[PATCH] Update:Change format specifiers to fix 4 warnings.
gedare at rtems.org
Fri Apr 7 14:58:39 UTC 2017
On Fri, Apr 7, 2017 at 10:05 AM, Joel Sherrill <joel at rtems.org> wrote:
> On Fri, Apr 7, 2017 at 4:15 AM, Cillian O'Donnell <cpodonnell8 at gmail.com>
>> Great! That looks like a cleaner, more standardized solution. I was just
>> hacking away until the warnings stopped. So can I go ahead and #include
>> <inttypes.h> or should I wait until <rtems/inttypes.h> is defined before
>> doing any more? Does this work for printk() too?
> First the easy answer. printk() supports a subset of printf() formats. So
> as long as it is a supported printf format, then yes. But I just looked for
> printk() documentation and I don't see any. The code is in
> cpukit/libcsupport/src/vprintk.c for a double check. It is supposed to
> support everything required by any test.
> If a file only uses C99 standard PRI constants, then use <inttypes.h>
> If a file needs the extra ones being collected into <rtems/inttypes.h>,
> then it should include that.
> No need to introduce an unneeded dependency.
> Funny, how this simple set of warnings has rippled. It also makes me
> want to file a ticket for printk() documentation and adding code style
> rules about <inttypes.h>.
Also we should consider converting our code style into a doc instead
of wiki page.
>> Thanks Joel.
>> devel mailing list
>> devel at rtems.org
> devel mailing list
> devel at rtems.org
More information about the devel