[PATCH] rt2x00: Use ops name instead of device name

[PATCH] rt2x00: Use ops name instead of device name

[Steven Rostedt]

Recently, someone complained that they see the PCI address of a device
in the interrupts file instead of the device name.

19:   73474106          0   IO-APIC-fasteoi   ata_piix, uhci_hcd:usb6, 0000:05:01.0

>From lspci:

05:01.0 Network controller: RaLink RT2561/RT61 rev B 802.11g
  Subsystem: D-Link System Inc AirPlus G DWL-G510 Wireless Network Adapter (Rev.C)
  Flags: bus master, slow devsel, latency 32,
  IRQ 19  Memory at f9100000 (32-bit, non-prefetchable) [size=32K]
  Capabilities: <access denied>
  Kernel driver in use: rt61pci
  Kernel modules: rt61pci

Investigating this, it seems to be the result of the pci_name() used in
rt2x00pci_probe().

Since the struct rt2x00_ops records the module name for the device as
well as the ops, it would make more sense to show that in the interrupt
output instead.

Signed-off-by: Steven Rostedt <rostedt@goodmis.org>

diff --git a/drivers/net/wireless/rt2x00/rt2x00pci.c b/drivers/net/wireless/rt2x00/rt2x00pci.c
index ace0b66..06575dc 100644
--- a/drivers/net/wireless/rt2x00/rt2x00pci.c
+++ b/drivers/net/wireless/rt2x00/rt2x00pci.c
@@ -284,7 +284,7 @@ int rt2x00pci_probe(struct pci_dev *pci_dev, const struct pci_device_id *id)
 	rt2x00dev->ops = ops;
 	rt2x00dev->hw = hw;
 	rt2x00dev->irq = pci_dev->irq;
-	rt2x00dev->name = pci_name(pci_dev);
+	rt2x00dev->name = ops->name;
 
 	if (pci_is_pcie(pci_dev))
 		rt2x00_set_chip_intf(rt2x00dev, RT2X00_CHIP_INTF_PCIE);

Re: [PATCH] rt2x00: Use ops name instead of device name

[Ben Hutchings]

On Tue, 2011-02-22 at 08:12 -0500, Steven Rostedt wrote:
> Recently, someone complained that they see the PCI address of a device
> in the interrupts file instead of the device name.
> 
> 19:   73474106          0   IO-APIC-fasteoi   ata_piix, uhci_hcd:usb6, 0000:05:01.0
[...]
> diff --git a/drivers/net/wireless/rt2x00/rt2x00pci.c b/drivers/net/wireless/rt2x00/rt2x00pci.c
> index ace0b66..06575dc 100644
> --- a/drivers/net/wireless/rt2x00/rt2x00pci.c
> +++ b/drivers/net/wireless/rt2x00/rt2x00pci.c
> @@ -284,7 +284,7 @@ int rt2x00pci_probe(struct pci_dev *pci_dev, const struct pci_device_id *id)
>  	rt2x00dev->ops = ops;
>  	rt2x00dev->hw = hw;
>  	rt2x00dev->irq = pci_dev->irq;
> -	rt2x00dev->name = pci_name(pci_dev);
> +	rt2x00dev->name = ops->name;
>  
>  	if (pci_is_pcie(pci_dev))
>  		rt2x00_set_chip_intf(rt2x00dev, RT2X00_CHIP_INTF_PCIE);

But then how can users distinguish the IRQs for multiple devices handled
by the same driver?  (Probably unusual for WLAN devices, but still
possible.)

I assume you can't use a net device name as there may be multiple net
devices per bus device?

Ben.

-- 
Ben Hutchings, Senior Software Engineer, Solarflare Communications
Not speaking for my employer; that's the marketing department's job.
They asked us to note that Solarflare product names are trademarked.

Re: [PATCH] rt2x00: Use ops name instead of device name

[Steven Rostedt]

On Tue, 2011-02-22 at 13:34 +0000, Ben Hutchings wrote:

> > 19:   73474106          0   IO-APIC-fasteoi   ata_piix, uhci_hcd:usb6, 0000:05:01.0

> >  	rt2x00dev->irq = pci_dev->irq;
> > -	rt2x00dev->name = pci_name(pci_dev);
> > +	rt2x00dev->name = ops->name;


> But then how can users distinguish the IRQs for multiple devices handled
> by the same driver?  (Probably unusual for WLAN devices, but still
> possible.)
> 
> I assume you can't use a net device name as there may be multiple net
> devices per bus device?

Honestly, I do not know this code well enough, but this patch seemed to
solve the problem at hand. Hence I sent it out to the experts hoping
they either take this patch or come up with a proper solution ;)

In any case, just posting the pci address is not a pretty answer.

-- Steve

Re: [PATCH] rt2x00: Use ops name instead of device name

[Ivo Van Doorn]

Hi,

On Tue, Feb 22, 2011 at 2:44 PM, Steven Rostedt <rostedt@goodmis.org> wrote:
> On Tue, 2011-02-22 at 13:34 +0000, Ben Hutchings wrote:
>
>> > 19:   73474106          0   IO-APIC-fasteoi   ata_piix, uhci_hcd:usb6, 0000:05:01.0
>
>> >     rt2x00dev->irq = pci_dev->irq;
>> > -   rt2x00dev->name = pci_name(pci_dev);
>> > +   rt2x00dev->name = ops->name;
>
>
>> But then how can users distinguish the IRQs for multiple devices handled
>> by the same driver?  (Probably unusual for WLAN devices, but still
>> possible.)
>>
>> I assume you can't use a net device name as there may be multiple net
>> devices per bus device?
>
> Honestly, I do not know this code well enough, but this patch seemed to
> solve the problem at hand. Hence I sent it out to the experts hoping
> they either take this patch or come up with a proper solution ;)
>
> In any case, just posting the pci address is not a pretty answer.

