diff mbox

Make __ASSUME_UTIMES hppa-specific

Message ID Pine.LNX.4.64.1406251318060.27516@digraph.polyomino.org.uk
State Committed
Headers show

Commit Message

Joseph Myers June 25, 2014, 1:19 p.m. UTC
This patch makes __ASSUME_UTIMES hppa-specific, removing mentions of
the macro from architecture-independent code and code for other
architectures.  (All other architectures either have the utimes
syscall in all relevant kernel versions, or use the asm-generic
interface so only have utimensat and won't get the utimes syscall.)  A
similar approach is used to that used for futimesat for MicroBlaze: if
the kernel is recent enough that the utimes syscall can be assumed to
be present, use the implementation in terms of the utimes syscall, and
otherwise use the linux/generic implementation in terms of utimensat.

Tested x86_64 that the disassembly of installed shared libraries is
unchanged by the patch.  Not tested for hppa.

2014-06-25  Joseph Myers  <joseph@codesourcery.com>

	* sysdeps/unix/sysv/linux/kernel-features.h (__ASSUME_UTIMES): Do
	not define.
	* sysdeps/unix/sysv/linux/utimes.c: Do not include
	<kernel-features.h>.
	(__utimes) [__NR_utimes]: Make code unconditional.
	(__utimes) [!__ASSUME_UTIMES]: Remove conditional code.
	* sysdeps/unix/sysv/linux/aarch64/kernel-features.h
	(__ASSUME_UTIMES): Do not undefine.
	* sysdeps/unix/sysv/linux/tile/kernel-features.h
	(__ASSUME_UTIMES): Likewise.
	* sysdeps/unix/sysv/linux/hppa/kernel-features.h
	(__ASSUME_UTIMES): Define for [__LINUX_KERNEL_VERSION >= 0x030e00]
	instead of undefining for [__LINUX_KERNEL_VERSION < 0x030e00].
	* sysdeps/unix/sysv/linux/hppa/utimes.c: New file.

Comments

Joseph Myers June 30, 2014, 3:22 p.m. UTC | #1
Ping (last ping for this patch for 2.20).  This patch 
<https://sourceware.org/ml/libc-alpha/2014-06/msg00758.html> is pending 
review.
Joseph Myers July 17, 2014, 2:39 p.m. UTC | #2
Ping^2.  This patch 
<https://sourceware.org/ml/libc-alpha/2014-06/msg00758.html> is still 
pending review.
Mike Frysinger Aug. 4, 2014, 2:54 p.m. UTC | #3
On Wed 25 Jun 2014 13:19:48 Joseph S. Myers wrote:
> This patch makes __ASSUME_UTIMES hppa-specific, removing mentions of
> the macro from architecture-independent code and code for other
> architectures.  (All other architectures either have the utimes
> syscall in all relevant kernel versions, or use the asm-generic
> interface so only have utimensat and won't get the utimes syscall.)  A
> similar approach is used to that used for futimesat for MicroBlaze: if
> the kernel is recent enough that the utimes syscall can be assumed to
> be present, use the implementation in terms of the utimes syscall, and
> otherwise use the linux/generic implementation in terms of utimensat.
> 
> Tested x86_64 that the disassembly of installed shared libraries is
> unchanged by the patch.  Not tested for hppa.

looks sane to me, but i'm guessing you're looking for Carlos to sign off
-mike
Joseph Myers Sept. 8, 2014, 3:35 p.m. UTC | #4
Ping^3.  This patch 
<https://sourceware.org/ml/libc-alpha/2014-06/msg00758.html> is still 
pending review for hppa.
Joseph Myers Sept. 15, 2014, 5:24 p.m. UTC | #5
Ping^4.  This patch 
<https://sourceware.org/ml/libc-alpha/2014-06/msg00758.html> is still 
pending review for hppa.
Joseph Myers Sept. 22, 2014, 3:02 p.m. UTC | #6
Ping^5.  This patch 
<https://sourceware.org/ml/libc-alpha/2014-06/msg00758.html> is still 
pending review for hppa.
Joseph Myers Sept. 29, 2014, 2:51 p.m. UTC | #7
Ping^6.  This patch 
<https://sourceware.org/ml/libc-alpha/2014-06/msg00758.html> is still 
pending review for hppa.
Joseph Myers Oct. 6, 2014, 10 a.m. UTC | #8
Ping^7.  This patch 
<https://sourceware.org/ml/libc-alpha/2014-06/msg00758.html> is still 
pending review for hppa.
Joseph Myers Oct. 13, 2014, 11:33 a.m. UTC | #9
Ping^8.  This patch 
<https://sourceware.org/ml/libc-alpha/2014-06/msg00758.html> is still 
pending review for hppa.
Joseph Myers Oct. 20, 2014, 3:27 p.m. UTC | #10
Ping^9.  This patch 
<https://sourceware.org/ml/libc-alpha/2014-06/msg00758.html> is still 
pending review for hppa.
Joseph Myers Oct. 27, 2014, 12:27 p.m. UTC | #11
Ping^10.  This patch 
<https://sourceware.org/ml/libc-alpha/2014-06/msg00758.html> is still 
pending review for hppa.
Joseph Myers Nov. 3, 2014, 6:15 p.m. UTC | #12
Ping^11.  This patch 
<https://sourceware.org/ml/libc-alpha/2014-06/msg00758.html> is still 
pending review for hppa.
Joseph Myers Nov. 10, 2014, 1:54 p.m. UTC | #13
Ping^12.  This patch 
<https://sourceware.org/ml/libc-alpha/2014-06/msg00758.html> is still 
pending review for hppa.
Joseph Myers Nov. 18, 2014, 3:14 p.m. UTC | #14
Ping^13.  This patch 
<https://sourceware.org/ml/libc-alpha/2014-06/msg00758.html> is still 
pending review for hppa.
Joseph Myers Nov. 24, 2014, 4:48 p.m. UTC | #15
Ping^14.  This patch 
<https://sourceware.org/ml/libc-alpha/2014-06/msg00758.html> is still 
pending review for hppa.
Joseph Myers Dec. 1, 2014, 6:15 p.m. UTC | #16
Ping^15.  This patch 
<https://sourceware.org/ml/libc-alpha/2014-06/msg00758.html> is still 
pending review for hppa.
Joseph Myers Dec. 8, 2014, 4:14 p.m. UTC | #17
On Mon, 1 Dec 2014, Joseph Myers wrote:

Ping^16.  This patch 
<https://sourceware.org/ml/libc-alpha/2014-06/msg00758.html> is still 
pending review for hppa.
Joseph Myers Dec. 16, 2014, 3:34 p.m. UTC | #18
Ping^17.  This patch 
<https://sourceware.org/ml/libc-alpha/2014-06/msg00758.html> is still 
pending review for hppa.
Joseph Myers Dec. 22, 2014, 1:08 p.m. UTC | #19
Ping^18.  This patch 
<https://sourceware.org/ml/libc-alpha/2014-06/msg00758.html> is still 
pending review for hppa.
Kyle McMartin Dec. 22, 2014, 9:30 p.m. UTC | #20
On Mon, Dec 22, 2014 at 01:08:37PM +0000, Joseph Myers wrote:
> Ping^18.  This patch 
> <https://sourceware.org/ml/libc-alpha/2014-06/msg00758.html> is still 
> pending review for hppa.
> 

for what good it is, your patch looks fine... though i question the
utility of even bothering to wire up utimes() when asm-generic/unistd.h
using architectures are never going to have it... it seems hppa'd be
better off sticking to their guns and let arm64 deal with it, since
that'll have more clout as a non-boat-anchor architecture.

--kyle
Carlos O'Donell Dec. 23, 2014, 3:11 p.m. UTC | #21
On 12/22/2014 04:30 PM, Kyle McMartin wrote:
> On Mon, Dec 22, 2014 at 01:08:37PM +0000, Joseph Myers wrote:
>> Ping^18.  This patch 
>> <https://sourceware.org/ml/libc-alpha/2014-06/msg00758.html> is still 
>> pending review for hppa.
>>
> 
> for what good it is, your patch looks fine... though i question the
> utility of even bothering to wire up utimes() when asm-generic/unistd.h
> using architectures are never going to have it... it seems hppa'd be
> better off sticking to their guns and let arm64 deal with it, since
> that'll have more clout as a non-boat-anchor architecture.

Thanks for the review. I kept meaning to get to this, but to be honest
I'm just barely getting through the -Werror fixes for hppa (and even
then had to hack a set of new fans to boot my box).

Cheers,
Carlos.
diff mbox

Patch

diff --git a/sysdeps/unix/sysv/linux/aarch64/kernel-features.h b/sysdeps/unix/sysv/linux/aarch64/kernel-features.h
index bd94fe1..b9963e3 100644
--- a/sysdeps/unix/sysv/linux/aarch64/kernel-features.h
+++ b/sysdeps/unix/sysv/linux/aarch64/kernel-features.h
@@ -29,6 +29,3 @@ 
 #define __ASSUME_SENDMMSG_SYSCALL       1
 
 #include_next <kernel-features.h>
-
-/* asm-generic architectures do not have the utimes syscall.  */
-#undef __ASSUME_UTIMES
diff --git a/sysdeps/unix/sysv/linux/hppa/kernel-features.h b/sysdeps/unix/sysv/linux/hppa/kernel-features.h
index 25c3e36..cf0c574 100644
--- a/sysdeps/unix/sysv/linux/hppa/kernel-features.h
+++ b/sysdeps/unix/sysv/linux/hppa/kernel-features.h
@@ -32,9 +32,9 @@ 
 # define __ASSUME_SENDMMSG_SYSCALL	1
 #endif
 
