+
+ kfree(req);
+ }
+
+ DBG("traffic thread exiting\n");
+ worker->thread = NULL;
+ return 0;
+}
+
+static int qcnet_startxmit(struct sk_buff *skb, struct net_device *netdev)
+{
+ unsigned long listflags;
+ struct qcusbnet *dev;
+ struct worker *worker;
+ struct urbreq *req;
+ void *data;
+ struct usbnet *usbnet = netdev_priv(netdev);
+
+ DBG("\n");
+
+ if (!usbnet || !usbnet->net) {
+ DBG("failed to get usbnet device\n");
+ return NETDEV_TX_BUSY;
+ }
+
+ dev = (struct qcusbnet *)usbnet->data[0];
+ if (!dev) {
+ DBG("failed to get QMIDevice\n");
+ return NETDEV_TX_BUSY;
+ }
+ worker = &dev->worker;
+
+ if (qc_isdown(dev, DOWN_DRIVER_SUSPENDED)) {
+ DBG("device is suspended\n");
+ dump_stack();
+ return NETDEV_TX_BUSY;
+ }
+
+ req = kmalloc(sizeof(*req), GFP_ATOMIC);
+ if (!req) {
+ DBG("unable to allocate URBList memory\n");
+ return NETDEV_TX_BUSY;
+ }
+
+ req->urb = usb_alloc_urb(0, GFP_ATOMIC);
+
+ if (!req->urb) {
+ kfree(req);
+ DBG("unable to allocate URB\n");
+ return NETDEV_TX_BUSY;
+ }
+
+ data = kmalloc(skb->len, GFP_ATOMIC);
+ if (!data) {
+ usb_free_urb(req->urb);
+ kfree(req);
+ DBG("unable to allocate URB data\n");
+ return NETDEV_TX_BUSY;
+ }
+ memcpy(data, skb->data, skb->len);
+
+ usb_fill_bulk_urb(req->urb, dev->usbnet->udev, dev->usbnet->out,
+ data, skb->len, qcnet_urbhook, worker);
+
+ spin_lock_irqsave(&worker->urbs_lock, listflags);
+ list_add_tail(&req->node, &worker->urbs);
+ spin_unlock_irqrestore(&worker->urbs_lock, listflags);
+
+ complete(&worker->work);
+
+ netdev->trans_start = jiffies;
+ dev_kfree_skb_any(skb);
+
+ return NETDEV_TX_OK;
+}
+
+static int qcnet_open(struct net_device *netdev)
+{
+ int status = 0;
+ struct qcusbnet *dev;
+ struct usbnet *usbnet = netdev_priv(netdev);
+
+ if (!usbnet) {
+ DBG("failed to get usbnet device\n");
+ return -ENXIO;
+ }
+
+ dev = (struct qcusbnet *)usbnet->data[0];
+ if (!dev) {
+ DBG("failed to get QMIDevice\n");
+ return -ENXIO;
+ }
+
+ DBG("\n");
+
+ dev->worker.iface = dev->iface;
+ INIT_LIST_HEAD(&dev->worker.urbs);
+ dev->worker.active = NULL;
+ spin_lock_init(&dev->worker.urbs_lock);
+ spin_lock_init(&dev->worker.active_lock);
+ init_completion(&dev->worker.work);
+
+ dev->worker.thread = kthread_run(qcnet_worker, &dev->worker, "qcnet_worker");
+ if (IS_ERR(dev->worker.thread)) {
+ DBG("AutoPM thread creation error\n");
+ return PTR_ERR(dev->worker.thread);
+ }
+
+ qc_cleardown(dev, DOWN_NET_IFACE_STOPPED);
+ if (dev->open) {
+ status = dev->open(netdev);
+ if (status == 0) {
+ usb_autopm_put_interface(dev->iface);
+ }
+ } else {
+ DBG("no USBNetOpen defined\n");
+ }
+
+ return status;
+}
+
+int qcnet_stop(struct net_device *netdev)
+{
+ struct qcusbnet *dev;
+ struct usbnet *usbnet = netdev_priv(netdev);
+
+ if (!usbnet || !usbnet->net) {
+ DBG("failed to get netdevice\n");
+ return -ENXIO;
+ }
+
+ dev = (struct qcusbnet *)usbnet->data[0];
+ if (!dev) {
+ DBG("failed to get QMIDevice\n");
+ return -ENXIO;
+ }
+
+ qc_setdown(dev, DOWN_NET_IFACE_STOPPED);
+ complete(&dev->worker.work);
+ kthread_stop(dev->worker.thread);
+ DBG("thread stopped\n");
+
+ if (dev->stop != NULL)
+ return dev->stop(netdev);
+ return 0;
+}
+
+static const struct driver_info qc_netinfo = {
+ .description = "QCUSBNet Ethernet Device",
+ .flags = FLAG_ETHER,
+ .bind = qcnet_bind,
+ .unbind = qcnet_unbind,
+ .data = 0,
+};
+
+#define MKVIDPID(v, p) \
+{ \
+ USB_DEVICE(v, p), \
+ .driver_info = (unsigned long)&qc_netinfo, \
+}
+
+static const struct usb_device_id qc_vidpids[] = {
+ MKVIDPID(0x05c6, 0x9215), /* Acer Gobi 2000 */
+ MKVIDPID(0x05c6, 0x9265), /* Asus Gobi 2000 */
+ MKVIDPID(0x16d8, 0x8002), /* CMOTech Gobi 2000 */
+ MKVIDPID(0x413c, 0x8186), /* Dell Gobi 2000 */
+ MKVIDPID(0x1410, 0xa010), /* Entourage Gobi 2000 */
+ MKVIDPID(0x1410, 0xa011), /* Entourage Gobi 2000 */
+ MKVIDPID(0x1410, 0xa012), /* Entourage Gobi 2000 */
+ MKVIDPID(0x1410, 0xa013), /* Entourage Gobi 2000 */
+ MKVIDPID(0x03f0, 0x251d), /* HP Gobi 2000 */
+ MKVIDPID(0x05c6, 0x9205), /* Lenovo Gobi 2000 */
+ MKVIDPID(0x05c6, 0x920b), /* Generic Gobi 2000 */
+ MKVIDPID(0x04da, 0x250f), /* Panasonic Gobi 2000 */
+ MKVIDPID(0x05c6, 0x9245), /* Samsung Gobi 2000 */
+ MKVIDPID(0x1199, 0x9001), /* Sierra Wireless Gobi 2000 */
+ MKVIDPID(0x1199, 0x9002), /* Sierra Wireless Gobi 2000 */
+ MKVIDPID(0x1199, 0x9003), /* Sierra Wireless Gobi 2000 */
+ MKVIDPID(0x1199, 0x9004), /* Sierra Wireless Gobi 2000 */
+ MKVIDPID(0x1199, 0x9005), /* Sierra Wireless Gobi 2000 */
+ MKVIDPID(0x1199, 0x9006), /* Sierra Wireless Gobi 2000 */
+ MKVIDPID(0x1199, 0x9007), /* Sierra Wireless Gobi 2000 */
+ MKVIDPID(0x1199, 0x9008), /* Sierra Wireless Gobi 2000 */
+ MKVIDPID(0x1199, 0x9009), /* Sierra Wireless Gobi 2000 */
+ MKVIDPID(0x1199, 0x900a), /* Sierra Wireless Gobi 2000 */
+ MKVIDPID(0x05c6, 0x9225), /* Sony Gobi 2000 */
+ MKVIDPID(0x05c6, 0x9235), /* Top Global Gobi 2000 */
+ MKVIDPID(0x05c6, 0x9275), /* iRex Technologies Gobi 2000 */
+
+ MKVIDPID(0x05c6, 0x920d), /* Qualcomm Gobi 3000 */
+ MKVIDPID(0x1410, 0xa021), /* Novatel Gobi 3000 */
+ { }
+};
+
+MODULE_DEVICE_TABLE(usb, qc_vidpids);
+
+static u8 nibble(unsigned char c)
+{
+ if (likely(isdigit(c)))
+ return c - '0';
+ c = toupper(c);
+ if (likely(isxdigit(c)))
+ return 10 + c - 'A';
+ return 0;
+}
+
+int qcnet_probe(struct usb_interface *iface, const struct usb_device_id *vidpids)
+{
+ int status;
+ struct usbnet *usbnet;
+ struct qcusbnet *dev;
+ struct net_device_ops *netdevops;
+ int i;
+ u8 *addr;
+
+ status = usbnet_probe(iface, vidpids);
+ if (status < 0) {
+ DBG("usbnet_probe failed %d\n", status);
+ return status;
+ }
+
+ usbnet = usb_get_intfdata(iface);
+
+ if (!usbnet || !usbnet->net) {
+ DBG("failed to get netdevice\n");
+ return -ENXIO;
+ }
+
+ dev = kmalloc(sizeof(struct qcusbnet), GFP_KERNEL);
+ if (!dev) {
+ DBG("failed to allocate device buffers\n");
+ return -ENOMEM;
+ }
+
+ usbnet->data[0] = (unsigned long)dev;
+
+ dev->usbnet = usbnet;
+
+ netdevops = kmalloc(sizeof(struct net_device_ops), GFP_KERNEL);
+ if (!netdevops) {
+ DBG("failed to allocate net device ops\n");
+ return -ENOMEM;
+ }
+ memcpy(netdevops, usbnet->net->netdev_ops, sizeof(struct net_device_ops));
+
+ dev->open = netdevops->ndo_open;
+ netdevops->ndo_open = qcnet_open;
+ dev->stop = netdevops->ndo_stop;
+ netdevops->ndo_stop = qcnet_stop;
+ netdevops->ndo_start_xmit = qcnet_startxmit;
+ netdevops->ndo_tx_timeout = qcnet_txtimeout;
+
+ usbnet->net->netdev_ops = netdevops;
+
+ memset(&(dev->usbnet->net->stats), 0, sizeof(struct net_device_stats));
+
+ dev->iface = iface;
+ memset(&(dev->meid), '0', 14);
+
+ dev->valid = false;
+ memset(&dev->qmi, 0, sizeof(dev->qmi));
+
+ dev->qmi.devclass = devclass;
+
+ kref_init(&dev->refcount);
+ INIT_LIST_HEAD(&dev->node);
+ INIT_LIST_HEAD(&dev->qmi.clients);
+ init_completion(&dev->worker.work);
+ spin_lock_init(&dev->qmi.clients_lock);
+
+ dev->down = 0;
+ qc_setdown(dev, DOWN_NO_NDIS_CONNECTION);
+ qc_setdown(dev, DOWN_NET_IFACE_STOPPED);
+
+ status = qc_register(dev);
+ if (status) {
+ qc_deregister(dev);
+ } else {
+ mutex_lock(&qcusbnet_lock);
+ /* Give our initial ref to the list */
+ list_add(&dev->node, &qcusbnet_list);
+ mutex_unlock(&qcusbnet_lock);
+ }
+ /* After calling qc_register, MEID is valid */
+ addr = &usbnet->net->dev_addr[0];
+ for (i = 0; i < 6; i++)
+ addr[i] = (nibble(dev->meid[i*2+2]) << 4)+
+ nibble(dev->meid[i*2+3]);
+ addr[0] &= 0xfe; /* clear multicast bit */
+ addr[0] |= 0x02; /* set local assignment bit (IEEE802) */
+
+ return status;
+}
+EXPORT_SYMBOL_GPL(qcnet_probe);
+
+static struct usb_driver qcusbnet = {
+ .name = "gobi",
+ .id_table = qc_vidpids,
+ .probe = qcnet_probe,
+ .disconnect = usbnet_disconnect,
+ .suspend = qc_suspend,
+ .resume = qc_resume,
+ .supports_autosuspend = true,
+};
+
+static int __init modinit(void)
+{
+ devclass = class_create(THIS_MODULE, "QCQMI");
+ if (IS_ERR(devclass)) {
+ DBG("error at class_create %ld\n", PTR_ERR(devclass));
+ return -ENOMEM;
+ }
+ printk(KERN_INFO "%s: %s\n", DRIVER_DESC, DRIVER_VERSION);
+ return usb_register(&qcusbnet);
+}
+module_init(modinit);
+
+static void __exit modexit(void)
+{
+ usb_deregister(&qcusbnet);
+ class_destroy(devclass);
+}
+module_exit(modexit);
+
+MODULE_VERSION(DRIVER_VERSION);
+MODULE_AUTHOR(DRIVER_AUTHOR);
+MODULE_DESCRIPTION(DRIVER_DESC);
+MODULE_LICENSE("Dual BSD/GPL");
+
+module_param(qcusbnet_debug, bool, S_IRUGO | S_IWUSR);
+MODULE_PARM_DESC(qcusbnet_debug, "Debugging enabled or not");
diff --git a/drivers/net/usb/gobi/qcusbnet.h b/drivers/net/usb/gobi/qcusbnet.h
new file mode 100644
index 0000000..00c3a7e
--- /dev/null
+++ b/drivers/net/usb/gobi/qcusbnet.h
@@ -0,0 +1,28 @@
+/* qcusbnet.h - gobi network device header
+ * Copyright (c) 2010, Code Aurora Forum. All rights reserved.
