[PATCH] man3/getaddrinfo.3: Document results ordering and /etc/gai.conf

[PATCH] man3/getaddrinfo.3: Document results ordering and /etc/gai.conf

[Petr Baudis]

This patch documents the order of the getaddrinfo(3) results (RFC 3484), how
should the application deal with that, mentions the extremely common cause of
having multiple results per query (both IPv4 and IPv6 addresses available)
and mentions /etc/gai.conf.

Signed-off-by: Petr Baudis <pasky-AlSwsSmVLrQ@public.gmane.org>

diff --git a/man3/getaddrinfo.3 b/man3/getaddrinfo.3
index b3b0cce..e1f1f9c 100644
--- a/man3/getaddrinfo.3
+++ b/man3/getaddrinfo.3
@@ -33,10 +33,12 @@
 .\" 2008-02-26, mtk; clarify discussion of NULL 'hints' argument; other
 .\"     minor rewrites.
 .\" 2008-06-18, mtk: many parts rewritten
+.\" 2008-12-04, Petr Baudis <pasky-AlSwsSmVLrQ@public.gmane.org>
+.\"	Describe results ordering and reference /etc/gai.conf.
 .\" FIXME . glibc's 2.9 NEWS file documents DCCP and UDP-lite support
 .\"           and is SCTP support now also there?
 .\"
-.TH GETADDRINFO 3 2008-10-28 "GNU" "Linux Programmer's Manual"
+.TH GETADDRINFO 3 2008-12-04 "GNU" "Linux Programmer's Manual"
 .SH NAME
 getaddrinfo, freeaddrinfo, gai_strerror \- network address and
 service translation
@@ -293,15 +295,26 @@ and returns a pointer to the start of the list in
 The items in the linked list are linked by the
 .I ai_next
 field.
+.PP
 There are several reasons why
 the linked list may have more than one
 .I addrinfo
-structure, including: the network host is multi-homed; or the same service
-is available from multiple socket protocols (one
+structure, including: the network host is multi-homed, accessible
+over multiple protocols (e.g. both
+.BR AF_INET
+and
+.BR AF_INET6 );
+or the same service is available from multiple socket types (one
 .B SOCK_STREAM
 address and another
 .B SOCK_DGRAM
-address, for example).
+address, for example).  Normally, the application should try
+using the addresses in the order they are returned in.
+The sorting function used within
+.BR getaddrinfo ()
+is defined in RFC\ 3484; the order can be tweaked for a particular
+system by editing
+.IR /etc/gai.conf .
 .PP
 If
 .I hints.ai_flags
@@ -558,6 +571,8 @@ The
 .BR gai_strerror ()
 function translates these error codes to a human readable string,
 suitable for error reporting.
+.SH "FILES"
+.I /etc/gai.conf
 .SH "CONFORMING TO"
 POSIX.1-2001.
 The
--
To unsubscribe from this list: send the line "unsubscribe linux-man" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH] man3/getaddrinfo.3: Document results ordering and /etc/gai.conf

[Michael Kerrisk]

Hi Petr,

On Thu, Dec 4, 2008 at 6:22 PM, Petr Baudis <pasky-AlSwsSmVLrQ@public.gmane.org> wrote:
> This patch documents the order of the getaddrinfo(3) results (RFC 3484), how
> should the application deal with that, mentions the extremely common cause of
> having multiple results per query (both IPv4 and IPv6 addresses available)
> and mentions /etc/gai.conf.

Thanks for this!  I've applied this for 3.15, but with tweaks as noted below.

> Signed-off-by: Petr Baudis <pasky-AlSwsSmVLrQ@public.gmane.org>
>
> diff --git a/man3/getaddrinfo.3 b/man3/getaddrinfo.3
> index b3b0cce..e1f1f9c 100644
> --- a/man3/getaddrinfo.3
> +++ b/man3/getaddrinfo.3
> @@ -33,10 +33,12 @@
>  .\" 2008-02-26, mtk; clarify discussion of NULL 'hints' argument; other
>  .\"     minor rewrites.
>  .\" 2008-06-18, mtk: many parts rewritten
> +.\" 2008-12-04, Petr Baudis <pasky-AlSwsSmVLrQ@public.gmane.org>
> +.\"    Describe results ordering and reference /etc/gai.conf.

Yep.  This is a case where in-source changelog makes sense.

