* Re: [PATCH net] failover: eliminate callback hell
From: Michael S. Tsirkin @ 2018-06-05 19:38 UTC (permalink / raw)
To: Stephen Hemminger
Cc: kys, haiyangz, davem, sridhar.samudrala, netdev,
Stephen Hemminger
In-Reply-To: <20180605115305.502a7ebb@xeon-e3>
On Tue, Jun 05, 2018 at 11:53:05AM -0700, Stephen Hemminger wrote:
> > > * Now, netvsc and net_failover use the same delayed work type
> > > mechanism for setup. Previously, net_failover code was triggering off
> > > name change but a similar policy was rejected for netvsc.
> > > "what is good for the goose is good for the gander"
> >
> > I don't really understand what you are saying here. I think the delayed
> > hack is kind of ugly and seems racy. Current failover code was rejected
> > by whom? Why is new one good and for whom? Did you want to do a name
> > change in netvsc but it was rejected? Could you clarify please?
>
> See:
> https://patchwork.ozlabs.org/patch/851711/
Let me try to summarize that:
You wanted to speed up the delayed link up. You had an idea to
additionally take link up when userspace renames the interface (standby
one which is also the failover for netvsc).
But userspace might not do any renames, in which case there will
still be the delay, and so this never got applied.
Is this a good summary?
Davem said delay should go away completely as it's not robust, and I
think I agree. So I don't think we should make all failover users use
delay. IIUC failover kept a delay option especially for netvsc to
minimize the surprise factor. Hopefully we can come up with
something more robust and drop that option completely.
> > > * Set permanent and current address of net_failover device
> > > to match the primary.
> > >
> > > * Carrier should be marked off before registering device
> > > the net_failover device.
> >
> > Are above two bugfixes?
>
> Yes.
Maybe fix these two as first patches in the set?
> > > Although this patch needs to go into 4.18 (linux-net),
> >
> > I'd rather we focused on fixing bugs in 4.18, and left refactoring to
> > 4.19.
> >
>
> Either we fix or revert the current code in 4.18.
> Sorry, I am not having callback hell code in any vendor or upstream kernel.
I agree callbacks add complexity which often isn't necessary, so
removing them where possible is a good cleanup. But maybe a patch
shouldn't mix bugfixes, cleanups and behaviour changes all together. If
nothing else it makes review harder. Splitting patches up might make it
more likely they can go into 4.18 which seems to be what you want.
HTH,
--
MST
^ permalink raw reply
* Re: [PATCH] r8169: Reinstate ALDPS and ASPM support
From: Florian Fainelli @ 2018-06-05 19:27 UTC (permalink / raw)
To: Heiner Kallweit, Bjorn Helgaas, Ryankao
Cc: Kai Heng Feng, jrg.otte@gmail.com, David Miller, Hayes Wang,
romieu@fr.zoreil.com, Linux Netdev List,
Linux Kernel Mailing List, Hau, linux-pci
In-Reply-To: <6c6b579a-5be4-ccc0-f64c-9998e51145f5@gmail.com>
On 06/05/2018 12:17 PM, Heiner Kallweit wrote:
> On 05.06.2018 21:11, Bjorn Helgaas wrote:
>> [+cc linux-pci]
>>
>> On Tue, Jun 05, 2018 at 12:28:05PM -0500, Bjorn Helgaas wrote:
>>> On Tue, Jun 05, 2018 at 06:34:09AM +0000, Ryankao wrote:
>>>> Add realtek folk Hau
>>>>
>>>> -----Original Message-----
>>>> From: Kai Heng Feng [mailto:kai.heng.feng@canonical.com]
>>>> Sent: Tuesday, June 05, 2018 1:02 PM
>>>> To: jrg.otte@gmail.com
>>>> Cc: David Miller <davem@davemloft.net>; Hayes Wang <hayeswang@realtek.com>; hkallweit1@gmail.com; romieu@fr.zoreil.com; Linux Netdev List <netdev@vger.kernel.org>; Linux Kernel Mailing List <linux-kernel@vger.kernel.org>; Ryankao <ryankao@realtek.com>
>>>> Subject: Re: [PATCH] r8169: Reinstate ALDPS and ASPM support
>>>>
>>>> Hi Jörg Otte,
>>>>
>>>> Can you give this patch a try?
>>>>
>>>> Since you are the only one that reported ALDPS/ASPM regression,
>>>>
>>>> And I think this patch should solve the issue you had [1].
>>>>
>>>> Hopefully we don't need to go down the rabbit hole of blacklist/whitelist...
>>>>
>>>> Kai-Heng
>>>>
>>>> [1] https://lkml.org/lkml/2013/1/5/36
>>>
>>> I have no idea what ALDPS is. It's not mentioned in the PCIe spec, so
>>> presumably it's some Realtek-specific thing. ASPM is a generic PCIe
>>> thing. Changes to these two things should be in separate patches so
>>> they don't get tangled up.
>>>
> ALDPS = Advanced Link Down Power Saving
> And yes, it's a Realtek feature.
Link as in Ethernet link or PCI(e) link? Sorry too lazy to let me google
that for myself :)
--
Florian
^ permalink raw reply
* Re: [PATCH] r8169: Reinstate ALDPS and ASPM support
From: Heiner Kallweit @ 2018-06-05 19:24 UTC (permalink / raw)
To: Bjorn Helgaas, Ryankao
Cc: Kai Heng Feng, jrg.otte@gmail.com, David Miller, Hayes Wang,
romieu@fr.zoreil.com, Linux Netdev List,
Linux Kernel Mailing List, Hau
In-Reply-To: <20180605172805.GD30381@bhelgaas-glaptop.roam.corp.google.com>
On 05.06.2018 19:28, Bjorn Helgaas wrote:
> On Tue, Jun 05, 2018 at 06:34:09AM +0000, Ryankao wrote:
>> Add realtek folk Hau
>>
>> -----Original Message-----
>> From: Kai Heng Feng [mailto:kai.heng.feng@canonical.com]
>> Sent: Tuesday, June 05, 2018 1:02 PM
>> To: jrg.otte@gmail.com
>> Cc: David Miller <davem@davemloft.net>; Hayes Wang <hayeswang@realtek.com>; hkallweit1@gmail.com; romieu@fr.zoreil.com; Linux Netdev List <netdev@vger.kernel.org>; Linux Kernel Mailing List <linux-kernel@vger.kernel.org>; Ryankao <ryankao@realtek.com>
>> Subject: Re: [PATCH] r8169: Reinstate ALDPS and ASPM support
>>
>> Hi Jörg Otte,
>>
>> Can you give this patch a try?
>>
>> Since you are the only one that reported ALDPS/ASPM regression,
>>
>> And I think this patch should solve the issue you had [1].
>>
>> Hopefully we don't need to go down the rabbit hole of blacklist/whitelist...
>>
>> Kai-Heng
>>
>> [1] https://lkml.org/lkml/2013/1/5/36
>
> I have no idea what ALDPS is. It's not mentioned in the PCIe spec, so
> presumably it's some Realtek-specific thing. ASPM is a generic PCIe
> thing. Changes to these two things should be in separate patches so
> they don't get tangled up.
>
>>> On Jun 5, 2018, at 12:58 PM, Kai-Heng Feng
>>> <kai.heng.feng@canonical.com>
>>> wrote:
>>>
>>> This patch reinstate ALDPS and ASPM support on r8169.
>>>
>>> On some Intel platforms, ASPM support on r8169 is the key factor to
>>> let Package C-State achieve PC8. Without ASPM support, the deepest
>>> Package C-State can hit is PC3. PC8 can save additional ~3W in
>>> comparison with PC3.
>>>
>>> This patch is from Realtek.
>>>
>>> Fixes: e0c075577965 ("r8169: enable ALDPS for power saving")
>>> Fixes: d64ec841517a ("r8169: enable internal ASPM and clock request
>>> settings")
>
>>> +3507,15 @@ static void rtl8168e_1_hw_phy_config(struct
>>> rtl8169_private *tp)
>>> rtl_writephy(tp, 0x0d, 0x4007);
>>> rtl_writephy(tp, 0x0e, 0x0000);
>>> rtl_writephy(tp, 0x0d, 0x0000);
>>> +
>>> + /* Check ALDPS bit, disable it if enabled */
>>> + rtl_writephy(tp, 0x1f, 0x0000);
>>> + if (enable_aldps)
>>> + rtl_w0w1_phy(tp, 0x15, 0x1000, 0x0000);
>>> + else if (rtl_readphy(tp, 0x15) & 0x1000)
>>> + rtl_w0w1_phy(tp, 0x15, 0x0000, 0x1000);
>
> There's a lot of repetition of this code with minor variations. You
> could probably factor it out and make it more concise and more
> readable.
>
>>> +static void rtl8169_check_link_status(struct net_device *dev,
>>> + struct rtl8169_private *tp) {
>>> + struct device *d = tp_to_dev(tp);
>>> +
>>> + if (tp->link_ok(tp)) {
>>> + rtl_link_chg_patch(tp);
>>> + /* This is to cancel a scheduled suspend if there's one. */
>>> + if (pm_request_resume(d))
>>> + _rtl_reset_work(tp);
>>> + netif_carrier_on(dev);
>>> + if (net_ratelimit())
>>> + netif_info(tp, ifup, dev, "link up\n");
>>> + } else {
>>> + netif_carrier_off(dev);
>>> + netif_info(tp, ifdown, dev, "link down\n");
>>> + pm_runtime_idle(d);
>>> + }
>>> +}
>
> This function apparently just got moved around without changing
> anything. That's fine, but the move should be in a separate patch to
> make the real changes easier to review.
>
>>> @@ -7649,8 +7757,12 @@ static int rtl_init_one(struct pci_dev *pdev,
>>> const struct pci_device_id *ent)
>>>
>>> /* disable ASPM completely as that cause random device stop working
>>> * problems as well as full system hangs for some PCIe devices users */
>>> - pci_disable_link_state(pdev, PCIE_LINK_STATE_L0S | PCIE_LINK_STATE_L1 |
>>> - PCIE_LINK_STATE_CLKPM);
>>> + if (!enable_aspm) {
>>> + pci_disable_link_state(pdev, PCIE_LINK_STATE_L0S |
>>> + PCIE_LINK_STATE_L1 |
>>> + PCIE_LINK_STATE_CLKPM);
>>> + netif_info(tp, probe, dev, "ASPM disabled\n");
>>> + }
>
> ASPM is a generic PCIe feature that should be configured by the PCI
> core without any help from the device driver.
>
> If code in the driver is needed, that means either the PCI core is
> doing it wrong and we should fix it there, or the device is broken and
> the driver is working around the erratum.
>
> If this is an erratum, you should include details about exactly what's
> broken and (ideally) a URL to the published erratum. Otherwise this
> is just unmaintainable black magic and likely to be broken by future
> ASPM changes in the PCI core.
>
Fully agree, but: There are no publicly available datasheets and only
source for such magic is the r8168 vendor driver and trial&error.
In addition the driver supports ~ 50 chip variants and not all variants
(unfortunately nobody except Realtek knows which) are affected by the
problem. Maybe the involved Realtek guys can shed some light on this.
> ASPM configuration is done by the PCI core before drivers are bound to
> the device. If you need device-specific workarounds, they should
> probably be in quirks so they're done before the core does that ASPM
> configuration.
>
>>> /* enable device (incl. PCI PM wakeup and hotplug setup) */
>>> rc = pcim_enable_device(pdev);
>>> --
>>> 2.17.0
>>
>> ------Please consider the environment before printing this e-mail.
>
^ permalink raw reply
* Re: pull-request: bpf-next 2018-06-05
From: David Miller @ 2018-06-05 19:22 UTC (permalink / raw)
To: daniel; +Cc: ast, netdev
In-Reply-To: <20180605163916.2922-1-daniel@iogearbox.net>
From: Daniel Borkmann <daniel@iogearbox.net>
Date: Tue, 5 Jun 2018 18:39:16 +0200
> The following pull-request contains BPF updates for your *net-next* tree.
>
> The main changes are:
...
> Please consider pulling these changes from:
>
> git://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf-next.git
I've pulled this in and will push back out after some build testing.
Thanks!
^ permalink raw reply
* Re: [PATCH] r8169: Reinstate ALDPS and ASPM support
From: Heiner Kallweit @ 2018-06-05 19:17 UTC (permalink / raw)
To: Bjorn Helgaas, Ryankao
Cc: Kai Heng Feng, jrg.otte@gmail.com, David Miller, Hayes Wang,
romieu@fr.zoreil.com, Linux Netdev List,
Linux Kernel Mailing List, Hau, linux-pci
In-Reply-To: <20180605191142.GA214338@bhelgaas-glaptop.roam.corp.google.com>
On 05.06.2018 21:11, Bjorn Helgaas wrote:
> [+cc linux-pci]
>
> On Tue, Jun 05, 2018 at 12:28:05PM -0500, Bjorn Helgaas wrote:
>> On Tue, Jun 05, 2018 at 06:34:09AM +0000, Ryankao wrote:
>>> Add realtek folk Hau
>>>
>>> -----Original Message-----
>>> From: Kai Heng Feng [mailto:kai.heng.feng@canonical.com]
>>> Sent: Tuesday, June 05, 2018 1:02 PM
>>> To: jrg.otte@gmail.com
>>> Cc: David Miller <davem@davemloft.net>; Hayes Wang <hayeswang@realtek.com>; hkallweit1@gmail.com; romieu@fr.zoreil.com; Linux Netdev List <netdev@vger.kernel.org>; Linux Kernel Mailing List <linux-kernel@vger.kernel.org>; Ryankao <ryankao@realtek.com>
>>> Subject: Re: [PATCH] r8169: Reinstate ALDPS and ASPM support
>>>
>>> Hi Jörg Otte,
>>>
>>> Can you give this patch a try?
>>>
>>> Since you are the only one that reported ALDPS/ASPM regression,
>>>
>>> And I think this patch should solve the issue you had [1].
>>>
>>> Hopefully we don't need to go down the rabbit hole of blacklist/whitelist...
>>>
>>> Kai-Heng
>>>
>>> [1] https://lkml.org/lkml/2013/1/5/36
>>
>> I have no idea what ALDPS is. It's not mentioned in the PCIe spec, so
>> presumably it's some Realtek-specific thing. ASPM is a generic PCIe
>> thing. Changes to these two things should be in separate patches so
>> they don't get tangled up.
>>
ALDPS = Advanced Link Down Power Saving
And yes, it's a Realtek feature.
>>>> On Jun 5, 2018, at 12:58 PM, Kai-Heng Feng
>>>> <kai.heng.feng@canonical.com>
>>>> wrote:
>>>>
>>>> This patch reinstate ALDPS and ASPM support on r8169.
>>>>
>>>> On some Intel platforms, ASPM support on r8169 is the key factor to
>>>> let Package C-State achieve PC8. Without ASPM support, the deepest
>>>> Package C-State can hit is PC3. PC8 can save additional ~3W in
>>>> comparison with PC3.
>>>>
>>>> This patch is from Realtek.
>>>>
>>>> Fixes: e0c075577965 ("r8169: enable ALDPS for power saving")
>>>> Fixes: d64ec841517a ("r8169: enable internal ASPM and clock request
>>>> settings")
>>
>>>> +3507,15 @@ static void rtl8168e_1_hw_phy_config(struct
>>>> rtl8169_private *tp)
>>>> rtl_writephy(tp, 0x0d, 0x4007);
>>>> rtl_writephy(tp, 0x0e, 0x0000);
>>>> rtl_writephy(tp, 0x0d, 0x0000);
>>>> +
>>>> + /* Check ALDPS bit, disable it if enabled */
>>>> + rtl_writephy(tp, 0x1f, 0x0000);
>>>> + if (enable_aldps)
>>>> + rtl_w0w1_phy(tp, 0x15, 0x1000, 0x0000);
>>>> + else if (rtl_readphy(tp, 0x15) & 0x1000)
>>>> + rtl_w0w1_phy(tp, 0x15, 0x0000, 0x1000);
>>
>> There's a lot of repetition of this code with minor variations. You
>> could probably factor it out and make it more concise and more
>> readable.
>>
>>>> +static void rtl8169_check_link_status(struct net_device *dev,
>>>> + struct rtl8169_private *tp) {
>>>> + struct device *d = tp_to_dev(tp);
>>>> +
>>>> + if (tp->link_ok(tp)) {
>>>> + rtl_link_chg_patch(tp);
>>>> + /* This is to cancel a scheduled suspend if there's one. */
>>>> + if (pm_request_resume(d))
>>>> + _rtl_reset_work(tp);
>>>> + netif_carrier_on(dev);
>>>> + if (net_ratelimit())
>>>> + netif_info(tp, ifup, dev, "link up\n");
>>>> + } else {
>>>> + netif_carrier_off(dev);
>>>> + netif_info(tp, ifdown, dev, "link down\n");
>>>> + pm_runtime_idle(d);
>>>> + }
>>>> +}
>>
>> This function apparently just got moved around without changing
>> anything. That's fine, but the move should be in a separate patch to
>> make the real changes easier to review.
>>
>>>> @@ -7649,8 +7757,12 @@ static int rtl_init_one(struct pci_dev *pdev,
>>>> const struct pci_device_id *ent)
>>>>
>>>> /* disable ASPM completely as that cause random device stop working
>>>> * problems as well as full system hangs for some PCIe devices users */
>>>> - pci_disable_link_state(pdev, PCIE_LINK_STATE_L0S | PCIE_LINK_STATE_L1 |
>>>> - PCIE_LINK_STATE_CLKPM);
>>>> + if (!enable_aspm) {
>>>> + pci_disable_link_state(pdev, PCIE_LINK_STATE_L0S |
>>>> + PCIE_LINK_STATE_L1 |
>>>> + PCIE_LINK_STATE_CLKPM);
>>>> + netif_info(tp, probe, dev, "ASPM disabled\n");
>>>> + }
>>
>> ASPM is a generic PCIe feature that should be configured by the PCI
>> core without any help from the device driver.
>>
>> If code in the driver is needed, that means either the PCI core is
>> doing it wrong and we should fix it there, or the device is broken and
>> the driver is working around the erratum.
>>
>> If this is an erratum, you should include details about exactly what's
>> broken and (ideally) a URL to the published erratum. Otherwise this
>> is just unmaintainable black magic and likely to be broken by future
>> ASPM changes in the PCI core.
>>
>> ASPM configuration is done by the PCI core before drivers are bound to
>> the device. If you need device-specific workarounds, they should
>> probably be in quirks so they're done before the core does that ASPM
>> configuration.
>>
>>>> /* enable device (incl. PCI PM wakeup and hotplug setup) */
>>>> rc = pcim_enable_device(pdev);
>>>> --
>>>> 2.17.0
>>>
>>> ------Please consider the environment before printing this e-mail.
>
^ permalink raw reply
* Re: suspicius csum initialization in vmxnet3_rx_csum
From: Ronak Doshi @ 2018-06-05 19:15 UTC (permalink / raw)
To: Paolo Abeni; +Cc: Neil Horman, Guolin Yang, Boon Ang, Louis Luo, netdev
In-Reply-To: <9296894c5f1ffb300ec18d7902751f68d914615b.camel@redhat.com>
On Tue, 5 Jun 2018, Paolo Abeni wrote:
> Hi,
>
> I'm sorry for the long delay in my answer, I've been travelling.
>
> On Fri, 2018-06-01 at 11:10 -0700, Ronak Doshi wrote:
> > On Thu, 31 May 2018, Neil Horman wrote:
> > > What packet types will rcd.csum be set for?
> > > Neil
> >
> > I looked thorugh the emulation code and found that rcd.csum is not set.
> > For valid v4/v6, TCP/UDP packets the code block above the mentioend "if"
> > block will be executed or else it will go through checksum none.
> >
> > That's why I wanted to know (in previous emails) which ESX build is being
> > used while this was tested. The code block under "if (gdesc->rcd.csum)"
> > block might seem incorrect but it shouldn't be hit as rcd.csum is not set.
>
> I'm unsure if I read the above correctly. Do you mean that the relevant
> code-path is never hit? If so, can we simply drop it, as we agreed that
> such code is uncorrect? Elsewhere, could you plese specify under which
> circumstances gdesc->rcd.csum is filled by the hypervisor?
>
I do not see hypervisor populating rcd.csum field or may be the code has
changed over the years. So, the codepath should not be hit as it is not
populated. I will check and fix it or remove the block if not required.
But as far as your issue is concerned, that code block is not hit.
Thanks,
Ronak
^ permalink raw reply
* Re: general protection fault in sockfs_setattr
From: Cong Wang @ 2018-06-05 19:14 UTC (permalink / raw)
To: shankarapailoor
Cc: David Miller, LKML, syzkaller, Linux Kernel Network Developers
In-Reply-To: <CAB+yDabFuBpT5UU1Hy0s4kY5UKJzA84=6fNieNcdTjjZNq5SHQ@mail.gmail.com>
On Mon, Jun 4, 2018 at 9:53 PM, shankarapailoor
<shankarapailoor@gmail.com> wrote:
> Hi,
>
> I have been fuzzing Linux 4.17-rc7 with Syzkaller and found the
> following crash: https://pastebin.com/ixX3RB9j
>
> Syzkaller isolated the cause of the bug to the following program:
>
> socketpair$unix(0x1, 0x1, 0x0,
> &(0x7f0000000000)={<r0=>0xffffffffffffffff, <r1=>0xffffffffffffffff})
> getresuid(&(0x7f0000000080)=<r2=>0x0, &(0x7f00000000c0),
> &(0x7f0000000700))r3 = getegid()
> fchownat(r0, &(0x7f0000000040)='\x00', r2, r3, 0x1000)
> dup3(r1, r0, 0x80000)
>
>
> The problematic area appears to be here:
>
> static int sockfs_setattr(struct dentry *dentry, struct iattr *iattr)
> {
> int err = simple_setattr(dentry, iattr);
>
> if (!err && (iattr->ia_valid & ATTR_UID)) {
> struct socket *sock = SOCKET_I(d_inode(dentry));
>
> sock->sk->sk_uid = iattr->ia_uid; //KASAN GPF
> }
> return err;
> }
>
> If dup3 is called concurrently with fchownat then can sock->sk be NULL?
Although dup3() implies a close(), fd is refcnt'ted, if dup3() runs
concurrently with fchownat() it should not be closed until whoever
the last closes it.
Or maybe fchownat() doesn't even hold refcnt of fd, since it aims
to change the file backed.
Not sure if the following is sufficient, inode might need to be protected
with some lock...
diff --git a/net/socket.c b/net/socket.c
index f10f1d947c78..6294b4b3132e 100644
--- a/net/socket.c
+++ b/net/socket.c
@@ -537,7 +537,10 @@ static int sockfs_setattr(struct dentry *dentry,
struct iattr *iattr)
if (!err && (iattr->ia_valid & ATTR_UID)) {
struct socket *sock = SOCKET_I(d_inode(dentry));
- sock->sk->sk_uid = iattr->ia_uid;
+ if (sock->sk)
+ sock->sk->sk_uid = iattr->ia_uid;
+ else
+ err = -ENOENT;
}
return err;
^ permalink raw reply related
* Re: [PATCH] r8169: Reinstate ALDPS and ASPM support
From: Heiner Kallweit @ 2018-06-05 19:13 UTC (permalink / raw)
To: Kai-Heng Feng, davem; +Cc: hayeswang, romieu, netdev, linux-kernel, Ryankao
In-Reply-To: <20180605045812.17977-1-kai.heng.feng@canonical.com>
On 05.06.2018 06:58, Kai-Heng Feng wrote:
> This patch reinstate ALDPS and ASPM support on r8169.
>
> On some Intel platforms, ASPM support on r8169 is the key factor to let
> Package C-State achieve PC8. Without ASPM support, the deepest Package
> C-State can hit is PC3. PC8 can save additional ~3W in comparison with
> PC3.
>
> This patch is from Realtek.
>
> Fixes: e0c075577965 ("r8169: enable ALDPS for power saving")
> Fixes: d64ec841517a ("r8169: enable internal ASPM and clock request settings")
>
> Cc: Ryankao <ryankao@realtek.com>
> Signed-off-by: Kai-Heng Feng <kai.heng.feng@canonical.com>
> ---
> drivers/net/ethernet/realtek/r8169.c | 190 +++++++++++++++++++++------
> 1 file changed, 151 insertions(+), 39 deletions(-)
>
> diff --git a/drivers/net/ethernet/realtek/r8169.c b/drivers/net/ethernet/realtek/r8169.c
> index 75dfac0248f4..a28ef20be221 100644
> --- a/drivers/net/ethernet/realtek/r8169.c
> +++ b/drivers/net/ethernet/realtek/r8169.c
> @@ -319,6 +319,8 @@ static const struct pci_device_id rtl8169_pci_tbl[] = {
>
> MODULE_DEVICE_TABLE(pci, rtl8169_pci_tbl);
>
> +static int enable_aspm = 1;
> +static int enable_aldps = 1;
> static int use_dac = -1;
> static struct {
> u32 msg_enable;
> @@ -817,6 +819,10 @@ struct rtl8169_private {
>
> MODULE_AUTHOR("Realtek and the Linux r8169 crew <netdev@vger.kernel.org>");
> MODULE_DESCRIPTION("RealTek RTL-8169 Gigabit Ethernet driver");
> +module_param(enable_aspm, int, 0);
> +MODULE_PARM_DESC(enable_aspm, "Enable ASPM");
> +module_param(enable_aldps, int, 0);
> +MODULE_PARM_DESC(enable_aldps, "Enable ALDPS");
> module_param(use_dac, int, 0);
> MODULE_PARM_DESC(use_dac, "Enable PCI DAC. Unsafe on 32 bit PCI slot.");
> module_param_named(debug, debug.msg_enable, int, 0);
> @@ -1567,25 +1573,6 @@ static void rtl_link_chg_patch(struct rtl8169_private *tp)
> }
> }
>
> -static void rtl8169_check_link_status(struct net_device *dev,
> - struct rtl8169_private *tp)
> -{
> - struct device *d = tp_to_dev(tp);
> -
> - if (tp->link_ok(tp)) {
> - rtl_link_chg_patch(tp);
> - /* This is to cancel a scheduled suspend if there's one. */
> - pm_request_resume(d);
> - netif_carrier_on(dev);
> - if (net_ratelimit())
> - netif_info(tp, ifup, dev, "link up\n");
> - } else {
> - netif_carrier_off(dev);
> - netif_info(tp, ifdown, dev, "link down\n");
> - pm_runtime_idle(d);
> - }
> -}
> -
> #define WAKE_ANY (WAKE_PHY | WAKE_MAGIC | WAKE_UCAST | WAKE_BCAST | WAKE_MCAST)
>
> static u32 __rtl8169_get_wol(struct rtl8169_private *tp)
> @@ -3520,6 +3507,15 @@ static void rtl8168e_1_hw_phy_config(struct rtl8169_private *tp)
> rtl_writephy(tp, 0x0d, 0x4007);
> rtl_writephy(tp, 0x0e, 0x0000);
> rtl_writephy(tp, 0x0d, 0x0000);
> +
> + /* Check ALDPS bit, disable it if enabled */
> + rtl_writephy(tp, 0x1f, 0x0000);
> + if (enable_aldps)
> + rtl_w0w1_phy(tp, 0x15, 0x1000, 0x0000);
> + else if (rtl_readphy(tp, 0x15) & 0x1000)
> + rtl_w0w1_phy(tp, 0x15, 0x0000, 0x1000);
> +
> + rtl_writephy(tp, 0x1f, 0x0000);
Few remarks:
- The comment isn't applicable any longer.
- The second rtl_writephy(tp, 0x1f, 0x0000) isn't needed because you don't
switch the page in between.
- The code is a little hard to read, instead you could use the following
and create a helper, ideally with register and bit number as
parameters so that you can use it for all affected chip types.
val = rtl_readphy(tp, 0x15);
val &= ~BIT(12);
if (enable_aldps)
val |= BIT(12);
rtl_writephy(tp, 0x15, val);
> }
>
> static void rtl_rar_exgmac_set(struct rtl8169_private *tp, u8 *addr)
> @@ -3627,6 +3623,15 @@ static void rtl8168e_2_hw_phy_config(struct rtl8169_private *tp)
>
> /* Broken BIOS workaround: feed GigaMAC registers with MAC address. */
> rtl_rar_exgmac_set(tp, tp->dev->dev_addr);
> +
> + /* Check ALDPS bit, disable it if enabled */
> + rtl_writephy(tp, 0x1f, 0x0000);
> + if (enable_aldps)
> + rtl_w0w1_phy(tp, 0x15, 0x1000, 0x0000);
> + else if (rtl_readphy(tp, 0x15) & 0x1000)
> + rtl_w0w1_phy(tp, 0x15, 0x0000, 0x1000);
> +
> + rtl_writephy(tp, 0x1f, 0x0000);
> }
>
> static void rtl8168f_hw_phy_config(struct rtl8169_private *tp)
> @@ -3649,6 +3654,15 @@ static void rtl8168f_hw_phy_config(struct rtl8169_private *tp)
> rtl_writephy(tp, 0x05, 0x8b86);
> rtl_w0w1_phy(tp, 0x06, 0x0001, 0x0000);
> rtl_writephy(tp, 0x1f, 0x0000);
> +
> + /* Check ALDPS bit, disable it if enabled */
> + rtl_writephy(tp, 0x1f, 0x0000);
> + if (enable_aldps)
> + rtl_w0w1_phy(tp, 0x15, 0x1000, 0x0000);
> + else if (rtl_readphy(tp, 0x15) & 0x1000)
> + rtl_w0w1_phy(tp, 0x15, 0x0000, 0x1000);
> +
> + rtl_writephy(tp, 0x1f, 0x0000);
> }
>
> static void rtl8168f_1_hw_phy_config(struct rtl8169_private *tp)
> @@ -3865,7 +3879,9 @@ static void rtl8168g_1_hw_phy_config(struct rtl8169_private *tp)
>
> /* Check ALDPS bit, disable it if enabled */
> rtl_writephy(tp, 0x1f, 0x0a43);
> - if (rtl_readphy(tp, 0x10) & 0x0004)
> + if (enable_aldps)
> + rtl_w0w1_phy(tp, 0x10, 0x0004, 0x0000);
> + else if (rtl_readphy(tp, 0x10) & 0x0004)
> rtl_w0w1_phy(tp, 0x10, 0x0000, 0x0004);
>
> rtl_writephy(tp, 0x1f, 0x0000);
> @@ -3874,6 +3890,14 @@ static void rtl8168g_1_hw_phy_config(struct rtl8169_private *tp)
> static void rtl8168g_2_hw_phy_config(struct rtl8169_private *tp)
> {
> rtl_apply_firmware(tp);
> +
> + rtl_writephy(tp, 0x1f, 0x0a43);
> + if (enable_aldps)
> + rtl_w0w1_phy(tp, 0x10, 0x0004, 0x0000);
> + else if (rtl_readphy(tp, 0x10) & 0x0004)
> + rtl_w0w1_phy(tp, 0x10, 0x0000, 0x0004);
> +
> + rtl_writephy(tp, 0x1f, 0x0000);
> }
>
> static void rtl8168h_1_hw_phy_config(struct rtl8169_private *tp)
> @@ -3980,7 +4004,9 @@ static void rtl8168h_1_hw_phy_config(struct rtl8169_private *tp)
>
> /* Check ALDPS bit, disable it if enabled */
> rtl_writephy(tp, 0x1f, 0x0a43);
> - if (rtl_readphy(tp, 0x10) & 0x0004)
> + if (enable_aldps)
> + rtl_w0w1_phy(tp, 0x10, 0x0004, 0x0000);
> + else if (rtl_readphy(tp, 0x10) & 0x0004)
> rtl_w0w1_phy(tp, 0x10, 0x0000, 0x0004);
>
> rtl_writephy(tp, 0x1f, 0x0000);
> @@ -4053,7 +4079,9 @@ static void rtl8168h_2_hw_phy_config(struct rtl8169_private *tp)
>
> /* Check ALDPS bit, disable it if enabled */
> rtl_writephy(tp, 0x1f, 0x0a43);
> - if (rtl_readphy(tp, 0x10) & 0x0004)
> + if (enable_aldps)
> + rtl_w0w1_phy(tp, 0x10, 0x0004, 0x0000);
> + else if (rtl_readphy(tp, 0x10) & 0x0004)
> rtl_w0w1_phy(tp, 0x10, 0x0000, 0x0004);
>
> rtl_writephy(tp, 0x1f, 0x0000);
> @@ -4095,7 +4123,9 @@ static void rtl8168ep_1_hw_phy_config(struct rtl8169_private *tp)
>
> /* Check ALDPS bit, disable it if enabled */
> rtl_writephy(tp, 0x1f, 0x0a43);
> - if (rtl_readphy(tp, 0x10) & 0x0004)
> + if (enable_aldps)
> + rtl_w0w1_phy(tp, 0x10, 0x0004, 0x0000);
> + else if (rtl_readphy(tp, 0x10) & 0x0004)
> rtl_w0w1_phy(tp, 0x10, 0x0000, 0x0004);
>
> rtl_writephy(tp, 0x1f, 0x0000);
> @@ -4186,7 +4216,9 @@ static void rtl8168ep_2_hw_phy_config(struct rtl8169_private *tp)
>
> /* Check ALDPS bit, disable it if enabled */
> rtl_writephy(tp, 0x1f, 0x0a43);
> - if (rtl_readphy(tp, 0x10) & 0x0004)
> + if (enable_aldps)
> + rtl_w0w1_phy(tp, 0x10, 0x0004, 0x0000);
> + else if (rtl_readphy(tp, 0x10) & 0x0004)
> rtl_w0w1_phy(tp, 0x10, 0x0000, 0x0004);
>
> rtl_writephy(tp, 0x1f, 0x0000);
> @@ -4233,6 +4265,15 @@ static void rtl8105e_hw_phy_config(struct rtl8169_private *tp)
> rtl_apply_firmware(tp);
>
> rtl_writephy_batch(tp, phy_reg_init, ARRAY_SIZE(phy_reg_init));
> +
> + /* Check ALDPS bit, disable it if enabled */
> + rtl_writephy(tp, 0x1f, 0x0000);
> + if (enable_aldps)
> + rtl_w0w1_phy(tp, 0x18, 0x1000, 0x0000);
> + else if (rtl_readphy(tp, 0x18) & 0x1000)
> + rtl_w0w1_phy(tp, 0x18, 0x0000, 0x1000);
> +
> + rtl_writephy(tp, 0x1f, 0x0000);
> }
>
> static void rtl8402_hw_phy_config(struct rtl8169_private *tp)
> @@ -4250,6 +4291,15 @@ static void rtl8402_hw_phy_config(struct rtl8169_private *tp)
> rtl_writephy(tp, 0x10, 0x401f);
> rtl_writephy(tp, 0x19, 0x7030);
> rtl_writephy(tp, 0x1f, 0x0000);
> +
> + /* Check ALDPS bit, disable it if enabled */
> + rtl_writephy(tp, 0x1f, 0x0000);
> + if (enable_aldps)
> + rtl_w0w1_phy(tp, 0x18, 0x1000, 0x0000);
> + else if (rtl_readphy(tp, 0x18) & 0x1000)
> + rtl_w0w1_phy(tp, 0x18, 0x0000, 0x1000);
> +
> + rtl_writephy(tp, 0x1f, 0x0000);
> }
>
> static void rtl8106e_hw_phy_config(struct rtl8169_private *tp)
> @@ -4272,6 +4322,15 @@ static void rtl8106e_hw_phy_config(struct rtl8169_private *tp)
> rtl_writephy_batch(tp, phy_reg_init, ARRAY_SIZE(phy_reg_init));
>
> rtl_eri_write(tp, 0x1d0, ERIAR_MASK_0011, 0x0000, ERIAR_EXGMAC);
> +
> + /* Check ALDPS bit, disable it if enabled */
> + rtl_writephy(tp, 0x1f, 0x0000);
> + if (enable_aldps)
> + rtl_w0w1_phy(tp, 0x18, 0x1000, 0x0000);
> + else if (rtl_readphy(tp, 0x18) & 0x1000)
> + rtl_w0w1_phy(tp, 0x18, 0x0000, 0x1000);
> +
> + rtl_writephy(tp, 0x1f, 0x0000);
> }
>
> static void rtl_hw_phy_config(struct net_device *dev)
> @@ -5290,6 +5349,18 @@ static void rtl_pcie_state_l2l3_enable(struct rtl8169_private *tp, bool enable)
> RTL_W8(tp, Config3, data);
> }
>
> +static void rtl_hw_internal_aspm_clkreq_enable(struct rtl8169_private *tp,
> + bool enable)
> +{
> + if (enable) {
> + RTL_W8(tp, Config2, RTL_R8(tp, Config2) | ClkReqEn);
> + RTL_W8(tp, Config5, RTL_R8(tp, Config5) | ASPM_en);
> + } else {
> + RTL_W8(tp, Config2, RTL_R8(tp, Config2) & ~ClkReqEn);
> + RTL_W8(tp, Config5, RTL_R8(tp, Config5) & ~ASPM_en);
> + }
> +}
> +
> static void rtl_hw_start_8168bb(struct rtl8169_private *tp)
> {
> RTL_W8(tp, Config3, RTL_R8(tp, Config3) & ~Beacon_en);
> @@ -5646,9 +5717,10 @@ static void rtl_hw_start_8168g_1(struct rtl8169_private *tp)
> rtl_hw_start_8168g(tp);
>
> /* disable aspm and clock request before access ephy */
> - RTL_W8(tp, Config2, RTL_R8(tp, Config2) & ~ClkReqEn);
> - RTL_W8(tp, Config5, RTL_R8(tp, Config5) & ~ASPM_en);
> + rtl_hw_internal_aspm_clkreq_enable(tp, false);
> rtl_ephy_init(tp, e_info_8168g_1, ARRAY_SIZE(e_info_8168g_1));
> + if (enable_aspm)
> + rtl_hw_internal_aspm_clkreq_enable(tp, true);
> }
>
> static void rtl_hw_start_8168g_2(struct rtl8169_private *tp)
> @@ -5681,9 +5753,10 @@ static void rtl_hw_start_8411_2(struct rtl8169_private *tp)
> rtl_hw_start_8168g(tp);
>
> /* disable aspm and clock request before access ephy */
> - RTL_W8(tp, Config2, RTL_R8(tp, Config2) & ~ClkReqEn);
> - RTL_W8(tp, Config5, RTL_R8(tp, Config5) & ~ASPM_en);
> + rtl_hw_internal_aspm_clkreq_enable(tp, false);
> rtl_ephy_init(tp, e_info_8411_2, ARRAY_SIZE(e_info_8411_2));
> + if (enable_aspm)
> + rtl_hw_internal_aspm_clkreq_enable(tp, true);
> }
>
> static void rtl_hw_start_8168h_1(struct rtl8169_private *tp)
> @@ -5700,8 +5773,7 @@ static void rtl_hw_start_8168h_1(struct rtl8169_private *tp)
> };
>
> /* disable aspm and clock request before access ephy */
> - RTL_W8(tp, Config2, RTL_R8(tp, Config2) & ~ClkReqEn);
> - RTL_W8(tp, Config5, RTL_R8(tp, Config5) & ~ASPM_en);
> + rtl_hw_internal_aspm_clkreq_enable(tp, false);
> rtl_ephy_init(tp, e_info_8168h_1, ARRAY_SIZE(e_info_8168h_1));
>
> RTL_W32(tp, TxConfig, RTL_R32(tp, TxConfig) | TXCFG_AUTO_FIFO);
> @@ -5780,6 +5852,9 @@ static void rtl_hw_start_8168h_1(struct rtl8169_private *tp)
> r8168_mac_ocp_write(tp, 0xe63e, 0x0000);
> r8168_mac_ocp_write(tp, 0xc094, 0x0000);
> r8168_mac_ocp_write(tp, 0xc09e, 0x0000);
> +
> + if (enable_aspm)
> + rtl_hw_internal_aspm_clkreq_enable(tp, true);
> }
>
> static void rtl_hw_start_8168ep(struct rtl8169_private *tp)
> @@ -5831,11 +5906,13 @@ static void rtl_hw_start_8168ep_1(struct rtl8169_private *tp)
> };
>
> /* disable aspm and clock request before access ephy */
> - RTL_W8(tp, Config2, RTL_R8(tp, Config2) & ~ClkReqEn);
> - RTL_W8(tp, Config5, RTL_R8(tp, Config5) & ~ASPM_en);
> + rtl_hw_internal_aspm_clkreq_enable(tp, false);
> rtl_ephy_init(tp, e_info_8168ep_1, ARRAY_SIZE(e_info_8168ep_1));
>
> rtl_hw_start_8168ep(tp);
> +
> + if (enable_aspm)
> + rtl_hw_internal_aspm_clkreq_enable(tp, true);
> }
>
> static void rtl_hw_start_8168ep_2(struct rtl8169_private *tp)
> @@ -5847,14 +5924,16 @@ static void rtl_hw_start_8168ep_2(struct rtl8169_private *tp)
> };
>
> /* disable aspm and clock request before access ephy */
> - RTL_W8(tp, Config2, RTL_R8(tp, Config2) & ~ClkReqEn);
> - RTL_W8(tp, Config5, RTL_R8(tp, Config5) & ~ASPM_en);
> + rtl_hw_internal_aspm_clkreq_enable(tp, false);
> rtl_ephy_init(tp, e_info_8168ep_2, ARRAY_SIZE(e_info_8168ep_2));
>
> rtl_hw_start_8168ep(tp);
>
> RTL_W8(tp, DLLPR, RTL_R8(tp, DLLPR) & ~PFM_EN);
> RTL_W8(tp, MISC_1, RTL_R8(tp, MISC_1) & ~PFM_D3COLD_EN);
> +
> + if (enable_aspm)
> + rtl_hw_internal_aspm_clkreq_enable(tp, true);
> }
>
> static void rtl_hw_start_8168ep_3(struct rtl8169_private *tp)
> @@ -5868,8 +5947,7 @@ static void rtl_hw_start_8168ep_3(struct rtl8169_private *tp)
> };
>
> /* disable aspm and clock request before access ephy */
> - RTL_W8(tp, Config2, RTL_R8(tp, Config2) & ~ClkReqEn);
> - RTL_W8(tp, Config5, RTL_R8(tp, Config5) & ~ASPM_en);
> + rtl_hw_internal_aspm_clkreq_enable(tp, false);
> rtl_ephy_init(tp, e_info_8168ep_3, ARRAY_SIZE(e_info_8168ep_3));
>
> rtl_hw_start_8168ep(tp);
> @@ -5889,6 +5967,9 @@ static void rtl_hw_start_8168ep_3(struct rtl8169_private *tp)
> data = r8168_mac_ocp_read(tp, 0xe860);
> data |= 0x0080;
> r8168_mac_ocp_write(tp, 0xe860, data);
> +
> + if (enable_aspm)
> + rtl_hw_internal_aspm_clkreq_enable(tp, true);
> }
>
> static void rtl_hw_start_8168(struct rtl8169_private *tp)
> @@ -6364,7 +6445,7 @@ static void rtl8169_tx_clear(struct rtl8169_private *tp)
> tp->cur_tx = tp->dirty_tx = 0;
> }
>
> -static void rtl_reset_work(struct rtl8169_private *tp)
> +static void _rtl_reset_work(struct rtl8169_private *tp)
> {
> struct net_device *dev = tp->dev;
> int i;
> @@ -6384,6 +6465,33 @@ static void rtl_reset_work(struct rtl8169_private *tp)
> napi_enable(&tp->napi);
> rtl_hw_start(tp);
> netif_wake_queue(dev);
> +}
> +
> +static void rtl8169_check_link_status(struct net_device *dev,
> + struct rtl8169_private *tp)
> +{
> + struct device *d = tp_to_dev(tp);
> +
> + if (tp->link_ok(tp)) {
> + rtl_link_chg_patch(tp);
> + /* This is to cancel a scheduled suspend if there's one. */
> + if (pm_request_resume(d))
> + _rtl_reset_work(tp);
This reset was added, what is it good for and how is it related to
ASPM/ALDPS? It looks a little bogus, especially considering that
pm_request_resume() can return also positive values.
