GNU bug report logs - #9594
doc: expand 27.3 Numeric modes (File permissions)

Previous Next

To add a comment to this bug, you must first unarchive it, by sending
a message to control AT debbugs.gnu.org, with unarchive 9594 in the body.
You can then email your comments to 9594 AT debbugs.gnu.org in the normal way.

Toggle the display of automated, internal messages from the tracker.

Message #5 received at submit <at> debbugs.gnu.org (full text, mbox):

From: Filipus Klutiero <chealer <at> gmail.com>
To: bug-coreutils <at> gnu.org
Subject: 27.3 Numeric modes (File permissions): Special mode bits assume file
	is regular
Date: Sat, 24 Sep 2011 17:33:19 -0400

Section 27.3 Numeric modes contains:

> Here is how the bits are arranged, starting with
> the lowest valued bit:
>
>      Value in  Corresponding
>      Mode      Mode Bit
>
>                Other users not in the file's group:
>         1      Execute/search
>         2      Write
>         4      Read
>
>                Other users in the file's group:
>        10      Execute/search
>        20      Write
>        40      Read
>
>                The file's owner:
>       100      Execute/search
>       200      Write
>       400      Read
>
>                Special mode bits:
>      1000      Restricted deletion flag or sticky bit
>      2000      Set group ID on execution
>      4000      Set user ID on execution

These correspondences assume that the file is a regular file. For 
example, for a directory (where execution doesn't make sense), 2000 
causes files created in the directory to inherit the directory's group. 
http://en.wikipedia.org/wiki/Setgid#setuid_and_setgid_on_directories has 
more information.

Message #8 received at 9594 <at> debbugs.gnu.org (full text, mbox):

From: Paul Eggert <eggert <at> cs.ucla.edu>
To: Filipus Klutiero <chealer <at> gmail.com>
Cc: 9594 <at> debbugs.gnu.org
Subject: Re: bug#9594: 27.3 Numeric modes (File permissions): Special mode
	bits	assume file is regular
Date: Sat, 24 Sep 2011 21:11:20 -0700

On 09/24/11 14:33, Filipus Klutiero wrote:

> These correspondences assume that the file is a regular file.

Not really: for example, the phrase "execute/search"
refers to execution (for regular files) and search
(for directories), and "Restricted deletion flag or sticky bit"
is talking about directories (for deletion) and regular
files (for sticky bit).

The phrase "Set group ID on execution" is not wrong, either,
as it is saying that if you execute the file, then you set
the group ID.  This phrase is used with the same meaning in
the table in

http://pubs.opengroup.org/onlinepubs/9699919799/basedefs/sys_stat.h.html

In such a small table, one can't explain *everything*.  Since
the very next section talks about setuid and setgid on directories,
the topic seems to be fairly well covered already.  However, if
you have a specific suggestion for improving the wording while
staying within the tight bounds established by the table, then
that would help.

Message #11 received at 9594 <at> debbugs.gnu.org (full text, mbox):

From: Filipus Klutiero <chealer <at> gmail.com>
To: 9594 <at> debbugs.gnu.org
Subject: Re: bug#9594: 27.3 Numeric modes (File permissions): Special mode
	bits	assume file is regular
Date: Sun, 25 Sep 2011 09:57:15 -0400

Le 2011-09-25 00:11, Paul Eggert a écrit :
> On 09/24/11 14:33, Filipus Klutiero wrote:
>
>> These correspondences assume that the file is a regular file.
> Not really: for example, the phrase "execute/search"
> refers to execution (for regular files) and search
> (for directories), and "Restricted deletion flag or sticky bit"
> is talking about directories (for deletion) and regular
> files (for sticky bit).
Sorry if the body didn't specify, but I was only referring to special 
mode bits. And indeed, one of them is OK.
> The phrase "Set group ID on execution" is not wrong, either,
> as it is saying that if you execute the file, then you set
> the group ID.  This phrase is used with the same meaning in
> the table in
>
> http://pubs.opengroup.org/onlinepubs/9699919799/basedefs/sys_stat.h.html
It may not be wrong, but it's misleading. On a directory, S_ISGID does 
not mean to set the group ID on execution. The meaning is OS-dependent.
> In such a small table, one can't explain *everything*.  Since
> the very next section talks about setuid and setgid on directories,
> the topic seems to be fairly well covered already.
Note that the manual is read by section, not sequentially. For example, 
I was reading 13.3 chmod and followed the link to 27.3.

