From patchwork Wed Jul 9 16:14:18 2014 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Florian Weimer X-Patchwork-Id: 1983 Received: (qmail 31033 invoked by alias); 9 Jul 2014 16:14:43 -0000 Mailing-List: contact libc-alpha-help@sourceware.org; run by ezmlm Precedence: bulk List-Id: List-Unsubscribe: List-Subscribe: List-Archive: List-Post: List-Help: , Sender: libc-alpha-owner@sourceware.org Delivered-To: mailing list libc-alpha@sourceware.org Received: (qmail 30986 invoked by uid 89); 9 Jul 2014 16:14:38 -0000 Authentication-Results: sourceware.org; auth=none X-Virus-Found: No X-Spam-SWARE-Status: No, score=-3.0 required=5.0 tests=AWL, BAYES_05, RP_MATCHES_RCVD, SPF_HELO_PASS, SPF_PASS autolearn=ham version=3.3.2 X-HELO: mx1.redhat.com Message-ID: <53BD6A5A.2020001@redhat.com> Date: Wed, 09 Jul 2014 18:14:18 +0200 From: Florian Weimer User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:24.0) Gecko/20100101 Thunderbird/24.6.0 MIME-Version: 1.0 To: GNU C Library Subject: [PATCH] locale directory traversal (CVE-2014-0475, bug 17137) These patches fix a security vulnerability reported privately by Stephane Chazelas. Pre-notification was sent out through linux-distros. I'd like to thank Stephane for reporting these issues, and Carlos for reviewing my patches. Before committing, I will add the bug number to the NEWS file and the ChangeLog. For obvious reasons, I did not have a bug number when I prepared those patches. The meat of the fixes is in the second patch. The first one is just the usual alloca hardening (technically not covered by the CVE assignment), and the third patch consists solely of documentation updates. From e175c6ba72a77cc3de4a7080f3cb9c072fc87353 Mon Sep 17 00:00:00 2001 From: Florian Weimer Date: Wed, 28 May 2014 14:05:03 +0200 Subject: [PATCH 3/3] manual: Update the locale documentation --- ChangeLog | 13 +++++ manual/locale.texi | 146 +++++++++++++++++++++++++++++++++++++++++------------ 2 files changed, 127 insertions(+), 32 deletions(-) diff --git a/ChangeLog b/ChangeLog index 207597c..d37328e 100644 --- a/ChangeLog +++ b/ChangeLog @@ -1,5 +1,18 @@ 2014-07-02 Florian Weimer + * manual/locale.texi (Locale Names): New section documenting + locale name syntax. Adjust menu and node chaining accordingly. + (Choosing Locale): Reference Locale Names, Locale Categories. + Mention setting LC_ALL=C. Reflect that name syntax is now + documented. + (Locale Categories): New section title. Reference Locale Names. + LC_ALL is an environment variable, but not a category. + (Setting the Locale): Remove "locale -a" invocation and LOCPATH + description, now in Locale Name. Reference that section. Locale + name syntax is now documented. + +2014-07-02 Florian Weimer + * locale/findlocale.c (name_present, valid_locale_name): New functions. (_nl_find_locale): Use the loc_name variable to store name diff --git a/manual/locale.texi b/manual/locale.texi index 45f1e94..ee1c3a1 100644 --- a/manual/locale.texi +++ b/manual/locale.texi @@ -29,6 +29,7 @@ will follow the conventions preferred by the user. * Setting the Locale:: How a program specifies the locale with library functions. * Standard Locales:: Locale names available on all systems. +* Locale Names:: Format of system-specific locale names. * Locale Information:: How to access the information for the locale. * Formatting Numbers:: A dedicated function to format numbers. * Yes-or-No Questions:: Check a Response against the locale. @@ -99,14 +100,16 @@ locale named @samp{espana-castellano} to use the standard conventions of most of Spain. The set of locales supported depends on the operating system you are -using, and so do their names. We can't make any promises about what -locales will exist, except for one standard locale called @samp{C} or -@samp{POSIX}. Later we will describe how to construct locales. -@comment (@pxref{Building Locale Files}). +using, and so do their names, except that the standard locale called +@samp{C} or @samp{POSIX} always exist. @xref{Locale Names}. + +In order to force the system to always use the default locale, the +user can set the @code{LC_ALL} environment variable to @samp{C}. @cindex combining locales -A user also has the option of specifying different locales for different -purposes---in effect, choosing a mixture of multiple locales. +A user also has the option of specifying different locales for +different purposes---in effect, choosing a mixture of multiple +locales. @xref{Locale Categories}. For example, the user might specify the locale @samp{espana-castellano} for most purposes, but specify the locale @samp{usa-english} for @@ -120,7 +123,7 @@ which locales apply. However, the user can choose to use each locale for a particular subset of those purposes. @node Locale Categories, Setting the Locale, Choosing Locale, Locales -@section Categories of Activities that Locales Affect +@section Locale Categories @cindex categories for locales @cindex locale categories @@ -128,7 +131,11 @@ The purposes that locales serve are grouped into @dfn{categories}, so that a user or a program can choose the locale for each category independently. Here is a table of categories; each name is both an environment variable that a user can set, and a macro name that you can -use as an argument to @code{setlocale}. +use as the first argument to @code{setlocale}. + +The contents of the environment variable (or the string in the second +argument to @code{setlocale}) has to be a valid locale name. +@xref{Locale Names}. @vtable @code @comment locale.h @@ -172,7 +179,7 @@ for affirmative and negative responses. @comment locale.h @comment ISO @item LC_ALL -This is not an environment variable; it is only a macro that you can use +This is not a category; it is only a macro that you can use with @code{setlocale} to set a single locale for all purposes. Setting this environment variable overwrites all selections by the other @code{LC_*} variables or @code{LANG}. @@ -355,13 +362,7 @@ The symbols in this section are defined in the header file @file{locale.h}. @c strndup @ascuheap @acsmem @c strcasecmp_l