> + netif_carrier_on(dev);
> + if (net_ratelimit())
> + netif_info(tp, ifup, dev, "link up\n");
> + } else {
> + netif_carrier_off(dev);
> + netif_info(tp, ifdown, dev, "link down\n");
> + pm_runtime_idle(d);
> + }
> +}
> +
> +static void rtl_reset_work(struct rtl8169_private *tp)
> +{
> + struct net_device *dev = tp->dev;
> +
> + _rtl_reset_work(tp);
> rtl8169_check_link_status(dev, tp);
> }
>
> @@ -7649,8 +7757,12 @@ static int rtl_init_one(struct pci_dev *pdev, const struct pci_device_id *ent)
>
> /* disable ASPM completely as that cause random device stop working
> * problems as well as full system hangs for some PCIe devices users */
> - pci_disable_link_state(pdev, PCIE_LINK_STATE_L0S | PCIE_LINK_STATE_L1 |
> - PCIE_LINK_STATE_CLKPM);
> + if (!enable_aspm) {
> + pci_disable_link_state(pdev, PCIE_LINK_STATE_L0S |
> + PCIE_LINK_STATE_L1 |
> + PCIE_LINK_STATE_CLKPM);
> + netif_info(tp, probe, dev, "ASPM disabled\n");
You should use dev_info() here because the net_device isn't registered yet.
> + }
>
> /* enable device (incl. PCI PM wakeup and hotplug setup) */
> rc = pcim_enable_device(pdev);
>
^ permalink raw reply
* Re: [PATCH] r8169: Reinstate ALDPS and ASPM support
From: Bjorn Helgaas @ 2018-06-05 19:11 UTC (permalink / raw)
To: Ryankao
Cc: Kai Heng Feng, jrg.otte@gmail.com, David Miller, Hayes Wang,
hkallweit1@gmail.com, romieu@fr.zoreil.com, Linux Netdev List,
Linux Kernel Mailing List, Hau, linux-pci
In-Reply-To: <20180605172805.GD30381@bhelgaas-glaptop.roam.corp.google.com>
[+cc linux-pci]
On Tue, Jun 05, 2018 at 12:28:05PM -0500, Bjorn Helgaas wrote:
> On Tue, Jun 05, 2018 at 06:34:09AM +0000, Ryankao wrote:
> > Add realtek folk Hau
> >
> > -----Original Message-----
> > From: Kai Heng Feng [mailto:kai.heng.feng@canonical.com]
> > Sent: Tuesday, June 05, 2018 1:02 PM
> > To: jrg.otte@gmail.com
> > Cc: David Miller <davem@davemloft.net>; Hayes Wang <hayeswang@realtek.com>; hkallweit1@gmail.com; romieu@fr.zoreil.com; Linux Netdev List <netdev@vger.kernel.org>; Linux Kernel Mailing List <linux-kernel@vger.kernel.org>; Ryankao <ryankao@realtek.com>
> > Subject: Re: [PATCH] r8169: Reinstate ALDPS and ASPM support
> >
> > Hi Jörg Otte,
> >
> > Can you give this patch a try?
> >
> > Since you are the only one that reported ALDPS/ASPM regression,
> >
> > And I think this patch should solve the issue you had [1].
> >
> > Hopefully we don't need to go down the rabbit hole of blacklist/whitelist...
> >
> > Kai-Heng
> >
> > [1] https://lkml.org/lkml/2013/1/5/36
>
> I have no idea what ALDPS is. It's not mentioned in the PCIe spec, so
> presumably it's some Realtek-specific thing. ASPM is a generic PCIe
> thing. Changes to these two things should be in separate patches so
> they don't get tangled up.
>
> > > On Jun 5, 2018, at 12:58 PM, Kai-Heng Feng
> > > <kai.heng.feng@canonical.com>
> > > wrote:
> > >
> > > This patch reinstate ALDPS and ASPM support on r8169.
> > >
> > > On some Intel platforms, ASPM support on r8169 is the key factor to
> > > let Package C-State achieve PC8. Without ASPM support, the deepest
> > > Package C-State can hit is PC3. PC8 can save additional ~3W in
> > > comparison with PC3.
> > >
> > > This patch is from Realtek.
> > >
> > > Fixes: e0c075577965 ("r8169: enable ALDPS for power saving")
> > > Fixes: d64ec841517a ("r8169: enable internal ASPM and clock request
> > > settings")
>
> > > +3507,15 @@ static void rtl8168e_1_hw_phy_config(struct
> > > rtl8169_private *tp)
> > > rtl_writephy(tp, 0x0d, 0x4007);
> > > rtl_writephy(tp, 0x0e, 0x0000);
> > > rtl_writephy(tp, 0x0d, 0x0000);
> > > +
> > > + /* Check ALDPS bit, disable it if enabled */
> > > + rtl_writephy(tp, 0x1f, 0x0000);
> > > + if (enable_aldps)
> > > + rtl_w0w1_phy(tp, 0x15, 0x1000, 0x0000);
> > > + else if (rtl_readphy(tp, 0x15) & 0x1000)
> > > + rtl_w0w1_phy(tp, 0x15, 0x0000, 0x1000);
>
> There's a lot of repetition of this code with minor variations. You
> could probably factor it out and make it more concise and more
> readable.
>
> > > +static void rtl8169_check_link_status(struct net_device *dev,
> > > + struct rtl8169_private *tp) {
> > > + struct device *d = tp_to_dev(tp);
> > > +
> > > + if (tp->link_ok(tp)) {
> > > + rtl_link_chg_patch(tp);
> > > + /* This is to cancel a scheduled suspend if there's one. */
> > > + if (pm_request_resume(d))
> > > + _rtl_reset_work(tp);
> > > + netif_carrier_on(dev);
> > > + if (net_ratelimit())
> > > + netif_info(tp, ifup, dev, "link up\n");
> > > + } else {
> > > + netif_carrier_off(dev);
> > > + netif_info(tp, ifdown, dev, "link down\n");
> > > + pm_runtime_idle(d);
> > > + }
> > > +}
>
> This function apparently just got moved around without changing
> anything. That's fine, but the move should be in a separate patch to
> make the real changes easier to review.
>
> > > @@ -7649,8 +7757,12 @@ static int rtl_init_one(struct pci_dev *pdev,
> > > const struct pci_device_id *ent)
> > >
> > > /* disable ASPM completely as that cause random device stop working
> > > * problems as well as full system hangs for some PCIe devices users */
> > > - pci_disable_link_state(pdev, PCIE_LINK_STATE_L0S | PCIE_LINK_STATE_L1 |
> > > - PCIE_LINK_STATE_CLKPM);
> > > + if (!enable_aspm) {
> > > + pci_disable_link_state(pdev, PCIE_LINK_STATE_L0S |
> > > + PCIE_LINK_STATE_L1 |
> > > + PCIE_LINK_STATE_CLKPM);
> > > + netif_info(tp, probe, dev, "ASPM disabled\n");
> > > + }
>
> ASPM is a generic PCIe feature that should be configured by the PCI
> core without any help from the device driver.
>
> If code in the driver is needed, that means either the PCI core is
> doing it wrong and we should fix it there, or the device is broken and
> the driver is working around the erratum.
>
> If this is an erratum, you should include details about exactly what's
> broken and (ideally) a URL to the published erratum. Otherwise this
> is just unmaintainable black magic and likely to be broken by future
> ASPM changes in the PCI core.
>
> ASPM configuration is done by the PCI core before drivers are bound to
> the device. If you need device-specific workarounds, they should
> probably be in quirks so they're done before the core does that ASPM
> configuration.
>
> > > /* enable device (incl. PCI PM wakeup and hotplug setup) */
> > > rc = pcim_enable_device(pdev);
> > > --
> > > 2.17.0
> >
> > ------Please consider the environment before printing this e-mail.
^ permalink raw reply
* Re: [PATCH net] net: sched: cls: Fix offloading when ingress dev is vxlan
From: David Miller @ 2018-06-05 19:06 UTC (permalink / raw)
To: kubakici
Cc: paulb, jiri, xiyou.wangcong, jhs, netdev, kliteyn, roid, shahark,
markb, ogerlitz
In-Reply-To: <20180605115747.0e939ac4@cakuba.netronome.com>
From: Jakub Kicinski <kubakici@wp.pl>
Date: Tue, 5 Jun 2018 11:57:47 -0700
> Do we still care about correctness and not breaking backward
> compatibility?
Jakub let me know if you want me to revert this change.
^ permalink raw reply
* Re: [PATCH net] net: sched: cls: Fix offloading when ingress dev is vxlan
From: Jakub Kicinski @ 2018-06-05 18:59 UTC (permalink / raw)
To: Paul Blakey
Cc: Jiri Pirko, Cong Wang, Jamal Hadi Salim, David Miller, netdev,
Yevgeny Kliteynik, Roi Dayan, Shahar Klein, Mark Bloch,
Or Gerlitz
In-Reply-To: <20180605115747.0e939ac4@cakuba.netronome.com>
On Tue, 5 Jun 2018 11:57:47 -0700, Jakub Kicinski wrote:
> On Tue, 5 Jun 2018 11:04:03 +0300, Paul Blakey wrote:
> > When using a vxlan device as the ingress dev, we count it as a
> > "no offload dev", so when such a rule comes and err stop is true,
> > we fail early and don't try the egdev route which can offload it
> > through the egress device.
> >
> > Fix that by not calling the block offload if one of the devices
> > attached to it is not offload capable, but make sure egress on such case
> > is capable instead.
> >
> > Fixes: caa7260156eb ("net: sched: keep track of offloaded filters [..]")
> > Reviewed-by: Roi Dayan <roid@mellanox.com>
> > Acked-by: Jiri Pirko <jiri@mellanox.com>
> > Signed-off-by: Paul Blakey <paulb@mellanox.com>
>
> Very poor commit message. What you're doing is re-enabling skip_sw
> filters on tunnel devices which semantically make no sense and
> shouldn't have been allowed in the first place.
>
> This will breaks block sharing between tunnels and HW netdevs (because
> you skip the tcf_block_cb_call() completely). The entire egdev idea
> remains badly broken accepting filters like this:
>
> # tc filter add dev vxlan0 ingress \
> matchall action skip_sw \
> mirred egress redirect dev lo \
> mirred egress redirect dev sw1np0
For above we probably need something like this (untested):
diff --git a/net/sched/act_api.c b/net/sched/act_api.c
index 3f4cf930f809..71e5eebec31a 100644
--- a/net/sched/act_api.c
+++ b/net/sched/act_api.c
@@ -1511,7 +1511,7 @@ int tc_setup_cb_egdev_call(const struct net_device *dev,
struct tcf_action_egdev *egdev = tcf_action_egdev_lookup(dev);
if (!egdev)
- return 0;
+ return err_stop ? -EOPNOTSUPP : 0;
return tcf_action_egdev_cb_call(egdev, type, type_data, err_stop);
}
EXPORT_SYMBOL_GPL(tc_setup_cb_egdev_call);
But the correct fix is to remove egdev crutch completely IMO.
^ permalink raw reply related
* Re: [PATCH net] net: sched: cls: Fix offloading when ingress dev is vxlan
From: Jakub Kicinski @ 2018-06-05 18:57 UTC (permalink / raw)
To: Paul Blakey
Cc: Jiri Pirko, Cong Wang, Jamal Hadi Salim, David Miller, netdev,
Yevgeny Kliteynik, Roi Dayan, Shahar Klein, Mark Bloch,
Or Gerlitz
In-Reply-To: <1528185843-18645-1-git-send-email-paulb@mellanox.com>
On Tue, 5 Jun 2018 11:04:03 +0300, Paul Blakey wrote:
> When using a vxlan device as the ingress dev, we count it as a
> "no offload dev", so when such a rule comes and err stop is true,
> we fail early and don't try the egdev route which can offload it
> through the egress device.
>
> Fix that by not calling the block offload if one of the devices
> attached to it is not offload capable, but make sure egress on such case
> is capable instead.
>
> Fixes: caa7260156eb ("net: sched: keep track of offloaded filters [..]")
> Reviewed-by: Roi Dayan <roid@mellanox.com>
> Acked-by: Jiri Pirko <jiri@mellanox.com>
> Signed-off-by: Paul Blakey <paulb@mellanox.com>
Very poor commit message. What you're doing is re-enabling skip_sw
filters on tunnel devices which semantically make no sense and
shouldn't have been allowed in the first place.
This will breaks block sharing between tunnels and HW netdevs (because
you skip the tcf_block_cb_call() completely). The entire egdev idea
remains badly broken accepting filters like this:
# tc filter add dev vxlan0 ingress \
matchall action skip_sw \
mirred egress redirect dev lo \
mirred egress redirect dev sw1np0
Do we still care about correctness and not breaking backward
compatibility?
^ permalink raw reply
* Re: [PATCH net] failover: eliminate callback hell
From: Stephen Hemminger @ 2018-06-05 18:53 UTC (permalink / raw)
To: Michael S. Tsirkin
Cc: kys, haiyangz, davem, sridhar.samudrala, netdev,
Stephen Hemminger
In-Reply-To: <20180605211927-mutt-send-email-mst@kernel.org>
On Tue, 5 Jun 2018 21:35:26 +0300
"Michael S. Tsirkin" <mst@redhat.com> wrote:
> Thanks, I think this is nice patch but I wonder whether it can be split
> up somewhat. Not all of it is uncontroversial.
I started that way, but then I was fixing code that was later deleted.
The big change was eliminating the callbacks.
>
> On Mon, Jun 04, 2018 at 08:42:31PM -0700, Stephen Hemminger wrote:
> > * The matching of secondary device to primary device policy
> > is up to the network device. Both net_failover and netvsc
> > will use MAC for now but can change separately.
>
> I actually suspect both will change to a serial number
> down the road.
>
> > * The match policy is only used during initial discovery; after
> > that the secondary device knows what the upper device is because
> > of the parent/child relationship; no searching is required.
>
> That would obviously be an improvement - does it have to be tied with
> rest of changes?
This was not possible with the version of the common code that
is in net now.
>
> > * Now, netvsc and net_failover use the same delayed work type
> > mechanism for setup. Previously, net_failover code was triggering off
> > name change but a similar policy was rejected for netvsc.
> > "what is good for the goose is good for the gander"
>
> I don't really understand what you are saying here. I think the delayed
> hack is kind of ugly and seems racy. Current failover code was rejected
> by whom? Why is new one good and for whom? Did you want to do a name
> change in netvsc but it was rejected? Could you clarify please?
See:
https://patchwork.ozlabs.org/patch/851711/
>
> > * The net_failover private device info 'struct net_failover_info'
> > should have been private to the driver file, not a visible
> > API.
> >
> > * The net_failover device should use SET_NETDEV_DEV
> > that is intended only for physical devices not virtual devices.
>
> You mean should not.
Yes. Virtual device should not set device parent.
>
> > * No point in having DocBook style comments on a driver file.
> > They only make sense on an external exposed API.
> >
> > * net_failover only supports Ethernet, so use ether_addr_copy.
>
> It is since you need to know about all the things you need to copy, and
> because of mac matching. But it isn't too much effort to add more
> transports and I don't see value in going in the reverse direction and
> making it more ethernet specific that it already is.
Sure, then do memcpy base on addr_len
>
> > * Set permanent and current address of net_failover device
> > to match the primary.
> >
> > * Carrier should be marked off before registering device
> > the net_failover device.
>
> Are above two bugfixes?
Yes.
>
> > * Use netdev_XXX for log messages, in net_failover (not dev_xxx)
> >
> > * Since failover infrastructure is about linking devices just
> > use RTNL no need for other locking in init and teardown.
> >
> > * Don't bother with ERR_PTR() style return if only possible
> > return is success or no memory.
> >
> > * As much as possible, the terms master and slave should be avoided
> > because of their cultural connotations.
>
> Also for consistency, failover is calling these primary and standby now.
Good, let's standardize on that.
>
> > Note; this code has been tested on Hyper-V
> > but is compile tested only on virtio.
> >
> > Fixes: 30c8bd5aa8b2 ("net: Introduce generic failover module")
> > Signed-off-by: Stephen Hemminger <sthemmin@microsoft.com>
> > ---
> >
> > Although this patch needs to go into 4.18 (linux-net),
>
> I'd rather we focused on fixing bugs in 4.18, and left refactoring to
> 4.19.
>
Either we fix or revert the current code in 4.18.
Sorry, I am not having callback hell code in any vendor or upstream kernel.
^ permalink raw reply
* Re: [PATCH V4] mlx4_core: allocate ICM memory in page size chunks
From: Qing Huang @ 2018-06-05 18:51 UTC (permalink / raw)
To: Vlastimil Babka, Michal Hocko
Cc: Eric Dumazet, David Miller, tariqt, haakon.bugge, yanjun.zhu,
netdev, linux-rdma, linux-kernel, gi-oh.kim,
santosh.shilimkar@oracle.com, rama nichanamatlu
In-Reply-To: <c67934a5-8858-0de7-a7b5-532d14b5c881@suse.cz>
On 6/4/2018 5:40 AM, Vlastimil Babka wrote:
> On 06/04/2018 08:27 AM, Michal Hocko wrote:
>> On Fri 01-06-18 15:05:26, Qing Huang wrote:
>>>
>>> On 6/1/2018 12:31 AM, Michal Hocko wrote:
>>>> On Thu 31-05-18 19:04:46, Qing Huang wrote:
>>>>> On 5/31/2018 2:10 AM, Michal Hocko wrote:
>>>>>> On Thu 31-05-18 10:55:32, Michal Hocko wrote:
>>>>>>> On Thu 31-05-18 04:35:31, Eric Dumazet wrote:
>>>>>> [...]
>>>>>>>> I merely copied/pasted from alloc_skb_with_frags() :/
>>>>>>> I will have a look at it. Thanks!
>>>>>> OK, so this is an example of an incremental development ;).
>>>>>>
>>>>>> __GFP_NORETRY was added by ed98df3361f0 ("net: use __GFP_NORETRY for
>>>>>> high order allocations") to prevent from OOM killer. Yet this was
>>>>>> not enough because fb05e7a89f50 ("net: don't wait for order-3 page
>>>>>> allocation") didn't want an excessive reclaim for non-costly orders
>>>>>> so it made it completely NOWAIT while it preserved __GFP_NORETRY in
>>>>>> place which is now redundant. Should I send a patch?
>>>>>>
>>>>> Just curious, how about GFP_ATOMIC flag? Would it work in a similar fashion?
>>>>> We experimented
>>>>> with it a bit in the past but it seemed to cause other issue in our tests.
>>>>> :-)
>>>> GFP_ATOMIC is a non-sleeping (aka no reclaim) context with an access to
>>>> memory reserves. So the risk is that you deplete those reserves and
>>>> cause issues to other subsystems which need them as well.
>>>>
>>>>> By the way, we didn't encounter any OOM killer events. It seemed that the
>>>>> mlx4_alloc_icm() triggered slowpath.
>>>>> We still had about 2GB free memory while it was highly fragmented.
>>>> The compaction was able to make a reasonable forward progress for you.
>>>> But considering mlx4_alloc_icm is called with GFP_KERNEL resp. GFP_HIGHUSER
>>>> then the OOM killer is clearly possible as long as the order is lower
>>>> than 4.
>>> The allocation was 256KB so the order was much higher than 4. The compaction
>>> seemed to be the root
>>> cause for our problem. It took too long to finish its work while putting
>>> mlx4_alloc_icm to sleep in a heavily
>>> fragmented memory situation . Will NORETRY flag avoid the compaction ops and
>>> fail the 256KB allocation
>>> immediately so mlx4_alloc_icm can enter adjustable lower order allocation
>>> code path quickly?
>> Costly orders should only perform a light compaction attempt unless
>> __GFP_RETRY_MAY_FAIL is used IIRC. CCing Vlastimil. So __GFP_NORETRY
>> shouldn't make any difference.
> It's a bit more complicated. Costly allocations will try the light
> compaction attempt first, even before reclaim. This is followed by
> reclaim and a more costly compaction attempt. With __GFP_NORETRY, the
> second compaction attempt is also only the light one, so the flag does
> make a difference here.
Thanks for the clarification!
Looks like our production kernel is kinda old, neither
__GFP_DIRECT_RECLAIM nor __GFP_NORETRY
has been used in __alloc_pages_slowpath() in our kernel.
^ permalink raw reply
* [PATCH 5/5] spi,can,char: add companion DT binding documentation
From: Mark Jonas @ 2018-06-05 18:44 UTC (permalink / raw)
To: Wolfgang Grandegger, Marc Kleine-Budde
Cc: linux-can, netdev, linux-kernel, hs, yi.zhu5, Mark Jonas
In-Reply-To: <1528224240-30786-1-git-send-email-mark.jonas@de.bosch.com>
From: Zhu Yi <yi.zhu5@cn.bosch.com>
Signed-off-by: Zhu Yi <yi.zhu5@cn.bosch.com>
Signed-off-by: Mark Jonas <mark.jonas@de.bosch.com>
---
.../devicetree/bindings/spi/bosch,companion.txt | 82 ++++++++++++++++++++++
1 file changed, 82 insertions(+)
create mode 100644 Documentation/devicetree/bindings/spi/bosch,companion.txt
diff --git a/Documentation/devicetree/bindings/spi/bosch,companion.txt b/Documentation/devicetree/bindings/spi/bosch,companion.txt
new file mode 100644
index 0000000..5ded325
--- /dev/null
+++ b/Documentation/devicetree/bindings/spi/bosch,companion.txt
@@ -0,0 +1,82 @@
+Bosch Companion SPI slave device
+
+The functionality bases on an external peripheral chip named Companion.
+It offers two CAN interfaces, each has 8 prioritized transmit FIFOs as
+well as one receive FIFO. Besides CAN, undisclosed additional functions
+can be accessed through the char device.
+
+A standard SPI interface with two additional lines for flow control is
+used. The Companion chip is the SPI slave.
+
+The driver suite consists of three separate drivers. The following
+diagram illustrates the dependencies in layers.
+
+ /dev/companion SocketCAN User Space
+-------------------------------------------------------------------
+ +----------------+ +---------------+
+ | companion-char | | companion-can |
+ +----------------+ +---------------+
+ +----------------------------------+
+ | companion-spi |
+ +----------------------------------+
+ +----------------------------------+
+ | standard SPI subsystem |
+ +----------------------------------+ Linux Kernel
+-------------------------------------------------------------------
+ | | | | | | Hardware
+ CS-+ | | | | +-BUSY
+ CLK--+ | | +---REQUEST
+ MOSI---+ |
+ MISO-----+
+
+Required properties:
+
+- compatible : must be "bosch,companion-spi"
+- interrupt-parent : the phandle of the GPIO controller
+- interrupts : (GPIO) interrupt to which 'request-gpios' is
+ connected to
+- request-gpios : GPIO pin to request SPI master to receive data
+- busy-gpios : GPIO pin to indicate SPI slave is busy
+- cs-gpios : GPIO pin to select SPI slave
+
+Optional properties:
+
+The controller supports at most 2 CAN and 1 char device subnodes. When
+optionally specify the subnodes, the following properties are required:
+
+- CAN subnode
+ - compatible : must be "bosch,companion-can"
+ - clock-frequency: CAN device clock in Hz
+ - port : must be 0 or 1
+
+- Char device subnode
+ - compatible : must be "bosch,companion-char"
+
+Example:
+
+&ecspi1 {
+ companion-spi@0 {
+ compatible = "bosch,companion-spi";
+ interrupt-parent = <&gpio1>;
+ interrupts = <26 IRQ_TYPE_EDGE_FALLING>;
+ request-gpios = <&gpio1 26 GPIO_ACTIVE_LOW>;
+ busy-gpios = <&gpio1 27 GPIO_ACTIVE_LOW>;
+ cs-gpios = <&gpio4 9 GPIO_ACTIVE_LOW>;
+
+ companion-can0 {
+ compatible = "bosch,companion-can";
+ clock-frequency = <28000000>;
+ port = <0>;
+ };
+
+ companion-can1 {
+ compatible = "bosch,companion-can";
+ clock-frequency = <28000000>;
+ port = <1>;
+ };
+
+ companion-char {
+ compatible = "bosch,companion-char";
+ };
+ };
+};
--
2.7.4
^ permalink raw reply related
* [PATCH 4/5] can: implement companion-can driver
From: Mark Jonas @ 2018-06-05 18:43 UTC (permalink / raw)
To: Wolfgang Grandegger, Marc Kleine-Budde
Cc: linux-can, netdev, linux-kernel, hs, yi.zhu5, Mark Jonas
In-Reply-To: <1528224240-30786-1-git-send-email-mark.jonas@de.bosch.com>
From: Zhu Yi <yi.zhu5@cn.bosch.com>
The upper level companion-can driver provides SocketCAN interface to
userspace for communicate CAN messages with the companion processor.
Signed-off-by: Zhu Yi <yi.zhu5@cn.bosch.com>
Signed-off-by: Mark Jonas <mark.jonas@de.bosch.com>
---
drivers/net/can/Kconfig | 8 +
drivers/net/can/Makefile | 1 +
drivers/net/can/companion-can.c | 694 ++++++++++++++++++++++++++++++++++++++++
3 files changed, 703 insertions(+)
create mode 100644 drivers/net/can/companion-can.c
diff --git a/drivers/net/can/Kconfig b/drivers/net/can/Kconfig
index ac4ff39..e403a7e 100644
--- a/drivers/net/can/Kconfig
+++ b/drivers/net/can/Kconfig
@@ -155,6 +155,14 @@ config PCH_CAN
is an IOH for x86 embedded processor (Intel Atom E6xx series).
This driver can access CAN bus.
+config COMPANION_CAN
+ tristate "Network device for companion communication (Bosch)"
+ depends on COMPANION_SPI
+ ---help---
+ The network device allows the userspace to use SocketCAN interface
+ to communicate with the Bosch companion processor via the companion
+ SPI driver.
