One chart on pci bridge and its bus and their children

One chart on pci bridge and its bus and their children

[Richard Yang]

All, 

I draw a chart on the relationship of bridge/bus/devics.

I am not sure it is exact, so please correct me if it is not correct.

BTW, the pci_bus->self, pci_bus->bridge seem to point the same thing.
So this is really necessary to do this?

                            Bus 0(pci_bus)                      
                            +--------------------+<----+
                            |                    |     |
                            +--------------------+     |
                                                       |
                                                       |
                             Bridge(pci_dev)           |                     
                            +--------------------+<----|---+--------------+
                            |subordinate(pci_bus)|-+   |   |              |
                            +--------------------+ |   |   |              |
                                                   |   |   |              |
                                                   |   |   |              |
                                                   |   |   |              |
                            Bus 1(pci_bus)         |   |   |              |
                       +--->+-------------------+<-+   |   |              |
                       |    |parent(pci_bus)    |------+   |              |
                       |    +-------------------+          |              |
                       |    |self(pci_dev)      |----------+              |
                       |    |bridge(device)     |                         |
                       |    +-------------------+                         |
                       |    |devices(list_head) |                         |
                       |    +--------+----------+                         |
                       |             |                                    |
                       |      ----+--+---+----                            |
   pci_dev             |          |      |          pci_dev               |
   +---------------+<--|----------+      +--------->+-----------------+   |
   |bus(pci_bus)   |---+                            |bus(pci_bus)     |   |
   +---------------+                                +-----------------+   |
   |dev.parent     |                                |dev.parent       |---+
   +---------------+                                +-----------------+
-- 
Richard Yang
Help you, Help me

Re: One chart on pci bridge and its bus and their children

[Ram Pai]

On Tue, Jun 05, 2012 at 11:26:27AM +0800, Richard Yang wrote:
> 
> All, 
> 
> I draw a chart on the relationship of bridge/bus/devics.
> 
> I am not sure it is exact, so please correct me if it is not correct.

A clearer representation can be found at http://tldp.org/LDP/tlk/dd/pci.html		
in section 6.6.1

> 
> BTW, the pci_bus->self, pci_bus->bridge seem to point the same thing.
> So this is really necessary to do this?

pci_bus->self points to the pci_dev structure, and pci_bus->bridge
points to the device structure. Both correspond to the same bridge. BTW: device
structure is the basic structure to represent any device in the system,
not necessarily pci device.


> 
>                             Bus 0(pci_bus)                      
>                             +--------------------+<----+
>                             |                    |     |
>                             +--------------------+     |
>                                                        |
>                                                        |
>                              Bridge(pci_dev)           |                     
>                             +--------------------+<----|---+--------------+
>                             |subordinate(pci_bus)|-+   |   |              |
>                             +--------------------+ |   |   |              |
>                                                    |   |   |              |
>                                                    |   |   |              |
>                                                    |   |   |              |
>                             Bus 1(pci_bus)         |   |   |              |
>                        +--->+-------------------+<-+   |   |              |
>                        |    |parent(pci_bus)    |------+   |              |
>                        |    +-------------------+          |              |
>                        |    |self(pci_dev)      |----------+              |
>                        |    |bridge(device)     |                         |
>                        |    +-------------------+                         |
>                        |    |devices(list_head) |                         |
>                        |    +--------+----------+                         |
>                        |             |                                    |
>                        |      ----+--+---+----                            |
>    pci_dev             |          |      |          pci_dev               |
>    +---------------+<--|----------+      +--------->+-----------------+   |
>    |bus(pci_bus)   |---+                            |bus(pci_bus)     |   |
>    +---------------+                                +-----------------+   |
>    |dev.parent     |                                |dev.parent       |---+
>    +---------------+                                +-----------------+

Re: One chart on pci bridge and its bus and their children

[Bjorn Helgaas]

On Mon, Jun 4, 2012 at 9:11 PM, Ram Pai <linuxram@us.ibm.com> wrote:
> On Tue, Jun 05, 2012 at 11:26:27AM +0800, Richard Yang wrote:

>> BTW, the pci_bus->self, pci_bus->bridge seem to point the same thing.
>> So this is really necessary to do this?
>
> pci_bus->self points to the pci_dev structure, and pci_bus->bridge
> points to the device structure. Both correspond to the same bridge. BTW: device
> structure is the basic structure to represent any device in the system,
> not necessarily pci device.

