Description of problem: When an unsigned int is passed to the error_message() function, it is sign-extended so that the error_message() function no longer matches the right table. This happens frequently with error codes returned by gssapi libraries, which are returned in unsigned variables. Version-Release number of selected component (if applicable): e2fsprogs-libs-1.39-23.el5 How reproducible: Always Steps to Reproduce: cat > com_err.c << EOF #include <sys/types.h> #include <stdio.h> #include <stdlib.h> #include <krb5.h> #include <gssapi/gssapi.h> int main(int argc, char **argv) { int i; unsigned int u; krb5_context ctx; ctx = NULL; krb5_init_context(&ctx); if (argc > 1) { for (i = 1; i < argc; i++) { u = atol(argv[i]); printf("Error message (%08x): \"%s\".\n", u, error_message(u)); } } else { u = KRB5KDC_ERR_S_PRINCIPAL_UNKNOWN; printf("Error message: \"%s\".\n", error_message(u)); } return 0; } EOF cc -Wall -o com_err `krb5-config --cflags` com_err.c `krb5-config --libs` ./com_err Actual results: Error message: "Unknown code krb5 7". Expected results: Error message: "Server not found in Kerberos database".
Straightforward upstream patch is: http://git.kernel.org/?p=fs/ext2/e2fsprogs.git;a=commitdiff;h=6b6c27fb8a45f264194d8dd637643d8b4898271a libcom_err: Fix sign-extension problem on 64-bit systems in error_message() On 64-bit systems (or anything with sizeof(long) > sizeof(int)), we sometimes get error codes passed to error_message which have been cast from an (int) to an (unsigned int). This almost always happens if you're using libgssapi_krb5, which returns an error code which is less than 0 but is returned in an (unsigned int). For example, -1765328377L gets cast to 2529638919, which is 0x96c73a07, not 0xffffffff96c73a07, so error_message() fails to find a matching error table. When error_message() then calls the error_table_name() function to get a name to use in the "unknown code" message, it gets a correct value back. This happens because error_table_name() drops most of the higher bits of the parameter it's passed before doing anything else with it (& 077777777f, or & 0xffffff). If we did the same thing in error_message(), we wouldn't have a problem there, either. Problem reported and fixed by: Nalin Dahyabhai Addresses-Sourceforge-Bug: #1809658 Signed-off-by: "Theodore Ts'o" <tytso> --- diff --git a/lib/et/error_message.c b/lib/et/error_message.c index 4e84e35..708dd2f 100644 --- a/lib/et/error_message.c +++ b/lib/et/error_message.c @@ -69,7 +69,7 @@ const char * error_message (errcode_t code) #endif } for (et = _et_list; et; et = et->next) { - if (et->table->base == table_num) { + if ((et->table->base & 0xffffffL) == (table_num & 0xffffffL)) { /* This is the right table */ if (et->table->n_msgs <= offset) goto oops; @@ -77,7 +77,7 @@ const char * error_message (errcode_t code) } } for (et = _et_dynamic_list; et; et = et->next) { - if (et->table->base == table_num) { + if ((et->table->base & 0xffffffL) == (table_num & 0xffffffL)) { /* This is the right table */ if (et->table->n_msgs <= offset) goto oops;
Built & tagged in e2fsprogs-1.39-28.el5
An advisory has been issued which should help the problem described in this bug report. This report is therefore being closed with a resolution of ERRATA. For more information on therefore solution and/or where to find the updated files, please follow the link below. You may reopen this bug report if the solution does not work for you. http://rhn.redhat.com/errata/RHBA-2011-1080.html