This is the mail archive of the libc-alpha@sources.redhat.com mailing list for the glibc project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: mon_grouping bug


On Friday 14 March 2003 17:47, you wrote:
> Gerhard Tonn <GerhardTonn at gammatau dot de> writes:
> |> Hi,
> |>
> |> running the LSB 1.3 testsuite on a platform that has unsigned char as
> |> default, gives a failure during the localisation tests. The mon_grouping
> |> parameter of the LOCALE being tested is defined as 1;2;-1.  After
> |> installation of the LOCALE the command 'locale -k mon_grouping' returns
> |> mon_grouping=1;2;127.
> |>
> |> This seems to be a bug in the glibc. The following code fragment in
> |> ld-monetary.c maps the '-1' to 127 which is the value of SCHAR_MAX which
> |> is equal to CHAR_MAX, if char is signed per default.
> |>
> |> if (now->tok == tok_minus1)
> |>                     {
> |>                       if (!ignore_content)
> |>                         grouping[act++] = '\177';
> |>                     }
> |>
> |> The following code fragment in locale.c prints '-1' only if the value
> |> equals CHAR_MAX which is not true on platforms that have unsigned char
> |> as default.
> |>
> |>  printf ("%d", *val == CHAR_MAX ? -1 : *val);
>
> IMHO the right fix is to change CHAR_MAX to '\177' since that's what's
> actually being stored.
>

The description of the grouping and mon_grouping members says
"An element value of CHAR_MAX ends any further grouping (and hence ends the 
string)". 

The best solution would probably be the substitution of '\177' by CHAR_MAX in 
ld-monetary, but I am not sure what the implications are. Therefore I decided 
not to change any running code but only to add a code path.

Gerhard


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]