That's true, but I don't think it answers the question.  If we have:

    pci_bus->bridge == pci_bus->self->dev

why would we need both "self" and "bridge"?  It would be interesting
to try to remove "bridge" and replace uses of it with "self->dev".

Re: One chart on pci bridge and its bus and their children

[Yinghai Lu]

On Mon, Jun 4, 2012 at 9:37 PM, Bjorn Helgaas <bhelgaas@google.com> wrote:
> That's true, but I don't think it answers the question.  If we have:
>
>    pci_bus->bridge == pci_bus->self->dev
>
> why would we need both "self" and "bridge"?  It would be interesting
> to try to remove "bridge" and replace uses of it with "self->dev".

then how about root bus?

root bus ->self should be NULL.

root bus ->bridge is to the hostbridge->dev.

Yinghai

Re: One chart on pci bridge and its bus and their children

[Yinghai Lu]

On Mon, Jun 4, 2012 at 10:18 PM, Yinghai Lu <yinghai@kernel.org> wrote:
> On Mon, Jun 4, 2012 at 9:37 PM, Bjorn Helgaas <bhelgaas@google.com> wrote:
>> That's true, but I don't think it answers the question.  If we have:
>>
>>    pci_bus->bridge == pci_bus->self->dev
>>
>> why would we need both "self" and "bridge"?  It would be interesting
>> to try to remove "bridge" and replace uses of it with "self->dev".
>
> then how about root bus?
>
> root bus ->self should be NULL.
>
> root bus ->bridge is to the hostbridge->dev.

also

b->bridge = get_device(&bridge->dev);

so there is one reference held by pci bus.

Re: One chart on pci bridge and its bus and their children

[Richard Yang]

On Mon, Jun 04, 2012 at 10:20:51PM -0700, Yinghai Lu wrote:
>On Mon, Jun 4, 2012 at 10:18 PM, Yinghai Lu <yinghai@kernel.org> wrote:
>> On Mon, Jun 4, 2012 at 9:37 PM, Bjorn Helgaas <bhelgaas@google.com> wrote:
>>> That's true, but I don't think it answers the question.  If we have:
>>>
>>>    pci_bus->bridge == pci_bus->self->dev
>>>
>>> why would we need both "self" and "bridge"?  It would be interesting
>>> to try to remove "bridge" and replace uses of it with "self->dev".
>>
>> then how about root bus?
>>
>> root bus ->self should be NULL.
>>
>> root bus ->bridge is to the hostbridge->dev.
>
>also
>
>b->bridge = get_device(&bridge->dev);
>
>so there is one reference held by pci bus.

Hmm... so can we just have b->bridge, and then convert bridge to self if
necessary?

-- 
Richard Yang
Help you, Help me

Re: One chart on pci bridge and its bus and their children

[Ram Pai]

On Mon, Jun 04, 2012 at 10:20:51PM -0700, Yinghai Lu wrote:
> On Mon, Jun 4, 2012 at 10:18 PM, Yinghai Lu <yinghai@kernel.org> wrote:
> > On Mon, Jun 4, 2012 at 9:37 PM, Bjorn Helgaas <bhelgaas@google.com> wrote:
> >> That's true, but I don't think it answers the question.  If we have:
> >>
> >>    pci_bus->bridge == pci_bus->self->dev
> >>
> >> why would we need both "self" and "bridge"?  It would be interesting
> >> to try to remove "bridge" and replace uses of it with "self->dev".
> >
> > then how about root bus?
> >
> > root bus ->self should be NULL.
> >
> > root bus ->bridge is to the hostbridge->dev.