+
source "drivers/net/can/c_can/Kconfig"
source "drivers/net/can/cc770/Kconfig"
source "drivers/net/can/ifi_canfd/Kconfig"
diff --git a/drivers/net/can/Makefile b/drivers/net/can/Makefile
index 02b8ed7..a66a1f9 100644
--- a/drivers/net/can/Makefile
+++ b/drivers/net/can/Makefile
@@ -34,5 +34,6 @@ obj-$(CONFIG_CAN_SUN4I) += sun4i_can.o
obj-$(CONFIG_CAN_TI_HECC) += ti_hecc.o
obj-$(CONFIG_CAN_XILINXCAN) += xilinx_can.o
obj-$(CONFIG_PCH_CAN) += pch_can.o
+obj-$(CONFIG_COMPANION_CAN) += companion-can.o
subdir-ccflags-$(CONFIG_CAN_DEBUG_DEVICES) += -DDEBUG
diff --git a/drivers/net/can/companion-can.c b/drivers/net/can/companion-can.c
new file mode 100644
index 0000000..5078640
--- /dev/null
+++ b/drivers/net/can/companion-can.c
@@ -0,0 +1,694 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Companion upper level can network device
+ *
+ * Copyright (C) 2015-2018 Bosch Sicherheitssysteme GmbH
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#include <linux/can/dev.h>
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/netdevice.h>
+#include <linux/of.h>
+#include <linux/of_device.h>
+#include <linux/companion.h>
+
+#define TX_QUEUE_DEPTH 16
+#define NUM_TX_QUEUES 8
+#define NUM_RX_QUEUES 1
+#define TX_ECHO_SKB_MAX NUM_TX_QUEUES * TX_QUEUE_DEPTH
+#define DRIVER_NAME "bosch,companion-can"
+
+/**
+ * struct companion_can_priv - companion-can private data structure
+ * @can: standard common CAN private data, must be first member
+ * @parent: address of the associated parent device
+ * @dev: address of the associated network device
+ * @port: the companion CAN port number
+ * @tx_head: array of all tx queue head
+ * @tx_tail: arrat of all tx queue tail
+ */
+struct companion_can_priv {
+ struct can_priv can;
+ struct device *parent;
+ struct net_device *dev;
+ u8 port;
+ u8 tx_head[NUM_TX_QUEUES];
+ u8 tx_tail[NUM_TX_QUEUES];
+};
+
+/**
+ * companion_can_put_echo_skb() - put echo skb into ring buffer
+ * @priv: address of companion-can private data
+ * @prio: which CAN queue to put
+ * @skb: address of the packet to put
+ */
+static void companion_can_put_echo_skb(struct companion_can_priv *priv,
+ u8 prio,
+ struct sk_buff *skb)
+{
+ u8 offset = prio * TX_QUEUE_DEPTH;
+ u8 index = priv->tx_head[prio] % TX_QUEUE_DEPTH;
+ can_put_echo_skb(skb, priv->dev, offset + index);
+ priv->tx_head[prio]++;
+}
+
+/**
+ * companion_can_get_echo_skb() - get echo skb from ring buffer
+ * @priv: address of companion-can private data
+ * @prio: which CAN queue to get
+ */
+static u8 companion_can_get_echo_skb(struct companion_can_priv *priv,
+ u8 prio)
+{
+ u8 offset, index, result = 0;
+
+ if (priv->tx_head[prio] != priv->tx_tail[prio]) {
+ offset = prio * TX_QUEUE_DEPTH;
+ index = priv->tx_tail[prio] % TX_QUEUE_DEPTH;
+ result = can_get_echo_skb(priv->dev, offset + index);
+ priv->tx_tail[prio]++;
+ }
+ return result;
+}
+
+/**
+ * companion_can_free_echo_skb() - free echo skb from ring buffer
+ * @priv: address of companion-can private data
+ * @prio: which CAN queue to free
+ */
+static void companion_can_free_echo_skb(struct companion_can_priv *priv,
+ u8 prio)
+{
+ u8 offset, index;
+
+ if (priv->tx_head[prio] != priv->tx_tail[prio]) {
+ offset = prio * TX_QUEUE_DEPTH;
+ index = priv->tx_tail[prio] % TX_QUEUE_DEPTH;
+ can_free_echo_skb(priv->dev, offset + index);
+ priv->tx_tail[prio]++;
+ }
+}
+
+/**
+ * companion_can_set_bittiming() - set CAN bittiming
+ * @dev: address of the associated network device
+ */
+static int companion_can_set_bittiming(struct net_device *dev)
+{
+ struct companion_can_priv *priv = netdev_priv(dev);
+ const struct can_bittiming *bt = &priv->can.bittiming;
+ u32 ctrl = priv->can.ctrlmode;
+ int err;
+
+ err = companion_do_set_can_bittiming(priv->parent, priv->port, bt);
+ if (err)
+ return err;
+
+ if (ctrl & CAN_CTRLMODE_LISTENONLY) {
+ err = companion_do_set_can_ctrlmode(priv->parent,
+ priv->port,
+ ctrl);
+ if (err)
+ return err;
+ }
+ return 0;
+}
+
+/**
+ * companion_can_set_mode() - set CAN mode
+ * @dev: address of the associated network device
+ * @mode: the CAN mode to set
+ */
+static int companion_can_set_mode(struct net_device *dev,
+ enum can_mode mode)
+{
+ struct companion_can_priv *priv = netdev_priv(dev);
+ int err;
+
+ switch (mode) {
+ case CAN_MODE_START:
+ err = companion_can_set_bittiming(dev);
+ if (err)
+ return err;
+ /* fall through */
+
+ case CAN_MODE_STOP:
+ err = companion_do_set_can_mode(priv->parent,
+ priv->port,
+ mode);
+ if (err)
+ return err;
+ break;
+
+ default:
+ return -EOPNOTSUPP;
+ }
+ return 0;
+}
+
+/**
+ * companion_can_get_berr_counter() - get CAN error counter
+ * @dev: address of the associated network device
+ * @bec: address of the CAN error counter to store
+ */
+static int companion_can_get_berr_counter(const struct net_device *dev,
+ struct can_berr_counter *bec)
+{
+ struct companion_can_priv *priv = netdev_priv(dev);
+ return companion_do_get_can_status(priv->parent, priv->port, bec);
+}
+
+/**
+ * companion_can_handle_state() - handle CAN state transition
+ * @dev: address of the associated network device
+ * @cf: address of the CAN frame to store CAN state
+ * @bec: address of the CAN error counter
+ * @state: the companion CAN state
+ */
+static void companion_can_handle_state(struct net_device *dev,
+ struct can_frame *cf,
+ struct can_berr_counter *bec,
+ u8 state)
+{
+ struct companion_can_priv *priv = netdev_priv(dev);
+ enum can_state new_state = CAN_STATE_ERROR_ACTIVE;
+ enum can_state rx_state = CAN_STATE_ERROR_ACTIVE;
+ enum can_state tx_state = CAN_STATE_ERROR_ACTIVE;
+
+ if (state & COMPANION_CAN_STATE_BUS_OFF) {
+ new_state = CAN_STATE_BUS_OFF;
+ rx_state = bec->rxerr >= bec->txerr ? new_state : rx_state;
+ tx_state = bec->txerr >= bec->rxerr ? new_state : tx_state;
+ } else if (state & COMPANION_CAN_STATE_PASSIVE) {
+ new_state = CAN_STATE_ERROR_PASSIVE;
+ rx_state = bec->rxerr > 127 ? new_state : rx_state;
+ tx_state = bec->txerr > 127 ? new_state : tx_state;
+ } else if (state & COMPANION_CAN_STATE_WARNING) {
+ new_state = CAN_STATE_ERROR_WARNING;
+ rx_state = bec->rxerr >= bec->txerr ? new_state : rx_state;
+ tx_state = bec->txerr >= bec->rxerr ? new_state : tx_state;
+ }
+
+ if (new_state != priv->can.state) {
+ can_change_state(dev, cf, tx_state, rx_state);
+
+ if (new_state == CAN_STATE_BUS_OFF)
+ can_bus_off(dev);
+ }
+}
+
+/**
+ * companion_can_handle_error() - handle CAN error
+ * @dev: address of the associated network device
+ * @cf: address of the CAN frame to store CAN error
+ * @code: the companion CAN error code
+ */
+static void companion_can_handle_error(struct net_device *dev,
+ struct can_frame *cf,
+ u8 code)
+{
+ struct companion_can_priv *priv = netdev_priv(dev);
+
+ if (code & COMPANION_CAN_ERROR_RXOV) {
+ cf->can_id |= CAN_ERR_CRTL;
+ cf->data[1] |= CAN_ERR_CRTL_RX_OVERFLOW;
+ dev->stats.rx_over_errors++;
+ dev->stats.rx_errors++;
+ }
+
+ if (code & (COMPANION_CAN_ERROR_STUFF |
+ COMPANION_CAN_ERROR_FORM |
+ COMPANION_CAN_ERROR_ACK |
+ COMPANION_CAN_ERROR_BIT1 |
+ COMPANION_CAN_ERROR_BIT0 |
+ COMPANION_CAN_ERROR_CRC))
+ {
+ cf->can_id |= CAN_ERR_PROT | CAN_ERR_BUSERROR;
+
+ if (code & COMPANION_CAN_ERROR_STUFF) {
+ cf->data[2] |= CAN_ERR_PROT_STUFF;
+ dev->stats.rx_errors++;
+ }
+
+ if (code & COMPANION_CAN_ERROR_FORM) {
+ cf->data[2] |= CAN_ERR_PROT_FORM;
+ dev->stats.rx_errors++;
+ }
+
+ if (code & COMPANION_CAN_ERROR_ACK) {
+ cf->can_id |= CAN_ERR_ACK;
+ cf->data[3] = CAN_ERR_PROT_LOC_ACK;
+ dev->stats.tx_errors++;
+ }
+
+ if (code & COMPANION_CAN_ERROR_BIT1) {
+ cf->data[2] |= CAN_ERR_PROT_BIT1;
+ dev->stats.tx_errors++;
+ }
+
+ if (code & COMPANION_CAN_ERROR_BIT0) {
+ cf->data[2] |= CAN_ERR_PROT_BIT0;
+ dev->stats.tx_errors++;
+ }
+
+ if (code & COMPANION_CAN_ERROR_CRC) {
+ cf->data[2] |= CAN_ERR_PROT_BIT;
+ cf->data[3] = CAN_ERR_PROT_LOC_CRC_SEQ;
+ dev->stats.rx_errors++;
+ }
+
+ priv->can.can_stats.bus_error++;
+ }
+}
+
+/**
+ * companion_can_poll_err() - poll CAN error packet from companion
+ * @dev: address of the associated network device
+ */
+static bool companion_can_poll_err(struct net_device *dev)
+{
+ struct companion_can_priv *priv = netdev_priv(dev);
+ struct can_berr_counter bec;
+ u8 state;
+ u8 code;
+ struct sk_buff *skb;
+ struct can_frame *cf;
+
+ if (companion_do_can_err(priv->parent,
+ priv->port,
+ &bec,
+ &state,
+ &code) != 0)
+ return false;
+
+ skb = alloc_can_err_skb(dev, &cf);
+ if (!skb) {
+ dev_err(&dev->dev, "cannot alloc err skb\n");
+ return false;
+ }
+
+ companion_can_handle_state(dev, cf, &bec, state);
+ companion_can_handle_error(dev, cf, code);
+
+ dev->stats.rx_bytes += cf->can_dlc;
+ dev->stats.rx_packets++;
+ netif_rx(skb);
+ return true;
+}
+
+/**
+ * companion_can_poll_data() - poll CAN data packet from companion
+ * @dev: address of the associated network device
+ */
+static bool companion_can_poll_data(struct net_device *dev)
+{
+ struct companion_can_priv *priv = netdev_priv(dev);
+ struct sk_buff *skb;
+ struct can_frame *cf;
+
+ skb = alloc_can_skb(dev, &cf);
+ if (!skb) {
+ dev_err(&dev->dev, "cannot alloc rx skb\n");
+ dev->stats.rx_dropped++;
+ return false;
+ }
+
+ if (companion_do_can_rx(priv->parent, priv->port, cf) != 0) {
+ dev_kfree_skb_any(skb);
+ return false;
+ }
+
+ dev->stats.rx_bytes += cf->can_dlc;
+ dev->stats.rx_packets++;
+ netif_rx(skb);
+ can_led_event(dev, CAN_LED_EVENT_RX);
+ return true;
+}
+
+/**
+ * companion_can_on_tx_done() - CAN tx done callback
+ * @data: address of user supplied callback data
+ * @prio: which CAN queue is done
+ * @lost_seq_sync: flag indicate lost sequence happened
+ * @success: flag indicate last send is succeed or not
+ */
+static void companion_can_on_tx_done(void *data,
+ u8 prio,
+ bool lost_seq_sync,
+ bool success)
+{
+ struct companion_can_priv *priv = data;
+ struct net_device *dev = priv->dev;
+ struct net_device_stats *stats = &dev->stats;
+ int err;
+
+ if (success) {
+ stats->tx_bytes += companion_can_get_echo_skb(priv, prio);
+ stats->tx_packets++;
+ can_led_event(dev, CAN_LED_EVENT_TX);
+ } else {
+ companion_can_free_echo_skb(priv, prio);
+ dev_err(&dev->dev, "on_tx_done(%d) failed\n", prio);
+ }
+
+ /*TODO: what else action should take in case lost sequence?*/
+ if (lost_seq_sync)
+ dev_err(&dev->dev, "txq[%d] lost sequence sync\n", prio);
+
+ err = companion_do_can_stop_tx_timer(priv->parent, priv->port, prio);
+ if (err)
+ dev_err(&dev->dev,
+ "stop txq[%d] tx timer failed: %d\n",
+ prio, err);
+
+ netif_wake_subqueue(dev, prio);
+}
+
+/**
+ * companion_can_on_rx_done() - CAN rx done callback
+ * @data: address of user supplied callback data
+ */
+static void companion_can_on_rx_done(void *data)
+{
+ struct companion_can_priv *priv = data;
+ while (companion_can_poll_data(priv->dev));
+}
+
+/**
+ * companion_can_on_error() - CAN error callback
+ * @data: address of user supplied callback data
+ */
+static void companion_can_on_error(void *data)
+{
+ struct companion_can_priv *priv = data;
+ while (companion_can_poll_err(priv->dev));
+}
+
+/**
+ * companion_can_on_tx_timeout() - CAN tx timeout callback
+ * @data: address of user supplied callback data
+ * @prio: which CAN queue tx timed out
+ */
+static void companion_can_on_tx_timeout(void *data, u8 prio)
+{
+ struct companion_can_priv *priv = data;
+ bool lost_txq_sync = false;
+ int err;
+
+ err = companion_do_get_can_txq_status(priv->parent,
+ priv->port,
+ prio,
+ &lost_txq_sync);
+ if (err) {
+ dev_err(&priv->dev->dev,
+ "get can txq[%d] status failed: %d\n", prio, err);
+
+ if (err != -EINVAL)
+ companion_do_can_start_tx_timer(priv->parent,
+ priv->port,
+ prio);
+ return;
+ }
+
+ if (lost_txq_sync) {
+ dev_err(&priv->dev->dev,
+ "txq[%d] out of sync, restart data flow\n", prio);
+ companion_can_free_echo_skb(priv, prio);
+ netif_wake_subqueue(priv->dev, prio);
+ } else {
+ dev_err(&priv->dev->dev,
+ "txq[%d] is sync'd, but no ack, wait again\n", prio);
+ companion_do_can_start_tx_timer(priv->parent, priv->port, prio);
+ }
+}
+
+static struct companion_can_ops companion_can_can_ops = {
+ .on_tx_done = companion_can_on_tx_done,
+ .on_rx_done = companion_can_on_rx_done,
+ .on_error = companion_can_on_error,
+ .on_tx_timeout = companion_can_on_tx_timeout,
+};
+
+/**
+ * companion_can_open() - ndo_open callback
+ * @dev: address of the associated network device
+ */
+static int companion_can_open(struct net_device *dev)
+{
+ struct companion_can_priv *priv = netdev_priv(dev);
+ bool has_space = false;
+ int err, i;
+
+ err = companion_can_ops_register(priv->parent,
+ priv->port,
+ &companion_can_can_ops,
+ priv);
+ if (err) {
+ dev_err(&dev->dev,
+ "companion_can_ops_register() failed: %d\n", err);
+ goto out;
+ }
+
+ err = companion_can_set_mode(dev, CAN_MODE_START);
+ if (err) {
+ dev_err(&dev->dev,
+ "companion_can_set_mode() failed: %d\n", err);
+ goto out_register;
+ }
+
+ err = companion_do_get_can_txq_status_all(priv->parent, priv->port);
+ if (err) {
+ dev_err(&dev->dev,
+ "companion_do_get_can_txq_status_all() failed: %d\n",
+ err);
+ goto out_mode;
+ }
+
+ err = open_candev(dev);
+ if (err) {
+ dev_err(&dev->dev, "open_candev() failed: %d\n", err);
+ goto out_mode;
+ }
+
+ priv->can.state = CAN_STATE_ERROR_ACTIVE;
+ can_led_event(dev, CAN_LED_EVENT_OPEN);
+
+ /*TODO: start all here or start depends on queue space?*/
+ for (i = 0; i < NUM_TX_QUEUES; ++i) {
+ err = companion_do_can_txq_has_space(priv->parent,
+ priv->port,
+ i,
+ &has_space);
+
+ if (!err && has_space) {
+ netif_tx_start_queue(netdev_get_tx_queue(dev, i));
+ } else {
+ netif_tx_stop_queue(netdev_get_tx_queue(dev, i));
+ dev_err(&dev->dev, "txq[%d] is not started\n", i);
+ }
+ }
+
+ return 0;
+
+out_mode:
+ companion_can_set_mode(dev, CAN_MODE_STOP);
+out_register:
+ companion_can_ops_unregister(priv->parent, priv->port);
+out:
+ return err;
+}
+
+/**
+ * companion_can_release() - ndo_close callback
+ * @dev: address of the associated network device
+ */
+static int companion_can_release(struct net_device *dev)
+{
+ struct companion_can_priv *priv = netdev_priv(dev);
+ int result;
+
+ netif_tx_stop_all_queues(dev);
+ can_led_event(dev, CAN_LED_EVENT_STOP);
+ priv->can.state = CAN_STATE_STOPPED;
+ close_candev(dev);
+ result = companion_can_set_mode(dev, CAN_MODE_STOP);
+ companion_can_ops_unregister(priv->parent, priv->port);
+ return result;
+}
+
+/**
+ * companion_can_start_xmit() - ndo_start_xmit callback
+ * @skb: address of the packet to send
+ * @dev: address of the associated network device
+ */
+static int companion_can_start_xmit(struct sk_buff *skb,
+ struct net_device *dev)
+{
+ struct companion_can_priv *priv = netdev_priv(dev);
+ struct can_frame *cf = (struct can_frame*)skb->data;
+ u16 prio = skb_get_queue_mapping(skb);
+ bool is_full = false;
+ int err;
+
+ if (can_dropped_invalid_skb(dev, skb)) {
+ dev_err(&dev->dev, "dropped invalid skb on txq[%d]\n", prio);
+ return NETDEV_TX_OK;
+ }
+
+ err = companion_do_can_tx(priv->parent, priv->port, prio, cf);
+ if (err) {
+ dev_err(&dev->dev, "dropped packet on txq[%d]\n", prio);
+ dev_kfree_skb_any(skb);
+ dev->stats.tx_dropped++;
+ return NETDEV_TX_OK;
+ }
+
+ err = companion_do_can_txq_is_full(priv->parent,
+ priv->port,
+ prio,
+ &is_full);
+ if (!err && is_full) {
+ netif_stop_subqueue(dev, prio);
+ err = companion_do_can_start_tx_timer(priv->parent,
+ priv->port,
+ prio);
+ if (err)
+ dev_err(&dev->dev,
+ "start txq[%d] tx timer failed: %d\n",
+ prio, err);
+ }
+
+ companion_can_put_echo_skb(priv, prio, skb);
+ return NETDEV_TX_OK;
+}
+
+static const struct net_device_ops companion_can_netdev_ops = {
+ .ndo_open = companion_can_open,
+ .ndo_stop = companion_can_release,
+ .ndo_start_xmit = companion_can_start_xmit,
+};
+
+static const struct of_device_id companion_can_of_match[] = {
+ { .compatible = DRIVER_NAME, .data = NULL, },
+ { /* sentinel */ },
+};
+MODULE_DEVICE_TABLE(of, companion_can_of_match);
+
+static const struct can_bittiming_const companion_can_bittiming_const = {
+ .name = "bosch,companion",
+ .tseg1_min = 2,
+ .tseg1_max = 16,
+ .tseg2_min = 1,
+ .tseg2_max = 8,
+ .sjw_max = 4,
+ .brp_min = 1,
+ .brp_max = 1024,
+ .brp_inc = 1,
+};
+
+/**
+ * companion_can_probe() - probe callback
+ * @pdev: address of the platform device
+ */
+static int companion_can_probe(struct platform_device *pdev)
+{
+ struct device_node *node = pdev->dev.of_node;
+ struct net_device *dev;
+ struct companion_can_priv *priv;
+ u32 port, freq;
+ int err;
+
+ if (!node) {
+ dev_err(&pdev->dev, "no device tree data\n");
+ return -ENODEV;
+ }
+
+ if (of_property_read_u32(node, "port", &port)) {
+ dev_err(&pdev->dev, "no port property\n");
+ return -ENODEV;
+ }
+
+ if ((port != 0) && (port != 1)) {
+ dev_err(&pdev->dev,
+ "invalid port %d, valid range is [0,1]\n", port);
+ return -EINVAL;
+ }
+
+ if (of_property_read_u32(node, "clock-frequency", &freq)) {
+ dev_err(&pdev->dev, "no clock-frequency property\n");
+ return -ENODEV;
+ }
+
+ if (!pdev->dev.parent) {
+ dev_err(&pdev->dev, "no parent device\n");
+ return -ENODEV;
+ }
+
+ dev = alloc_candev_mqs(sizeof(*priv),
+ TX_ECHO_SKB_MAX,
+ NUM_TX_QUEUES,
+ NUM_RX_QUEUES);
+ if (!dev)
+ return -ENOMEM;
+
+ dev->netdev_ops = &companion_can_netdev_ops;
+ dev->flags |= IFF_ECHO;
+ dev->real_num_tx_queues = NUM_TX_QUEUES;
+
+ priv = netdev_priv(dev);
+ priv->can.clock.freq = freq;
+ priv->can.bittiming_const = &companion_can_bittiming_const;
+ priv->can.do_set_mode = companion_can_set_mode;
+ priv->can.do_get_berr_counter = companion_can_get_berr_counter;
+ priv->can.ctrlmode_supported = CAN_CTRLMODE_LISTENONLY |
+ CAN_CTRLMODE_BERR_REPORTING;
+ priv->parent = pdev->dev.parent;
+ priv->dev = dev;
+ priv->port = port;
+
+ platform_set_drvdata(pdev, dev);
+ SET_NETDEV_DEV(dev, &pdev->dev);
+
+ err = register_candev(dev);
+ if (err) {
+ dev_err(&pdev->dev, "register_candev() failed: %d\n", err);
+ free_candev(dev);
+ return err;
+ }
+
+ devm_can_led_init(dev);
+ return 0;
+}
+
+/**
+ * companion_can_remove() - remove callback
+ * @pdev: address of the platform device
+ */
+static int companion_can_remove(struct platform_device *pdev)
+{
+ struct net_device *dev = platform_get_drvdata(pdev);
+
+ unregister_candev(dev);
+ free_candev(dev);
+ return 0;
+}
+
+static struct platform_driver companion_can_driver = {
+ .driver = {
+ .name = DRIVER_NAME,
+ .owner = THIS_MODULE,
+ .of_match_table = of_match_ptr(companion_can_of_match),
+ },
+ .probe = companion_can_probe,
+ .remove = companion_can_remove,
+};
+module_platform_driver(companion_can_driver);
+
+MODULE_AUTHOR("Zhu Yi <yi.zhu5@cn.bosch.com>");
+MODULE_DESCRIPTION("Companion upper level can network device");
+MODULE_LICENSE("GPL v2");
--
2.7.4
^ permalink raw reply related
* [PATCH 3/5] char: implement companion-char driver
From: Mark Jonas @ 2018-06-05 18:43 UTC (permalink / raw)
To: Wolfgang Grandegger, Marc Kleine-Budde
Cc: linux-can, netdev, linux-kernel, hs, yi.zhu5, Mark Jonas
In-Reply-To: <1528224240-30786-1-git-send-email-mark.jonas@de.bosch.com>
From: Zhu Yi <yi.zhu5@cn.bosch.com>
The upper level companion-char driver provides character device
interface to userspace for communicate IO messages with the
companion processor.
Signed-off-by: Zhu Yi <yi.zhu5@cn.bosch.com>
Signed-off-by: Mark Jonas <mark.jonas@de.bosch.com>
---
drivers/char/Kconfig | 7 +
drivers/char/Makefile | 2 +
drivers/char/companion-char.c | 367 ++++++++++++++++++++++++++++++++++++++++++
3 files changed, 376 insertions(+)
create mode 100644 drivers/char/companion-char.c
diff --git a/drivers/char/Kconfig b/drivers/char/Kconfig
index c28dca0..e878d56 100644
--- a/drivers/char/Kconfig
+++ b/drivers/char/Kconfig
@@ -588,5 +588,12 @@ config TILE_SROM
source "drivers/char/xillybus/Kconfig"
+config COMPANION_CHAR
+ tristate "Character device for companion communication (Bosch)"
+ depends on COMPANION_SPI
+ help
+ The character device allows the userspace to exchange IO messages
+ with the Bosch companion processor via the companion SPI driver.
+
endmenu
diff --git a/drivers/char/Makefile b/drivers/char/Makefile
index 7dc3abe..dfe4fc1 100644
--- a/drivers/char/Makefile
+++ b/drivers/char/Makefile
@@ -60,3 +60,5 @@ js-rtc-y = rtc.o
obj-$(CONFIG_TILE_SROM) += tile-srom.o
obj-$(CONFIG_XILLYBUS) += xillybus/
obj-$(CONFIG_POWERNV_OP_PANEL) += powernv-op-panel.o
+
+obj-$(CONFIG_COMPANION_CHAR) += companion-char.o
diff --git a/drivers/char/companion-char.c b/drivers/char/companion-char.c
new file mode 100644
index 0000000..3c198f2
--- /dev/null
+++ b/drivers/char/companion-char.c
@@ -0,0 +1,367 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Companion upper level character device
+ *
+ * Copyright (C) 2015-2018 Bosch Sicherheitssysteme GmbH
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#include <linux/cdev.h>
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include <linux/platform_device.h>
+#include <linux/poll.h>
+#include <linux/companion.h>
+
+#define DRIVER_NAME "bosch,companion-char"
+
+/*TODO: get from protocol.h*/
+#define COMPANION_PACKET_SIZE 16
+
+static struct class *companion_char_class;
+static dev_t devt;
+
+/**
+ * struct companion_char_minor - companion-char minor structure
+ * @dev: address of the associated device
+ * @writelock: mutex to protect write
+ * @readlock: mutex to protect read
+ * @writewait: wait queue head of write
+ * @readwait: wait queue head of read
+ */
+struct companion_char_minor {
+ struct device *dev;
+ struct mutex writelock;
+ struct mutex readlock;
+ wait_queue_head_t writewait;
+ wait_queue_head_t readwait;
+};
+
+/**
+ * struct companion_char_priv - companion-char private data structure
+ * @cdev: char device
+ * @parent: address of the associated parent device
+ * @minors: address of the companion-char minor
+ */
+struct companion_char_priv {
+ struct cdev cdev;
+ struct device *parent;
+ struct companion_char_minor *minors;
+};
+
+/**
+ * companion_char_read() - read callback
+ * @filp: address of the associated virtual file
+ * @buf: address of the user space buffer to receive
+ * @count: number of bytes to read
+ * @offset: address of the read offset
+ */
+static ssize_t companion_char_read(struct file *filp,
+ char __user *buf,
+ size_t count,
+ loff_t *offset)
+{
+ unsigned int number = MINOR(file_inode(filp)->i_rdev);
+ struct companion_char_priv *priv = filp->private_data;
+ struct companion_char_minor *minor = &priv->minors[number];
+ int status;
+
+ if (count != COMPANION_PACKET_SIZE)
+ return -EMSGSIZE;
+
+ if (mutex_lock_interruptible(&minor->readlock))
+ return -ERESTARTSYS;
+
+ while (companion_io_rxq_is_empty(priv->parent)) {
+ mutex_unlock(&minor->readlock);
+ if (filp->f_flags & O_NONBLOCK)
+ return -EAGAIN;
+ if (wait_event_interruptible(minor->readwait,
+ !companion_io_rxq_is_empty(priv->parent)))
+ return -ERESTARTSYS;
+ if (mutex_lock_interruptible(&minor->readlock))
+ return -ERESTARTSYS;
+ }
+
+ status = companion_do_io_rx(priv->parent, buf, count);
+ mutex_unlock(&minor->readlock);
+ return status;
+}
+
+/**
+ * companion_char_write() - write callback
+ * @filp: address of the associated virtual file
+ * @buf: address of the user space buffer to transfer
+ * @count: number of bytes to write
+ * @offset: address of the write offset
+ */
+static ssize_t companion_char_write(struct file *filp,
+ const char __user *buf,
+ size_t count,
+ loff_t *offset)
+{
+ unsigned int number = MINOR(file_inode(filp)->i_rdev);
+ struct companion_char_priv *priv = filp->private_data;
+ struct companion_char_minor *minor = &priv->minors[number];
+ int status;
+
+ if (count != COMPANION_PACKET_SIZE)
+ return -EMSGSIZE;
+
+ if (mutex_lock_interruptible(&minor->writelock))
+ return -ERESTARTSYS;
+
+ while (companion_io_txq_is_full(priv->parent)) {
+ mutex_unlock(&minor->writelock);
+ if (filp->f_flags & O_NONBLOCK)
+ return -EAGAIN;
+ if (wait_event_interruptible(minor->writewait,
+ !companion_io_txq_is_full(priv->parent)))
+ return -ERESTARTSYS;
+ if (mutex_lock_interruptible(&minor->writelock))
+ return -ERESTARTSYS;
+ }
+
+ status = companion_do_io_tx(priv->parent, buf, count);
+ mutex_unlock(&minor->writelock);
+ return status;
+}
+
+/**
+ * companion_char_poll() - poll callback
+ * @filp: address of the associated virtual file
+ * @wait: address of the associated poll table
+ */
+static unsigned int companion_char_poll(struct file *filp, poll_table *wait)
+{
+ unsigned int number = MINOR(file_inode(filp)->i_rdev);
+ struct companion_char_priv *priv = filp->private_data;
+ struct companion_char_minor *minor = &priv->minors[number];
+ unsigned int mask = 0;
+
+ poll_wait(filp, &minor->writewait, wait);
+ poll_wait(filp, &minor->readwait, wait);
+
+ mutex_lock(&minor->writelock);
+ if (!companion_io_txq_is_full(priv->parent))
+ mask |= POLLOUT | POLLWRNORM;
+ mutex_unlock(&minor->writelock);
+
+ mutex_lock(&minor->readlock);
+ if (!companion_io_rxq_is_empty(priv->parent))
+ mask |= POLLIN | POLLRDNORM;
+ mutex_unlock(&minor->readlock);
+
+ return mask;
+}
+
+/**
+ * companion_char_open() - open callback
+ * @inode: address of the associated inode
+ * @filp: address of the associated virtual file
+ */
+static int companion_char_open(struct inode *inode, struct file *filp)
+{
+ struct companion_char_priv *priv = container_of(
+ inode->i_cdev,
+ struct companion_char_priv,
+ cdev);
+ filp->private_data = priv;
+ nonseekable_open(inode, filp);
+ return 0;
+}
+
+/**
+ * companion_char_release() - release callback
+ * @inode: address of the associated inode
+ * @filp: address of the associated virtual file
+ */
+static int companion_char_release(struct inode *inode, struct file *filp)
+{
+ filp->private_data = NULL;
+ return 0;
+}
+
+static const struct file_operations companion_char_ops = {
+ .owner = THIS_MODULE,
+ .llseek = no_llseek,
+ .read = companion_char_read,
+ .write = companion_char_write,
+ .poll = companion_char_poll,
+ .open = companion_char_open,
+ .release = companion_char_release,
+};
+
+/**
+ * companion_char_on_tx_done() - tx done callback
+ * @data: address of user supplied callback data
+ */
+static void companion_char_on_tx_done(void *data)
+{
+ struct companion_char_priv *priv = data;
+ struct companion_char_minor *minor = &priv->minors[0];
+
+ wake_up_interruptible(&minor->writewait);
+}
+
+/**
+ * companion_char_on_rx_done() - rx done callback
+ * @data: address of user supplied callback data
+ */
+static void companion_char_on_rx_done(void *data)
+{
+ struct companion_char_priv *priv = data;
+ struct companion_char_minor *minor = &priv->minors[0];
+
+ wake_up_interruptible(&minor->readwait);
+}
+
+static struct companion_io_ops companion_char_io_ops = {
+ .on_tx_done = companion_char_on_tx_done,
+ .on_rx_done = companion_char_on_rx_done,
+};
+
+/**
+ * companion_char_probe() - probe callback
+ * @pdev: address of the platform device
+ */
+static int companion_char_probe(struct platform_device *pdev)
+{
+ struct companion_char_priv *priv;
+ struct companion_char_minor *minors;
+ int err;
+
+ if (!pdev->dev.parent) {
+ dev_err(&pdev->dev, "no parent device found\n");
+ return -ENODEV;
+ }
+
+ priv = devm_kzalloc(&pdev->dev, sizeof(*priv), GFP_KERNEL);
+ if (!priv)
+ return -ENOMEM;
+ priv->parent = pdev->dev.parent;
+
+ minors = devm_kzalloc(&pdev->dev, sizeof(*minors), GFP_KERNEL);
+ if (!minors)
+ return -ENOMEM;
+
+ minors->dev = device_create(companion_char_class,
+ &pdev->dev,
+ MKDEV(MAJOR(devt), 0),
+ priv,
+ "companion%d",
+ 0);
+ if (IS_ERR_OR_NULL(minors->dev))
+ return PTR_ERR_OR_ZERO(minors->dev);
+ priv->minors = minors;
+
+ mutex_init(&minors->writelock);
+ mutex_init(&minors->readlock);
+ init_waitqueue_head(&minors->writewait);
+ init_waitqueue_head(&minors->readwait);
+
+ cdev_init(&priv->cdev, &companion_char_ops);
+ err = cdev_add(&priv->cdev, MKDEV(MAJOR(devt), 0), 1);
+ if (err) {
+ dev_err(&pdev->dev, "cdev_add() failed: %d\n", err);
+ goto on_error;
+ }
+
+ dev_set_drvdata(&pdev->dev, priv);
+
+ err = companion_io_ops_register(priv->parent,
+ &companion_char_io_ops,
+ priv);
+ if (err) {
+ dev_err(&pdev->dev, "companion_io_ops_register() failed: %d\n",
+ err);
+ goto on_error;
+ }
+ return 0;
+
+on_error:
+ device_destroy(companion_char_class, MKDEV(MAJOR(devt), 0));
+ cdev_del(&priv->cdev);
+ return err;
+
+}
+
+/**
+ * companion_char_remove() - remove callback
+ * @pdev: address of the platform device
+ */
+static int companion_char_remove(struct platform_device *pdev)
+{
+ struct companion_char_priv *priv = dev_get_drvdata(&pdev->dev);
+
+ companion_io_ops_unregister(priv->parent);
+ device_destroy(companion_char_class, MKDEV(MAJOR(devt), 0));
+ cdev_del(&priv->cdev);
+ return 0;
+}
+
+static const struct of_device_id companion_char_of_match[] = {
+ { .compatible = DRIVER_NAME, .data = NULL, },
+ { /* sentinel */ },
+};
+MODULE_DEVICE_TABLE(of, companion_char_of_match);
+
+static struct platform_driver companion_char_driver = {
+ .driver = {
+ .name = DRIVER_NAME,
+ .owner = THIS_MODULE,
+ .of_match_table = of_match_ptr(companion_char_of_match),
+ },
+ .probe = companion_char_probe,
+ .remove = companion_char_remove,
+};
+
+/**
+ * companion_char_init() - module init
+ */
+static int __init companion_char_init(void)
+{
+ int err;
+
+ companion_char_class = class_create(THIS_MODULE, DRIVER_NAME);
+ if (IS_ERR_OR_NULL(companion_char_class))
+ return PTR_ERR_OR_ZERO(companion_char_class);
+
+ err = alloc_chrdev_region(&devt, 0, 1, DRIVER_NAME);
+ if (err) {
+ printk("companion:alloc_chrdev_region() failed: %d\n", err);
+ class_destroy(companion_char_class);
+ return err;
+ }
+
+ err = platform_driver_register(&companion_char_driver);
+ if (err) {
+ printk("companion:platform_driver_register() failed: %d\n",
+ err);
+ class_destroy(companion_char_class);
+ unregister_chrdev_region(devt, 1);
+ }
+
+ return err;
+}
+
+/**
+ * companion_char_exit() - module exit
+ */
+static void __exit companion_char_exit(void)
+{
+ platform_driver_unregister(&companion_char_driver);
+ class_destroy(companion_char_class);
+ unregister_chrdev_region(devt, 1);
+}
+
+module_init(companion_char_init);
+module_exit(companion_char_exit);
+
+MODULE_AUTHOR("Zhu Yi <yi.zhu5@cn.bosch.com>");
+MODULE_DESCRIPTION("Companion upper level character device");
+MODULE_LICENSE("GPL v2");
--
2.7.4
^ permalink raw reply related
* [PATCH 2/5] spi: implement companion-spi driver
From: Mark Jonas @ 2018-06-05 18:43 UTC (permalink / raw)
To: Wolfgang Grandegger, Marc Kleine-Budde
Cc: linux-can, netdev, linux-kernel, hs, yi.zhu5, Mark Jonas
In-Reply-To: <1528224240-30786-1-git-send-email-mark.jonas@de.bosch.com>
From: Zhu Yi <yi.zhu5@cn.bosch.com>
The low level companion-spi driver encapsulates the communication
details with the companion processor, and provides interface for
the upper level drivers to access.