>  .\" FIXME . glibc's 2.9 NEWS file documents DCCP and UDP-lite support
>  .\"           and is SCTP support now also there?
>  .\"
> -.TH GETADDRINFO 3 2008-10-28 "GNU" "Linux Programmer's Manual"
> +.TH GETADDRINFO 3 2008-12-04 "GNU" "Linux Programmer's Manual"
>  .SH NAME
>  getaddrinfo, freeaddrinfo, gai_strerror \- network address and
>  service translation
> @@ -293,15 +295,26 @@ and returns a pointer to the start of the list in
>  The items in the linked list are linked by the
>  .I ai_next
>  field.
> +.PP

This change is not relevant to the subject of the patch, and unneeded.  Dropped.

>  There are several reasons why
>  the linked list may have more than one
>  .I addrinfo
> -structure, including: the network host is multi-homed; or the same service
> -is available from multiple socket protocols (one
> +structure, including: the network host is multi-homed, accessible
> +over multiple protocols (e.g. both
> +.BR AF_INET
> +and
> +.BR AF_INET6 );
> +or the same service is available from multiple socket types (one
>  .B SOCK_STREAM
>  address and another
>  .B SOCK_DGRAM
> -address, for example).
> +address, for example).  Normally, the application should try

My preference (see man-pages(7) is to start new sentences on new line.
 This is because (later) diffs at least sometimes operate at the
sentence level.

> +using the addresses in the order they are returned in.
> +The sorting function used within
> +.BR getaddrinfo ()
> +is defined in RFC\ 3484; the order can be tweaked for a particular
> +system by editing
> +.IR /etc/gai.conf .

Here, it is good to mention the glibc version that first started
supporting gai.conf.  It looks like it was 2.5, and I've added that
information.

>  .PP
>  If
>  .I hints.ai_flags
> @@ -558,6 +571,8 @@ The
>  .BR gai_strerror ()
>  function translates these error codes to a human readable string,
>  suitable for error reporting.
> +.SH "FILES"
> +.I /etc/gai.conf
>  .SH "CONFORMING TO"
>  POSIX.1-2001.
>  The

Thanks!

Michael


-- 
Michael Kerrisk
Linux man-pages maintainer; http://www.kernel.org/doc/man-pages/
git://git.kernel.org/pub/scm/docs/man-pages/man-pages.git
man-pages online: http://www.kernel.org/doc/man-pages/online_pages.html
Found a bug? http://www.kernel.org/doc/man-pages/reporting_bugs.html
--
To unsubscribe from this list: send the line "unsubscribe linux-man" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH] man3/getaddrinfo.3: Document results ordering and /etc/gai.conf

[Petr Baudis]

  Hi,

On Fri, Dec 05, 2008 at 12:09:44PM -0500, Michael Kerrisk wrote:
> Hi Petr,
> 
> On Thu, Dec 4, 2008 at 6:22 PM, Petr Baudis <pasky-AlSwsSmVLrQ@public.gmane.org> wrote:
> > This patch documents the order of the getaddrinfo(3) results (RFC 3484), how
> > should the application deal with that, mentions the extremely common cause of
> > having multiple results per query (both IPv4 and IPv6 addresses available)
> > and mentions /etc/gai.conf.
> 
> Thanks for this!  I've applied this for 3.15, but with tweaks as noted below.

  thanks! (Purely out of curiosity, how often do you push out your git
tree? Only on releases?)

> > @@ -293,15 +295,26 @@ and returns a pointer to the start of the list in
> >  The items in the linked list are linked by the
> >  .I ai_next
> >  field.
> > +.PP
> 
> This change is not relevant to the subject of the patch, and unneeded.  Dropped.

  I simply think that with the addition, the whole paragraph is too big
and hard to read. But this is of course matter of personal preference.

> >  There are several reasons why
> >  the linked list may have more than one
> >  .I addrinfo
> > -structure, including: the network host is multi-homed; or the same service
> > -is available from multiple socket protocols (one
> > +structure, including: the network host is multi-homed, accessible
> > +over multiple protocols (e.g. both
> > +.BR AF_INET
> > +and
> > +.BR AF_INET6 );
> > +or the same service is available from multiple socket types (one
> >  .B SOCK_STREAM
> >  address and another
> >  .B SOCK_DGRAM
> > -address, for example).
> > +address, for example).  Normally, the application should try
> 
> My preference (see man-pages(7) is to start new sentences on new line.
>  This is because (later) diffs at least sometimes operate at the
> sentence level.