Yinghai,

is there a reason why this can't be fixed in
pci_create_root_bus() by having
   bus->self = bridge; /* 'bridge' is the pci_device of the corresponding host bridge */
   get_device(&bridge->dev);


RP

Re: One chart on pci bridge and its bus and their children

[Jiang Liu]

On 06/05/2012 02:35 PM, Ram Pai wrote:
> On Mon, Jun 04, 2012 at 10:20:51PM -0700, Yinghai Lu wrote:
>> On Mon, Jun 4, 2012 at 10:18 PM, Yinghai Lu <yinghai@kernel.org> wrote:
>>> On Mon, Jun 4, 2012 at 9:37 PM, Bjorn Helgaas <bhelgaas@google.com> wrote:
>>>> That's true, but I don't think it answers the question.  If we have:
>>>>
>>>>    pci_bus->bridge == pci_bus->self->dev
>>>>
>>>> why would we need both "self" and "bridge"?  It would be interesting
>>>> to try to remove "bridge" and replace uses of it with "self->dev".
>>>
>>> then how about root bus?
>>>
>>> root bus ->self should be NULL.
>>>
>>> root bus ->bridge is to the hostbridge->dev.
> 
> Yinghai,
> 
> is there a reason why this can't be fixed in
> pci_create_root_bus() by having
>    bus->self = bridge; /* 'bridge' is the pci_device of the corresponding host bridge */
Hi Ram,
	A host bridge has no associated pci_device because it's out of the
PCI domain. PCI domain only covers the hierarchy starting from the host bridge,
but doesn't include the host bridge itself.
Thanks!

>    get_device(&bridge->dev);
> 
> 
> RP
> 
> --
> To unsubscribe from this list: send the line "unsubscribe linux-pci" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: One chart on pci bridge and its bus and their children

[Ram Pai]

On Tue, Jun 05, 2012 at 10:41:39PM +0800, Jiang Liu wrote:
> On 06/05/2012 02:35 PM, Ram Pai wrote:
> > On Mon, Jun 04, 2012 at 10:20:51PM -0700, Yinghai Lu wrote:
> >> On Mon, Jun 4, 2012 at 10:18 PM, Yinghai Lu <yinghai@kernel.org> wrote:
> >>> On Mon, Jun 4, 2012 at 9:37 PM, Bjorn Helgaas <bhelgaas@google.com> wrote:
> >>>> That's true, but I don't think it answers the question.  If we have:
> >>>>
> >>>>    pci_bus->bridge == pci_bus->self->dev
> >>>>
> >>>> why would we need both "self" and "bridge"?  It would be interesting
> >>>> to try to remove "bridge" and replace uses of it with "self->dev".
> >>>
> >>> then how about root bus?
> >>>
> >>> root bus ->self should be NULL.
> >>>
> >>> root bus ->bridge is to the hostbridge->dev.
> > 
> > Yinghai,
> > 
> > is there a reason why this can't be fixed in
> > pci_create_root_bus() by having
> >    bus->self = bridge; /* 'bridge' is the pci_device of the corresponding host bridge */
> Hi Ram,
> 	A host bridge has no associated pci_device because it's out of the
> PCI domain. PCI domain only covers the hierarchy starting from the host bridge,
> but doesn't include the host bridge itself.

That makes it even easier to kill pci_bus->bridge.

for non-root buses, we can access the bridge's device using
bus->self->dev and for root bridges we can silently ignore
the bridge's dev since we know that it does not exist; because
pci_bus->self is null.