+
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 and
+ * only version 2 as published by the Free Software Foundation.
+
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
+ * 02110-1301, USA.
+ */
+
+#ifndef QCUSBNET_QCUSBNET_H
+#define QCUSBNET_QCUSBNET_H
+
+#include "structs.h"
+
+extern int qc_suspend(struct usb_interface *iface, pm_message_t event);
+extern void qcusbnet_put(struct qcusbnet *dev);
+extern struct qcusbnet *qcusbnet_get(struct qcusbnet *dev);
+
+#endif /* !QCUSBNET_QCUSBNET_H */
diff --git a/drivers/net/usb/gobi/qmi.c b/drivers/net/usb/gobi/qmi.c
new file mode 100644
index 0000000..cdbdbaf
--- /dev/null
+++ b/drivers/net/usb/gobi/qmi.c
@@ -0,0 +1,358 @@
+/* qmi.c - QMI protocol implementation
+ * Copyright (c) 2010, Code Aurora Forum. All rights reserved.
+
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 and
+ * only version 2 as published by the Free Software Foundation.
+
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
+ * 02110-1301, USA.
+ */
+
+#include "qmi.h"
+
+#include <linux/slab.h>
+
+struct qmux {
+ u8 tf; /* always 1 */
+ u16 len;
+ u8 ctrl;
+ u8 service;
+ u8 qmicid;
+} __attribute__((__packed__));
+
+struct getcid_req {
+ struct qmux header;
+ u8 req;
+ u8 tid;
+ u16 msgid;
+ u16 tlvsize;
+ u8 service;
+ u16 size;
+ u8 qmisvc;
+} __attribute__((__packed__));
+
+struct releasecid_req {
+ struct qmux header;
+ u8 req;
+ u8 tid;
+ u16 msgid;
+ u16 tlvsize;
+ u8 rlscid;
+ u16 size;
+ u16 cid;
+} __attribute__((__packed__));
+
+struct ready_req {
+ struct qmux header;
+ u8 req;
+ u8 tid;
+ u16 msgid;
+ u16 tlvsize;
+} __attribute__((__packed__));
+
+struct seteventreport_req {
+ struct qmux header;
+ u8 req;
+ u16 tid;
+ u16 msgid;
+ u16 tlvsize;
+ u8 reportchanrate;
+ u16 size;
+ u8 period;
+ u32 mask;
+} __attribute__((__packed__));
+
+struct getpkgsrvcstatus_req {
+ struct qmux header;
+ u8 req;
+ u16 tid;
+ u16 msgid;
+ u16 tlvsize;
+} __attribute__((__packed__));
+
+struct getmeid_req {
+ struct qmux header;
+ u8 req;
+ u16 tid;
+ u16 msgid;
+ u16 tlvsize;
+} __attribute__((__packed__));
+
+const size_t qmux_size = sizeof(struct qmux);
+
+void *qmictl_new_getcid(u8 tid, u8 svctype, size_t *size)
+{
+ struct getcid_req *req = kmalloc(sizeof(*req), GFP_KERNEL);
+ if (!req)
+ return NULL;
+ req->req = 0x00;
+ req->tid = tid;
+ req->msgid = 0x0022;
+ req->tlvsize = 0x0004;
+ req->service = 0x01;
+ req->size = 0x0001;
+ req->qmisvc = svctype;
+ *size = sizeof(*req);
+ return req;
+}
+
+void *qmictl_new_releasecid(u8 tid, u16 cid, size_t *size)
+{
+ struct releasecid_req *req = kmalloc(sizeof(*req), GFP_KERNEL);
+ if (!req)
+ return NULL;
+ req->req = 0x00;
+ req->tid = tid;
+ req->msgid = 0x0023;
+ req->tlvsize = 0x05;
+ req->rlscid = 0x01;
+ req->size = 0x0002;
+ req->cid = cid;
+ *size = sizeof(*req);
+ return req;
+}
+
+void *qmictl_new_ready(u8 tid, size_t *size)
+{
+ struct ready_req *req = kmalloc(sizeof(*req), GFP_KERNEL);
+ if (!req)
+ return NULL;
+ req->req = 0x00;
+ req->tid = tid;
+ req->msgid = 0x21;
+ req->tlvsize = 0;
+ *size = sizeof(*req);
+ return req;
+}
+
+void *qmiwds_new_seteventreport(u8 tid, size_t *size)
+{
+ struct seteventreport_req *req = kmalloc(sizeof(*req), GFP_KERNEL);
+ req->req = 0x00;
+ req->tid = tid;
+ req->msgid = 0x0001;
+ req->tlvsize = 0x0008;
+ req->reportchanrate = 0x11;
+ req->size = 0x0005;
+ req->period = 0x01;
+ req->mask = 0x000000ff;
+ *size = sizeof(*req);
+ return req;
+}
+
+void *qmiwds_new_getpkgsrvcstatus(u8 tid, size_t *size)
+{
+ struct getpkgsrvcstatus_req *req = kmalloc(sizeof(*req), GFP_KERNEL);
+ if (!req)
+ return NULL;
+ req->req = 0x00;
+ req->tid = tid;
+ req->msgid = 0x22;
+ req->tlvsize = 0x0000;
+ *size = sizeof(*req);
+ return req;
+}
+
+void *qmidms_new_getmeid(u8 tid, size_t *size)
+{
+ struct getmeid_req *req = kmalloc(sizeof(*req), GFP_KERNEL);
+ if (!req)
+ return NULL;
+ req->req = 0x00;
+ req->tid = tid;
+ req->msgid = 0x25;
+ req->tlvsize = 0x0000;
+ *size = sizeof(*req);
+ return req;
+}
+
+int qmux_parse(u16 *cid, void *buf, size_t size)
+{
+ struct qmux *qmux = buf;
+
+ if (!buf || size < 12)
+ return -ENOMEM;
+
+ if (qmux->tf != 1 || qmux->len != size - 1 || qmux->ctrl != 0x80)
+ return -EINVAL;
+
+ *cid = (qmux->qmicid << 8) + qmux->service;
+ return sizeof(*qmux);
+}
+
+int qmux_fill(u16 cid, void *buf, size_t size)
+{
+ struct qmux *qmux = buf;
+
+ if (!buf || size < sizeof(*qmux))
+ return -ENOMEM;
+
+ qmux->tf = 1;
+ qmux->len = size - 1;
+ qmux->ctrl = 0;
+ qmux->service = cid & 0xff;
+ qmux->qmicid = cid >> 8;
+ return 0;
+}
+
+static u16 tlv_get(void *msg, u16 msgsize, u8 type, void *buf, u16 bufsize)
+{
+ u16 pos;
+ u16 msize = 0;
+
+ if (!msg || !buf)
+ return -ENOMEM;
+
+ for (pos = 4; pos + 3 < msgsize; pos += msize + 3) {
+ msize = *(u16 *)(msg + pos + 1);
+ if (*(u8 *)(msg + pos) == type) {
+ if (bufsize < msize)
+ return -ENOMEM;
+
+ memcpy(buf, msg + pos + 3, msize);
+ return msize;
+ }
+ }
+
+ return -ENOMSG;
+}
+
+int qmi_msgisvalid(void *msg, u16 size)
+{
+ char tlv[4];
+
+ if (tlv_get(msg, size, 2, &tlv[0], 4) == 4) {
+ if (*(u16 *)&tlv[0] != 0)
+ return *(u16 *)&tlv[2];
+ else
+ return 0;
+ }
+ return -ENOMSG;
+}
+
+int qmi_msgid(void *msg, u16 size)
+{
+ return size < 2 ? -ENODATA : *(u16 *)msg;
+}
+
+int qmictl_alloccid_resp(void *buf, u16 size, u16 *cid)
+{
+ int result;
+ u8 offset = sizeof(struct qmux) + 2;
+
+ if (!buf || size < offset)
+ return -ENOMEM;
+
+ buf = buf + offset;
+ size -= offset;
+
+ result = qmi_msgid(buf, size);
+ if (result != 0x22)
+ return -EFAULT;
+
+ result = qmi_msgisvalid(buf, size);
+ if (result != 0)
+ return -EFAULT;
+
+ result = tlv_get(buf, size, 0x01, cid, 2);
+ if (result != 2)
+ return -EFAULT;
+
+ return 0;
+}
+
+int qmictl_freecid_resp(void *buf, u16 size)
+{
+ int result;
+ u8 offset = sizeof(struct qmux) + 2;
+
+ if (!buf || size < offset)
+ return -ENOMEM;
+
+ buf = buf + offset;
+ size -= offset;
+
+ result = qmi_msgid(buf, size);
+ if (result != 0x23)
+ return -EFAULT;
+
+ result = qmi_msgisvalid(buf, size);
+ if (result != 0)
+ return -EFAULT;
+
+ return 0;
+}
+
+int qmiwds_event_resp(void *buf, u16 size, struct qmiwds_stats *stats)
+{
+ int result;
+ u8 status[2];
+
+ u8 offset = sizeof(struct qmux) + 3;
+
+ if (!buf || size < offset || !stats)
+ return -ENOMEM;
+
+ buf = buf + offset;
+ size -= offset;
+
+ result = qmi_msgid(buf, size);
+ if (result == 0x01) {
+ tlv_get(buf, size, 0x10, &stats->txok, 4);
+ tlv_get(buf, size, 0x11, &stats->rxok, 4);
+ tlv_get(buf, size, 0x12, &stats->txerr, 4);
+ tlv_get(buf, size, 0x13, &stats->rxerr, 4);
+ tlv_get(buf, size, 0x14, &stats->txofl, 4);
+ tlv_get(buf, size, 0x15, &stats->rxofl, 4);
+ tlv_get(buf, size, 0x19, &stats->txbytesok, 8);
+ tlv_get(buf, size, 0x1A, &stats->rxbytesok, 8);
+ } else if (result == 0x22) {
+ result = tlv_get(buf, size, 0x01, &status[0], 2);
+ if (result >= 1)
+ stats->linkstate = status[0] == 0x02;
+ if (result == 2)
+ stats->reconfigure = status[1] == 0x01;
+
+ if (result < 0)
+ return result;
+ } else {
+ return -EFAULT;
+ }
+
+ return 0;
+}
+
+int qmidms_meid_resp(void *buf, u16 size, char *meid, int meidsize)
+{
+ int result;
+
+ u8 offset = sizeof(struct qmux) + 3;
+
+ if (!buf || size < offset || meidsize < 14)
+ return -ENOMEM;
+
+ buf = buf + offset;
+ size -= offset;
+
+ result = qmi_msgid(buf, size);
+ if (result != 0x25)
+ return -EFAULT;
+
+ result = qmi_msgisvalid(buf, size);
+ if (result)
+ return -EFAULT;
+
+ result = tlv_get(buf, size, 0x12, meid, 14);
+ if (result != 14)
+ return -EFAULT;
+
+ return 0;
+}
diff --git a/drivers/net/usb/gobi/qmi.h b/drivers/net/usb/gobi/qmi.h
new file mode 100644
index 0000000..7954790
--- /dev/null
+++ b/drivers/net/usb/gobi/qmi.h
@@ -0,0 +1,67 @@
+/* qmi.h - QMI protocol header
+ * Copyright (c) 2010, Code Aurora Forum. All rights reserved.
+
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 and
+ * only version 2 as published by the Free Software Foundation.
+
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
+ * 02110-1301, USA.
+ */
+
+#ifndef QCUSBNET_QMI_H
+#define QCUSBNET_QMI_H
+
+#include <linux/types.h>
+
+#define QMICTL 0
+#define QMIWDS 1
+#define QMIDMS 2
+
+#define true 1
+#define false 0
+
+#define ENOMEM 12
+#define EFAULT 14
+#define EINVAL 22
+#define ENOMSG 42
+#define ENODATA 61
+
+int qmux_parse(u16 *cid, void *buf, size_t size);
+int qmux_fill(u16 cid, void *buf, size_t size);
+
+extern const size_t qmux_size;
+
+void *qmictl_new_getcid(u8 tid, u8 svctype, size_t *size);