Signed-off-by: Zhu Yi <yi.zhu5@cn.bosch.com>
Signed-off-by: Mark Jonas <mark.jonas@de.bosch.com>
---
drivers/spi/Kconfig | 2 +
drivers/spi/Makefile | 2 +
drivers/spi/companion/Kconfig | 5 +
drivers/spi/companion/Makefile | 2 +
drivers/spi/companion/core.c | 1189 ++++++++++++++++++++++++++++++
drivers/spi/companion/protocol-manager.c | 1035 ++++++++++++++++++++++++++
drivers/spi/companion/protocol-manager.h | 348 +++++++++
drivers/spi/companion/protocol.h | 273 +++++++
drivers/spi/companion/queue-manager.c | 146 ++++
drivers/spi/companion/queue-manager.h | 245 ++++++
include/linux/companion.h | 258 +++++++
11 files changed, 3505 insertions(+)
create mode 100644 drivers/spi/companion/Kconfig
create mode 100644 drivers/spi/companion/Makefile
create mode 100644 drivers/spi/companion/core.c
create mode 100644 drivers/spi/companion/protocol-manager.c
create mode 100644 drivers/spi/companion/protocol-manager.h
create mode 100644 drivers/spi/companion/protocol.h
create mode 100644 drivers/spi/companion/queue-manager.c
create mode 100644 drivers/spi/companion/queue-manager.h
create mode 100644 include/linux/companion.h
diff --git a/drivers/spi/Kconfig b/drivers/spi/Kconfig
index a75f2a2..8b575ec 100644
--- a/drivers/spi/Kconfig
+++ b/drivers/spi/Kconfig
@@ -799,6 +799,8 @@ config SPI_TLE62X0
# Add new SPI protocol masters in alphabetical order above this line
#
+source "drivers/spi/companion/Kconfig"
+
endif # SPI_MASTER
#
diff --git a/drivers/spi/Makefile b/drivers/spi/Makefile
index 8e0cda7..ae369d9 100644
--- a/drivers/spi/Makefile
+++ b/drivers/spi/Makefile
@@ -112,3 +112,5 @@ obj-$(CONFIG_SPI_ZYNQMP_GQSPI) += spi-zynqmp-gqspi.o
# SPI slave protocol handlers
obj-$(CONFIG_SPI_SLAVE_TIME) += spi-slave-time.o
obj-$(CONFIG_SPI_SLAVE_SYSTEM_CONTROL) += spi-slave-system-control.o
+
+obj-y += companion/
diff --git a/drivers/spi/companion/Kconfig b/drivers/spi/companion/Kconfig
new file mode 100644
index 0000000..490a273
--- /dev/null
+++ b/drivers/spi/companion/Kconfig
@@ -0,0 +1,5 @@
+config COMPANION_SPI
+ tristate "Low level driver for companion communication (Bosch)"
+ depends on SPI
+ help
+ This driver communicates with the companion processor via SPI.
diff --git a/drivers/spi/companion/Makefile b/drivers/spi/companion/Makefile
new file mode 100644
index 0000000..e60e733
--- /dev/null
+++ b/drivers/spi/companion/Makefile
@@ -0,0 +1,2 @@
+obj-$(CONFIG_COMPANION_SPI) += companion-spi.o
+companion-spi-objs := core.o protocol-manager.o queue-manager.o
diff --git a/drivers/spi/companion/core.c b/drivers/spi/companion/core.c
new file mode 100644
index 0000000..435b215
--- /dev/null
+++ b/drivers/spi/companion/core.c
@@ -0,0 +1,1189 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Companion low level init/core code
+ *
+ * Copyright (C) 2015-2018 Bosch Sicherheitssysteme GmbH
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#include <linux/delay.h>
+#include <linux/gpio.h>
+#include <linux/interrupt.h>
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include <linux/of_device.h>
+#include <linux/of_gpio.h>
+#include <linux/spi/spi.h>
+#include <linux/kfifo.h>
+
+#include "protocol-manager.h"
+
+#define DRIVER_NAME "bosch,companion-spi"
+
+#define READY_POLL_US 80
+#define READY_POLL_US_GRAN 1
+#define READY_POLL_MS 100
+#define READY_POLL_MS_GRAN 10
+
+/**
+ * struct busy_signal_statistics - spi busy signal statistics
+ * @while_busy_ext: how many times while_busy loop been waited
+ * @while_busy_fail: how many times while_busy been timed out
+ * @until_busy_ext: how many times until_busy loop been waited
+ * @until_busy_fail: how many times until_busy been timed out
+ * @force_started: how many times of force started
+ * @force_started_failure: how many times of force started failure
+ * @ready_failure: how many times of ready failure
+ */
+struct busy_signal_statistics {
+ u32 while_busy_ext;
+ u32 while_busy_fail;
+ u32 until_busy_ext;
+ u32 until_busy_fail;
+ u32 force_started;
+ u32 force_started_failure;
+ u32 ready_failure;
+};
+
+/**
+ * struct companion_spi_priv - companion-spi private data structure
+ * @spi: address of spi device
+ * @task: address of task struct
+ * @wait: wait queue head
+ * @request_gpios: gpio line connect to request signal
+ * @request_gpios_assert: polarity of request signal
+ * @busy_gpios: gpio line connect to busy signal
+ * @busy_gpios_assert: polarity of busy signal
+ * @cs_gpios: gpio line connect to cs signal
+ * @cs_gpios_assert: polarity of cs signal
+ * @dump_packet: flag to control dump spi packet
+ * @stats: spi busy signal statistics
+ * @pm: companion protocol manager
+ */
+struct companion_spi_priv {
+ struct spi_device *spi;
+ struct task_struct *task;
+ wait_queue_head_t wait;
+
+ u32 request_gpios;
+ u32 request_gpios_assert;
+ u32 busy_gpios;
+ u32 busy_gpios_assert;
+ u32 cs_gpios;
+ u32 cs_gpios_assert;
+
+ bool dump_packets;
+ struct busy_signal_statistics stats;
+ struct companion_protocol_manager pm;
+};
+
+/**
+ * companion_io_ops_register() - register companion IO packets handler
+ * @parent: address of the parent device
+ * @ops: address of the IO callbacks
+ * @data: address of the data passed to the IO callbacks
+ */
+int companion_io_ops_register(struct device *parent,
+ struct companion_io_ops *ops,
+ void *data)
+{
+ struct companion_spi_priv *priv = dev_get_drvdata(parent);
+ return pm_io_ops_register(&priv->pm, ops, data);
+}
+EXPORT_SYMBOL_GPL(companion_io_ops_register);
+
+/**
+ * companion_io_ops_unregister() - unregister companion IO packets handler
+ * @parent: address of the parent device
+ */
+int companion_io_ops_unregister(struct device *parent)
+{
+ struct companion_spi_priv *priv = dev_get_drvdata(parent);
+ return pm_io_ops_unregister(&priv->pm);
+}
+EXPORT_SYMBOL_GPL(companion_io_ops_unregister);
+
+/**
+ * companion_can_ops_register() - register companion CAN packets handler
+ * @parent: address of the parent device
+ * @port: which CAN port to register
+ * @ops: address of the CAN callbacks
+ * @data: address of the data passed to the CAN callbacks
+ */
+int companion_can_ops_register(struct device *parent,
+ u8 port,
+ struct companion_can_ops *ops,
+ void *data)
+{
+ struct companion_spi_priv *priv = dev_get_drvdata(parent);
+ return pm_can_ops_register(&priv->pm, port, ops, data);
+}
+EXPORT_SYMBOL_GPL(companion_can_ops_register);
+
+/**
+ * companion_can_ops_unregister() - unregister companion CAN packets handler
+ * @parent: address of the parent device
+ * @port: which CAN port to unregister
+ */
+int companion_can_ops_unregister(struct device *parent, u8 port)
+{
+ struct companion_spi_priv *priv = dev_get_drvdata(parent);
+ return pm_can_ops_unregister(&priv->pm, port);
+}
+EXPORT_SYMBOL_GPL(companion_can_ops_unregister);
+
+/**
+ * companion_io_txq_is_full() - return true if IO tx queue is full
+ * @parent: address of the parent device
+ */
+bool companion_io_txq_is_full(struct device *parent)
+{
+ struct companion_spi_priv *priv = dev_get_drvdata(parent);
+ return qm_io_txq_is_full(&priv->pm.qm);
+}
+EXPORT_SYMBOL_GPL(companion_io_txq_is_full);
+
+/**
+ * companion_io_rxq_is_empty() - return true if IO rx queue is empty
+ * @parent: address of the parent device
+ */
+bool companion_io_rxq_is_empty(struct device *parent)
+{
+ struct companion_spi_priv *priv = dev_get_drvdata(parent);
+ return qm_io_rxq_is_empty(&priv->pm.qm);
+}
+EXPORT_SYMBOL_GPL(companion_io_rxq_is_empty);
+
+/**
+ * companion_do_io_tx() - send IO packet
+ * @parent: address of the parent device
+ * @buf: address of the user space buffer to send
+ * @count: number of bytes to copy
+ */
+int companion_do_io_tx(struct device *parent,
+ const char __user *buf,
+ size_t count)
+{
+ struct companion_spi_priv *priv = dev_get_drvdata(parent);
+ unsigned int copied;
+ int error;
+ struct companion_packet p;
+
+ /*TODO: support mutiple packets in one write in future*/
+ if (copy_from_user(p.data, buf, sizeof(p)) == 0) {
+ if (is_can_type(&p))
+ return -EINVAL;
+ } else {
+ dev_info(parent, "copy from user not succeed in one call\n");
+ }
+
+ error = qm_io_txq_in(&priv->pm.qm, buf, count, &copied);
+ if (!error) {
+ wake_up_interruptible(&priv->wait);
+ priv->pm.stats.io_tx++;
+ return copied;
+ } else {
+ priv->pm.stats.io_tx_overflows++;
+ }
+ return error;
+}
+EXPORT_SYMBOL_GPL(companion_do_io_tx);
+
+/**
+ * companion_do_io_rx() - receive IO packet
+ * @parent: address of the parent device
+ * @buf: address of the user space buffer to receive
+ * @count: number of bytes to copy
+ */
+int companion_do_io_rx(struct device *parent,
+ char __user *buf,
+ size_t count)
+{
+ struct companion_spi_priv *priv = dev_get_drvdata(parent);
+ unsigned int copied;
+ int error;
+
+ error = qm_io_rxq_out(&priv->pm.qm, buf, count, &copied);
+ return error ? error : copied;
+}
+EXPORT_SYMBOL_GPL(companion_do_io_rx);
+
+/**
+ * companion_do_can_tx() - send CAN packet
+ * @parent: address of the parent device
+ * @port: which CAN port to send
+ * @prio: priority of the CAN frame
+ * @cf: address of the CAN frame to send
+ */
+int companion_do_can_tx(struct device *parent,
+ u8 port,
+ u8 prio,
+ const struct can_frame *cf)
+{
+ struct companion_spi_priv *priv = dev_get_drvdata(parent);
+ int err = pm_can_data_tx(&priv->pm, port, prio, cf);
+ if (!err)
+ wake_up_interruptible(&priv->wait);
+ return err;
+}
+EXPORT_SYMBOL_GPL(companion_do_can_tx);
+
+/**
+ * companion_do_can_rx() - receive CAN packet
+ * @parent: address of the parent device
+ * @port: which CAN port to receive
+ * @cf: address of the CAN frame to receive
+ */
+int companion_do_can_rx(struct device *parent,
+ u8 port,
+ struct can_frame *cf)
+{
+ struct companion_spi_priv *priv = dev_get_drvdata(parent);
+ return pm_can_data_rx(&priv->pm, port, cf);
+}
+EXPORT_SYMBOL_GPL(companion_do_can_rx);
+
+/**
+ * companion_do_can_err() - receive CAN error packet
+ * @parent: address of the parent device
+ * @port: which CAN port to receive
+ * @bec: address to store CAN error counter
+ * @state: address to store CAN state
+ * @code: address to store CAN error code
+ */
+int companion_do_can_err(struct device *parent,
+ u8 port,
+ struct can_berr_counter *bec,
+ u8 *state,
+ u8 *code)
+{
+ struct companion_spi_priv *priv = dev_get_drvdata(parent);
+ return pm_can_err(&priv->pm, port, bec, state, code);
+}
+EXPORT_SYMBOL_GPL(companion_do_can_err);
+
+/**
+ * companion_do_set_can_bittiming() - set CAN bittiming
+ * @parent: address of the parent device
+ * @port: which CAN port to set
+ * @bittiming: address of the bittiming to set
+ */
+int companion_do_set_can_bittiming(struct device *parent,
+ u8 port,
+ const struct can_bittiming *bittiming)
+{
+ struct companion_spi_priv *priv = dev_get_drvdata(parent);
+ int err = pm_can_set_bittiming(&priv->pm, port, bittiming);
+ if (!err) {
+ wake_up_interruptible(&priv->wait);
+ err = pm_wait_for_response(&priv->pm, port, bcp_can_bittiming);
+ }
+ return err;
+}
+EXPORT_SYMBOL_GPL(companion_do_set_can_bittiming);
+
+/**
+ * companion_do_set_can_mode() - set CAN mode
+ * @parent: address of the parent device
+ * @port: which CAN port to set
+ * @mode: the CAN mode to set
+ */
+int companion_do_set_can_mode(struct device *parent,
+ u8 port,
+ enum can_mode mode)
+{
+ struct companion_spi_priv *priv = dev_get_drvdata(parent);
+ int err = pm_can_set_mode(&priv->pm, port, mode);
+ if (!err) {
+ wake_up_interruptible(&priv->wait);
+ err = pm_wait_for_response(&priv->pm, port, bcp_can_mode);
+ }
+ return err;
+}
+EXPORT_SYMBOL_GPL(companion_do_set_can_mode);
+
+/**
+ * companion_do_set_can_ctrlmode() - set CAN control mode
+ * @parent: address of the parent device
+ * @port: which CAN port to set
+ * @ctrl: the CAN control mode to set
+ */
+int companion_do_set_can_ctrlmode(struct device *parent,
+ u8 port,
+ u32 ctrl)
+{
+ struct companion_spi_priv *priv = dev_get_drvdata(parent);
+ int err = pm_can_set_ctrlmode(&priv->pm, port, ctrl);
+ if (!err) {
+ wake_up_interruptible(&priv->wait);
+ err = pm_wait_for_response(&priv->pm, port, bcp_can_mode);
+ }
+ return err;
+}
+EXPORT_SYMBOL_GPL(companion_do_set_can_ctrlmode);
+
+/**
+ * companion_do_get_can_status() - get CAN status
+ * @parent: address of the parent device
+ * @port: which CAN port to receive
+ * @bec: address of the CAN error counter to store
+ */
+int companion_do_get_can_status(struct device *parent,
+ u8 port,
+ struct can_berr_counter *bec)
+{
+ struct companion_spi_priv *priv = dev_get_drvdata(parent);
+ int err = pm_can_get_status(&priv->pm, port);
+ if (!err) {
+ wake_up_interruptible(&priv->wait);
+ err = pm_wait_for_response(&priv->pm, port, bcp_can_status);
+ if (!err) {
+ bec->rxerr = priv->pm.rx_err[port];
+ bec->txerr = priv->pm.tx_err[port];
+ }
+ }
+ return err;
+}
+EXPORT_SYMBOL_GPL(companion_do_get_can_status);
+
+/**
+ * companion_do_get_can_txq_status() - get single CAN tx queue status
+ * @parent: address of the parent device
+ * @port: which CAN port to inquiry
+ * @prio: which CAN queue to inquiry
+ * @lost_txq_sync: address of flag to store whether tx queue lost sync
+ */
+int companion_do_get_can_txq_status(struct device *parent,
+ u8 port,
+ u8 prio,
+ bool *lost_txq_sync)
+{
+ struct companion_spi_priv *priv = dev_get_drvdata(parent);
+ struct companion_protocol_manager *pm = &priv->pm;
+ u8 local, remote;
+ int err;
+
+ if (prio >= BCP_CAN_PRIOS)
+ return -EINVAL;
+
+ err = pm_can_get_txq_status(pm, port);
+ if (!err) {
+ wake_up_interruptible(&priv->wait);
+ err = pm_wait_for_response(pm, port, bcp_can_txq_status);
+ if (!err) {
+ local = pm->local_txq[port][prio];
+ remote = pm->remote_txq[port][prio];
+
+ if (local != remote) {
+ *lost_txq_sync = true;
+ pm->stats.can_lost_txq_sync[port][prio]++;
+ } else {
+ *lost_txq_sync = false;
+ pm->stats.can_ack_timeout[port][prio]++;
+ }
+
+ pm->local_txq[port][prio] = remote;
+ }
+ }
+ return err;
+}
+EXPORT_SYMBOL_GPL(companion_do_get_can_txq_status);
+
+/**
+ * companion_do_get_can_txq_status_all() - get all CAN tx queue status
+ * @parent: address of the parent device
+ * @port: which CAN port to inquiry
+ */
+int companion_do_get_can_txq_status_all(struct device *parent,
+ u8 port)
+{
+ struct companion_spi_priv *priv = dev_get_drvdata(parent);
+ int err;
+
+ err = pm_can_get_txq_status(&priv->pm, port);
+ if (!err) {
+ wake_up_interruptible(&priv->wait);
+ err = pm_wait_for_response(&priv->pm, port, bcp_can_txq_status);
+ if (!err)
+ memcpy(priv->pm.local_txq[port],
+ priv->pm.remote_txq[port],
+ BCP_CAN_PRIOS);
+ }
+ return err;
+}
+EXPORT_SYMBOL_GPL(companion_do_get_can_txq_status_all);
+
+/**
+ * companion_do_can_txq_is_full() - inquiry CAN tx queue is full
+ * @parent: address of the parent device
+ * @port: which CAN port to inquiry
+ * @prio: which CAN queue to inquiry
+ * @is_full: address of flag to store is full or not
+ */
+int companion_do_can_txq_is_full(struct device *parent,
+ u8 port,
+ u8 prio,
+ bool *is_full)
+{
+ struct companion_spi_priv *priv = dev_get_drvdata(parent);
+ return pm_can_txq_is_full(&priv->pm, port, prio, is_full);
+}
+EXPORT_SYMBOL_GPL(companion_do_can_txq_is_full);
+
+/**
+ * companion_do_can_txq_has_space() - inquiry CAN tx queue has space
+ * @parent: address of the parent device
+ * @port: which CAN port to inquiry
+ * @prio: which CAN queue to inquiry
+ * @has_space: address of flag to store has space or not
+ */
+int companion_do_can_txq_has_space(struct device *parent,
+ u8 port,
+ u8 prio,
+ bool *has_space)
+{
+ struct companion_spi_priv *priv = dev_get_drvdata(parent);
+ return pm_can_txq_has_space(&priv->pm, port, prio, has_space);
+}
+EXPORT_SYMBOL_GPL(companion_do_can_txq_has_space);
+
+/**
+ * companion_do_can_start_tx_timer() - start CAN tx timeout detection
+ * @parent: address of the parent device
+ * @port: which CAN port to start
+ * @prio: which CAN queue to start
+ */
+int companion_do_can_start_tx_timer(struct device *parent,
+ u8 port,
+ u8 prio)
+{
+ struct companion_spi_priv *priv = dev_get_drvdata(parent);
+ return pm_can_start_tx_timer(&priv->pm, port, prio);
+}
+EXPORT_SYMBOL_GPL(companion_do_can_start_tx_timer);
+
+/**
+ * companion_do_can_stop_tx_timer() - stop CAN tx timeout detection
+ * @parent: address of the parent device
+ * @port: which CAN port to stop
+ * @prio: which CAN queue to stop
+ */
+int companion_do_can_stop_tx_timer(struct device *parent,
+ u8 port,
+ u8 prio)
+{
+ struct companion_spi_priv *priv = dev_get_drvdata(parent);
+ return pm_can_stop_tx_timer(&priv->pm, port, prio);
+}
+EXPORT_SYMBOL_GPL(companion_do_can_stop_tx_timer);
+
+/**
+ * show_dump_packets() - display dump_packets value in sysfs entry
+ * @dev: address of the device associated with sysfs entry
+ * @attr: address of the device attribute
+ * @buf: address of the buffer to encode value
+ */
+static ssize_t show_dump_packets(struct device *dev,
+ struct device_attribute *attr,
+ char *buf)
+{
+ struct spi_device *spi = to_spi_device(dev);
+ struct companion_spi_priv *priv = spi_get_drvdata(spi);
+ return snprintf(buf, PAGE_SIZE, "%d\n", priv->dump_packets);
+}
+
+/**
+ * store_dump_packets() - store dump_packets value from sysfs entry
+ * @dev: address of the device associated with sysfs entry
+ * @attr: address of the device attribute
+ * @buf: address of the buffer to decode value
+ * @count: number of bytes in the buffer
+ */
+static ssize_t store_dump_packets(struct device *dev,
+ struct device_attribute *attr,
+ const char *buf,
+ size_t count)
+{
+ struct spi_device *spi = to_spi_device(dev);
+ struct companion_spi_priv *priv = spi_get_drvdata(spi);
+ int ret, value;
+
+ ret = sscanf(buf, "%d", &value);
+ if (ret != 1) {
+ dev_err(&spi->dev, "input invalid value: %s\n", buf);
+ return -EINVAL;
+ }
+
+ priv->dump_packets = (value != 0);
+ return count;
+}
+static DEVICE_ATTR(dump_packets, S_IRUGO | S_IWUSR,
+ show_dump_packets, store_dump_packets);
+
+/**
+ * show_overflows() - display overflows value in sysfs entry
+ * @dev: address of the device associated with sysfs entry
+ * @attr: address of the device attribute
+ * @buf: address of the buffer to encode value
+ */
+static ssize_t show_overflows(struct device *dev,
+ struct device_attribute *attr,
+ char *buf)
+{
+ struct spi_device *spi = to_spi_device(dev);
+ struct companion_spi_priv *priv = spi_get_drvdata(spi);
+ int ret, pos, i, j, total = 0;
+
+ ret = snprintf(buf, PAGE_SIZE, "io\ntx: %u, rx: %u\n\n",
+ priv->pm.stats.io_tx_overflows,
+ priv->pm.stats.io_rx_overflows);
+ pos = ret;
+
+ for (i = 0; i < BCP_CAN_PORTS; ++i) {
+ ret = snprintf(buf + pos, PAGE_SIZE - pos, "can%u\n", i);
+ pos += ret;
+
+ for (j = 0; j < BCP_CAN_PRIOS; ++j) {
+ ret = snprintf(buf + pos, PAGE_SIZE - pos,
+ "[%u]:%u ", j,
+ priv->pm.stats.can_tx_overflows[i][j]);
+ total += priv->pm.stats.can_tx_overflows[i][j];
+ pos += ret;
+ }
+
+ ret = snprintf(buf + pos, PAGE_SIZE - pos,
+ "\ntx: %u, rx: %u, err: %u\n\n",
+ total,
+ priv->pm.stats.can_rx_overflows[i],
+ priv->pm.stats.can_err_overflows[i]);
+ pos += ret;
+ }
+ return pos;
+}
+static DEVICE_ATTR(overflows, S_IRUGO, show_overflows, NULL);
+
+/**
+ * show_traffic() - display traffic of IO and CAN in sysfs entry
+ * @dev: address of the device associated with sysfs entry
+ * @attr: address of the device attribute
+ * @buf: address of the buffer to encode value
+ */
+static ssize_t show_traffic(struct device *dev,
+ struct device_attribute *attr,
+ char *buf)
+{
+ struct spi_device *spi = to_spi_device(dev);
+ struct companion_spi_priv *priv = spi_get_drvdata(spi);
+ int ret, pos, i, j;
+
+ ret = snprintf(buf, PAGE_SIZE, "io\ntx: %u, rx: %u\n\n",
+ priv->pm.stats.io_tx, priv->pm.stats.io_rx);
+ pos = ret;
+
+ for (i = 0; i < BCP_CAN_PORTS; ++i) {
+ ret = snprintf(buf + pos, PAGE_SIZE - pos, "can%u\n", i);
+ pos += ret;
+
+ ret = snprintf(buf + pos, PAGE_SIZE - pos, "tx : ");
+ pos += ret;
+ for (j = 0; j < BCP_CAN_PRIOS; ++j) {
+ ret = snprintf(buf + pos, PAGE_SIZE - pos,
+ "[%u]:%u ", j,
+ priv->pm.stats.can_tx[i][j]);
+ pos += ret;
+ }
+
+ ret = snprintf(buf + pos, PAGE_SIZE - pos, "\nack success: ");
+ pos += ret;
+ for (j = 0; j < BCP_CAN_PRIOS; ++j) {
+ ret = snprintf(buf + pos, PAGE_SIZE - pos,
+ "[%u]:%u ", j,
+ priv->pm.stats.can_ack_success[i][j]);
+ pos += ret;
+ }
+
+ ret = snprintf(buf + pos, PAGE_SIZE - pos, "\nack failure: ");
+ pos += ret;
+ for (j = 0; j < BCP_CAN_PRIOS; ++j) {
+ ret = snprintf(buf + pos, PAGE_SIZE - pos,
+ "[%u]:%u ", j,
+ priv->pm.stats.can_ack_failure[i][j]);
+ pos += ret;
+ }
+
+ ret = snprintf(buf + pos, PAGE_SIZE - pos, "\nlost seq : ");
+ pos += ret;
+ for (j = 0; j < BCP_CAN_PRIOS; ++j) {
+ ret = snprintf(buf + pos, PAGE_SIZE - pos,
+ "[%u]:%u ", j,
+ priv->pm.stats.can_lost_seq_sync[i][j]);
+ pos += ret;
+ }
+
+ ret = snprintf(buf + pos, PAGE_SIZE - pos, "\nlost txq : ");
+ pos += ret;
+ for (j = 0; j < BCP_CAN_PRIOS; ++j) {
+ ret = snprintf(buf + pos, PAGE_SIZE - pos,
+ "[%u]:%u ", j,
+ priv->pm.stats.can_lost_txq_sync[i][j]);
+ pos += ret;
+ }
+
+ ret = snprintf(buf + pos, PAGE_SIZE - pos, "\nack timeout: ");
+ pos += ret;
+ for (j = 0; j < BCP_CAN_PRIOS; ++j) {
+ ret = snprintf(buf + pos, PAGE_SIZE - pos,
+ "[%u]:%u ", j,
+ priv->pm.stats.can_ack_timeout[i][j]);
+ pos += ret;
+ }
+
+ ret = snprintf(buf + pos, PAGE_SIZE - pos, "\nack unexpect:");
+ pos += ret;
+ for (j = 0; j < BCP_CAN_PRIOS; ++j) {
+ ret = snprintf(buf + pos, PAGE_SIZE - pos,
+ "[%u]:%u ", j,
+ priv->pm.stats.can_ack_unexpect[i][j]);
+ pos += ret;
+ }
+
+ ret = snprintf(buf + pos, PAGE_SIZE - pos,
+ "\nrx : %u\nerr : %u\n\n",
+ priv->pm.stats.can_rx[i],
+ priv->pm.stats.can_err[i]);
+ pos += ret;
+ }
+ return pos;
+}
+static DEVICE_ATTR(traffic, S_IRUGO, show_traffic, NULL);
+
+/**
+ * show_can_space() - display CAN queue space in sysfs entry
+ * @dev: address of the device associated with sysfs entry
+ * @attr: address of the device attribute
+ * @buf: address of the buffer to encode value
+ */
+static ssize_t show_can_space(struct device *dev,
+ struct device_attribute *attr,
+ char *buf)
+{
+ struct spi_device *spi = to_spi_device(dev);
+ struct companion_spi_priv *priv = spi_get_drvdata(spi);
+ int i, j, ret, pos = 0;
+
+ for (i = 0; i < BCP_CAN_PORTS; ++i) {
+ ret = snprintf(buf + pos, PAGE_SIZE - pos, "can%u\n", i);
+ pos += ret;
+
+ ret = snprintf(buf + pos, PAGE_SIZE - pos, "local : ");
+ pos += ret;
+ for (j = 0; j < BCP_CAN_PRIOS; ++j) {
+ ret = snprintf(buf + pos, PAGE_SIZE - pos,
+ "[%u]:%u ", j,
+ priv->pm.local_txq[i][j]);
+ pos += ret;
+ }
+
+ ret = snprintf(buf + pos, PAGE_SIZE - pos, "\nremote: ");
+ pos += ret;
+ for (j = 0; j < BCP_CAN_PRIOS; ++j) {
+ ret = snprintf(buf + pos, PAGE_SIZE - pos,
+ "[%u]:%u ", j,
+ priv->pm.remote_txq[i][j]);
+ pos += ret;
+ }
+
+ ret = snprintf(buf + pos, PAGE_SIZE - pos, "\n\n");
+ pos += ret;
+ }
+ return pos;
+}
+static DEVICE_ATTR(can_space, S_IRUGO, show_can_space, NULL);
+
+/**
+ * show_busy() - display busy signal statisitics in sysfs entry
+ * @dev: address of the device associated with sysfs entry
+ * @attr: address of the device attribute
+ * @buf: address of the buffer to encode value
+ */
+static ssize_t show_busy(struct device *dev,
+ struct device_attribute *attr,
+ char *buf)
+{
+ struct spi_device *spi = to_spi_device(dev);
+ struct companion_spi_priv *priv = spi_get_drvdata(spi);
+ return snprintf(buf, PAGE_SIZE,
+ "while_busy_ext : %u\n"
+ "while_busy_fail : %u\n"
+ "until_busy_ext : %u\n"
+ "until_busy_fail : %u\n"
+ "force_started : %u\n"
+ "force_started_failure: %u\n"
+ "ready_failure : %u\n",
+ priv->stats.while_busy_ext,
+ priv->stats.while_busy_fail,
+ priv->stats.until_busy_ext,
+ priv->stats.until_busy_fail,
+ priv->stats.force_started,
+ priv->stats.force_started_failure,
+ priv->stats.ready_failure);
+}
+static DEVICE_ATTR(busy, S_IRUGO, show_busy, NULL);
+
+static struct attribute *companion_spi_sysfs_attrs[] = {
+ &dev_attr_dump_packets.attr,
+ &dev_attr_overflows.attr,
+ &dev_attr_traffic.attr,
+ &dev_attr_can_space.attr,
+ &dev_attr_busy.attr,
+ NULL
+};
+
+static struct attribute_group companion_spi_attribute_group = {
+ .attrs = companion_spi_sysfs_attrs
+};
+
+/**
+ * slave_has_request() - inquiry spi slave has request
+ * @priv: address of companion-spi private data
+ */
+static inline bool slave_has_request(struct companion_spi_priv *priv)
+{
+ int value = gpio_get_value(priv->request_gpios);
+ return value == priv->request_gpios_assert;
+}
+
+/**
+ * slave_is_busy() - inquiry spi slave is busy
+ * @priv: address of companion-spi private data
+ */
+static inline bool slave_is_busy(struct companion_spi_priv *priv)
+{
+ int value = gpio_get_value(priv->busy_gpios);
+ return value == priv->busy_gpios_assert;
+}
+
+/**
+ * slave_is_not_busy() - inquiry spi slave is not busy
+ * @priv: address of companion-spi private data
+ */
+static inline bool slave_is_not_busy(struct companion_spi_priv *priv)
+{
+ int value = gpio_get_value(priv->busy_gpios);
+ return value != priv->busy_gpios_assert;
+}
+
+/**
+ * slave_select() - select spi slave
+ * @priv: address of companion-spi private data
+ */
+static inline void slave_select(struct companion_spi_priv *priv)
+{
+ gpio_set_value(priv->cs_gpios, priv->cs_gpios_assert);
+}
+
+/**
+ * slave_deselect() - deselect spi slave
+ * @priv: address of companion-spi private data
+ */
+static inline void slave_deselect(struct companion_spi_priv *priv)
+{
+ gpio_set_value(priv->cs_gpios, !priv->cs_gpios_assert);
+}
+
+/**
+ * companion_spi_wait_while_busy() - wait while spi slave is busy
+ * @priv: address of companion-spi private data
+ */
+static int companion_spi_wait_while_busy(struct companion_spi_priv *priv)
+{
+ /*
+ * as short as possible wait while busy polling which shall
+ * succeed most of the times
+ */
+ unsigned int count = READY_POLL_US / READY_POLL_US_GRAN;
+ while (count--) {
+ if (slave_is_not_busy(priv))
+ return 0;
+
+ udelay(READY_POLL_US_GRAN);
+ }
+
+ /*
+ * wait while busy polling with sleeping, in case companion
+ * is busy with other things, this shall happen rarely
+ */
+ count = READY_POLL_MS / READY_POLL_MS_GRAN;
+ while (count--) {
+ if (slave_is_not_busy(priv)) {
+ priv->stats.while_busy_ext++;
+ dev_info(&priv->spi->dev,
+ "waited long while busy (%u)\n",
+ priv->stats.while_busy_ext);
+ return 0;
+ }
+
+ msleep(READY_POLL_MS_GRAN);
+ }
+
+ priv->stats.while_busy_fail++;
+ dev_err(&priv->spi->dev,
+ "time out waiting for busy deassertion (%u)\n",
+ priv->stats.while_busy_fail);
+ return -EBUSY;
+}
+
+/**
+ * companion_spi_wait_until_busy() - wait until spi slave is busy
+ * @priv: address of companion-spi private data
+ */
+static int companion_spi_wait_until_busy(struct companion_spi_priv *priv)
+{
+ /*
+ * as short as possible wait until busy polling which shall
+ * succeed most of the times
+ */
+ unsigned int count = READY_POLL_US / READY_POLL_US_GRAN;
+ while (count--) {
+ if (slave_is_busy(priv))
+ return 0;
+
+ udelay(READY_POLL_US_GRAN);
+ }
+
+ /*
+ * wait until busy polling with sleeping, in case companion
+ * is busy with other things, this shall happen rarely
+ */
+ count = READY_POLL_MS / READY_POLL_MS_GRAN;
+ while (count--) {
+ if (slave_is_busy(priv)) {
+ priv->stats.until_busy_ext++;
+ dev_info(&priv->spi->dev,
+ "waited long until busy (%u)\n",
+ priv->stats.until_busy_ext);
+ return 0;
+ }
+
+ msleep(READY_POLL_MS_GRAN);
+ }
+
+ priv->stats.until_busy_fail++;
+ dev_err(&priv->spi->dev,
+ "time out waiting for busy assertion (%u)\n",
+ priv->stats.until_busy_fail);
+ return -EBUSY;
+}
+
+/**
+ * companion_spi_cpu_to_be32() - convert companion packet to big endian 32 bit
+ * @buf: address of the packet to convert
+ */
+static void companion_spi_cpu_to_be32(char *buf)
+{
+ u32 *buf32 = (u32*)buf;
+ int i;
+
+ for (i = 0; i < (BCP_PACKET_SIZE / sizeof(u32)); i++, buf32++)
+ *buf32 = cpu_to_be32(*buf32);
+}
+
+/**
+ * companion_spi_be32_to_cpu() - convert companion packet from big endian 32 bit
+ * @buf: address of the packet to convert
+ */
+static void companion_spi_be32_to_cpu(char *buf)
+{
+ u32 *buf32 = (u32*)buf;
+ int i;
+
+ for (i = 0; i < (BCP_PACKET_SIZE / sizeof(u32)); i++, buf32++)
+ *buf32 = be32_to_cpu(*buf32);
+}
+
+/**
+ * companion_spi_transceive() - transceive spi message
+ * @priv: address of companion-spi private data
+ * @message: address of the spi message to transceive
+ * @transfer: address of the spi transfer
+ */
+static void companion_spi_transceive(struct companion_spi_priv *priv,
+ struct spi_message *message,
+ struct spi_transfer *transfer)
+{
+ const struct companion_packet *p;
+ int status;
+
+ if (priv->dump_packets) {
+ p = (const struct companion_packet*)transfer->tx_buf;
+ dump_packet(p, KERN_INFO, DRIVER_NAME" Tx: ");
+ }
+
+ companion_spi_cpu_to_be32((char*)transfer->tx_buf);
+
+ spi_message_init_with_transfers(message, transfer, 1);
+
+ status = companion_spi_wait_while_busy(priv);
+
+ slave_select(priv);
+
+ if (status != 0) {
+ priv->stats.force_started++;
+ dev_err(&priv->spi->dev,
+ "force started transfer (%u)\n",
+ priv->stats.force_started);
+
+ /* wait slave to pull up busy line in case force started */
+ status = companion_spi_wait_while_busy(priv);
+ if (status != 0) {
+ priv->stats.force_started_failure++;
+ dev_err(&priv->spi->dev,
+ "force started failed, continuing (%u)\n",
+ priv->stats.force_started_failure);
+ }
+ }
+
+ status = companion_spi_wait_until_busy(priv);
+ if (status != 0) {
+ priv->stats.ready_failure++;
+ dev_err(&priv->spi->dev,
+ "started transfer in case not ready (%u)\n",
+ priv->stats.ready_failure);
+ }
+
+ if (spi_sync(priv->spi, message) != 0)
+ dev_err(&priv->spi->dev,
+ "sending spi message failed: %d\n",
+ message->status);
+
+ slave_deselect(priv);
+
