Re: [PATCH RESEND] erofs: fix compile warnings when moving out include/trace/events/erofs.h

[Chao Yu]

On 2019-8-26 21:26, Gao Xiang wrote:
> As Stephon reported [1], many compile warnings are raised when
> moving out include/trace/events/erofs.h:
> 
> In file included from include/trace/events/erofs.h:8,
>                  from <command-line>:
> include/trace/events/erofs.h:28:37: warning: 'struct dentry' declared inside parameter list will not be visible outside of this definition or declaration
>   TP_PROTO(struct inode *dir, struct dentry *dentry, unsigned int flags),
>                                      ^~~~~~
> include/linux/tracepoint.h:233:34: note: in definition of macro '__DECLARE_TRACE'
>   static inline void trace_##name(proto)    \
>                                   ^~~~~
> include/linux/tracepoint.h:396:24: note: in expansion of macro 'PARAMS'
>   __DECLARE_TRACE(name, PARAMS(proto), PARAMS(args),  \
>                         ^~~~~~
> include/linux/tracepoint.h:532:2: note: in expansion of macro 'DECLARE_TRACE'
>   DECLARE_TRACE(name, PARAMS(proto), PARAMS(args))
>   ^~~~~~~~~~~~~
> include/linux/tracepoint.h:532:22: note: in expansion of macro 'PARAMS'
>   DECLARE_TRACE(name, PARAMS(proto), PARAMS(args))
>                       ^~~~~~
> include/trace/events/erofs.h:26:1: note: in expansion of macro 'TRACE_EVENT'
>  TRACE_EVENT(erofs_lookup,
>  ^~~~~~~~~~~
> include/trace/events/erofs.h:28:2: note: in expansion of macro 'TP_PROTO'
>   TP_PROTO(struct inode *dir, struct dentry *dentry, unsigned int flags),
>   ^~~~~~~~
> 
> That makes me very confused since most original EROFS tracepoint code
> was taken from f2fs, and finally I found
> 
> commit 43c78d88036e ("kbuild: compile-test kernel headers to ensure they are self-contained")
> 
> It seems these warnings are generated from KERNEL_HEADER_TEST feature and
> ext4/f2fs tracepoint files were in blacklist.

For f2fs.h, it will be only used by f2fs module, I guess it's okay to let it
stay in blacklist...

> 
> Anyway, let's fix these issues for KERNEL_HEADER_TEST feature instead
> of adding to blacklist...
> 
> [1] https://lore.kernel.org/lkml/20190826162432.11100665@xxxxxxxxxxxxxxxx/
> Reported-by: Stephen Rothwell <sfr@xxxxxxxxxxxxxxxx>
> Signed-off-by: Gao Xiang <gaoxiang25@xxxxxxxxxx>

Reviewed-by: Chao Yu <yuchao0@xxxxxxxxxx>

Thanks,