+void *qmictl_new_releasecid(u8 tid, u16 cid, size_t *size);
+void *qmictl_new_ready(u8 tid, size_t *size);
+void *qmiwds_new_seteventreport(u8 tid, size_t *size);
+void *qmiwds_new_getpkgsrvcstatus(u8 tid, size_t *size);
+void *qmidms_new_getmeid(u8 tid, size_t *size);
+
+struct qmiwds_stats {
+ u32 txok;
+ u32 rxok;
+ u32 txerr;
+ u32 rxerr;
+ u32 txofl;
+ u32 rxofl;
+ u64 txbytesok;
+ u64 rxbytesok;
+ bool linkstate;
+ bool reconfigure;
+};
+
+int qmictl_alloccid_resp(void *buf, u16 size, u16 *cid);
+int qmictl_freecid_resp(void *buf, u16 size);
+int qmiwds_event_resp(void *buf, u16 size, struct qmiwds_stats *stats);
+int qmidms_meid_resp(void *buf, u16 size, char *meid, int meidsize);
+
+#endif /* !QCUSBNET_QMI_H */
diff --git a/drivers/net/usb/gobi/qmidevice.c b/drivers/net/usb/gobi/qmidevice.c
new file mode 100644
index 0000000..be61dda
--- /dev/null
+++ b/drivers/net/usb/gobi/qmidevice.c
@@ -0,0 +1,1562 @@
+/* qmidevice.c - gobi QMI device
+ * Copyright (c) 2010, Code Aurora Forum. All rights reserved.
+
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 and
+ * only version 2 as published by the Free Software Foundation.
+
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
+ * 02110-1301, USA.
+ */
+
+#include "qmidevice.h"
+#include "qcusbnet.h"
+
+struct readreq {
+ struct list_head node;
+ void *data;
+ u16 tid;
+ u16 size;
+};
+
+struct notifyreq {
+ struct list_head node;
+ void (*func)(struct qcusbnet *, u16, void *);
+ u16 tid;
+ void *data;
+};
+
+struct client {
+ struct list_head node;
+ u16 cid;
+ struct list_head reads;
+ struct list_head notifies;
+ struct list_head urbs;
+};
+
+struct urbsetup {
+ u8 type;
+ u8 code;
+ u16 value;
+ u16 index;
+ u16 len;
+};
+
+struct qmihandle {
+ u16 cid;
+ struct qcusbnet *dev;
+};
+
+extern int qcusbnet_debug;
+static int qcusbnet2k_fwdelay;
+
+static bool device_valid(struct qcusbnet *dev);
+static struct client *client_bycid(struct qcusbnet *dev, u16 cid);
+static bool client_addread(struct qcusbnet *dev, u16 cid, u16 tid, void *data, u16 size);
+static bool client_delread(struct qcusbnet *dev, u16 cid, u16 tid, void **data, u16 *size);
+static bool client_addnotify(struct qcusbnet *dev, u16 cid, u16 tid,
+ void (*hook)(struct qcusbnet *, u16 cid, void *),
+ void *data);
+static bool client_notify(struct qcusbnet *dev, u16 cid, u16 tid);
+static bool client_addurb(struct qcusbnet *dev, u16 cid, struct urb *urb);
+static struct urb *client_delurb(struct qcusbnet *dev, u16 cid);
+
+static int resubmit_int_urb(struct urb *urb);
+
+static int devqmi_open(struct inode *inode, struct file *file);
+static int devqmi_ioctl(struct inode *inode, struct file *file, unsigned int cmd, unsigned long arg);
+static int devqmi_release(struct inode *inode, struct file *file);
+static ssize_t devqmi_read(struct file *file, char __user *buf, size_t size,
+ loff_t *pos);
+static ssize_t devqmi_write(struct file *file, const char __user *buf,
+ size_t size, loff_t *pos);
+
+static bool qmi_ready(struct qcusbnet *dev, u16 timeout);
+static void wds_callback(struct qcusbnet *dev, u16 cid, void *data);
+static int setup_wds_callback(struct qcusbnet *dev);
+static int qmidms_getmeid(struct qcusbnet *dev);
+
+#define IOCTL_QMI_GET_SERVICE_FILE (0x8BE0 + 1)
+#define IOCTL_QMI_GET_DEVICE_VIDPID (0x8BE0 + 2)
+#define IOCTL_QMI_GET_DEVICE_MEID (0x8BE0 + 3)
+#define IOCTL_QMI_CLOSE (0x8BE0 + 4)
+#define CDC_GET_ENCAPSULATED_RESPONSE 0x01A1ll
+#define CDC_CONNECTION_SPEED_CHANGE 0x08000000002AA1ll
+
+static const struct file_operations devqmi_fops = {
+ .owner = THIS_MODULE,
+ .read = devqmi_read,
+ .write = devqmi_write,
+ .ioctl = devqmi_ioctl,
+ .open = devqmi_open,
+ .release = devqmi_release,
+};
+
+#ifdef CONFIG_SMP
+static inline void assert_locked(struct qcusbnet *dev)
+{
+ BUG_ON(!spin_is_locked(&dev->qmi.clients_lock));
+}
+#else
+static inline void assert_locked(struct qcusbnet *dev)
+{
+
+}
+#endif
+
+static bool device_valid(struct qcusbnet *dev)
+{
+ return dev && dev->valid;
+}
+
+void qc_setdown(struct qcusbnet *dev, u8 reason)
+{
+ set_bit(reason, &dev->down);
+ netif_carrier_off(dev->usbnet->net);
+}
+
+void qc_cleardown(struct qcusbnet *dev, u8 reason)
+{
+ clear_bit(reason, &dev->down);
+ if (!dev->down)
+ netif_carrier_on(dev->usbnet->net);
+}
+
+bool qc_isdown(struct qcusbnet *dev, u8 reason)
+{
+ return test_bit(reason, &dev->down);
+}
+
+static int resubmit_int_urb(struct urb *urb)
+{
+ int status;
+ int interval;
+ if (!urb || !urb->dev)
+ return -EINVAL;
+ interval = urb->dev->speed == USB_SPEED_HIGH ? 7 : 3;
+ usb_fill_int_urb(urb, urb->dev, urb->pipe, urb->transfer_buffer,
+ urb->transfer_buffer_length, urb->complete,
+ urb->context, interval);
+ status = usb_submit_urb(urb, GFP_ATOMIC);
+ if (status)
+ DBG("status %d", status);
+ return status;
+}
+
+static void read_callback(struct urb *urb)
+{
+ struct list_head *node;
+ int result;
+ u16 cid;
+ struct client *client;
+ void *data;
+ void *copy;
+ u16 size;
+ struct qcusbnet *dev;
+ unsigned long flags;
+ u16 tid;
+
+ if (!urb) {
+ DBG("bad read URB\n");
+ return;
+ }
+
+ dev = urb->context;
+ if (!device_valid(dev)) {
+ DBG("Invalid device!\n");
+ return;
+ }
+
+ if (urb->status) {
+ DBG("Read status = %d\n", urb->status);
+ resubmit_int_urb(dev->qmi.inturb);
+ return;
+ }
+
+ DBG("Read %d bytes\n", urb->actual_length);
+
+ data = urb->transfer_buffer;
+ size = urb->actual_length;
+
+ if (qcusbnet_debug)
+ print_hex_dump(KERN_INFO, "gobi-read: ", DUMP_PREFIX_OFFSET,
+ 16, 1, data, size, true);
+
+ result = qmux_parse(&cid, data, size);
+ if (result < 0) {
+ DBG("Read error parsing QMUX %d\n", result);
+ resubmit_int_urb(dev->qmi.inturb);
+ return;
+ }
+
+ if (size < result + 3) {
+ DBG("Data buffer too small to parse\n");
+ resubmit_int_urb(dev->qmi.inturb);
+ return;
+ }
+
+ if (cid == QMICTL)
+ tid = *(u8 *)(data + result + 1);
+ else
+ tid = *(u16 *)(data + result + 1);
+ spin_lock_irqsave(&dev->qmi.clients_lock, flags);
+
+ list_for_each(node, &dev->qmi.clients) {
+ client = list_entry(node, struct client, node);
+ if (client->cid == cid || (client->cid | 0xff00) == cid) {
+ copy = kmalloc(size, GFP_ATOMIC);
+ memcpy(copy, data, size);
+ if (!client_addread(dev, client->cid, tid, copy, size)) {
+ DBG("Error allocating pReadMemListEntry "
+ "read will be discarded\n");
+ kfree(copy);
+ spin_unlock_irqrestore(&dev->qmi.clients_lock, flags);
+ resubmit_int_urb(dev->qmi.inturb);
+ return;
+ }
+
+ DBG("Creating new readListEntry for client 0x%04X, TID %x\n",
+ cid, tid);
+
+ client_notify(dev, client->cid, tid);
+
+ if (cid >> 8 != 0xff)
+ break;
+ }
+ }
+
+ spin_unlock_irqrestore(&dev->qmi.clients_lock, flags);
+ resubmit_int_urb(dev->qmi.inturb);
+}
+
+static void int_callback(struct urb *urb)
+{
+ int status;
+ int interval;
+ struct qcusbnet *dev = (struct qcusbnet *)urb->context;
+
+ if (!device_valid(dev)) {
+ DBG("Invalid device!\n");
+ return;
+ }
+
+ if (urb->status) {
+ DBG("Int status = %d\n", urb->status);
+ if (urb->status != -EOVERFLOW)
+ return;
+ } else {
+ if ((urb->actual_length == 8) &&
+ (*(u64 *)urb->transfer_buffer == CDC_GET_ENCAPSULATED_RESPONSE)) {
+ usb_fill_control_urb(dev->qmi.readurb, dev->usbnet->udev,
+ usb_rcvctrlpipe(dev->usbnet->udev, 0),
+ (unsigned char *)dev->qmi.readsetup,
+ dev->qmi.readbuf,
+ DEFAULT_READ_URB_LENGTH,
+ read_callback, dev);
+ status = usb_submit_urb(dev->qmi.readurb, GFP_ATOMIC);
+ if (status) {
+ DBG("Error submitting Read URB %d\n", status);
+ return;
+ }
+ } else if ((urb->actual_length == 16) &&
+ (*(u64 *)urb->transfer_buffer == CDC_CONNECTION_SPEED_CHANGE)) {
+ /* if upstream or downstream is 0, stop traffic.
+ * Otherwise resume it */
+ if ((*(u32 *)(urb->transfer_buffer + 8) == 0) ||
+ (*(u32 *)(urb->transfer_buffer + 12) == 0)) {
+ qc_setdown(dev, DOWN_CDC_CONNECTION_SPEED);
+ DBG("traffic stopping due to CONNECTION_SPEED_CHANGE\n");
+ } else {
+ qc_cleardown(dev, DOWN_CDC_CONNECTION_SPEED);
+ DBG("resuming traffic due to CONNECTION_SPEED_CHANGE\n");
+ }
+ } else {
+ DBG("ignoring invalid interrupt in packet\n");
+ if (qcusbnet_debug)
+ print_hex_dump(KERN_INFO, "gobi-int: ",
+ DUMP_PREFIX_OFFSET, 16, 1,
+ urb->transfer_buffer,
+ urb->actual_length, true);
+ }
+ }
+
+ resubmit_int_urb(dev->qmi.inturb);
+ return;
+}
+
+int qc_startread(struct qcusbnet *dev)
+{
+ int interval;
+
+ if (!device_valid(dev)) {
+ DBG("Invalid device!\n");
+ return -ENXIO;
+ }
+
+ dev->qmi.readurb = usb_alloc_urb(0, GFP_KERNEL);
+ if (!dev->qmi.readurb) {
+ DBG("Error allocating read urb\n");
