This is the mail archive of the
ecos-discuss@sources.redhat.com
mailing list for the eCos project.
Re: Will gcc3.3 will work with eCos?.
- From: Bart Veer <bartv at ecoscentric dot com>
- To: radhaamanji at yahoo dot co dot uk
- Cc: ecos-discuss at sources dot redhat dot com, Thomas dot Koeller at baslerweb dot com,fredrik at wespot dot com, satishkumar at sanyo dot co dot in
- Date: Tue, 13 Jan 2004 11:26:11 +0000 (GMT)
- Subject: Re: [ECOS] Will gcc3.3 will work with eCos?.
- References: <20040113042759.25390.qmail@web25007.mail.ukl.yahoo.com>
>>>>> " " == =?iso-8859-1?q?radhakrishnan=20R?= <iso-8859-1> writes:
> Hi All,
> I like thank you for answering my questions.
> Do you meant to say, if i use __attribute__
> ((init_priority (42))) with the required priority
> number, i will get the same behaviour as
> -finit-priority?.
Not quite. The syntax for giving a C++ static object a constructor
priority has always been __attribute__ ((init_priority (<num>))),
since that functionality was added back in 1997 or so. However early
on this syntax was only accepted if you also specified -finit-priority
on the command line. If you tried to use an init_priority attribute
without -finit-priority then the compiler would give an error or
something like that - I don't remember the exact details.
It did not take long before the init_priority attribute was considered
mainstream rather than experimental. Hence the compiler was changed so
that it would always accept the attribute, irrespective of whether or
not -finit-priority was specified on the command line. So
-finit-priority no longer served any purpose. It was kept around for a
while for compatibility reasons, but removed recently as part of a
more general clean-up of compiler command-line options.
Bart
--
Bart Veer eCos Configuration Architect
http://www.ecoscentric.com/ The eCos and RedBoot experts
--
Before posting, please read the FAQ: http://sources.redhat.com/fom/ecos
and search the list archive: http://sources.redhat.com/ml/ecos-discuss