+ companion_spi_be32_to_cpu(transfer->rx_buf);
+
+ if (priv->dump_packets) {
+ p = (const struct companion_packet*)transfer->rx_buf;
+ dump_packet(p, KERN_INFO, DRIVER_NAME" Rx: ");
+ }
+}
+
+/**
+ * companion_spi_request_irq() - irq handler of request signal
+ * @irq: irq number of request signal
+ * @data: address of user supplied data for irq handler
+ */
+static irqreturn_t companion_spi_request_irq(int irq, void *data)
+{
+ struct companion_spi_priv *priv = data;
+ wake_up_interruptible(&priv->wait);
+ return IRQ_HANDLED;
+}
+
+/**
+ * companion_spi_thread() - main thread to drive spi communication
+ * @data: address of user supplied data for thread
+ */
+static int companion_spi_thread(void *data)
+{
+ struct companion_spi_priv *priv = data;
+ struct companion_packet tx_packet;
+ struct companion_packet rx_packet;
+ struct spi_message message;
+ struct spi_transfer transfer;
+
+ memset(&transfer, 0, sizeof(transfer));
+ transfer.tx_buf = tx_packet.data;
+ transfer.rx_buf = rx_packet.data;
+ transfer.len = sizeof(struct companion_packet);
+ transfer.cs_change = 0;
+ transfer.bits_per_word = 32;
+
+ for (;;) {
+ if (wait_event_interruptible(priv->wait,
+ kthread_should_stop() ||
+ slave_has_request(priv) ||
+ qm_has_tx_data(&priv->pm.qm)))
+ continue;
+
+ if (kthread_should_stop())
+ break;
+
+ pm_prepare_tx(&priv->pm, &tx_packet);
+ companion_spi_transceive(priv, &message, &transfer);
+ pm_on_tx_done(&priv->pm);
+ pm_on_rx_done(&priv->pm, &rx_packet);
+ }
+
+ return 0;
+}
+
+static const struct of_device_id companion_spi_of_match[] = {
+ { .compatible = DRIVER_NAME, .data = NULL, },
+ { /* sentinel */ },
+};
+MODULE_DEVICE_TABLE(of, companion_spi_of_match);
+
+/**
+ * companion_spi_parse_dt() - parse device tree
+ * @priv: address of companion-spi private data
+ */
+static int companion_spi_parse_dt(struct companion_spi_priv *priv)
+{
+ struct device *dev = &priv->spi->dev;
+ struct device_node *np = dev->of_node;
+ int gpio;
+ enum of_gpio_flags flags;
+
+ if (!np) {
+ dev_err(dev, "no device tree data\n");
+ return -EINVAL;
+ }
+
+ gpio = of_get_named_gpio_flags(np, "request-gpios", 0, &flags);
+ if (!gpio_is_valid(gpio)) {
+ dev_err(dev, "invalid 'request-gpios' supplied\n");
+ return -EINVAL;
+ }
+ priv->request_gpios = gpio;
+ priv->request_gpios_assert = (flags & OF_GPIO_ACTIVE_LOW) ? 0 : 1;
+
+ gpio = of_get_named_gpio_flags(np, "busy-gpios", 0, &flags);
+ if (!gpio_is_valid(gpio)) {
+ dev_err(dev, "invalid 'busy-gpios' supplied\n");
+ return -EINVAL;
+ }
+ priv->busy_gpios = gpio;
+ priv->busy_gpios_assert = (flags & OF_GPIO_ACTIVE_LOW) ? 0 : 1;
+
+ gpio = of_get_named_gpio_flags(np, "cs-gpios", 0, &flags);
+ if (!gpio_is_valid(gpio)) {
+ dev_err(dev, "invalid 'cs-gpios' supplied\n");
+ return -EINVAL;
+ }
+ priv->cs_gpios = gpio;
+ priv->cs_gpios_assert = (flags & OF_GPIO_ACTIVE_LOW) ? 0 : 1;
+
+ return 0;
+}
+
+/**
+ * companion_spi_probe() - probe callback
+ * @spi: address of the spi device
+ */
+static int companion_spi_probe(struct spi_device *spi)
+{
+ struct companion_spi_priv *priv;
+ u8 null_packet[BCP_PACKET_SIZE] = {0};
+ int err;
+
+ priv = devm_kzalloc(&spi->dev, sizeof(*priv), GFP_KERNEL);
+ if (!priv)
+ return -ENOMEM;
+
+ priv->spi = spi;
+ init_waitqueue_head(&priv->wait);
+ pm_init(&priv->pm);
+
+ err = companion_spi_parse_dt(priv);
+ if (err)
+ return err;
+
+ err = devm_gpio_request_one(&spi->dev,
+ priv->request_gpios,
+ GPIOF_IN,
+ DRIVER_NAME);
+ if (err) {
+ dev_err(&spi->dev, "request 'request-gpios' failed: %d\n", err);
+ return err;
+ }
+
+ err = devm_gpio_request_one(&spi->dev,
+ priv->busy_gpios,
+ GPIOF_IN,
+ DRIVER_NAME);
+ if (err) {
+ dev_err(&spi->dev, "request 'busy-gpios' failed: %d\n", err);
+ return err;
+ }
+
+ err = devm_gpio_request_one(&spi->dev,
+ priv->cs_gpios,
+ GPIOF_OUT_INIT_HIGH,
+ DRIVER_NAME);
+ if (err) {
+ dev_err(&spi->dev, "request 'cs-gpios'failed: %d\n", err);
+ return err;
+ }
+
+ spi->mode = SPI_MODE_1;
+
+ err = spi_setup(spi);
+ if (err) {
+ dev_err(&spi->dev, "spi_setup() returns: %d\n", err);
+ return err;
+ }
+
+ err = spi_write(spi, null_packet, sizeof(null_packet));
+ if (err) {
+ dev_err(&spi->dev, "dummy transfer failed: %d\n", err);
+ return err;
+ }
+
+ spi_set_drvdata(spi, priv);
+
+ err = sysfs_create_group(&spi->dev.kobj,
+ &companion_spi_attribute_group);
+ if (err) {
+ dev_err(&spi->dev, "sysfs_create_group() returns: %d\n", err);
+ return err;
+ }
+
+ priv->task = kthread_run(companion_spi_thread, priv, DRIVER_NAME);
+ if (!priv->task)
+ return -EIO;
+
+ err = devm_request_irq(&spi->dev,
+ gpio_to_irq(priv->request_gpios),
+ companion_spi_request_irq,
+ IRQF_TRIGGER_FALLING,
+ "companion-spi-request",
+ priv);
+ if (err)
+ return -ENODEV;
+
+ return of_platform_populate(spi->dev.of_node, NULL, NULL, &spi->dev);
+}
+
+/**
+ * companion_spi_remove() - remove callback
+ * @spi: address of the spi device
+ */
+static int companion_spi_remove(struct spi_device *spi)
+{
+ struct companion_spi_priv *priv = spi_get_drvdata(spi);
+
+ qm_reset(&priv->pm.qm);
+ kthread_stop(priv->task);
+ sysfs_remove_group(&spi->dev.kobj,
+ &companion_spi_attribute_group);
+ of_platform_depopulate(&spi->dev);
+ return 0;
+}
+
+static struct spi_driver companion_spi_driver = {
+ .driver = {
+ .name = DRIVER_NAME,
+ .owner = THIS_MODULE,
+ .of_match_table = of_match_ptr(companion_spi_of_match),
+ },
+ .probe = companion_spi_probe,
+ .remove = companion_spi_remove,
+};
+module_spi_driver(companion_spi_driver);
+
+MODULE_AUTHOR("Zhu Yi <yi.zhu5@cn.bosch.com>");
+MODULE_DESCRIPTION("Companion low level init/core code");
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/spi/companion/protocol-manager.c b/drivers/spi/companion/protocol-manager.c
new file mode 100644
index 0000000..3a9dc40
--- /dev/null
+++ b/drivers/spi/companion/protocol-manager.c
@@ -0,0 +1,1035 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Companion protocol manager code
+ *
+ * Copyright (C) 2015-2018 Bosch Sicherheitssysteme GmbH
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#include "protocol-manager.h"
+
+#define PM_RESPONSE_TIMEOUT HZ
+#define PM_CAN_TX_TIMEOUT msecs_to_jiffies(30000)
+
+/**
+ * struct companion_filter - companion packet filter
+ * @node: filter list node
+ * @match: address of match callback
+ * @process: address of process callback
+ */
+struct companion_filter {
+ struct list_head node;
+ bool (* match)(const struct companion_packet *p);
+ void (* process)(struct companion_protocol_manager *pm,
+ const struct companion_packet *p);
+};
+
+/**
+ * null_match() - match null packet
+ * @p: address of the packet to handle
+ */
+static bool null_match(const struct companion_packet *p)
+{
+ return is_null_type(p);
+}
+
+/**
+ * io_match() - match IO packet
+ * @p: address of the packet to handle
+ */
+static bool io_match(const struct companion_packet *p)
+{
+ return is_io_type(p);
+}
+
+/**
+ * io_process() - process IO packet
+ * @pm: address of the protocol manager
+ * @p: address of the packet to handle
+ */
+static void io_process(struct companion_protocol_manager *pm,
+ const struct companion_packet *p)
+{
+ if (qm_io_rxq_in(&pm->qm, p)) {
+ down_read(&pm->io_lock);
+ if (pm->io_ops && pm->io_ops->on_rx_done)
+ pm->io_ops->on_rx_done(pm->io_data);
+ up_read(&pm->io_lock);
+
+ pm->stats.io_rx++;
+ } else {
+ pm->stats.io_rx_overflows++;
+ }
+}
+
+/**
+ * can_data_match() - match CAN data packet
+ * @p: address of the packet to handle
+ */
+static bool can_data_match(const struct companion_packet *p)
+{
+ return ((const struct can_data_frame*)p)->type == BCP_CAN_DATA;
+}
+
+/**
+ * can_data_process() - process CAN data packet
+ * @pm: address of the protocol manager
+ * @p: address of the packet to handle
+ */
+static void can_data_process(struct companion_protocol_manager *pm,
+ const struct companion_packet *p)
+{
+ u8 port = ((const struct can_data_frame*)p)->port - 1;
+
+ if (port >= BCP_CAN_PORTS)
+ return;
+
+ if (qm_can_rxq_in(&pm->qm, p, port)) {
+ down_read(&pm->can_lock[port]);
+ if (pm->can_ops[port] && pm->can_ops[port]->on_rx_done)
+ pm->can_ops[port]->on_rx_done(pm->can_data[port]);
+ up_read(&pm->can_lock[port]);
+
+ pm->stats.can_rx[port]++;
+ } else {
+ pm->stats.can_rx_overflows[port]++;
+ }
+}
+
+/**
+ * can_bittiming_match() - match CAN bittiming packet
+ * @p: address of the packet to handle
+ */
+static bool can_bittiming_match(const struct companion_packet *p)
+{
+ return ((const struct can_bittiming_response*)p)->type ==
+ BCP_CAN_BITTIMING;
+}
+
+/**
+ * can_bittiming_process() - process CAN bittiming packet
+ * @pm: address of the protocol manager
+ * @p: address of the packet to handle
+ */
+static void can_bittiming_process(struct companion_protocol_manager *pm,
+ const struct companion_packet *p)
+{
+ u8 port = ((const struct can_bittiming_response*)p)->port - 1;
+ u8 status = ((const struct can_bittiming_response*)p)->status;
+
+ if (port >= BCP_CAN_PORTS)
+ return;
+
+ if (status == BCP_STATUS_SUCCESS) {
+ pm->response[port][bcp_can_bittiming] = 0;
+ pm->stats.can_ack_success[port][0]++;
+ } else {
+ pm->response[port][bcp_can_bittiming] = -EINVAL;
+ pm->stats.can_ack_failure[port][0]++;
+ }
+
+ if (test_and_clear_bit(bcp_can_bittiming, &pm->flags[port]))
+ wake_up_interruptible_all(&pm->wait[port]);
+}
+
+/**
+ * can_mode_match() - match CAN mode packet
+ * @p: address of the packet to handle
+ */
+static bool can_mode_match(const struct companion_packet *p)
+{
+ return ((const struct can_mode_response*)p)->type ==
+ BCP_CAN_MODE;
+}
+
+/**
+ * can_mode_process() - process CAN mode packet
+ * @pm: address of the protocol manager
+ * @p: address of the packet to handle
+ */
+static void can_mode_process(struct companion_protocol_manager *pm,
+ const struct companion_packet *p)
+{
+ u8 port = ((const struct can_mode_response*)p)->port - 1;
+ u8 status = ((const struct can_mode_response*)p)->status;
+
+ if (port >= BCP_CAN_PORTS)
+ return;
+
+ if (status == BCP_STATUS_SUCCESS) {
+ pm->response[port][bcp_can_mode] = 0;
+ pm->stats.can_ack_success[port][0]++;
+ } else {
+ pm->response[port][bcp_can_mode] = -EINVAL;
+ pm->stats.can_ack_failure[port][0]++;
+ }
+
+ if (test_and_clear_bit(bcp_can_mode, &pm->flags[port]))
+ wake_up_interruptible_all(&pm->wait[port]);
+}
+
+/**
+ * can_status_match() - match CAN status packet
+ * @p: address of the packet to handle
+ */
+static bool can_status_match(const struct companion_packet *p)
+{
+ return ((const struct can_status_response*)p)->type ==
+ BCP_CAN_STATUS;
+}
+
+/**
+ * can_status_process() - process CAN status packet
+ * @pm: address of the protocol manager
+ * @p: address of the packet to handle
+ */
+static void can_status_process(struct companion_protocol_manager *pm,
+ const struct companion_packet *p)
+{
+ u8 port = ((const struct can_status_response*)p)->port - 1;
+ u8 rx_err = ((const struct can_status_response*)p)->rx_err;
+ u8 tx_err = ((const struct can_status_response*)p)->tx_err;
+ u8 status = ((const struct can_status_response*)p)->status;
+
+ if (port >= BCP_CAN_PORTS)
+ return;
+
+ if (status == BCP_STATUS_SUCCESS) {
+ pm->response[port][bcp_can_status] = 0;
+ pm->rx_err[port] = rx_err;
+ pm->tx_err[port] = tx_err;
+
+ if (test_bit(bcp_can_status, &pm->flags[port])) {
+ pm->stats.can_ack_success[port][0]++;
+ goto polling_out;
+ }
+
+ if (qm_can_err_in(&pm->qm, p, port)) {
+ down_read(&pm->can_lock[port]);
+ if (pm->can_ops[port] && pm->can_ops[port]->on_error)
+ pm->can_ops[port]->on_error(pm->can_data[port]);
+ up_read(&pm->can_lock[port]);
+
+ pm->stats.can_err[port]++;
+ } else {
+ pm->stats.can_err_overflows[port]++;
+ }
+ } else {
+ pm->response[port][bcp_can_status] = -EINVAL;
+ if (test_bit(bcp_can_status, &pm->flags[port]))
+ pm->stats.can_ack_failure[port][0]++;
+ }
+
+polling_out:
+ if (test_and_clear_bit(bcp_can_status, &pm->flags[port]))
+ wake_up_interruptible_all(&pm->wait[port]);
+}
+
+/**
+ * can_tx_ack_match() - match CAN tx acknowledge packet
+ * @p: address of the packet to handle
+ */
+static bool can_tx_ack_match(const struct companion_packet *p)
+{
+ return ((const struct can_tx_acknowledge*)p)->type ==
+ BCP_CAN_TX_ACK;
+}
+
+/**
+ * can_tx_ack_process() - process CAN tx acknowledge packet
+ * @pm: address of the protocol manager
+ * @p: address of the packet to handle
+ */
+static void can_tx_ack_process(struct companion_protocol_manager *pm,
+ const struct companion_packet *p)
+{
+ u8 port = ((const struct can_tx_acknowledge*)p)->port - 1;
+ u8 prio = ((const struct can_tx_acknowledge*)p)->prio;
+ u8 sequence = ((const struct can_tx_acknowledge*)p)->sequence;
+ u8 status = ((const struct can_tx_acknowledge*)p)->status;
+ u8 space = ((const struct can_tx_acknowledge*)p)->space;
+ bool lost_seq = false;
+
+ if (port >= BCP_CAN_PORTS || prio >= BCP_CAN_PRIOS)
+ return;
+
+ /* local_txq will be decreased after kernel sent data to companion,
+ * remote_txq will be increased after companion send ack to kernel,
+ * so local_txq < remote_txq should be guaranteed. Otherwise, kernel
+ * received unexpected ack, hence ignore it.
+ */
+ pm->remote_txq[port][prio] = space;
+ if (pm->local_txq[port][prio] >= space) {
+ pm->stats.can_ack_unexpect[port][prio]++;
+ return;
+ }
+
+ pm->local_txq[port][prio]++;
+ pm->sequence[port][prio]++;
+ if (pm->sequence[port][prio] != sequence) {
+ lost_seq = true;
+ pm->sequence[port][prio] = sequence;
+ pm->stats.can_lost_seq_sync[port][prio]++;
+ }
+
+ down_read(&pm->can_lock[port]);
+ if (pm->can_ops[port] && pm->can_ops[port]->on_tx_done)
+ pm->can_ops[port]->on_tx_done(pm->can_data[port],
+ prio,
+ lost_seq,
+ BCP_STATUS_SUCCESS == status);
+ up_read(&pm->can_lock[port]);
+
+ if (BCP_STATUS_SUCCESS == status)
+ pm->stats.can_ack_success[port][prio]++;
+ else
+ pm->stats.can_ack_failure[port][prio]++;
+}
+
+/**
+ * can_txq_status_match() - match CAN tx queue status packet
+ * @p: address of the packet to handle
+ */
+static bool can_txq_status_match(const struct companion_packet *p)
+{
+ return ((const struct can_txq_status_response*)p)->type ==
+ BCP_CAN_TX_QUEUE_STATUS;
+}
+
+/**
+ * can_txq_status_process() - process CAN tx queue status packet
+ * @pm: address of the protocol manager
+ * @p: address of the packet to handle
+ */
+static void can_txq_status_process(struct companion_protocol_manager *pm,
+ const struct companion_packet *p)
+{
+ u8 port = ((const struct can_txq_status_response*)p)->port - 1;
+ const u8 *space = ((const struct can_txq_status_response*)p)->space;
+ u8 status = ((const struct can_txq_status_response*)p)->status;
+
+ if (port >= BCP_CAN_PORTS)
+ return;
+
+ if (BCP_STATUS_SUCCESS == status) {
+ memcpy(pm->remote_txq[port], space, BCP_CAN_PRIOS);
+ pm->response[port][bcp_can_txq_status] = 0;
+ pm->stats.can_ack_success[port][0]++;
+ } else {
+ pm->response[port][bcp_can_txq_status] = -EINVAL;
+ pm->stats.can_ack_failure[port][0]++;
+ }
+
+ if (test_and_clear_bit(bcp_can_txq_status, &pm->flags[port]))
+ wake_up_interruptible_all(&pm->wait[port]);
+}
+
+/**
+ * unknown_match() - match unknown packet
+ * @p: address of the packet to handle
+ */
+static bool unknown_match(const struct companion_packet *p)
+{
+ return true;
+}
+
+/**
+ * unknown_process() - process unknown packet
+ * @pm: address of the protocol manager
+ * @p: address of the packet to handle
+ */
+static void unknown_process(struct companion_protocol_manager *pm,
+ const struct companion_packet *p)
+{
+ dump_packet(p, KERN_ERR, "Unkown packet: ");
+}
+
+static struct companion_filter null_filter = {
+ .node = LIST_HEAD_INIT(null_filter.node),
+ .match = null_match,
+ .process = NULL,
+};
+
+static struct companion_filter io_filter = {
+ .node = LIST_HEAD_INIT(io_filter.node),
+ .match = io_match,
+ .process = io_process,
+};
+
+static struct companion_filter can_data_filter = {
+ .node = LIST_HEAD_INIT(can_data_filter.node),
+ .match = can_data_match,
+ .process = can_data_process,
+};
+
+static struct companion_filter can_bittiming_filter = {
+ .node = LIST_HEAD_INIT(can_bittiming_filter.node),
+ .match = can_bittiming_match,
+ .process = can_bittiming_process,
+};
+
+static struct companion_filter can_mode_filter = {
+ .node = LIST_HEAD_INIT(can_mode_filter.node),
+ .match = can_mode_match,
+ .process = can_mode_process,
+};
+
+static struct companion_filter can_status_filter = {
+ .node = LIST_HEAD_INIT(can_status_filter.node),
+ .match = can_status_match,
+ .process = can_status_process,
+};
+
+static struct companion_filter can_tx_ack_filter = {
+ .node = LIST_HEAD_INIT(can_tx_ack_filter.node),
+ .match = can_tx_ack_match,
+ .process = can_tx_ack_process,
+};
+
+static struct companion_filter can_txq_status_filter = {
+ .node = LIST_HEAD_INIT(can_txq_status_filter.node),
+ .match = can_txq_status_match,
+ .process = can_txq_status_process,
+};
+
+static struct companion_filter unknown_filter = {
+ .node = LIST_HEAD_INIT(unknown_filter.node),
+ .match = unknown_match,
+ .process = unknown_process,
+};
+
+/**
+ * to_timer_data() - helper to access companion_timer_data in work_struct
+ * @ws: address of the work_struct object
+ */
+static struct companion_timer_data* to_timer_data(struct work_struct *ws)
+{
+ return container_of(ws, struct companion_timer_data, work);
+}
+
+/**
+ * pm_can_tx_timeout_callback() - CAN tx timeout callback
+ * @ws: address of the work_struct object
+ */
+static void pm_can_tx_timeout_callback(struct work_struct *ws)
+{
+ struct companion_timer_data *td = to_timer_data(ws);
+ struct companion_protocol_manager *pm = td->pm;
+ u8 port = td->port;
+ u8 prio = td->prio;
+
+ down_read(&pm->can_lock[port]);
+ if (pm->can_ops[port] && pm->can_ops[port]->on_tx_timeout)
+ pm->can_ops[port]->on_tx_timeout(pm->can_data[port], prio);
+ up_read(&pm->can_lock[port]);
+}
+
+/**
+ * pm_can_on_tx_timeout() - CAN tx timeout handler
+ * @tl: address of the timer_list object
+ */
+static void pm_can_on_tx_timeout(struct timer_list *tl)
+{
+ struct companion_timer *ct = from_timer(ct, tl, timer);
+ schedule_work(&ct->data.work);
+}
+
+#define CHECK_SIZE(x) BUILD_BUG_ON(sizeof(struct companion_packet) != \
+ sizeof(x))
+
+/**
+ * pm_init() - initialize the protocol manager
+ * @pm: address of the protocol manager to be initialized
+ */
+void pm_init(struct companion_protocol_manager *pm)
+{
+ int i;
+
+ /* sanity check for correct packet size at compile time */
+ CHECK_SIZE(struct can_data_frame);
+ CHECK_SIZE(struct can_bittiming_request);
+ CHECK_SIZE(struct can_bittiming_response);
+ CHECK_SIZE(struct can_mode_request);
+ CHECK_SIZE(struct can_mode_response);
+ CHECK_SIZE(struct can_status_request);
+ CHECK_SIZE(struct can_status_response);
+ CHECK_SIZE(struct can_tx_acknowledge);
+ CHECK_SIZE(struct can_txq_status_request);
+ CHECK_SIZE(struct can_txq_status_response);
+
+
+ init_rwsem(&pm->io_lock);
+ for (i = 0; i < BCP_CAN_PORTS; ++i) {
+ init_rwsem(&pm->can_lock[i]);
+ init_waitqueue_head(&pm->wait[i]);
+ }
+
+ qm_init(&pm->qm);
+
+ INIT_LIST_HEAD(&pm->filters);
+ list_add_tail(&null_filter.node, &pm->filters);
+ list_add_tail(&io_filter.node, &pm->filters);
+ list_add_tail(&can_data_filter.node, &pm->filters);
+ list_add_tail(&can_tx_ack_filter.node, &pm->filters);
+ list_add_tail(&can_bittiming_filter.node, &pm->filters);
+ list_add_tail(&can_mode_filter.node, &pm->filters);
+ list_add_tail(&can_status_filter.node, &pm->filters);
+ list_add_tail(&can_txq_status_filter.node, &pm->filters);
+ list_add_tail(&unknown_filter.node, &pm->filters);
+}
+
+/**
+ * pm_io_ops_register() - register companion IO packets handler
+ * @pm: address of the protocol manager to be registered
+ * @ops: address of the IO packets callback
+ * @data: address of the IO packets callback argument
+ */
+int pm_io_ops_register(struct companion_protocol_manager *pm,
+ struct companion_io_ops *ops,
+ void *data)
+{
+ int result = 0;
+
+ down_write(&pm->io_lock);
+ if (pm->io_ops) {
+ result = -EEXIST;
+ goto out;
+ }
+
+ qm_reset_io(&pm->qm);
+ pm->io_ops = ops;
+ pm->io_data = data;
+
+out:
+ up_write(&pm->io_lock);
+ return result;
+}
+
+/**
+ * pm_io_ops_unregister() - unregister companion IO packets handler
+ * @pm: address of the protocol manager to be unregistered
+ */
+int pm_io_ops_unregister(struct companion_protocol_manager *pm)
+{
+ int result = 0;
+
+ down_write(&pm->io_lock);
+ if (!pm->io_ops) {
+ result = -ENODEV;
+ goto out;
+ }
+
+ pm->io_ops = NULL;
+ pm->io_data = NULL;
+ qm_reset_io(&pm->qm);
+
+out:
+ up_write(&pm->io_lock);
+ return result;
+}
+
+/**
+ * pm_can_ops_register() - register companion CAN packets handler
+ * @pm: address of the protocol manager to be registered
+ * @port: port number of which CAN to be registered
+ * @ops: address of the CAN packets callback
+ * @data: address of the CAN packets callback argument
+ */
+int pm_can_ops_register(struct companion_protocol_manager *pm,
+ u8 port,
+ struct companion_can_ops *ops,
+ void *data)
+{
+ int i, result = 0;
+
+ if (port >= BCP_CAN_PORTS)
+ return -EINVAL;
+
+ down_write(&pm->can_lock[port]);
+ if (pm->can_ops[port]) {
+ result = -EEXIST;
+ goto out;
+ }
+
+ qm_reset_can(&pm->qm, port);
+ pm->can_ops[port] = ops;
+ pm->can_data[port] = data;
+
+ for (i = 0; i < BCP_CAN_PRIOS; ++i) {
+ pm->timer[port][i].data.pm = pm;
+ pm->timer[port][i].data.port = port;
+ pm->timer[port][i].data.prio = i;
+ INIT_WORK(&pm->timer[port][i].data.work,
+ pm_can_tx_timeout_callback);
+ timer_setup(&pm->timer[port][i].timer, pm_can_on_tx_timeout, 0);
+ }
+
+out:
+ up_write(&pm->can_lock[port]);
+ return result;
+}
+
+/**
+ * pm_can_ops_unregister() - unregister companion CAN packets handler
+ * @pm: address of the protocol manager to be unregistered
+ * @port: port number of which CAN to be unregistered
+ */
+int pm_can_ops_unregister(struct companion_protocol_manager *pm,
+ u8 port)
+{
+ int i, result = 0;
+
+ if (port >= BCP_CAN_PORTS)
+ return -EINVAL;
+
+ down_write(&pm->can_lock[port]);
+ if (!pm->can_ops[port]) {
+ result = -ENODEV;
+ goto out;
+ }
+
+ pm->can_ops[port] = NULL;
+ pm->can_data[port] = NULL;
+ qm_reset_can(&pm->qm, port);
+
+ for (i = 0; i < BCP_CAN_PRIOS; ++i) {
+ del_timer_sync(&pm->timer[port][i].timer);
+ cancel_work_sync(&pm->timer[port][i].data.work);
+ pm->timer[port][i].data.pm = NULL;
+ pm->timer[port][i].data.port = 0;
+ pm->timer[port][i].data.prio = 0;
+ }
+
+out:
+ up_write(&pm->can_lock[port]);
+ return result;
+}
+
+/**
+ * pm_prepare_tx() - prepare tx data
+ * @pm: address of the protocol manager to be used
+ * @p: address of the data to be sent
+ */
+void pm_prepare_tx(struct companion_protocol_manager *pm,
+ struct companion_packet *p)
+{
+ pm->is_io_type = false;
+
+ if (qm_get_tx_data(&pm->qm, p)) {
+ if (is_io_type(p))
+ pm->is_io_type = true;
+ } else {
+ memset(p, BCP_NOOP, sizeof(*p));
+ }
+}
+
+/**
+ * pm_on_tx_done() - handle tx done
+ * @pm: address of the protocol manager to be used
+ */
+void pm_on_tx_done(struct companion_protocol_manager *pm)
+{
+ if (!pm->is_io_type)
+ return;
+
+ down_read(&pm->io_lock);
+ if (pm->io_ops && pm->io_ops->on_tx_done)
+ pm->io_ops->on_tx_done(pm->io_data);
+ up_read(&pm->io_lock);
+}
+
+/**
+ * pm_on_rx_done() - handle rx done
+ * @pm: address of the protocol manager to be used
+ * @p: address of the recevied data
+ */
+void pm_on_rx_done(struct companion_protocol_manager *pm,
+ const struct companion_packet *p)
+{
+ struct companion_filter *filter;
+
+ list_for_each_entry(filter, &pm->filters, node) {
+ if (filter->match && filter->match(p)) {
+ if (filter->process)
+ filter->process(pm, p);
+
+ break;
+ }
+ }
+}
+
+/**
+ * pm_can_data_tx() - send CAN data according to protocol
+ * @pm: address of the protocol manager to be used
+ * @port: port number of which CAN to be sent
+ * @prio: priority of the data to be sent
+ * @cf: the raw CAN frame to be send
+ */
+int pm_can_data_tx(struct companion_protocol_manager *pm,
+ u8 port,
+ u8 prio,
+ const struct can_frame *cf)
+{
+ struct can_data_frame p;
+
+ if (port >= BCP_CAN_PORTS || prio >= BCP_CAN_PRIOS)
+ return -EINVAL;
+
+ if (pm->local_txq[port][prio] == 0)
+ return -ENOSPC;
+
+ p.type = BCP_CAN_DATA;
+ p.port = port + 1;
+ p.prio = prio;
+ p.dlc = cf->can_dlc;
+ p.id = cf->can_id;
+ memcpy(p.data, cf->data, sizeof(cf->data));
+
+ if (!qm_can_txq_in(&pm->qm,
+ (struct companion_packet*)&p,
+ port,
+ prio)) {
+ pm->stats.can_tx_overflows[port][prio]++;
+ return -ENOSPC;
+ }
+
+ pm->local_txq[port][prio]--;
+
+ pm->stats.can_tx[port][prio]++;
+ return 0;
+}
+
+/**
+ * pm_can_data_rx() - receive CAN data according to protocol
+ * @pm: address of the protocol manager to be used
+ * @port: port number of which CAN to be received
+ * @cf: address of the raw CAN frame to be copied
+ */
+int pm_can_data_rx(struct companion_protocol_manager *pm,
+ u8 port,
+ struct can_frame *cf)
+{
+ struct can_data_frame p;
+
+ if (port >= BCP_CAN_PORTS)
+ return -EINVAL;
+
+ if (!qm_can_rxq_out(&pm->qm, (struct companion_packet*)&p, port))
+ return -EIO;
+
+ cf->can_id = p.id;
+ cf->can_dlc = p.dlc;
+ memcpy(cf->data, p.data, sizeof(cf->data));
+ return 0;
+}
+
+/**
+ * pm_can_err() - receive CAN error according to protocol
+ * @pm: address of the protocol manager to be used
+ * @port: port number of which CAN to be received
+ * @bec: address of the error counter to be copied
+ * @state: address of the error state to be copied
+ * @code: address of the error code to be copied
+ */
+int pm_can_err(struct companion_protocol_manager *pm,
+ u8 port,
+ struct can_berr_counter *berr,
+ u8 *state,
+ u8 *code)
+{
+ struct can_status_response p;
+
+ if (port >= BCP_CAN_PORTS)
+ return -EINVAL;
+
+ if (!qm_can_err_out(&pm->qm, (struct companion_packet*)&p, port))
+ return -EIO;
+
+ berr->rxerr = p.rx_err;
+ berr->txerr = p.tx_err;
+ *state = p.state;
+ *code = p.code;
+ return 0;
+}
+
+/**
+ * pm_wait_for_response() - wait for CAN packets response
+ * @pm: address of the protocol manager to be used
+ * @port: port number of which CAN to be wait
+ * @flag: flag to be wait
+ */
+int pm_wait_for_response(struct companion_protocol_manager *pm,
+ u8 port,
+ enum bcp_can_flag flag)
+{
+ unsigned long *flags = &pm->flags[port];
+ long ret;
+
+ if (test_bit(flag, flags)) {
+ ret = wait_event_interruptible_timeout(pm->wait[port],
+ !test_bit(flag, flags),
+ PM_RESPONSE_TIMEOUT);
+
+ if (ret < 0)
+ return ret;
+
+ if (ret == 0)
+ return -ETIMEDOUT;
+ }
+
+ return pm->response[port][flag];
+}
+
+/**
+ * pm_can_set_bittiming() - set CAN bittiming according to protocol
+ * @pm: address of the protocol manager to be used
+ * @port: port number of which CAN to be set
+ * @bittiming: the bittiming to be set
+ */
+int pm_can_set_bittiming(struct companion_protocol_manager *pm,
+ u8 port,
+ const struct can_bittiming *bittiming)
+{
+ struct can_bittiming_request p;
+
+ if (port >= BCP_CAN_PORTS)
+ return -EINVAL;
+
+ memset(&p, 0, sizeof(p));
+ p.type = BCP_CAN_BITTIMING;
+ p.port = port + 1;
+ p.prescaler = bittiming->brp;
+ p.prop_seg = bittiming->prop_seg;
+ p.phase_seg1 = bittiming->phase_seg1;
+ p.phase_seg2 = bittiming->phase_seg2;
+ p.sjw = bittiming->sjw;
+
+ if (!qm_can_txq_in(&pm->qm,
+ (struct companion_packet*)&p,
+ port,
+ 0)) {
+ pm->stats.can_tx_overflows[port][0]++;
+ return -ENOSPC;
+ }
+
+ set_bit(bcp_can_bittiming, &pm->flags[port]);
+ pm->stats.can_tx[port][0]++;
+ return 0;
+}
+
+/**
+ * pm_can_set_mode() - set CAN mode according to protocol
+ * @pm: address of the protocol manager to be used
+ * @port: port number of which CAN to be set
+ * @mode: the mode to be set
+ */
+int pm_can_set_mode(struct companion_protocol_manager *pm,
+ u8 port,
+ enum can_mode mode)
+{
+ struct can_mode_request p;
+
+ if (port >= BCP_CAN_PORTS)
+ return -EINVAL;
+
+ memset(&p, 0, sizeof(p));
+ p.type = BCP_CAN_MODE;
+ p.port = port + 1;
+
+ switch (mode) {
+ case CAN_MODE_START:
+ p.mode = BCP_CAN_MODE_NORMAL;
+ break;
+
+ case CAN_MODE_STOP:
+ p.mode = BCP_CAN_MODE_OFF;
+ break;
+
+ default:
+ return -EOPNOTSUPP;
+ }
+
+ if (!qm_can_txq_in(&pm->qm,
+ (struct companion_packet*)&p,
+ port,
+ 0)) {
+ pm->stats.can_tx_overflows[port][0]++;
+ return -ENOSPC;
+ }
+
+ set_bit(bcp_can_mode, &pm->flags[port]);
+ pm->stats.can_tx[port][0]++;
+ return 0;
+}
+
+/**
+ * pm_can_set_ctrlmode() - set CAN control mode according to protocol
+ * @pm: address of the protocol manager to be used
+ * @port: port number of which CAN to be set
+ * @ctrl: the control mode to be set
+ */
+int pm_can_set_ctrlmode(struct companion_protocol_manager *pm,
+ u8 port,
+ u32 ctrl)
+{
+ struct can_mode_request p;
+
+ if (port >= BCP_CAN_PORTS)
+ return -EINVAL;
+
+ memset(&p, 0, sizeof(p));
+ p.type = BCP_CAN_MODE;
+ p.port = port + 1;
+
+ if (ctrl & CAN_CTRLMODE_LISTENONLY)
+ p.mode = BCP_CAN_MODE_LISTEN;
+ else
+ return -EOPNOTSUPP;
+
+ if (!qm_can_txq_in(&pm->qm,
+ (struct companion_packet*)&p,
+ port,
+ 0)) {
+ pm->stats.can_tx_overflows[port][0]++;
+ return -ENOSPC;
+ }
+
+ set_bit(bcp_can_mode, &pm->flags[port]);
+ pm->stats.can_tx[port][0]++;
+ return 0;
+}
+
+/**
+ * pm_can_get_status() - get CAN status according to protocol
+ * @pm: address of the protocol manager to be used
+ * @port: port number of which CAN to be inquiry
+ */
+int pm_can_get_status(struct companion_protocol_manager *pm,
+ u8 port)
+{
+ struct can_status_request p;
+
+ if (port >= BCP_CAN_PORTS)
+ return -EINVAL;
+
+ memset(&p, 0, sizeof(p));
+ p.type = BCP_CAN_STATUS;
+ p.port = port + 1;
+
+ if (!qm_can_txq_in(&pm->qm,
+ (struct companion_packet*)&p,
+ port,
+ 0)) {
+ pm->stats.can_tx_overflows[port][0]++;
+ return -ENOSPC;
+ }
+
+ set_bit(bcp_can_status, &pm->flags[port]);
+ pm->stats.can_tx[port][0]++;
+ return 0;
+}
+
+/**