No?
RP

Re: One chart on pci bridge and its bus and their children

[Ram Pai]

On Wed, Jun 06, 2012 at 12:40:19AM +0800, Ram Pai wrote:
> On Tue, Jun 05, 2012 at 10:41:39PM +0800, Jiang Liu wrote:
> > On 06/05/2012 02:35 PM, Ram Pai wrote:
> > > On Mon, Jun 04, 2012 at 10:20:51PM -0700, Yinghai Lu wrote:
> > >> On Mon, Jun 4, 2012 at 10:18 PM, Yinghai Lu <yinghai@kernel.org> wrote:
> > >>> On Mon, Jun 4, 2012 at 9:37 PM, Bjorn Helgaas <bhelgaas@google.com> wrote:
> > >>>> That's true, but I don't think it answers the question.  If we have:
> > >>>>
> > >>>>    pci_bus->bridge == pci_bus->self->dev
> > >>>>
> > >>>> why would we need both "self" and "bridge"?  It would be interesting
> > >>>> to try to remove "bridge" and replace uses of it with "self->dev".
> > >>>
> > >>> then how about root bus?
> > >>>
> > >>> root bus ->self should be NULL.
> > >>>
> > >>> root bus ->bridge is to the hostbridge->dev.
> > > 
> > > Yinghai,
> > > 
> > > is there a reason why this can't be fixed in
> > > pci_create_root_bus() by having
> > >    bus->self = bridge; /* 'bridge' is the pci_device of the corresponding host bridge */
> > Hi Ram,
> > 	A host bridge has no associated pci_device because it's out of the
> > PCI domain. PCI domain only covers the hierarchy starting from the host bridge,
> > but doesn't include the host bridge itself.
> 
> That makes it even easier to kill pci_bus->bridge.
> 
> for non-root buses, we can access the bridge's device using
> bus->self->dev and for root bridges we can silently ignore
> the bridge's dev since we know that it does not exist; because
> pci_bus->self is null.

hmm..ok. I realize the struct device of the host bridge exists and needs
to be accessed, even though it is not a pci_device. So killing 
bus->bridge is not trivial.

RP

Re: One chart on pci bridge and its bus and their children

[Yinghai Lu]

On Mon, Jun 4, 2012 at 11:35 PM, Ram Pai <linuxram@us.ibm.com> wrote:
> On Mon, Jun 04, 2012 at 10:20:51PM -0700, Yinghai Lu wrote:
>> On Mon, Jun 4, 2012 at 10:18 PM, Yinghai Lu <yinghai@kernel.org> wrote:
>> > On Mon, Jun 4, 2012 at 9:37 PM, Bjorn Helgaas <bhelgaas@google.com> wrote:
>> >> That's true, but I don't think it answers the question.  If we have:
>> >>
>> >>    pci_bus->bridge == pci_bus->self->dev
>> >>
>> >> why would we need both "self" and "bridge"?  It would be interesting
>> >> to try to remove "bridge" and replace uses of it with "self->dev".
>> >
>> > then how about root bus?
>> >
>> > root bus ->self should be NULL.
>> >
>> > root bus ->bridge is to the hostbridge->dev.
>
> Yinghai,
>
> is there a reason why this can't be fixed in
> pci_create_root_bus() by having
>   bus->self = bridge; /* 'bridge' is the pci_device of the corresponding host bridge */
>   get_device(&bridge->dev);

self is pointer to pci_dev bridge.

and for root bus, it is NULL. that mean no pci bridge device related.

bridge is device, and have that assigned from host bridge dev, or pci
bridge dev.

also for virtfn of iov, new bus from there will not have self assigned too.
for it even bridge is not assigned.

so bridge is just generic way to save bridge (pci, host) device ...

Re: One chart on pci bridge and its bus and their children

[Richard Yang]

On Tue, Jun 05, 2012 at 12:38:52PM -0700, Yinghai Lu wrote:
>>
>> Yinghai,
>>
>> is there a reason why this can't be fixed in
>> pci_create_root_bus() by having
>>   bus->self = bridge; /* 'bridge' is the pci_device of the corresponding host bridge */
>>   get_device(&bridge->dev);
>
>self is pointer to pci_dev bridge.
>
>and for root bus, it is NULL. that mean no pci bridge device related.
>
>bridge is device, and have that assigned from host bridge dev, or pci
>bridge dev.
>
>also for virtfn of iov, new bus from there will not have self assigned too.
>for it even bridge is not assigned.
>
>so bridge is just generic way to save bridge (pci, host) device ...

