Re: [boost] HP-UX_pa_risc_aCC test failure

Subject: Re: [boost] HP-UX_pa_risc_aCC test failure
From: Emil Dotchevski (emil_at_[hidden])
Date: 2008-09-11 12:50:29

On Thu, Sep 11, 2008 at 8:41 AM, Matus Chochlik <chochlik_at_[hidden]> wrote:
> On Thu, Sep 11, 2008 at 4:02 PM, Bruno Lalande <bruno.lalande_at_[hidden]> wrote:
>> Hello,
>>
>> I've been using this compiler a few years ago and I remember it wasn't
>> possible to write a function that only throws an exception without
>> having a warning complaining that the function never returns.
>> Basically, the compiler seemed to "ignore" the fact that a throw is
>> supposed to return. It might be the reason of your error. My
>> workaround used to be to put a dummy return statement just after the
>> throw...
>
> if this is the reason, then using the BOOST_UNREACHABLE_RETURN(result)
> macro is (AFAICT) the best option as some other compilers complain about
> returns (or any other code) behind unconditional throw being unreachable.

No, this is not the reason.

This test:

http://beta.boost.org/development/tests/trunk/developer/output/HP-UX_pa_risc_aCC-boost-bin-v2-libs-exception-test-throw_exception_test-test-acc-pa_risc-debug.html

fails only on that platform (if you ignore stale results in the test
matrix, and Borland which isn't supported), and the problem being
reported is that a function which is supposed to throw an exception,
returns without throwing.

Emil Dotchevski
Reverge Studios, Inc.
http://www.revergestudios.com/reblog/index.php?n=ReCode