I just checked the other wireless drivers, and they all seem to use
the modulename. So I guess your patch is correct. You can add my:

Acked-by: Ivo van Doorn <IvDoorn@gmail.com>

Re: [PATCH] rt2x00: Use ops name instead of device name

[John W. Linville]

On Tue, Feb 22, 2011 at 08:33:45PM +0100, Ivo Van Doorn wrote:
> Hi,
> 
> On Tue, Feb 22, 2011 at 2:44 PM, Steven Rostedt <rostedt@goodmis.org> wrote:
> > On Tue, 2011-02-22 at 13:34 +0000, Ben Hutchings wrote:
> >
> >> > 19:   73474106          0   IO-APIC-fasteoi   ata_piix, uhci_hcd:usb6, 0000:05:01.0
> >
> >> >     rt2x00dev->irq = pci_dev->irq;
> >> > -   rt2x00dev->name = pci_name(pci_dev);
> >> > +   rt2x00dev->name = ops->name;
> >
> >
> >> But then how can users distinguish the IRQs for multiple devices handled
> >> by the same driver?  (Probably unusual for WLAN devices, but still
> >> possible.)
> >>
> >> I assume you can't use a net device name as there may be multiple net
> >> devices per bus device?
> >
> > Honestly, I do not know this code well enough, but this patch seemed to
> > solve the problem at hand. Hence I sent it out to the experts hoping
> > they either take this patch or come up with a proper solution ;)
> >
> > In any case, just posting the pci address is not a pretty answer.
> 
> I just checked the other wireless drivers, and they all seem to use
> the modulename. So I guess your patch is correct. You can add my:
> 
> Acked-by: Ivo van Doorn <IvDoorn@gmail.com>

I completely missed this series -- please be sure to Cc:
linux-wireless@vger.kernel.org if you resend, and/or resend the series
directly to me?

Thanks!

John
-- 
John W. Linville		Someday the world will need a hero, and you
linville@tuxdriver.com			might be all we have.  Be ready.

Re: [PATCH] rt2x00: Use ops name instead of device name

[Steven Rostedt]

On Tue, 2011-02-22 at 14:47 -0500, John W. Linville wrote:

> I completely missed this series -- please be sure to Cc:
> linux-wireless@vger.kernel.org if you resend, and/or resend the series
> directly to me?

Hi John,

Not sure which series you were talking about. I sent out a single patch
and Cc'd you on it. But I should have sent it to linux-wireless, instead
of the net-dev list.

I don't plan on sending out any new patches. This should fix the issue.

Thanks,

-- Steve

Re: [PATCH] rt2x00: Use ops name instead of device name

[Steven Rostedt]

On Tue, 2011-02-22 at 20:33 +0100, Ivo Van Doorn wrote:
> Hi,
> 
> On Tue, Feb 22, 2011 at 2:44 PM, Steven Rostedt <rostedt@goodmis.org> wrote:
> > On Tue, 2011-02-22 at 13:34 +0000, Ben Hutchings wrote:
> >
> >> > 19:   73474106          0   IO-APIC-fasteoi   ata_piix, uhci_hcd:usb6, 0000:05:01.0
> >
> >> >     rt2x00dev->irq = pci_dev->irq;
> >> > -   rt2x00dev->name = pci_name(pci_dev);
> >> > +   rt2x00dev->name = ops->name;
> >
> >
> >> But then how can users distinguish the IRQs for multiple devices handled
> >> by the same driver?  (Probably unusual for WLAN devices, but still
> >> possible.)
> >>
> >> I assume you can't use a net device name as there may be multiple net
> >> devices per bus device?
> >
> > Honestly, I do not know this code well enough, but this patch seemed to
> > solve the problem at hand. Hence I sent it out to the experts hoping
> > they either take this patch or come up with a proper solution ;)
> >
> > In any case, just posting the pci address is not a pretty answer.
> 
> I just checked the other wireless drivers, and they all seem to use
> the modulename. So I guess your patch is correct. You can add my:
> 
> Acked-by: Ivo van Doorn <IvDoorn@gmail.com>

Thanks!

Actually, to answer Ben's question. lspci should give you what you want.

On my local box (not the one mentioned above):

$ lspci -vv
[...]
00:0a.0 Bridge: nVidia Corporation CK804 Ethernet Controller (rev a3)
        Subsystem: Tyan Computer Tomcat K8E (S2865)
        Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B- DisINTx-
        Status: Cap+ 66MHz+ UDF- FastB2B+ ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR- INTx+
        Latency: 0 (250ns min, 5000ns max)

        Interrupt: pin A routed to IRQ 23

        Region 0: Memory at febf9000 (32-bit, non-prefetchable) [size=4K]
        Region 1: I/O ports at b400 [size=8]
        Capabilities: <access denied>
        Kernel driver in use: forcedeth


(spaces added by me)

$ cat /proc/interrupts
[...]
 23:        547     648564   IO-APIC-fasteoi   ohci_hcd:usb2, eth0


-- Steve