head	1.3;
access;
symbols
	pkgsrc-2025Q3:1.1.0.2
	pkgsrc-2025Q3-base:1.1;
locks; strict;
comment	@// @;


1.3
date	2026.09.29.10.49.58;	author tnn;	state Exp;
branches;
next	1.2;
commitid	ztbPCqeqk1L5qvXG;

1.2
date	2025.11.15.12.17.12;	author ryoon;	state dead;
branches;
next	1.1;
commitid	SHodt4IOcVSG7EiG;

1.1
date	2025.08.11.18.06.32;	author pho;	state Exp;
branches;
next	;
commitid	bmUAUprLlDrX8l6G;


desc
@@


1.3
log
@openjdk21: reduce diffs w/ openjdk25, add Darwin and Linux targets
@
text
@$NetBSD$

Work around SIGBUS on macOS

--- src/hotspot/os_cpu/bsd_aarch64/os_bsd_aarch64.cpp.orig	2026-09-29 10:45:05.907015432 +0000
+++ src/hotspot/os_cpu/bsd_aarch64/os_bsd_aarch64.cpp
@@@@ -472,6 +472,9 @@@@ static inline void atomic_copy64(const v
 
 extern "C" {
   int SpinPause() {
+#if defined(__APPLE__)
+   return 0; // XXX stub broken; See JDK-8278241, JDK-8321371
+#endif
     using spin_wait_func_ptr_t = void (*)();
     spin_wait_func_ptr_t func = CAST_TO_FN_PTR(spin_wait_func_ptr_t, StubRoutines::aarch64::spin_wait());
     assert(func != nullptr, "StubRoutines::aarch64::spin_wait must not be null.");
@


1.2
log
@lang/openjdk21: Update to 1.21.0.9.10

Changelog:
Updated BSD port of JDK 21

Additional features include:

    Update to 21.0.8 GA
@
text
@d1 1
a1 1
$NetBSD: patch-src_hotspot_os__cpu_bsd__aarch64_os__bsd__aarch64.cpp,v 1.1 2025/08/11 18:06:32 pho Exp $
d3 1
a3 2
Workaround for the Hotspot VM dying on Apple Silicon chips. See
the patch to threadWXSetters.inline.hpp for details.
d5 1
a5 1
--- src/hotspot/os_cpu/bsd_aarch64/os_bsd_aarch64.cpp.orig	2025-04-19 18:18:28.000000000 +0000
d7 1
a7 11
@@@@ -694,6 +694,11 @@@@ bool PosixSignals::pd_hotspot_signal_han
     assert(sig == info->si_signo, "bad siginfo");
   }
 */
+
+  // Enable WXWrite: this function is called by the signal handler at arbitrary
+  // point of execution.
+  ThreadWXEnable wx(WXWrite, thread);
+
   // decide if this trap can be handled by a stub
   address stub = nullptr;
d9 8
@


1.1
log
@lang/openjdk21: Workaround for the VM dying on Apple Silicon chips

See the patch for threadWXSetters.inline.hpp for details. I rebuilt the
bootkit for aarch64 with new patches applied. This may cause some
performance regression on Cortex series, but as I stated in the patch
comment, relying on implementation details of a particular chip is
fundamentally unsound, and that's why it didn't work on Apple Silicon in
the first place.

But since the upstream has stated they are knowingly doing this, the chance
of these patches getting accepted is slim, if not zero.
@
text
@d1 1
a1 1
$NetBSD$
@