  Thanks, I did not actually read man-pages(7) yet, I will do that.

> > +using the addresses in the order they are returned in.
> > +The sorting function used within
> > +.BR getaddrinfo ()
> > +is defined in RFC\ 3484; the order can be tweaked for a particular
> > +system by editing
> > +.IR /etc/gai.conf .
> 
> Here, it is good to mention the glibc version that first started
> supporting gai.conf.  It looks like it was 2.5, and I've added that
> information.

  Yes, glibc-2.5 introduced this.

-- 
				Petr "Pasky" Baudis
People who take cold baths never have rheumatism, but they have
cold baths.
--
To unsubscribe from this list: send the line "unsubscribe linux-man" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH] man3/getaddrinfo.3: Document results ordering and /etc/gai.conf

[Michael Kerrisk]

Hi Petr,

On Fri, Dec 5, 2008 at 12:15 PM, Petr Baudis <pasky-AlSwsSmVLrQ@public.gmane.org> wrote:
>  Hi,
>
> On Fri, Dec 05, 2008 at 12:09:44PM -0500, Michael Kerrisk wrote:
>> Hi Petr,
>>
>> On Thu, Dec 4, 2008 at 6:22 PM, Petr Baudis <pasky-AlSwsSmVLrQ@public.gmane.org> wrote:
>> > This patch documents the order of the getaddrinfo(3) results (RFC 3484), how
>> > should the application deal with that, mentions the extremely common cause of
>> > having multiple results per query (both IPv4 and IPv6 addresses available)
>> > and mentions /etc/gai.conf.
>>
>> Thanks for this!  I've applied this for 3.15, but with tweaks as noted below.
>
>  thanks! (Purely out of curiosity, how often do you push out your git
> tree? Only on releases?)

It varies.  (Maybe I don't push it out quite as often as I should.
I'm open to input on that point.)  It does get pushed out at least on
every release, and I try to do it mor often if there's a lot of stuff
since the last push.

>> > @@ -293,15 +295,26 @@ and returns a pointer to the start of the list in
>> >  The items in the linked list are linked by the
>> >  .I ai_next
>> >  field.
>> > +.PP
>>
>> This change is not relevant to the subject of the patch, and unneeded.  Dropped.
>  I simply think that with the addition, the whole paragraph is too big
> and hard to read. But this is of course matter of personal preference.

D'oh!  I agree.  Changed now.

>> >  There are several reasons why
>> >  the linked list may have more than one
>> >  .I addrinfo
>> > -structure, including: the network host is multi-homed; or the same service
>> > -is available from multiple socket protocols (one
>> > +structure, including: the network host is multi-homed, accessible
>> > +over multiple protocols (e.g. both
>> > +.BR AF_INET
>> > +and
>> > +.BR AF_INET6 );
>> > +or the same service is available from multiple socket types (one
>> >  .B SOCK_STREAM
>> >  address and another
>> >  .B SOCK_DGRAM
>> > -address, for example).
>> > +address, for example).  Normally, the application should try
>>
>> My preference (see man-pages(7) is to start new sentences on new line.
>>  This is because (later) diffs at least sometimes operate at the
>> sentence level.
>
>  Thanks, I did not actually read man-pages(7) yet, I will do that.

Thanks.

>> > +using the addresses in the order they are returned in.
>> > +The sorting function used within
>> > +.BR getaddrinfo ()
>> > +is defined in RFC\ 3484; the order can be tweaked for a particular
>> > +system by editing
>> > +.IR /etc/gai.conf .
>>
>> Here, it is good to mention the glibc version that first started
>> supporting gai.conf.  It looks like it was 2.5, and I've added that
>> information.
>
>  Yes, glibc-2.5 introduced this.

Thanks for the ACK.

Cheers,

Michael

-- 
Michael Kerrisk
Linux man-pages maintainer; http://www.kernel.org/doc/man-pages/
git://git.kernel.org/pub/scm/docs/man-pages/man-pages.git
man-pages online: http://www.kernel.org/doc/man-pages/online_pages.html
Found a bug? http://www.kernel.org/doc/man-pages/reporting_bugs.html
--
To unsubscribe from this list: send the line "unsubscribe linux-man" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html