[PATCH] seq_buf: avoid type mismatch for seq_buf_init

[PATCH] seq_buf: avoid type mismatch for seq_buf_init

[Arnd Bergmann]

From: Arnd Bergmann <arnd@arndb.de>

Building with W=2 prints a number of warnings for one function that
has a pointer type mismatch:

linux/seq_buf.h: In function 'seq_buf_init':
linux/seq_buf.h:35:12: warning: pointer targets in assignment from 'unsigned char *' to 'char *' differ in signedness [-Wpointer-sign]

Change the type in the function prototype according to the type in
the structure.

Fixes: 9a7777935c34 ("tracing: Convert seq_buf fields to be like seq_file fields")
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
---
 include/linux/seq_buf.h   | 2 +-
 include/linux/trace_seq.h | 4 ++--
 2 files changed, 3 insertions(+), 3 deletions(-)

diff --git a/include/linux/seq_buf.h b/include/linux/seq_buf.h
index fb0205d87d3c..9d6c28cc4d8f 100644
--- a/include/linux/seq_buf.h
+++ b/include/linux/seq_buf.h
@@ -30,7 +30,7 @@ static inline void seq_buf_clear(struct seq_buf *s)
 }
 
 static inline void
-seq_buf_init(struct seq_buf *s, unsigned char *buf, unsigned int size)
+seq_buf_init(struct seq_buf *s, char *buf, unsigned int size)
 {
 	s->buffer = buf;
 	s->size = size;
diff --git a/include/linux/trace_seq.h b/include/linux/trace_seq.h
index 6c30508fca19..5a2c650d9e1c 100644
--- a/include/linux/trace_seq.h
+++ b/include/linux/trace_seq.h
@@ -12,7 +12,7 @@
  */
 
 struct trace_seq {
-	unsigned char		buffer[PAGE_SIZE];
+	char			buffer[PAGE_SIZE];
 	struct seq_buf		seq;
 	int			full;
 };
@@ -51,7 +51,7 @@ static inline int trace_seq_used(struct trace_seq *s)
  * that is about to be written to and then return the result
  * of that write.
  */
-static inline unsigned char *
+static inline char *
 trace_seq_buffer_ptr(struct trace_seq *s)
 {
 	return s->buffer + seq_buf_used(&s->seq);
-- 
2.27.0

RE: [PATCH] seq_buf: avoid type mismatch for seq_buf_init

[Rojewski, Cezary]

On 2020-10-26 5:10 PM, Arnd Bergmann wrote:
> From: Arnd Bergmann <arnd@arndb.de>
> 
> Building with W=2 prints a number of warnings for one function that
> has a pointer type mismatch:
> 
> linux/seq_buf.h: In function 'seq_buf_init':
> linux/seq_buf.h:35:12: warning: pointer targets in assignment from 'unsigned char *' to 'char *' differ in signedness [-Wpointer-sign]
> 
> Change the type in the function prototype according to the type in
> the structure.
> 
> Fixes: 9a7777935c34 ("tracing: Convert seq_buf fields to be like seq_file fields")
> Signed-off-by: Arnd Bergmann <arnd@arndb.de>
> ---

Reviewed-by: Cezary Rojewski <cezary.rojewski@intel.com>

Thanks,
Czarek

Re: [PATCH] seq_buf: avoid type mismatch for seq_buf_init

[Steven Rostedt]

On Mon, 26 Oct 2020 17:10:58 +0100
Arnd Bergmann <arnd@kernel.org> wrote:

> From: Arnd Bergmann <arnd@arndb.de>
> 
> Building with W=2 prints a number of warnings for one function that
> has a pointer type mismatch:
> 
> linux/seq_buf.h: In function 'seq_buf_init':
> linux/seq_buf.h:35:12: warning: pointer targets in assignment from 'unsigned char *' to 'char *' differ in signedness [-Wpointer-sign]

I've always hated the warning about char * and unsigned char *, as they are
mostly meaningless. Yes, bugs happen with int to unsigned int conversions,
but this is dealing with strings, where unsigned char * and char * are
basically equivalent, except when it comes to one thing, which is why I
prefer unsigned char * over char *, and that is printing out the numerical
values of a buffer, if they go above 177, the char * prints the negative
value, but unsigned char * keeps printing what you would expect.

As this is just an annoyance (extra warnings), and not really a "fix", I'll
queue it up for the next merge window.

-- Steve

> 
> Change the type in the function prototype according to the type in
> the structure.
> 
> Fixes: 9a7777935c34 ("tracing: Convert seq_buf fields to be like seq_file fields")
> Signed-off-by: Arnd Bergmann <arnd@arndb.de>
> ---

Re: [PATCH] seq_buf: avoid type mismatch for seq_buf_init

[Arnd Bergmann]

On Mon, Oct 26, 2020 at 6:23 PM Steven Rostedt <rostedt@goodmis.org> wrote:
>
> On Mon, 26 Oct 2020 17:10:58 +0100
> Arnd Bergmann <arnd@kernel.org> wrote:
>
> > From: Arnd Bergmann <arnd@arndb.de>
> >
> > Building with W=2 prints a number of warnings for one function that
> > has a pointer type mismatch:
> >
> > linux/seq_buf.h: In function 'seq_buf_init':
> > linux/seq_buf.h:35:12: warning: pointer targets in assignment from 'unsigned char *' to 'char *' differ in signedness [-Wpointer-sign]
>
> I've always hated the warning about char * and unsigned char *, as they are
> mostly meaningless. Yes, bugs happen with int to unsigned int conversions,
> but this is dealing with strings, where unsigned char * and char * are
> basically equivalent, except when it comes to one thing, which is why I
> prefer unsigned char * over char *, and that is printing out the numerical
> values of a buffer, if they go above 177, the char * prints the negative
> value, but unsigned char * keeps printing what you would expect.

I agree it's a super annoying warning, which is exactly why I sent the
fixes to shut it up in common headers. At least that way, building a specific
driver with W=2 will only show the warnings in that driver, rather than
those in header files as well.

> As this is just an annoyance (extra warnings), and not really a "fix", I'll
> queue it up for the next merge window.

Yes, that was the idea, thanks!

        Arnd