+ return -ENOMEM;
+ }
+
+ dev->qmi.inturb = usb_alloc_urb(0, GFP_KERNEL);
+ if (!dev->qmi.inturb) {
+ usb_free_urb(dev->qmi.readurb);
+ DBG("Error allocating int urb\n");
+ return -ENOMEM;
+ }
+
+ dev->qmi.readbuf = kmalloc(DEFAULT_READ_URB_LENGTH, GFP_KERNEL);
+ if (!dev->qmi.readbuf) {
+ usb_free_urb(dev->qmi.readurb);
+ usb_free_urb(dev->qmi.inturb);
+ DBG("Error allocating read buffer\n");
+ return -ENOMEM;
+ }
+
+ dev->qmi.intbuf = kmalloc(DEFAULT_READ_URB_LENGTH, GFP_KERNEL);
+ if (!dev->qmi.intbuf) {
+ usb_free_urb(dev->qmi.readurb);
+ usb_free_urb(dev->qmi.inturb);
+ kfree(dev->qmi.readbuf);
+ DBG("Error allocating int buffer\n");
+ return -ENOMEM;
+ }
+
+ dev->qmi.readsetup = kmalloc(sizeof(*dev->qmi.readsetup), GFP_KERNEL);
+ if (!dev->qmi.readsetup) {
+ usb_free_urb(dev->qmi.readurb);
+ usb_free_urb(dev->qmi.inturb);
+ kfree(dev->qmi.readbuf);
+ kfree(dev->qmi.intbuf);
+ DBG("Error allocating setup packet buffer\n");
+ return -ENOMEM;
+ }
+
+ dev->qmi.readsetup->type = 0xA1;
+ dev->qmi.readsetup->code = 1;
+ dev->qmi.readsetup->value = 0;
+ dev->qmi.readsetup->index = 0;
+ dev->qmi.readsetup->len = DEFAULT_READ_URB_LENGTH;
+
+ interval = (dev->usbnet->udev->speed == USB_SPEED_HIGH) ? 7 : 3;
+
+ usb_fill_int_urb(dev->qmi.inturb, dev->usbnet->udev,
+ usb_rcvintpipe(dev->usbnet->udev, 0x81),
+ dev->qmi.intbuf, DEFAULT_READ_URB_LENGTH,
+ int_callback, dev, interval);
+ return usb_submit_urb(dev->qmi.inturb, GFP_KERNEL);
+}
+
+void qc_stopread(struct qcusbnet *dev)
+{
+ if (dev->qmi.readurb) {
+ DBG("Killing read URB\n");
+ usb_kill_urb(dev->qmi.readurb);
+ }
+
+ if (dev->qmi.inturb) {
+ DBG("Killing int URB\n");
+ usb_kill_urb(dev->qmi.inturb);
+ }
+
+ kfree(dev->qmi.readsetup);
+ dev->qmi.readsetup = NULL;
+ kfree(dev->qmi.readbuf);
+ dev->qmi.readbuf = NULL;
+ kfree(dev->qmi.intbuf);
+ dev->qmi.intbuf = NULL;
+
+ usb_free_urb(dev->qmi.readurb);
+ dev->qmi.readurb = NULL;
+ usb_free_urb(dev->qmi.inturb);
+ dev->qmi.inturb = NULL;
+}
+
+static int read_async(struct qcusbnet *dev, u16 cid, u16 tid,
+ void (*hook)(struct qcusbnet *, u16, void *),
+ void *data)
+{
+ struct list_head *node;
+ struct client *client;
+ struct readreq *readreq;
+
+ unsigned long flags;
+
+ if (!device_valid(dev)) {
+ DBG("Invalid device!\n");
+ return -ENXIO;
+ }
+
+ spin_lock_irqsave(&dev->qmi.clients_lock, flags);
+
+ client = client_bycid(dev, cid);
+ if (!client) {
+ DBG("Could not find matching client ID 0x%04X\n", cid);
+ spin_unlock_irqrestore(&dev->qmi.clients_lock, flags);
+ return -ENXIO;
+ }
+
+ list_for_each(node, &client->reads) {
+ readreq = list_entry(node, struct readreq, node);
+ if (!tid || tid == readreq->tid) {
+ spin_unlock_irqrestore(&dev->qmi.clients_lock, flags);
+ hook(dev, cid, data);
+ return 0;
+ }
+ }
+
+ if (!client_addnotify(dev, cid, tid, hook, data))
+ DBG("Unable to register for notification\n");
+
+ spin_unlock_irqrestore(&dev->qmi.clients_lock, flags);
+ return 0;
+}
+
+static void upsem(struct qcusbnet *dev, u16 cid, void *data)
+{
+ DBG("0x%04X\n", cid);
+ up((struct semaphore *)data);
+}
+
+static int read_sync(struct qcusbnet *dev, void **buf, u16 cid, u16 tid)
+{
+ struct list_head *node;
+ int result;
+ struct client *client;
+ struct notifyreq *notify;
+ struct semaphore sem;
+ void *data;
+ unsigned long flags;
+ u16 size;
+
+ if (!device_valid(dev)) {
+ DBG("Invalid device!\n");
+ return -ENXIO;
+ }
+
+ spin_lock_irqsave(&dev->qmi.clients_lock, flags);
+
+ client = client_bycid(dev, cid);
+ if (!client) {
+ DBG("Could not find matching client ID 0x%04X\n", cid);
+ spin_unlock_irqrestore(&dev->qmi.clients_lock, flags);
+ return -ENXIO;
+ }
+
+ while (!client_delread(dev, cid, tid, &data, &size)) {
+ sema_init(&sem, 0);
+ if (!client_addnotify(dev, cid, tid, upsem, &sem)) {
+ DBG("unable to register for notification\n");
+ spin_unlock_irqrestore(&dev->qmi.clients_lock, flags);
+ return -EFAULT;
+ }
+
+ spin_unlock_irqrestore(&dev->qmi.clients_lock, flags);
+
+ result = down_interruptible(&sem);
+ if (result) {
+ DBG("Interrupted %d\n", result);
+ spin_lock_irqsave(&dev->qmi.clients_lock, flags);
+ list_for_each(node, &client->notifies) {
+ notify = list_entry(node, struct notifyreq, node);
+ if (notify->data == &sem) {
+ list_del(¬ify->node);
+ kfree(notify);
+ break;
+ }
+ }
+
+ spin_unlock_irqrestore(&dev->qmi.clients_lock, flags);
+ return -EINTR;
+ }
+
+ if (!device_valid(dev)) {
+ DBG("Invalid device!\n");
+ return -ENXIO;
+ }
+
+ spin_lock_irqsave(&dev->qmi.clients_lock, flags);
+ }
+
+ spin_unlock_irqrestore(&dev->qmi.clients_lock, flags);
+ *buf = data;
+ return size;
+}
+
+static void write_callback(struct urb *urb)
+{
+ if (!urb) {
+ DBG("null urb\n");
+ return;
+ }
+
+ DBG("Write status/size %d/%d\n", urb->status, urb->actual_length);
+ up((struct semaphore *)urb->context);
+}
+
+static int write_sync(struct qcusbnet *dev, char *buf, int size, u16 cid)
+{
+ int result;
+ struct semaphore sem;
+ struct urb *urb;
+ struct urbsetup setup;
+ unsigned long flags;
+
+ if (!device_valid(dev)) {
+ DBG("Invalid device!\n");
+ return -ENXIO;
+ }
+
+ urb = usb_alloc_urb(0, GFP_KERNEL);
+ if (!urb) {
+ DBG("URB mem error\n");
+ return -ENOMEM;
+ }
+
+ result = qmux_fill(cid, buf, size);
+ if (result < 0) {
+ usb_free_urb(urb);
+ return result;
+ }
+
+ /* CDC Send Encapsulated Request packet */
+ setup.type = 0x21;
+ setup.code = 0;
+ setup.value = 0;
+ setup.index = 0;
+ setup.len = 0;
+ setup.len = size;
+
+ usb_fill_control_urb(urb, dev->usbnet->udev,
+ usb_sndctrlpipe(dev->usbnet->udev, 0),
+ (unsigned char *)&setup, (void *)buf, size,
+ NULL, dev);
+
+ DBG("Actual Write:\n");
+ if (qcusbnet_debug)
+ print_hex_dump(KERN_INFO, "gobi-write: ", DUMP_PREFIX_OFFSET,
+ 16, 1, buf, size, true);
+
+ sema_init(&sem, 0);
+
+ urb->complete = write_callback;
+ urb->context = &sem;
+
+ result = usb_autopm_get_interface(dev->iface);
+ if (result < 0) {
+ DBG("unable to resume interface: %d\n", result);
+ if (result == -EPERM) {
+ qc_suspend(dev->iface, PMSG_SUSPEND);
+ }
+ return result;
+ }
+
+ spin_lock_irqsave(&dev->qmi.clients_lock, flags);
+
+ if (!client_addurb(dev, cid, urb)) {
+ usb_free_urb(urb);
+ spin_unlock_irqrestore(&dev->qmi.clients_lock, flags);
+ usb_autopm_put_interface(dev->iface);
+ return -EINVAL;
+ }
+
+ result = usb_submit_urb(urb, GFP_KERNEL);
+ if (result < 0) {
+ DBG("submit URB error %d\n", result);
+ if (client_delurb(dev, cid) != urb) {
+ DBG("Didn't get write URB back\n");
+ }
+
+ usb_free_urb(urb);
+
+ spin_unlock_irqrestore(&dev->qmi.clients_lock, flags);
+ usb_autopm_put_interface(dev->iface);
+ return result;
+ }
+
+ spin_unlock_irqrestore(&dev->qmi.clients_lock, flags);
+ result = down_interruptible(&sem);
+ if (!device_valid(dev)) {
+ DBG("Invalid device!\n");
+ return -ENXIO;
+ }
+
+ usb_autopm_put_interface(dev->iface);
+ spin_lock_irqsave(&dev->qmi.clients_lock, flags);
+ if (client_delurb(dev, cid) != urb) {
+ DBG("Didn't get write URB back\n");
+ spin_unlock_irqrestore(&dev->qmi.clients_lock, flags);
+ return -EINVAL;
+ }
+ spin_unlock_irqrestore(&dev->qmi.clients_lock, flags);
+
+ if (!result) {
+ if (!urb->status) {
+ result = size;
+ } else {
+ DBG("bad status = %d\n", urb->status);
+ result = urb->status;
+ }
+ } else {
+ DBG("Interrupted %d !!!\n", result);
+ DBG("Device may be in bad state and need reset !!!\n");
+ usb_kill_urb(urb);
+ }
+
+ usb_free_urb(urb);
+ return result;
+}
+
+static int client_alloc(struct qcusbnet *dev, u8 type)
+{
+ u16 cid;
+ struct client *client;
+ int result;
+ void *wbuf;
+ size_t wbufsize;
+ void *rbuf;
+ u16 rbufsize;
+ unsigned long flags;
+ u8 tid;
+
+ if (!device_valid(dev)) {
+ DBG("Invalid device!\n");
+ return -ENXIO;
+ }
+
+ if (type) {
+ tid = atomic_add_return(1, &dev->qmi.qmitid);
+ if (!tid)
+ atomic_add_return(1, &dev->qmi.qmitid);
+ wbuf = qmictl_new_getcid(tid, type, &wbufsize);
+ if (!wbuf)
+ return -ENOMEM;
+ result = write_sync(dev, wbuf, wbufsize, QMICTL);
+ kfree(wbuf);
+
+ if (result < 0)
+ return result;
+
+ result = read_sync(dev, &rbuf, QMICTL, tid);
+ if (result < 0) {
+ DBG("bad read data %d\n", result);
+ return result;
+ }
+ rbufsize = result;
+
+ result = qmictl_alloccid_resp(rbuf, rbufsize, &cid);
+ kfree(rbuf);
+
+ if (result < 0)
+ return result;
+ } else {
+ cid = 0;
+ }
+
+ spin_lock_irqsave(&dev->qmi.clients_lock, flags);
+ if (client_bycid(dev, cid)) {
+ DBG("Client memory already exists\n");
+ spin_unlock_irqrestore(&dev->qmi.clients_lock, flags);
+ return -ETOOMANYREFS;
+ }
+
