[SSSD] [PATCH] LDAP: Add a new error code for malformed access control filter

Simo Sorce simo at redhat.com
Thu Jan 9 16:58:28 UTC 2014


On Thu, 2014-01-09 at 17:17 +0100, Jakub Hrozek wrote:
> On Thu, Jan 09, 2014 at 10:55:39AM -0500, Simo Sorce wrote:
> > On Thu, 2014-01-09 at 15:58 +0100, Jakub Hrozek wrote:
> > > On Thu, Jan 09, 2014 at 08:45:08AM -0500, Simo Sorce wrote:
> > > > On Thu, 2014-01-09 at 10:50 +0100, Jakub Hrozek wrote:
> > > > > I was actually thinking about mass-replacing strerror for sss_strerror
> > > > > in master..since sss_strerror is backwards-compatible, there should be
> > > > > no
> > > > > issues even if strerror was used to print error message from a system
> > > > > call like write(2) where only errno-ranged error code was present.
> > > > > 
> > > > > Given that we already have one large automated patch (the DEBUG
> > > > > levels) coming up and the strerror() calls are largely in the same
> > > > > area
> > > > > the churn wouldn't be too high..
> > > > > 
> > > > It would be simpler to add this in header files:
> > > > #undef strerror
> > > > #define strerror sss_strerror
> > > > 
> > > > Simo.
> > > 
> > > Right, since sss_strerror is backwards compatible, this would work.
> > > 
> > > Should we convert the sss_strerror back to strerror, then, to not
> > > confuse developers with two similar calls?
> > 
> > No, we should rather encourage developers to slowly move from strerror()
> > to sss_strerror()
> > 
> > I just hate mass changes as it is easy to sneak nasty bugs inadvertently
> > (plus makes git blame harder to use :)
> > 
> 
> Right, I'm not fond of them either, but since Nikolai is already doing a
> mass-change of the DEBUG levels and strerror() calls are typically, if
> not always used in a DEBUG statement, this could be a good time to
> standardize on either strerror or sss_strerror.

Yeah, I guess so.

Simo.

-- 
Simo Sorce * Red Hat, Inc * New York




More information about the sssd-devel mailing list