You are viewing a plain text version of this content. The canonical link for it is here.

Posted to dev@apr.apache.org by Mladen Turk <mt...@apache.org> on 2005/01/14 12:18:04 UTC

resend [PATCH] set process attribute uid

Hi,

It's been couple of months since I proposed a patch
that will allow setting uid/gid for newly created
processes using apr_proc_create.

If no one objects I'll commit the patch.

Regards,
Mladen

Re: resend [PATCH] set process attribute uid

Posted by Mladen Turk <mt...@apache.org>.

Joe Orton wrote:
 > On Fri, Jan 14, 2005 at 12:18:04PM +0100, Mladen Turk wrote:
 >
 >>Hi,
 >>
 >>It's been couple of months since I proposed a patch
 >>that will allow setting uid/gid for newly created
 >>processes using apr_proc_create.
 >
 >
 > New warnings:
 >
 > threadproc/unix/proc.c: In function `apr_proc_create':
 > threadproc/unix/proc.c:421: warning: unused variable `rv'
 > threadproc/unix/proc.c:431: warning: unused variable `rv'
 >

I'll remove those two 'apr_status_t rv;' unused declarations.

 > I think it would be much better to define the apr_procattr_user_set
 > interface to fail with some particular (new?) APR_E* error code if the
 > password is not given on platforms which require a password, than to use
 > an apr.h feature flag for this.
 >

The flag APR_PROCATTR_USER_SET_REQUIRES_PASSWORD should be present
to give the users application hint whether to use or disregard
the password.
Also NULL password should mean empty password (I'll fix the win32
code to accept that), so there is no EINVAL.

Mladen.

Re: resend [PATCH] set process attribute uid

Posted by Joe Orton <jo...@redhat.com>.

On Fri, Jan 14, 2005 at 12:18:04PM +0100, Mladen Turk wrote:
> Hi,
> 
> It's been couple of months since I proposed a patch
> that will allow setting uid/gid for newly created
> processes using apr_proc_create.

New warnings:

threadproc/unix/proc.c: In function `apr_proc_create':
threadproc/unix/proc.c:421: warning: unused variable `rv'
threadproc/unix/proc.c:431: warning: unused variable `rv'

I think it would be much better to define the apr_procattr_user_set
interface to fail with some particular (new?) APR_E* error code if the
password is not given on platforms which require a password, than to use
an apr.h feature flag for this.

joe