svn commit: r349133 - head/share/man/man4

[Niclas Zeising]

On 2019-06-17 11:03, Rodney W. Grimes wrote:
>> On 2019-06-17 09:56, Benjamin Kaduk wrote:
>>> On Sun, Jun 16, 2019 at 10:42 PM Niclas Zeising <zeising at freebsd.org
>>> <mailto:zeising at freebsd.org>> wrote:
>>>
>>>      Author: zeising (doc,ports committer)
>>>      Date: Mon Jun 17 05:41:47 2019
>>>      New Revision: 349133
>>>      URL: https://svnweb.freebsd.org/changeset/base/349133
>>>
>>>      Log:
>>>       ? pci(4): Document PCIOCATTACHED
>>>
>>>       ? Document the PCIOCATTACHED ioctl(2) in the pci(4) manual.
>>>       ? PCIOCATTACHED is used to query if a driver has attached to a PCI.
>>>
>>>       ? Reviewed by:? bcr, imp
>>>       ? MFC after:? ? 2 weeks
>>>       ? Differential Revision: https://reviews.freebsd.org/D20652
>>>
>>>      Modified:
>>>       ? head/share/man/man4/pci.4
>>>
>>>      Modified: head/share/man/man4/pci.4
>>>      ==============================================================================
>>>      --- head/share/man/man4/pci.4? ?Mon Jun 17 03:48:44 2019
>>>      (r349132)
>>>      +++ head/share/man/man4/pci.4? ?Mon Jun 17 05:41:47 2019
>>>      (r349133)
>>>      @@ -24,7 +24,7 @@
>>>       ?.\"
>>>       ?.\" $FreeBSD$
>>>       ?.\"
>>>      -.Dd June 14, 2018
>>>      +.Dd June 17, 2019
>>>       ?.Dt PCI 4
>>>       ?.Os
>>>       ?.Sh NAME
>>>      @@ -333,6 +333,26 @@ The limitations on data width described for
>>>       ?reading registers, above, also apply to writing
>>>       ?.Tn PCI
>>>       ?configuration registers.
>>>      +.It PCIOCATTACHED
>>>      +This
>>>      +.Xr ioctl 2
>>>      +allows users to query if a driver is attached to the
>>>      +.Tn PCI
>>>      +specified in the passed-in
>>>
>>>
>>> Is there a missing word like "device" here?
> 
> Actally I think the missing word, in both cases, is register,
> unless I am misreading some part of the manual page and
> what a struct pci_io points at.  I guess if the pi_reg is null
> then this would be device.  Either way there is defanity a
> missing word.

I'll try to fix this.  In the PCIOCWRITE case, perhaps register is best, 
however, inthe PCIOCATTACHED case, device is best, I think.
I'll create a patch and put it for review, I'll get back to you once 
it's done.
Regards

> 
>>
>> I just followed the same syntax as for the PCIOCWRITE ioctl in the
>> paragraph before.
> 
> Probably a mistake there too.
> 
>> I found it a little strange when I read it, but I
>> kind of assumed there was a reason it said PCI and not PCI device,
>> perhaps similar to using RAM instead of RAM memory, if you understand
>> what I mean.
> 
> RAM memory == Random Access Memory Memory
> PCI == Peripheral Component Interconnect
> PCI {Register, Device, Bus} all make valid sense,
> PCI alone does not.
> 
>> Regards
>> Niclas
>>
>>>
>>>      +.Va pci_io
>>>      +structure.
>>>      +The
>>>      +.Va pci_io
>>>      +structure is described above, however, the
>>>      +.Va pi_reg
>>>      +and
>>>      +.Va pi_width
>>>      +fields are not used.
>>>      +The status of the device is stored in the
>>>      +.Va pi_data
>>>      +field.
>>>      +A value of 0 indicates no driver is attached, while a value larger
>>>      than 0
>>>      +indicates that a driver is attached.
>>>       ?.It PCIOCBARMMAP
>>>       ?This
>>>       ?.Xr ioctl 2
>>
>>
>> -- 
>> Niclas Zeising
>>
>>
>>
> 


-- 
Niclas Zeising