+ client = kmalloc(sizeof(*client), GFP_ATOMIC);
+ if (!client) {
+ DBG("Error allocating read list\n");
+ spin_unlock_irqrestore(&dev->qmi.clients_lock, flags);
+ return -ENOMEM;
+ }
+
+ list_add_tail(&client->node, &dev->qmi.clients);
+ client->cid = cid;
+ INIT_LIST_HEAD(&client->reads);
+ INIT_LIST_HEAD(&client->notifies);
+ INIT_LIST_HEAD(&client->urbs);
+ spin_unlock_irqrestore(&dev->qmi.clients_lock, flags);
+ return cid;
+}
+
+static void client_free(struct qcusbnet *dev, u16 cid)
+{
+ struct list_head *node, *tmp;
+ int result;
+ struct client *client;
+ struct urb *urb;
+ void *data;
+ u16 size;
+ void *wbuf;
+ size_t wbufsize;
+ void *rbuf;
+ u16 rbufsize;
+ unsigned long flags;
+ u8 tid;
+
+ DBG("releasing 0x%04X\n", cid);
+
+ if (cid != QMICTL) {
+ tid = atomic_add_return(1, &dev->qmi.qmitid);
+ if (!tid)
+ tid = atomic_add_return(1, &dev->qmi.qmitid);
+ wbuf = qmictl_new_releasecid(tid, cid, &wbufsize);
+ if (!wbuf) {
+ DBG("memory error\n");
+ } else {
+ result = write_sync(dev, wbuf, wbufsize, QMICTL);
+ kfree(wbuf);
+
+ if (result < 0) {
+ DBG("bad write status %d\n", result);
+ } else {
+ result = read_sync(dev, &rbuf, QMICTL, tid);
+ if (result < 0) {
+ DBG("bad read status %d\n", result);
+ } else {
+ rbufsize = result;
+ result = qmictl_freecid_resp(rbuf, rbufsize);
+ kfree(rbuf);
+ if (result < 0)
+ DBG("error %d parsing response\n", result);
+ }
+ }
+ }
+ }
+
+ spin_lock_irqsave(&dev->qmi.clients_lock, flags);
+ list_for_each_safe(node, tmp, &dev->qmi.clients) {
+ client = list_entry(node, struct client, node);
+ if (client->cid == cid) {
+ while (client_notify(dev, cid, 0)) {
+ ;
+ }
+
+ urb = client_delurb(dev, cid);
+ while (urb != NULL) {
+ usb_kill_urb(urb);
+ usb_free_urb(urb);
+ urb = client_delurb(dev, cid);
+ }
+
+ while (client_delread(dev, cid, 0, &data, &size))
+ kfree(data);
+
+ list_del(&client->node);
+ kfree(client);
+ }
+ }
+ spin_unlock_irqrestore(&dev->qmi.clients_lock, flags);
+}
+
+struct client *client_bycid(struct qcusbnet *dev, u16 cid)
+{
+ struct list_head *node;
+ struct client *client;
+
+ if (!device_valid(dev)) {
+ DBG("Invalid device\n");
+ return NULL;
+ }
+
+ assert_locked(dev);
+
+ list_for_each(node, &dev->qmi.clients) {
+ client = list_entry(node, struct client, node);
+ if (client->cid == cid)
+ return client;
+ }
+
+ DBG("Could not find client mem 0x%04X\n", cid);
+ return NULL;
+}
+
+static bool client_addread(struct qcusbnet *dev, u16 cid, u16 tid, void *data,
+ u16 size)
+{
+ struct client *client;
+ struct readreq *req;
+
+ assert_locked(dev);
+
+ client = client_bycid(dev, cid);
+ if (!client) {
+ DBG("Could not find this client's memory 0x%04X\n", cid);
+ return false;
+ }
+
+ req = kmalloc(sizeof(*req), GFP_ATOMIC);
+ if (!req) {
+ DBG("Mem error\n");
+ return false;
+ }
+
+ req->data = data;
+ req->size = size;
+ req->tid = tid;
+
+ list_add_tail(&req->node, &client->reads);
+
+ return true;
+}
+
+static bool client_delread(struct qcusbnet *dev, u16 cid, u16 tid, void **data,
+ u16 *size)
+{
+ struct client *client;
+ struct readreq *req;
+ struct list_head *node;
+
+ assert_locked(dev);
+
+ client = client_bycid(dev, cid);
+ if (!client) {
+ DBG("Could not find this client's memory 0x%04X\n", cid);
+ return false;
+ }
+
+ list_for_each(node, &client->reads) {
+ req = list_entry(node, struct readreq, node);
+ if (!tid || tid == req->tid) {
+ *data = req->data;
+ *size = req->size;
+ list_del(&req->node);
+ kfree(req);
+ return true;
+ }
+
+ DBG("skipping 0x%04X data TID = %x\n", cid, req->tid);
+ }
+
+ DBG("No read memory to pop, Client 0x%04X, TID = %x\n", cid, tid);
+ return false;
+}
+
+static bool client_addnotify(struct qcusbnet *dev, u16 cid, u16 tid,
+ void (*hook)(struct qcusbnet *, u16, void *),
+ void *data)
+{
+ struct client *client;
+ struct notifyreq *req;
+
+ assert_locked(dev);
+
+ client = client_bycid(dev, cid);
+ if (!client) {
+ DBG("Could not find this client's memory 0x%04X\n", cid);
+ return false;
+ }
+
+ req = kmalloc(sizeof(*req), GFP_ATOMIC);
+ if (!req) {
+ DBG("Mem error\n");
+ return false;
+ }
+
+ list_add_tail(&req->node, &client->notifies);
+ req->func = hook;
+ req->data = data;
+ req->tid = tid;
+
+ return true;
+}
+
+static bool client_notify(struct qcusbnet *dev, u16 cid, u16 tid)
+{
+ struct client *client;
+ struct notifyreq *delnotify, *notify;
+ struct list_head *node;
+
+ assert_locked(dev);
+
+ client = client_bycid(dev, cid);
+ if (!client) {
+ DBG("Could not find this client's memory 0x%04X\n", cid);
+ return false;
+ }
+
+ delnotify = NULL;
+
+ list_for_each(node, &client->notifies) {
+ notify = list_entry(node, struct notifyreq, node);
+ if (!tid || !notify->tid || tid == notify->tid) {
+ delnotify = notify;
+ break;
+ }
+
+ DBG("skipping data TID = %x\n", notify->tid);
+ }
+
+ if (delnotify) {
+ list_del(&delnotify->node);
+ if (delnotify->func) {
+ spin_unlock(&dev->qmi.clients_lock);
+ delnotify->func(dev, cid, delnotify->data);
+ spin_lock(&dev->qmi.clients_lock);
+ }
+ kfree(delnotify);
+ return true;
+ }
+
+ DBG("no one to notify for TID %x\n", tid);
+ return false;
+}
+
+static bool client_addurb(struct qcusbnet *dev, u16 cid, struct urb *urb)
+{
+ struct client *client;
+ struct urbreq *req;
+
+ assert_locked(dev);
+
+ client = client_bycid(dev, cid);
+ if (!client) {
+ DBG("Could not find this client's memory 0x%04X\n", cid);
+ return false;
+ }
+
+ req = kmalloc(sizeof(*req), GFP_ATOMIC);
+ if (!req) {
+ DBG("Mem error\n");
+ return false;
+ }
+
+ req->urb = urb;
+ list_add_tail(&req->node, &client->urbs);
+
+ return true;
+}
+
+static struct urb *client_delurb(struct qcusbnet *dev, u16 cid)
+{
+ struct client *client;
+ struct urbreq *req;
+ struct urb *urb;
+
+ assert_locked(dev);
+
+ client = client_bycid(dev, cid);
+ if (!client) {
+ DBG("Could not find this client's memory 0x%04X\n", cid);
+ return NULL;
+ }
+
+ if (list_empty(&client->urbs)) {
+ DBG("No URB's to pop\n");
+ return NULL;
+ }
+
+ req = list_first_entry(&client->urbs, struct urbreq, node);
+ list_del(&req->node);
+ urb = req->urb;
+ kfree(req);
+ return urb;
+}
+
+static int devqmi_open(struct inode *inode, struct file *file)
+{
+ struct qmihandle *handle;
+ struct qmidev *qmidev = container_of(inode->i_cdev, struct qmidev, cdev);
+ struct qcusbnet *dev = container_of(qmidev, struct qcusbnet, qmi);
+
+ /* We need an extra ref on the device per fd, since we stash a ref
+ * inside the handle. If qcusbnet_get() returns NULL, that means the
+ * device has been removed from the list - no new refs for us. */
+ struct qcusbnet *ref = qcusbnet_get(dev);
+
+ if (!ref)
+ return -ENXIO;
+
+ file->private_data = kmalloc(sizeof(struct qmihandle), GFP_KERNEL);
+ if (!file->private_data) {
+ DBG("Mem error\n");
+ return -ENOMEM;
+ }
+
+ handle = (struct qmihandle *)file->private_data;
+ handle->cid = (u16)-1;
+ handle->dev = ref;
+
+ DBG("%p %04x", handle, handle->cid);
+
+ return 0;
+}
+
+static int devqmi_ioctl(struct inode *inode, struct file *file, unsigned int cmd, unsigned long arg)
+{
+ int result;
+ u32 vidpid;
+
+ struct qmihandle *handle = (struct qmihandle *)file->private_data;
+
+ DBG("%p %04x %08x", handle, handle->cid, cmd);
+
+ if (!handle) {
+ DBG("Bad file data\n");
+ return -EBADF;
+ }
+
+ if (handle->dev->dying) {
+ DBG("Dying device");
+ return -ENXIO;
+ }
+
+ if (!device_valid(handle->dev)) {
+ DBG("Invalid device!\n");
+ return -ENXIO;
+ }
+
+ switch (cmd) {
+ case IOCTL_QMI_GET_SERVICE_FILE:
+
+ DBG("Setting up QMI for service %lu\n", arg);
+ if (!(u8)arg) {
+ DBG("Cannot use QMICTL from userspace\n");
+ return -EINVAL;
+ }
+
+ if (handle->cid != (u16)-1) {
+ DBG("Close the current connection before opening a new one\n");
+ return -EBADR;
+ }
+
+ result = client_alloc(handle->dev, (u8)arg);
+ if (result < 0)
+ return result;
+ handle->cid = result;
+
+ return 0;
+ break;
+
+ /* Okay, all aboard the nasty hack express. If we don't have this
+ * ioctl() (and we just rely on userspace to close() the file
+ * descriptors), if userspace has any refs left to this fd (like, say, a
+ * pending read()), then the read might hang around forever. Userspace
+ * needs a way to cause us to kick people off those waitqueues before
+ * closing the fd for good.
+ *
+ * If this driver used workqueues, the correct approach here would
+ * instead be to make the file descriptor select()able, and then just
+ * use select() instead of aio in userspace (thus allowing us to get
+ * away with one thread total and avoiding the recounting mess
+ * altogether).
