From mboxrd@z Thu Jan 1 00:00:00 1970 Date: Sun, 19 Oct 2014 14:30:13 +0200 From: Carsten Kunze To: 9fans@9fans.net Message-ID: <60682619.771899.1413721813507.JavaMail.ngmail@webmail24.arcor-online.net> MIME-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Subject: [9fans] libc struct Dir field type Topicbox-Message-UUID: 1e6a6a94-ead9-11e9-9d60-3106f5b1d025 Hello, where is the type field of libc struct Dir documented? In source I found a compare of field type with the char 'M'. Where can I loop up what 'M' means? Carsten From mboxrd@z Thu Jan 1 00:00:00 1970 MIME-Version: 1.0 In-Reply-To: <60682619.771899.1413721813507.JavaMail.ngmail@webmail24.arcor-online.net> References: <60682619.771899.1413721813507.JavaMail.ngmail@webmail24.arcor-online.net> Date: Sun, 19 Oct 2014 16:32:48 +0100 Message-ID: From: Charles Forsyth To: Fans of the OS Plan 9 from Bell Labs <9fans@9fans.net> Content-Type: multipart/alternative; boundary=f46d04426c5278e7130505c84f85 Subject: Re: [9fans] libc struct Dir field type Topicbox-Message-UUID: 1e6fc9f8-ead9-11e9-9d60-3106f5b1d025 --f46d04426c5278e7130505c84f85 Content-Type: text/plain; charset=UTF-8 On 19 October 2014 13:30, Carsten Kunze wrote: > where is the type field of libc struct Dir documented? In source I found > a compare of field type with the char 'M'. Where can I loop up what 'M' > means? > It's the Unicode character used to identify the device driver within a kernel, and are part of the device driver's Dev value. Only a few have any sort of global significance, and even that's probably a mistake. cat /dev/drivers for the map currently in use. /sys/src/9/port/master tries to keep track of them, but as you'll see some characters are used for several devices (not in use simultaneously in the same kernel) --f46d04426c5278e7130505c84f85 Content-Type: text/html; charset=UTF-8 Content-Transfer-Encoding: quoted-printable

= On 19 October 2014 13:30, Carsten Kunze <carsten.kunze@arcor.de&g= t; wrote:
where is the type field of libc struct D= ir documented?=C2=A0 In source I found a compare of field type with the cha= r 'M'.=C2=A0 Where can I loop up what 'M' means?


It's th= e Unicode character used to identify the device driver within a kernel,
and are part of the device driver's Dev va= lue. Only a few have any sort of global
sig= nificance, and even that's probably a mistake.

cat /dev/drivers
for the map currently in use. /sys/src/9/port/master tri= es to keep track of them,
but as you'll= see some characters are used for several devices (not in use simultaneousl= y in the same kernel)
--f46d04426c5278e7130505c84f85-- From mboxrd@z Thu Jan 1 00:00:00 1970 MIME-Version: 1.0 In-Reply-To: References: <60682619.771899.1413721813507.JavaMail.ngmail@webmail24.arcor-online.net> Date: Sun, 19 Oct 2014 16:36:19 +0100 Message-ID: From: Charles Forsyth To: Fans of the OS Plan 9 from Bell Labs <9fans@9fans.net> Content-Type: multipart/alternative; boundary=001a11c1b31212a4720505c85c2e Subject: Re: [9fans] libc struct Dir field type Topicbox-Message-UUID: 1e75bdea-ead9-11e9-9d60-3106f5b1d025 --001a11c1b31212a4720505c85c2e Content-Type: text/plain; charset=UTF-8 On 19 October 2014 16:32, Charles Forsyth wrote: > Only a few have any sort of global > significance, and even that's probably a mistake. > Ignore that: that statement's wrong since they appear in bind operations in bind(1), bind(2), /lib/namespace and similar contexts. I also forgot to add that ls -l shows you the type character: ls -l /dev shows you where different names are produced. --001a11c1b31212a4720505c85c2e Content-Type: text/html; charset=UTF-8 Content-Transfer-Encoding: quoted-printable

= On 19 October 2014 16:32, Charles Forsyth <charles.forsyth@gmail.c= om> wrote:
Only a few have any sort of global
significance, and even that's probably a mistake.

Ignore that: that statement&#= 39;s wrong since they appear in bind operations in bind(1), bind(2), /lib/n= amespace and similar contexts.

I also forgot to add that ls -l shows you the type= character: ls -l /dev shows you where different names are produced.
<= div class=3D"gmail_extra">
--001a11c1b31212a4720505c85c2e-- From mboxrd@z Thu Jan 1 00:00:00 1970 MIME-Version: 1.0 In-Reply-To: References: <60682619.771899.1413721813507.JavaMail.ngmail@webmail24.arcor-online.net> Date: Sun, 19 Oct 2014 16:38:54 +0100 Message-ID: From: Charles Forsyth To: Fans of the OS Plan 9 from Bell Labs <9fans@9fans.net> Content-Type: multipart/alternative; boundary=f46d04426c524ac4080505c8653f Subject: Re: [9fans] libc struct Dir field type Topicbox-Message-UUID: 1e879ef2-ead9-11e9-9d60-3106f5b1d025 --f46d04426c524ac4080505c8653f Content-Type: text/plain; charset=UTF-8 On 19 October 2014 16:36, Charles Forsyth wrote: > Only a few have any sort of global > significance, and even that's probably a mistake. > What I really meant was that few programs care about Dir.type since most don't care what serves the names as long as they behave as expected. Relatively few programs use the # names directly in bind(2) calls: it's more common to set up a name space and then invoke a program. --f46d04426c524ac4080505c8653f Content-Type: text/html; charset=UTF-8 Content-Transfer-Encoding: quoted-printable

= On 19 October 2014 16:36, Charles Forsyth <charles.forsyth@gmail.c= om> wrote:
Only a few have any sort of global
significance, and even that's probably a mistake.
=

What I really meant was that few programs care about Dir.type sin= ce most don't care what serves the names
as long as they behave as expected. Relatively few programs use the # nam= es directly in bind(2) calls:
it's more= common to set up a name space and then invoke a program.

--f46d04426c524ac4080505c8653f-- From mboxrd@z Thu Jan 1 00:00:00 1970 Date: Sun, 19 Oct 2014 18:01:36 +0200 From: Carsten Kunze To: 9fans@9fans.net Message-ID: <902251034.773510.1413734496857.JavaMail.ngmail@webmail24.arcor-online.net> In-Reply-To: References: <60682619.771899.1413721813507.JavaMail.ngmail@webmail24.arcor-online.net> MIME-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Subject: Re: [9fans] libc struct Dir field type Topicbox-Message-UUID: 1e8d059a-ead9-11e9-9d60-3106f5b1d025 > On 19 October 2014 16:36, Charles Forsyth > wrote: > > What I really meant was that few programs care about Dir.type since most > don't care what serves the names > as long as they behave as expected. Relatively few programs use the # names > directly in bind(2) calls: > it's more common to set up a name space and then invoke a program. Thank you for your in-depth information! Carsten