Justin Erenkrantz wrote:
> On 2/4/07, Garrett Rooney <[EMAIL PROTECTED]> wrote:
>> I was under the impression that apr_status_t is a superset of errno,
>> so returning the current errno should be perfectly fine...
> 
> It is.  -- justin

Sort of.  Posix errno's.  Other platforms need to munge their values
with apr_get_os_error/apr_set_os_error.  (Hmmm - it seems long past
time to rename these apr_os_error_get/apr_os_error_set.)

Other ranges are reserved for other errors ... getaddr results, and
dlerror results.

It gets complex - maybe some invocation of apr-config aught to spell out
what ranges have been mapped to what?

Bill

Reply via email to