+ */
+ case IOCTL_QMI_CLOSE:
+ DBG("Tearing down QMI for service %lu", arg);
+ if (handle->cid == (u16)-1) {
+ DBG("no qmi cid");
+ return -EBADR;
+ }
+
+ file->private_data = NULL;
+ client_free(handle->dev, handle->cid);
+ kfree(handle);
+ return 0;
+ break;
+
+ case IOCTL_QMI_GET_DEVICE_VIDPID:
+ if (!arg) {
+ DBG("Bad VIDPID buffer\n");
+ return -EINVAL;
+ }
+
+ if (!handle->dev->usbnet) {
+ DBG("Bad usbnet\n");
+ return -ENOMEM;
+ }
+
+ if (!handle->dev->usbnet->udev) {
+ DBG("Bad udev\n");
+ return -ENOMEM;
+ }
+
+ vidpid = ((le16_to_cpu(handle->dev->usbnet->udev->descriptor.idVendor) << 16)
+ + le16_to_cpu(handle->dev->usbnet->udev->descriptor.idProduct));
+
+ result = copy_to_user((unsigned int *)arg, &vidpid, 4);
+ if (result)
+ DBG("Copy to userspace failure\n");
+
+ return result;
+ break;
+
+ case IOCTL_QMI_GET_DEVICE_MEID:
+ if (!arg) {
+ DBG("Bad MEID buffer\n");
+ return -EINVAL;
+ }
+
+ result = copy_to_user((unsigned int *)arg, &handle->dev->meid[0], 14);
+ if (result)
+ DBG("copy to userspace failure\n");
+
+ return result;
+ break;
+ default:
+ return -EBADRQC;
+ }
+}
+
+static int devqmi_release(struct inode *inode, struct file *file)
+{
+ struct qmihandle *handle = (struct qmihandle *)file->private_data;
+ if (!handle)
+ return 0;
+ file->private_data = NULL;
+ if (handle->cid != (u16)-1)
+ client_free(handle->dev, handle->cid);
+ qcusbnet_put(handle->dev);
+ kfree(handle);
+ return 0;
+}
+
+static ssize_t devqmi_read(struct file *file, char __user *buf, size_t size,
+ loff_t *pos)
+{
+ int result;
+ void *data = NULL;
+ void *smalldata;
+ struct qmihandle *handle = (struct qmihandle *)file->private_data;
+
+ if (!handle) {
+ DBG("Bad file data\n");
+ return -EBADF;
+ }
+
+ if (handle->dev->dying) {
+ DBG("Dying device");
+ return -ENXIO;
+ }
+
+ if (!device_valid(handle->dev)) {
+ DBG("Invalid device!\n");
+ return -ENXIO;
+ }
+
+ if (handle->cid == (u16)-1) {
+ DBG("Client ID must be set before reading 0x%04X\n",
+ handle->cid);
+ return -EBADR;
+ }
+
+ result = read_sync(handle->dev, &data, handle->cid, 0);
+ if (result <= 0)
+ return result;
+
+ result -= qmux_size;
+ smalldata = data + qmux_size;
+
+ if (result > size) {
+ DBG("Read data is too large for amount user has requested\n");
+ kfree(data);
+ return -EOVERFLOW;
+ }
+
+ if (copy_to_user(buf, smalldata, result)) {
+ DBG("Error copying read data to user\n");
+ result = -EFAULT;
+ }
+
+ kfree(data);
+ return result;
+}
+
+static ssize_t devqmi_write(struct file *file, const char __user * buf,
+ size_t size, loff_t *pos)
+{
+ int status;
+ void *wbuf;
+ struct qmihandle *handle = (struct qmihandle *)file->private_data;
+
+ if (!handle) {
+ DBG("Bad file data\n");
+ return -EBADF;
+ }
+
+ if (!device_valid(handle->dev)) {
+ DBG("Invalid device! Updating f_ops\n");
+ file->f_op = file->f_dentry->d_inode->i_fop;
+ return -ENXIO;
+ }
+
+ if (handle->cid == (u16)-1) {
+ DBG("Client ID must be set before writing 0x%04X\n",
+ handle->cid);
+ return -EBADR;
+ }
+
+ wbuf = kmalloc(size + qmux_size, GFP_KERNEL);
+ if (!wbuf)
+ return -ENOMEM;
+ status = copy_from_user(wbuf + qmux_size, buf, size);
+ if (status) {
+ DBG("Unable to copy data from userspace %d\n", status);
+ kfree(wbuf);
+ return status;
+ }
+
+ status = write_sync(handle->dev, wbuf, size + qmux_size,
+ handle->cid);
+
+ kfree(wbuf);
+ if (status == size + qmux_size)
+ return size;
+ return status;
+}
+
+int qc_register(struct qcusbnet *dev)
+{
+ int result;
+ int qmiidx = 0;
+ dev_t devno;
+ char *name;
+
+ dev->valid = true;
+ dev->dying = false;
+ result = client_alloc(dev, QMICTL);
+ if (result) {
+ dev->valid = false;
+ return result;
+ }
+ atomic_set(&dev->qmi.qmitid, 1);
+
+ result = qc_startread(dev);
+ if (result) {
+ dev->valid = false;
+ return result;
+ }
+
+ if (!qmi_ready(dev, 30000)) {
+ DBG("Device unresponsive to QMI\n");
+ return -ETIMEDOUT;
+ }
+
+ result = setup_wds_callback(dev);
+ if (result) {
+ dev->valid = false;
+ return result;
+ }
+
+ result = qmidms_getmeid(dev);
+ if (result) {
+ dev->valid = false;
+ return result;
+ }
+
+ result = alloc_chrdev_region(&devno, 0, 1, "qcqmi");
+ if (result < 0)
+ return result;
+
+ cdev_init(&dev->qmi.cdev, &devqmi_fops);
+ dev->qmi.cdev.owner = THIS_MODULE;
+ dev->qmi.cdev.ops = &devqmi_fops;
+
+ result = cdev_add(&dev->qmi.cdev, devno, 1);
+ if (result) {
+ DBG("error adding cdev\n");
+ return result;
+ }
+
+ name = strstr(dev->usbnet->net->name, "usb");
+ if (!name) {
+ DBG("Bad net name: %s\n", dev->usbnet->net->name);
+ return -ENXIO;
+ }
+ name += strlen("usb");
+ qmiidx = simple_strtoul(name, NULL, 10);
+ if (qmiidx < 0) {
+ DBG("Bad minor number\n");
+ return -ENXIO;
+ }
+
+ printk(KERN_INFO "creating qcqmi%d\n", qmiidx);
+ device_create(dev->qmi.devclass, &dev->iface->dev, devno, NULL, "qcqmi%d", qmiidx);
+
+ dev->qmi.devnum = devno;
+ return 0;
+}
+
+void qc_deregister(struct qcusbnet *dev)
+{
+ struct list_head *node, *tmp;
+ struct client *client;
+
+ dev->dying = true;
+ list_for_each_safe(node, tmp, &dev->qmi.clients) {
+ client = list_entry(node, struct client, node);
+ DBG("release 0x%04X\n", client->cid);
+ client_free(dev, client->cid);
+ }
+
+ qc_stopread(dev);
+ dev->valid = false;
+ if (!IS_ERR(dev->qmi.devclass))
+ device_destroy(dev->qmi.devclass, dev->qmi.devnum);
+ cdev_del(&dev->qmi.cdev);
+ unregister_chrdev_region(dev->qmi.devnum, 1);
+}
+
+static bool qmi_ready(struct qcusbnet *dev, u16 timeout)
+{
+ int result;
+ void *wbuf;
+ size_t wbufsize;
+ void *rbuf;
+ u16 rbufsize;
+ struct semaphore sem;
+ u16 now;
+ unsigned long flags;
