This is the mail archive of the
glibc-bugs@sourceware.org
mailing list for the glibc project.
[Bug libc/14821] New: sunrpc code problems for 64-bit
- From: "jsm28 at gcc dot gnu.org" <sourceware-bugzilla at sourceware dot org>
- To: glibc-bugs at sources dot redhat dot com
- Date: Thu, 08 Nov 2012 22:41:27 +0000
- Subject: [Bug libc/14821] New: sunrpc code problems for 64-bit
- Auto-submitted: auto-generated
http://sourceware.org/bugzilla/show_bug.cgi?id=14821
Bug #: 14821
Summary: sunrpc code problems for 64-bit
Product: glibc
Version: 2.16
Status: NEW
Severity: normal
Priority: P2
Component: libc
AssignedTo: unassigned@sourceware.org
ReportedBy: jsm28@gcc.gnu.org
CC: drepper.fsp@gmail.com
Classification: Unclassified
sunrpc/clnt_tcp.c:clnttcp_control writes u_long values into a buffer at offsets
both 12 and 16 (BYTES_PER_XDR_UNIT is always 4). This has several problems:
* When u_long is 64 bits, these values overlap.
* When u_long is 64 bits, one of them will not be 8-byte-aligned but the
processor / ABI may require such values to be 8-byte aligned.
* When u_long is 64 bits and the processor is bit-endian, the value will be
written in a different location from where the other end of the connection, on
another system, expects it (htonl deals with getting the right byte order of
the low 4 bytes, but nothing ensures that the low 4 bytes of the value go in
the right place in the buffer).
clnt_udp.c and clnt_unix.c have the same issue.
This appears to cause NFS mounts, with an n64 MIPS mount.nfs binary, to fail; I
don't have a better testcase than that.
Testing a patch.
--
Configure bugmail: http://sourceware.org/bugzilla/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.