public inbox for glibc-cvs@sourceware.org
help / color / mirror / Atom feed
From: John David Anglin <danglin@sourceware.org>
To: glibc-cvs@sourceware.org
Subject: [glibc] Revert "hppa: Drop 16-byte pthread lock alignment"
Date: Thu,  6 Jul 2023 15:48:42 +0000 (GMT)	[thread overview]
Message-ID: <20230706154842.5928D3858D33@sourceware.org> (raw)

https://sourceware.org/git/gitweb.cgi?p=glibc.git;h=500054974667be3153ed760152ea0153df33c3d0

commit 500054974667be3153ed760152ea0153df33c3d0
Author: John David Anglin <danglin@gcc.gnu.org>
Date:   Thu Jul 6 15:47:50 2023 +0000

    Revert "hppa: Drop 16-byte pthread lock alignment"
    
    This change reverts commits c4468cd3995b4236ea886901109b194641132b08
    and ab991a3d1b401ded6bd4f027352da8262b021a11.

Diff:
---
 sysdeps/hppa/nptl/bits/pthreadtypes-arch.h |  2 +-
 sysdeps/hppa/nptl/bits/struct_rwlock.h     | 10 ++--------
 2 files changed, 3 insertions(+), 9 deletions(-)

diff --git a/sysdeps/hppa/nptl/bits/pthreadtypes-arch.h b/sysdeps/hppa/nptl/bits/pthreadtypes-arch.h
index c1a46d66d0..999195c5b0 100644
--- a/sysdeps/hppa/nptl/bits/pthreadtypes-arch.h
+++ b/sysdeps/hppa/nptl/bits/pthreadtypes-arch.h
@@ -40,7 +40,7 @@
 #define __SIZEOF_PTHREAD_RWLOCK_T 64
 #define __SIZEOF_PTHREAD_RWLOCKATTR_T 8
 
-#define __LOCK_ALIGNMENT
+#define __LOCK_ALIGNMENT __attribute__ ((__aligned__(16)))
 #define __ONCE_ALIGNMENT
 
 #endif	/* bits/pthreadtypes.h */
diff --git a/sysdeps/hppa/nptl/bits/struct_rwlock.h b/sysdeps/hppa/nptl/bits/struct_rwlock.h
index f91460b7f8..e83b4aab52 100644
--- a/sysdeps/hppa/nptl/bits/struct_rwlock.h
+++ b/sysdeps/hppa/nptl/bits/struct_rwlock.h
@@ -25,14 +25,8 @@ struct __pthread_rwlock_arch_t
   /* In the old Linuxthreads pthread_rwlock_t, this is the
      start of the 4-word 16-byte aligned lock structure. The
      next four words are all set to 1 by the Linuxthreads
-     PTHREAD_RWLOCK_INITIALIZER. We ignore them in NPTL.
-
-     The 16-byte aligned lock structure is not needed for NPTL.
-     It causes some package builds to fail (e.g., protobuf) and
-     issues with dynamic memory allocation in C++ versions prior
-     to C++17. Removing it allows for more efficient allocators
-     and better compatibility with other architectures.  */
-  int __compat_padding[4];
+     PTHREAD_RWLOCK_INITIALIZER. We ignore them in NPTL.  */
+  int __compat_padding[4] __attribute__ ((__aligned__(16)));
   unsigned int __readers;
   unsigned int __writers;
   unsigned int __wrphase_futex;

                 reply	other threads:[~2023-07-06 15:48 UTC|newest]

Thread overview: [no followups] expand[flat|nested]  mbox.gz  Atom feed

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=20230706154842.5928D3858D33@sourceware.org \
    --to=danglin@sourceware.org \
    --cc=glibc-cvs@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).