ok (C locale) The function @code{setlocale} sets the current locale for category -@var{category} to @var{locale}. A list of all the locales the system -provides can be created by running - -@pindex locale -@smallexample - locale -a -@end smallexample +@var{category} to @var{locale}. If @var{category} is @code{LC_ALL}, this specifies the locale for all purposes. The other possible values of @var{category} specify an @@ -386,10 +387,9 @@ is passed in as @var{locale} parameter. When you read the current locale for category @code{LC_ALL}, the value encodes the entire combination of selected locales for all categories. -In this case, the value is not just a single locale name. In fact, we -don't make any promises about what it looks like. But if you specify -the same ``locale name'' with @code{LC_ALL} in a subsequent call to -@code{setlocale}, it restores the same combination of locale selections. +If you specify the same ``locale name'' with @code{LC_ALL} in a +subsequent call to @code{setlocale}, it restores the same combination +of locale selections. To be sure you can use the returned string encoding the currently selected locale at a later time, you must make a copy of the string. It is not @@ -405,20 +405,15 @@ for @var{category}. If a nonempty string is given for @var{locale}, then the locale of that name is used if possible. +The effective locale name (either the second argument to +@code{setlocale}, or if the argument is an empty string, the name +obtained from the process environment) must be valid locale name. +@xref{Locale Names}. + If you specify an invalid locale name, @code{setlocale} returns a null pointer and leaves the current locale unchanged. @end deftypefun -The path used for finding locale data can be set using the -@code{LOCPATH} environment variable. The default path for finding -locale data is system specific. It is computed from the value given -as the prefix while configuring the C library. This value normally is -@file{/usr} or @file{/}. For the former the complete path is: - -@smallexample -/usr/lib/locale -@end smallexample - Here is an example showing how you might use @code{setlocale} to temporarily switch to a new locale. @@ -458,7 +453,7 @@ locale categories, and future versions of the library will do so. For portability, assume that any symbol beginning with @samp{LC_} might be defined in @file{locale.h}. -@node Standard Locales, Locale Information, Setting the Locale, Locales +@node Standard Locales, Locale Names, Setting the Locale, Locales @section Standard Locales The only locale names you can count on finding on all operating systems @@ -492,7 +487,94 @@ with the environment, rather than trying to specify some non-standard locale explicitly by name. Remember, different machines might have different sets of locales installed. -@node Locale Information, Formatting Numbers, Standard Locales, Locales +@node Locale Names, Locale Information, Standard Locales, Locales +@section Locale Names + +The following command prints a list of locales supported by the +system: + +@pindex locale +@smallexample + locale -a +@end smallexample + +@strong{Portability Note:} With the notable exception of the standard +locale names @samp{C} and @samp{POSIX}, locale names are +system-specific. + +Most locale names follow XPG syntax and consist of up to four parts: + +@smallexample +@var{language}[_@var{territory}[.@var{codeset}]][@@@var{modifier}] +@end smallexample + +Beside the first part, all of them are allowed to be missing. If the +full specified locale is not found, less specific ones are looked for. +The various parts will be stripped off, in the following order: + +@enumerate +@item +codeset +@item +normalized codeset +@item +territory +@item +modifier +@end enumerate + +For example, the locale name @samp{de_AT.iso885915@@euro} denotes a +German-language locale for use in Austria, using the ISO-8859-15 +(Latin-9) character set, and with the Euro as the currency symbol. + +In addition to locale names which follow XPG syntax, systems may +provide aliases such as @samp{german}. Both categories of names must +not contain the slash character @samp{/}. + +If the locale name starts with a slash @samp{/}, it is treated as a +path relative to the configured locale directories; see @code{LOCPATH} +below. The specified path must not contain a component @samp{..}, or +the name is invalid, and @code{setlocale} will fail. + +@strong{Portability Note:} POSIX suggests that if a locale name starts +with a slash @samp{/}, it is resolved as an absolute path. However, +@theglibc{} treats it as a relative path under the directories listed +in @code{LOCPATH} (or the default locale directory if @code{LOCPATH} +is unset). + +Locale names which are longer than an implementation-defined limit are +invalid and cause @code{setlocale} to fail. + +As a special case, locale names used with @code{LC_ALL} can combine +several locales, reflecting different locale settings for different +categories. For example, you might want to use a U.S. locale with ISO +A4 paper format, so you set @code{LANG} to @samp{en_US.UTF-8}, and +@code{LC_PAPER} to @samp{de_DE.UTF-8}. In this case, the +@code{LC_ALL}-style combined locale name is + +@smallexample +LC_CTYPE=en_US.UTF-8;LC_TIME=en_US.UTF-8;LC_PAPER=de_DE.UTF-8;@dots{} +@end smallexample + +followed by other category settings not shown here. + +@vindex LOCPATH +The path used for finding locale data can be set using the +@code{LOCPATH} environment variable. This variable lists the +directories in which to search for locale definitions, separated by a +colon @samp{:}. + +The default path for finding locale data is system specific. A typical +value for the @code{LOCPATH} default is: + +@smallexample +/usr/share/locale +@end smallexample + +The value of @code{LOCPATH} is ignored by privileged programs for +security reasons, and only the default directory is used. + +@node Locale Information, Formatting Numbers, Locale Names, Locales @section Accessing Locale Information There are several ways to access locale information. The simplest -- 1.9.3