I did a summarize, we have three types of pci_bus.


Can we identify the type by following method?

+--------------------+--------------------------------------------------+
| bus type           |   identification                                 |
+--------------------+--------------------------------------------------+
| root_bus           |parent is NULL, bridge point to host_bridge dev   |
+--------------------+--------------------------------------------------+
| general_bus        |parent to parent bus, bridge point to the bridge  |
+--------------------+--------------------------------------------------+
| iov_bus            |parent to parent bus, bridge point to NULL        |
+--------------------+--------------------------------------------------+

-- 
Richard Yang
Help you, Help me

Re: One chart on pci bridge and its bus and their children

[Richard Yang]

On Wed, Jun 06, 2012 at 01:28:35PM +0800, Richard Yang wrote:
>On Tue, Jun 05, 2012 at 12:38:52PM -0700, Yinghai Lu wrote:
>>>
>>> Yinghai,
>>>
>>> is there a reason why this can't be fixed in
>>> pci_create_root_bus() by having
>>>   bus->self = bridge; /* 'bridge' is the pci_device of the corresponding host bridge */
>>>   get_device(&bridge->dev);
>>
>>self is pointer to pci_dev bridge.
>>
>>and for root bus, it is NULL. that mean no pci bridge device related.
>>
>>bridge is device, and have that assigned from host bridge dev, or pci
>>bridge dev.
>>
>>also for virtfn of iov, new bus from there will not have self assigned too.
>>for it even bridge is not assigned.
>>
>>so bridge is just generic way to save bridge (pci, host) device ...
>
I did a summarize, we have three types of pci_bus.


Can we identify the type by following method?

+--------------------+--------------------------------------------------+
| bus type           |   identification                                 |
+--------------------+--------------------------------------------------+
| root_bus           |parent is NULL, bridge point to host_bridge dev   |
+--------------------+--------------------------------------------------+
| general_bus        |parent to parent bus, bridge point to the bridge  |
+--------------------+--------------------------------------------------+
| iov_bus            |parent to parent bus, bridge point to NULL        |
+--------------------+--------------------------------------------------+

Hmm... if we can determine which type a pci_bus is from this criteria. 
Sounds we don't need the self in pci_bus?

-- 
Richard Yang
Help you, Help me

Re: One chart on pci bridge and its bus and their children

[Gavin Shan]

Hi YingHai,

>>> That's true, but I don't think it answers the question.  If we have:
>>>
>>>    pci_bus->bridge == pci_bus->self->dev
>>>
>>> why would we need both "self" and "bridge"?  It would be interesting
>>> to try to remove "bridge" and replace uses of it with "self->dev".
>>
>> then how about root bus?
>>
>> root bus ->self should be NULL.
>>
>> root bus ->bridge is to the hostbridge->dev.

As I understood, the current implementation is to trace the PCI host bridge
through the PCI root bus. The problem would be avoided if you change it for
a little bit to trace the PCI host bridge through (PCI domain number), which
will simplify the current implementation in 3.5.RC1 as well even though not
much. For example: When looking for the PCI host bridge for the child bus,
we only need get the PCI domain number, then get the corresponding PCI host
bridge. We needn't go through the whole PCI bus tree from the bottom to the
top until the PCI root bus.

I'm not sure that will introduce performance improvement or not.

>
>also
>
>b->bridge = get_device(&bridge->dev);
>
>so there is one reference held by pci bus.
>

If we ignore the return value of get_device() here, that would be possible
to merge "bridge"/"self" for child PCI bus. However, we still need differentiate
the PCI root bus and child bus because the "bridge"/"self" of PCI root bus
would be "NULL", but "bridge"/"self" (might be merged to "bridge") would point
pci_dev of the P2P bridge.

