bug-gnulib@gnu.org mirror (unofficial)
 help / color / mirror / Atom feed
From: Paul Eggert <eggert@cs.ucla.edu>
To: Bruno Haible <bruno@clisp.org>
Cc: bug-gnulib@gnu.org,
	"libc-alpha@sourceware.org" <libc-alpha@sourceware.org>,
	Vaseeharan Vinayagamoorthy <Vaseeharan.Vinayagamoorthy@arm.com>,
	Adhemerval Zanella <adhemerval.zanella@linaro.org>
Subject: Re: [PATCH 1/5] posix: Sync regex code with gnulib
Date: Tue, 19 Jan 2021 18:55:51 -0800	[thread overview]
Message-ID: <05f5e913-f1d1-b2f9-beb7-f226588733c0@cs.ucla.edu> (raw)
In-Reply-To: <2108975.lN1YhbmyTB@omega>

[-- Attachment #1: Type: text/plain, Size: 1459 bytes --]

On 1/19/21 7:43 AM, Bruno Haible wrote:
> Adhemerval Zanella wrote:
>> -# if (__GNUC__ >= 7) || (__clang_major__ >= 10)
>> +# if (__GNUC__ >= 7) || (defined __clang_major__ &&__clang_major__ >= 10)
> I would write it as:
> 
> +# if (__GNUC__ >= 7) || (defined __clang__ && __clang_major__ >= 10)

This line should be used only if _LIBC is defined, so we can simplify it 
to just "#if __GNUC__ >= 7" and thus not worry about Clang.

I see that Gnulib wasn't consistent about this, so I installed the 
attached patch to Gnulib to fix the issue here and elsewhere. The idea 
is that the Gnulib regex_internal.h can be copied back to glibc, and 
that the other uses in Gnulib should be similar.

On 1/19/21 6:43 AM, Adhemerval Zanella wrote:

> Paul, this seemed to a common pattern scatter on multiple file in gnulib.
> Wouldn't be better to consolidate it on cdefs.h?

We could append something like the following to cdefs.h, and switch to 
__attribute_fallthrough__ for modules shared between the two systems. Is 
that something you'd like to pursue? (We should also sync Gnulib cdefs.h 
back to glibc of course.)

#if defined __STDC_VERSION__ && 201710L < __STDC_VERSION__
# define __attribute_fallthrough__ [[__fallthrough__]]
#elif __GNUC_PREREQ (7, 0) || __glibc_has_attribute (__fallthrough__)
# define __attribute_fallthrough__ __attribute__ ((__fallthrough__))
#else
# define __attribute_fallthrough__ ((void) 0)
#endif

[-- Attachment #2: 0001-fnmatch-regex-fts-FALLTHROUGH-consistency.patch --]
[-- Type: text/x-patch, Size: 3792 bytes --]

From 5c52f00c69f39fe86ec087654893087a83290ee7 Mon Sep 17 00:00:00 2001
From: Paul Eggert <eggert@cs.ucla.edu>
Date: Tue, 19 Jan 2021 18:35:30 -0800
Subject: [PATCH] fnmatch, regex, fts: FALLTHROUGH consistency
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Be more consistent about how FALLTHROUGH is defined.
For Gnulib, use attribute.h.  For glibc, use __GNUC__ >= 7.
Problem for glibc reported by Vaseeharan Vinayagamoorthy in:
https://sourceware.org/pipermail/libc-alpha/2021-January/121778.html
* lib/fnmatch.c (FALLTHROUGH) [_LIBC]:
* lib/regex_internal.h (FALLTHROUGH) [_LIBC]:
Don’t worry about Clang, as it’s not needed and provokes GCC.
* lib/fts.c (FALLTHROUGH) [!_LIBC]:
* lib/regex_internal.h (FALLTHROUGH) [!_LIBC]:
Rely on attribute.h for FALLTHROUGH
* modules/regex: Depend on attribute module.
---
 ChangeLog            | 15 +++++++++++++++
 lib/fnmatch.c        |  2 +-
 lib/fts.c            |  4 ++--
 lib/regex_internal.h |  6 ++++--
 modules/regex        |  1 +
 5 files changed, 23 insertions(+), 5 deletions(-)

diff --git a/ChangeLog b/ChangeLog
index 269577caa..290fa1b4c 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,3 +1,18 @@
+2021-01-19  Paul Eggert  <eggert@cs.ucla.edu>
+
+	fnmatch, regex, fts: FALLTHROUGH consistency
+	Be more consistent about how FALLTHROUGH is defined.
+	For Gnulib, use attribute.h.  For glibc, use __GNUC__ >= 7.
+	Problem for glibc reported by Vaseeharan Vinayagamoorthy in:
+	https://sourceware.org/pipermail/libc-alpha/2021-January/121778.html
+	* lib/fnmatch.c (FALLTHROUGH) [_LIBC]:
+	* lib/regex_internal.h (FALLTHROUGH) [_LIBC]:
+	Don’t worry about Clang, as it’s not needed and provokes GCC.
+	* lib/fts.c (FALLTHROUGH) [!_LIBC]:
+	* lib/regex_internal.h (FALLTHROUGH) [!_LIBC]:
+	Rely on attribute.h for FALLTHROUGH
+	* modules/regex: Depend on attribute module.
+
 2021-01-19  KO Myung-Hun  <komh78@gmail.com>
 
 	spawn-pipe: Fix SIGSEGV on OS/2 kLIBC.
diff --git a/lib/fnmatch.c b/lib/fnmatch.c
index 5896812c9..b8a71f164 100644
--- a/lib/fnmatch.c
+++ b/lib/fnmatch.c
@@ -64,7 +64,7 @@ extern int fnmatch (const char *pattern, const char *string, int flags);
 #endif
 
 #ifdef _LIBC
-# if (__GNUC__ >= 7) || (__clang_major__ >= 10)
+# if __GNUC__ >= 7
 #  define FALLTHROUGH __attribute__ ((__fallthrough__))
 # else
 #  define FALLTHROUGH ((void) 0)
diff --git a/lib/fts.c b/lib/fts.c
index 8a9b5ed96..e6603f40e 100644
--- a/lib/fts.c
+++ b/lib/fts.c
@@ -200,8 +200,8 @@ enum Fts_stat
     while (false)
 #endif
 
-#ifndef FALLTHROUGH
-# if (__GNUC__ >= 7) || (__clang_major__ >= 10)
+#ifdef _LIBC
+# if __GNUC__ >= 7
 #  define FALLTHROUGH __attribute__ ((__fallthrough__))
 # else
 #  define FALLTHROUGH ((void) 0)
diff --git a/lib/regex_internal.h b/lib/regex_internal.h
index b4f91d9ec..3fa2bf1aa 100644
--- a/lib/regex_internal.h
+++ b/lib/regex_internal.h
@@ -830,12 +830,14 @@ re_string_elem_size_at (const re_string_t *pstr, Idx idx)
 }
 #endif /* RE_ENABLE_I18N */
 
-#ifndef FALLTHROUGH
-# if (__GNUC__ >= 7) || (__clang_major__ >= 10)
+#ifdef _LIBC
+# if __GNUC__ >= 7
 #  define FALLTHROUGH __attribute__ ((__fallthrough__))
 # else
 #  define FALLTHROUGH ((void) 0)
 # endif
+#else
+# include "attribute.h"
 #endif
 
 #endif /*  _REGEX_INTERNAL_H */
diff --git a/modules/regex b/modules/regex
index 20cbe375a..a32c46e18 100644
--- a/modules/regex
+++ b/modules/regex
@@ -16,6 +16,7 @@ Depends-on:
 c99
 extensions
 ssize_t
+attribute       [test $ac_use_included_regex = yes]
 btowc           [test $ac_use_included_regex = yes]
 builtin-expect  [test $ac_use_included_regex = yes]
 dynarray        [test $ac_use_included_regex = yes]
-- 
2.27.0


  reply	other threads:[~2021-01-20  2:56 UTC|newest]

Thread overview: 22+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-12-30 20:15 [PATCH 1/5] posix: Sync regex code with gnulib Adhemerval Zanella
2020-12-30 20:15 ` [PATCH 2/5] posix: Sync glob " Adhemerval Zanella
2020-12-31 21:47   ` Paul Eggert
2020-12-30 20:15 ` [PATCH 3/5] Sync intprops.h " Adhemerval Zanella
2020-12-31 21:47   ` Paul Eggert
2020-12-30 20:15 ` [PATCH 4/5] Sync flexmember.h " Adhemerval Zanella
2020-12-31 21:48   ` Paul Eggert
2020-12-30 20:15 ` [PATCH 5/5] posix: Sync fnmatch " Adhemerval Zanella
2020-12-31 21:54   ` Paul Eggert
2020-12-31 21:37 ` [PATCH 1/5] posix: Sync regex code " Paul Eggert
2021-01-19 14:16 ` Vaseeharan Vinayagamoorthy
2021-01-19 14:43   ` Adhemerval Zanella
2021-01-19 15:43     ` Bruno Haible
2021-01-20  2:55       ` Paul Eggert [this message]
2021-01-20 11:27         ` Adhemerval Zanella
2021-01-20 15:32           ` Vaseeharan Vinayagamoorthy
2021-01-20 16:05             ` Adhemerval Zanella
2021-01-20 17:46           ` Paul Eggert
2021-01-19 16:52     ` Florian Weimer
2021-01-19 17:11       ` Adhemerval Zanella
2021-01-19 17:16         ` Florian Weimer
2021-01-19 17:18           ` Adhemerval Zanella

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

  List information: https://lists.gnu.org/mailman/listinfo/bug-gnulib

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

  git send-email \
    --in-reply-to=05f5e913-f1d1-b2f9-beb7-f226588733c0@cs.ucla.edu \
    --to=eggert@cs.ucla.edu \
    --cc=Vaseeharan.Vinayagamoorthy@arm.com \
    --cc=adhemerval.zanella@linaro.org \
    --cc=bruno@clisp.org \
    --cc=bug-gnulib@gnu.org \
    --cc=libc-alpha@sourceware.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 read-only IMAP folder(s) and NNTP newsgroup(s).