-#include_next <kernel-features.h>
-
-/* hppa did not get the utimes syscall until 3.14.  */
-#if __LINUX_KERNEL_VERSION < 0x030e00
-# undef __ASSUME_UTIMES
+/* Support for the utimes syscall was added in 3.14.  */
+#if __LINUX_KERNEL_VERSION >= 0x030e00
+# define __ASSUME_UTIMES		1
 #endif
+
+#include_next <kernel-features.h>
diff --git a/sysdeps/unix/sysv/linux/hppa/utimes.c b/sysdeps/unix/sysv/linux/hppa/utimes.c
new file mode 100644
index 0000000..f62dd05
--- /dev/null
+++ b/sysdeps/unix/sysv/linux/hppa/utimes.c
@@ -0,0 +1,30 @@ 
+/* Implement utimes for hppa.
+   Copyright (C) 2014 Free Software Foundation, Inc.
+   This file is part of the GNU C Library.
+
+   The GNU C Library is free software; you can redistribute it and/or
+   modify it under the terms of the GNU Lesser General Public
+   License as published by the Free Software Foundation; either
+   version 2.1 of the License, or (at your option) any later version.
+
+   The GNU C Library is distributed in the hope that it will be useful,
+   but WITHOUT ANY WARRANTY; without even the implied warranty of
+   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+   Lesser General Public License for more details.
+
+   You should have received a copy of the GNU Lesser General Public
+   License along with the GNU C Library; if not, see
+   <http://www.gnu.org/licenses/>.  */
+
+/* hppa has the utimensat syscall in all supported kernel versions but
+   gained the utimes syscall later, so use the linux-generic
+   implementation of utimes in terms of the utimensat syscall unless
+   the utimes syscall is known to be available.  */
+
+#include <kernel-features.h>
+
+#ifdef __ASSUME_UTIMES
+# include <sysdeps/unix/sysv/linux/utimes.c>
+#else
+# include <sysdeps/unix/sysv/linux/generic/utimes.c>
+#endif
diff --git a/sysdeps/unix/sysv/linux/kernel-features.h b/sysdeps/unix/sysv/linux/kernel-features.h
index 3c8ea50..2de5775 100644
--- a/sysdeps/unix/sysv/linux/kernel-features.h
+++ b/sysdeps/unix/sysv/linux/kernel-features.h
@@ -72,12 +72,6 @@ 
 /* The statfs64 syscalls are available in 2.5.74 (but not for alpha).  */
 #define __ASSUME_STATFS64	1
 
-/* The utimes syscall has been available for some architectures
-   forever.  For x86 it was introduced after 2.5.75, for x86-64,
-   ppc, and ppc64 it was introduced in 2.6.0-test3, for s390 it was
-   introduced in 2.6.21-rc5.  */
-#define __ASSUME_UTIMES	1
-
 /* pselect/ppoll were introduced just after 2.6.16-rc1.  On x86_64 and
    SH this appeared first in 2.6.19-rc1, on ia64 in 2.6.22-rc1.  */
 #define __ASSUME_PSELECT	1
diff --git a/sysdeps/unix/sysv/linux/tile/kernel-features.h b/sysdeps/unix/sysv/linux/tile/kernel-features.h
index 78aacac..3dc0dfc 100644
--- a/sysdeps/unix/sysv/linux/tile/kernel-features.h
+++ b/sysdeps/unix/sysv/linux/tile/kernel-features.h
@@ -28,9 +28,6 @@ 
 
 #include_next <kernel-features.h>
 
-/* asm-generic architectures do not have the utimes syscall.  */
-#undef __ASSUME_UTIMES
-
 /* Define this if your 32-bit syscall API requires 64-bit register
    pairs to start with an even-number register.  */
 #define __ASSUME_ALIGNED_REGISTER_PAIRS	1
diff --git a/sysdeps/unix/sysv/linux/utimes.c b/sysdeps/unix/sysv/linux/utimes.c
index 2bcebeb..5bea069 100644
--- a/sysdeps/unix/sysv/linux/utimes.c
+++ b/sysdeps/unix/sysv/linux/utimes.c
@@ -20,40 +20,16 @@ 
 #include <utime.h>
 #include <sys/time.h>
 #include <sysdep.h>
-#include <kernel-features.h>
 
 
+/* Consider moving to syscalls.list.  */
+
 /* Change the access time of FILE to TVP[0] and
    the modification time of FILE to TVP[1].  */
 int
 __utimes (const char *file, const struct timeval tvp[2])
 {
-#ifdef __NR_utimes
-  int result = INLINE_SYSCALL (utimes, 2, file, tvp);
-# ifndef __ASSUME_UTIMES
-  if (result != -1 || errno != ENOSYS)
-# endif
-    return result;
-#endif
-
-  /* The utimes() syscall does not exist or is not available in the
-     used kernel.  Use utime().  For this we have to convert to the
-     data format utime() expects.  */
-#ifndef __ASSUME_UTIMES
-  struct utimbuf buf;
-  struct utimbuf *times;
-
-  if (tvp != NULL)
-    {
-      times = &buf;
-      buf.actime = tvp[0].tv_sec + tvp[0].tv_usec / 1000000;
-      buf.modtime = tvp[1].tv_sec + tvp[1].tv_usec / 1000000;
-    }
-  else
-    times = NULL;
-
-  return INLINE_SYSCALL (utime, 2, file, times);
-#endif
+  return INLINE_SYSCALL (utimes, 2, file, tvp);
 }
 
 weak_alias (__utimes, utimes)