+ u8 tid;
+
+ if (!device_valid(dev)) {
+ DBG("Invalid device\n");
+ return -EFAULT;
+ }
+
+
+ for (now = 0; now < timeout; now += 100) {
+ sema_init(&sem, 0);
+
+ tid = atomic_add_return(1, &dev->qmi.qmitid);
+ if (!tid)
+ tid = atomic_add_return(1, &dev->qmi.qmitid);
+ kfree(wbuf);
+ wbuf = qmictl_new_ready(tid, &wbufsize);
+ if (!wbuf)
+ return -ENOMEM;
+
+ result = read_async(dev, QMICTL, tid, upsem, &sem);
+ if (result) {
+ kfree(wbuf);
+ return false;
+ }
+
+ write_sync(dev, wbuf, wbufsize, QMICTL);
+
+ msleep(100);
+ if (!down_trylock(&sem)) {
+ spin_lock_irqsave(&dev->qmi.clients_lock, flags);
+ if (client_delread(dev, QMICTL, tid, &rbuf, &rbufsize)) {
+ spin_unlock_irqrestore(&dev->qmi.clients_lock, flags);
+ kfree(rbuf);
+ break;
+ } else {
+ spin_unlock_irqrestore(&dev->qmi.clients_lock, flags);
+ }
+ } else {
+ spin_lock_irqsave(&dev->qmi.clients_lock, flags);
+ client_notify(dev, QMICTL, tid);
+ spin_unlock_irqrestore(&dev->qmi.clients_lock, flags);
+ }
+ }
+
+ kfree(wbuf);
+
+ if (now >= timeout)
+ return false;
+
+ DBG("QMI Ready after %u milliseconds\n", now);
+
+ /* 3580 and newer doesn't need a delay; older needs 5000ms */
+ if (qcusbnet2k_fwdelay)
+ msleep(qcusbnet2k_fwdelay * 1000);
+
+ return true;
+}
+
+static void wds_callback(struct qcusbnet *dev, u16 cid, void *data)
+{
+ bool ret;
+ int result;
+ void *rbuf;
+ u16 rbufsize;
+
+ struct net_device_stats *stats = &(dev->usbnet->net->stats);
+
+ struct qmiwds_stats dstats = {
+ .txok = (u32)-1,
+ .rxok = (u32)-1,
+ .txerr = (u32)-1,
+ .rxerr = (u32)-1,
+ .txofl = (u32)-1,
+ .rxofl = (u32)-1,
+ .txbytesok = (u64)-1,
+ .rxbytesok = (u64)-1,
+ };
+ unsigned long flags;
+
+ if (!device_valid(dev)) {
+ DBG("Invalid device\n");
+ return;
+ }
+
+ spin_lock_irqsave(&dev->qmi.clients_lock, flags);
+ ret = client_delread(dev, cid, 0, &rbuf, &rbufsize);
+ spin_unlock_irqrestore(&dev->qmi.clients_lock, flags);
+
+ if (!ret) {
+ DBG("WDS callback failed to get data\n");
+ return;
+ }
+
+ dstats.linkstate = !qc_isdown(dev, DOWN_NO_NDIS_CONNECTION);
+ dstats.reconfigure = false;
+
+ result = qmiwds_event_resp(rbuf, rbufsize, &dstats);
+ if (result < 0) {
+ DBG("bad WDS packet\n");
+ } else {
+ if (dstats.txofl != (u32)-1)
+ stats->tx_fifo_errors = dstats.txofl;
+
+ if (dstats.rxofl != (u32)-1)
+ stats->rx_fifo_errors = dstats.rxofl;
+
+ if (dstats.txerr != (u32)-1)
+ stats->tx_errors = dstats.txerr;
+
+ if (dstats.rxerr != (u32)-1)
+ stats->rx_errors = dstats.rxerr;
+
+ if (dstats.txok != (u32)-1)
+ stats->tx_packets = dstats.txok + stats->tx_errors;
+
+ if (dstats.rxok != (u32)-1)
+ stats->rx_packets = dstats.rxok + stats->rx_errors;
+
+ if (dstats.txbytesok != (u64)-1)
+ stats->tx_bytes = dstats.txbytesok;
+
+ if (dstats.rxbytesok != (u64)-1)
+ stats->rx_bytes = dstats.rxbytesok;
+
+ if (dstats.reconfigure) {
+ DBG("Net device link reset\n");
+ qc_setdown(dev, DOWN_NO_NDIS_CONNECTION);
+ qc_cleardown(dev, DOWN_NO_NDIS_CONNECTION);
+ } else {
+ if (dstats.linkstate) {
+ DBG("Net device link is connected\n");
+ qc_cleardown(dev, DOWN_NO_NDIS_CONNECTION);
+ } else {
+ DBG("Net device link is disconnected\n");
+ qc_setdown(dev, DOWN_NO_NDIS_CONNECTION);
+ }
+ }
+ }
+
+ kfree(rbuf);
+
+ result = read_async(dev, cid, 0, wds_callback, data);
+ if (result != 0)
+ DBG("unable to setup next async read\n");
+}
+
+static int setup_wds_callback(struct qcusbnet *dev)
+{
+ int result;
+ void *buf;
+ size_t size;
+ u16 cid;
+
+ if (!device_valid(dev)) {
+ DBG("Invalid device\n");
+ return -EFAULT;
+ }
+
+ result = client_alloc(dev, QMIWDS);
+ if (result < 0)
+ return result;
+ cid = result;
+
+ buf = qmiwds_new_seteventreport(1, &size);
+ if (!buf)
+ return -ENOMEM;
+
+ result = write_sync(dev, buf, size, cid);
+ kfree(buf);
+
+ if (result < 0) {
+ return result;
+ }
+
+ buf = qmiwds_new_getpkgsrvcstatus(2, &size);
+ if (buf == NULL)
+ return -ENOMEM;
+
+ result = write_sync(dev, buf, size, cid);
+ kfree(buf);
+
+ if (result < 0)
+ return result;
+
+ result = read_async(dev, cid, 0, wds_callback, NULL);
+ if (result) {
+ DBG("unable to setup async read\n");
+ return result;
+ }
+
+ result = usb_control_msg(dev->usbnet->udev,
+ usb_sndctrlpipe(dev->usbnet->udev, 0),
+ 0x22, 0x21, 1, 0, NULL, 0, 100);
+ if (result < 0) {
+ DBG("Bad SetControlLineState status %d\n", result);
+ return result;
+ }
+
+ return 0;
+}
+
+static int qmidms_getmeid(struct qcusbnet *dev)
+{
+ int result;
+ void *wbuf;
+ size_t wbufsize;
+ void *rbuf;
+ u16 rbufsize;
+ u16 cid;
+
+ if (!device_valid(dev)) {
+ DBG("Invalid device\n");
+ return -EFAULT;
+ }
+
+ result = client_alloc(dev, QMIDMS);
+ if (result < 0)
+ return result;
+ cid = result;
+
+ wbuf = qmidms_new_getmeid(1, &wbufsize);
+ if (!wbuf)
+ return -ENOMEM;
+
+ result = write_sync(dev, wbuf, wbufsize, cid);
+ kfree(wbuf);
+
+ if (result < 0)
+ return result;
+
+ result = read_sync(dev, &rbuf, cid, 1);
+ if (result < 0)
+ return result;
+ rbufsize = result;
+
+ result = qmidms_meid_resp(rbuf, rbufsize, &dev->meid[0], 14);
+ kfree(rbuf);
+
+ if (result < 0) {
+ DBG("bad get MEID resp\n");
+ memset(&dev->meid[0], '0', 14);
+ }
+
+ client_free(dev, cid);
+ return 0;
+}
+
+module_param(qcusbnet2k_fwdelay, int, S_IRUGO | S_IWUSR);
+MODULE_PARM_DESC(qcusbnet2k_fwdelay, "Delay for old firmware");
diff --git a/drivers/net/usb/gobi/qmidevice.h b/drivers/net/usb/gobi/qmidevice.h
new file mode 100644
index 0000000..5274a0d
--- /dev/null
+++ b/drivers/net/usb/gobi/qmidevice.h
@@ -0,0 +1,35 @@
+/* qmidevice.h - gobi QMI device header
+ * Copyright (c) 2010, Code Aurora Forum. All rights reserved.