+ * pm_can_get_txq_status() - get CAN tx queue status according to protocol
+ * @pm: address of the protocol manager to be used
+ * @port: port number of which CAN to be inquiry
+ */
+int pm_can_get_txq_status(struct companion_protocol_manager *pm,
+ u8 port)
+{
+ struct can_txq_status_request p;
+
+ if (port >= BCP_CAN_PORTS)
+ return -EINVAL;
+
+ memset(&p, 0, sizeof(p));
+ p.type = BCP_CAN_TX_QUEUE_STATUS;
+ p.port = port + 1;
+
+ if (!qm_can_txq_in(&pm->qm,
+ (struct companion_packet*)&p,
+ port,
+ 0)) {
+ pm->stats.can_tx_overflows[port][0]++;
+ return -ENOSPC;
+ }
+
+ set_bit(bcp_can_txq_status, &pm->flags[port]);
+ pm->stats.can_tx[port][0]++;
+ return 0;
+}
+
+/**
+ * pm_can_txq_is_full() - inquiry CAN tx queue is full
+ * @pm: address of the protocol manager to be used
+ * @port: port number of which CAN to be inquiry
+ * @prio: queue number of which tx queue to be inquiry
+ * @is_full: address of the is full result to be copied
+ */
+int pm_can_txq_is_full(struct companion_protocol_manager *pm,
+ u8 port,
+ u8 prio,
+ bool *is_full)
+{
+ if (port >= BCP_CAN_PORTS || prio >= BCP_CAN_PRIOS)
+ return -EINVAL;
+
+ *is_full = (pm->local_txq[port][prio] == 0);
+ return 0;
+}
+
+/**
+ * pm_can_txq_has_space() - inquiry CAN tx queue has space
+ * @pm: address of the protocol manager to be used
+ * @port: port number of which CAN to be inquiry
+ * @prio: queue number of which tx queue to be inquiry
+ * @has_space: address of the has space result to be copied
+ */
+int pm_can_txq_has_space(struct companion_protocol_manager *pm,
+ u8 port,
+ u8 prio,
+ bool *has_space)
+{
+ if (port >= BCP_CAN_PORTS || prio >= BCP_CAN_PRIOS)
+ return -EINVAL;
+
+ *has_space = (pm->local_txq[port][prio] > 0);
+ return 0;
+}
+
+/**
+ * pm_can_start_tx_timer() - start CAN tx timeout detection
+ * @pm: address of the protocol manager to be used
+ * @port: port number of which CAN to be used
+ * @prio: queue number of which tx queue's timer to be used
+ */
+int pm_can_start_tx_timer(struct companion_protocol_manager *pm,
+ u8 port,
+ u8 prio)
+{
+ if (port >= BCP_CAN_PORTS || prio >= BCP_CAN_PRIOS)
+ return -EINVAL;
+
+ mod_timer(&pm->timer[port][prio].timer, jiffies + PM_CAN_TX_TIMEOUT);
+ return 0;
+}
+
+/**
+ * pm_can_stop_tx_timer() - stop CAN tx timeout detection
+ * @pm: address of the protocol manager to be used
+ * @port: port number of which CAN to be used
+ * @prio: queue number of which tx queue's timer to be used
+ */
+int pm_can_stop_tx_timer(struct companion_protocol_manager *pm,
+ u8 port,
+ u8 prio)
+{
+ if (port >= BCP_CAN_PORTS || prio >= BCP_CAN_PRIOS)
+ return -EINVAL;
+
+ del_timer_sync(&pm->timer[port][prio].timer);
+ return 0;
+}
diff --git a/drivers/spi/companion/protocol-manager.h b/drivers/spi/companion/protocol-manager.h
new file mode 100644
index 0000000..09d32ad
--- /dev/null
+++ b/drivers/spi/companion/protocol-manager.h
@@ -0,0 +1,348 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Companion protocol manager code
+ *
+ * Copyright (C) 2015-2018 Bosch Sicherheitssysteme GmbH
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#ifndef _BOSCH_COMPANION_PROTOCOL_MANAGER_H
+#define _BOSCH_COMPANION_PROTOCOL_MANAGER_H
+
+#include <linux/can/dev.h>
+#include <linux/list.h>
+#include <linux/rwsem.h>
+#include <linux/timer.h>
+#include <linux/workqueue.h>
+#include <linux/companion.h>
+#include "queue-manager.h"
+
+/**
+ * enum bcp_can_flag - companion CAN packet event flag
+ */
+enum bcp_can_flag {
+ bcp_can_bittiming,
+ bcp_can_mode,
+ bcp_can_status,
+ bcp_can_txq_status,
+
+ bcp_can_max
+};
+
+/**
+ * struct companion_statistics - statistics information of companion
+ * @io_tx_overflows: counter of IO packets transmit overflow
+ * @io_rx_overflows: counter of IO packets receive overflow
+ * @can_tx_overflows: counter of CAN packets transmit overflow
+ * @can_rx_overflows: counter of CAN packets receive overflow
+ * @can_err_overflows: counter of CAN error packets receive overflow
+ * @io_tx: counter of IO packets transmitted
+ * @io_rx: counter of IO packets received
+ * @can_tx: counter of CAN packets transmitted
+ * @can_ack_success: counter of CAN response success packets received
+ * @can_ack_failure: counter of CAN response failure packets received
+ * @can_rx: counter of CAN packets received
+ * @can_err: counter of CAN error packets received
+ * @can_lost_seq_sync: counter of CAN packets sequence lost sync
+ * @can_lost_txq_sync: counter of CAN tx queue status lost sync
+ * @can_ack_timeout: counter of CAN tx ack timeout
+ * @can_ack_unexpect: counter of CAN unexpected tx ack
+ *
+ * TODO: add more statistics fields and export to sysfs
+ */
+struct companion_statistics {
+ u32 io_tx_overflows;
+ u32 io_rx_overflows;
+ u32 can_tx_overflows[BCP_CAN_PORTS][BCP_CAN_PRIOS];
+ u32 can_rx_overflows[BCP_CAN_PORTS];
+ u32 can_err_overflows[BCP_CAN_PORTS];
+
+ u32 io_tx;
+ u32 io_rx;
+ u32 can_tx[BCP_CAN_PORTS][BCP_CAN_PRIOS];
+ u32 can_ack_success[BCP_CAN_PORTS][BCP_CAN_PRIOS];
+ u32 can_ack_failure[BCP_CAN_PORTS][BCP_CAN_PRIOS];
+ u32 can_rx[BCP_CAN_PORTS];
+ u32 can_err[BCP_CAN_PORTS];
+ u32 can_lost_seq_sync[BCP_CAN_PORTS][BCP_CAN_PRIOS];
+ u32 can_lost_txq_sync[BCP_CAN_PORTS][BCP_CAN_PRIOS];
+ u32 can_ack_timeout[BCP_CAN_PORTS][BCP_CAN_PRIOS];
+ u32 can_ack_unexpect[BCP_CAN_PORTS][BCP_CAN_PRIOS];
+};
+
+struct companion_protocol_manager;
+/**
+ * struct companion_timer_data - encapsulate data passed to timer
+ * @pm: companion protocol manager
+ * @port: port number of which CAN associated with the timer
+ * @prio: priority of which CAN tx queue associated with the timer
+ * @work: work for execute timeout callback in thread context
+ */
+struct companion_timer_data {
+ struct companion_protocol_manager *pm;
+ u8 port;
+ u8 prio;
+ struct work_struct work;
+};
+
+/**
+ * struct companion_timer - encapsulate timer and data
+ * @timer: timer_list object for timeout
+ * @data: data associated with the timer
+ */
+struct companion_timer {
+ struct timer_list timer;
+ struct companion_timer_data data;
+};
+
+/**
+ * struct companion_protocol_manager - encapsulate companion protocol handling
+ * @io_ops: callback of IO packet handling
+ * @io_data: callback argument of io_ops
+ * @io_lock: lock to protect IO callback
+ * @can_ops: callback of CAN packet handling
+ * @can_data: callback argument of can_ops
+ * @can_lock: lock to protect CAN callback
+ * @wait: wait queue head for CAN packet response
+ * @flags: event flags for CAN packet response
+ * @response: response result of each CAN packet event flag
+ * @rx_err: receive error counter of CAN port
+ * @tx_err: transmit error counter of CAN port
+ * @sequence: sequence counter of each CAN tx queue
+ * @remote_txq: CAN tx queue space status at remote
+ * @local_txq: CAN tx queue space status at local
+ * @timer: timer for detect CAN tx ack timeout for each CAN tx queue
+ * @stats: statistics information of companion
+ * @is_io_type: flag to record the sent packet is IO type or not
+ * @qm: queue manager
+ * @filters: filter list to handle receveid companion packets
+ *
+ * TODO: re-think the data structure for handle CAN response
+ */
+struct companion_protocol_manager {
+ struct companion_io_ops *io_ops;
+ void *io_data;
+ struct rw_semaphore io_lock;
+ struct companion_can_ops *can_ops[BCP_CAN_PORTS];
+ void *can_data[BCP_CAN_PORTS];
+ struct rw_semaphore can_lock[BCP_CAN_PORTS];
+ wait_queue_head_t wait[BCP_CAN_PORTS];
+ unsigned long flags[BCP_CAN_PORTS];
+ int response[BCP_CAN_PORTS][bcp_can_max];
+ u8 rx_err[BCP_CAN_PORTS];
+ u8 tx_err[BCP_CAN_PORTS];
+ u8 sequence[BCP_CAN_PORTS][BCP_CAN_PRIOS];
+ u8 remote_txq[BCP_CAN_PORTS][BCP_CAN_PRIOS];
+ u8 local_txq[BCP_CAN_PORTS][BCP_CAN_PRIOS];
+ struct companion_timer timer[BCP_CAN_PORTS][BCP_CAN_PRIOS];
+
+ struct companion_statistics stats;
+ bool is_io_type;
+
+ struct companion_queue_manager qm;
+ struct list_head filters;
+};
+
+/**
+ * pm_init() - initialize the protocol manager
+ * @pm: address of the protocol manager to be initialized
+ */
+void pm_init(struct companion_protocol_manager *pm);
+
+/**
+ * pm_io_ops_register() - register companion IO packets handler
+ * @pm: address of the protocol manager to be registered
+ * @ops: address of the IO packets callback
+ * @data: address of the IO packets callback argument
+ */
+int pm_io_ops_register(struct companion_protocol_manager *pm,
+ struct companion_io_ops *ops,
+ void *data);
+
+/**
+ * pm_io_ops_unregister() - unregister companion IO packets handler
+ * @pm: address of the protocol manager to be unregistered
+ */
+int pm_io_ops_unregister(struct companion_protocol_manager *pm);
+
+/**
+ * pm_can_ops_register() - register companion CAN packets hanler
+ * @pm: address of the protocol manager to be registered
+ * @port: port number of which CAN to be registered
+ * @ops: address of the CAN packets callback
+ * @data: address of the CAN packets callback argument
+ */
+int pm_can_ops_register(struct companion_protocol_manager *pm,
+ u8 port,
+ struct companion_can_ops *ops,
+ void *data);
+
+/**
+ * pm_can_ops_unregister() - unregister companion CAN packets handler
+ * @pm: address of the protocol manager to be unregistered
+ * @port: port number of which CAN to be unregistered
+ */
+int pm_can_ops_unregister(struct companion_protocol_manager *pm,
+ u8 port);
+
+/**
+ * pm_prepare_tx() - prepare tx data
+ * @pm: address of the protocol manager to be used
+ * @p: address of the data to be sent
+ */
+void pm_prepare_tx(struct companion_protocol_manager *pm,
+ struct companion_packet *p);
+
+/**
+ * pm_on_tx_done() - handle tx done
+ * @pm: address of the protocol manager to be used
+ */
+void pm_on_tx_done(struct companion_protocol_manager *pm);
+
+/**
+ * pm_on_rx_done() - handle rx done
+ * @pm: address of the protocol manager to be used
+ * @p: address of the recevied data
+ */
+void pm_on_rx_done(struct companion_protocol_manager *pm,
+ const struct companion_packet *p);
+
+/**
+ * pm_can_data_tx() - send CAN data according to protocol
+ * @pm: address of the protocol manager to be used
+ * @port: port number of which CAN to be sent
+ * @prio: priority of the data to be sent
+ * @cf: the raw CAN frame to be send
+ */
+int pm_can_data_tx(struct companion_protocol_manager *pm,
+ u8 port,
+ u8 prio,
+ const struct can_frame *cf);
+
+/**
+ * pm_can_data_rx() - receive CAN data according to protocol
+ * @pm: address of the protocol manager to be used
+ * @port: port number of which CAN to be received
+ * @cf: address of the raw CAN frame to be copied
+ */
+int pm_can_data_rx(struct companion_protocol_manager *pm,
+ u8 port,
+ struct can_frame *cf);
+
+/**
+ * pm_can_err() - receive CAN error according to protocol
+ * @pm: address of the protocol manager to be used
+ * @port: port number of which CAN to be received
+ * @bec: address of the error counter to be copied
+ * @state: address of the error state to be copied
+ * @code: address of the error code to be copied
+ */
+int pm_can_err(struct companion_protocol_manager *pm,
+ u8 port,
+ struct can_berr_counter *bec,
+ u8 *state,
+ u8 *code);
+
+/**
+ * pm_wait_for_response() - wait for CAN packets response
+ * @pm: address of the protocol manager to be used
+ * @port: port number of which CAN to be wait
+ * @flag: flag to be wait
+ */
+int pm_wait_for_response(struct companion_protocol_manager *pm,
+ u8 port,
+ enum bcp_can_flag flag);
+
+/**
+ * pm_can_set_bittiming() - set CAN bittiming according to protocol
+ * @pm: address of the protocol manager to be used
+ * @port: port number of which CAN to be set
+ * @bittiming: the bittiming to be set
+ */
+int pm_can_set_bittiming(struct companion_protocol_manager *pm,
+ u8 port,
+ const struct can_bittiming *bittiming);
+
+/**
+ * pm_can_set_mode() - set CAN mode according to protocol
+ * @pm: address of the protocol manager to be used
+ * @port: port number of which CAN to be set
+ * @mode: the mode to be set
+ */
+int pm_can_set_mode(struct companion_protocol_manager *pm,
+ u8 port,
+ enum can_mode mode);
+
+/**
+ * pm_can_set_ctrlmode() - set CAN control mode according to protocol
+ * @pm: address of the protocol manager to be used
+ * @port: port number of which CAN to be set
+ * @ctrl: the control mode to be set
+ */
+int pm_can_set_ctrlmode(struct companion_protocol_manager *pm,
+ u8 port,
+ u32 ctrl);
+
+/**
+ * pm_can_get_status() - get CAN status according to protocol
+ * @pm: address of the protocol manager to be used
+ * @port: port number of which CAN to be inquiry
+ */
+int pm_can_get_status(struct companion_protocol_manager *pm,
+ u8 port);
+
+/**
+ * pm_can_get_txq_status() - get CAN tx queue status according to protocol
+ * @pm: address of the protocol manager to be used
+ * @port: port number of which CAN to be inquiry
+ */
+int pm_can_get_txq_status(struct companion_protocol_manager *pm,
+ u8 port);
+
+/**
+ * pm_can_txq_is_full() - inquiry CAN tx queue is full
+ * @pm: address of the protocol manager to be used
+ * @port: port number of which CAN to be inquiry
+ * @prio: queue number of which tx queue to be inquiry
+ * @is_full: address of the is full result to be copied
+ */
+int pm_can_txq_is_full(struct companion_protocol_manager *pm,
+ u8 port,
+ u8 prio,
+ bool *is_full);
+
+/**
+ * pm_can_txq_has_space() - inquiry CAN tx queue has space
+ * @pm: address of the protocol manager to be used
+ * @port: port number of which CAN to be inquiry
+ * @prio: queue number of which tx queue to be inquiry
+ * @has_space: address of the has space result to be copied
+ */
+int pm_can_txq_has_space(struct companion_protocol_manager *pm,
+ u8 port,
+ u8 prio,
+ bool *has_space);
+
+/**
+ * pm_can_start_tx_timer() - start CAN tx timeout detection
+ * @pm: address of the protocol manager to be used
+ * @port: port number of which CAN to be used
+ * @prio: queue number of which tx queue's timer to be used
+ */
+int pm_can_start_tx_timer(struct companion_protocol_manager *pm,
+ u8 port,
+ u8 prio);
+
+/**
+ * pm_can_stop_tx_timer() - stop CAN tx timeout detection
+ * @pm: address of the protocol manager to be used
+ * @port: port number of which CAN to be used
+ * @prio: queue number of which tx queue's timer to be used
+ */
+int pm_can_stop_tx_timer(struct companion_protocol_manager *pm,
+ u8 port,
+ u8 prio);
+#endif
diff --git a/drivers/spi/companion/protocol.h b/drivers/spi/companion/protocol.h
new file mode 100644
index 0000000..f426cf0
--- /dev/null
+++ b/drivers/spi/companion/protocol.h
@@ -0,0 +1,273 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Companion protocol kernel space API
+ *
+ * Copyright (C) 2015-2018 Bosch Sicherheitssysteme GmbH
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#ifndef _BOSCH_COMPANION_PROTOCOL_KERNEL_H
+#define _BOSCH_COMPANION_PROTOCOL_KERNEL_H
+
+#include <linux/kernel.h>
+
+#define BCP_CAN_PORTS 2u
+#define BCP_CAN_PRIOS 8u
+
+/**
+ * BCP type field definitions (CAN)
+ */
+#define BCP_NOOP 0x00u
+#define BCP_CAN_DATA 0x01u
+#define BCP_CAN_BITTIMING 0x03u
+#define BCP_CAN_MODE 0x04u
+#define BCP_CAN_STATUS 0x06u
+#define BCP_CAN_TX_ACK 0x07u
+#define BCP_CAN_TX_QUEUE_STATUS 0x0Fu
+
+/**
+ * BCP status field definitions
+ */
+#define BCP_STATUS_SUCCESS 0x00u
+#define BCP_STATUS_UNKNOWN 0x01u
+#define BCP_STATUS_OTHER 0x02u
+
+/**
+ * BCP packet size definition
+ */
+#define BCP_PACKET_SIZE 16u
+
+/**
+ * struct companion_packet - companion packet general format
+ * @data: contents of the packet
+ */
+struct companion_packet {
+ __u8 data[BCP_PACKET_SIZE];
+};
+
+/**
+ * struct can_data_frame - companion can data frame packet
+ * @type: packet type
+ * @port: can port
+ * @prio: priority of the can frame
+ * @dlc: can frame payload in bytes
+ * @id: can frame id
+ * @data: can frame payload
+ */
+struct can_data_frame {
+ u8 type;
+ u8 port;
+ u8 prio;
+ u8 dlc;
+ u32 id;
+ u8 data[8];
+};
+
+/**
+ * struct can_bittiming_request - companion can bittiming request packet
+ * @type: packet type
+ * @port: can port
+ * @prescaler: bitrate prescaler
+ * @prop_seg: propagation segment in TQs
+ * @phase_seg1: phase buffer segment 1 in TQs
+ * @phase_seg2: phase buffer segment 2 in TQs
+ * @sjw: synchronisation jump width in TQs
+ * @reserved: reserved
+ */
+struct can_bittiming_request {
+ u8 type;
+ u8 port;
+ u16 prescaler;
+ u8 prop_seg;
+ u8 phase_seg1;
+ u8 phase_seg2;
+ u8 sjw;
+ u8 reserved[8];
+};
+
+/**
+ * struct can_bittiming_response - companion can bittiming response packet
+ * @type: packet type
+ * @port: can port
+ * @prescaler: bitrate prescaler
+ * @prop_seg: propagation segment in TQs
+ * @phase_seg1: phase buffer segment 1 in TQs
+ * @phase_seg2: phase buffer segment 2 in TQs
+ * @sjw: synchronisation jump width in TQs
+ * @reserved: reserved
+ * @status: process status
+ */
+struct can_bittiming_response {
+ u8 type;
+ u8 port;
+ u16 prescaler;
+ u8 prop_seg;
+ u8 phase_seg1;
+ u8 phase_seg2;
+ u8 sjw;
+ u8 reserved[7];
+ u8 status;
+};
+
+/**
+ * struct can_mode_request - companion can mode request packet
+ * @type: packet type
+ * @port: can port
+ * @mode: can mode
+ * @reserved: reserved
+ */
+struct can_mode_request {
+ u8 type;
+ u8 port;
+ u8 mode;
+ u8 reserved[13];
+};
+#define BCP_CAN_MODE_OFF 0x00u
+#define BCP_CAN_MODE_NORMAL 0x01u
+#define BCP_CAN_MODE_LISTEN 0x02u
+
+/**
+ * struct can_mode_response - companion can mode response packet
+ * @type: packet type
+ * @port: can port
+ * @mode: can mode
+ * @reserved: reserved
+ * @status: process status
+ */
+struct can_mode_response {
+ u8 type;
+ u8 port;
+ u8 mode;
+ u8 reserved[12];
+ u8 status;
+};
+
+/**
+ * struct can_status_request - companion can status request packet
+ * @type: packet type
+ * @port: can port
+ * @reserved: reserved
+ */
+struct can_status_request {
+ u8 type;
+ u8 port;
+ u8 reserved[14];
+};
+
+/**
+ * struct can_status_response - companion can status response packet
+ * @type: packet type
+ * @port: can port
+ * @rx_err: rx error counter
+ * @tx_err: tx error counter
+ * @err1: can controller error status 1
+ * @err2: can controller error status 2
+ * @reserved: reserved
+ * @status: process status
+ */
+struct can_status_response {
+ u8 type;
+ u8 port;
+ u8 rx_err;
+ u8 tx_err;
+ u8 state;
+ u8 code;
+ u8 reserved[9];
+ u8 status;
+};
+
+/**
+ * struct can_tx_acknowledge - companion can tx acknowledge packet
+ * @type: packet type
+ * @port: can port
+ * @prio: priority of the can frame
+ * @dlc: payload length of the can frame
+ * @sequence: monotonic increasing sequence counter of sent can frames
+ * @space: queue space left of this priority
+ * @reserved: reserved
+ * @status: process status
+ */
+struct can_tx_acknowledge {
+ u8 type;
+ u8 port;
+ u8 prio;
+ u8 dlc;
+ u8 sequence;
+ u8 space;
+ u8 reserved[9];
+ u8 status;
+};
+
+/**
+ * struct can_txq_status_request - companion can txq status request packet
+ * @type: packet type
+ * @port: can port
+ * @reserved: reserved
+ */
+struct can_txq_status_request {
+ u8 type;
+ u8 port;
+ u8 reserved[14];
+};
+
+/**
+ * struct can_txq_status_response - companion can txq status response packet
+ * @type: packet type
+ * @port: can port
+ * @space: queue space left of each priority
+ * @reserved: reserved
+ * @status: process status
+ */
+struct can_txq_status_response {
+ u8 type;
+ u8 port;
+ u8 space[8];
+ u8 reserved[5];
+ u8 status;
+};
+
+/**
+ * is_null_type() - return true if the packet is null type
+ * @p: the packet to test
+ */
+static inline bool is_null_type(const struct companion_packet *p)
+{
+ return p->data[0] == BCP_NOOP;
+}
+
+/**
+ * is_io_type() - return true if the packet is io type
+ * @p: the packet to test
+ */
+static inline bool is_io_type(const struct companion_packet *p)
+{
+ return (p->data[0] & 0x80);
+}
+
+/**
+ * is_can_type() - return true if the packet is can type
+ * @p: the packet to test
+ */
+static inline bool is_can_type(const struct companion_packet *p)
+{
+ return !is_io_type(p) && !is_null_type(p);
+}
+
+/**
+ * dump_packet() - dump raw packet data in hexadecimal format
+ * @p: the packet to dump
+ * @level: the log level of the dump
+ * @prefix: the prefix string of the dump
+ */
+static inline void dump_packet(const struct companion_packet *p,
+ const char *level,
+ const char *prefix)
+{
+ print_hex_dump(level, prefix, DUMP_PREFIX_NONE, 16, 1,
+ p->data, sizeof(p->data), false);
+}
+
+#endif
diff --git a/drivers/spi/companion/queue-manager.c b/drivers/spi/companion/queue-manager.c
new file mode 100644
index 0000000..a50646c
--- /dev/null
+++ b/drivers/spi/companion/queue-manager.c
@@ -0,0 +1,146 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Companion queue management code
+ *
+ * Copyright (C) 2015-2018 Bosch Sicherheitssysteme GmbH
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#include "queue-manager.h"
+
+/**
+ * qm_init() - initialize all managed queues
+ * @qm: address of the queue manager to be initialized
+ */
+void qm_init(struct companion_queue_manager *qm)
+{
+ int i, j;
+
+ INIT_KFIFO(qm->io_txq.fifo);
+ INIT_KFIFO(qm->io_rxq.fifo);
+
+ for (i = 0; i < BCP_CAN_PORTS; ++i) {
+ for (j = 0; j < BCP_CAN_PRIOS; ++j)
+ INIT_KFIFO(qm->can_txq[i][j].fifo);
+
+ INIT_KFIFO(qm->can_rxq[i].fifo);
+ INIT_KFIFO(qm->can_err[i].fifo);
+ mutex_init(&qm->can_txq_lock[i]);
+ }
+}
+
+/**
+ * qm_reset() - reset all managed queues
+ * @qm: address of the queue manager to be reset
+ */
+void qm_reset(struct companion_queue_manager *qm)
+{
+ int i;
+
+ qm_reset_io(qm);
+
+ for (i = 0; i < BCP_CAN_PORTS; ++i)
+ qm_reset_can(qm, i);
+}
+
+/**
+ * qm_reset_io() - reset managed IO queues
+ * @qm: address of the queue manager to be reset
+ */
+void qm_reset_io(struct companion_queue_manager *qm)
+{
+ kfifo_reset(&qm->io_txq.fifo);
+ kfifo_reset(&qm->io_rxq.fifo);
+}
+
+/**
+ * qm_reset_can() - reset managed CAN queues
+ * @qm: address of the queue manager to be reset
+ * @port: port number of which CAN queue should be reset
+ */
+void qm_reset_can(struct companion_queue_manager *qm, u8 port)
+{
+ int i;
+
+ for (i = 0; i < BCP_CAN_PRIOS; ++i)
+ kfifo_reset(&qm->can_txq[port][i].fifo);
+
+ kfifo_reset(&qm->can_rxq[port].fifo);
+ kfifo_reset(&qm->can_err[port].fifo);
+}
+
+/**
+ * qm_has_tx_data() - return true if has tx data
+ * @qm: address of the queue manager to be used
+ */
+bool qm_has_tx_data(struct companion_queue_manager *qm)
+{
+ int i, j;
+
+ for (i = 0; i < BCP_CAN_PORTS; ++i)
+ for (j = 0; j < BCP_CAN_PRIOS; ++j)
+ if (!kfifo_is_empty(&qm->can_txq[i][j].fifo))
+ return true;
+
+ return !kfifo_is_empty(&qm->io_txq.fifo);
+}
+
+/*
+ * Define maximum CAN packets can be sent in a row in case there is IO packet
+ * pending or coming, which specifies the minimal bandwidth for IO packets.
+ */
+#define CAN_MAX_IN_A_ROW 8
+
+
+
+/**
+ * qm_get_tx_data() - return true if got the tx data
+ * @qm: address of the queue manager to be used
+ * @p: where the data to be copied
+ */
+bool qm_get_tx_data(struct companion_queue_manager *qm,
+ struct companion_packet *p)
+{
+ int i, j;
+
+ /*
+ * Implement the companion packet scheduling algorithm which guarantees
+ * IO packets share minimal 1 / (CAN_MAX_IN_A_ROW + 1) bandwidth, and
+ * the rest bandwidth is shared equally for all CAN ports.
+ *
+ * The purpose is to ensure fairness between all CAN ports and also keep
+ * CAN packets have higher priority than IO packets in general, but
+ * avoid IO packets starvation in case CAN is very busy.
+ *
+ * The bandwidth is not statically allocated, so the active user (IO or
+ * CAN) can use up to 100% bandwidth if there are no other active users.
+ */
+
+ if (qm->io_promoted && qm_io_txq_out(qm, p)) {
+ qm->io_promoted = false;
+ return true;
+ }
+
+ for (i = 0; i < BCP_CAN_PORTS; ++i) {
+ /* ensure fairness for all can ports */
+ qm->can_current++;
+ if (qm->can_current >= BCP_CAN_PORTS)
+ qm->can_current = 0;
+
+ for (j = 0; j < BCP_CAN_PRIOS; ++j) {
+ if (qm_can_txq_out(qm, p, qm->can_current, j)) {
+ qm->can_sched++;
+ if (qm->can_sched >= CAN_MAX_IN_A_ROW) {
+ qm->io_promoted = true;
+ qm->can_sched = 0;
+ }
+ return true;
+ }
+ }
+ }
+
+ return qm_io_txq_out(qm, p);
+}
diff --git a/drivers/spi/companion/queue-manager.h b/drivers/spi/companion/queue-manager.h
new file mode 100644
index 0000000..3bd77d7
--- /dev/null
+++ b/drivers/spi/companion/queue-manager.h
@@ -0,0 +1,245 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Companion queue management code
+ *
+ * Copyright (C) 2015-2018 Bosch Sicherheitssysteme GmbH
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#ifndef _BOSCH_COMPANION_QUEUE_MANAGEMENT_H
+#define _BOSCH_COMPANION_QUEUE_MANAGEMENT_H
+
+#include <linux/kernel.h>
+#include <linux/kfifo.h>
+#include "protocol.h"
+
+#define QUEUE_SIZE 16u
+
+/**
+ * struct companion_queue - encapsulate kfifo as companion queue
+ * @fifo: the kfifo object for companion packets
+ */
+struct companion_queue {
+ DECLARE_KFIFO(fifo, struct companion_packet, QUEUE_SIZE);
+};
+
+/**
+ * struct companion_queue_manager - manage all queues for companion
+ * @io_txq: the tx queue for IO messages
+ * @io_rxq: the rx queue for IO messages
+ * @can_txq: the tx queues for CAN messages
+ * @can_rxq: the rx queues for CAN messages
+ * @can_err: the queues for CAN error messages
+ * @can_txq_lock: lock for protect CAN tx queue 0
+ * @io_promoted: flag to indicate promoted IO messages priority
+ * @can_current: the currently scheduled CAN port
+ * @can_sched: counter of how many times CAN messages be scheduled
+ */
+struct companion_queue_manager {
+ struct companion_queue io_txq;
+ struct companion_queue io_rxq;
+ struct companion_queue can_txq[BCP_CAN_PORTS][BCP_CAN_PRIOS];
+ struct companion_queue can_rxq[BCP_CAN_PORTS];
+ struct companion_queue can_err[BCP_CAN_PORTS];
+ struct mutex can_txq_lock[BCP_CAN_PORTS];
+
+ bool io_promoted;
+ u8 can_current;
+ u32 can_sched;
+};
+
+/**
+ * qm_init() - initialize all managed queues
+ * @qm: address of the queue manager to be initialized
+ */
+void qm_init(struct companion_queue_manager *qm);
+
+/**
+ * qm_reset() - reset all managed queues
+ * @qm: address of the queue manager to be reset
+ */
+void qm_reset(struct companion_queue_manager *qm);
+
+/**
+ * qm_reset_io() - reset managed IO queues
+ * @qm: address of the queue manager to be reset
+ */
+void qm_reset_io(struct companion_queue_manager *qm);
+
+/**
+ * qm_reset_can() - reset managed CAN queues
+ * @qm: address of the queue manager to be reset
+ * @port: port number of which CAN queue should be reset
+ */
+void qm_reset_can(struct companion_queue_manager *qm, u8 port);
+
+/**
+ * qm_has_tx_data() - return true if has tx data
+ * @qm: address of the queue manager to be used
+ */
+bool qm_has_tx_data(struct companion_queue_manager *qm);
+
+/**
+ * qm_get_tx_data() - return true if got the tx data
+ * @qm: address of the queue manager to be used
+ * @p: where the data to be copied
+ */
+bool qm_get_tx_data(struct companion_queue_manager *qm,
+ struct companion_packet *p);
+
+/**
+ * qm_io_txq_is_full() - return true if IO tx queue is full
+ * @qm: address of the queue manager to be used
+ */
+static inline bool qm_io_txq_is_full(struct companion_queue_manager *qm)
+{
+ return kfifo_is_full(&qm->io_txq.fifo);
+}
+
+/**
+ * qm_io_rxq_is_empty() - return true if IO rx queue is empty
+ * @qm: address of the queue manager to be used
+ */
+static inline bool qm_io_rxq_is_empty(struct companion_queue_manager *qm)
+{
+ return kfifo_is_empty(&qm->io_rxq.fifo);
+}
+
+/**
+ * qm_io_txq_in() - put data from user sapce into IO tx queue
+ * @qm: address of the queue manager to be used
+ * @buf: address of the data to be put
+ * @count: number of bytes to be put
+ * @copied: address to store the number of copied bytes
+ */
+static inline int qm_io_txq_in(struct companion_queue_manager *qm,
+ const char __user *buf,
+ size_t count,
+ unsigned int *copied)
+{
+ return kfifo_from_user(&qm->io_txq.fifo, buf, count, copied);
+}
+
+/**
+ * qm_io_txq_out() - get data from the IO tx queue
+ * @qm: address of the queue manager to be used
+ * @p: address of the data to be copied
+ */
+static inline bool qm_io_txq_out(struct companion_queue_manager *qm,
+ struct companion_packet *p)
+{
+ return kfifo_out(&qm->io_txq.fifo, p, 1) == 1;
+}
+
+/**
+ * qm_io_rxq_in() - put data into IO rx queue
+ * @qm: address of the queue manager to be used
+ * @p: address of the data to be put
+ */
+static inline bool qm_io_rxq_in(struct companion_queue_manager *qm,
+ const struct companion_packet *p)
+{
+ return kfifo_in(&qm->io_rxq.fifo, p, 1) == 1;
+}
+
+/**
+ * qm_io_rxq_out() - copy data from the IO rx queue into user space
+ * @qm: address of the queue manager to be used
+ * @buf: address of the data to be copied
+ * @count: number of the bytes to be copied
+ * @copied: address to store the number of copied bytes
+ */
+static inline int qm_io_rxq_out(struct companion_queue_manager *qm,
+ char __user *buf,
+ size_t count,
+ unsigned int *copied)
+{
+ return kfifo_to_user(&qm->io_rxq.fifo, buf, count, copied);
+}
+
+/**
+ * qm_can_txq_in() - put data into CAN tx queue
+ * @qm: address of the queue manager to be used
+ * @p: address of the data to be put
+ * @port: port number of which CAN queue array to be put
+ * @prio: priority of which CAN queue to be put
+ */
+static inline bool qm_can_txq_in(struct companion_queue_manager *qm,
+ const struct companion_packet *p,
+ u8 port,
+ u8 prio)
+{
+ bool result = false;
+
+ if (prio > 0)
+ return kfifo_in(&qm->can_txq[port][prio].fifo, p, 1) == 1;
+
+ /* queue 0 has multiple writers due to it sends both data and
+ * adminstrative frames, while queue 1-7 only send data frame
+ * (single writer), hence only queue 0 needs lock.