It may be better to have 2 tables, or a table of common meanings and 2 
tables of specific meanings. I believe I would go for 2 tables.
>    However, if
> you have a specific suggestion for improving the wording while
> staying within the tight bounds established by the table, then
> that would help.
If kept in a single table, the wording should clarify that it only 
applies to regular files. It would be good to mention that the meaning 
on directories is OS-dependent, or to link to 27.4.

Message #14 received at 9594 <at> debbugs.gnu.org (full text, mbox):

From: Assaf Gordon <assafgordon <at> gmail.com>
To: Filipus Klutiero <chealer <at> gmail.com>, 9594 <at> debbugs.gnu.org
Subject: Re: bug#9594: 27.3 Numeric modes (File permissions): Special mode
 bits assume file is regular
Date: Mon, 15 Oct 2018 08:07:53 -0600

close 9594
retitle 9594 doc: expand 27.3 Numeric modes (File permissions)
stop

(triaging old bugs)

Hello,

On 25/09/11 07:57 AM, Filipus Klutiero wrote:
> Le 2011-09-25 00:11, Paul Eggert a écrit :
>> On 09/24/11 14:33, Filipus Klutiero wrote:
>>
>>> These correspondences assume that the file is a regular file.
>> Not really: for example, the phrase "execute/search"
>> refers to execution (for regular files) and search
>> (for directories), and "Restricted deletion flag or sticky bit"
>> is talking about directories (for deletion) and regular
>> files (for sticky bit).
> Sorry if the body didn't specify, but I was only referring to special 
> mode bits. And indeed, one of them is OK.

With no further correspondence in 7 years, I'm closing this bug.
Discussion can continue by replying to this thread.

regards,
 - assaf

Message #21 received at 9594 <at> debbugs.gnu.org (full text, mbox):

From: Philippe Cloutier <chealer <at> gmail.com>
To: Assaf Gordon <assafgordon <at> gmail.com>, 9594 <at> debbugs.gnu.org
Subject: Re: bug#9594: 27.3 Numeric modes (File permissions): Special mode
 bits assume file is regular
Date: Sat, 20 Oct 2018 11:30:51 -0400

Hi Assaf,

On 18-10-15 10 h 07, Assaf Gordon wrote:
> close 9594
> retitle 9594 doc: expand 27.3 Numeric modes (File permissions)
> stop
>
> (triaging old bugs)
>
> Hello,
>
> On 25/09/11 07:57 AM, Filipus Klutiero wrote:
>> Le 2011-09-25 00:11, Paul Eggert a écrit :
>>> On 09/24/11 14:33, Filipus Klutiero wrote:
>>>
>>>> These correspondences assume that the file is a regular file.
>>> Not really: for example, the phrase "execute/search"
>>> refers to execution (for regular files) and search
>>> (for directories), and "Restricted deletion flag or sticky bit"
>>> is talking about directories (for deletion) and regular
>>> files (for sticky bit).
>> Sorry if the body didn't specify, but I was only referring to special mode bits. And indeed, one of them is OK.
>
> With no further correspondence in 7 years, I'm closing this bug.


Please clarify why this was closed. I still see the bug in Debian 9's coreutils 8.26-3. Was this fixed since?


> Discussion can continue by replying to this thread.
>
> regards,
>  - assaf


-- 
Philippe Cloutier
http://www.philippecloutier.com

Message #24 received at 9594 <at> debbugs.gnu.org (full text, mbox):

From: Paul Eggert <eggert <at> cs.ucla.edu>
To: Philippe Cloutier <chealer <at> gmail.com>,
 Assaf Gordon <assafgordon <at> gmail.com>, 9594 <at> debbugs.gnu.org
Subject: Re: bug#9594: 27.3 Numeric modes (File permissions): Special mode
 bits assume file is regular
Date: Sat, 20 Oct 2018 11:04:24 -0700

[Message part 1 (text/plain, inline)]

Philippe Cloutier wrote:

> Please clarify why this was closed.

It was closed primarily because no specific wording fix was proposed.

I looked at the bug report again, and don't agree that the old wording was wrong 
or even misleading. It's merely a summary table, there's a lot of explanatory 
text before the table, and we shouldn't clutter the table with another copy of 
the explanation. That being said, the explanatory text can be improved and the 
summary table can be abbreviated even further by removing the phrase "on 
execution" which seems to be the primary point of the original bug report. I 
installed the attached.

[0001-doc-tidy-up-setuid-commentary.txt (text/plain, attachment)]

Previous Next

GNU bug tracking system
Copyright (C) 1999 Darren O. Benham, 1997,2003 nCipher Corporation Ltd, 1994-97 Ian Jackson.