Hi,
we try to squeeze a 64bit type into a 32bit one which kind of works on little-endian because we checked that the value is not larger then UINT32_MAX before but fails on big-endian. This patch tries to fix it.
bye, Sumit
On Thu, Jan 29, 2015 at 09:17:28PM +0100, Sumit Bose wrote:
Hi,
we try to squeeze a 64bit type into a 32bit one which kind of works on little-endian because we checked that the value is not larger then UINT32_MAX before but fails on big-endian. This patch tries to fix it.
bye, Sumit
ACK for this patch because it fixes a bug, but we should check the usage of ldb_msg_find_attr_as_uint64() more systematically. There's a couple more places where I saw the same issue.
I filed https://fedorahosted.org/sssd/ticket/2578 to track this.
On Fri, Jan 30, 2015 at 03:36:39PM +0100, Jakub Hrozek wrote:
On Thu, Jan 29, 2015 at 09:17:28PM +0100, Sumit Bose wrote:
Hi,
we try to squeeze a 64bit type into a 32bit one which kind of works on little-endian because we checked that the value is not larger then UINT32_MAX before but fails on big-endian. This patch tries to fix it.
bye, Sumit
ACK for this patch because it fixes a bug, but we should check the usage of ldb_msg_find_attr_as_uint64() more systematically. There's a couple more places where I saw the same issue.
I filed https://fedorahosted.org/sssd/ticket/2578 to track this.
There patches were pushed some time ago, but I forgot to send the pushmail while working on the downstream builds. Pushed to master: * 866ab45027c83fafb7f7f45d34d3e1e7721b77dc and sssd-1-12: * e3e6a3f87496667bf6456de8855ff831bf222abb
sssd-devel@lists.fedorahosted.org