On Mon, Jun 08, 2020 at 03:46:57PM +0800, Li Zhijian wrote:

On 6/6/20 12:35 AM, Mark Brown wrote:

quoted

quoted

arch/arm64/kernel/vdso/vgettimeofday.c:9:5: warning: no previous prototype for '__kernel_clock_gettime' [-Wmissing-prototypes]
9 | int __kernel_clock_gettime(clockid_t clock,
|     ^~~~~~~~~~~~~~~~~~~~~~

quoted

I'm not seeing this here and what we're doing is in line with the
existing idiom as can be seen from the __kernel_clock_gettime() code
flagging the same thing.  Possibly an old/outdated toolchain?

Actually, the origin mail contained 4 warnings,  and only one is *new*, as
the mail mentioned that

quoted

All warnings (new ones prefixed by >>, old ones prefixed by <<):

0Day/LKP had marked the *new* ones prefixed by '>>'

My point here is that all the warnings, both old and new, look spurious
and the new warning comes from following the pattern that generates the
existing warnings.  I'm not seeing any of those warnings locally or in
other build services like KernelCI.