This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: Testsuite adaptions for the avr target / nested function support
- From: Janis Johnson <janis187 at us dot ibm dot com>
- To: Björn Haase <bjoern dot m dot haase at web dot de>
- Cc: Janis Johnson <janis187 at us dot ibm dot com>, gcc-patches at gcc dot gnu dot org
- Date: Thu, 7 Apr 2005 16:20:53 -0700
- Subject: Re: Testsuite adaptions for the avr target / nested function support
- References: <200502222300.12426.bjoern.m.haase@web.de> <20050223011202.GA7898@us.ibm.com> <200503061136.41018.bjoern.m.haase@web.de>
On Sun, Mar 06, 2005 at 11:36:40AM +0100, Björn Haase wrote:
> when preparing to prepare the testsuite to be better adapted also for small
> targets, I have made use of the approach that you had suggested. You will
> find attached a patch that defines three new effective-target keywords:
This was sent directly to me and copied to gcc-patches, so obviously I
have seen it; I'm very sorry to have forgotten about it. I probably put
off looking at it because it was an attachment rathern than included in
the body of the message.
> nested_functions
> trampolines
> int_larger_than_16_bits
>
> All of those make use of the knowledge stored in the board information
> variables. With the new keyword "trampolines" it now will also be possible to
> uprightly skip tests that use trampolines instead of continuing to report the
> testcases to pass.
Using the board information variables seems like a good idea, but instead
of checking those in is-effective-keyword, add new procs
check_effective_target_nested_functions
check_effective_target_trampolines
check_effective_target_int_larger_than_16_bits
That allows the effective-target keywords to be recognized in selector
expressions.
> The attached patch also adds "avr-*-*" to the list of targets that do not
> support profiling.
That part is fine.
Janis