git.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Jeff King <peff@peff.net>
To: Felipe Contreras <felipe.contreras@gmail.com>
Cc: git@vger.kernel.org
Subject: Re: [PATCH 1/3] Revert "make error()'s constant return value more visible"
Date: Mon, 5 May 2014 01:49:01 -0400	[thread overview]
Message-ID: <20140505054901.GA19331@sigill.intra.peff.net> (raw)
In-Reply-To: <1399183975-2346-2-git-send-email-felipe.contreras@gmail.com>

On Sun, May 04, 2014 at 01:12:53AM -0500, Felipe Contreras wrote:

> So it looks like gcc is smarter now, and in trying to fix a few warnings
> we generated hundreds more.
> 
> This reverts commit e208f9cc7574f5980faba498d0aa30b4defeb34f.

And now we've gone the other way, and re-enabled the initial warnings.
Can we come up with a solution that helps both cases?

I could not find a way to annotate a value as "maybe unused", but we can
hide it inside a function, like:

-- >8 --
Subject: inline error() function

The error() function does two things: it prints an error,
and it returns "-1" as a convenience to allow:

  return error("foo");

Commit e208f9c converted this to a macro to make the
constant return value more visible to the compiler. However,
recent versions of gcc complain when error is used in a void
context, as the constant "-1" ends up unused.

Instead, let's convert error() to a static inline, which
should accomplish the same thing without the extra warnings
(because gcc will not warn about unused return values unless
warn_unused_result is specified for the particular
function).

Signed-off-by: Jeff King <peff@peff.net>
---
 git-compat-util.h | 20 ++++++++++++--------
 usage.c           |  8 +-------
 2 files changed, 13 insertions(+), 15 deletions(-)

diff --git a/git-compat-util.h b/git-compat-util.h
index f6d3a46..3aef0d3 100644
--- a/git-compat-util.h
+++ b/git-compat-util.h
@@ -310,7 +310,6 @@ extern NORETURN void usage(const char *err);
 extern NORETURN void usagef(const char *err, ...) __attribute__((format (printf, 1, 2)));
 extern NORETURN void die(const char *err, ...) __attribute__((format (printf, 1, 2)));
 extern NORETURN void die_errno(const char *err, ...) __attribute__((format (printf, 1, 2)));
-extern int error(const char *err, ...) __attribute__((format (printf, 1, 2)));
 extern void warning(const char *err, ...) __attribute__((format (printf, 1, 2)));
 
 #ifndef NO_OPENSSL
@@ -325,14 +324,19 @@ extern void warning(const char *err, ...) __attribute__((format (printf, 1, 2)))
 
 /*
  * Let callers be aware of the constant return value; this can help
- * gcc with -Wuninitialized analysis. We restrict this trick to gcc, though,
- * because some compilers may not support variadic macros. Since we're only
- * trying to help gcc, anyway, it's OK; other compilers will fall back to
- * using the function as usual.
+ * gcc with -Wuninitialized analysis.
  */
-#if defined(__GNUC__) && ! defined(__clang__)
-#define error(...) (error(__VA_ARGS__), -1)
-#endif
+__attribute__((format (printf, 1, 0)))
+extern void error_impl(const char *err, va_list params);
+__attribute__((format (printf, 1, 2)))
+static inline int error(const char *err, ...)
+{
+    va_list params;
+    va_start(params, err);
+    error_impl(err, params);
+    va_end(params);
+    return -1;
+}
 
 extern void set_die_routine(NORETURN_PTR void (*routine)(const char *err, va_list params));
 extern void set_error_routine(void (*routine)(const char *err, va_list params));
diff --git a/usage.c b/usage.c
index ed14645..5456e4b 100644
--- a/usage.c
+++ b/usage.c
@@ -138,15 +138,9 @@ void NORETURN die_errno(const char *fmt, ...)
 	va_end(params);
 }
 
-#undef error
-int error(const char *err, ...)
+void error_impl(const char *err, va_list params)
 {
-	va_list params;
-
-	va_start(params, err);
 	error_routine(err, params);
-	va_end(params);
-	return -1;
 }
 
 void warning(const char *warn, ...)
-- 
2.0.0.rc1.436.g03cb729

  reply	other threads:[~2014-05-06 16:06 UTC|newest]

Thread overview: 26+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2014-05-04  6:12 [PATCH 0/3] Fix a ton of compiler warnings Felipe Contreras
2014-05-04  6:12 ` [PATCH 1/3] Revert "make error()'s constant return value more visible" Felipe Contreras
2014-05-05  5:49   ` Jeff King [this message]
2014-05-05  5:45     ` Felipe Contreras
2014-05-05  6:02       ` Jeff King
2014-05-05  6:14         ` Felipe Contreras
2014-05-05  6:29           ` Jeff King
2014-05-05  7:30             ` Felipe Contreras
2014-05-05 21:29               ` Jeff King
2014-05-06 15:14                 ` [PATCH 1/2] inline constant return from error() function Jeff King
2014-05-06 22:29                   ` Junio C Hamano
2014-05-07  3:02                     ` Jeff King
2014-05-11 17:22                       ` Junio C Hamano
2014-05-12 18:13                         ` Jeff King
2014-05-11  7:13                     ` Felipe Contreras
2014-05-12 18:44                   ` Jonathan Nieder
2014-05-12 18:56                     ` Jeff King
2014-05-06 15:17                 ` [PATCH 2/2] let clang use the constant-return error() macro Jeff King
2014-05-04  6:12 ` [PATCH 2/3] Revert "silence some -Wuninitialized false positives" Felipe Contreras
2014-05-04  6:12 ` [PATCH 3/3] Silence a bunch of format-zero-length warnings Felipe Contreras
2014-05-04 19:01   ` brian m. carlson
2014-05-04 20:13     ` Felipe Contreras
2014-05-05  5:21     ` Jeff King
2014-05-07 18:19       ` Junio C Hamano
2014-05-07 20:05         ` Heiko Voigt
2014-05-07 20:31           ` Junio C Hamano

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20140505054901.GA19331@sigill.intra.peff.net \
    --to=peff@peff.net \
    --cc=felipe.contreras@gmail.com \
    --cc=git@vger.kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).