Re: [Qemu-devel] [PULL 00/20] QAPI patches

[Peter Maydell]

On 12 October 2015 at 14:25, Peter Maydell <address@hidden> wrote:
> On 9 October 2015 at 16:14, Markus Armbruster <address@hidden> wrote:
>> The following changes since commit b37686f7e84b22cfaf7fd01ac5133f2617cc3027:
>>
>>   Merge remote-tracking branch 'remotes/stefanha/tags/tracing-pull-request' 
>> into staging (2015-10-09 12:18:14 +0100)
>>
>> are available in the git repository at:
>>
>>   git://repo.or.cz/qemu/armbru.git tags/pull-qapi-2015-10-09
>>
>> for you to fetch changes up to 51c605bab50f1b9e0766aa9ed14c1abe58eb7f48:
>>
>>   qapi: Simplify gen_visit_fields() error handling (2015-10-09 16:44:10 
>> +0200)
>>
>> ----------------------------------------------------------------
>> QAPI patches
>>
>> ----------------------------------------------------------------
>
> Hi. I see a clang compile failure due to this warning (OSX and Linux):
>
> /Users/pm215/src/qemu-for-merges/tests/test-qmp-input-visitor.c:500:33:
> warning: implicit conversion from 'double' to 'gint64' (aka 'long')
> changes value from 42.5 to 42 [-Wliteral-conversion]

Whoops, failed to quote the whole message:
/home/petmay01/linaro/qemu-for-merges/tests/test-qmp-input-visitor.c:500:33:
error: implicit conversion from 'double' to 'gint64' (aka 'long')
changes value from 42.5 to 42 [-Werror,-Wliteral-conversion]
    g_assert_cmpint(ani->n, ==, 42.5);
    ~~~~~~~~~~~~~~~~~~~~~~~~~~~~^~~~~
/usr/include/glib-2.0/glib/gtestutils.h:45:74: note: expanded from
macro 'g_assert_cmpint'
#define g_assert_cmpint(n1, cmp, n2)    do { gint64 __n1 = (n1), __n2 = (n2); \
                                                                         ^
1 error generated.

Trying to compare floating point values with g_assert_cmpint
doesn't seem very likely to do what the author intended.

-- PMM