Thanks,
Gavin

Re: One chart on pci bridge and its bus and their children

[Richard Yang]

On Mon, Jun 04, 2012 at 10:18:46PM -0700, Yinghai Lu wrote:
>On Mon, Jun 4, 2012 at 9:37 PM, Bjorn Helgaas <bhelgaas@google.com> wrote:
>> That's true, but I don't think it answers the question.  If we have:
>>
>>    pci_bus->bridge == pci_bus->self->dev
>>
>> why would we need both "self" and "bridge"?  It would be interesting
>> to try to remove "bridge" and replace uses of it with "self->dev".
>
>then how about root bus?
>
>root bus ->self should be NULL.
>
>root bus ->bridge is to the hostbridge->dev.
>
>Yinghai

Agree, root_pci_bus->self is NULL, and root_pci_bus->bridge point to a
solo device structure, not in hostbridge. 

BTW, if the dev is the hostbrige->dev, why not pci_host_bridge contain a
field "dev"?

And, why root bus differs with other bus a lot?
Both root bus and other bus use self/bridge point to the "PCI BRIDGE"
device.
I think we can determine whether this is a root bus by bus->parent?

-- 
Richard Yang
Help you, Help me

Re: One chart on pci bridge and its bus and their children

[Jiang Liu]

On 06/05/2012 02:02 PM, Richard Yang wrote:
> On Mon, Jun 04, 2012 at 10:18:46PM -0700, Yinghai Lu wrote:
>> On Mon, Jun 4, 2012 at 9:37 PM, Bjorn Helgaas <bhelgaas@google.com> wrote:
>>> That's true, but I don't think it answers the question.  If we have:
>>>
>>>    pci_bus->bridge == pci_bus->self->dev
>>>
>>> why would we need both "self" and "bridge"?  It would be interesting
>>> to try to remove "bridge" and replace uses of it with "self->dev".
>>
>> then how about root bus?
>>
>> root bus ->self should be NULL.
>>
>> root bus ->bridge is to the hostbridge->dev.
>>
>> Yinghai
> 
> Agree, root_pci_bus->self is NULL, and root_pci_bus->bridge point to a
> solo device structure, not in hostbridge. 
> 
> BTW, if the dev is the hostbrige->dev, why not pci_host_bridge contain a
> field "dev"?
> 
> And, why root bus differs with other bus a lot?
> Both root bus and other bus use self/bridge point to the "PCI BRIDGE"
> device.
> I think we can determine whether this is a root bus by bus->parent?
Hi Richard,

There's fundamental difference between root bridge and pci bridge.  

A PCI bridge is a PCI to PCI bridge. It's presented as a PCI device
on the primary bus and hosts the secondary bus. So the secondary bus is
associated with the PCI device on the primary bus through pci_bus->self
and pci_bus->device.

A root bridge is a system bus to PCI bus bridge, which connects PCI domain
to system bus domain. The primary side is the system bus and the secondary
side is a PCI bus, so there's a PCI bus associated with a root bridge, 
but there's no PCI device for the root bridge because it's primary side
is system bus, out of the PCI domain.

pci_bus->bridge is a common abstraction of the device associated with
a PCI bus, no matter it's a host bridge or P2P bridge. So we don't need
to distinguish between host bridge and pci bridge when we want to hold
a reference to the device associated with a PCI bus.

Thanks!
Gerry

Re: One chart on pci bridge and its bus and their children

[Jiang Liu]

On 06/05/2012 02:02 PM, Richard Yang wrote:
> On Mon, Jun 04, 2012 at 10:18:46PM -0700, Yinghai Lu wrote:
>> On Mon, Jun 4, 2012 at 9:37 PM, Bjorn Helgaas <bhelgaas@google.com> wrote:
>>> That's true, but I don't think it answers the question.  If we have:
>>>
>>>    pci_bus->bridge == pci_bus->self->dev
>>>
>>> why would we need both "self" and "bridge"?  It would be interesting
>>> to try to remove "bridge" and replace uses of it with "self->dev".
>>
>> then how about root bus?
>>
>> root bus ->self should be NULL.
>>
>> root bus ->bridge is to the hostbridge->dev.
>>
>> Yinghai
> 
> Agree, root_pci_bus->self is NULL, and root_pci_bus->bridge point to a
> solo device structure, not in hostbridge. 
> 
> BTW, if the dev is the hostbrige->dev, why not pci_host_bridge contain a
> field "dev"?
With latest code from Yinghai, the host bridge is defined as:
struct pci_host_bridge {        
        struct device dev;
        struct pci_bus *bus;            /* root bus */
        struct list_head windows;       /* pci_host_bridge_windows */
        void (*release_fn)(struct pci_host_bridge *);
        void *release_data;
};

So I think it's already working in the way suggested by you.