+ */
+ mutex_lock(&qm->can_txq_lock[port]);
+ result = (kfifo_in(&qm->can_txq[port][prio].fifo, p, 1) == 1);
+ mutex_unlock(&qm->can_txq_lock[port]);
+ return result;
+}
+
+/**
+ * qm_can_txq_out() - get data from the CAN tx queue
+ * @qm: address of the queue manager to be used
+ * @p: address of the data to be copied
+ * @port: port number of which CAN queue array to be copied
+ * @prio: priority of which CAN queue to be copied
+ */
+static inline bool qm_can_txq_out(struct companion_queue_manager *qm,
+ struct companion_packet *p,
+ u8 port,
+ u8 prio)
+{
+ return kfifo_out(&qm->can_txq[port][prio].fifo, p, 1) == 1;
+}
+
+/**
+ * qm_can_rxq_in() - put data into CAN rx queue
+ * @qm: address of the queue manager to be used
+ * @p: address of the data to be put
+ * @port: port number of which CAN queue to be put
+ */
+static inline bool qm_can_rxq_in(struct companion_queue_manager *qm,
+ const struct companion_packet *p,
+ u8 port)
+{
+ return kfifo_in(&qm->can_rxq[port].fifo, p, 1) == 1;
+}
+
+/**
+ * qm_can_rxq_out() - get data from the CAN rx queue
+ * @qm: address of the queue manager to be used
+ * @p: address of the data to be copied
+ * @port: port number of which CAN queue to be copied
+ */
+static inline bool qm_can_rxq_out(struct companion_queue_manager *qm,
+ struct companion_packet *p,
+ u8 port)
+{
+ return kfifo_out(&qm->can_rxq[port].fifo, p, 1) == 1;
+}
+
+static inline bool qm_can_err_in(struct companion_queue_manager *qm,
+ const struct companion_packet *p,
+ u8 port)
+{
+ return kfifo_in(&qm->can_err[port].fifo, p, 1) == 1;
+}
+
+static inline bool qm_can_err_out(struct companion_queue_manager *qm,
+ struct companion_packet *p,
+ u8 port)
+{
+ return kfifo_out(&qm->can_err[port].fifo, p, 1) == 1;
+}
+
+#endif
diff --git a/include/linux/companion.h b/include/linux/companion.h
new file mode 100644
index 0000000..8f9c876
--- /dev/null
+++ b/include/linux/companion.h
@@ -0,0 +1,258 @@
+/*
+ * Companion low level driver interface
+ *
+ * Copyright (C) 2017 Bosch Sicherheitssysteme GmbH
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#ifndef _BOSCH_COMPANION_H
+#define _BOSCH_COMPANION_H
+
+#include <linux/can/dev.h>
+#include <linux/device.h>
+
+/**
+ * struct companion_io_ops - callbacks of companion IO packets handling
+ * @on_tx_done: called when IO packets tx is done
+ * @on_rx_done: called when IO packets rx is done
+ */
+struct companion_io_ops
+{
+ void (* on_tx_done)(void *data);
+ void (* on_rx_done)(void *data);
+};
+
+/**
+ * struct companion_can_ops - callbacks of companion CAN packets handling
+ * @on_tx_done: called when CAN packets tx is done
+ * @on_rx_done: called when CAN packets rx is done
+ * @on_error: called when CAN error detected
+ * @on_tx_timeout: called when CAN packets tx timeout
+ */
+struct companion_can_ops
+{
+ void (* on_tx_done)(void *data, u8 prio, bool lost_seq, bool success);
+ void (* on_rx_done)(void *data);
+ void (* on_error)(void *data);
+ void (* on_tx_timeout)(void *data, u8 prio);
+};
+
+/**
+ * companion_io_ops_register() - register companion IO packets handler
+ * @parent: address of the caller parent device to be registered
+ * @ops: address of the IO packets callback
+ * @data: address of the IO packets callback argument
+ */
+int companion_io_ops_register(struct device *parent,
+ struct companion_io_ops *ops,
+ void *data);
+
+/**
+ * companion_io_ops_unregister() - unregister companion IO packets handler
+ * @parent: address of the caller parent device to be unregistered
+ */
+int companion_io_ops_unregister(struct device *parent);
+
+/**
+ * companion_can_ops_register() - register companion CAN packets handler
+ * @parent: address of the caller parent device to be registered
+ * @ops: address of the CAN packets callback
+ * @data: address of the CAN packets callback argument
+ */
+int companion_can_ops_register(struct device *parent,
+ u8 port,
+ struct companion_can_ops *ops,
+ void *data);
+
+/**
+ * companion_can_ops_unregister() - unregister comapnion CAN packets handler
+ * @parent: address of the caller parent device to be unregistered
+ * @port: port number of which CAN to be unregistered
+ */
+int companion_can_ops_unregister(struct device *parent, u8 port);
+
+/**
+ * companion_io_txq_is_full() - return true if IO tx queue is full
+ * @parent: address of the caller parent device to be used
+ */
+bool companion_io_txq_is_full(struct device *parent);
+
+/**
+ * companion_io_rxq_is_empty() - return true if IO rx queue is empty
+ * @parent: address of the caller parent device to be used
+ */
+bool companion_io_rxq_is_empty(struct device *parent);
+
+/**
+ * companion_do_io_tx() - send IO packets from user space to companion
+ * @parent: address of the caller parent device to be used
+ * @buf: address of the user space data to be sent
+ * @count: number of bytes to be sent
+ */
+int companion_do_io_tx(struct device *parent,
+ const char __user *buf,
+ size_t count);
+
+/**
+ * companion_do_io_rx() - receive IO packets from companion to user space
+ * @parent: address of the caller parent device to be used
+ * @buf: address of the data to be copied
+ * @count: number of bytes to be copied
+ */
+int companion_do_io_rx(struct device *parent,
+ char __user *buf,
+ size_t count);
+
+/**
+ * companion_do_can_tx() - send CAN data to companion
+ * @parent: address of the caller parent device to be used
+ * @port: port number of which CAN to be sent
+ * @prio: priority of the raw CAN frame to be sent
+ * @cf: address of the raw CAN frame to be sent
+ */
+int companion_do_can_tx(struct device *parent,
+ u8 port,
+ u8 prio,
+ const struct can_frame *cf);
+
+/**
+ * companion_do_can_rx() - receive CAN data from companion
+ * @parent: address of the caller parent device to be used
+ * @port: port number of which CAN to be received
+ * @cf: address of the raw CAN frame to be copied
+ */
+int companion_do_can_rx(struct device *parent,
+ u8 port,
+ struct can_frame *cf);
+
+/**
+ * companion_do_can_err() - receive CAN error from companion
+ * @parent: address of the caller parent device to be used
+ * @port: port number of which CAN to be received
+ * @bec: address of the error counter to be copied
+ * @state: address of the error state to be copied
+ * @code: address of the error code to be copied
+ */
+int companion_do_can_err(struct device *parent,
+ u8 port,
+ struct can_berr_counter *bec,
+ u8 *state,
+ u8 *code);
+#define COMPANION_CAN_STATE_WARNING 0x01u
+#define COMPANION_CAN_STATE_PASSIVE 0x02u
+#define COMPANION_CAN_STATE_BUS_OFF 0x04u
+#define COMPANION_CAN_ERROR_STUFF 0x01u
+#define COMPANION_CAN_ERROR_FORM 0x02u
+#define COMPANION_CAN_ERROR_ACK 0x04u
+#define COMPANION_CAN_ERROR_BIT1 0x08u
+#define COMPANION_CAN_ERROR_BIT0 0x10u
+#define COMPANION_CAN_ERROR_CRC 0x20u
+#define COMPANION_CAN_ERROR_RXOV 0x80u
+
+/**
+ * companion_do_set_can_bittiming() - set companion CAN bittiming
+ * @parent: address of the caller parent device to be used
+ * @port: port number of which CAN to be set
+ * @bittiming: the bittiming to be set
+ */
+int companion_do_set_can_bittiming(struct device *parent,
+ u8 port,
+ const struct can_bittiming *bittiming);
+
+/**
+ * companion_do_set_can_mode() - set companion CAN mode
+ * @parent: address of the caller parent device to be used
+ * @port: port number of which CAN to be set
+ * @mode: the mode to be set
+ */
+int companion_do_set_can_mode(struct device *parent,
+ u8 port,
+ enum can_mode mode);
+
+/**
+ * companion_do_set_can_ctrlmode() - set companion CAN control mode
+ * @parent: address of the caller parent device to be used
+ * @port: port number of which CAN to be set
+ * @ctrl: the control mode to be set
+ */
+int companion_do_set_can_ctrlmode(struct device *parent,
+ u8 port,
+ u32 ctrl);
+
+/**
+ * companion_do_get_can_status() - get companion CAN status
+ * @parent: address of the caller parent device to be used
+ * @port: port number of which CAN to be inquiry
+ * @bec: address of the error counter to be copied
+ */
+int companion_do_get_can_status(struct device *parent,
+ u8 port,
+ struct can_berr_counter *bec);
+
+/**
+ * companion_do_get_can_txq_status() - get companion CAN tx queue status
+ * @parent: address of the caller parent device to be used
+ * @port: port number of which CAN to be inquiry
+ * @prio: queue number of which tx queue to be inquiry
+ * @lost_txq_sync: flag of the given CAN tx queue lost sync or not
+ */
+int companion_do_get_can_txq_status(struct device *parent,
+ u8 port,
+ u8 prio,
+ bool *lost_txq_sync);
+
+/**
+ * companion_do_get_can_txq_status_all() - get all companion CAN tx queue status
+ * @parent: address of the caller parent device to be used
+ * @port: port number of which CAN to be inquiry
+ */
+int companion_do_get_can_txq_status_all(struct device *parent,
+ u8 port);
+
+/**
+ * companion_do_can_txq_is_full() - inquiry companion CAN tx queue is full
+ * @parent: address of the caller parent device to be used
+ * @port: port number of which CAN to be inquiry
+ * @prio: queue number of which tx queue to be inquiry
+ * @is_full: address of the is full result to be copied
+ */
+int companion_do_can_txq_is_full(struct device *parent,
+ u8 port,
+ u8 prio,
+ bool *is_full);
+
+/**
+ * companion_do_can_txq_has_space() - inquiry companion CAN tx queue has space
+ * @parent: address of the caller parent device to be used
+ * @port: port number of which CAN to be inquiry
+ * @prio: queue number of which tx queue to be inquiry
+ * @has_space: address of the has_space result to be copied
+ */
+int companion_do_can_txq_has_space(struct device *parent,
+ u8 port,
+ u8 prio,
+ bool *has_space);
+
+/**
+ * companion_do_can_start_tx_timer() - start companioin CAN tx timeout detection
+ * @parent: address of the caller parent device to be used
+ * @port: port number of which CAN to be used
+ * @prio: queue number of which tx queue's timer to be used
+ */
+int companion_do_can_start_tx_timer(struct device *parent,
+ u8 port,
+ u8 prio);
+
+/**
+ * companion_do_can_stop_tx_timer() - stop companion CAN tx timeout detection
+ * @parent: address of the caller parent device to be used
+ * @port: port number of which CAN to be used
+ * @prio: queue number of which tx queue's timer to be used
+ */
+int companion_do_can_stop_tx_timer(struct device *parent,
+ u8 port,
+ u8 prio);
+#endif
--
2.7.4
^ permalink raw reply related
* [PATCH 1/5] can: enable multi-queue for SocketCAN devices
From: Mark Jonas @ 2018-06-05 18:43 UTC (permalink / raw)
To: Wolfgang Grandegger, Marc Kleine-Budde
Cc: linux-can, netdev, linux-kernel, hs, yi.zhu5, Mark Jonas
In-Reply-To: <1528224240-30786-1-git-send-email-mark.jonas@de.bosch.com>
From: Zhu Yi <yi.zhu5@cn.bosch.com>
The existing SocketCAN implementation provides alloc_candev() to
allocate a CAN device using a single Tx and Rx queue. This can lead to
priority inversion in case the single Tx queue is already full with low
priority messages and a high priority message needs to be sent while the
bus is fully loaded with medium priority messages.
This problem can be solved by using the existing multi-queue support of
the network subsytem. The commit makes it possible to use multi-queue in
the CAN subsystem in the same way it is used in the Ethernet subsystem
by adding an alloc_candev_mqs() call and accompanying macros. With this
support a CAN device can use multi-queue qdisc (e.g. mqprio) to avoid
the aforementioned priority inversion.
The exisiting functionality of alloc_candev() is the same as before.
CAN devices need to have prioritized multiple hardware queues or are
able to abort waiting for arbitration to make sensible use of
multi-queues.
Signed-off-by: Zhu Yi <yi.zhu5@cn.bosch.com>
Signed-off-by: Mark Jonas <mark.jonas@de.bosch.com>
Reviewed-by: Heiko Schocher <hs@denx.de>
---
drivers/net/can/dev.c | 8 +++++---
include/linux/can/dev.h | 7 ++++++-
2 files changed, 11 insertions(+), 4 deletions(-)
diff --git a/drivers/net/can/dev.c b/drivers/net/can/dev.c
index 365a8cc..ac8270c 100644
--- a/drivers/net/can/dev.c
+++ b/drivers/net/can/dev.c
@@ -702,7 +702,8 @@ EXPORT_SYMBOL_GPL(alloc_can_err_skb);
/*
* Allocate and setup space for the CAN network device
*/
-struct net_device *alloc_candev(int sizeof_priv, unsigned int echo_skb_max)
+struct net_device *alloc_candev_mqs(int sizeof_priv, unsigned int echo_skb_max,
+ unsigned int txqs, unsigned int rxqs)
{
struct net_device *dev;
struct can_priv *priv;
@@ -714,7 +715,8 @@ struct net_device *alloc_candev(int sizeof_priv, unsigned int echo_skb_max)
else
size = sizeof_priv;
- dev = alloc_netdev(size, "can%d", NET_NAME_UNKNOWN, can_setup);
+ dev = alloc_netdev_mqs(size, "can%d", NET_NAME_UNKNOWN, can_setup,
+ txqs, rxqs);
if (!dev)
return NULL;
@@ -733,7 +735,7 @@ struct net_device *alloc_candev(int sizeof_priv, unsigned int echo_skb_max)
return dev;
}
-EXPORT_SYMBOL_GPL(alloc_candev);
+EXPORT_SYMBOL_GPL(alloc_candev_mqs);
/*
* Free space of the CAN network device
diff --git a/include/linux/can/dev.h b/include/linux/can/dev.h
index 61f1cf2..07b73d2 100644
--- a/include/linux/can/dev.h
+++ b/include/linux/can/dev.h
@@ -142,7 +142,12 @@ u8 can_dlc2len(u8 can_dlc);
/* map the sanitized data length to an appropriate data length code */
u8 can_len2dlc(u8 len);
-struct net_device *alloc_candev(int sizeof_priv, unsigned int echo_skb_max);
+struct net_device *alloc_candev_mqs(int sizeof_priv, unsigned int echo_skb_max,
+ unsigned int txqs, unsigned int rxqs);
+#define alloc_candev(sizeof_priv, echo_skb_max) \
+ alloc_candev_mqs(sizeof_priv, echo_skb_max, 1, 1)
+#define alloc_candev_mq(sizeof_priv, echo_skb_max, count) \
+ alloc_candev_mqs(sizeof_priv, echo_skb_max, count, count)
void free_candev(struct net_device *dev);
/* a candev safe wrapper around netdev_priv */
--
2.7.4
^ permalink raw reply related
* [PATCH 0/5] can: enable multi-queue for SocketCAN devices
From: Mark Jonas @ 2018-06-05 18:43 UTC (permalink / raw)
To: Wolfgang Grandegger, Marc Kleine-Budde
Cc: linux-can, netdev, linux-kernel, hs, yi.zhu5, Mark Jonas
Upon request by Marc Kleine-Budde this patch series does not only
contain our patch to enable enable multi-queue for SocketCAN devices
but also a driver (Companion driver suite) which makes active use of
this feature.
The driver suite implements
- two CAN interfaces
- one generic command interfaces
and offers a SocketCAN as well as a char device interface. The
SocketCAN interface supports multi-queue.
The functionality bases on an external peripheral chip named Companion.
It offers two CAN interfaces, each has 8 prioritized transmit FIFOs as
well as one receive FIFO. Besides CAN, undisclosed additional functions
can be accessed through the char device.
A standard SPI interface with two additional lines for flow control is
used. The Companion chip is the SPI slave.
The driver suite consists of three separate drivers. The following
diagram illustrates the dependencies in layers.
/dev/companion SocketCAN User Space
-------------------------------------------------------------------
+----------------+ +---------------+
| companion-char | | companion-can |
+----------------+ +---------------+
+----------------------------------+
| companion-spi |
+----------------------------------+
+----------------------------------+
| standard SPI subsystem |
+----------------------------------+ Linux Kernel
-------------------------------------------------------------------
| | | | | | Hardware
CS-+ | | | | +-BUSY
CLK--+ | | +---REQUEST
MOSI---+ |
MISO-----+
companion-spi
core.c: handles SPI, sysfs entry and interface to upper layer
protocol-manager.c: handles protocol with the SPI HW
queue-manager.c: handles buffering and packets scheduling
companion-can
makes use of multi-queue support and allows to use tc to configure
the queuing discipline (e.g. mqprio). Together with the SO_PRIORITY
socket option this allows to specify the FIFO a CAN frame shall be
sent to.
companion-char
handles messages to other undisclosed functionality beyond CAN.
Zhu Yi (5):
can: enable multi-queue for SocketCAN devices
spi: implement companion-spi driver
char: implement companion-char driver
can: implement companion-can driver
spi,can,char: add companion DT binding documentation
.../devicetree/bindings/spi/bosch,companion.txt | 82 ++
drivers/char/Kconfig | 7 +
drivers/char/Makefile | 2 +
drivers/char/companion-char.c | 367 ++++++
drivers/net/can/Kconfig | 8 +
drivers/net/can/Makefile | 1 +
drivers/net/can/companion-can.c | 694 ++++++++++++
drivers/net/can/dev.c | 8 +-
drivers/spi/Kconfig | 2 +
drivers/spi/Makefile | 2 +
drivers/spi/companion/Kconfig | 5 +
drivers/spi/companion/Makefile | 2 +
drivers/spi/companion/core.c | 1189 ++++++++++++++++++++
drivers/spi/companion/protocol-manager.c | 1035 +++++++++++++++++
drivers/spi/companion/protocol-manager.h | 348 ++++++
drivers/spi/companion/protocol.h | 273 +++++
drivers/spi/companion/queue-manager.c | 146 +++
drivers/spi/companion/queue-manager.h | 245 ++++
include/linux/can/dev.h | 7 +-
include/linux/companion.h | 258 +++++
20 files changed, 4677 insertions(+), 4 deletions(-)
create mode 100644 Documentation/devicetree/bindings/spi/bosch,companion.txt
create mode 100644 drivers/char/companion-char.c
create mode 100644 drivers/net/can/companion-can.c
create mode 100644 drivers/spi/companion/Kconfig
create mode 100644 drivers/spi/companion/Makefile
create mode 100644 drivers/spi/companion/core.c
create mode 100644 drivers/spi/companion/protocol-manager.c
create mode 100644 drivers/spi/companion/protocol-manager.h
create mode 100644 drivers/spi/companion/protocol.h
create mode 100644 drivers/spi/companion/queue-manager.c
create mode 100644 drivers/spi/companion/queue-manager.h
create mode 100644 include/linux/companion.h
--
2.7.4
^ permalink raw reply
* Re: [PATCH net-next 2/3] net: hns3: Fix for VF mailbox receiving unknown message
From: David Miller @ 2018-06-05 18:40 UTC (permalink / raw)
To: salil.mehta
Cc: yisen.zhuang, lipeng321, mehta.salil, netdev, linux-kernel,
linuxarm, wangxi11
In-Reply-To: <20180605114201.29900-3-salil.mehta@huawei.com>
From: Salil Mehta <salil.mehta@huawei.com>
Date: Tue, 5 Jun 2018 12:42:00 +0100
> + if (unlikely(!hnae3_get_bit(flag, HCLGEVF_CMDQ_RX_OUTVLD_B))) {
This breaks the build, there is no such symbol named hnae3_get_bit().
^ permalink raw reply
* Re: [PATCH net] failover: eliminate callback hell
From: Michael S. Tsirkin @ 2018-06-05 18:35 UTC (permalink / raw)
To: Stephen Hemminger
Cc: kys, haiyangz, davem, sridhar.samudrala, netdev,
Stephen Hemminger
In-Reply-To: <20180605034231.31610-1-sthemmin@microsoft.com>
Thanks, I think this is nice patch but I wonder whether it can be split
up somewhat. Not all of it is uncontroversial.
On Mon, Jun 04, 2018 at 08:42:31PM -0700, Stephen Hemminger wrote:
> * The matching of secondary device to primary device policy
> is up to the network device. Both net_failover and netvsc
> will use MAC for now but can change separately.
I actually suspect both will change to a serial number
down the road.
> * The match policy is only used during initial discovery; after
> that the secondary device knows what the upper device is because
> of the parent/child relationship; no searching is required.
That would obviously be an improvement - does it have to be tied with
rest of changes?
> * Now, netvsc and net_failover use the same delayed work type
> mechanism for setup. Previously, net_failover code was triggering off
> name change but a similar policy was rejected for netvsc.
> "what is good for the goose is good for the gander"
I don't really understand what you are saying here. I think the delayed
hack is kind of ugly and seems racy. Current failover code was rejected
by whom? Why is new one good and for whom? Did you want to do a name
change in netvsc but it was rejected? Could you clarify please?
> * The net_failover private device info 'struct net_failover_info'
> should have been private to the driver file, not a visible
> API.
>
> * The net_failover device should use SET_NETDEV_DEV
> that is intended only for physical devices not virtual devices.
You mean should not.
> * No point in having DocBook style comments on a driver file.
> They only make sense on an external exposed API.
>
> * net_failover only supports Ethernet, so use ether_addr_copy.
It is since you need to know about all the things you need to copy, and
because of mac matching. But it isn't too much effort to add more
transports and I don't see value in going in the reverse direction and
making it more ethernet specific that it already is.
> * Set permanent and current address of net_failover device
> to match the primary.
>
> * Carrier should be marked off before registering device
> the net_failover device.
Are above two bugfixes?
> * Use netdev_XXX for log messages, in net_failover (not dev_xxx)
>
> * Since failover infrastructure is about linking devices just
> use RTNL no need for other locking in init and teardown.
>
> * Don't bother with ERR_PTR() style return if only possible
> return is success or no memory.
>
> * As much as possible, the terms master and slave should be avoided
> because of their cultural connotations.
Also for consistency, failover is calling these primary and standby now.
> Note; this code has been tested on Hyper-V
> but is compile tested only on virtio.
>
> Fixes: 30c8bd5aa8b2 ("net: Introduce generic failover module")
> Signed-off-by: Stephen Hemminger <sthemmin@microsoft.com>
> ---
>
> Although this patch needs to go into 4.18 (linux-net),
I'd rather we focused on fixing bugs in 4.18, and left refactoring to
4.19.
At some point you said refactoring is needed to support matching using
the serial number, but I see this didn't make 4.18. So no rush IMHO.
> this version is based against net-next because net-next
> hasn't been merged into linux-net yet.
>
>
> drivers/net/hyperv/hyperv_net.h | 3 +-
> drivers/net/hyperv/netvsc_drv.c | 173 +++++++++++------
> drivers/net/net_failover.c | 312 ++++++++++++++++++++-----------
> drivers/net/virtio_net.c | 9 +-
> include/net/failover.h | 31 +---
> include/net/net_failover.h | 32 +---
> net/Kconfig | 13 +-
> net/core/failover.c | 316 ++++----------------------------
> 8 files changed, 373 insertions(+), 516 deletions(-)
>
> diff --git a/drivers/net/hyperv/hyperv_net.h b/drivers/net/hyperv/hyperv_net.h
> index 99d8e7398a5b..c7d25d10765e 100644
> --- a/drivers/net/hyperv/hyperv_net.h
> +++ b/drivers/net/hyperv/hyperv_net.h
> @@ -902,6 +902,8 @@ struct net_device_context {
> struct hv_device *device_ctx;
> /* netvsc_device */
> struct netvsc_device __rcu *nvdev;
> + /* list of netvsc net_devices */
> + struct list_head list;
> /* reconfigure work */
> struct delayed_work dwork;
> /* last reconfig time */
> @@ -933,7 +935,6 @@ struct net_device_context {
> /* Serial number of the VF to team with */
> u32 vf_serial;
>
> - struct failover *failover;
> };
>
> /* Per channel data */
> diff --git a/drivers/net/hyperv/netvsc_drv.c b/drivers/net/hyperv/netvsc_drv.c
> index bef4d55a108c..074e6b8578df 100644
> --- a/drivers/net/hyperv/netvsc_drv.c
> +++ b/drivers/net/hyperv/netvsc_drv.c
> @@ -70,6 +70,8 @@ static int debug = -1;
> module_param(debug, int, 0444);
> MODULE_PARM_DESC(debug, "Debug level (0=none,...,16=all)");
>
> +static LIST_HEAD(netvsc_dev_list);
> +
> static void netvsc_change_rx_flags(struct net_device *net, int change)
> {
> struct net_device_context *ndev_ctx = netdev_priv(net);
> @@ -1846,101 +1848,120 @@ static void netvsc_vf_setup(struct work_struct *w)
> }
>
> vf_netdev = rtnl_dereference(ndev_ctx->vf_netdev);
> - if (vf_netdev)
> + if (vf_netdev) {
> __netvsc_vf_setup(ndev, vf_netdev);
> -
> + dev_put(vf_netdev);
> + }
> rtnl_unlock();
> }
>
> -static int netvsc_pre_register_vf(struct net_device *vf_netdev,
> - struct net_device *ndev)
> +static struct net_device *get_netvsc_bymac(const u8 *mac)
> {
> - struct net_device_context *net_device_ctx;
> - struct netvsc_device *netvsc_dev;
> + struct net_device_context *ndev_ctx;
>
> - net_device_ctx = netdev_priv(ndev);
> - netvsc_dev = rtnl_dereference(net_device_ctx->nvdev);
> - if (!netvsc_dev || rtnl_dereference(net_device_ctx->vf_netdev))
> - return -ENODEV;
> + ASSERT_RTNL();
>
> - return 0;
> + list_for_each_entry(ndev_ctx, &netvsc_dev_list, list) {
> + struct net_device *dev = hv_get_drvdata(ndev_ctx->device_ctx);
> +
> + if (ether_addr_equal(mac, dev->perm_addr))
> + return dev;
> + }
> +
> + return NULL;
> }
>
> -static int netvsc_register_vf(struct net_device *vf_netdev,
> - struct net_device *ndev)
> +static int netvsc_register_vf(struct net_device *vf_netdev)
> {
> - struct net_device_context *ndev_ctx = netdev_priv(ndev);
> + struct net_device *ndev;
> + struct net_device_context *ndev_ctx;
> +
> + /* Must use Ethernet addresses */
> + if (vf_netdev->addr_len != ETH_ALEN)
> + return NOTIFY_DONE;
> +
> + /* VF must be a physical device not VLAN, etc */
> + if (!vf_netdev->dev.parent)
> + return NOTIFY_DONE;
> +
> + /* Use the MAC address to locate the synthetic interface to
> + * associate with the VF interface.
> + */
> + ndev = get_netvsc_bymac(vf_netdev->perm_addr);
> + if (!ndev)
> + return NOTIFY_DONE;
> +
> + /* If network device is being removed, don't do anything */
> + ndev_ctx = netdev_priv(ndev);
> + if (!rtnl_dereference(ndev_ctx->nvdev))
> + return NOTIFY_DONE;
> +
> + if (netdev_failover_join(vf_netdev, ndev, netvsc_vf_handle_frame)) {
> + netdev_err(vf_netdev, "could not join: %s", ndev->name);
> + return NOTIFY_DONE;
> + }
>
> /* set slave flag before open to prevent IPv6 addrconf */
> vf_netdev->flags |= IFF_SLAVE;
>
> + dev_hold(vf_netdev);
> +
> schedule_delayed_work(&ndev_ctx->vf_takeover, VF_TAKEOVER_INT);
>
> call_netdevice_notifiers(NETDEV_JOIN, vf_netdev);
>
> netdev_info(vf_netdev, "joined to %s\n", ndev->name);
>
> - dev_hold(vf_netdev);
> rcu_assign_pointer(ndev_ctx->vf_netdev, vf_netdev);
>
> - return 0;
> + return NOTIFY_OK;
> }
>
> /* VF up/down change detected, schedule to change data path */
> -static int netvsc_vf_changed(struct net_device *vf_netdev,
> - struct net_device *ndev)
> +static int netvsc_vf_changed(struct net_device *vf_netdev)
> {
> struct net_device_context *net_device_ctx;
> struct netvsc_device *netvsc_dev;
> + struct net_device *ndev;
> bool vf_is_up = netif_running(vf_netdev);
>
> + ndev = netdev_failover_upper_get(vf_netdev);
> + if (!ndev)
> + return NOTIFY_DONE;
> +
> net_device_ctx = netdev_priv(ndev);
> netvsc_dev = rtnl_dereference(net_device_ctx->nvdev);
> if (!netvsc_dev)
> - return -ENODEV;
> + return NOTIFY_DONE;
>
> netvsc_switch_datapath(ndev, vf_is_up);
> netdev_info(ndev, "Data path switched %s VF: %s\n",
> vf_is_up ? "to" : "from", vf_netdev->name);
>
> - return 0;
> + return NOTIFY_OK;
> }
>
> -static int netvsc_pre_unregister_vf(struct net_device *vf_netdev,
> - struct net_device *ndev)
> +static int netvsc_unregister_vf(struct net_device *vf_netdev)
> {
> struct net_device_context *net_device_ctx;
> + struct net_device *ndev;
>
> - net_device_ctx = netdev_priv(ndev);
> - cancel_delayed_work_sync(&net_device_ctx->vf_takeover);
> -
> - return 0;
> -}
> -
> -static int netvsc_unregister_vf(struct net_device *vf_netdev,
> - struct net_device *ndev)
> -{
> - struct net_device_context *net_device_ctx;
> + ndev = netdev_failover_upper_get(vf_netdev);
> + if (!ndev)
> + return NOTIFY_DONE;
>
> net_device_ctx = netdev_priv(ndev);
> + if (cancel_delayed_work_sync(&net_device_ctx->vf_takeover))
> + dev_put(vf_netdev);
>
> netdev_info(ndev, "VF unregistering: %s\n", vf_netdev->name);
>
> + netdev_failover_unjoin(vf_netdev, ndev);
> RCU_INIT_POINTER(net_device_ctx->vf_netdev, NULL);
> - dev_put(vf_netdev);
>
> - return 0;
> + return NOTIFY_OK;
> }
>
> -static struct failover_ops netvsc_failover_ops = {
> - .slave_pre_register = netvsc_pre_register_vf,
> - .slave_register = netvsc_register_vf,
> - .slave_pre_unregister = netvsc_pre_unregister_vf,
> - .slave_unregister = netvsc_unregister_vf,
> - .slave_link_change = netvsc_vf_changed,
> - .slave_handle_frame = netvsc_vf_handle_frame,
> -};
> -
> static int netvsc_probe(struct hv_device *dev,
> const struct hv_vmbus_device_id *dev_id)
> {
> @@ -2009,6 +2030,8 @@ static int netvsc_probe(struct hv_device *dev,
>
> memcpy(net->dev_addr, device_info.mac_adr, ETH_ALEN);
>
> + net->priv_flags |= IFF_FAILOVER;
> +
> /* hw_features computed in rndis_netdev_set_hwcaps() */
> net->features = net->hw_features |
> NETIF_F_HIGHDMA | NETIF_F_SG |
> @@ -2024,23 +2047,19 @@ static int netvsc_probe(struct hv_device *dev,
> else
> net->max_mtu = ETH_DATA_LEN;
>
> - ret = register_netdev(net);
> + rtnl_lock();
> + ret = register_netdevice(net);
> if (ret != 0) {
> pr_err("Unable to register netdev.\n");
> goto register_failed;
> }
>
> - net_device_ctx->failover = failover_register(net, &netvsc_failover_ops);
> - if (IS_ERR(net_device_ctx->failover)) {
> - ret = PTR_ERR(net_device_ctx->failover);
> - goto err_failover;
> - }
> -
> - return ret;
> + list_add(&net_device_ctx->list, &netvsc_dev_list);
> + rtnl_unlock();
> + return 0;
>
> -err_failover:
> - unregister_netdev(net);
> register_failed:
> + rtnl_unlock();
> rndis_filter_device_remove(dev, nvdev);
> rndis_failed:
> free_percpu(net_device_ctx->vf_stats);
> @@ -2079,15 +2098,17 @@ static int netvsc_remove(struct hv_device *dev)
> */
> rtnl_lock();
> vf_netdev = rtnl_dereference(ndev_ctx->vf_netdev);
> - if (vf_netdev)
> - failover_slave_unregister(vf_netdev);
> + if (vf_netdev) {
> + netdev_failover_unjoin(vf_netdev, net);
> + dev_put(vf_netdev);
> + }
>
> if (nvdev)
> rndis_filter_device_remove(dev, nvdev);
>
> unregister_netdevice(net);
>
> - failover_unregister(ndev_ctx->failover);
> + list_del(&ndev_ctx->list);
>
> rtnl_unlock();
> rcu_read_unlock();
> @@ -2115,8 +2136,47 @@ static struct hv_driver netvsc_drv = {
> .remove = netvsc_remove,
> };
>
> +/* On Hyper-V, every VF interface is matched with a corresponding
> + * synthetic interface. The synthetic interface is presented first
> + * to the guest. When the corresponding VF instance is registered,
> + * we will take care of switching the data path.
