This is the mail archive of the
libc-alpha@sourceware.org
mailing list for the glibc project.
Re: Fix x86_64 memrchr namespace (bug 17719)
- From: "H.J. Lu" <hjl dot tools at gmail dot com>
- To: Joseph Myers <joseph at codesourcery dot com>
- Cc: GNU C Library <libc-alpha at sourceware dot org>
- Date: Tue, 16 Dec 2014 10:25:28 -0800
- Subject: Re: Fix x86_64 memrchr namespace (bug 17719)
- Authentication-results: sourceware.org; auth=none
- References: <alpine dot DEB dot 2 dot 10 dot 1412161815590 dot 15637 at digraph dot polyomino dot org dot uk>
On Tue, Dec 16, 2014 at 10:16 AM, Joseph Myers <joseph@codesourcery.com> wrote:
> On x86_64, memrchr (not a standard function) is defined as a strong
> symbol, instead of a weak alias of __memrchr as on other
> architectures. This results in linknamespace test failures from the
> use of __memrchr from dirname. (Not a conformance issue because of
> the mem* reservation, but contrary to glibc conventions.) This patch
> makes x86_64 follow other architectures by defining memrchr as a weak
> alias.
>
> Tested for x86_64 (testsuite, and that disassembly of installed shared
> libraries is unchanged by the patch).
>
> 2014-12-16 Joseph Myers <joseph@codesourcery.com>
>
> [BZ #17719]
> * sysdeps/x86_64/memrchr.S (memrchr): Rename to __memrchr and
> define as weak alias of __memrchr.
> (__memrchr): Do not define as strong alias of memrchr.
> * conform/Makefile (test-xfail-XPG4/libgen.h/linknamespace):
> Remove variable.
> (test-xfail-UNIX98/libgen.h/linknamespace): Likewise.
> (test-xfail-XOPEN2K/libgen.h/linknamespace): Likewise.
> (test-xfail-XOPEN2K8/libgen.h/linknamespace): Likewise.
>
> diff --git a/conform/Makefile b/conform/Makefile
> index e0412de..45a17fb 100644
> --- a/conform/Makefile
> +++ b/conform/Makefile
> @@ -359,7 +359,6 @@ test-xfail-XPG4/ctype.h/linknamespace = yes
> test-xfail-XPG4/fmtmsg.h/linknamespace = yes
> test-xfail-XPG4/fnmatch.h/linknamespace = yes
> test-xfail-XPG4/glob.h/linknamespace = yes
> -test-xfail-XPG4/libgen.h/linknamespace = yes
> test-xfail-XPG4/netdb.h/linknamespace = yes
> test-xfail-XPG4/regex.h/linknamespace = yes
> test-xfail-XPG4/search.h/linknamespace = yes
> @@ -383,7 +382,6 @@ test-xfail-UNIX98/ctype.h/linknamespace = yes
> test-xfail-UNIX98/fmtmsg.h/linknamespace = yes
> test-xfail-UNIX98/fnmatch.h/linknamespace = yes
> test-xfail-UNIX98/glob.h/linknamespace = yes
> -test-xfail-UNIX98/libgen.h/linknamespace = yes
> test-xfail-UNIX98/mqueue.h/linknamespace = yes
> test-xfail-UNIX98/netdb.h/linknamespace = yes
> test-xfail-UNIX98/regex.h/linknamespace = yes
> @@ -398,7 +396,6 @@ test-xfail-XOPEN2K/fcntl.h/linknamespace = yes
> test-xfail-XOPEN2K/fmtmsg.h/linknamespace = yes
> test-xfail-XOPEN2K/fnmatch.h/linknamespace = yes
> test-xfail-XOPEN2K/glob.h/linknamespace = yes
> -test-xfail-XOPEN2K/libgen.h/linknamespace = yes
> test-xfail-XOPEN2K/mqueue.h/linknamespace = yes
> test-xfail-XOPEN2K/netdb.h/linknamespace = yes
> test-xfail-XOPEN2K/regex.h/linknamespace = yes
> @@ -425,7 +422,6 @@ test-xfail-XOPEN2K8/dirent.h/linknamespace = yes
> test-xfail-XOPEN2K8/fcntl.h/linknamespace = yes
> test-xfail-XOPEN2K8/fmtmsg.h/linknamespace = yes
> test-xfail-XOPEN2K8/grp.h/linknamespace = yes
> -test-xfail-XOPEN2K8/libgen.h/linknamespace = yes
> test-xfail-XOPEN2K8/mqueue.h/linknamespace = yes
> test-xfail-XOPEN2K8/netdb.h/linknamespace = yes
> test-xfail-XOPEN2K8/pwd.h/linknamespace = yes
> diff --git a/sysdeps/x86_64/memrchr.S b/sysdeps/x86_64/memrchr.S
> index ff875f4..3b558b5 100644
> --- a/sysdeps/x86_64/memrchr.S
> +++ b/sysdeps/x86_64/memrchr.S
> @@ -21,7 +21,7 @@
> #include <sysdep.h>
>
> .text
> -ENTRY (memrchr)
> +ENTRY (__memrchr)
> movd %rsi, %xmm1
>
> sub $16, %rdx
> @@ -378,5 +378,5 @@ L(length_less16_part2_return):
> lea 16(%rax, %rdi), %rax
> ret
>
> -END (memrchr)
> -strong_alias (memrchr, __memrchr)
> +END (__memrchr)
> +weak_alias (__memrchr, memrchr)
>
> --
> Joseph S. Myers
> joseph@codesourcery.com
Looks good to me.
Thanks.
--
H.J.