> 
> And, why root bus differs with other bus a lot?
> Both root bus and other bus use self/bridge point to the "PCI BRIDGE"
> device.
> I think we can determine whether this is a root bus by bus->parent?
> 

Re: One chart on pci bridge and its bus and their children

[Yinghai Lu]

On Mon, Jun 4, 2012 at 11:02 PM, Richard Yang
<weiyang@linux.vnet.ibm.com> wrote:
> On Mon, Jun 04, 2012 at 10:18:46PM -0700, Yinghai Lu wrote:
>
> BTW, if the dev is the hostbrige->dev, why not pci_host_bridge contain a
> field "dev"?

we have that from 3.5-rc1

>
> And, why root bus differs with other bus a lot?
> Both root bus and other bus use self/bridge point to the "PCI BRIDGE"
> device.
> I think we can determine whether this is a root bus by bus->parent?

self is for real pci dev bridge.
bridge is the for pci_dev bridge generic device pointer. or host
bridge device pointer.

Yinghai

Re: One chart on pci bridge and its bus and their children

[Richard Yang]

On Tue, Jun 05, 2012 at 12:11:46PM +0800, Ram Pai wrote:
>On Tue, Jun 05, 2012 at 11:26:27AM +0800, Richard Yang wrote:
>> 
>> All, 
>> 
>> I draw a chart on the relationship of bridge/bus/devics.
>> 
>> I am not sure it is exact, so please correct me if it is not correct.
>
>A clearer representation can be found at http://tldp.org/LDP/tlk/dd/pci.html		
>in section 6.6.1
Thanks, Ram.

I found I missed one important field in pci_bus.

>
>> 
>> BTW, the pci_bus->self, pci_bus->bridge seem to point the same thing.
>> So this is really necessary to do this?
>
>pci_bus->self points to the pci_dev structure, and pci_bus->bridge
>points to the device structure. Both correspond to the same bridge. BTW: device
>structure is the basic structure to represent any device in the system,
>not necessarily pci device.
>
>
>> 
>>                             Bus 0(pci_bus)                      
>>                             +--------------------+<----+
>>                             |                    |     |
>>                             +--------------------+     |
>>                                                        |
>>                                                        |
>>                              Bridge(pci_dev)           |                     
>>                             +--------------------+<----|---+--------------+
>>                             |subordinate(pci_bus)|-+   |   |              |
>>                             +--------------------+ |   |   |              |
>>                                                    |   |   |              |
>>                                                    |   |   |              |
>>                                                    |   |   |              |
>>                             Bus 1(pci_bus)         |   |   |              |
>>                        +--->+-------------------+<-+   |   |              |
>>                        |    |parent(pci_bus)    |------+   |              |
>>                        |    +-------------------+          |              |
>>                        |    |self(pci_dev)      |----------+              |
>>                        |    |bridge(device)     |                         |
>>                        |    +-------------------+                         |
>>                        |    |devices(list_head) |                         |
>>                        |    +--------+----------+                         |
>>                        |             |                                    |
>>                        |      ----+--+---+----                            |
>>    pci_dev             |          |      |          pci_dev               |
>>    +---------------+<--|----------+      +--------->+-----------------+   |
>>    |bus(pci_bus)   |---+                            |bus(pci_bus)     |   |
>>    +---------------+                                +-----------------+   |
>>    |dev.parent     |                                |dev.parent       |---+
>>    +---------------+                                +-----------------+

-- 
Richard Yang
Help you, Help me