+
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 and
+ * only version 2 as published by the Free Software Foundation.
+
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
+ * 02110-1301, USA.
+ */
+
+#ifndef QCUSBNET_QMIDEVICE_H
+#define QCUSBNET_QMIDEVICE_H
+
+#include "structs.h"
+#include "qmi.h"
+
+void qc_setdown(struct qcusbnet *dev, u8 reason);
+void qc_cleardown(struct qcusbnet *dev, u8 reason);
+bool qc_isdown(struct qcusbnet *dev, u8 reason);
+
+int qc_startread(struct qcusbnet *dev);
+void qc_stopread(struct qcusbnet *dev);
+
+int qc_register(struct qcusbnet *dev);
+void qc_deregister(struct qcusbnet *dev);
+
+#endif /* !QCUSBNET_QMIDEVICE_H */
diff --git a/drivers/net/usb/gobi/structs.h b/drivers/net/usb/gobi/structs.h
new file mode 100644
index 0000000..13b3788
--- /dev/null
+++ b/drivers/net/usb/gobi/structs.h
@@ -0,0 +1,96 @@
+/* structs.h - shared structures
+ * Copyright (c) 2010, Code Aurora Forum. All rights reserved.
+
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 and
+ * only version 2 as published by the Free Software Foundation.
+
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
+ * 02110-1301, USA.
+ */
+
+#ifndef QCUSBNET_STRUCTS_H
+#define QCUSBNET_STRUCTS_H
+
+#include <linux/etherdevice.h>
+#include <linux/ethtool.h>
+#include <linux/mii.h>
+#include <linux/usb.h>
+#include <linux/version.h>
+#include <linux/cdev.h>
+#include <linux/kobject.h>
+#include <linux/kthread.h>
+
+#include <linux/usb/usbnet.h>
+
+#include <linux/fdtable.h>
+
+#define DBG(fmt, arg...) \
+do { \
+ if (qcusbnet_debug == 1) \
+ printk(KERN_INFO "QCUSBNet2k::%s " fmt, __func__, ##arg); \
+} while (0)
+
+struct qcusbnet;
+
+struct urbreq {
+ struct list_head node;
+ struct urb *urb;
+};
+
+#define DEFAULT_READ_URB_LENGTH 0x1000
+
+struct worker {
+ struct task_struct *thread;
+ struct completion work;
+ struct list_head urbs;
+ spinlock_t urbs_lock;
+ struct urb *active;
+ spinlock_t active_lock;
+ struct usb_interface *iface;
+};
+
+struct qmidev {
+ dev_t devnum;
+ struct class *devclass;
+ struct cdev cdev;
+ struct urb *readurb;
+ struct urbsetup *readsetup;
+ void *readbuf;
+ struct urb *inturb;
+ void *intbuf;
+ struct list_head clients;
+ spinlock_t clients_lock;
+ atomic_t qmitid;
+};
+
+enum {
+ DOWN_NO_NDIS_CONNECTION = 0,
+ DOWN_CDC_CONNECTION_SPEED = 1,
+ DOWN_DRIVER_SUSPENDED = 2,
+ DOWN_NET_IFACE_STOPPED = 3,
+};
+
+struct qcusbnet {
+ struct list_head node;
+ struct kref refcount;
+ struct usbnet *usbnet;
+ struct usb_interface *iface;
+ int (*open)(struct net_device *);
+ int (*stop)(struct net_device *);
+ unsigned long down;
+ bool valid;
+ bool dying;
+ struct qmidev qmi;
+ char meid[14];
+ struct worker worker;
+};
+
+#endif /* !QCUSBNET_STRUCTS_H */
--
1.7.3.1
^ permalink raw reply related
* Re: [Bugme-new] [Bug 32832] New: shutdown(2) does not fully shut down socket any more
From: David Miller @ 2011-04-13 19:09 UTC (permalink / raw)
To: eric.dumazet; +Cc: akpm, netdev, bugzilla-daemon, bugme-daemon, kees
In-Reply-To: <1302663608.2811.62.camel@edumazet-laptop>
From: Eric Dumazet <eric.dumazet@gmail.com>
Date: Wed, 13 Apr 2011 05:00:08 +0200
> Le mercredi 13 avril 2011 à 04:55 +0200, Eric Dumazet a écrit :
>
>> Since SO_REUSEPORT is not a 'stable fix', I suggest we revert the patch,
>> and eventually work on SO_REUSEPORT on net-next-2.6
>>
>> What do you think ?
>>
>
> Sorry, I should have mentioned commit id : c191a836a908d1dd6
> (tcp: disallow bind() to reuse addr/port)
I'm commiting the revert as follows to net-2.6, and will queue
it up for -stable as well:
--------------------
Revert "tcp: disallow bind() to reuse addr/port"
This reverts commit c191a836a908d1dd6b40c503741f91b914de3348.
It causes known regressions for programs that expect to be able to use
SO_REUSEADDR to shutdown a socket, then successfully rebind another
socket to the same ID.
Programs such as haproxy and amavisd expect this to work.
This should fix kernel bugzilla 32832.
Signed-off-by: David S. Miller <davem@davemloft.net>
---
net/ipv4/inet_connection_sock.c | 5 ++---
net/ipv6/inet6_connection_sock.c | 2 +-
2 files changed, 3 insertions(+), 4 deletions(-)
diff --git a/net/ipv4/inet_connection_sock.c b/net/ipv4/inet_connection_sock.c
index 6c0b7f4..38f23e7 100644
--- a/net/ipv4/inet_connection_sock.c
+++ b/net/ipv4/inet_connection_sock.c
@@ -73,7 +73,7 @@ int inet_csk_bind_conflict(const struct sock *sk,
!sk2->sk_bound_dev_if ||
sk->sk_bound_dev_if == sk2->sk_bound_dev_if)) {
if (!reuse || !sk2->sk_reuse ||
- ((1 << sk2->sk_state) & (TCPF_LISTEN | TCPF_CLOSE))) {
+ sk2->sk_state == TCP_LISTEN) {
const __be32 sk2_rcv_saddr = sk_rcv_saddr(sk2);
if (!sk2_rcv_saddr || !sk_rcv_saddr(sk) ||
sk2_rcv_saddr == sk_rcv_saddr(sk))
@@ -122,8 +122,7 @@ again:
(tb->num_owners < smallest_size || smallest_size == -1)) {
smallest_size = tb->num_owners;
smallest_rover = rover;
- if (atomic_read(&hashinfo->bsockets) > (high - low) + 1 &&
- !inet_csk(sk)->icsk_af_ops->bind_conflict(sk, tb)) {
+ if (atomic_read(&hashinfo->bsockets) > (high - low) + 1) {
spin_unlock(&head->lock);
snum = smallest_rover;
goto have_snum;
diff --git a/net/ipv6/inet6_connection_sock.c b/net/ipv6/inet6_connection_sock.c
index 1660546..f2c5b0f 100644
--- a/net/ipv6/inet6_connection_sock.c
+++ b/net/ipv6/inet6_connection_sock.c
@@ -44,7 +44,7 @@ int inet6_csk_bind_conflict(const struct sock *sk,
!sk2->sk_bound_dev_if ||
sk->sk_bound_dev_if == sk2->sk_bound_dev_if) &&
(!sk->sk_reuse || !sk2->sk_reuse ||
- ((1 << sk2->sk_state) & (TCPF_LISTEN | TCPF_CLOSE))) &&
+ sk2->sk_state == TCP_LISTEN) &&
ipv6_rcv_saddr_equal(sk, sk2))
break;
}
--
1.7.4.3
^ 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