> + */
> +static int netvsc_netdev_event(struct notifier_block *this,
> + unsigned long event, void *ptr)
> +{
> + struct net_device *event_dev = netdev_notifier_info_to_dev(ptr);
> +
> + /* Skip parent events */
> + if (netif_is_failover(event_dev))
> + return NOTIFY_DONE;
> +
> + /* Avoid non-Ethernet type devices */
> + if (event_dev->type != ARPHRD_ETHER)
> + return NOTIFY_DONE;
> +
> + switch (event) {
> + case NETDEV_REGISTER:
> + return netvsc_register_vf(event_dev);
> +
> + case NETDEV_UNREGISTER:
> + return netvsc_unregister_vf(event_dev);
> +
> + case NETDEV_UP:
> + case NETDEV_DOWN:
> + return netvsc_vf_changed(event_dev);
> +
> + default:
> + return NOTIFY_DONE;
> + }
> +}
> +
> +static struct notifier_block netvsc_netdev_notifier = {
> + .notifier_call = netvsc_netdev_event,
> +};
> +
> static void __exit netvsc_drv_exit(void)
> {
> + unregister_netdevice_notifier(&netvsc_netdev_notifier);
> vmbus_driver_unregister(&netvsc_drv);
> }
>
> @@ -2136,6 +2196,7 @@ static int __init netvsc_drv_init(void)
> if (ret)
> return ret;
>
> + register_netdevice_notifier(&netvsc_netdev_notifier);
> return 0;
> }
>
> diff --git a/drivers/net/net_failover.c b/drivers/net/net_failover.c
> index 83f7420ddea5..e0d30527f748 100644
> --- a/drivers/net/net_failover.c
> +++ b/drivers/net/net_failover.c
> @@ -28,6 +28,46 @@
> #include <uapi/linux/if_arp.h>
> #include <net/net_failover.h>
>
> +static LIST_HEAD(net_failover_list);
> +
> +/* failover state */
> +struct net_failover_info {
> + struct net_device *failover_dev;
> +
> + /* list of failover virtual devices */
> + struct list_head list;
> +
> + /* primary netdev with same MAC */
> + struct net_device __rcu *primary_dev;
> +
> + /* standby netdev */
> + struct net_device __rcu *standby_dev;
> +
> + /* primary netdev stats */
> + struct rtnl_link_stats64 primary_stats;
> +
> + /* standby netdev stats */
> + struct rtnl_link_stats64 standby_stats;
> +
> + /* aggregated stats */
> + struct rtnl_link_stats64 failover_stats;
> +
> + /* spinlock while updating stats */
> + spinlock_t stats_lock;
> +
> + /* delayed setup of slave */
> + struct delayed_work standby_init;
> +};
> +
> +#define FAILOVER_VLAN_FEATURES (NETIF_F_HW_CSUM | NETIF_F_SG | \
> + NETIF_F_FRAGLIST | NETIF_F_ALL_TSO | \
> + NETIF_F_HIGHDMA | NETIF_F_LRO)
> +
> +#define FAILOVER_ENC_FEATURES (NETIF_F_HW_CSUM | NETIF_F_SG | \
> + NETIF_F_RXCSUM | NETIF_F_ALL_TSO)
> +
> +#define FAILOVER_SETUP_INTERVAL (HZ / 10)
> +
> static bool net_failover_xmit_ready(struct net_device *dev)
> {
> return netif_running(dev) && netif_carrier_ok(dev);
> @@ -460,22 +500,42 @@ static void net_failover_lower_state_changed(struct net_device *slave_dev,
> netdev_lower_state_changed(slave_dev, &info);
> }
>
> -static int net_failover_slave_pre_register(struct net_device *slave_dev,
> - struct net_device *failover_dev)
> +static struct net_device *get_net_failover_bymac(const u8 *mac)
> {
> - struct net_device *standby_dev, *primary_dev;
> + struct net_failover_info *nfo_info;
> +
> + ASSERT_RTNL();
> +
> + list_for_each_entry(nfo_info, &net_failover_list, list) {
> + struct net_device *failover_dev = nfo_info->failover_dev;
> +
> + if (ether_addr_equal(mac, failover_dev->perm_addr))
> + return failover_dev;
> + }
> +
> + return NULL;
> +}
> +
> +static int net_failover_register_event(struct net_device *slave_dev)
> +{
> + struct net_device *failover_dev, *standby_dev, *primary_dev;
> struct net_failover_info *nfo_info;
> bool slave_is_standby;
>
> + failover_dev = get_net_failover_bymac(slave_dev->perm_addr);
> + if (!failover_dev)
> + return NOTIFY_DONE;
> +
> nfo_info = netdev_priv(failover_dev);
> standby_dev = rtnl_dereference(nfo_info->standby_dev);
> primary_dev = rtnl_dereference(nfo_info->primary_dev);
> slave_is_standby = slave_dev->dev.parent == failover_dev->dev.parent;
> if (slave_is_standby ? standby_dev : primary_dev) {
> - netdev_err(failover_dev, "%s attempting to register as slave dev when %s already present\n",
> + netdev_err(failover_dev,
> + "%s attempting to register as slave dev when %s already present\n",
> slave_dev->name,
> slave_is_standby ? "standby" : "primary");
> - return -EINVAL;
> + return NOTIFY_DONE;
> }
>
> /* We want to allow only a direct attached VF device as a primary
> @@ -484,23 +544,33 @@ static int net_failover_slave_pre_register(struct net_device *slave_dev,
> */
> if (!slave_is_standby && (!slave_dev->dev.parent ||
> !dev_is_pci(slave_dev->dev.parent)))
> - return -EINVAL;
> + return NOTIFY_DONE;
>
> if (failover_dev->features & NETIF_F_VLAN_CHALLENGED &&
> vlan_uses_dev(failover_dev)) {
> - netdev_err(failover_dev, "Device %s is VLAN challenged and failover device has VLAN set up\n",
> + netdev_err(failover_dev,
> + "Device %s is VLAN challenged and failover device has VLAN set up\n",
> failover_dev->name);
> - return -EINVAL;
> + return NOTIFY_DONE;
> }
>
> - return 0;
> + if (netdev_failover_join(slave_dev, failover_dev,
> + net_failover_handle_frame)) {
> + netdev_err(failover_dev, "could not join: %s", slave_dev->name);
> + return NOTIFY_DONE;
> + }
> +
> + /* Trigger rest of setup in process context */
> + schedule_delayed_work(&nfo_info->standby_init, FAILOVER_SETUP_INTERVAL);
> +
> + return NOTIFY_OK;
> }
>
> -static int net_failover_slave_register(struct net_device *slave_dev,
> - struct net_device *failover_dev)
> +static void __net_failover_setup(struct net_device *failover_dev)
> {
> + struct net_failover_info *nfo_info = netdev_priv(failover_dev);
> + struct net_device *slave_dev = rtnl_dereference(nfo_info->standby_dev);
> struct net_device *standby_dev, *primary_dev;
> - struct net_failover_info *nfo_info;
> bool slave_is_standby;
> u32 orig_mtu;
> int err;
> @@ -509,13 +579,12 @@ static int net_failover_slave_register(struct net_device *slave_dev,
> orig_mtu = slave_dev->mtu;
> err = dev_set_mtu(slave_dev, failover_dev->mtu);
> if (err) {
> - netdev_err(failover_dev, "unable to change mtu of %s to %u register failed\n",
> + netdev_err(failover_dev,
> + "unable to change mtu of %s to %u register failed\n",
> slave_dev->name, failover_dev->mtu);
> goto done;
> }
>
> - dev_hold(slave_dev);
> -
> if (netif_running(failover_dev)) {
> err = dev_open(slave_dev);
> if (err && (err != -EBUSY)) {
> @@ -537,7 +606,6 @@ static int net_failover_slave_register(struct net_device *slave_dev,
> goto err_vlan_add;
> }
>
> - nfo_info = netdev_priv(failover_dev);
> standby_dev = rtnl_dereference(nfo_info->standby_dev);
> primary_dev = rtnl_dereference(nfo_info->primary_dev);
> slave_is_standby = slave_dev->dev.parent == failover_dev->dev.parent;
> @@ -562,52 +630,56 @@ static int net_failover_slave_register(struct net_device *slave_dev,
> netdev_info(failover_dev, "failover %s slave:%s registered\n",
> slave_is_standby ? "standby" : "primary", slave_dev->name);
>
> - return 0;
> + return;
>
> err_vlan_add:
> dev_uc_unsync(slave_dev, failover_dev);
> dev_mc_unsync(slave_dev, failover_dev);
> dev_close(slave_dev);
> err_dev_open:
> - dev_put(slave_dev);
> dev_set_mtu(slave_dev, orig_mtu);
> done:
> - return err;
> + return;
> }
>
> -static int net_failover_slave_pre_unregister(struct net_device *slave_dev,
> - struct net_device *failover_dev)
> +static void net_failover_setup(struct work_struct *w)
> {
> - struct net_device *standby_dev, *primary_dev;
> - struct net_failover_info *nfo_info;
> + struct net_failover_info *nfo_info
> + = container_of(w, struct net_failover_info, standby_init.work);
> + struct net_device *failover_dev = nfo_info->failover_dev;
>
> - nfo_info = netdev_priv(failover_dev);
> - primary_dev = rtnl_dereference(nfo_info->primary_dev);
> - standby_dev = rtnl_dereference(nfo_info->standby_dev);
> -
> - if (slave_dev != primary_dev && slave_dev != standby_dev)
> - return -ENODEV;
> + /* handle race with cancel delayed work on removal */
> + if (!rtnl_trylock()) {
> + schedule_delayed_work(&nfo_info->standby_init, 0);
> + return;
> + }
>
> - return 0;
> + __net_failover_setup(failover_dev);
> + rtnl_unlock();
> }
>
> -static int net_failover_slave_unregister(struct net_device *slave_dev,
> - struct net_device *failover_dev)
> +static int net_failover_unregister_event(struct net_device *slave_dev)
> {
> - struct net_device *standby_dev, *primary_dev;
> + struct net_device *failover_dev, *primary_dev, *standby_dev;
> struct net_failover_info *nfo_info;
> bool slave_is_standby;
>
> + failover_dev = netdev_failover_upper_get(slave_dev);
> + if (!failover_dev)
> + return NOTIFY_DONE;
> +
> nfo_info = netdev_priv(failover_dev);
> primary_dev = rtnl_dereference(nfo_info->primary_dev);
> standby_dev = rtnl_dereference(nfo_info->standby_dev);
>
> + if (slave_dev != primary_dev && slave_dev != standby_dev)
> + return NOTIFY_DONE;
> +
> vlan_vids_del_by_dev(slave_dev, failover_dev);
> dev_uc_unsync(slave_dev, failover_dev);
> dev_mc_unsync(slave_dev, failover_dev);
> dev_close(slave_dev);
>
> - nfo_info = netdev_priv(failover_dev);
> dev_get_stats(failover_dev, &nfo_info->failover_stats);
>
> slave_is_standby = slave_dev->dev.parent == failover_dev->dev.parent;
> @@ -628,22 +700,25 @@ static int net_failover_slave_unregister(struct net_device *slave_dev,
> netdev_info(failover_dev, "failover %s slave:%s unregistered\n",
> slave_is_standby ? "standby" : "primary", slave_dev->name);
>
> - return 0;
> + return NOTIFY_OK;
> }
>
> -static int net_failover_slave_link_change(struct net_device *slave_dev,
> - struct net_device *failover_dev)
> +static int net_failover_link_event(struct net_device *slave_dev)
> +
> {
> - struct net_device *primary_dev, *standby_dev;
> + struct net_device *failover_dev, *primary_dev, *standby_dev;
> struct net_failover_info *nfo_info;
>
> - nfo_info = netdev_priv(failover_dev);
> + failover_dev = netdev_failover_upper_get(slave_dev);
> + if (!failover_dev)
> + return NOTIFY_DONE;
>
> + nfo_info = netdev_priv(failover_dev);
> primary_dev = rtnl_dereference(nfo_info->primary_dev);
> standby_dev = rtnl_dereference(nfo_info->standby_dev);
>
> if (slave_dev != primary_dev && slave_dev != standby_dev)
> - return -ENODEV;
> + return NOTIFY_DONE;
>
> if ((primary_dev && net_failover_xmit_ready(primary_dev)) ||
> (standby_dev && net_failover_xmit_ready(standby_dev))) {
> @@ -657,43 +732,11 @@ static int net_failover_slave_link_change(struct net_device *slave_dev,
>
> net_failover_lower_state_changed(slave_dev, primary_dev, standby_dev);
>
> - return 0;
> + return NOTIFY_DONE;
> }
>
> -static int net_failover_slave_name_change(struct net_device *slave_dev,
> - struct net_device *failover_dev)
> -{
> - struct net_device *primary_dev, *standby_dev;
> - struct net_failover_info *nfo_info;
> -
> - nfo_info = netdev_priv(failover_dev);
> -
> - primary_dev = rtnl_dereference(nfo_info->primary_dev);
> - standby_dev = rtnl_dereference(nfo_info->standby_dev);
> -
> - if (slave_dev != primary_dev && slave_dev != standby_dev)
> - return -ENODEV;
> -
> - /* We need to bring up the slave after the rename by udev in case
> - * open failed with EBUSY when it was registered.
> - */
> - dev_open(slave_dev);
> -
> - return 0;
> -}
> -
> -static struct failover_ops net_failover_ops = {
> - .slave_pre_register = net_failover_slave_pre_register,
> - .slave_register = net_failover_slave_register,
> - .slave_pre_unregister = net_failover_slave_pre_unregister,
> - .slave_unregister = net_failover_slave_unregister,
> - .slave_link_change = net_failover_slave_link_change,
> - .slave_name_change = net_failover_slave_name_change,
> - .slave_handle_frame = net_failover_handle_frame,
> -};
> -
> /**
> - * net_failover_create - Create and register a failover instance
> + * net_failover_create - Create and register a failover device
> *
> * @dev: standby netdev
> *
> @@ -703,13 +746,12 @@ static struct failover_ops net_failover_ops = {
> * the original standby netdev and a VF netdev with the same MAC gets
> * registered as primary netdev.
> *
> - * Return: pointer to failover instance
> + * Return: pointer to failover network device
> */
> -struct failover *net_failover_create(struct net_device *standby_dev)
> +struct net_device *net_failover_create(struct net_device *standby_dev)
> {
> - struct device *dev = standby_dev->dev.parent;
> + struct net_failover_info *nfo_info;
> struct net_device *failover_dev;
> - struct failover *failover;
> int err;
>
> /* Alloc at least 2 queues, for now we are going with 16 assuming
> @@ -717,18 +759,22 @@ struct failover *net_failover_create(struct net_device *standby_dev)
> */
> failover_dev = alloc_etherdev_mq(sizeof(struct net_failover_info), 16);
> if (!failover_dev) {
> - dev_err(dev, "Unable to allocate failover_netdev!\n");
> - return ERR_PTR(-ENOMEM);
> + netdev_err(standby_dev, "Unable to allocate failover_netdev!\n");
> + return NULL;
> }
>
> + nfo_info = netdev_priv(failover_dev);
> dev_net_set(failover_dev, dev_net(standby_dev));
> - SET_NETDEV_DEV(failover_dev, dev);
> + nfo_info->failover_dev = failover_dev;
> + INIT_DELAYED_WORK(&nfo_info->standby_init, net_failover_setup);
>
> failover_dev->netdev_ops = &failover_dev_ops;
> failover_dev->ethtool_ops = &failover_ethtool_ops;
>
> /* Initialize the device options */
> - failover_dev->priv_flags |= IFF_UNICAST_FLT | IFF_NO_QUEUE;
> + failover_dev->priv_flags |= IFF_UNICAST_FLT |
> + IFF_NO_QUEUE |
> + IFF_FAILOVER;
> failover_dev->priv_flags &= ~(IFF_XMIT_DST_RELEASE |
> IFF_TX_SKB_SHARING);
>
> @@ -746,29 +792,38 @@ struct failover *net_failover_create(struct net_device *standby_dev)
> failover_dev->hw_features |= NETIF_F_GSO_ENCAP_ALL;
> failover_dev->features |= failover_dev->hw_features;
>
> - memcpy(failover_dev->dev_addr, standby_dev->dev_addr,
> - failover_dev->addr_len);
> + ether_addr_copy(failover_dev->dev_addr, standby_dev->dev_addr);
> + ether_addr_copy(failover_dev->perm_addr, standby_dev->perm_addr);
>
> failover_dev->min_mtu = standby_dev->min_mtu;
> failover_dev->max_mtu = standby_dev->max_mtu;
>
> - err = register_netdev(failover_dev);
> + netif_carrier_off(failover_dev);
> +
> + rtnl_lock();
> + err = register_netdevice(failover_dev);
> if (err) {
> - dev_err(dev, "Unable to register failover_dev!\n");
> + netdev_err(standby_dev, "Unable to register failover_dev!\n");
> goto err_register_netdev;
> }
>
> - netif_carrier_off(failover_dev);
> + err = netdev_failover_join(standby_dev, failover_dev,
> + net_failover_handle_frame);
> + if (err) {
> + netdev_err(failover_dev, "Unable to join with %s\n",
> + standby_dev->name);
> + goto err_failover_join;
> + }
>
> - failover = failover_register(failover_dev, &net_failover_ops);
> - if (IS_ERR(failover))
> - goto err_failover_register;
> + list_add(&nfo_info->list, &net_failover_list);
> + rtnl_unlock();
>
> - return failover;
> + return failover_dev;
>
> -err_failover_register:
> - unregister_netdev(failover_dev);
> +err_failover_join:
> + unregister_netdevice(failover_dev);
> err_register_netdev:
> + rtnl_unlock();
> free_netdev(failover_dev);
>
> return ERR_PTR(err);
> @@ -786,31 +841,27 @@ EXPORT_SYMBOL_GPL(net_failover_create);
> * netdev. Used by paravirtual drivers that use 3-netdev model.
> *
> */
> -void net_failover_destroy(struct failover *failover)
> +void net_failover_destroy(struct net_device *failover_dev)
> {
> - struct net_failover_info *nfo_info;
> - struct net_device *failover_dev;
> + struct net_failover_info *nfo_info = netdev_priv(failover_dev);
> struct net_device *slave_dev;
>
> - if (!failover)
> - return;
> -
> - failover_dev = rcu_dereference(failover->failover_dev);
> - nfo_info = netdev_priv(failover_dev);
> -
> netif_device_detach(failover_dev);
>
> rtnl_lock();
> -
> slave_dev = rtnl_dereference(nfo_info->primary_dev);
> - if (slave_dev)
> - failover_slave_unregister(slave_dev);
> + if (slave_dev) {
> + netdev_failover_unjoin(slave_dev, failover_dev);
> + dev_put(slave_dev);
> + }
>
> slave_dev = rtnl_dereference(nfo_info->standby_dev);
> - if (slave_dev)
> - failover_slave_unregister(slave_dev);
> + if (slave_dev) {
> + netdev_failover_unjoin(slave_dev, failover_dev);
> + dev_put(slave_dev);
> + }
>
> - failover_unregister(failover);
> + list_del(&nfo_info->list);
>
> unregister_netdevice(failover_dev);
>
> @@ -820,9 +871,53 @@ void net_failover_destroy(struct failover *failover)
> }
> EXPORT_SYMBOL_GPL(net_failover_destroy);
>
> +static int net_failover_event(struct notifier_block *this,
> + unsigned long event, void *ptr)
> +{
> + struct net_device *event_dev = netdev_notifier_info_to_dev(ptr);
> +
> + /* Skip parent events */
> + if (netif_is_failover(event_dev))
> + return NOTIFY_DONE;
> +
> + /* Avoid non-Ethernet type devices */
> + if (event_dev->type != ARPHRD_ETHER)
> + return NOTIFY_DONE;
> +
> + /* Avoid Vlan dev with same MAC registering as VF */
> + if (is_vlan_dev(event_dev))
> + return NOTIFY_DONE;
> +
> + /* Avoid Bonding master dev with same MAC registering as VF */
> + if ((event_dev->priv_flags & IFF_BONDING) &&
> + (event_dev->flags & IFF_MASTER))
> + return NOTIFY_DONE;
> +
> + switch (event) {
> + case NETDEV_REGISTER:
> + return net_failover_register_event(event_dev);
> +
> + case NETDEV_UNREGISTER:
> + return net_failover_unregister_event(event_dev);
> +
> + case NETDEV_UP:
> + case NETDEV_DOWN:
> + case NETDEV_CHANGE:
> + return net_failover_link_event(event_dev);
> +
> + default:
> + return NOTIFY_DONE;
> + }
> +}
> +
> +static struct notifier_block net_failover_notifier = {
> + .notifier_call = net_failover_event,
> +};
> +
> static __init int
> net_failover_init(void)
> {
> + register_netdevice_notifier(&net_failover_notifier);
> return 0;
> }
> module_init(net_failover_init);
> @@ -830,6 +925,7 @@ module_init(net_failover_init);
> static __exit
> void net_failover_exit(void)
> {
> + unregister_netdevice_notifier(&net_failover_notifier);
> }
> module_exit(net_failover_exit);
>
> diff --git a/drivers/net/virtio_net.c b/drivers/net/virtio_net.c
> index 6d710b8b41c5..b40ae28dac93 100644
> --- a/drivers/net/virtio_net.c
> +++ b/drivers/net/virtio_net.c
> @@ -215,7 +215,7 @@ struct virtnet_info {
> unsigned long guest_offloads;
>
> /* failover when STANDBY feature enabled */
> - struct failover *failover;
> + struct net_device *failover;
> };
>
> struct padded_vnet_hdr {
> @@ -2930,11 +2930,10 @@ static int virtnet_probe(struct virtio_device *vdev)
> virtnet_init_settings(dev);
>
> if (virtio_has_feature(vdev, VIRTIO_NET_F_STANDBY)) {
> - vi->failover = net_failover_create(vi->dev);
> - if (IS_ERR(vi->failover)) {
> - err = PTR_ERR(vi->failover);
> + err = -ENOMEM;
> + vi->failover = net_failover_create(dev);
> + if (!vi->failover)
> goto free_vqs;
> - }
> }
>
> err = register_netdev(dev);
> diff --git a/include/net/failover.h b/include/net/failover.h
> index bb15438f39c7..22d6c1369101 100644
> --- a/include/net/failover.h
> +++ b/include/net/failover.h
> @@ -6,31 +6,10 @@
>
> #include <linux/netdevice.h>
>
> -struct failover_ops {
> - int (*slave_pre_register)(struct net_device *slave_dev,
> - struct net_device *failover_dev);
> - int (*slave_register)(struct net_device *slave_dev,
> - struct net_device *failover_dev);
> - int (*slave_pre_unregister)(struct net_device *slave_dev,
> - struct net_device *failover_dev);
> - int (*slave_unregister)(struct net_device *slave_dev,
> - struct net_device *failover_dev);
> - int (*slave_link_change)(struct net_device *slave_dev,
> - struct net_device *failover_dev);
> - int (*slave_name_change)(struct net_device *slave_dev,
> - struct net_device *failover_dev);
> - rx_handler_result_t (*slave_handle_frame)(struct sk_buff **pskb);
> -};
> -
> -struct failover {
> - struct list_head list;
> - struct net_device __rcu *failover_dev;
> - struct failover_ops __rcu *ops;
> -};
> -
> -struct failover *failover_register(struct net_device *dev,
> - struct failover_ops *ops);
> -void failover_unregister(struct failover *failover);
> -int failover_slave_unregister(struct net_device *slave_dev);
> +int netdev_failover_join(struct net_device *lower, struct net_device *upper,
> + rx_handler_func_t *rx_handler);
> +struct net_device *netdev_failover_upper_get(struct net_device *lower);
> +void netdev_failover_unjoin(struct net_device *lower,
> + struct net_device *upper);
>
> #endif /* _FAILOVER_H */
> diff --git a/include/net/net_failover.h b/include/net/net_failover.h
> index b12a1c469d1c..a99b3b00b4e3 100644
> --- a/include/net/net_failover.h
> +++ b/include/net/net_failover.h
> @@ -6,35 +6,7 @@
>
> #include <net/failover.h>
>
> -/* failover state */
> -struct net_failover_info {
> - /* primary netdev with same MAC */
> - struct net_device __rcu *primary_dev;
> -
> - /* standby netdev */
> - struct net_device __rcu *standby_dev;
> -
> - /* primary netdev stats */
> - struct rtnl_link_stats64 primary_stats;
> -
> - /* standby netdev stats */
> - struct rtnl_link_stats64 standby_stats;
> -
> - /* aggregated stats */
> - struct rtnl_link_stats64 failover_stats;
> -
> - /* spinlock while updating stats */
> - spinlock_t stats_lock;
> -};
> -
> -struct failover *net_failover_create(struct net_device *standby_dev);
> -void net_failover_destroy(struct failover *failover);
> -
> -#define FAILOVER_VLAN_FEATURES (NETIF_F_HW_CSUM | NETIF_F_SG | \
> - NETIF_F_FRAGLIST | NETIF_F_ALL_TSO | \
> - NETIF_F_HIGHDMA | NETIF_F_LRO)
> -
> -#define FAILOVER_ENC_FEATURES (NETIF_F_HW_CSUM | NETIF_F_SG | \
> - NETIF_F_RXCSUM | NETIF_F_ALL_TSO)
> +struct net_device *net_failover_create(struct net_device *standby_dev);
> +void net_failover_destroy(struct net_device *failover_dev);
>
> #endif /* _NET_FAILOVER_H */
> diff --git a/net/Kconfig b/net/Kconfig
> index f738a6f27665..697d84202695 100644
> --- a/net/Kconfig
> +++ b/net/Kconfig
> @@ -433,17 +433,8 @@ config PAGE_POOL
> bool
>
> config FAILOVER
> - tristate "Generic failover module"
> - help
> - The failover module provides a generic interface for paravirtual
> - drivers to register a netdev and a set of ops with a failover
> - instance. The ops are used as event handlers that get called to
> - handle netdev register/unregister/link change/name change events
> - on slave pci ethernet devices with the same mac address as the
> - failover netdev. This enables paravirtual drivers to use a
> - VF as an accelerated low latency datapath. It also allows live
> - migration of VMs with direct attached VFs by failing over to the
> - paravirtual datapath when the VF is unplugged.
> + bool
> + default n
>
> endif # if NET
>
> diff --git a/net/core/failover.c b/net/core/failover.c
> index 4a92a98ccce9..499f0fd7e4d3 100644
> --- a/net/core/failover.c
> +++ b/net/core/failover.c
> @@ -1,10 +1,8 @@
> // SPDX-License-Identifier: GPL-2.0
> /* Copyright (c) 2018, Intel Corporation. */
>
> -/* A common module to handle registrations and notifications for paravirtual
> +/* A library for managing chained upper/oower devices such as
> * drivers to enable accelerated datapath and support VF live migration.
> - *
> - * The notifier and event handling code is based on netvsc driver.
> */
>
> #include <linux/module.h>
> @@ -14,302 +12,62 @@
> #include <linux/if_vlan.h>
> #include <net/failover.h>
>
> -static LIST_HEAD(failover_list);
> -static DEFINE_SPINLOCK(failover_lock);
> -
> -static struct net_device *failover_get_bymac(u8 *mac, struct failover_ops **ops)
> -{
> - struct net_device *failover_dev;
> - struct failover *failover;
> -
> - spin_lock(&failover_lock);
> - list_for_each_entry(failover, &failover_list, list) {
> - failover_dev = rtnl_dereference(failover->failover_dev);
> - if (ether_addr_equal(failover_dev->perm_addr, mac)) {
> - *ops = rtnl_dereference(failover->ops);
> - spin_unlock(&failover_lock);
> - return failover_dev;
> - }
> - }
> - spin_unlock(&failover_lock);
> - return NULL;
> -}
> -
> -/**
> - * failover_slave_register - Register a slave netdev
> - *
> - * @slave_dev: slave netdev that is being registered
> - *
> - * Registers a slave device to a failover instance. Only ethernet devices
> - * are supported.
> - */
> -static int failover_slave_register(struct net_device *slave_dev)
> +/* failover_join - Join an lower netdev with an upper device. */
> +int netdev_failover_join(struct net_device *lower_dev,
> + struct net_device *upper_dev,
> + rx_handler_func_t *rx_handler)
> {
> - struct netdev_lag_upper_info lag_upper_info;
> - struct net_device *failover_dev;
> - struct failover_ops *fops;
> int err;
>
> - if (slave_dev->type != ARPHRD_ETHER)
> - goto done;
> -
> ASSERT_RTNL();
>
> - failover_dev = failover_get_bymac(slave_dev->perm_addr, &fops);
> - if (!failover_dev)
> - goto done;
> + /* Don't allow joining devices of different protocols */
> + if (upper_dev->type != lower_dev->type)
> + return -EINVAL;
>
> - if (fops && fops->slave_pre_register &&
> - fops->slave_pre_register(slave_dev, failover_dev))
> - goto done;
> -
> - err = netdev_rx_handler_register(slave_dev, fops->slave_handle_frame,
> - failover_dev);
> + err = netdev_rx_handler_register(lower_dev, rx_handler, upper_dev);
> if (err) {
> - netdev_err(slave_dev, "can not register failover rx handler (err = %d)\n",
> + netdev_err(lower_dev,
> + "can not register failover rx handler (err = %d)\n",
> err);
> - goto done;
> + return err;
> }
>
> - lag_upper_info.tx_type = NETDEV_LAG_TX_TYPE_ACTIVEBACKUP;
> - err = netdev_master_upper_dev_link(slave_dev, failover_dev, NULL,
> - &lag_upper_info, NULL);
> + err = netdev_master_upper_dev_link(lower_dev, upper_dev, NULL,
> + NULL, NULL);
> if (err) {
> - netdev_err(slave_dev, "can not set failover device %s (err = %d)\n",
> - failover_dev->name, err);
> - goto err_upper_link;
> + netdev_err(lower_dev,
> + "can not set failover device %s (err = %d)\n",
> + upper_dev->name, err);
> + netdev_rx_handler_unregister(lower_dev);
> + return err;
> }
>
> - slave_dev->priv_flags |= IFF_FAILOVER_SLAVE;
> -
> - if (fops && fops->slave_register &&
> - !fops->slave_register(slave_dev, failover_dev))
> - return NOTIFY_OK;
> -
> - netdev_upper_dev_unlink(slave_dev, failover_dev);
> - slave_dev->priv_flags &= ~IFF_FAILOVER_SLAVE;
> -err_upper_link:
> - netdev_rx_handler_unregister(slave_dev);
> -done:
> - return NOTIFY_DONE;
> -}
> -
> -/**
> - * failover_slave_unregister - Unregister a slave netdev
> - *
> - * @slave_dev: slave netdev that is being unregistered
> - *
> - * Unregisters a slave device from a failover instance.
> - */
> -int failover_slave_unregister(struct net_device *slave_dev)
> -{
> - struct net_device *failover_dev;
> - struct failover_ops *fops;
> -
> - if (!netif_is_failover_slave(slave_dev))
> - goto done;
> -
> - ASSERT_RTNL();
> -
> - failover_dev = failover_get_bymac(slave_dev->perm_addr, &fops);
> - if (!failover_dev)
> - goto done;
> -
> - if (fops && fops->slave_pre_unregister &&
> - fops->slave_pre_unregister(slave_dev, failover_dev))
> - goto done;
> -
> - netdev_rx_handler_unregister(slave_dev);
> - netdev_upper_dev_unlink(slave_dev, failover_dev);
> - slave_dev->priv_flags &= ~IFF_FAILOVER_SLAVE;
> -
> - if (fops && fops->slave_unregister &&
> - !fops->slave_unregister(slave_dev, failover_dev))
> - return NOTIFY_OK;
> -
> -done:
> - return NOTIFY_DONE;
> + dev_hold(lower_dev);
> + lower_dev->priv_flags |= IFF_FAILOVER_SLAVE;
> + return 0;
> }
> -EXPORT_SYMBOL_GPL(failover_slave_unregister);
> +EXPORT_SYMBOL_GPL(netdev_failover_join);
>
> -static int failover_slave_link_change(struct net_device *slave_dev)
> +/* Find upper network device for failover slave device */
> +struct net_device *netdev_failover_upper_get(struct net_device *lower_dev)
> {
> - struct net_device *failover_dev;
> - struct failover_ops *fops;
> -
> - if (!netif_is_failover_slave(slave_dev))
> - goto done;
> -
> - ASSERT_RTNL();
> -
> - failover_dev = failover_get_bymac(slave_dev->perm_addr, &fops);
> - if (!failover_dev)
> - goto done;
> -
> - if (!netif_running(failover_dev))
> - goto done;
> + if (!netif_is_failover_slave(lower_dev))
> + return NULL;
>
> - if (fops && fops->slave_link_change &&
> - !fops->slave_link_change(slave_dev, failover_dev))
> - return NOTIFY_OK;
> -
> -done:
> - return NOTIFY_DONE;
> + return netdev_master_upper_dev_get(lower_dev);
> }
> +EXPORT_SYMBOL_GPL(netdev_failover_upper_get);
>
> -static int failover_slave_name_change(struct net_device *slave_dev)
> +/* failover_unjoin - Break connection between lower and upper device. */
> +void netdev_failover_unjoin(struct net_device *lower_dev,
> + struct net_device *upper_dev)
> {
> - struct net_device *failover_dev;
> - struct failover_ops *fops;
> -
> - if (!netif_is_failover_slave(slave_dev))
> - goto done;
> -
> ASSERT_RTNL();
>
> - failover_dev = failover_get_bymac(slave_dev->perm_addr, &fops);
> - if (!failover_dev)
> - goto done;
> -
> - if (!netif_running(failover_dev))
> - goto done;
> -
> - if (fops && fops->slave_name_change &&
> - !fops->slave_name_change(slave_dev, failover_dev))
> - return NOTIFY_OK;
> -
> -done:
> - return NOTIFY_DONE;
> -}
> -
> -static int
> -failover_event(struct notifier_block *this, unsigned long event, void *ptr)
> -{
> - struct net_device *event_dev = netdev_notifier_info_to_dev(ptr);
> -
> - /* Skip parent events */
> - if (netif_is_failover(event_dev))
> - return NOTIFY_DONE;
> -
> - switch (event) {
> - case NETDEV_REGISTER:
> - return failover_slave_register(event_dev);
> - case NETDEV_UNREGISTER:
> - return failover_slave_unregister(event_dev);
> - case NETDEV_UP:
> - case NETDEV_DOWN:
> - case NETDEV_CHANGE:
> - return failover_slave_link_change(event_dev);
> - case NETDEV_CHANGENAME:
> - return failover_slave_name_change(event_dev);
> - default:
> - return NOTIFY_DONE;
> - }
> -}
> -
> -static struct notifier_block failover_notifier = {
> - .notifier_call = failover_event,
> -};
> -
> -static void
> -failover_existing_slave_register(struct net_device *failover_dev)
> -{
> - struct net *net = dev_net(failover_dev);
> - struct net_device *dev;
> -
> - rtnl_lock();
> - for_each_netdev(net, dev) {
> - if (netif_is_failover(dev))
> - continue;
> - if (ether_addr_equal(failover_dev->perm_addr, dev->perm_addr))
> - failover_slave_register(dev);
> - }
> - rtnl_unlock();
> -}
> -
> -/**
> - * failover_register - Register a failover instance
> - *
> - * @dev: failover netdev
> - * @ops: failover ops
> - *
> - * Allocate and register a failover instance for a failover netdev. ops
> - * provides handlers for slave device register/unregister/link change/
> - * name change events.
> - *
> - * Return: pointer to failover instance
> - */
> -struct failover *failover_register(struct net_device *dev,
> - struct failover_ops *ops)
> -{
> - struct failover *failover;
> -
> - if (dev->type != ARPHRD_ETHER)
> - return ERR_PTR(-EINVAL);
> -
> - failover = kzalloc(sizeof(*failover), GFP_KERNEL);
> - if (!failover)
> - return ERR_PTR(-ENOMEM);
> -
> - rcu_assign_pointer(failover->ops, ops);
> - dev_hold(dev);
> - dev->priv_flags |= IFF_FAILOVER;
> - rcu_assign_pointer(failover->failover_dev, dev);
> -
> - spin_lock(&failover_lock);
> - list_add_tail(&failover->list, &failover_list);
> - spin_unlock(&failover_lock);
> -
> - netdev_info(dev, "failover master:%s registered\n", dev->name);
> -
> - failover_existing_slave_register(dev);
> -
> - return failover;
> -}
> -EXPORT_SYMBOL_GPL(failover_register);
> -
> -/**
> - * failover_unregister - Unregister a failover instance
> - *
> - * @failover: pointer to failover instance
> - *
> - * Unregisters and frees a failover instance.
> - */
> -void failover_unregister(struct failover *failover)
> -{
> - struct net_device *failover_dev;
> -
> - failover_dev = rcu_dereference(failover->failover_dev);
> -
> - netdev_info(failover_dev, "failover master:%s unregistered\n",
> - failover_dev->name);
> -
> - failover_dev->priv_flags &= ~IFF_FAILOVER;
> - dev_put(failover_dev);
> -
> - spin_lock(&failover_lock);
> - list_del(&failover->list);
> - spin_unlock(&failover_lock);
> -
> - kfree(failover);
> + netdev_rx_handler_unregister(lower_dev);
> + netdev_upper_dev_unlink(lower_dev, upper_dev);
> + dev_put(lower_dev);
> + lower_dev->priv_flags &= ~IFF_FAILOVER_SLAVE;
> }
> -EXPORT_SYMBOL_GPL(failover_unregister);
> -
> -static __init int
> -failover_init(void)
> -{
> - register_netdevice_notifier(&failover_notifier);
> -
> - return 0;
> -}
> -module_init(failover_init);
> -
> -static __exit
> -void failover_exit(void)
> -{
> - unregister_netdevice_notifier(&failover_notifier);
> -}
> -module_exit(failover_exit);
> -
> -MODULE_DESCRIPTION("Generic failover infrastructure/interface");
> -MODULE_LICENSE("GPL v2");
> +EXPORT_SYMBOL_GPL(netdev_failover_unjoin);
> --
> 2.17.1
^ permalink raw reply
* Re: [Patch net v2] netdev-FAQ: clarify DaveM's position for stable backports
From: David Miller @ 2018-06-05 18:19 UTC (permalink / raw)
To: xiyou.wangcong; +Cc: netdev, stable, gregkh
In-Reply-To: <20180605164813.12525-1-xiyou.wangcong@gmail.com>
From: Cong Wang <xiyou.wangcong@gmail.com>
Date: Tue, 5 Jun 2018 09:48:13 -0700
> Per discussion with David at netconf 2018, let's clarify
> DaveM's position of handling stable backports in netdev-FAQ.
>
> This is important for people relying on upstream -stable
> releases.
>
> Cc: stable@vger.kernel.org
> Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
> Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com>
Applied and queued up for -stable, thanks Cong.
^ permalink raw reply
* Re: [PATCH iproute2] configure: require libmnl
From: Simon Horman @ 2018-06-05 18:15 UTC (permalink / raw)
To: Stephen Hemminger; +Cc: netdev
In-Reply-To: <20180531193209.27406-1-stephen@networkplumber.org>
On Thu, May 31, 2018 at 03:32:09PM -0400, Stephen Hemminger wrote:
> Several users of BPF and other features are trying to build without
> libmnl, then complaining that features don't work. The time has
> come to require libmnl to build iproute2.
>
> Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Reviewed-by: Simon Horman <simon.horman@netronome.com>
^ permalink raw reply
* Re: [PATCH net] failover: eliminate callback hell
From: David Miller @ 2018-06-05 18:14 UTC (permalink / raw)
To: stephen
Cc: sridhar.samudrala, kys, haiyangz, mst, alexander.h.duyck, jiri,
netdev, sthemmin, jesse.brandeburg
In-Reply-To: <20180605104510.611bd247@xeon-e3>
From: Stephen Hemminger <stephen@networkplumber.org>
Date: Tue, 5 Jun 2018 10:45:10 -0700
> I said it wasn't tested. Not surprising. Don't have a version of KVM
> that supports standby (and not going to build KVM from scratch for
> this).
It would definitely help me if you put "RFC" in the subject line
for patches which aren't tested :-)
Thanks.
^ permalink raw reply
page: next (older) | prev (newer) | latest
- recent:[subjects (threaded)|topics (new)|topics (active)]
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox