[PATCH 00/30] RFC: x32 support

[PATCH 00/30] RFC: x32 support

[H. Peter Anvin]

This is an initial RFC patchset for x32 support.  This is largely
complete and should be able to boot an x32 userspace.

The patch authors are assigned to myself and H. J. in a somewhat
haphazard fashion; the kernel side of x32 was very much a
collaborative effort, although H. J. ended up doing most of the latter
coding due to my being occupied with kernel.org in late 2011.

This patchset applies on top of tip:core/types which in turn is based
on v3.3-rc3; there may be some trivially resolved conflicts with -rc4.
I am planning to push this out as tip:x86/x32 after the RFC.

Controversial issue in this patchset:

There are a lot of ABIs that expose "long" or "unsigned long"; the
traditional way to deal with that is compat, but we don't want to use
that in most cases for x32.  As a result, I have introduced
__kernel_[u]long_t as an additional type in posix_types.h.

This type largely duplicates __kernel_[s]size_t, but the latter is
defined as [unsigned] int on a lot of platforms, so changing types
over to __kernel_[s]size_t would have changed the type, if not the
size, on a lot of platforms.  I have personally no strong opinion on
which way is preferrable.

The use of a flag to signal x32 rather than using a completely new set
of system call numbers.  This was discussed back in August.

Incompleteness:

There are a number of (especially) ioctl paths which don't yet handle
a 64-bit time_t in the compat path.  This is necessary because of
Linus' (quite correct) dictat that new ABIs should use a 64-bit
time_t, but this also means this is the first 32-bit ABI with a 64-bit
time_t, and some problems are expected to crop up.

The header-exported ABI is not correct everywhere yet.


The total diffstat is attached.

 arch/x86/Kconfig                       |   21 +-
 arch/x86/ia32/ia32_signal.c            |   12 +-
 arch/x86/ia32/sys_ia32.c               |   40 ---
 arch/x86/include/asm/Kbuild            |    2 +
 arch/x86/include/asm/compat.h          |   39 ++-
 arch/x86/include/asm/elf.h             |   29 ++-
 arch/x86/include/asm/ia32.h            |    9 +
 arch/x86/include/asm/posix_types.h     |    4 +-
 arch/x86/include/asm/posix_types_x32.h |   19 +
 arch/x86/include/asm/processor.h       |   10 +-
 arch/x86/include/asm/ptrace.h          |    1 -
 arch/x86/include/asm/sigcontext.h      |   57 ++--
 arch/x86/include/asm/sigframe.h        |   13 +
 arch/x86/include/asm/sighandling.h     |   24 ++
 arch/x86/include/asm/sys_ia32.h        |    7 +-
 arch/x86/include/asm/syscall.h         |    5 +-
 arch/x86/include/asm/thread_info.h     |    6 +-
 arch/x86/include/asm/unistd.h          |   15 +-
 arch/x86/kernel/asm-offsets_64.c       |    6 +
 arch/x86/kernel/cpu/perf_event.c       |    4 +-
 arch/x86/kernel/entry_64.S             |   44 +++
 arch/x86/kernel/process_64.c           |   25 +-
 arch/x86/kernel/signal.c               |  138 +++++++-
 arch/x86/kernel/sys_x86_64.c           |    6 +-
 arch/x86/kernel/syscall_64.c           |    8 +
 arch/x86/oprofile/backtrace.c          |    2 +-
 arch/x86/syscalls/Makefile             |   19 +-
 arch/x86/syscalls/syscall_32.tbl       |    2 +-
 arch/x86/syscalls/syscall_64.tbl       |  579 +++++++++++++++++---------------
 arch/x86/um/sys_call_table_64.c        |    3 +
 arch/x86/um/user-offsets.c             |    2 +
 arch/x86/vdso/.gitignore               |    2 +
 arch/x86/vdso/Makefile                 |   46 +++-
 arch/x86/vdso/vdso32-setup.c           |    6 +
 arch/x86/vdso/vdsox32.S                |   22 ++
 arch/x86/vdso/vdsox32.lds.S            |   32 ++
 arch/x86/vdso/vma.c                    |   78 ++++-
 drivers/char/lp.c                      |   20 +-
 drivers/input/input-compat.c           |    4 +-
 drivers/input/input-compat.h           |    2 +-
 fs/binfmt_elf.c                        |   24 ++-
 fs/compat.c                            |   26 ++-
 include/asm-generic/posix_types.h      |   23 +-
 include/linux/Kbuild                   |    1 +
 include/linux/aio_abi.h                |    2 +-
 include/linux/compat.h                 |    4 +
 include/linux/kernel.h                 |   21 +-
 include/linux/sysinfo.h                |   24 ++
 net/bluetooth/hci_sock.c               |    3 +-
 49 files changed, 1027 insertions(+), 464 deletions(-)
 
It is probably worth noting that this work directly spawned the
x86/syscall and core/types cleanups:

 33 files changed, 1075 insertions(+), 1978 deletions(-)
 22 files changed, 190 insertions(+), 1588 deletions(-)

[PATCH 01/30] x86: Factor out TIF_IA32 from 32-bit address space

[H. Peter Anvin]

From: "H. Peter Anvin" <hpa@zytor.com>

Factor out IA32 (compatibility instruction set) from 32-bit address
space in the thread_info flags; this is a precondition patch for x32
support.

Originally-by: H. J. Lu <hjl.tools@gmail.com>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Link: http://lkml.kernel.org/n/tip-4pr1xnnksprt7t0h3w5fw4rv@git.kernel.org
---
 arch/x86/include/asm/elf.h         |    4 ++--
 arch/x86/include/asm/processor.h   |    4 ++--
 arch/x86/include/asm/thread_info.h |    4 +++-
 arch/x86/kernel/process_64.c       |    2 ++
 arch/x86/kernel/sys_x86_64.c       |    6 +++---
 arch/x86/oprofile/backtrace.c      |    2 +-
 6 files changed, 13 insertions(+), 9 deletions(-)

diff --git a/arch/x86/include/asm/elf.h b/arch/x86/include/asm/elf.h
index 5f962df..410fa6a 100644
--- a/arch/x86/include/asm/elf.h
+++ b/arch/x86/include/asm/elf.h
@@ -287,7 +287,7 @@ do {									\
 #define VDSO_HIGH_BASE		0xffffe000U /* CONFIG_COMPAT_VDSO address */
 
 /* 1GB for 64bit, 8MB for 32bit */
-#define STACK_RND_MASK (test_thread_flag(TIF_IA32) ? 0x7ff : 0x3fffff)
+#define STACK_RND_MASK (test_thread_flag(TIF_ADDR32) ? 0x7ff : 0x3fffff)
 
 #define ARCH_DLINFO							\
 do {									\
@@ -330,7 +330,7 @@ static inline int mmap_is_ia32(void)
 	return 1;
 #endif
 #ifdef CONFIG_IA32_EMULATION
-	if (test_thread_flag(TIF_IA32))
+	if (test_thread_flag(TIF_ADDR32))
 		return 1;
 #endif
 	return 0;
diff --git a/arch/x86/include/asm/processor.h b/arch/x86/include/asm/processor.h
index aa9088c..9f748b5 100644
--- a/arch/x86/include/asm/processor.h
+++ b/arch/x86/include/asm/processor.h
@@ -924,9 +924,9 @@ extern unsigned long thread_saved_pc(struct task_struct *tsk);
 #define IA32_PAGE_OFFSET	((current->personality & ADDR_LIMIT_3GB) ? \
 					0xc0000000 : 0xFFFFe000)
 
-#define TASK_SIZE		(test_thread_flag(TIF_IA32) ? \
+#define TASK_SIZE		(test_thread_flag(TIF_ADDR32) ? \
 					IA32_PAGE_OFFSET : TASK_SIZE_MAX)
-#define TASK_SIZE_OF(child)	((test_tsk_thread_flag(child, TIF_IA32)) ? \
+#define TASK_SIZE_OF(child)	((test_tsk_thread_flag(child, TIF_ADDR32)) ? \
 					IA32_PAGE_OFFSET : TASK_SIZE_MAX)
 
 #define STACK_TOP		TASK_SIZE
diff --git a/arch/x86/include/asm/thread_info.h b/arch/x86/include/asm/thread_info.h
index bc817cd..d1803a4 100644
--- a/arch/x86/include/asm/thread_info.h
+++ b/arch/x86/include/asm/thread_info.h
@@ -86,7 +86,7 @@ struct thread_info {
 #define TIF_MCE_NOTIFY		10	/* notify userspace of an MCE */
 #define TIF_USER_RETURN_NOTIFY	11	/* notify kernel of userspace return */
 #define TIF_NOTSC		16	/* TSC is not accessible in userland */
-#define TIF_IA32		17	/* 32bit process */
+#define TIF_IA32		17	/* IA32 compatibility process */
 #define TIF_FORK		18	/* ret_from_fork */
 #define TIF_MEMDIE		20	/* is terminating due to OOM killer */
 #define TIF_DEBUG		21	/* uses debug registers */
@@ -95,6 +95,7 @@ struct thread_info {
 #define TIF_BLOCKSTEP		25	/* set when we want DEBUGCTLMSR_BTF */
 #define TIF_LAZY_MMU_UPDATES	27	/* task is updating the mmu lazily */
 #define TIF_SYSCALL_TRACEPOINT	28	/* syscall tracepoint instrumentation */
+#define TIF_ADDR32		29	/* 32-bit address space on 64 bits */
 
 #define _TIF_SYSCALL_TRACE	(1 << TIF_SYSCALL_TRACE)
 #define _TIF_NOTIFY_RESUME	(1 << TIF_NOTIFY_RESUME)
@@ -116,6 +117,7 @@ struct thread_info {
 #define _TIF_BLOCKSTEP		(1 << TIF_BLOCKSTEP)
 #define _TIF_LAZY_MMU_UPDATES	(1 << TIF_LAZY_MMU_UPDATES)
 #define _TIF_SYSCALL_TRACEPOINT	(1 << TIF_SYSCALL_TRACEPOINT)
+#define _TIF_ADDR32		(1 << TIF_ADDR32)
 
 /* work to do in syscall_trace_enter() */
 #define _TIF_WORK_SYSCALL_ENTRY	\
diff --git a/arch/x86/kernel/process_64.c b/arch/x86/kernel/process_64.c
index 9b9fe4a..0e900d0 100644
--- a/arch/x86/kernel/process_64.c
+++ b/arch/x86/kernel/process_64.c
@@ -508,6 +508,7 @@ void set_personality_64bit(void)
 
 	/* Make sure to be in 64bit mode */
 	clear_thread_flag(TIF_IA32);
+	clear_thread_flag(TIF_ADDR32);
 
 	/* Ensure the corresponding mm is not marked. */
 	if (current->mm)
@@ -526,6 +527,7 @@ void set_personality_ia32(void)
 
 	/* Make sure to be in 32bit mode */
 	set_thread_flag(TIF_IA32);
+	set_thread_flag(TIF_ADDR32);
 	current->personality |= force_personality32;
 
 	/* Mark the associated mm as containing 32-bit tasks. */
diff --git a/arch/x86/kernel/sys_x86_64.c b/arch/x86/kernel/sys_x86_64.c
index 0514890..f921df8 100644
--- a/arch/x86/kernel/sys_x86_64.c
+++ b/arch/x86/kernel/sys_x86_64.c
@@ -98,7 +98,7 @@ out:
 static void find_start_end(unsigned long flags, unsigned long *begin,
 			   unsigned long *end)
 {
-	if (!test_thread_flag(TIF_IA32) && (flags & MAP_32BIT)) {
+	if (!test_thread_flag(TIF_ADDR32) && (flags & MAP_32BIT)) {
 		unsigned long new_begin;
 		/* This is usually used needed to map code in small
 		   model, so it needs to be in the first 31bit. Limit
@@ -144,7 +144,7 @@ arch_get_unmapped_area(struct file *filp, unsigned long addr,
 		    (!vma || addr + len <= vma->vm_start))
 			return addr;
 	}
-	if (((flags & MAP_32BIT) || test_thread_flag(TIF_IA32))
+	if (((flags & MAP_32BIT) || test_thread_flag(TIF_ADDR32))
 	    && len <= mm->cached_hole_size) {
 		mm->cached_hole_size = 0;
 		mm->free_area_cache = begin;
@@ -205,7 +205,7 @@ arch_get_unmapped_area_topdown(struct file *filp, const unsigned long addr0,
 		return addr;
 
 	/* for MAP_32BIT mappings we force the legact mmap base */
-	if (!test_thread_flag(TIF_IA32) && (flags & MAP_32BIT))
+	if (!test_thread_flag(TIF_ADDR32) && (flags & MAP_32BIT))
 		goto bottomup;
 
 	/* requesting a specific address */
diff --git a/arch/x86/oprofile/backtrace.c b/arch/x86/oprofile/backtrace.c
index bff89df..d6aa6e8 100644
--- a/arch/x86/oprofile/backtrace.c
+++ b/arch/x86/oprofile/backtrace.c
@@ -67,7 +67,7 @@ x86_backtrace_32(struct pt_regs * const regs, unsigned int depth)
 {
 	struct stack_frame_ia32 *head;
 
-	/* User process is 32-bit */
+	/* User process is IA32 */
 	if (!current || !test_thread_flag(TIF_IA32))
 		return 0;
 
-- 
1.7.6.5

[PATCH 02/30] x86-64: Use explicit sizes in sigcontext.h, prepare for x32

[H. Peter Anvin]

From: "H. Peter Anvin" <hpa@zytor.com>

Use explicit sizes (__u64) instead of implicit sizes (unsigned long)
in the definition for sigcontext.h; this will allow this structure to
be shared between the x86-64 native ABI and the x32 ABI.

Originally-by: H. J. Lu <hjl.tools@gmail.com>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Link: http://lkml.kernel.org/n/tip-4pr1xnnksprt7t0h3w5fw4rv@git.kernel.org
---
 arch/x86/include/asm/sigcontext.h |   57 +++++++++++++++++++-----------------
 1 files changed, 30 insertions(+), 27 deletions(-)

diff --git a/arch/x86/include/asm/sigcontext.h b/arch/x86/include/asm/sigcontext.h
index 04459d2..4a08538 100644
--- a/arch/x86/include/asm/sigcontext.h
+++ b/arch/x86/include/asm/sigcontext.h
@@ -230,34 +230,37 @@ struct sigcontext {
  * User-space might still rely on the old definition:
  */
 struct sigcontext {
-	unsigned long r8;
-	unsigned long r9;
-	unsigned long r10;
-	unsigned long r11;
-	unsigned long r12;
-	unsigned long r13;
-	unsigned long r14;
-	unsigned long r15;
-	unsigned long rdi;
-	unsigned long rsi;
-	unsigned long rbp;
-	unsigned long rbx;
-	unsigned long rdx;
-	unsigned long rax;
-	unsigned long rcx;
-	unsigned long rsp;
-	unsigned long rip;
-	unsigned long eflags;		/* RFLAGS */
-	unsigned short cs;
-	unsigned short gs;
-	unsigned short fs;
-	unsigned short __pad0;
-	unsigned long err;
-	unsigned long trapno;
-	unsigned long oldmask;
-	unsigned long cr2;
+	__u64 r8;
+	__u64 r9;
+	__u64 r10;
+	__u64 r11;
+	__u64 r12;
+	__u64 r13;
+	__u64 r14;
+	__u64 r15;
+	__u64 rdi;
+	__u64 rsi;
+	__u64 rbp;
+	__u64 rbx;
+	__u64 rdx;
+	__u64 rax;
+	__u64 rcx;
+	__u64 rsp;
+	__u64 rip;
+	__u64 eflags;		/* RFLAGS */
+	__u16 cs;
+	__u16 gs;
+	__u16 fs;
+	__u16 __pad0;
+	__u64 err;
+	__u64 trapno;
+	__u64 oldmask;
+	__u64 cr2;
 	struct _fpstate __user *fpstate;	/* zero when no FPU context */
-	unsigned long reserved1[8];
+#ifndef __LP64__
+	__u32 __fpstate_pad;
+#endif
+	__u64 reserved1[8];
 };
 #endif /* !__KERNEL__ */
 
-- 
1.7.6.5

[PATCH 03/30] sysinfo: Move struct sysinfo to a separate header file

[H. Peter Anvin]

From: "H. Peter Anvin" <hpa@zytor.com>

struct sysinfo is just about the only thing exported to userspace from
<linux/kernel.h>, so move it into a separate header file with a
residual #include in <linux/kernel.h>.

Originally-by: H. J. Lu <hjl.tools@gmail.com>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Link: http://lkml.kernel.org/n/tip-4pr1xnnksprt7t0h3w5fw4rv@git.kernel.org
---
 include/linux/Kbuild    |    1 +
 include/linux/kernel.h  |   21 ++-------------------
 include/linux/sysinfo.h |   22 ++++++++++++++++++++++
 3 files changed, 25 insertions(+), 19 deletions(-)
 create mode 100644 include/linux/sysinfo.h

diff --git a/include/linux/Kbuild b/include/linux/Kbuild
index c94e717..8446086 100644
--- a/include/linux/Kbuild
+++ b/include/linux/Kbuild
@@ -355,6 +355,7 @@ header-y += suspend_ioctls.h
 header-y += swab.h
 header-y += synclink.h
 header-y += sysctl.h
+header-y += sysinfo.h
 header-y += taskstats.h
 header-y += tcp.h
 header-y += telephony.h
diff --git a/include/linux/kernel.h b/include/linux/kernel.h
index e834342..dc6a50f 100644
--- a/include/linux/kernel.h
+++ b/include/linux/kernel.h
@@ -1,6 +1,8 @@
 #ifndef _LINUX_KERNEL_H
 #define _LINUX_KERNEL_H
 
+#include <linux/sysinfo.h>
+
 /*
  * 'kernel.h' contains some often-used function prototypes etc
  */
@@ -745,27 +747,8 @@ extern int __build_bug_on_failed;
 # define REBUILD_DUE_TO_FTRACE_MCOUNT_RECORD
 #endif
 
-struct sysinfo;
 extern int do_sysinfo(struct sysinfo *info);
 
 #endif /* __KERNEL__ */
 
-#define SI_LOAD_SHIFT	16
-struct sysinfo {
-	long uptime;			/* Seconds since boot */
-	unsigned long loads[3];		/* 1, 5, and 15 minute load averages */
-	unsigned long totalram;		/* Total usable main memory size */
-	unsigned long freeram;		/* Available memory size */
-	unsigned long sharedram;	/* Amount of shared memory */
-	unsigned long bufferram;	/* Memory used by buffers */
-	unsigned long totalswap;	/* Total swap space size */
-	unsigned long freeswap;		/* swap space still available */
-	unsigned short procs;		/* Number of current processes */
-	unsigned short pad;		/* explicit padding for m68k */
-	unsigned long totalhigh;	/* Total high memory size */
-	unsigned long freehigh;		/* Available high memory size */
-	unsigned int mem_unit;		/* Memory unit size in bytes */
-	char _f[20-2*sizeof(long)-sizeof(int)];	/* Padding: libc5 uses this.. */
-};
-
 #endif
diff --git a/include/linux/sysinfo.h b/include/linux/sysinfo.h
new file mode 100644
index 0000000..ec4fc22
--- /dev/null
+++ b/include/linux/sysinfo.h
@@ -0,0 +1,22 @@
+#ifndef _LINUX_SYSINFO_H
+#define _LINUX_SYSINFO_H
+
+#define SI_LOAD_SHIFT	16
+struct sysinfo {
+	long uptime;			/* Seconds since boot */
+	unsigned long loads[3];		/* 1, 5, and 15 minute load averages */
+	unsigned long totalram;		/* Total usable main memory size */
+	unsigned long freeram;		/* Available memory size */
+	unsigned long sharedram;	/* Amount of shared memory */
+	unsigned long bufferram;	/* Memory used by buffers */
+	unsigned long totalswap;	/* Total swap space size */
+	unsigned long freeswap;		/* swap space still available */
+	unsigned short procs;		/* Number of current processes */
+	unsigned short pad;		/* explicit padding for m68k */
+	unsigned long totalhigh;	/* Total high memory size */
+	unsigned long freehigh;		/* Available high memory size */
+	unsigned int mem_unit;		/* Memory unit size in bytes */
+	char _f[20-2*sizeof(long)-sizeof(int)];	/* Padding: libc5 uses this.. */
+};
+
+#endif /* _LINUX_SYSINFO_H */
-- 
1.7.6.5

[PATCH 03/30] sysinfo: Move struct sysinfo to a separate header file

[H. Peter Anvin]

From: "H. Peter Anvin" <hpa@zytor.com>

struct sysinfo is just about the only thing exported to userspace from
<linux/kernel.h>, so move it into a separate header file with a
residual #include in <linux/kernel.h>.

Originally-by: H. J. Lu <hjl.tools@gmail.com>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Link: http://lkml.kernel.org/n/tip-4pr1xnnksprt7t0h3w5fw4rv@git.kernel.org
---
 include/linux/Kbuild    |    1 +
 include/linux/kernel.h  |   21 ++-------------------
 include/linux/sysinfo.h |   22 ++++++++++++++++++++++
 3 files changed, 25 insertions(+), 19 deletions(-)
 create mode 100644 include/linux/sysinfo.h

diff --git a/include/linux/Kbuild b/include/linux/Kbuild
index c94e717..8446086 100644
--- a/include/linux/Kbuild
+++ b/include/linux/Kbuild
@@ -355,6 +355,7 @@ header-y += suspend_ioctls.h
 header-y += swab.h
 header-y += synclink.h
 header-y += sysctl.h
+header-y += sysinfo.h
 header-y += taskstats.h
 header-y += tcp.h
 header-y += telephony.h
diff --git a/include/linux/kernel.h b/include/linux/kernel.h
index e834342..dc6a50f 100644
--- a/include/linux/kernel.h
+++ b/include/linux/kernel.h
@@ -1,6 +1,8 @@
 #ifndef _LINUX_KERNEL_H
 #define _LINUX_KERNEL_H
 
+#include <linux/sysinfo.h>
+
 /*
  * 'kernel.h' contains some often-used function prototypes etc
  */
@@ -745,27 +747,8 @@ extern int __build_bug_on_failed;
 # define REBUILD_DUE_TO_FTRACE_MCOUNT_RECORD
 #endif
 
-struct sysinfo;
 extern int do_sysinfo(struct sysinfo *info);
 
 #endif /* __KERNEL__ */
 
-#define SI_LOAD_SHIFT	16
-struct sysinfo {
-	long uptime;			/* Seconds since boot */
-	unsigned long loads[3];		/* 1, 5, and 15 minute load averages */
-	unsigned long totalram;		/* Total usable main memory size */
-	unsigned long freeram;		/* Available memory size */
-	unsigned long sharedram;	/* Amount of shared memory */
-	unsigned long bufferram;	/* Memory used by buffers */
-	unsigned long totalswap;	/* Total swap space size */
-	unsigned long freeswap;		/* swap space still available */
-	unsigned short procs;		/* Number of current processes */
-	unsigned short pad;		/* explicit padding for m68k */
-	unsigned long totalhigh;	/* Total high memory size */
-	unsigned long freehigh;		/* Available high memory size */
-	unsigned int mem_unit;		/* Memory unit size in bytes */
-	char _f[20-2*sizeof(long)-sizeof(int)];	/* Padding: libc5 uses this.. */
-};
-
 #endif
diff --git a/include/linux/sysinfo.h b/include/linux/sysinfo.h
new file mode 100644
index 0000000..ec4fc22
--- /dev/null
+++ b/include/linux/sysinfo.h
@@ -0,0 +1,22 @@
+#ifndef _LINUX_SYSINFO_H
+#define _LINUX_SYSINFO_H
+
+#define SI_LOAD_SHIFT	16
+struct sysinfo {
+	long uptime;			/* Seconds since boot */
+	unsigned long loads[3];		/* 1, 5, and 15 minute load averages */
+	unsigned long totalram;		/* Total usable main memory size */
+	unsigned long freeram;		/* Available memory size */
+	unsigned long sharedram;	/* Amount of shared memory */
+	unsigned long bufferram;	/* Memory used by buffers */
+	unsigned long totalswap;	/* Total swap space size */
+	unsigned long freeswap;		/* swap space still available */
+	unsigned short procs;		/* Number of current processes */
+	unsigned short pad;		/* explicit padding for m68k */
+	unsigned long totalhigh;	/* Total high memory size */
+	unsigned long freehigh;		/* Available high memory size */
+	unsigned int mem_unit;		/* Memory unit size in bytes */
+	char _f[20-2*sizeof(long)-sizeof(int)];	/* Padding: libc5 uses this.. */
+};
+
+#endif /* _LINUX_SYSINFO_H */
-- 
1.7.6.5

[PATCH 04/30] posix_types: Introduce __kernel_[u]long_t

[H. Peter Anvin]

From: "H. Peter Anvin" <hpa@zytor.com>

Introduce __kernel_[u]long_t, which allows an ABI to override all
defaults of type [unsigned] long.

This enables x32 and potentially other 32-bit userspace on 64-bit
kernel ABIs.

Signed-off-by: H. Peter Anvin <hpa@zytor.com>
---
 include/asm-generic/posix_types.h |   23 ++++++++++++++---------
 1 files changed, 14 insertions(+), 9 deletions(-)

diff --git a/include/asm-generic/posix_types.h b/include/asm-generic/posix_types.h
index e294fe6..91d44bd 100644
--- a/include/asm-generic/posix_types.h
+++ b/include/asm-generic/posix_types.h
@@ -10,8 +10,13 @@
  * architectures, so that you can override them.
  */
 
+#ifndef __kernel_long_t
+typedef long		__kernel_long_t;
+typedef unsigned long	__kernel_ulong_t;
+#endif
+
 #ifndef __kernel_ino_t
-typedef unsigned long	__kernel_ino_t;
+typedef __kernel_ulong_t __kernel_ino_t;
 #endif
 
 #ifndef __kernel_mode_t
@@ -19,7 +24,7 @@ typedef unsigned int	__kernel_mode_t;
 #endif
 
 #ifndef __kernel_nlink_t
-typedef unsigned long	__kernel_nlink_t;
+typedef __kernel_ulong_t __kernel_nlink_t;
 #endif
 
 #ifndef __kernel_pid_t
@@ -36,7 +41,7 @@ typedef unsigned int	__kernel_gid_t;
 #endif
 
 #ifndef __kernel_suseconds_t
-typedef long		__kernel_suseconds_t;
+typedef __kernel_long_t		__kernel_suseconds_t;
 #endif
 
 #ifndef __kernel_daddr_t
@@ -67,9 +72,9 @@ typedef unsigned int	__kernel_size_t;
 typedef int		__kernel_ssize_t;
 typedef int		__kernel_ptrdiff_t;
 #else
-typedef unsigned long	__kernel_size_t;
-typedef long		__kernel_ssize_t;
-typedef long		__kernel_ptrdiff_t;
+typedef __kernel_ulong_t __kernel_size_t;
+typedef __kernel_long_t	__kernel_ssize_t;
+typedef __kernel_long_t	__kernel_ptrdiff_t;
 #endif
 #endif
 
@@ -82,10 +87,10 @@ typedef struct {
 /*
  * anything below here should be completely generic
  */
-typedef long		__kernel_off_t;
+typedef __kernel_long_t	__kernel_off_t;
 typedef long long	__kernel_loff_t;
-typedef long		__kernel_time_t;
-typedef long		__kernel_clock_t;
+typedef __kernel_long_t	__kernel_time_t;
+typedef __kernel_long_t	__kernel_clock_t;
 typedef int		__kernel_timer_t;
 typedef int		__kernel_clockid_t;
 typedef char *		__kernel_caddr_t;
-- 
1.7.6.5

[PATCH 05/30] x32: Create posix_types_x32.h

[H. Peter Anvin]

From: "H. Peter Anvin" <hpa@zytor.com>

This is the same as the 64-bit posix_types.h, except that
__kernel_[u]long_t is defined to be [unsigned] long long and therefore
64 bits.

Signed-off-by: H. Peter Anvin <hpa@zytor.com>
---
 arch/x86/include/asm/Kbuild            |    1 +
 arch/x86/include/asm/posix_types.h     |    4 +++-
 arch/x86/include/asm/posix_types_x32.h |   19 +++++++++++++++++++
 3 files changed, 23 insertions(+), 1 deletions(-)
 create mode 100644 arch/x86/include/asm/posix_types_x32.h

diff --git a/arch/x86/include/asm/Kbuild b/arch/x86/include/asm/Kbuild
index b57e6a4..986954f 100644
--- a/arch/x86/include/asm/Kbuild
+++ b/arch/x86/include/asm/Kbuild
@@ -14,6 +14,7 @@ header-y += msr.h
 header-y += mtrr.h
 header-y += posix_types_32.h
 header-y += posix_types_64.h
+header-y += posix_types_x32.h
 header-y += prctl.h
 header-y += processor-flags.h
 header-y += ptrace-abi.h
diff --git a/arch/x86/include/asm/posix_types.h b/arch/x86/include/asm/posix_types.h
index bb7133d..3427b77 100644
--- a/arch/x86/include/asm/posix_types.h
+++ b/arch/x86/include/asm/posix_types.h
@@ -7,7 +7,9 @@
 #else
 # ifdef __i386__
 #  include "posix_types_32.h"
-# else
+# elif defined(__LP64__)
 #  include "posix_types_64.h"
+# else
+#  include "posix_types_x32.h"
 # endif
 #endif
diff --git a/arch/x86/include/asm/posix_types_x32.h b/arch/x86/include/asm/posix_types_x32.h
new file mode 100644
index 0000000..85f9bda
--- /dev/null
+++ b/arch/x86/include/asm/posix_types_x32.h
@@ -0,0 +1,19 @@
+#ifndef _ASM_X86_POSIX_TYPES_X32_H
+#define _ASM_X86_POSIX_TYPES_X32_H
+
+/*
+ * This file is only used by user-level software, so you need to
+ * be a little careful about namespace pollution etc.  Also, we cannot
+ * assume GCC is being used.
+ *
+ * These types should generally match the ones used by the 64-bit kernel,
+ *
+ */
+
+typedef long long __kernel_long_t;
+typedef unsigned long long __kernel_ulong_t;
+#define __kernel_long_t __kernel_long_t
+
+#include <asm/posix_types_64.h>
+
+#endif /* _ASM_X86_POSIX_TYPES_X32_H */
-- 
1.7.6.5

[PATCH 06/30] sysinfo: Use explicit types in <linux/sysinfo.h>

[H. Peter Anvin]

From: "H. Peter Anvin" <hpa@zytor.com>

Change <linux/sysinfo.h> to use explicitly sized types.  Replace
long/unsigned long with __kernel_[u]long_t so that a non-legacy 32-bit
ABI running on a 64-bit kernel can export those as 64-bit types.

Originally-by: H. J. Lu <hjl.tools@gmail.com>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
---
 include/linux/sysinfo.h |   30 ++++++++++++++++--------------
 1 files changed, 16 insertions(+), 14 deletions(-)

diff --git a/include/linux/sysinfo.h b/include/linux/sysinfo.h
index ec4fc22..934335a 100644
--- a/include/linux/sysinfo.h
+++ b/include/linux/sysinfo.h
@@ -1,22 +1,24 @@
 #ifndef _LINUX_SYSINFO_H
 #define _LINUX_SYSINFO_H
 
+#include <linux/types.h>
+
 #define SI_LOAD_SHIFT	16
 struct sysinfo {
-	long uptime;			/* Seconds since boot */
-	unsigned long loads[3];		/* 1, 5, and 15 minute load averages */
-	unsigned long totalram;		/* Total usable main memory size */
-	unsigned long freeram;		/* Available memory size */
-	unsigned long sharedram;	/* Amount of shared memory */
-	unsigned long bufferram;	/* Memory used by buffers */
-	unsigned long totalswap;	/* Total swap space size */
-	unsigned long freeswap;		/* swap space still available */
-	unsigned short procs;		/* Number of current processes */
-	unsigned short pad;		/* explicit padding for m68k */
-	unsigned long totalhigh;	/* Total high memory size */
-	unsigned long freehigh;		/* Available high memory size */
-	unsigned int mem_unit;		/* Memory unit size in bytes */
-	char _f[20-2*sizeof(long)-sizeof(int)];	/* Padding: libc5 uses this.. */
+	__kernel_long_t uptime;		/* Seconds since boot */
+	__kernel_ulong_t loads[3];	/* 1, 5, and 15 minute load averages */
+	__kernel_ulong_t totalram;	/* Total usable main memory size */
+	__kernel_ulong_t freeram;	/* Available memory size */
+	__kernel_ulong_t sharedram;	/* Amount of shared memory */
+	__kernel_ulong_t bufferram;	/* Memory used by buffers */
+	__kernel_ulong_t totalswap;	/* Total swap space size */
+	__kernel_ulong_t freeswap;	/* swap space still available */
+	__u16 procs;		   	/* Number of current processes */
+	__u16 pad;		   	/* Explicit padding for m68k */
+	__kernel_ulong_t totalhigh;	/* Total high memory size */
+	__kernel_ulong_t freehigh;	/* Available high memory size */
+	__u32 mem_unit;			/* Memory unit size in bytes */
+	char _f[20-2*sizeof(__kernel_ulong_t)-sizeof(__u32)];	/* Padding: libc5 uses this.. */
 };
 
 #endif /* _LINUX_SYSINFO_H */
-- 
1.7.6.5

[PATCH 07/30] compat: Introduce COMPAT_USE_64BIT_TIME

[H. Peter Anvin]

From: "H. J. Lu" <hjl.tools@gmail.com>

Allow a compatibility ABI to use a 64-bit time_t and 64-bit members in
struct timeval and struct timespec to avoid the Y2038 problem.

This will be used for the x32 ABI.

Signed-off-by: H. Peter Anvin <hpa@zytor.com>
---
 include/linux/compat.h |    4 ++++
 1 files changed, 4 insertions(+), 0 deletions(-)

diff --git a/include/linux/compat.h b/include/linux/compat.h
index 41c9f65..1be91c0 100644
--- a/include/linux/compat.h
+++ b/include/linux/compat.h
@@ -19,6 +19,10 @@
 #include <asm/siginfo.h>
 #include <asm/signal.h>
 
+#ifndef COMPAT_USE_64BIT_TIME
+#define COMPAT_USE_64BIT_TIME 0
+#endif
+
 #define compat_jiffies_to_clock_t(x)	\
 		(((unsigned long)(x) * COMPAT_USER_HZ) / HZ)
 
-- 
1.7.6.5

[PATCH 07/30] compat: Introduce COMPAT_USE_64BIT_TIME

[H. Peter Anvin]

From: "H. J. Lu" <hjl.tools@gmail.com>

Allow a compatibility ABI to use a 64-bit time_t and 64-bit members in
struct timeval and struct timespec to avoid the Y2038 problem.

This will be used for the x32 ABI.

Signed-off-by: H. Peter Anvin <hpa@zytor.com>
---
 include/linux/compat.h |    4 ++++
 1 files changed, 4 insertions(+), 0 deletions(-)

diff --git a/include/linux/compat.h b/include/linux/compat.h
index 41c9f65..1be91c0 100644
--- a/include/linux/compat.h
+++ b/include/linux/compat.h
@@ -19,6 +19,10 @@
 #include <asm/siginfo.h>
 #include <asm/signal.h>
 
+#ifndef COMPAT_USE_64BIT_TIME
+#define COMPAT_USE_64BIT_TIME 0
+#endif
+
 #define compat_jiffies_to_clock_t(x)	\
 		(((unsigned long)(x) * COMPAT_USER_HZ) / HZ)
 
-- 
1.7.6.5

[PATCH 08/30] compat: Use COMPAT_USE_64BIT_TIME in the lp driver

[H. Peter Anvin]

From: "H. J. Lu" <hjl.tools@gmail.com>

Enable the lp driver to be used with a compat ABI with 64-bit time.

Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Greg Kroah-Hartman <greg@kroah.com>
---
 drivers/char/lp.c |   20 ++++++++++++++------
 1 files changed, 14 insertions(+), 6 deletions(-)

diff --git a/drivers/char/lp.c b/drivers/char/lp.c
index f434856..a111ff2 100644
--- a/drivers/char/lp.c
+++ b/drivers/char/lp.c
@@ -706,18 +706,26 @@ static long lp_compat_ioctl(struct file *file, unsigned int cmd,
 {
 	unsigned int minor;
 	struct timeval par_timeout;
-	struct compat_timeval __user *tc;
 	int ret;
 
 	minor = iminor(file->f_path.dentry->d_inode);
 	mutex_lock(&lp_mutex);
 	switch (cmd) {
 	case LPSETTIMEOUT:
-		tc = compat_ptr(arg);
-		if (get_user(par_timeout.tv_sec, &tc->tv_sec) ||
-		    get_user(par_timeout.tv_usec, &tc->tv_usec)) {
-			ret = -EFAULT;
-			break;
+		if (COMPAT_USE_64BIT_TIME) {
+			if (copy_from_user(&par_timeout, (void __user *)arg,
+						sizeof (struct timeval))) {
+				ret = -EFAULT;
+				break;
+			}
+		} else {
+			struct compat_timeval __user *tc;
+			tc = compat_ptr(arg);
+			if (get_user(par_timeout.tv_sec, &tc->tv_sec) ||
+			    get_user(par_timeout.tv_usec, &tc->tv_usec)) {
+				ret = -EFAULT;
+				break;
+			}
 		}
 		ret = lp_set_timeout(minor, &par_timeout);
 		break;
-- 
1.7.6.5

[PATCH 09/30] compat: Use COMPAT_USE_64BIT_TIME in the input subsystem

[H. Peter Anvin]

From: "H. J. Lu" <hjl.tools@gmail.com>

Enable the input system to be used with a compat ABI with 64-bit time.

Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Cc: Dmitry Torokhov <dmitry.torokhov@gmail.com>
---
 drivers/input/input-compat.c |    4 ++--
 drivers/input/input-compat.h |    2 +-
 2 files changed, 3 insertions(+), 3 deletions(-)

diff --git a/drivers/input/input-compat.c b/drivers/input/input-compat.c
index e46a867..64ca711 100644
--- a/drivers/input/input-compat.c
+++ b/drivers/input/input-compat.c
@@ -17,7 +17,7 @@
 int input_event_from_user(const char __user *buffer,
 			  struct input_event *event)
 {
-	if (INPUT_COMPAT_TEST) {
+	if (INPUT_COMPAT_TEST && !COMPAT_USE_64BIT_TIME) {
 		struct input_event_compat compat_event;
 
 		if (copy_from_user(&compat_event, buffer,
@@ -41,7 +41,7 @@ int input_event_from_user(const char __user *buffer,
 int input_event_to_user(char __user *buffer,
 			const struct input_event *event)
 {
-	if (INPUT_COMPAT_TEST) {
+	if (INPUT_COMPAT_TEST && !COMPAT_USE_64BIT_TIME) {
 		struct input_event_compat compat_event;
 
 		compat_event.time.tv_sec = event->time.tv_sec;
diff --git a/drivers/input/input-compat.h b/drivers/input/input-compat.h
index 22be27b..148f66f 100644
--- a/drivers/input/input-compat.h
+++ b/drivers/input/input-compat.h
@@ -67,7 +67,7 @@ struct ff_effect_compat {
 
 static inline size_t input_event_size(void)
 {
-	return INPUT_COMPAT_TEST ?
+	return (INPUT_COMPAT_TEST && !COMPAT_USE_64BIT_TIME) ?
 		sizeof(struct input_event_compat) : sizeof(struct input_event);
 }
 
-- 
1.7.6.5

[PATCH 10/30] compat: Use COMPAT_USE_64BIT_TIME in the Bluetooth subsystem

[H. Peter Anvin]

From: "H. J. Lu" <hjl.tools@gmail.com>

Enable the Bluetooth subsystem to be used with a compat ABI with
64-bit time.

Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Cc: Marcel Holtmann <marcel@holtmann.org>
Cc: Gustavo F. Padovan <padovan@profusion.mobi>
Cc: David S. Miller <davem@davemloft.net>
---
 net/bluetooth/hci_sock.c |    3 ++-
 1 files changed, 2 insertions(+), 1 deletions(-)

diff --git a/net/bluetooth/hci_sock.c b/net/bluetooth/hci_sock.c
index 0dcc962..b2eb2b9 100644
--- a/net/bluetooth/hci_sock.c
+++ b/net/bluetooth/hci_sock.c
@@ -418,7 +418,8 @@ static inline void hci_sock_cmsg(struct sock *sk, struct msghdr *msg, struct sk_
 		data = &tv;
 		len = sizeof(tv);
 #ifdef CONFIG_COMPAT
-		if (msg->msg_flags & MSG_CMSG_COMPAT) {
+		if (!COMPAT_USE_64BIT_TIME &&
+		    (msg->msg_flags & MSG_CMSG_COMPAT)) {
 			ctv.tv_sec = tv.tv_sec;
 			ctv.tv_usec = tv.tv_usec;
 			data = &ctv;
-- 
1.7.6.5

[PATCH 11/30] aio: Use __kernel_ulong_t to define aio_context_t

[H. Peter Anvin]

From: "H. Peter Anvin" <hpa@zytor.com>

Rather than using "unsigned long" which is ABI-dependent, use
__kernel_ulong_t to define the externally visible type aio_context_t.

Note: the change in this form will cause unsigned long/unsigned int
differences on existing ABIs.  If that is unacceptable we may have to
define a new type.

Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Cc: Benjamin LaHaise <bcrl@kvack.org>
---
 include/linux/aio_abi.h |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/include/linux/aio_abi.h b/include/linux/aio_abi.h
index 2c87316..86fa7a7 100644
--- a/include/linux/aio_abi.h
+++ b/include/linux/aio_abi.h
@@ -30,7 +30,7 @@
 #include <linux/types.h>
 #include <asm/byteorder.h>
 
-typedef unsigned long	aio_context_t;
+typedef __kernel_ulong_t aio_context_t;
 
 enum {
 	IOCB_CMD_PREAD = 0,
-- 
1.7.6.5

[PATCH 12/30] compat: Create compat_sys_p{read,write}v64

[H. Peter Anvin]

From: "H. J. Lu" <hjl.tools@gmail.com>

For 32-bit ABIs which have real 64-bit registers, we don't want to
break the position argument into two.  However, we still need compat
support to deal with 32-bit pointers, so we can't just use
sys_p{read,write} directly.

Signed-off-by: H. Peter Anvin <hpa@zytor.com>
---
 fs/compat.c |   26 ++++++++++++++++++++------
 1 files changed, 20 insertions(+), 6 deletions(-)

diff --git a/fs/compat.c b/fs/compat.c
index fa9d721..83d751c 100644
--- a/fs/compat.c
+++ b/fs/compat.c
@@ -1177,10 +1177,9 @@ compat_sys_readv(unsigned long fd, const struct compat_iovec __user *vec,
 }
 
 asmlinkage ssize_t
-compat_sys_preadv(unsigned long fd, const struct compat_iovec __user *vec,
-		  unsigned long vlen, u32 pos_low, u32 pos_high)
+compat_sys_preadv64(unsigned long fd, const struct compat_iovec __user *vec,
+		    unsigned long vlen, loff_t pos)
 {
-	loff_t pos = ((loff_t)pos_high << 32) | pos_low;
 	struct file *file;
 	int fput_needed;
 	ssize_t ret;
@@ -1197,6 +1196,14 @@ compat_sys_preadv(unsigned long fd, const struct compat_iovec __user *vec,
 	return ret;
 }
 
+asmlinkage ssize_t
+compat_sys_preadv(unsigned long fd, const struct compat_iovec __user *vec,
+		  unsigned long vlen, u32 pos_low, u32 pos_high)
+{
+	loff_t pos = ((loff_t)pos_high << 32) | pos_low;
+	return compat_sys_preadv64(fd, vec, vlen, pos);
+}
+
 static size_t compat_writev(struct file *file,
 			    const struct compat_iovec __user *vec,
 			    unsigned long vlen, loff_t *pos)
@@ -1236,10 +1243,9 @@ compat_sys_writev(unsigned long fd, const struct compat_iovec __user *vec,
 }
 
 asmlinkage ssize_t
-compat_sys_pwritev(unsigned long fd, const struct compat_iovec __user *vec,
-		   unsigned long vlen, u32 pos_low, u32 pos_high)
+compat_sys_pwritev64(unsigned long fd, const struct compat_iovec __user *vec,
+		     unsigned long vlen, loff_t pos)
 {
-	loff_t pos = ((loff_t)pos_high << 32) | pos_low;
 	struct file *file;
 	int fput_needed;
 	ssize_t ret;
@@ -1256,6 +1262,14 @@ compat_sys_pwritev(unsigned long fd, const struct compat_iovec __user *vec,
 	return ret;
 }
 
+asmlinkage ssize_t
+compat_sys_pwritev(unsigned long fd, const struct compat_iovec __user *vec,
+		   unsigned long vlen, u32 pos_low, u32 pos_high)
+{
+	loff_t pos = ((loff_t)pos_high << 32) | pos_low;
+	return compat_sys_pwritev64(fd, vec, vlen, pos);
+}
+
 asmlinkage long
 compat_sys_vmsplice(int fd, const struct compat_iovec __user *iov32,
 		    unsigned int nr_segs, unsigned int flags)
-- 
1.7.6.5

[PATCH 13/30] elf: Allow core dump-related fields to be overridden

[H. Peter Anvin]

From: "H. J. Lu" <hjl.tools@gmail.com>

Allow some core dump-related fields to be overridden.  This allows
core dumps to work correctly for x32.

Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Cc: Alexander Viro <viro@zeniv.linux.org.uk>
Cc: Roland McGrath <roland@redhat.com>
Cc: Oleg Nesterov <oleg@redhat.com>
---
 fs/binfmt_elf.c |   24 ++++++++++++++++++++----
 1 files changed, 20 insertions(+), 4 deletions(-)

diff --git a/fs/binfmt_elf.c b/fs/binfmt_elf.c
index bcb884e..43ba478 100644
--- a/fs/binfmt_elf.c
+++ b/fs/binfmt_elf.c
@@ -1390,6 +1390,22 @@ static void do_thread_regset_writeback(struct task_struct *task,
 		regset->writeback(task, regset, 1);
 }
 
+#ifndef PR_REG_SIZE
+#define PR_REG_SIZE(S) sizeof(S)
+#endif
+
+#ifndef PRSTATUS_SIZE
+#define PRSTATUS_SIZE(S) sizeof(S)
+#endif
+
+#ifndef PR_REG_PTR
+#define PR_REG_PTR(S) (&((S)->pr_reg))
+#endif
+
+#ifndef SET_PR_FPVALID
+#define SET_PR_FPVALID(S, V) ((S)->pr_fpvalid = (V))
+#endif
+
 static int fill_thread_core_info(struct elf_thread_core_info *t,
 				 const struct user_regset_view *view,
 				 long signr, size_t *total)
@@ -1404,11 +1420,11 @@ static int fill_thread_core_info(struct elf_thread_core_info *t,
 	 */
 	fill_prstatus(&t->prstatus, t->task, signr);
 	(void) view->regsets[0].get(t->task, &view->regsets[0],
-				    0, sizeof(t->prstatus.pr_reg),
-				    &t->prstatus.pr_reg, NULL);
+				    0, PR_REG_SIZE(t->prstatus.pr_reg),
+				    PR_REG_PTR(&t->prstatus), NULL);
 
 	fill_note(&t->notes[0], "CORE", NT_PRSTATUS,
-		  sizeof(t->prstatus), &t->prstatus);
+		  PRSTATUS_SIZE(t->prstatus), &t->prstatus);
 	*total += notesize(&t->notes[0]);
 
 	do_thread_regset_writeback(t->task, &view->regsets[0]);
@@ -1438,7 +1454,7 @@ static int fill_thread_core_info(struct elf_thread_core_info *t,
 						  regset->core_note_type,
 						  size, data);
 				else {
-					t->prstatus.pr_fpvalid = 1;
+					SET_PR_FPVALID(&t->prstatus, 1);
 					fill_note(&t->notes[i], "CORE",
 						  NT_PRFPREG, size, data);
 				}
-- 
1.7.6.5

[PATCH 14/30] x86-64: Add prototype for old_rsp to a header file

[H. Peter Anvin]

From: "H. J. Lu" <hjl.tools@gmail.com>

So far this has only been used in process_64.c, but the x32 code will
need it in additional code.

Signed-off-by: H. Peter Anvin <hpa@zytor.com>
---
 arch/x86/include/asm/processor.h |    6 ++++++
 1 files changed, 6 insertions(+), 0 deletions(-)

diff --git a/arch/x86/include/asm/processor.h b/arch/x86/include/asm/processor.h
index 9f748b5..e34f951 100644
--- a/arch/x86/include/asm/processor.h
+++ b/arch/x86/include/asm/processor.h
@@ -948,6 +948,12 @@ extern unsigned long thread_saved_pc(struct task_struct *tsk);
 
 #define task_pt_regs(tsk)	((struct pt_regs *)(tsk)->thread.sp0 - 1)
 extern unsigned long KSTK_ESP(struct task_struct *task);
+
+/*
+ * User space RSP while inside the SYSCALL fast path
+ */
+DECLARE_PER_CPU(unsigned long, old_rsp);
+
 #endif /* CONFIG_X86_64 */
 
 extern void start_thread(struct pt_regs *regs, unsigned long new_ip,
-- 
1.7.6.5

[PATCH 15/30] x32: Add a thread flag for x32 processes

[H. Peter Anvin]

From: "H. Peter Anvin" <hpa@zytor.com>

An x32 process is *almost* the same thing as a 64-bit process with a
32-bit address limit, but there are a few minor differences -- in
particular core dumps are 32 bits and signal handling is different.

Signed-off-by: H. Peter Anvin <hpa@zytor.com>
---
 arch/x86/include/asm/thread_info.h |    2 ++
 arch/x86/kernel/process_64.c       |    2 ++
 2 files changed, 4 insertions(+), 0 deletions(-)

diff --git a/arch/x86/include/asm/thread_info.h b/arch/x86/include/asm/thread_info.h
index d1803a4..912e935 100644
--- a/arch/x86/include/asm/thread_info.h
+++ b/arch/x86/include/asm/thread_info.h
@@ -96,6 +96,7 @@ struct thread_info {
 #define TIF_LAZY_MMU_UPDATES	27	/* task is updating the mmu lazily */
 #define TIF_SYSCALL_TRACEPOINT	28	/* syscall tracepoint instrumentation */
 #define TIF_ADDR32		29	/* 32-bit address space on 64 bits */
+#define TIF_X32			30	/* 32-bit native x86-64 binary */
 
 #define _TIF_SYSCALL_TRACE	(1 << TIF_SYSCALL_TRACE)
 #define _TIF_NOTIFY_RESUME	(1 << TIF_NOTIFY_RESUME)
@@ -118,6 +119,7 @@ struct thread_info {
 #define _TIF_LAZY_MMU_UPDATES	(1 << TIF_LAZY_MMU_UPDATES)
 #define _TIF_SYSCALL_TRACEPOINT	(1 << TIF_SYSCALL_TRACEPOINT)
 #define _TIF_ADDR32		(1 << TIF_ADDR32)
+#define _TIF_X32		(1 << TIF_X32)
 
 /* work to do in syscall_trace_enter() */
 #define _TIF_WORK_SYSCALL_ENTRY	\
diff --git a/arch/x86/kernel/process_64.c b/arch/x86/kernel/process_64.c
index 0e900d0..5fe2fba 100644
--- a/arch/x86/kernel/process_64.c
+++ b/arch/x86/kernel/process_64.c
@@ -509,6 +509,7 @@ void set_personality_64bit(void)
 	/* Make sure to be in 64bit mode */
 	clear_thread_flag(TIF_IA32);
 	clear_thread_flag(TIF_ADDR32);
+	clear_thread_flag(TIF_X32);
 
 	/* Ensure the corresponding mm is not marked. */
 	if (current->mm)
@@ -528,6 +529,7 @@ void set_personality_ia32(void)
 	/* Make sure to be in 32bit mode */
 	set_thread_flag(TIF_IA32);
 	set_thread_flag(TIF_ADDR32);
+	clear_thread_flag(TIF_X32);
 	current->personality |= force_personality32;
 
 	/* Mark the associated mm as containing 32-bit tasks. */
-- 
1.7.6.5

[PATCH 16/30] x86-64, ia32: Drop sys32_rt_sigprocmask

[H. Peter Anvin]

From: "H. Peter Anvin" <hpa@zytor.com>

On x86, the only difference between sys_rt_sigprocmask and
sys32_rt_sigprocmask is the alignment of the data structures.
However, x86 allows data accesses with arbitrary alignment, and
therefore there is no reason for this code to be different.

Reported-by: Gregory M. Lueck <gregory.m.lueck@intel.com>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
---
 arch/x86/ia32/sys_ia32.c         |   40 --------------------------------------
 arch/x86/include/asm/sys_ia32.h  |    2 -
 arch/x86/syscalls/syscall_32.tbl |    2 +-
 3 files changed, 1 insertions(+), 43 deletions(-)

diff --git a/arch/x86/ia32/sys_ia32.c b/arch/x86/ia32/sys_ia32.c
index f6f5c53..aec2202 100644
--- a/arch/x86/ia32/sys_ia32.c
+++ b/arch/x86/ia32/sys_ia32.c
@@ -287,46 +287,6 @@ asmlinkage long sys32_sigaction(int sig, struct old_sigaction32 __user *act,
 	return ret;
 }
 
-asmlinkage long sys32_rt_sigprocmask(int how, compat_sigset_t __user *set,
-				     compat_sigset_t __user *oset,
-				     unsigned int sigsetsize)
-{
-	sigset_t s;
-	compat_sigset_t s32;
-	int ret;
-	mm_segment_t old_fs = get_fs();
-
-	if (set) {
-		if (copy_from_user(&s32, set, sizeof(compat_sigset_t)))
-			return -EFAULT;
-		switch (_NSIG_WORDS) {
-		case 4: s.sig[3] = s32.sig[6] | (((long)s32.sig[7]) << 32);
-		case 3: s.sig[2] = s32.sig[4] | (((long)s32.sig[5]) << 32);
-		case 2: s.sig[1] = s32.sig[2] | (((long)s32.sig[3]) << 32);
-		case 1: s.sig[0] = s32.sig[0] | (((long)s32.sig[1]) << 32);
-		}
-	}
-	set_fs(KERNEL_DS);
-	ret = sys_rt_sigprocmask(how,
-				 set ? (sigset_t __user *)&s : NULL,
-				 oset ? (sigset_t __user *)&s : NULL,
-				 sigsetsize);
-	set_fs(old_fs);
-	if (ret)
-		return ret;
-	if (oset) {
-		switch (_NSIG_WORDS) {
-		case 4: s32.sig[7] = (s.sig[3] >> 32); s32.sig[6] = s.sig[3];
-		case 3: s32.sig[5] = (s.sig[2] >> 32); s32.sig[4] = s.sig[2];
-		case 2: s32.sig[3] = (s.sig[1] >> 32); s32.sig[2] = s.sig[1];
-		case 1: s32.sig[1] = (s.sig[0] >> 32); s32.sig[0] = s.sig[0];
-		}
-		if (copy_to_user(oset, &s32, sizeof(compat_sigset_t)))
-			return -EFAULT;
-	}
-	return 0;
-}
-
 asmlinkage long sys32_alarm(unsigned int seconds)
 {
 	return alarm_setitimer(seconds);
diff --git a/arch/x86/include/asm/sys_ia32.h b/arch/x86/include/asm/sys_ia32.h
index cb23852..68da87b 100644
--- a/arch/x86/include/asm/sys_ia32.h
+++ b/arch/x86/include/asm/sys_ia32.h
@@ -36,8 +36,6 @@ asmlinkage long sys32_rt_sigaction(int, struct sigaction32 __user *,
 				   struct sigaction32 __user *, unsigned int);
 asmlinkage long sys32_sigaction(int, struct old_sigaction32 __user *,
 				struct old_sigaction32 __user *);
-asmlinkage long sys32_rt_sigprocmask(int, compat_sigset_t __user *,
-				     compat_sigset_t __user *, unsigned int);
 asmlinkage long sys32_alarm(unsigned int);
 
 asmlinkage long sys32_waitpid(compat_pid_t, unsigned int *, int);
diff --git a/arch/x86/syscalls/syscall_32.tbl b/arch/x86/syscalls/syscall_32.tbl
index ce98e28..031cef8 100644
--- a/arch/x86/syscalls/syscall_32.tbl
+++ b/arch/x86/syscalls/syscall_32.tbl
@@ -181,7 +181,7 @@
 172	i386	prctl			sys_prctl
 173	i386	rt_sigreturn		ptregs_rt_sigreturn		stub32_rt_sigreturn
 174	i386	rt_sigaction		sys_rt_sigaction		sys32_rt_sigaction
-175	i386	rt_sigprocmask		sys_rt_sigprocmask		sys32_rt_sigprocmask
+175	i386	rt_sigprocmask		sys_rt_sigprocmask
 176	i386	rt_sigpending		sys_rt_sigpending		sys32_rt_sigpending
 177	i386	rt_sigtimedwait		sys_rt_sigtimedwait		compat_sys_rt_sigtimedwait
 178	i386	rt_sigqueueinfo		sys_rt_sigqueueinfo		sys32_rt_sigqueueinfo
-- 
1.7.6.5

[PATCH 17/30] x32: Add x32 system calls to syscall/syscall_64.tbl

[H. Peter Anvin]

From: "H. Peter Anvin" <hpa@zytor.com>

Split the 64-bit system calls into "64" (64-bit only) and "common"
(64-bit or x32) and add the x32 system call numbers.

Signed-off-by: H. Peter Anvin <hpa@zytor.com>
---
 arch/x86/kernel/asm-offsets_64.c |    2 +
 arch/x86/kernel/syscall_64.c     |    3 +
 arch/x86/syscalls/Makefile       |    2 +-
 arch/x86/syscalls/syscall_64.tbl |  579 ++++++++++++++++++++------------------
 arch/x86/um/sys_call_table_64.c  |    3 +
 arch/x86/um/user-offsets.c       |    2 +
 6 files changed, 317 insertions(+), 274 deletions(-)

diff --git a/arch/x86/kernel/asm-offsets_64.c b/arch/x86/kernel/asm-offsets_64.c
index 834e897..c3354f7 100644
--- a/arch/x86/kernel/asm-offsets_64.c
+++ b/arch/x86/kernel/asm-offsets_64.c
@@ -1,6 +1,8 @@
 #include <asm/ia32.h>
 
 #define __SYSCALL_64(nr, sym, compat) [nr] = 1,
+#define __SYSCALL_COMMON(nr, sym, compat) [nr] = 1,
+#define __SYSCALL_X32(nr, sym, compat) /* Not yet */
 static char syscalls_64[] = {
 #include <asm/syscalls_64.h>
 };
diff --git a/arch/x86/kernel/syscall_64.c b/arch/x86/kernel/syscall_64.c
index 7ac7943..26c4ca1 100644
--- a/arch/x86/kernel/syscall_64.c
+++ b/arch/x86/kernel/syscall_64.c
@@ -5,6 +5,9 @@
 #include <linux/cache.h>
 #include <asm/asm-offsets.h>
 
+#define __SYSCALL_COMMON(nr, sym, compat) __SYSCALL_64(nr, sym, compat)
+#define __SYSCALL_X32(nr, sym, compat) /* Not yet */
+
 #define __SYSCALL_64(nr, sym, compat) extern asmlinkage void sym(void) ;
 #include <asm/syscalls_64.h>
 #undef __SYSCALL_64
diff --git a/arch/x86/syscalls/Makefile b/arch/x86/syscalls/Makefile
index 564b247..89dd958 100644
--- a/arch/x86/syscalls/Makefile
+++ b/arch/x86/syscalls/Makefile
@@ -24,7 +24,7 @@ syshdr_pfx_unistd_32_ia32 := ia32_
 $(out)/unistd_32_ia32.h: $(syscall32) $(syshdr)
 	$(call if_changed,syshdr)
 
-syshdr_abi_unistd_64 := 64
+syshdr_abi_unistd_64 := common,64
 $(out)/unistd_64.h: $(syscall64) $(syshdr)
 	$(call if_changed,syshdr)
 
diff --git a/arch/x86/syscalls/syscall_64.tbl b/arch/x86/syscalls/syscall_64.tbl
index b440a8f..4aecc7e 100644
--- a/arch/x86/syscalls/syscall_64.tbl
+++ b/arch/x86/syscalls/syscall_64.tbl
@@ -4,317 +4,350 @@
 # The format is:
 # <number> <abi> <name> <entry point>
 #
-# The abi is always "64" for this file (for now.)
+# The abi is "common", "64" or "x32" for this file.
 #
-0	64	read			sys_read
-1	64	write			sys_write
-2	64	open			sys_open
-3	64	close			sys_close
-4	64	stat			sys_newstat
-5	64	fstat			sys_newfstat
-6	64	lstat			sys_newlstat
-7	64	poll			sys_poll
-8	64	lseek			sys_lseek
-9	64	mmap			sys_mmap
-10	64	mprotect		sys_mprotect
-11	64	munmap			sys_munmap
-12	64	brk			sys_brk
+0	common	read			sys_read
+1	common	write			sys_write
+2	common	open			sys_open
+3	common	close			sys_close
+4	common	stat			sys_newstat
+5	common	fstat			sys_newfstat
+6	common	lstat			sys_newlstat
+7	common	poll			sys_poll
+8	common	lseek			sys_lseek
+9	common	mmap			sys_mmap
+10	common	mprotect		sys_mprotect
+11	common	munmap			sys_munmap
+12	common	brk			sys_brk
 13	64	rt_sigaction		sys_rt_sigaction
-14	64	rt_sigprocmask		sys_rt_sigprocmask
+14	common	rt_sigprocmask		sys_rt_sigprocmask
 15	64	rt_sigreturn		stub_rt_sigreturn
 16	64	ioctl			sys_ioctl
-17	64	pread64			sys_pread64
-18	64	pwrite64		sys_pwrite64
+17	common	pread64			sys_pread64
+18	common	pwrite64		sys_pwrite64
 19	64	readv			sys_readv
 20	64	writev			sys_writev
-21	64	access			sys_access
-22	64	pipe			sys_pipe
-23	64	select			sys_select
-24	64	sched_yield		sys_sched_yield
-25	64	mremap			sys_mremap
-26	64	msync			sys_msync
-27	64	mincore			sys_mincore
-28	64	madvise			sys_madvise
-29	64	shmget			sys_shmget
-30	64	shmat			sys_shmat
-31	64	shmctl			sys_shmctl
-32	64	dup			sys_dup
-33	64	dup2			sys_dup2
-34	64	pause			sys_pause
-35	64	nanosleep		sys_nanosleep
-36	64	getitimer		sys_getitimer
-37	64	alarm			sys_alarm
-38	64	setitimer		sys_setitimer
-39	64	getpid			sys_getpid
-40	64	sendfile		sys_sendfile64
-41	64	socket			sys_socket
-42	64	connect			sys_connect
-43	64	accept			sys_accept
-44	64	sendto			sys_sendto
+21	common	access			sys_access
+22	common	pipe			sys_pipe
+23	common	select			sys_select
+24	common	sched_yield		sys_sched_yield
+25	common	mremap			sys_mremap
+26	common	msync			sys_msync
+27	common	mincore			sys_mincore
+28	common	madvise			sys_madvise
+29	common	shmget			sys_shmget
+30	common	shmat			sys_shmat
+31	common	shmctl			sys_shmctl
+32	common	dup			sys_dup
+33	common	dup2			sys_dup2
+34	common	pause			sys_pause
+35	common	nanosleep		sys_nanosleep
+36	common	getitimer		sys_getitimer
+37	common	alarm			sys_alarm
+38	common	setitimer		sys_setitimer
+39	common	getpid			sys_getpid
+40	common	sendfile		sys_sendfile64
+41	common	socket			sys_socket
+42	common	connect			sys_connect
+43	common	accept			sys_accept
+44	common	sendto			sys_sendto
 45	64	recvfrom		sys_recvfrom
 46	64	sendmsg			sys_sendmsg
 47	64	recvmsg			sys_recvmsg
-48	64	shutdown		sys_shutdown
-49	64	bind			sys_bind
-50	64	listen			sys_listen
-51	64	getsockname		sys_getsockname
-52	64	getpeername		sys_getpeername
-53	64	socketpair		sys_socketpair
-54	64	setsockopt		sys_setsockopt
-55	64	getsockopt		sys_getsockopt
-56	64	clone			stub_clone
-57	64	fork			stub_fork
-58	64	vfork			stub_vfork
+48	common	shutdown		sys_shutdown
+49	common	bind			sys_bind
+50	common	listen			sys_listen
+51	common	getsockname		sys_getsockname
+52	common	getpeername		sys_getpeername
+53	common	socketpair		sys_socketpair
+54	common	setsockopt		sys_setsockopt
+55	common	getsockopt		sys_getsockopt
+56	common	clone			stub_clone
+57	common	fork			stub_fork
+58	common	vfork			stub_vfork
 59	64	execve			stub_execve
-60	64	exit			sys_exit
-61	64	wait4			sys_wait4
-62	64	kill			sys_kill
-63	64	uname			sys_newuname
-64	64	semget			sys_semget
-65	64	semop			sys_semop
-66	64	semctl			sys_semctl
-67	64	shmdt			sys_shmdt
-68	64	msgget			sys_msgget
-69	64	msgsnd			sys_msgsnd
-70	64	msgrcv			sys_msgrcv
-71	64	msgctl			sys_msgctl
-72	64	fcntl			sys_fcntl
-73	64	flock			sys_flock
-74	64	fsync			sys_fsync
-75	64	fdatasync		sys_fdatasync
-76	64	truncate		sys_truncate
-77	64	ftruncate		sys_ftruncate
-78	64	getdents		sys_getdents
-79	64	getcwd			sys_getcwd
-80	64	chdir			sys_chdir
-81	64	fchdir			sys_fchdir
-82	64	rename			sys_rename
-83	64	mkdir			sys_mkdir
-84	64	rmdir			sys_rmdir
-85	64	creat			sys_creat
-86	64	link			sys_link
-87	64	unlink			sys_unlink
-88	64	symlink			sys_symlink
-89	64	readlink		sys_readlink
-90	64	chmod			sys_chmod
-91	64	fchmod			sys_fchmod
-92	64	chown			sys_chown
-93	64	fchown			sys_fchown
-94	64	lchown			sys_lchown
-95	64	umask			sys_umask
-96	64	gettimeofday		sys_gettimeofday
-97	64	getrlimit		sys_getrlimit
-98	64	getrusage		sys_getrusage
-99	64	sysinfo			sys_sysinfo
+60	common	exit			sys_exit
+61	common	wait4			sys_wait4
+62	common	kill			sys_kill
+63	common	uname			sys_newuname
+64	common	semget			sys_semget
+65	common	semop			sys_semop
+66	common	semctl			sys_semctl
+67	common	shmdt			sys_shmdt
+68	common	msgget			sys_msgget
+69	common	msgsnd			sys_msgsnd
+70	common	msgrcv			sys_msgrcv
+71	common	msgctl			sys_msgctl
+72	common	fcntl			sys_fcntl
+73	common	flock			sys_flock
+74	common	fsync			sys_fsync
+75	common	fdatasync		sys_fdatasync
+76	common	truncate		sys_truncate
+77	common	ftruncate		sys_ftruncate
+78	common	getdents		sys_getdents
+79	common	getcwd			sys_getcwd
+80	common	chdir			sys_chdir
+81	common	fchdir			sys_fchdir
+82	common	rename			sys_rename
+83	common	mkdir			sys_mkdir
+84	common	rmdir			sys_rmdir
+85	common	creat			sys_creat
+86	common	link			sys_link
+87	common	unlink			sys_unlink
+88	common	symlink			sys_symlink
+89	common	readlink		sys_readlink
+90	common	chmod			sys_chmod
+91	common	fchmod			sys_fchmod
+92	common	chown			sys_chown
+93	common	fchown			sys_fchown
+94	common	lchown			sys_lchown
+95	common	umask			sys_umask
+96	common	gettimeofday		sys_gettimeofday
+97	common	getrlimit		sys_getrlimit
+98	common	getrusage		sys_getrusage
+99	common	sysinfo			sys_sysinfo
 100	64	times			sys_times
-101	64	ptrace			sys_ptrace
-102	64	getuid			sys_getuid
-103	64	syslog			sys_syslog
-104	64	getgid			sys_getgid
-105	64	setuid			sys_setuid
-106	64	setgid			sys_setgid
-107	64	geteuid			sys_geteuid
-108	64	getegid			sys_getegid
-109	64	setpgid			sys_setpgid
-110	64	getppid			sys_getppid
-111	64	getpgrp			sys_getpgrp
-112	64	setsid			sys_setsid
-113	64	setreuid		sys_setreuid
-114	64	setregid		sys_setregid
-115	64	getgroups		sys_getgroups
-116	64	setgroups		sys_setgroups
-117	64	setresuid		sys_setresuid
-118	64	getresuid		sys_getresuid
-119	64	setresgid		sys_setresgid
-120	64	getresgid		sys_getresgid
-121	64	getpgid			sys_getpgid
-122	64	setfsuid		sys_setfsuid
-123	64	setfsgid		sys_setfsgid
-124	64	getsid			sys_getsid
-125	64	capget			sys_capget
-126	64	capset			sys_capset
+101	common	ptrace			sys_ptrace
+102	common	getuid			sys_getuid
+103	common	syslog			sys_syslog
+104	common	getgid			sys_getgid
+105	common	setuid			sys_setuid
+106	common	setgid			sys_setgid
+107	common	geteuid			sys_geteuid
+108	common	getegid			sys_getegid
+109	common	setpgid			sys_setpgid
+110	common	getppid			sys_getppid
+111	common	getpgrp			sys_getpgrp
+112	common	setsid			sys_setsid
+113	common	setreuid		sys_setreuid
+114	common	setregid		sys_setregid
+115	common	getgroups		sys_getgroups
+116	common	setgroups		sys_setgroups
+117	common	setresuid		sys_setresuid
+118	common	getresuid		sys_getresuid
+119	common	setresgid		sys_setresgid
+120	common	getresgid		sys_getresgid
+121	common	getpgid			sys_getpgid
+122	common	setfsuid		sys_setfsuid
+123	common	setfsgid		sys_setfsgid
+124	common	getsid			sys_getsid
+125	common	capget			sys_capget
+126	common	capset			sys_capset
 127	64	rt_sigpending		sys_rt_sigpending
 128	64	rt_sigtimedwait		sys_rt_sigtimedwait
 129	64	rt_sigqueueinfo		sys_rt_sigqueueinfo
-130	64	rt_sigsuspend		sys_rt_sigsuspend
+130	common	rt_sigsuspend		sys_rt_sigsuspend
 131	64	sigaltstack		stub_sigaltstack
-132	64	utime			sys_utime
-133	64	mknod			sys_mknod
+132	common	utime			sys_utime
+133	common	mknod			sys_mknod
 134	64	uselib
-135	64	personality		sys_personality
-136	64	ustat			sys_ustat
-137	64	statfs			sys_statfs
-138	64	fstatfs			sys_fstatfs
-139	64	sysfs			sys_sysfs
-140	64	getpriority		sys_getpriority
-141	64	setpriority		sys_setpriority
-142	64	sched_setparam		sys_sched_setparam
-143	64	sched_getparam		sys_sched_getparam
-144	64	sched_setscheduler	sys_sched_setscheduler
-145	64	sched_getscheduler	sys_sched_getscheduler
-146	64	sched_get_priority_max	sys_sched_get_priority_max
-147	64	sched_get_priority_min	sys_sched_get_priority_min
-148	64	sched_rr_get_interval	sys_sched_rr_get_interval
-149	64	mlock			sys_mlock
-150	64	munlock			sys_munlock
-151	64	mlockall		sys_mlockall
-152	64	munlockall		sys_munlockall
-153	64	vhangup			sys_vhangup
-154	64	modify_ldt		sys_modify_ldt
-155	64	pivot_root		sys_pivot_root
+135	common	personality		sys_personality
+136	common	ustat			sys_ustat
+137	common	statfs			sys_statfs
+138	common	fstatfs			sys_fstatfs
+139	common	sysfs			sys_sysfs
+140	common	getpriority		sys_getpriority
+141	common	setpriority		sys_setpriority
+142	common	sched_setparam		sys_sched_setparam
+143	common	sched_getparam		sys_sched_getparam
+144	common	sched_setscheduler	sys_sched_setscheduler
+145	common	sched_getscheduler	sys_sched_getscheduler
+146	common	sched_get_priority_max	sys_sched_get_priority_max
+147	common	sched_get_priority_min	sys_sched_get_priority_min
+148	common	sched_rr_get_interval	sys_sched_rr_get_interval
+149	common	mlock			sys_mlock
+150	common	munlock			sys_munlock
+151	common	mlockall		sys_mlockall
+152	common	munlockall		sys_munlockall
+153	common	vhangup			sys_vhangup
+154	common	modify_ldt		sys_modify_ldt
+155	common	pivot_root		sys_pivot_root
 156	64	_sysctl			sys_sysctl
-157	64	prctl			sys_prctl
-158	64	arch_prctl		sys_arch_prctl
-159	64	adjtimex		sys_adjtimex
-160	64	setrlimit		sys_setrlimit
-161	64	chroot			sys_chroot
-162	64	sync			sys_sync
-163	64	acct			sys_acct
-164	64	settimeofday		sys_settimeofday
-165	64	mount			sys_mount
-166	64	umount2			sys_umount
-167	64	swapon			sys_swapon
-168	64	swapoff			sys_swapoff
-169	64	reboot			sys_reboot
-170	64	sethostname		sys_sethostname
-171	64	setdomainname		sys_setdomainname
-172	64	iopl			stub_iopl
-173	64	ioperm			sys_ioperm
+157	common	prctl			sys_prctl
+158	common	arch_prctl		sys_arch_prctl
+159	common	adjtimex		sys_adjtimex
+160	common	setrlimit		sys_setrlimit
+161	common	chroot			sys_chroot
+162	common	sync			sys_sync
+163	common	acct			sys_acct
+164	common	settimeofday		sys_settimeofday
+165	common	mount			sys_mount
+166	common	umount2			sys_umount
+167	common	swapon			sys_swapon
+168	common	swapoff			sys_swapoff
+169	common	reboot			sys_reboot
+170	common	sethostname		sys_sethostname
+171	common	setdomainname		sys_setdomainname
+172	common	iopl			stub_iopl
+173	common	ioperm			sys_ioperm
 174	64	create_module
-175	64	init_module		sys_init_module
-176	64	delete_module		sys_delete_module
+175	common	init_module		sys_init_module
+176	common	delete_module		sys_delete_module
 177	64	get_kernel_syms
 178	64	query_module
-179	64	quotactl		sys_quotactl
+179	common	quotactl		sys_quotactl
 180	64	nfsservctl
-181	64	getpmsg
-182	64	putpmsg
-183	64	afs_syscall
-184	64	tuxcall
-185	64	security
-186	64	gettid			sys_gettid
-187	64	readahead		sys_readahead
-188	64	setxattr		sys_setxattr
-189	64	lsetxattr		sys_lsetxattr
-190	64	fsetxattr		sys_fsetxattr
-191	64	getxattr		sys_getxattr
-192	64	lgetxattr		sys_lgetxattr
-193	64	fgetxattr		sys_fgetxattr
-194	64	listxattr		sys_listxattr
-195	64	llistxattr		sys_llistxattr
-196	64	flistxattr		sys_flistxattr
-197	64	removexattr		sys_removexattr
-198	64	lremovexattr		sys_lremovexattr
-199	64	fremovexattr		sys_fremovexattr
-200	64	tkill			sys_tkill
-201	64	time			sys_time
-202	64	futex			sys_futex
-203	64	sched_setaffinity	sys_sched_setaffinity
-204	64	sched_getaffinity	sys_sched_getaffinity
+181	common	getpmsg
+182	common	putpmsg
+183	common	afs_syscall
+184	common	tuxcall
+185	common	security
+186	common	gettid			sys_gettid
+187	common	readahead		sys_readahead
+188	common	setxattr		sys_setxattr
+189	common	lsetxattr		sys_lsetxattr
+190	common	fsetxattr		sys_fsetxattr
+191	common	getxattr		sys_getxattr
+192	common	lgetxattr		sys_lgetxattr
+193	common	fgetxattr		sys_fgetxattr
+194	common	listxattr		sys_listxattr
+195	common	llistxattr		sys_llistxattr
+196	common	flistxattr		sys_flistxattr
+197	common	removexattr		sys_removexattr
+198	common	lremovexattr		sys_lremovexattr
+199	common	fremovexattr		sys_fremovexattr
+200	common	tkill			sys_tkill
+201	common	time			sys_time
+202	common	futex			sys_futex
+203	common	sched_setaffinity	sys_sched_setaffinity
+204	common	sched_getaffinity	sys_sched_getaffinity
 205	64	set_thread_area
-206	64	io_setup		sys_io_setup
-207	64	io_destroy		sys_io_destroy
-208	64	io_getevents		sys_io_getevents
-209	64	io_submit		sys_io_submit
-210	64	io_cancel		sys_io_cancel
+206	common	io_setup		sys_io_setup
+207	common	io_destroy		sys_io_destroy
+208	common	io_getevents		sys_io_getevents
+209	common	io_submit		sys_io_submit
+210	common	io_cancel		sys_io_cancel
 211	64	get_thread_area
-212	64	lookup_dcookie		sys_lookup_dcookie
-213	64	epoll_create		sys_epoll_create
+212	common	lookup_dcookie		sys_lookup_dcookie
+213	common	epoll_create		sys_epoll_create
 214	64	epoll_ctl_old
 215	64	epoll_wait_old
-216	64	remap_file_pages	sys_remap_file_pages
-217	64	getdents64		sys_getdents64
-218	64	set_tid_address		sys_set_tid_address
-219	64	restart_syscall		sys_restart_syscall
-220	64	semtimedop		sys_semtimedop
-221	64	fadvise64		sys_fadvise64
+216	common	remap_file_pages	sys_remap_file_pages
+217	common	getdents64		sys_getdents64
+218	common	set_tid_address		sys_set_tid_address
+219	common	restart_syscall		sys_restart_syscall
+220	common	semtimedop		sys_semtimedop
+221	common	fadvise64		sys_fadvise64
 222	64	timer_create		sys_timer_create
-223	64	timer_settime		sys_timer_settime
-224	64	timer_gettime		sys_timer_gettime
-225	64	timer_getoverrun	sys_timer_getoverrun
-226	64	timer_delete		sys_timer_delete
-227	64	clock_settime		sys_clock_settime
-228	64	clock_gettime		sys_clock_gettime
-229	64	clock_getres		sys_clock_getres
-230	64	clock_nanosleep		sys_clock_nanosleep
-231	64	exit_group		sys_exit_group
-232	64	epoll_wait		sys_epoll_wait
-233	64	epoll_ctl		sys_epoll_ctl
-234	64	tgkill			sys_tgkill
-235	64	utimes			sys_utimes
+223	common	timer_settime		sys_timer_settime
+224	common	timer_gettime		sys_timer_gettime
+225	common	timer_getoverrun	sys_timer_getoverrun
+226	common	timer_delete		sys_timer_delete
+227	common	clock_settime		sys_clock_settime
+228	common	clock_gettime		sys_clock_gettime
+229	common	clock_getres		sys_clock_getres
+230	common	clock_nanosleep		sys_clock_nanosleep
+231	common	exit_group		sys_exit_group
+232	common	epoll_wait		sys_epoll_wait
+233	common	epoll_ctl		sys_epoll_ctl
+234	common	tgkill			sys_tgkill
+235	common	utimes			sys_utimes
 236	64	vserver
-237	64	mbind			sys_mbind
-238	64	set_mempolicy		sys_set_mempolicy
-239	64	get_mempolicy		sys_get_mempolicy
-240	64	mq_open			sys_mq_open
-241	64	mq_unlink		sys_mq_unlink
-242	64	mq_timedsend		sys_mq_timedsend
-243	64	mq_timedreceive		sys_mq_timedreceive
+237	common	mbind			sys_mbind
+238	common	set_mempolicy		sys_set_mempolicy
+239	common	get_mempolicy		sys_get_mempolicy
+240	common	mq_open			sys_mq_open
+241	common	mq_unlink		sys_mq_unlink
+242	common	mq_timedsend		sys_mq_timedsend
+243	common	mq_timedreceive		sys_mq_timedreceive
 244	64	mq_notify		sys_mq_notify
-245	64	mq_getsetattr		sys_mq_getsetattr
+245	common	mq_getsetattr		sys_mq_getsetattr
 246	64	kexec_load		sys_kexec_load
 247	64	waitid			sys_waitid
-248	64	add_key			sys_add_key
-249	64	request_key		sys_request_key
-250	64	keyctl			sys_keyctl
-251	64	ioprio_set		sys_ioprio_set
-252	64	ioprio_get		sys_ioprio_get
-253	64	inotify_init		sys_inotify_init
-254	64	inotify_add_watch	sys_inotify_add_watch
-255	64	inotify_rm_watch	sys_inotify_rm_watch
-256	64	migrate_pages		sys_migrate_pages
-257	64	openat			sys_openat
-258	64	mkdirat			sys_mkdirat
-259	64	mknodat			sys_mknodat
-260	64	fchownat		sys_fchownat
-261	64	futimesat		sys_futimesat
-262	64	newfstatat		sys_newfstatat
-263	64	unlinkat		sys_unlinkat
-264	64	renameat		sys_renameat
-265	64	linkat			sys_linkat
-266	64	symlinkat		sys_symlinkat
-267	64	readlinkat		sys_readlinkat
-268	64	fchmodat		sys_fchmodat
-269	64	faccessat		sys_faccessat
-270	64	pselect6		sys_pselect6
-271	64	ppoll			sys_ppoll
-272	64	unshare			sys_unshare
+248	common	add_key			sys_add_key
+249	common	request_key		sys_request_key
+250	common	keyctl			sys_keyctl
+251	common	ioprio_set		sys_ioprio_set
+252	common	ioprio_get		sys_ioprio_get
+253	common	inotify_init		sys_inotify_init
+254	common	inotify_add_watch	sys_inotify_add_watch
+255	common	inotify_rm_watch	sys_inotify_rm_watch
+256	common	migrate_pages		sys_migrate_pages
+257	common	openat			sys_openat
+258	common	mkdirat			sys_mkdirat
+259	common	mknodat			sys_mknodat
+260	common	fchownat		sys_fchownat
+261	common	futimesat		sys_futimesat
+262	common	newfstatat		sys_newfstatat
+263	common	unlinkat		sys_unlinkat
+264	common	renameat		sys_renameat
+265	common	linkat			sys_linkat
+266	common	symlinkat		sys_symlinkat
+267	common	readlinkat		sys_readlinkat
+268	common	fchmodat		sys_fchmodat
+269	common	faccessat		sys_faccessat
+270	common	pselect6		sys_pselect6
+271	common	ppoll			sys_ppoll
+272	common	unshare			sys_unshare
 273	64	set_robust_list		sys_set_robust_list
 274	64	get_robust_list		sys_get_robust_list
-275	64	splice			sys_splice
-276	64	tee			sys_tee
-277	64	sync_file_range		sys_sync_file_range
+275	common	splice			sys_splice
+276	common	tee			sys_tee
+277	common	sync_file_range		sys_sync_file_range
 278	64	vmsplice		sys_vmsplice
 279	64	move_pages		sys_move_pages
-280	64	utimensat		sys_utimensat
-281	64	epoll_pwait		sys_epoll_pwait
-282	64	signalfd		sys_signalfd
-283	64	timerfd_create		sys_timerfd_create
-284	64	eventfd			sys_eventfd
-285	64	fallocate		sys_fallocate
-286	64	timerfd_settime		sys_timerfd_settime
-287	64	timerfd_gettime		sys_timerfd_gettime
-288	64	accept4			sys_accept4
-289	64	signalfd4		sys_signalfd4
-290	64	eventfd2		sys_eventfd2
-291	64	epoll_create1		sys_epoll_create1
-292	64	dup3			sys_dup3
-293	64	pipe2			sys_pipe2
-294	64	inotify_init1		sys_inotify_init1
+280	common	utimensat		sys_utimensat
+281	common	epoll_pwait		sys_epoll_pwait
+282	common	signalfd		sys_signalfd
+283	common	timerfd_create		sys_timerfd_create
+284	common	eventfd			sys_eventfd
+285	common	fallocate		sys_fallocate
+286	common	timerfd_settime		sys_timerfd_settime
+287	common	timerfd_gettime		sys_timerfd_gettime
+288	common	accept4			sys_accept4
+289	common	signalfd4		sys_signalfd4
+290	common	eventfd2		sys_eventfd2
+291	common	epoll_create1		sys_epoll_create1
+292	common	dup3			sys_dup3
+293	common	pipe2			sys_pipe2
+294	common	inotify_init1		sys_inotify_init1
 295	64	preadv			sys_preadv
 296	64	pwritev			sys_pwritev
 297	64	rt_tgsigqueueinfo	sys_rt_tgsigqueueinfo
-298	64	perf_event_open		sys_perf_event_open
+298	common	perf_event_open		sys_perf_event_open
 299	64	recvmmsg		sys_recvmmsg
-300	64	fanotify_init		sys_fanotify_init
-301	64	fanotify_mark		sys_fanotify_mark
-302	64	prlimit64		sys_prlimit64
-303	64	name_to_handle_at	sys_name_to_handle_at
-304	64	open_by_handle_at	sys_open_by_handle_at
-305	64	clock_adjtime		sys_clock_adjtime
-306	64	syncfs			sys_syncfs
+300	common	fanotify_init		sys_fanotify_init
+301	common	fanotify_mark		sys_fanotify_mark
+302	common	prlimit64		sys_prlimit64
+303	common	name_to_handle_at	sys_name_to_handle_at
+304	common	open_by_handle_at	sys_open_by_handle_at
+305	common	clock_adjtime		sys_clock_adjtime
+306	common	syncfs			sys_syncfs
 307	64	sendmmsg		sys_sendmmsg
-308	64	setns			sys_setns
-309	64	getcpu			sys_getcpu
+308	common	setns			sys_setns
+309	common	getcpu			sys_getcpu
 310	64	process_vm_readv	sys_process_vm_readv
 311	64	process_vm_writev	sys_process_vm_writev
+#
+# x32-specific system call numbers start at 512 to avoid cache impact
+# for native 64-bit operation.
+#
+512	x32	rt_sigaction		sys32_rt_sigaction
+513	x32	rt_sigreturn		stub_x32_rt_sigreturn
+514	x32	ioctl			compat_sys_ioctl
+515	x32	readv			compat_sys_readv
+516	x32	writev			compat_sys_writev
+517	x32	recvfrom		compat_sys_recvfrom
+518	x32	sendmsg			compat_sys_sendmsg
+519	x32	recvmsg			compat_sys_recvmsg
+520	x32	execve			stub_x32_execve
+521	x32	times			compat_sys_times
+522	x32	rt_sigpending		sys32_rt_sigpending
+523	x32	rt_sigtimedwait		compat_sys_rt_sigtimedwait
+524	x32	rt_sigqueueinfo		sys32_rt_sigqueueinfo
+525	x32	sigaltstack		stub_x32_sigaltstack
+526	x32	timer_create		compat_sys_timer_create
+527	x32	mq_notify		compat_sys_mq_notify
+528	x32	kexec_load		compat_sys_kexec_load
+529	x32	waitid			compat_sys_waitid
+530	x32	set_robust_list		compat_sys_set_robust_list
+531	x32	get_robust_list		compat_sys_get_robust_list
+532	x32	vmsplice		compat_sys_vmsplice
+533	x32	move_pages		compat_sys_move_pages
+534	x32	preadv			compat_sys_preadv64
+535	x32	pwritev			compat_sys_pwritev64
+536	x32	rt_tgsigqueueinfo	compat_sys_rt_tgsigqueueinfo
+537	x32	recvmmsg		compat_sys_recvmmsg
+538	x32	sendmmsg		compat_sys_sendmmsg
+539	x32	process_vm_readv	compat_sys_process_vm_readv
+540	x32	process_vm_writev	compat_sys_process_vm_writev
diff --git a/arch/x86/um/sys_call_table_64.c b/arch/x86/um/sys_call_table_64.c
index fe626c3..9924776 100644
--- a/arch/x86/um/sys_call_table_64.c
+++ b/arch/x86/um/sys_call_table_64.c
@@ -35,6 +35,9 @@
 #define stub_sigaltstack sys_sigaltstack
 #define stub_rt_sigreturn sys_rt_sigreturn
 
+#define __SYSCALL_COMMON(nr, sym, compat) __SYSCALL_64(nr, sym, compat)
+#define __SYSCALL_X32(nr, sym, compat) /* Not supported */
+
 #define __SYSCALL_64(nr, sym, compat) extern asmlinkage void sym(void) ;
 #include <asm/syscalls_64.h>
 
diff --git a/arch/x86/um/user-offsets.c b/arch/x86/um/user-offsets.c
index 5edf4f4..ce7e360 100644
--- a/arch/x86/um/user-offsets.c
+++ b/arch/x86/um/user-offsets.c
@@ -15,6 +15,8 @@ static char syscalls[] = {
 };
 #else
 #define __SYSCALL_64(nr, sym, compat) [nr] = 1,
+#define __SYSCALL_COMMON(nr, sym, compat) [nr] = 1,
+#define __SYSCALL_X32(nr, sym, compat) /* Not supported */
 static char syscalls[] = {
 #include <asm/syscalls_64.h>
 };
-- 
1.7.6.5

[PATCH 18/30] x32: Generate <asm/unistd_x32.h>

[H. Peter Anvin]

From: "H. Peter Anvin" <hpa@zytor.com>

Generate <asm/unistd_x32.h>; this exports x32 system call numbers to
user space.

Signed-off-by: H. Peter Anvin <hpa@zytor.com>
---
 arch/x86/include/asm/Kbuild   |    1 +
 arch/x86/include/asm/unistd.h |    7 ++++++-
 arch/x86/syscalls/Makefile    |   10 ++++++++--
 3 files changed, 15 insertions(+), 3 deletions(-)

diff --git a/arch/x86/include/asm/Kbuild b/arch/x86/include/asm/Kbuild
index 986954f..f9c0d3b 100644
--- a/arch/x86/include/asm/Kbuild
+++ b/arch/x86/include/asm/Kbuild
@@ -25,3 +25,4 @@ header-y += vsyscall.h
 
 genhdr-y += unistd_32.h
 genhdr-y += unistd_64.h
+genhdr-y += unistd_x32.h
diff --git a/arch/x86/include/asm/unistd.h b/arch/x86/include/asm/unistd.h
index 21f77b8..dab5349 100644
--- a/arch/x86/include/asm/unistd.h
+++ b/arch/x86/include/asm/unistd.h
@@ -1,6 +1,9 @@
 #ifndef _ASM_X86_UNISTD_H
 #define _ASM_X86_UNISTD_H 1
 
+/* x32 syscall flag bit */
+#define __X32_SYSCALL_BIT	0x40000000
+
 #ifdef __KERNEL__
 # ifdef CONFIG_X86_32
 
@@ -52,8 +55,10 @@
 #else
 # ifdef __i386__
 #  include <asm/unistd_32.h>
-# else
+# elif defined(__LP64__)
 #  include <asm/unistd_64.h>
+# else
+#  include <asm/unistd_x32.h>
 # endif
 #endif
 
diff --git a/arch/x86/syscalls/Makefile b/arch/x86/syscalls/Makefile
index 89dd958..54bdbd7 100644
--- a/arch/x86/syscalls/Makefile
+++ b/arch/x86/syscalls/Makefile
@@ -11,7 +11,8 @@ systbl := $(srctree)/$(src)/syscalltbl.sh
 
 quiet_cmd_syshdr = SYSHDR  $@
       cmd_syshdr = $(CONFIG_SHELL) '$(syshdr)' $< $@ \
-		   $(syshdr_abi_$(basetarget)) $(syshdr_pfx_$(basetarget))
+		   $(syshdr_abi_$(basetarget)) $(syshdr_pfx_$(basetarget)) \
+		   $(syshdr_offset_$(basetarget))
 quiet_cmd_systbl = SYSTBL  $@
       cmd_systbl = $(CONFIG_SHELL) '$(systbl)' $< $@
 
@@ -24,6 +25,11 @@ syshdr_pfx_unistd_32_ia32 := ia32_
 $(out)/unistd_32_ia32.h: $(syscall32) $(syshdr)
 	$(call if_changed,syshdr)
 
+syshdr_abi_unistd_x32 := common,x32
+syshdr_offset_unistd_x32 := __X32_SYSCALL_BIT
+$(out)/unistd_x32.h: $(syscall64) $(syshdr)
+	$(call if_changed,syshdr)
+
 syshdr_abi_unistd_64 := common,64
 $(out)/unistd_64.h: $(syscall64) $(syshdr)
 	$(call if_changed,syshdr)
@@ -33,7 +39,7 @@ $(out)/syscalls_32.h: $(syscall32) $(systbl)
 $(out)/syscalls_64.h: $(syscall64) $(systbl)
 	$(call if_changed,systbl)
 
-syshdr-y			+= unistd_32.h unistd_64.h
+syshdr-y			+= unistd_32.h unistd_64.h unistd_x32.h
 syshdr-y			+= syscalls_32.h
 syshdr-$(CONFIG_X86_64)		+= unistd_32_ia32.h
 syshdr-$(CONFIG_X86_64)		+= syscalls_64.h
-- 
1.7.6.5

[PATCH 19/30] x32: Generate <asm/unistd_64_x32.h>

[H. Peter Anvin]

From: "H. Peter Anvin" <hpa@zytor.com>

Generate macros for the *kernel* code to use to refer to x32 system
calls.  These have an __NR_x32_ prefix and do not include
__X32_SYSCALL_BIT.

Signed-off-by: H. Peter Anvin <hpa@zytor.com>
---
 arch/x86/include/asm/unistd.h |    1 +
 arch/x86/syscalls/Makefile    |    7 ++++++-
 2 files changed, 7 insertions(+), 1 deletions(-)

diff --git a/arch/x86/include/asm/unistd.h b/arch/x86/include/asm/unistd.h
index dab5349..7a48a55 100644
--- a/arch/x86/include/asm/unistd.h
+++ b/arch/x86/include/asm/unistd.h
@@ -17,6 +17,7 @@
 # else
 
 #  include <asm/unistd_64.h>
+#  include <asm/unistd_64_x32.h>
 #  define __ARCH_WANT_COMPAT_SYS_TIME
 
 # endif
diff --git a/arch/x86/syscalls/Makefile b/arch/x86/syscalls/Makefile
index 54bdbd7..414d402 100644
--- a/arch/x86/syscalls/Makefile
+++ b/arch/x86/syscalls/Makefile
@@ -34,6 +34,11 @@ syshdr_abi_unistd_64 := common,64
 $(out)/unistd_64.h: $(syscall64) $(syshdr)
 	$(call if_changed,syshdr)
 
+syshdr_abi_unistd_64_x32 := x32
+syshdr_pfx_unistd_64_x32 := x32_
+$(out)/unistd_64_x32.h: $(syscall64) $(syshdr)
+	$(call if_changed,syshdr)
+
 $(out)/syscalls_32.h: $(syscall32) $(systbl)
 	$(call if_changed,systbl)
 $(out)/syscalls_64.h: $(syscall64) $(systbl)
@@ -41,7 +46,7 @@ $(out)/syscalls_64.h: $(syscall64) $(systbl)
 
 syshdr-y			+= unistd_32.h unistd_64.h unistd_x32.h
 syshdr-y			+= syscalls_32.h
-syshdr-$(CONFIG_X86_64)		+= unistd_32_ia32.h
+syshdr-$(CONFIG_X86_64)		+= unistd_32_ia32.h unistd_64_x32.h
 syshdr-$(CONFIG_X86_64)		+= syscalls_64.h
 
 targets	+= $(syshdr-y)
-- 
1.7.6.5

[PATCH 20/30] x86: Move some signal-handling definitions to a common header

[H. Peter Anvin]

From: "H. Peter Anvin" <hpa@zytor.com>

There are some definitions which are duplicated between
kernel/signal.c and ia32/ia32_signal.c; move them to a common header
file.

Rather than adding stuff to existing header files which contain data
structures, create a new header file; hence the slightly odd name
("all the good ones were taken.")

Note: nothing relied on signal_fault() being defined in
<asm/ptrace.h>.

Signed-off-by: H. Peter Anvin <hpa@zytor.com>
---
 arch/x86/ia32/ia32_signal.c        |   12 ++----------
 arch/x86/include/asm/ptrace.h      |    1 -
 arch/x86/include/asm/sighandling.h |   19 +++++++++++++++++++
 arch/x86/kernel/signal.c           |   10 +---------
 4 files changed, 22 insertions(+), 20 deletions(-)
 create mode 100644 arch/x86/include/asm/sighandling.h

diff --git a/arch/x86/ia32/ia32_signal.c b/arch/x86/ia32/ia32_signal.c
index 6557769..25d80f3 100644
--- a/arch/x86/ia32/ia32_signal.c
+++ b/arch/x86/ia32/ia32_signal.c
@@ -12,10 +12,8 @@
 #include <linux/mm.h>
 #include <linux/smp.h>
 #include <linux/kernel.h>
-#include <linux/signal.h>
 #include <linux/errno.h>
 #include <linux/wait.h>
-#include <linux/ptrace.h>
 #include <linux/unistd.h>
 #include <linux/stddef.h>
 #include <linux/personality.h>
@@ -31,16 +29,10 @@
 #include <asm/proto.h>
 #include <asm/vdso.h>
 #include <asm/sigframe.h>
+#include <asm/sighandling.h>
 #include <asm/sys_ia32.h>
 
-#define _BLOCKABLE (~(sigmask(SIGKILL) | sigmask(SIGSTOP)))
-
-#define FIX_EFLAGS	(X86_EFLAGS_AC | X86_EFLAGS_OF | \
-			 X86_EFLAGS_DF | X86_EFLAGS_TF | X86_EFLAGS_SF | \
-			 X86_EFLAGS_ZF | X86_EFLAGS_AF | X86_EFLAGS_PF | \
-			 X86_EFLAGS_CF)
-
-void signal_fault(struct pt_regs *regs, void __user *frame, char *where);
+#define FIX_EFLAGS	__FIX_EFLAGS
 
 int copy_siginfo_to_user32(compat_siginfo_t __user *to, siginfo_t *from)
 {
diff --git a/arch/x86/include/asm/ptrace.h b/arch/x86/include/asm/ptrace.h
index 3566454..dcfde52 100644
--- a/arch/x86/include/asm/ptrace.h
+++ b/arch/x86/include/asm/ptrace.h
@@ -145,7 +145,6 @@ extern unsigned long
 convert_ip_to_linear(struct task_struct *child, struct pt_regs *regs);
 extern void send_sigtrap(struct task_struct *tsk, struct pt_regs *regs,
 			 int error_code, int si_code);
-void signal_fault(struct pt_regs *regs, void __user *frame, char *where);
 
 extern long syscall_trace_enter(struct pt_regs *);
 extern void syscall_trace_leave(struct pt_regs *);
diff --git a/arch/x86/include/asm/sighandling.h b/arch/x86/include/asm/sighandling.h
new file mode 100644
index 0000000..843e299
--- /dev/null
+++ b/arch/x86/include/asm/sighandling.h
@@ -0,0 +1,19 @@
+#ifndef _ASM_X86_SIGHANDLING_H
+#define _ASM_X86_SIGHANDLING_H
+
+#include <linux/compiler.h>
+#include <linux/ptrace.h>
+#include <linux/signal.h>
+
+#include <asm/processor-flags.h>
+
+#define _BLOCKABLE (~(sigmask(SIGKILL) | sigmask(SIGSTOP)))
+
+#define __FIX_EFLAGS	(X86_EFLAGS_AC | X86_EFLAGS_OF | \
+			 X86_EFLAGS_DF | X86_EFLAGS_TF | X86_EFLAGS_SF | \
+			 X86_EFLAGS_ZF | X86_EFLAGS_AF | X86_EFLAGS_PF | \
+			 X86_EFLAGS_CF)
+
+void signal_fault(struct pt_regs *regs, void __user *frame, char *where);
+
+#endif /* _ASM_X86_SIGHANDLING_H */
diff --git a/arch/x86/kernel/signal.c b/arch/x86/kernel/signal.c
index 46a01bd..c432dc0 100644
--- a/arch/x86/kernel/signal.c
+++ b/arch/x86/kernel/signal.c
@@ -10,10 +10,8 @@
 #include <linux/mm.h>
 #include <linux/smp.h>
 #include <linux/kernel.h>
-#include <linux/signal.h>
 #include <linux/errno.h>
 #include <linux/wait.h>
-#include <linux/ptrace.h>
 #include <linux/tracehook.h>
 #include <linux/unistd.h>
 #include <linux/stddef.h>
@@ -26,6 +24,7 @@
 #include <asm/i387.h>
 #include <asm/vdso.h>
 #include <asm/mce.h>
+#include <asm/sighandling.h>
 
 #ifdef CONFIG_X86_64
 #include <asm/proto.h>
@@ -37,13 +36,6 @@
 
 #include <asm/sigframe.h>
 
-#define _BLOCKABLE (~(sigmask(SIGKILL) | sigmask(SIGSTOP)))
-
-#define __FIX_EFLAGS	(X86_EFLAGS_AC | X86_EFLAGS_OF | \
-			 X86_EFLAGS_DF | X86_EFLAGS_TF | X86_EFLAGS_SF | \
-			 X86_EFLAGS_ZF | X86_EFLAGS_AF | X86_EFLAGS_PF | \
-			 X86_EFLAGS_CF)
-
 #ifdef CONFIG_X86_32
 # define FIX_EFLAGS	(__FIX_EFLAGS | X86_EFLAGS_RF)
 #else
-- 
1.7.6.5

[PATCH 21/30] x32: Export setup/restore_sigcontext from signal.c

[H. Peter Anvin]

From: "H. Peter Anvin" <hpa@zytor.com>

Export setup_sigcontext() and restore_sigcontext() from signal.c, so
we can use the 64-bit versions verbatim for x32.

Signed-off-by: H. Peter Anvin <hpa@zytor.com>
---
 arch/x86/include/asm/sighandling.h |    5 +++++
 arch/x86/kernel/signal.c           |   10 ++++------
 2 files changed, 9 insertions(+), 6 deletions(-)

diff --git a/arch/x86/include/asm/sighandling.h b/arch/x86/include/asm/sighandling.h
index 843e299..ada93b3 100644
--- a/arch/x86/include/asm/sighandling.h
+++ b/arch/x86/include/asm/sighandling.h
@@ -16,4 +16,9 @@
 
 void signal_fault(struct pt_regs *regs, void __user *frame, char *where);
 
+int restore_sigcontext(struct pt_regs *regs, struct sigcontext __user *sc,
+		       unsigned long *pax);
+int setup_sigcontext(struct sigcontext __user *sc, void __user *fpstate,
+		     struct pt_regs *regs, unsigned long mask);
+
 #endif /* _ASM_X86_SIGHANDLING_H */
diff --git a/arch/x86/kernel/signal.c b/arch/x86/kernel/signal.c
index c432dc0..450fb25 100644
--- a/arch/x86/kernel/signal.c
+++ b/arch/x86/kernel/signal.c
@@ -60,9 +60,8 @@
 	regs->seg = GET_SEG(seg) | 3;			\
 } while (0)
 
-static int
-restore_sigcontext(struct pt_regs *regs, struct sigcontext __user *sc,
-		   unsigned long *pax)
+int restore_sigcontext(struct pt_regs *regs, struct sigcontext __user *sc,
+		       unsigned long *pax)
 {
 	void __user *buf;
 	unsigned int tmpflags;
@@ -117,9 +116,8 @@ restore_sigcontext(struct pt_regs *regs, struct sigcontext __user *sc,
 	return err;
 }
 
-static int
-setup_sigcontext(struct sigcontext __user *sc, void __user *fpstate,
-		 struct pt_regs *regs, unsigned long mask)
+int setup_sigcontext(struct sigcontext __user *sc, void __user *fpstate,
+		     struct pt_regs *regs, unsigned long mask)
 {
 	int err = 0;
 
-- 
1.7.6.5

[PATCH 22/30] x32: Add struct ucontext_x32

[H. Peter Anvin]

From: "H. J. Lu" <hjl.tools@gmail.com>

Add a definition for struct ucontext_x32; this is inherently a mix of
the 32- and 64-bit versions.

Signed-off-by: H. Peter Anvin <hpa@zytor.com>
---
 arch/x86/include/asm/ia32.h |    9 +++++++++
 1 files changed, 9 insertions(+), 0 deletions(-)

diff --git a/arch/x86/include/asm/ia32.h b/arch/x86/include/asm/ia32.h
index 1f7e625..c6435ab 100644
--- a/arch/x86/include/asm/ia32.h
+++ b/arch/x86/include/asm/ia32.h
@@ -43,6 +43,15 @@ struct ucontext_ia32 {
 	compat_sigset_t	  uc_sigmask;	/* mask last for extensibility */
 };
 
+struct ucontext_x32 {
+	unsigned int	  uc_flags;
+	unsigned int 	  uc_link;
+	stack_ia32_t	  uc_stack;
+	unsigned int	  uc__pad0;     /* needed for alignment */
+	struct sigcontext uc_mcontext;  /* the 64-bit sigcontext type */
+	compat_sigset_t	  uc_sigmask;	/* mask last for extensibility */
+};
+
 /* This matches struct stat64 in glibc2.2, hence the absolutely
  * insane amounts of padding around dev_t's.
  */
-- 
1.7.6.5

[PATCH 23/30] x32: Add rt_sigframe_x32

[H. Peter Anvin]

From: "H. Peter Anvin" <hpa@zytor.com>

Add rt_sigframe_x32 to <asm/sigframe.h>.  Unfortunately we can't just
define all the data structures unconditionally, due to the #ifdef
CONFIG_COMPAT in <linux/compat.h> and its trickle-down effects, hence
the #ifdef mess.

Signed-off-by: H. Peter Anvin <hpa@zytor.com>
---
 arch/x86/include/asm/sigframe.h |   13 +++++++++++++
 1 files changed, 13 insertions(+), 0 deletions(-)

diff --git a/arch/x86/include/asm/sigframe.h b/arch/x86/include/asm/sigframe.h
index 4e0fe26..7c7c27c 100644
--- a/arch/x86/include/asm/sigframe.h
+++ b/arch/x86/include/asm/sigframe.h
@@ -59,12 +59,25 @@ struct rt_sigframe_ia32 {
 #endif /* defined(CONFIG_X86_32) || defined(CONFIG_IA32_EMULATION) */
 
 #ifdef CONFIG_X86_64
+
 struct rt_sigframe {
 	char __user *pretcode;
 	struct ucontext uc;
 	struct siginfo info;
 	/* fp state follows here */
 };
+
+#ifdef CONFIG_X86_X32_ABI
+
+struct rt_sigframe_x32 {
+	u64 pretcode;
+	struct ucontext_x32 uc;
+	compat_siginfo_t info;
+	/* fp state follows here */
+};
+
+#endif /* CONFIG_X86_X32_ABI */
+
 #endif /* CONFIG_X86_64 */
 
 #endif /* _ASM_X86_SIGFRAME_H */
-- 
1.7.6.5

[PATCH 24/30] x32: Handle the x32 system call flag

[H. Peter Anvin]

From: "H. Peter Anvin" <hpa@zytor.com>

x32 shares most system calls with x86-64, but unfortunately some
subsystem (the input subsystem is the chief offender) which require
is_compat() when operating with a 32-bit userspace.  The input system
actually has text files in sysfs whose meaning is dependent on
sizeof(long) in userspace!

We could solve this by having two completely disjoint system call
tables; requiring that each system call be duplicated.  This patch
takes a different approach: we add a flag to the system call number;
this flag doesn't affect the system call dispatch but requests compat
treatment from affected subsystems for the duration of the system call.

The change of cmpq to cmpl is safe since it immediately follows the
and.

Signed-off-by: H. Peter Anvin <hpa@zytor.com>
---
 arch/x86/include/asm/compat.h  |   13 +++++++++++--
 arch/x86/include/asm/syscall.h |    5 +++--
 arch/x86/include/asm/unistd.h  |    7 +++++++
 arch/x86/kernel/entry_64.S     |   10 ++++++++++
 4 files changed, 31 insertions(+), 4 deletions(-)

diff --git a/arch/x86/include/asm/compat.h b/arch/x86/include/asm/compat.h
index 30d737e..7938b84 100644
--- a/arch/x86/include/asm/compat.h
+++ b/arch/x86/include/asm/compat.h
@@ -7,6 +7,7 @@
 #include <linux/types.h>
 #include <linux/sched.h>
 #include <asm/user32.h>
+#include <asm/unistd.h>
 
 #define COMPAT_USER_HZ		100
 #define COMPAT_UTS_MACHINE	"i686\0\0"
@@ -212,9 +213,17 @@ static inline void __user *arch_compat_alloc_user_space(long len)
 	return (void __user *)regs->sp - len;
 }
 
-static inline int is_compat_task(void)
+static inline bool is_compat_task(void)
 {
-	return current_thread_info()->status & TS_COMPAT;
+#ifdef CONFIG_IA32_EMULATION
+	if (current_thread_info()->status & TS_COMPAT)
+		return true;
+#endif
+#ifdef CONFIG_X86_X32_ABI
+	if (task_pt_regs(current)->orig_ax & __X32_SYSCALL_BIT)
+		return true;
+#endif
+	return false;
 }
 
 #endif /* _ASM_X86_COMPAT_H */
diff --git a/arch/x86/include/asm/syscall.h b/arch/x86/include/asm/syscall.h
index d962e56..386b786 100644
--- a/arch/x86/include/asm/syscall.h
+++ b/arch/x86/include/asm/syscall.h
@@ -16,6 +16,7 @@
 #include <linux/sched.h>
 #include <linux/err.h>
 #include <asm/asm-offsets.h>	/* For NR_syscalls */
+#include <asm/unistd.h>
 
 extern const unsigned long sys_call_table[];
 
@@ -26,13 +27,13 @@ extern const unsigned long sys_call_table[];
  */
 static inline int syscall_get_nr(struct task_struct *task, struct pt_regs *regs)
 {
-	return regs->orig_ax;
+	return regs->orig_ax & __SYSCALL_MASK;
 }
 
 static inline void syscall_rollback(struct task_struct *task,
 				    struct pt_regs *regs)
 {
-	regs->ax = regs->orig_ax;
+	regs->ax = regs->orig_ax & __SYSCALL_MASK;
 }
 
 static inline long syscall_get_error(struct task_struct *task,
diff --git a/arch/x86/include/asm/unistd.h b/arch/x86/include/asm/unistd.h
index 7a48a55..37cdc9d 100644
--- a/arch/x86/include/asm/unistd.h
+++ b/arch/x86/include/asm/unistd.h
@@ -5,6 +5,13 @@
 #define __X32_SYSCALL_BIT	0x40000000
 
 #ifdef __KERNEL__
+
+# ifdef CONFIG_X86_X32_ABI
+#  define __SYSCALL_MASK (~(__X32_SYSCALL_BIT))
+# else
+#  define __SYSCALL_MASK (~0)
+# endif
+
 # ifdef CONFIG_X86_32
 
 #  include <asm/unistd_32.h>
diff --git a/arch/x86/kernel/entry_64.S b/arch/x86/kernel/entry_64.S
index 3fe8239..a17b342 100644
--- a/arch/x86/kernel/entry_64.S
+++ b/arch/x86/kernel/entry_64.S
@@ -482,7 +482,12 @@ GLOBAL(system_call_after_swapgs)
 	testl $_TIF_WORK_SYSCALL_ENTRY,TI_flags+THREAD_INFO(%rsp,RIP-ARGOFFSET)
 	jnz tracesys
 system_call_fastpath:
+#if __SYSCALL_MASK == ~0
 	cmpq $__NR_syscall_max,%rax
+#else
+	andl $__SYSCALL_MASK,%eax
+	cmpl $__NR_syscall_max,%eax
+#endif
 	ja badsys
 	movq %r10,%rcx
 	call *sys_call_table(,%rax,8)  # XXX:	 rip relative
@@ -596,7 +601,12 @@ tracesys:
 	 */
 	LOAD_ARGS ARGOFFSET, 1
 	RESTORE_REST
+#if __SYSCALL_MASK == ~0
 	cmpq $__NR_syscall_max,%rax
+#else
+	andl $__SYSCALL_MASK,%eax
+	cmpl $__NR_syscall_max,%eax
+#endif
 	ja   int_ret_from_sys_call	/* RAX(%rsp) set to -ENOSYS above */
 	movq %r10,%rcx	/* fixup for C */
 	call *sys_call_table(,%rax,8)
-- 
1.7.6.5

[PATCH 25/30] x86: Add #ifdef CONFIG_COMPAT to <asm/sys_ia32.h>

[H. Peter Anvin]

From: "H. Peter Anvin" <hpa@zytor.com>

Unfortunately a lot of the compat types are guarded with CONFIG_COMPAT
or the equivalent, so add a similar guard to <asm/sys_ia32.h> to avoid
compilation failures when CONFIG_COMPAT=n.

Signed-off-by: H. Peter Anvin <hpa@zytor.com>
---
 arch/x86/include/asm/sys_ia32.h |    5 +++++
 1 files changed, 5 insertions(+), 0 deletions(-)

diff --git a/arch/x86/include/asm/sys_ia32.h b/arch/x86/include/asm/sys_ia32.h
index 68da87b..3fda9db4 100644
--- a/arch/x86/include/asm/sys_ia32.h
+++ b/arch/x86/include/asm/sys_ia32.h
@@ -10,6 +10,8 @@
 #ifndef _ASM_X86_SYS_IA32_H
 #define _ASM_X86_SYS_IA32_H
 
+#ifdef CONFIG_COMPAT
+
 #include <linux/compiler.h>
 #include <linux/linkage.h>
 #include <linux/types.h>
@@ -81,4 +83,7 @@ asmlinkage long sys32_ipc(u32, int, int, int, compat_uptr_t, u32);
 
 asmlinkage long sys32_fanotify_mark(int, unsigned int, u32, u32, int,
 				    const char __user *);
+
+#endif /* CONFIG_COMPAT */
+
 #endif /* _ASM_X86_SYS_IA32_H */
-- 
1.7.6.5

[PATCH 26/30] x32: Signal-related system calls

[H. Peter Anvin]

From: "H. Peter Anvin" <hpa@zytor.com>

x32 uses the 64-bit signal frame format, obviously, but there are some
structures which mixes that with pointers or sizeof(long) types, as
such we have to create a handful of system calls specific to x32.  By
and large these are a mixture of the 64-bit and the compat system
calls.

Originally-by: H. J. Lu <hjl.tools@gmail.com>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
---
 arch/x86/kernel/entry_64.S |   19 +++++++
 arch/x86/kernel/signal.c   |  118 +++++++++++++++++++++++++++++++++++++++++++-
 2 files changed, 136 insertions(+), 1 deletions(-)

diff --git a/arch/x86/kernel/entry_64.S b/arch/x86/kernel/entry_64.S
index a17b342..53dc821 100644
--- a/arch/x86/kernel/entry_64.S
+++ b/arch/x86/kernel/entry_64.S
@@ -746,6 +746,25 @@ ENTRY(stub_rt_sigreturn)
 	CFI_ENDPROC
 END(stub_rt_sigreturn)
 
+#ifdef CONFIG_X86_X32_ABI
+	PTREGSCALL stub_x32_sigaltstack, sys32_sigaltstack, %rdx
+
+ENTRY(stub_x32_rt_sigreturn)
+	CFI_STARTPROC
+	addq $8, %rsp
+	PARTIAL_FRAME 0
+	SAVE_REST
+	movq %rsp,%rdi
+	FIXUP_TOP_OF_STACK %r11
+	call sys32_x32_rt_sigreturn
+	movq %rax,RAX(%rsp) # fixme, this could be done at the higher layer
+	RESTORE_REST
+	jmp int_ret_from_sys_call
+	CFI_ENDPROC
+END(stub_x32_rt_sigreturn)
+
+#endif
+
 /*
  * Build the entry stubs and pointer table with some assembler magic.
  * We pack 7 stubs into a single 32-byte chunk, which will fit in a
diff --git a/arch/x86/kernel/signal.c b/arch/x86/kernel/signal.c
index 450fb25..c3846b6 100644
--- a/arch/x86/kernel/signal.c
+++ b/arch/x86/kernel/signal.c
@@ -29,6 +29,7 @@
 #ifdef CONFIG_X86_64
 #include <asm/proto.h>
 #include <asm/ia32_unistd.h>
+#include <asm/sys_ia32.h>
 #endif /* CONFIG_X86_64 */
 
 #include <asm/syscall.h>
@@ -632,6 +633,16 @@ static int signr_convert(int sig)
 #define is_ia32	0
 #endif /* CONFIG_IA32_EMULATION */
 
+#ifdef CONFIG_X86_X32_ABI
+#define is_x32	test_thread_flag(TIF_X32)
+
+static int x32_setup_rt_frame(int sig, struct k_sigaction *ka,
+			      siginfo_t *info, compat_sigset_t *set,
+			      struct pt_regs *regs);
+#else /* !CONFIG_X86_X32_ABI */
+#define is_x32	0
+#endif /* CONFIG_X86_X32_ABI */
+
 int ia32_setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info,
 		sigset_t *set, struct pt_regs *regs);
 int ia32_setup_frame(int sig, struct k_sigaction *ka,
@@ -656,8 +667,14 @@ setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info,
 			ret = ia32_setup_rt_frame(usig, ka, info, set, regs);
 		else
 			ret = ia32_setup_frame(usig, ka, set, regs);
-	} else
+#ifdef CONFIG_X86_X32_ABI
+	} else if (is_x32) {
+		ret = x32_setup_rt_frame(usig, ka, info,
+					 (compat_sigset_t *)set, regs);
+#endif
+	} else {
 		ret = __setup_rt_frame(sig, ka, info, set, regs);
+	}
 
 	if (ret) {
 		force_sigsegv(sig, current);
@@ -840,3 +857,102 @@ void signal_fault(struct pt_regs *regs, void __user *frame, char *where)
 
 	force_sig(SIGSEGV, me);
 }
+
+#ifdef CONFIG_X86_X32_ABI
+static int x32_setup_rt_frame(int sig, struct k_sigaction *ka,
+			      siginfo_t *info, compat_sigset_t *set,
+			      struct pt_regs *regs)
+{
+	struct rt_sigframe_x32 __user *frame;
+	void __user *restorer;
+	int err = 0;
+	void __user *fpstate = NULL;
+
+	frame = get_sigframe(ka, regs, sizeof(*frame), &fpstate);
+
+	if (!access_ok(VERIFY_WRITE, frame, sizeof(*frame)))
+		return -EFAULT;
+
+	if (ka->sa.sa_flags & SA_SIGINFO) {
+		if (copy_siginfo_to_user32(&frame->info, info))
+			return -EFAULT;
+	}
+
+	put_user_try {
+		/* Create the ucontext.  */
+		if (cpu_has_xsave)
+			put_user_ex(UC_FP_XSTATE, &frame->uc.uc_flags);
+		else
+			put_user_ex(0, &frame->uc.uc_flags);
+		put_user_ex(0, &frame->uc.uc_link);
+		put_user_ex(current->sas_ss_sp, &frame->uc.uc_stack.ss_sp);
+		put_user_ex(sas_ss_flags(regs->sp),
+			    &frame->uc.uc_stack.ss_flags);
+		put_user_ex(current->sas_ss_size, &frame->uc.uc_stack.ss_size);
+		put_user_ex(0, &frame->uc.uc__pad0);
+		err |= setup_sigcontext(&frame->uc.uc_mcontext, fpstate,
+					regs, set->sig[0]);
+		err |= __copy_to_user(&frame->uc.uc_sigmask, set, sizeof(*set));
+
+		if (ka->sa.sa_flags & SA_RESTORER) {
+			restorer = ka->sa.sa_restorer;
+		} else {
+			/* could use a vstub here */
+			restorer = NULL;
+			err |= -EFAULT;
+		}
+		put_user_ex(restorer, &frame->pretcode);
+	} put_user_catch(err);
+
+	if (err)
+		return -EFAULT;
+
+	/* Set up registers for signal handler */
+	regs->sp = (unsigned long) frame;
+	regs->ip = (unsigned long) ka->sa.sa_handler;
+
+	/* We use the x32 calling convention here... */
+	regs->di = sig;
+	regs->si = (unsigned long) &frame->info;
+	regs->dx = (unsigned long) &frame->uc;
+
+	loadsegment(ds, __USER_DS);
+	loadsegment(es, __USER_DS);
+
+	regs->cs = __USER_CS;
+	regs->ss = __USER_DS;
+
+	return 0;
+}
+
+asmlinkage long sys32_x32_rt_sigreturn(struct pt_regs *regs)
+{
+	struct rt_sigframe_x32 __user *frame;
+	sigset_t set;
+	unsigned long ax;
+	struct pt_regs tregs;
+
+	frame = (struct rt_sigframe_x32 __user *)(regs->sp - 8);
+
+	if (!access_ok(VERIFY_READ, frame, sizeof(*frame)))
+		goto badframe;
+	if (__copy_from_user(&set, &frame->uc.uc_sigmask, sizeof(set)))
+		goto badframe;
+
+	sigdelsetmask(&set, ~_BLOCKABLE);
+	set_current_blocked(&set);
+
+	if (restore_sigcontext(regs, &frame->uc.uc_mcontext, &ax))
+		goto badframe;
+
+	tregs = *regs;
+	if (sys32_sigaltstack(&frame->uc.uc_stack, NULL, &tregs) == -EFAULT)
+		goto badframe;
+
+	return ax;
+
+badframe:
+	signal_fault(regs, frame, "x32 rt_sigreturn");
+	return 0;
+}
+#endif
-- 
1.7.6.5

[PATCH 27/30] x32: Handle process creation

[H. Peter Anvin]

From: "H. Peter Anvin" <hpa@zytor.com>

Allow an x32 process to be started.

Originally-by: H. J. Lu <hjl.tools@gmail.com>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
---
 arch/x86/include/asm/compat.h    |   26 ++++++++++++++++++++++++--
 arch/x86/include/asm/elf.h       |   25 +++++++++++++++++++++----
 arch/x86/kernel/cpu/perf_event.c |    4 +++-
 arch/x86/kernel/entry_64.S       |   15 +++++++++++++++
 arch/x86/kernel/process_64.c     |   23 ++++++++++++++++-------
 5 files changed, 79 insertions(+), 14 deletions(-)

diff --git a/arch/x86/include/asm/compat.h b/arch/x86/include/asm/compat.h
index 7938b84..e7f68b4 100644
--- a/arch/x86/include/asm/compat.h
+++ b/arch/x86/include/asm/compat.h
@@ -6,6 +6,7 @@
  */
 #include <linux/types.h>
 #include <linux/sched.h>
+#include <asm/processor.h>
 #include <asm/user32.h>
 #include <asm/unistd.h>
 
@@ -187,7 +188,20 @@ struct compat_shmid64_ds {
 /*
  * The type of struct elf_prstatus.pr_reg in compatible core dumps.
  */
+#ifdef CONFIG_X86_X32_ABI
+typedef struct user_regs_struct compat_elf_gregset_t;
+
+#define PR_REG_SIZE(S) (test_thread_flag(TIF_IA32) ? 68 : 216)
+#define PRSTATUS_SIZE(S) (test_thread_flag(TIF_IA32) ? 144 : 296)
+#define SET_PR_FPVALID(S,V) \
+  do { *(int *) (((void *) &((S)->pr_reg)) + PR_REG_SIZE(0)) = (V); } \
+  while (0)
+
+#define COMPAT_USE_64BIT_TIME \
+	(!!(task_pt_regs(current)->orig_ax & __X32_SYSCALL_BIT))
+#else
 typedef struct user_regs_struct32 compat_elf_gregset_t;
+#endif
 
 /*
  * A pointer passed in from user mode. This should not
@@ -209,8 +223,16 @@ static inline compat_uptr_t ptr_to_compat(void __user *uptr)
 
 static inline void __user *arch_compat_alloc_user_space(long len)
 {
-	struct pt_regs *regs = task_pt_regs(current);
-	return (void __user *)regs->sp - len;
+	compat_uptr_t sp;
+
+	if (test_thread_flag(TIF_IA32)) {
+		sp = task_pt_regs(current)->sp;
+	} else {
+		/* -128 for the x32 ABI redzone */
+		sp = percpu_read(old_rsp) - 128;
+	}
+
+	return (void __user *)round_down(sp - len, 16);
 }
 
 static inline bool is_compat_task(void)
diff --git a/arch/x86/include/asm/elf.h b/arch/x86/include/asm/elf.h
index 410fa6a..83aabea 100644
--- a/arch/x86/include/asm/elf.h
+++ b/arch/x86/include/asm/elf.h
@@ -156,7 +156,12 @@ do {						\
 #define elf_check_arch(x)			\
 	((x)->e_machine == EM_X86_64)
 
-#define compat_elf_check_arch(x)	elf_check_arch_ia32(x)
+#define compat_elf_check_arch(x)		\
+	(elf_check_arch_ia32(x) || (x)->e_machine == EM_X86_64)
+
+#if __USER32_DS != __USER_DS
+# error "The following code assumes __USER32_DS == __USER_DS"
+#endif
 
 static inline void elf_common_init(struct thread_struct *t,
 				   struct pt_regs *regs, const u16 ds)
@@ -179,8 +184,9 @@ static inline void elf_common_init(struct thread_struct *t,
 void start_thread_ia32(struct pt_regs *regs, u32 new_ip, u32 new_sp);
 #define compat_start_thread start_thread_ia32
 
-void set_personality_ia32(void);
-#define COMPAT_SET_PERSONALITY(ex) set_personality_ia32()
+void set_personality_ia32(bool);
+#define COMPAT_SET_PERSONALITY(ex)			\
+	set_personality_ia32((ex).e_machine == EM_X86_64)
 
 #define COMPAT_ELF_PLATFORM			("i686")
 
@@ -296,9 +302,20 @@ do {									\
 			    (unsigned long)current->mm->context.vdso);	\
 } while (0)
 
+#define ARCH_DLINFO_X32							\
+do {									\
+	if (vdso_enabled)						\
+		NEW_AUX_ENT(AT_SYSINFO_EHDR,				\
+			    (unsigned long)current->mm->context.vdso);	\
+} while (0)
+
 #define AT_SYSINFO		32
 
-#define COMPAT_ARCH_DLINFO	ARCH_DLINFO_IA32(sysctl_vsyscall32)
+#define COMPAT_ARCH_DLINFO						\
+if (test_thread_flag(TIF_X32))						\
+	ARCH_DLINFO_X32;						\
+else									\
+	ARCH_DLINFO_IA32(sysctl_vsyscall32)
 
 #define COMPAT_ELF_ET_DYN_BASE	(TASK_UNMAPPED_BASE + 0x1000000)
 
diff --git a/arch/x86/kernel/cpu/perf_event.c b/arch/x86/kernel/cpu/perf_event.c
index 5adce10..63c0e05 100644
--- a/arch/x86/kernel/cpu/perf_event.c
+++ b/arch/x86/kernel/cpu/perf_event.c
@@ -28,7 +28,6 @@
 #include <asm/apic.h>
 #include <asm/stacktrace.h>
 #include <asm/nmi.h>
-#include <asm/compat.h>
 #include <asm/smp.h>
 #include <asm/alternative.h>
 
@@ -1595,6 +1594,9 @@ perf_callchain_kernel(struct perf_callchain_entry *entry, struct pt_regs *regs)
 }
 
 #ifdef CONFIG_COMPAT
+
+#include <asm/compat.h>
+
 static inline int
 perf_callchain_user32(struct pt_regs *regs, struct perf_callchain_entry *entry)
 {
diff --git a/arch/x86/kernel/entry_64.S b/arch/x86/kernel/entry_64.S
index 53dc821..9e036f0 100644
--- a/arch/x86/kernel/entry_64.S
+++ b/arch/x86/kernel/entry_64.S
@@ -763,6 +763,21 @@ ENTRY(stub_x32_rt_sigreturn)
 	CFI_ENDPROC
 END(stub_x32_rt_sigreturn)
 
+ENTRY(stub_x32_execve)
+	CFI_STARTPROC
+	addq $8, %rsp
+	PARTIAL_FRAME 0
+	SAVE_REST
+	FIXUP_TOP_OF_STACK %r11
+	movq %rsp, %rcx
+	call sys32_execve
+	RESTORE_TOP_OF_STACK %r11
+	movq %rax,RAX(%rsp)
+	RESTORE_REST
+	jmp int_ret_from_sys_call
+	CFI_ENDPROC
+END(stub_x32_execve)
+
 #endif
 
 /*
diff --git a/arch/x86/kernel/process_64.c b/arch/x86/kernel/process_64.c
index 5fe2fba..a0701da 100644
--- a/arch/x86/kernel/process_64.c
+++ b/arch/x86/kernel/process_64.c
@@ -364,7 +364,9 @@ start_thread(struct pt_regs *regs, unsigned long new_ip, unsigned long new_sp)
 void start_thread_ia32(struct pt_regs *regs, u32 new_ip, u32 new_sp)
 {
 	start_thread_common(regs, new_ip, new_sp,
-			    __USER32_CS, __USER32_DS, __USER32_DS);
+			    test_thread_flag(TIF_X32)
+			    ? __USER_CS : __USER32_CS,
+			    __USER_DS, __USER_DS);
 }
 #endif
 
@@ -508,6 +510,7 @@ void set_personality_64bit(void)
 
 	/* Make sure to be in 64bit mode */
 	clear_thread_flag(TIF_IA32);
+	clear_thread_flag(TIF_X32);
 	clear_thread_flag(TIF_ADDR32);
 	clear_thread_flag(TIF_X32);
 
@@ -522,22 +525,28 @@ void set_personality_64bit(void)
 	current->personality &= ~READ_IMPLIES_EXEC;
 }
 
-void set_personality_ia32(void)
+void set_personality_ia32(bool x32)
 {
 	/* inherit personality from parent */
 
 	/* Make sure to be in 32bit mode */
-	set_thread_flag(TIF_IA32);
 	set_thread_flag(TIF_ADDR32);
-	clear_thread_flag(TIF_X32);
-	current->personality |= force_personality32;
 
 	/* Mark the associated mm as containing 32-bit tasks. */
 	if (current->mm)
 		current->mm->context.ia32_compat = 1;
 
-	/* Prepare the first "return" to user space */
-	current_thread_info()->status |= TS_COMPAT;
+	if (x32) {
+		clear_thread_flag(TIF_IA32);
+		set_thread_flag(TIF_X32);
+		current->personality &= ~READ_IMPLIES_EXEC;
+	} else {
+		set_thread_flag(TIF_IA32);
+		clear_thread_flag(TIF_X32);
+		current->personality |= force_personality32;
+		/* Prepare the first "return" to user space */
+		current_thread_info()->status |= TS_COMPAT;
+	}
 }
 
 unsigned long get_wchan(struct task_struct *p)
-- 
1.7.6.5

[PATCH 28/30] x32: If configured, add x32 system calls to system call tables

[H. Peter Anvin]

From: "H. Peter Anvin" <hpa@zytor.com>

If CONFIG_X86_X32_ABI is defined, add the x32 system calls to the
system call tables.

Signed-off-by: H. Peter Anvin <hpa@zytor.com>
---
 arch/x86/kernel/asm-offsets_64.c |    6 +++++-
 arch/x86/kernel/syscall_64.c     |    7 ++++++-
 2 files changed, 11 insertions(+), 2 deletions(-)

diff --git a/arch/x86/kernel/asm-offsets_64.c b/arch/x86/kernel/asm-offsets_64.c
index c3354f7..1b4754f 100644
--- a/arch/x86/kernel/asm-offsets_64.c
+++ b/arch/x86/kernel/asm-offsets_64.c
@@ -2,7 +2,11 @@
 
 #define __SYSCALL_64(nr, sym, compat) [nr] = 1,
 #define __SYSCALL_COMMON(nr, sym, compat) [nr] = 1,
-#define __SYSCALL_X32(nr, sym, compat) /* Not yet */
+#ifdef CONFIG_X86_X32_ABI
+# define __SYSCALL_X32(nr, sym, compat) [nr] = 1,
+#else
+# define __SYSCALL_X32(nr, sym, compat) /* nothing */
+#endif
 static char syscalls_64[] = {
 #include <asm/syscalls_64.h>
 };
diff --git a/arch/x86/kernel/syscall_64.c b/arch/x86/kernel/syscall_64.c
index 26c4ca1..5c7f8c2 100644
--- a/arch/x86/kernel/syscall_64.c
+++ b/arch/x86/kernel/syscall_64.c
@@ -6,7 +6,12 @@
 #include <asm/asm-offsets.h>
 
 #define __SYSCALL_COMMON(nr, sym, compat) __SYSCALL_64(nr, sym, compat)
-#define __SYSCALL_X32(nr, sym, compat) /* Not yet */
+
+#ifdef CONFIG_X86_X32_ABI
+# define __SYSCALL_X32(nr, sym, compat) __SYSCALL_64(nr, sym, compat)
+#else
+# define __SYSCALL_X32(nr, sym, compat) /* nothing */
+#endif
 
 #define __SYSCALL_64(nr, sym, compat) extern asmlinkage void sym(void) ;
 #include <asm/syscalls_64.h>
-- 
1.7.6.5

[PATCH 29/30] x32: Allow x32 to be configured

[H. Peter Anvin]

From: "H. J. Lu" <hjl.tools@gmail.com>

At this point, one should be able to build an x32 kernel.

Note that for now we depend on CONFIG_IA32_EMULATION.  Long term, x32
and IA32 should be detangled.

Signed-off-by: H. Peter Anvin <hpa@zytor.com>
---
 arch/x86/Kconfig |   21 +++++++++++++++++----
 1 files changed, 17 insertions(+), 4 deletions(-)

diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig
index 5bed94e..c9d6c9e 100644
--- a/arch/x86/Kconfig
+++ b/arch/x86/Kconfig
@@ -2165,9 +2165,9 @@ config IA32_EMULATION
 	depends on X86_64
 	select COMPAT_BINFMT_ELF
 	---help---
-	  Include code to run 32-bit programs under a 64-bit kernel. You should
-	  likely turn this on, unless you're 100% sure that you don't have any
-	  32-bit programs left.
+	  Include code to run legacy 32-bit programs under a
+	  64-bit kernel. You should likely turn this on, unless you're
+	  100% sure that you don't have any 32-bit programs left.
 
 config IA32_AOUT
 	tristate "IA32 a.out support"
@@ -2175,9 +2175,22 @@ config IA32_AOUT
 	---help---
 	  Support old a.out binaries in the 32bit emulation.
 
+config X86_X32_ABI
+	bool "x32 ABI for 64-bit mode (EXPERIMENTAL)"
+	depends on X86_64 && IA32_EMULATION && EXPERIMENTAL
+	---help---
+	  Include code to run binaries for the x32 native 32-bit ABI
+	  for 64-bit processors.  An x32 process gets access to the
+	  full 64-bit register file and wide data path while leaving
+	  pointers at 32 bits for smaller memory footprint.
+
+	  You will need a recent binutils (2.22 or later) with
+	  elf32_x86_64 support enabled to compile a kernel with this
+	  option set.
+
 config COMPAT
 	def_bool y
-	depends on IA32_EMULATION
+	depends on IA32_EMULATION || X86_X32_ABI
 
 config COMPAT_FOR_U64_ALIGNMENT
 	def_bool COMPAT
-- 
1.7.6.5

[PATCH 30/30] x32: Add x32 VDSO support

[H. Peter Anvin]

From: "H. J. Lu" <hjl.tools@gmail.com>

Add support for the x32 VDSO.  The x32 VDSO takes advantage of the
similarity between the x86-64 and the x32 ABIs to contain the same
content, only the container is different, as the x32 VDSO obviously is
an x32 shared object.

Signed-off-by: H. Peter Anvin <hpa@zytor.com>
---
 arch/x86/vdso/.gitignore     |    2 +
 arch/x86/vdso/Makefile       |   46 ++++++++++++++++++++++++-
 arch/x86/vdso/vdso32-setup.c |    6 +++
 arch/x86/vdso/vdsox32.S      |   22 ++++++++++++
 arch/x86/vdso/vdsox32.lds.S  |   32 +++++++++++++++++
 arch/x86/vdso/vma.c          |   78 +++++++++++++++++++++++++++++++++++++----
 6 files changed, 177 insertions(+), 9 deletions(-)
 create mode 100644 arch/x86/vdso/vdsox32.S
 create mode 100644 arch/x86/vdso/vdsox32.lds.S

diff --git a/arch/x86/vdso/.gitignore b/arch/x86/vdso/.gitignore
index 60274d5..3282874 100644
--- a/arch/x86/vdso/.gitignore
+++ b/arch/x86/vdso/.gitignore
@@ -1,5 +1,7 @@
 vdso.lds
 vdso-syms.lds
+vdsox32.lds
+vdsox32-syms.lds
 vdso32-syms.lds
 vdso32-syscall-syms.lds
 vdso32-sysenter-syms.lds
diff --git a/arch/x86/vdso/Makefile b/arch/x86/vdso/Makefile
index 5d17950..fd14be1 100644
--- a/arch/x86/vdso/Makefile
+++ b/arch/x86/vdso/Makefile
@@ -3,21 +3,29 @@
 #
 
 VDSO64-$(CONFIG_X86_64)		:= y
+VDSOX32-$(CONFIG_X86_X32_ABI)	:= y
 VDSO32-$(CONFIG_X86_32)		:= y
 VDSO32-$(CONFIG_COMPAT)		:= y
 
 vdso-install-$(VDSO64-y)	+= vdso.so
+vdso-install-$(VDSOX32-y)	+= vdsox32.so
 vdso-install-$(VDSO32-y)	+= $(vdso32-images)
 
 
 # files to link into the vdso
 vobjs-y := vdso-note.o vclock_gettime.o vgetcpu.o
 
+vobjs-$(VDSOX32-y) += $(vobjx32s-compat)
+
+# Filter out x32 objects.
+vobj64s := $(filter-out $(vobjx32s-compat),$(vobjs-y))
+
 # files to link into kernel
 obj-$(VDSO64-y)			+= vma.o vdso.o
+obj-$(VDSOX32-y)		+= vdsox32.o
 obj-$(VDSO32-y)			+= vdso32.o vdso32-setup.o
 
-vobjs := $(foreach F,$(vobjs-y),$(obj)/$F)
+vobjs := $(foreach F,$(vobj64s),$(obj)/$F)
 
 $(obj)/vdso.o: $(obj)/vdso.so
 
@@ -73,6 +81,42 @@ $(obj)/%-syms.lds: $(obj)/%.so.dbg FORCE
 	$(call if_changed,vdsosym)
 
 #
+# X32 processes use x32 vDSO to access 64bit kernel data.
+#
+# Build x32 vDSO image:
+# 1. Compile x32 vDSO as 64bit.
+# 2. Convert object files to x32.
+# 3. Build x32 VDSO image with x32 objects, which contains 64bit codes
+# so that it can reach 64bit address space with 64bit pointers.
+#
+
+targets += vdsox32-syms.lds
+obj-$(VDSOX32-y)		+= vdsox32-syms.lds
+
+CPPFLAGS_vdsox32.lds = $(CPPFLAGS_vdso.lds)
+VDSO_LDFLAGS_vdsox32.lds = -Wl,-m,elf32_x86_64 \
+			   -Wl,-soname=linux-vdso.so.1 \
+			   -Wl,-z,max-page-size=4096 \
+			   -Wl,-z,common-page-size=4096
+
+vobjx32s-y := $(vobj64s:.o=-x32.o)
+vobjx32s := $(foreach F,$(vobjx32s-y),$(obj)/$F)
+
+# Convert 64bit object file to x32 for x32 vDSO.
+quiet_cmd_x32 = X32     $@
+      cmd_x32 = $(OBJCOPY) -O elf32-x86-64 $< $@
+
+$(obj)/%-x32.o: $(obj)/%.o FORCE
+	$(call if_changed,x32)
+
+targets += vdsox32.so vdsox32.so.dbg vdsox32.lds $(vobjx32s-y)
+
+$(obj)/vdsox32.o: $(src)/vdsox32.S $(obj)/vdsox32.so
+
+$(obj)/vdsox32.so.dbg: $(src)/vdsox32.lds $(vobjx32s) FORCE
+	$(call if_changed,vdso)
+
+#
 # Build multiple 32-bit vDSO images to choose from at boot time.
 #
 obj-$(VDSO32-y)			+= vdso32-syms.lds
diff --git a/arch/x86/vdso/vdso32-setup.c b/arch/x86/vdso/vdso32-setup.c
index 468d591..01b8a0d 100644
--- a/arch/x86/vdso/vdso32-setup.c
+++ b/arch/x86/vdso/vdso32-setup.c
@@ -317,6 +317,12 @@ int arch_setup_additional_pages(struct linux_binprm *bprm, int uses_interp)
 	int ret = 0;
 	bool compat;
 
+#ifdef CONFIG_X86_X32_ABI
+	extern int x32_setup_additional_pages(struct linux_binprm *, int);
+	if (test_thread_flag(TIF_X32))
+		return x32_setup_additional_pages (bprm, uses_interp);
+#endif
+
 	if (vdso_enabled == VDSO_DISABLED)
 		return 0;
 
diff --git a/arch/x86/vdso/vdsox32.S b/arch/x86/vdso/vdsox32.S
new file mode 100644
index 0000000..d6b9a7f
--- /dev/null
+++ b/arch/x86/vdso/vdsox32.S
@@ -0,0 +1,22 @@
+#include <asm/page_types.h>
+#include <linux/linkage.h>
+#include <linux/init.h>
+
+__PAGE_ALIGNED_DATA
+
+	.globl vdsox32_start, vdsox32_end
+	.align PAGE_SIZE
+vdsox32_start:
+	.incbin "arch/x86/vdso/vdsox32.so"
+vdsox32_end:
+	.align PAGE_SIZE /* extra data here leaks to userspace. */
+
+.previous
+
+	.globl vdsox32_pages
+	.bss
+	.align 8
+	.type vdsox32_pages, @object
+vdsox32_pages:
+	.zero (vdsox32_end - vdsox32_start + PAGE_SIZE - 1) / PAGE_SIZE * 8
+	.size vdsox32_pages, .-vdsox32_pages
diff --git a/arch/x86/vdso/vdsox32.lds.S b/arch/x86/vdso/vdsox32.lds.S
new file mode 100644
index 0000000..373ca9a
--- /dev/null
+++ b/arch/x86/vdso/vdsox32.lds.S
@@ -0,0 +1,32 @@
+/*
+ * Linker script for x32 vDSO.
+ * We #include the file to define the layout details.
+ * Here we only choose the prelinked virtual address.
+ *
+ * This file defines the version script giving the user-exported symbols in
+ * the DSO.  We can define local symbols here called VDSO* to make their
+ * values visible using the asm-x86/vdso.h macros from the kernel proper.
+ */
+
+#define VDSO_PRELINK 0
+#include "vdso-layout.lds.S"
+
+/*
+ * This controls what userland symbols we export from the vDSO.
+ */
+VERSION {
+	LINUX_2.6 {
+	global:
+		clock_gettime;
+		__vdso_clock_gettime;
+		gettimeofday;
+		__vdso_gettimeofday;
+		getcpu;
+		__vdso_getcpu;
+		time;
+		__vdso_time;
+	local: *;
+	};
+}
+
+VDSOX32_PRELINK = VDSO_PRELINK;
diff --git a/arch/x86/vdso/vma.c b/arch/x86/vdso/vma.c
index 153407c..1bbcc62 100644
--- a/arch/x86/vdso/vma.c
+++ b/arch/x86/vdso/vma.c
@@ -24,7 +24,44 @@ extern unsigned short vdso_sync_cpuid;
 extern struct page *vdso_pages[];
 static unsigned vdso_size;
 
-static void __init patch_vdso(void *vdso, size_t len)
+#ifdef CONFIG_X86_X32_ABI
+extern char vdsox32_start[], vdsox32_end[];
+extern struct page *vdsox32_pages[];
+static unsigned vdsox32_size;
+
+static void __init patch_vdsox32(void *vdso, size_t len)
+{
+	Elf32_Ehdr *hdr = vdso;
+	Elf32_Shdr *sechdrs, *alt_sec = 0;
+	char *secstrings;
+	void *alt_data;
+	int i;
+
+	BUG_ON(len < sizeof(Elf32_Ehdr));
+	BUG_ON(memcmp(hdr->e_ident, ELFMAG, SELFMAG) != 0);
+
+	sechdrs = (void *)hdr + hdr->e_shoff;
+	secstrings = (void *)hdr + sechdrs[hdr->e_shstrndx].sh_offset;
+
+	for (i = 1; i < hdr->e_shnum; i++) {
+		Elf32_Shdr *shdr = &sechdrs[i];
+		if (!strcmp(secstrings + shdr->sh_name, ".altinstructions")) {
+			alt_sec = shdr;
+			goto found;
+		}
+	}
+
+	/* If we get here, it's probably a bug. */
+	pr_warning("patch_vdsox32: .altinstructions not found\n");
+	return;  /* nothing to patch */
+
+found:
+	alt_data = (void *)hdr + alt_sec->sh_offset;
+	apply_alternatives(alt_data, alt_data + alt_sec->sh_size);
+}
+#endif
+
+static void __init patch_vdso64(void *vdso, size_t len)
 {
 	Elf64_Ehdr *hdr = vdso;
 	Elf64_Shdr *sechdrs, *alt_sec = 0;
@@ -47,7 +84,7 @@ static void __init patch_vdso(void *vdso, size_t len)
 	}
 
 	/* If we get here, it's probably a bug. */
-	pr_warning("patch_vdso: .altinstructions not found\n");
+	pr_warning("patch_vdso64: .altinstructions not found\n");
 	return;  /* nothing to patch */
 
 found:
@@ -60,12 +97,20 @@ static int __init init_vdso(void)
 	int npages = (vdso_end - vdso_start + PAGE_SIZE - 1) / PAGE_SIZE;
 	int i;
 
-	patch_vdso(vdso_start, vdso_end - vdso_start);
+	patch_vdso64(vdso_start, vdso_end - vdso_start);
 
 	vdso_size = npages << PAGE_SHIFT;
 	for (i = 0; i < npages; i++)
 		vdso_pages[i] = virt_to_page(vdso_start + i*PAGE_SIZE);
 
+#ifdef CONFIG_X86_X32_ABI
+	patch_vdsox32(vdsox32_start, vdsox32_end - vdsox32_start);
+	npages = (vdsox32_end - vdsox32_start + PAGE_SIZE - 1) / PAGE_SIZE;
+	vdsox32_size = npages << PAGE_SHIFT;
+	for (i = 0; i < npages; i++)
+		vdsox32_pages[i] = virt_to_page(vdsox32_start + i*PAGE_SIZE);
+#endif
+
 	return 0;
 }
 subsys_initcall(init_vdso);
@@ -103,7 +148,10 @@ static unsigned long vdso_addr(unsigned long start, unsigned len)
 
 /* Setup a VMA at program startup for the vsyscall page.
    Not called for compat tasks */
-int arch_setup_additional_pages(struct linux_binprm *bprm, int uses_interp)
+static int setup_additional_pages(struct linux_binprm *bprm,
+				  int uses_interp,
+				  struct page **pages,
+				  unsigned size)
 {
 	struct mm_struct *mm = current->mm;
 	unsigned long addr;
@@ -113,8 +161,8 @@ int arch_setup_additional_pages(struct linux_binprm *bprm, int uses_interp)
 		return 0;
 
 	down_write(&mm->mmap_sem);
-	addr = vdso_addr(mm->start_stack, vdso_size);
-	addr = get_unmapped_area(NULL, addr, vdso_size, 0, 0);
+	addr = vdso_addr(mm->start_stack, size);
+	addr = get_unmapped_area(NULL, addr, size, 0, 0);
 	if (IS_ERR_VALUE(addr)) {
 		ret = addr;
 		goto up_fail;
@@ -122,11 +170,11 @@ int arch_setup_additional_pages(struct linux_binprm *bprm, int uses_interp)
 
 	current->mm->context.vdso = (void *)addr;
 
-	ret = install_special_mapping(mm, addr, vdso_size,
+	ret = install_special_mapping(mm, addr, size,
 				      VM_READ|VM_EXEC|
 				      VM_MAYREAD|VM_MAYWRITE|VM_MAYEXEC|
 				      VM_ALWAYSDUMP,
-				      vdso_pages);
+				      pages);
 	if (ret) {
 		current->mm->context.vdso = NULL;
 		goto up_fail;
@@ -137,6 +185,20 @@ up_fail:
 	return ret;
 }
 
+int arch_setup_additional_pages(struct linux_binprm *bprm, int uses_interp)
+{
+	return setup_additional_pages (bprm, uses_interp, vdso_pages,
+				       vdso_size);
+}
+
+#ifdef CONFIG_X86_X32_ABI
+int x32_setup_additional_pages(struct linux_binprm *bprm, int uses_interp)
+{
+	return setup_additional_pages (bprm, uses_interp, vdsox32_pages,
+				       vdsox32_size);
+}
+#endif
+
 static __init int vdso_setup(char *s)
 {
 	vdso_enabled = simple_strtoul(s, NULL, 0);
-- 
1.7.6.5

Re: [PATCH 02/30] x86-64: Use explicit sizes in sigcontext.h, prepare for x32

[Linus Torvalds]

On Sun, Feb 19, 2012 at 4:07 PM, H. Peter Anvin <hpa@zytor.com> wrote:
>
> Use explicit sizes (__u64) instead of implicit sizes (unsigned long)
> in the definition for sigcontext.h; this will allow this structure to
> be shared between the x86-64 native ABI and the x32 ABI.

Btw, since we had this issue just with autofs: what are the x32 ABI
alignment issues for __u64? Are they like x86-64 ("natural alignment")
or x86-32 ("4-byte alignment")?

I assume they are natural alignment, and as pointed out by Davem, we
do have the versions of u64 that make this explicit: "compat_u64" is
the 4-byte-aligned one, while "__aligned_u64" is the natively aligned
one.

Just plain "__u64" doesn't tell which it is, which is sad and wrong,
but we're likely stuck with it forever. Unless some shining knight
comes and says "__u64 is native alignment, and if you want anything
else, you need to use __compat_u64", and actually fixes the cases
where x86-32 depends on the 4-byte aligned one.

Which would be nice, but sounds unlikely. Shining knights tend to be
rare. But this *could* possibly be automated, so it's not entirely out
of the question.

                 Linus

Re: [PATCH 02/30] x86-64: Use explicit sizes in sigcontext.h, prepare for x32

[H. Peter Anvin]

On 02/19/2012 04:51 PM, Linus Torvalds wrote:
> On Sun, Feb 19, 2012 at 4:07 PM, H. Peter Anvin <hpa@zytor.com> wrote:
>>
>> Use explicit sizes (__u64) instead of implicit sizes (unsigned long)
>> in the definition for sigcontext.h; this will allow this structure to
>> be shared between the x86-64 native ABI and the x32 ABI.
> 
> Btw, since we had this issue just with autofs: what are the x32 ABI
> alignment issues for __u64? Are they like x86-64 ("natural alignment")
> or x86-32 ("4-byte alignment")?
> 
> I assume they are natural alignment, and as pointed out by Davem, we
> do have the versions of u64 that make this explicit: "compat_u64" is
> the 4-byte-aligned one, while "__aligned_u64" is the natively aligned
> one.
> 
> Just plain "__u64" doesn't tell which it is, which is sad and wrong,
> but we're likely stuck with it forever. Unless some shining knight
> comes and says "__u64 is native alignment, and if you want anything
> else, you need to use __compat_u64", and actually fixes the cases
> where x86-32 depends on the 4-byte aligned one.
> 
> Which would be nice, but sounds unlikely. Shining knights tend to be
> rare. But this *could* possibly be automated, so it's not entirely out
> of the question.
> 

We are using __u64 as x86-32 compatible since we are sharing most of the
really complex path (like ioctl) with i386 much more so than x86-64.  So
it is defined in userspace as:

typedef unsigned long long __u64 __attribute__((aligned(4)));

__aligned_u64 obviously is naturally aligned, which matches uint64_t is
userspace.

	-hpa

-- 
H. Peter Anvin, Intel Open Source Technology Center
I work for Intel.  I don't speak on their behalf.

Re: [PATCH 08/30] compat: Use COMPAT_USE_64BIT_TIME in the lp driver

[Linus Torvalds]

On Sun, Feb 19, 2012 at 4:07 PM, H. Peter Anvin <hpa@zytor.com> wrote:
> From: "H. J. Lu" <hjl.tools@gmail.com>
>
> Enable the lp driver to be used with a compat ABI with 64-bit time.

Ugh. Is this really the only case?

Because if it isn't, I suspect it would be much better off with a
helper function. In fact, even if this *does* end up being the only
place, a helper function to get/set a timeval from user space sounds
like a good idea, and makes things much more readable than that
ad-hoccery in the middle of code.

IOW, I'd like to see something like

   get_user_timeval(void __user *tv, struct timeval *res)

even if it would be only local to this file.

                   Linus

Re: [PATCH 08/30] compat: Use COMPAT_USE_64BIT_TIME in the lp driver

[H. Peter Anvin]

On 02/19/2012 04:56 PM, Linus Torvalds wrote:
> On Sun, Feb 19, 2012 at 4:07 PM, H. Peter Anvin <hpa@zytor.com> wrote:
>> From: "H. J. Lu" <hjl.tools@gmail.com>
>>
>> Enable the lp driver to be used with a compat ABI with 64-bit time.
> 
> Ugh. Is this really the only case?
> 

No, it's not the only case, and a helper function is a great idea.

	-hpa


-- 
H. Peter Anvin, Intel Open Source Technology Center
I work for Intel.  I don't speak on their behalf.

[PATCH 0/7] COMPAT_USE_64BIT_TIME v2

[H. Peter Anvin]

From: "H. Peter Anvin" <hpa@zytor.com>

This is a respin of the COMPAT_USE_64BIT_TIME bits of the x32
patchset.  It replaces patches 07/30-10/30 out of that patchset.

The differences in this respin are:

1. Use helper functions as suggested by Linus, and
2. Add support in the networking stack (this bit got inadvertently
   left out of the previous patchset.)

Linus, is this more what you were looking for?

This is the diffstat for this set:

 drivers/char/lp.c            |    5 +--
 drivers/input/input-compat.c |    4 +-
 drivers/input/input-compat.h |    2 +-
 include/linux/compat.h       |   20 ++++++++++++
 kernel/compat.c              |   70 +++++++++++++++++++++++++++++++++++++-----
 net/bluetooth/hci_sock.c     |    3 +-
 net/compat.c                 |   65 ++++++++++++++++++++++++---------------
 net/socket.c                 |   18 +++++------
 8 files changed, 136 insertions(+), 51 deletions(-)

[PATCH 1/7] compat: Introduce COMPAT_USE_64BIT_TIME

[H. Peter Anvin]

From: "H. J. Lu" <hjl.tools@gmail.com>

Allow a compatibility ABI to use a 64-bit time_t and 64-bit members in
struct timeval and struct timespec to avoid the Y2038 problem.

This will be used for the x32 ABI.

Signed-off-by: H. Peter Anvin <hpa@zytor.com>
---
 include/linux/compat.h |    4 ++++
 1 files changed, 4 insertions(+), 0 deletions(-)

diff --git a/include/linux/compat.h b/include/linux/compat.h
index 41c9f65..1be91c0 100644
--- a/include/linux/compat.h
+++ b/include/linux/compat.h
@@ -19,6 +19,10 @@
 #include <asm/siginfo.h>
 #include <asm/signal.h>
 
+#ifndef COMPAT_USE_64BIT_TIME
+#define COMPAT_USE_64BIT_TIME 0
+#endif
+
 #define compat_jiffies_to_clock_t(x)	\
 		(((unsigned long)(x) * COMPAT_USER_HZ) / HZ)
 
-- 
1.7.6.5

[PATCH 1/7] compat: Introduce COMPAT_USE_64BIT_TIME

[H. Peter Anvin]

From: "H. J. Lu" <hjl.tools@gmail.com>

Allow a compatibility ABI to use a 64-bit time_t and 64-bit members in
struct timeval and struct timespec to avoid the Y2038 problem.

This will be used for the x32 ABI.

Signed-off-by: H. Peter Anvin <hpa@zytor.com>
---
 include/linux/compat.h |    4 ++++
 1 files changed, 4 insertions(+), 0 deletions(-)

diff --git a/include/linux/compat.h b/include/linux/compat.h
index 41c9f65..1be91c0 100644
--- a/include/linux/compat.h
+++ b/include/linux/compat.h
@@ -19,6 +19,10 @@
 #include <asm/siginfo.h>
 #include <asm/signal.h>
 
+#ifndef COMPAT_USE_64BIT_TIME
+#define COMPAT_USE_64BIT_TIME 0
+#endif
+
 #define compat_jiffies_to_clock_t(x)	\
 		(((unsigned long)(x) * COMPAT_USER_HZ) / HZ)
 
-- 
1.7.6.5

[PATCH 2/7] compat: Add helper functions to read/write struct timeval, timespec

[H. Peter Anvin]

From: "H. Peter Anvin" <hpa@zytor.com>

Add helper functions to read and write struct timeval and struct
timespec from userspace.  We already had helper functions for reading
and writing struct compat_timespec; add a set of functions to do the
same with struct timeval, and add a second suite of functions which
can be sensitive to COMPAT_USE_64BIT_TIME and access either 32- or
64-bit time structures.

This also exports these helper functions to modules.

Rename the existing inlines for converting between struct
compat_timeval and native struct timespec so we can have a saner
naming convention for the exported functions.

Suggested-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
---
 include/linux/compat.h |   16 +++++++++++
 kernel/compat.c        |   70 ++++++++++++++++++++++++++++++++++++++++++-----
 2 files changed, 78 insertions(+), 8 deletions(-)

diff --git a/include/linux/compat.h b/include/linux/compat.h
index 1be91c0..a82e452 100644
--- a/include/linux/compat.h
+++ b/include/linux/compat.h
@@ -87,10 +87,26 @@ typedef struct {
 	compat_sigset_word	sig[_COMPAT_NSIG_WORDS];
 } compat_sigset_t;
 
+/*
+ * These functions operate strictly on struct compat_time*
+ */
 extern int get_compat_timespec(struct timespec *,
 			       const struct compat_timespec __user *);
 extern int put_compat_timespec(const struct timespec *,
 			       struct compat_timespec __user *);
+extern int get_compat_timeval(struct timeval *,
+			      const struct compat_timeval __user *);
+extern int put_compat_timeval(const struct timeval *,
+			      struct compat_timeval __user *);
+/*
+ * These functions operate on 32- or 64-bit specs depending on
+ * COMPAT_USE_64BIT_TIME, hence the void user pointer arguments and the
+ * naming as compat_get/put_ rather than get/put_compat_.
+ */
+extern int compat_get_timespec(struct timespec *, const void __user *);
+extern int compat_put_timespec(const struct timespec *, void __user *);
+extern int compat_get_timeval(struct timeval *, const void __user *);
+extern int compat_put_timeval(const struct timeval *, void __user *);
 
 struct compat_iovec {
 	compat_uptr_t	iov_base;
diff --git a/kernel/compat.c b/kernel/compat.c
index f346ced..1743d67 100644
--- a/kernel/compat.c
+++ b/kernel/compat.c
@@ -31,11 +31,10 @@
 #include <asm/uaccess.h>
 
 /*
- * Note that the native side is already converted to a timespec, because
- * that's what we want anyway.
+ * Get/set struct timeval with struct timespec on the native side
  */
-static int compat_get_timeval(struct timespec *o,
-		struct compat_timeval __user *i)
+static int compat_get_timeval_convert(struct timespec *o,
+				      struct compat_timeval __user *i)
 {
 	long usec;
 
@@ -46,8 +45,8 @@ static int compat_get_timeval(struct timespec *o,
 	return 0;
 }
 
-static int compat_put_timeval(struct compat_timeval __user *o,
-		struct timeval *i)
+static int compat_put_timeval_convert(struct compat_timeval __user *o,
+				      struct timeval *i)
 {
 	return (put_user(i->tv_sec, &o->tv_sec) ||
 		put_user(i->tv_usec, &o->tv_usec)) ? -EFAULT : 0;
@@ -117,7 +116,7 @@ asmlinkage long compat_sys_gettimeofday(struct compat_timeval __user *tv,
 	if (tv) {
 		struct timeval ktv;
 		do_gettimeofday(&ktv);
-		if (compat_put_timeval(tv, &ktv))
+		if (compat_put_timeval_convert(tv, &ktv))
 			return -EFAULT;
 	}
 	if (tz) {
@@ -135,7 +134,7 @@ asmlinkage long compat_sys_settimeofday(struct compat_timeval __user *tv,
 	struct timezone ktz;
 
 	if (tv) {
-		if (compat_get_timeval(&kts, tv))
+		if (compat_get_timeval_convert(&kts, tv))
 			return -EFAULT;
 	}
 	if (tz) {
@@ -146,12 +145,29 @@ asmlinkage long compat_sys_settimeofday(struct compat_timeval __user *tv,
 	return do_sys_settimeofday(tv ? &kts : NULL, tz ? &ktz : NULL);
 }
 
+int get_compat_timeval(struct timeval *tv, const struct compat_timeval __user *ctv)
+{
+	return (!access_ok(VERIFY_READ, ctv, sizeof(*ctv)) ||
+			__get_user(tv->tv_sec, &ctv->tv_sec) ||
+			__get_user(tv->tv_usec, &ctv->tv_usec)) ? -EFAULT : 0;
+}
+EXPORT_SYMBOL_GPL(get_compat_timeval);
+
+int put_compat_timeval(const struct timeval *tv, struct compat_timespec __user *ctv)
+{
+	return (!access_ok(VERIFY_WRITE, ctv, sizeof(*ctv)) ||
+			__put_user(tv->tv_sec, &ctv->tv_sec) ||
+			__put_user(tv->tv_usec, &ctv->tv_usec)) ? -EFAULT : 0;
+}
+EXPORT_SYMBOL_GPL(put_compat_timeval);
+
 int get_compat_timespec(struct timespec *ts, const struct compat_timespec __user *cts)
 {
 	return (!access_ok(VERIFY_READ, cts, sizeof(*cts)) ||
 			__get_user(ts->tv_sec, &cts->tv_sec) ||
 			__get_user(ts->tv_nsec, &cts->tv_nsec)) ? -EFAULT : 0;
 }
+EXPORT_SYMBOL_GPL(get_compat_timespec);
 
 int put_compat_timespec(const struct timespec *ts, struct compat_timespec __user *cts)
 {
@@ -161,6 +177,42 @@ int put_compat_timespec(const struct timespec *ts, struct compat_timespec __user
 }
 EXPORT_SYMBOL_GPL(put_compat_timespec);
 
+int compat_get_timeval(struct timeval *tv, const void __user *utv)
+{
+	if (COMPAT_USE_64BIT_TIME)
+		return get_user(tv, (const struct timeval __user *)utv);
+	else
+		return get_compat_timeval(tv, utv);
+}
+EXPORT_SYMBOL_GPL(compat_get_timeval);
+
+int compat_put_timeval(const struct timeval *tv, void __user *utv)
+{
+	if (COMPAT_USE_64BIT_TIME)
+		return put_user(tv, (struct timeval __user *)utv);
+	else
+		return put_compat_timeval(tv, utv);
+}
+EXPORT_SYMBOL_GPL(compat_put_timeval);
+
+int compat_get_timespec(struct timespec *ts, const void __user *uts)
+{
+	if (COMPAT_USE_64BIT_TIME)
+		return get_user(ts, (const struct timespec __user *)uts);
+	else
+		return get_compat_timespec(ts, uts);
+}
+EXPORT_SYMBOL_GPL(compat_get_timespec);
+
+int compat_put_timespec(const struct timespec *ts, void __user *uts)
+{
+	if (COMPAT_USE_64BIT_TIME)
+		return put_user(tv, (struct timeval __user *)ctv);
+	else
+		return put_compat_timespec(ts, uts);
+}
+EXPORT_SYMBOL_GPL(compat_put_timespec);
+
 static long compat_nanosleep_restart(struct restart_block *restart)
 {
 	struct compat_timespec __user *rmtp;
@@ -1162,3 +1214,5 @@ void __user *compat_alloc_user_space(unsigned long len)
 	return ptr;
 }
 EXPORT_SYMBOL_GPL(compat_alloc_user_space);
+
+/*
-- 
1.7.6.5

[PATCH 3/7] compat: Handle COMPAT_USE_64BIT_TIME in the lp driver

[H. Peter Anvin]

From: "H. Peter Anvin" <hpa@zytor.com>

Enable the lp driver to be used with a compat ABI with 64-bit time.

Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Greg Kroah-Hartman <greg@kroah.com>
---
 drivers/char/lp.c |    5 +----
 1 files changed, 1 insertions(+), 4 deletions(-)

diff --git a/drivers/char/lp.c b/drivers/char/lp.c
index f434856..64f5aeb 100644
--- a/drivers/char/lp.c
+++ b/drivers/char/lp.c
@@ -706,16 +706,13 @@ static long lp_compat_ioctl(struct file *file, unsigned int cmd,
 {
 	unsigned int minor;
 	struct timeval par_timeout;
-	struct compat_timeval __user *tc;
 	int ret;
 
 	minor = iminor(file->f_path.dentry->d_inode);
 	mutex_lock(&lp_mutex);
 	switch (cmd) {
 	case LPSETTIMEOUT:
-		tc = compat_ptr(arg);
-		if (get_user(par_timeout.tv_sec, &tc->tv_sec) ||
-		    get_user(par_timeout.tv_usec, &tc->tv_usec)) {
+		if (compat_get_timeval(par_timeout, compat_ptr(arg))) {
 			ret = -EFAULT;
 			break;
 		}
-- 
1.7.6.5

[PATCH 4/7] compat: Use COMPAT_USE_64BIT_TIME in the input subsystem

[H. Peter Anvin]

From: "H. J. Lu" <hjl.tools@gmail.com>

Enable the input system to be used with a compat ABI with 64-bit time.

Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Cc: Dmitry Torokhov <dmitry.torokhov@gmail.com>
---
 drivers/input/input-compat.c |    4 ++--
 drivers/input/input-compat.h |    2 +-
 2 files changed, 3 insertions(+), 3 deletions(-)

diff --git a/drivers/input/input-compat.c b/drivers/input/input-compat.c
index e46a867..64ca711 100644
--- a/drivers/input/input-compat.c
+++ b/drivers/input/input-compat.c
@@ -17,7 +17,7 @@
 int input_event_from_user(const char __user *buffer,
 			  struct input_event *event)
 {
-	if (INPUT_COMPAT_TEST) {
+	if (INPUT_COMPAT_TEST && !COMPAT_USE_64BIT_TIME) {
 		struct input_event_compat compat_event;
 
 		if (copy_from_user(&compat_event, buffer,
@@ -41,7 +41,7 @@ int input_event_from_user(const char __user *buffer,
 int input_event_to_user(char __user *buffer,
 			const struct input_event *event)
 {
-	if (INPUT_COMPAT_TEST) {
+	if (INPUT_COMPAT_TEST && !COMPAT_USE_64BIT_TIME) {
 		struct input_event_compat compat_event;
 
 		compat_event.time.tv_sec = event->time.tv_sec;
diff --git a/drivers/input/input-compat.h b/drivers/input/input-compat.h
index 22be27b..148f66f 100644
--- a/drivers/input/input-compat.h
+++ b/drivers/input/input-compat.h
@@ -67,7 +67,7 @@ struct ff_effect_compat {
 
 static inline size_t input_event_size(void)
 {
-	return INPUT_COMPAT_TEST ?
+	return (INPUT_COMPAT_TEST && !COMPAT_USE_64BIT_TIME) ?
 		sizeof(struct input_event_compat) : sizeof(struct input_event);
 }
 
-- 
1.7.6.5

[PATCH 5/7] compat: Use COMPAT_USE_64BIT_TIME in the Bluetooth subsystem

[H. Peter Anvin]

From: "H. J. Lu" <hjl.tools@gmail.com>

Enable the Bluetooth subsystem to be used with a compat ABI with
64-bit time.

Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Cc: Marcel Holtmann <marcel@holtmann.org>
Cc: Gustavo F. Padovan <padovan@profusion.mobi>
Cc: David S. Miller <davem@davemloft.net>
---
 net/bluetooth/hci_sock.c |    3 ++-
 1 files changed, 2 insertions(+), 1 deletions(-)

diff --git a/net/bluetooth/hci_sock.c b/net/bluetooth/hci_sock.c
index 0dcc962..b2eb2b9 100644
--- a/net/bluetooth/hci_sock.c
+++ b/net/bluetooth/hci_sock.c
@@ -418,7 +418,8 @@ static inline void hci_sock_cmsg(struct sock *sk, struct msghdr *msg, struct sk_
 		data = &tv;
 		len = sizeof(tv);
 #ifdef CONFIG_COMPAT
-		if (msg->msg_flags & MSG_CMSG_COMPAT) {
+		if (!COMPAT_USE_64BIT_TIME &&
+		    (msg->msg_flags & MSG_CMSG_COMPAT)) {
 			ctv.tv_sec = tv.tv_sec;
 			ctv.tv_usec = tv.tv_usec;
 			data = &ctv;
-- 
1.7.6.5

[PATCH 6/7] compat: Use COMPAT_USE_64BIT_TIME in net/compat.c

[H. Peter Anvin]

From: "H. J. Lu" <hjl.tools@gmail.com>

Handle 64-bit time structures in the networking core compat code.

Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Cc: David S. Miller <davem@davemloft.net>
---
 net/compat.c |   65 +++++++++++++++++++++++++++++++++++----------------------
 1 files changed, 40 insertions(+), 25 deletions(-)

diff --git a/net/compat.c b/net/compat.c
index 6def90e..73bf0e0 100644
--- a/net/compat.c
+++ b/net/compat.c
@@ -219,8 +219,6 @@ Efault:
 
 int put_cmsg_compat(struct msghdr *kmsg, int level, int type, int len, void *data)
 {
-	struct compat_timeval ctv;
-	struct compat_timespec cts[3];
 	struct compat_cmsghdr __user *cm = (struct compat_cmsghdr __user *) kmsg->msg_control;
 	struct compat_cmsghdr cmhdr;
 	int cmlen;
@@ -230,24 +228,28 @@ int put_cmsg_compat(struct msghdr *kmsg, int level, int type, int len, void *dat
 		return 0; /* XXX: return error? check spec. */
 	}
 
-	if (level == SOL_SOCKET && type == SCM_TIMESTAMP) {
-		struct timeval *tv = (struct timeval *)data;
-		ctv.tv_sec = tv->tv_sec;
-		ctv.tv_usec = tv->tv_usec;
-		data = &ctv;
-		len = sizeof(ctv);
-	}
-	if (level == SOL_SOCKET &&
-	    (type == SCM_TIMESTAMPNS || type == SCM_TIMESTAMPING)) {
-		int count = type == SCM_TIMESTAMPNS ? 1 : 3;
-		int i;
-		struct timespec *ts = (struct timespec *)data;
-		for (i = 0; i < count; i++) {
-			cts[i].tv_sec = ts[i].tv_sec;
-			cts[i].tv_nsec = ts[i].tv_nsec;
+	if (!COMPAT_USE_64BIT_TIME) {
+		struct compat_timeval ctv;
+		struct compat_timespec cts[3];
+		if (level == SOL_SOCKET && type == SCM_TIMESTAMP) {
+			struct timeval *tv = (struct timeval *)data;
+			ctv.tv_sec = tv->tv_sec;
+			ctv.tv_usec = tv->tv_usec;
+			data = &ctv;
+			len = sizeof(ctv);
+		}
+		if (level == SOL_SOCKET &&
+		    (type == SCM_TIMESTAMPNS || type == SCM_TIMESTAMPING)) {
+			int count = type == SCM_TIMESTAMPNS ? 1 : 3;
+			int i;
+			struct timespec *ts = (struct timespec *)data;
+			for (i = 0; i < count; i++) {
+				cts[i].tv_sec = ts[i].tv_sec;
+				cts[i].tv_nsec = ts[i].tv_nsec;
+			}
+			data = &cts;
+			len = sizeof(cts[0]) * count;
 		}
-		data = &cts;
-		len = sizeof(cts[0]) * count;
 	}
 
 	cmlen = CMSG_COMPAT_LEN(len);
@@ -454,11 +456,15 @@ static int compat_sock_getsockopt(struct socket *sock, int level, int optname,
 
 int compat_sock_get_timestamp(struct sock *sk, struct timeval __user *userstamp)
 {
-	struct compat_timeval __user *ctv =
-			(struct compat_timeval __user *) userstamp;
-	int err = -ENOENT;
+	struct compat_timeval __user *ctv;
+	int err;
 	struct timeval tv;
 
+	if (COMPAT_USE_64BIT_TIME)
+		return sock_get_timestamp(sk, userstamp);
+
+	ctv = (struct compat_timeval __user *) userstamp;
+	err = -ENOENT;
 	if (!sock_flag(sk, SOCK_TIMESTAMP))
 		sock_enable_timestamp(sk, SOCK_TIMESTAMP);
 	tv = ktime_to_timeval(sk->sk_stamp);
@@ -478,11 +484,15 @@ EXPORT_SYMBOL(compat_sock_get_timestamp);
 
 int compat_sock_get_timestampns(struct sock *sk, struct timespec __user *userstamp)
 {
-	struct compat_timespec __user *ctv =
-			(struct compat_timespec __user *) userstamp;
-	int err = -ENOENT;
+	struct compat_timespec __user *ctv;
+	int err;
 	struct timespec ts;
 
+	if (COMPAT_USE_64BIT_TIME)
+		return sock_get_timestampns (sk, userstamp);
+
+	ctv = (struct compat_timespec __user *) userstamp;
+	err = -ENOENT;
 	if (!sock_flag(sk, SOCK_TIMESTAMP))
 		sock_enable_timestamp(sk, SOCK_TIMESTAMP);
 	ts = ktime_to_timespec(sk->sk_stamp);
@@ -767,6 +777,11 @@ asmlinkage long compat_sys_recvmmsg(int fd, struct compat_mmsghdr __user *mmsg,
 	int datagrams;
 	struct timespec ktspec;
 
+	if (COMPAT_USE_64BIT_TIME)
+		return __sys_recvmmsg(fd, (struct mmsghdr __user *)mmsg, vlen,
+				      flags | MSG_CMSG_COMPAT,
+				      (struct timespec *) timeout);
+
 	if (timeout == NULL)
 		return __sys_recvmmsg(fd, (struct mmsghdr __user *)mmsg, vlen,
 				      flags | MSG_CMSG_COMPAT, NULL);
-- 
1.7.6.5

[PATCH 7/7] compat: Handle COMPAT_USE_64BIT_TIME in net/socket.c

[H. Peter Anvin]

From: "H. Peter Anvin" <hpa@zytor.com>

Use helper functions aware of COMPAT_USE_64BIT_TIME to write struct
timeval and struct timespec to userspace in net/socket.c.

Signed-off-by: H. Peter Anvin <hpa@zytor.com>
---
 net/socket.c |   18 ++++++++----------
 1 files changed, 8 insertions(+), 10 deletions(-)

diff --git a/net/socket.c b/net/socket.c
index 28a96af..57f5a25 100644
--- a/net/socket.c
+++ b/net/socket.c
@@ -2600,7 +2600,7 @@ void socket_seq_show(struct seq_file *seq)
 
 #ifdef CONFIG_COMPAT
 static int do_siocgstamp(struct net *net, struct socket *sock,
-			 unsigned int cmd, struct compat_timeval __user *up)
+			 unsigned int cmd, void __user *up)
 {
 	mm_segment_t old_fs = get_fs();
 	struct timeval ktv;
@@ -2609,15 +2609,14 @@ static int do_siocgstamp(struct net *net, struct socket *sock,
 	set_fs(KERNEL_DS);
 	err = sock_do_ioctl(net, sock, cmd, (unsigned long)&ktv);
 	set_fs(old_fs);
-	if (!err) {
-		err = put_user(ktv.tv_sec, &up->tv_sec);
-		err |= __put_user(ktv.tv_usec, &up->tv_usec);
-	}
+	if (!err)
+		err = compat_put_timeval(up, &ktv);
+
 	return err;
 }
 
 static int do_siocgstampns(struct net *net, struct socket *sock,
-			 unsigned int cmd, struct compat_timespec __user *up)
+			 unsigned int cmd, struct void __user *up)
 {
 	mm_segment_t old_fs = get_fs();
 	struct timespec kts;
@@ -2626,10 +2625,9 @@ static int do_siocgstampns(struct net *net, struct socket *sock,
 	set_fs(KERNEL_DS);
 	err = sock_do_ioctl(net, sock, cmd, (unsigned long)&kts);
 	set_fs(old_fs);
-	if (!err) {
-		err = put_user(kts.tv_sec, &up->tv_sec);
-		err |= __put_user(kts.tv_nsec, &up->tv_nsec);
-	}
+	if (!err)
+		err = compat_put_timespec(up, &kts);
+
 	return err;
 }
 
-- 
1.7.6.5

Re: [PATCH 0/7] COMPAT_USE_64BIT_TIME v2

[H. Peter Anvin]

On 02/19/2012 06:22 PM, H. Peter Anvin wrote:
> From: "H. Peter Anvin" <hpa@zytor.com>
> 
> This is a respin of the COMPAT_USE_64BIT_TIME bits of the x32
> patchset.  It replaces patches 07/30-10/30 out of that patchset.
> 
> The differences in this respin are:
> 
> 1. Use helper functions as suggested by Linus, and
> 2. Add support in the networking stack (this bit got inadvertently
>    left out of the previous patchset.)
> 
> Linus, is this more what you were looking for?
> 
> This is the diffstat for this set:
> 

(And no, it doesn't even compile... working on that ;)

	-hpa

-- 
H. Peter Anvin, Intel Open Source Technology Center
I work for Intel.  I don't speak on their behalf.

Re: [PATCH 0/7] COMPAT_USE_64BIT_TIME v2

[H. Peter Anvin]

On 02/19/2012 06:42 PM, H. Peter Anvin wrote:
> 
> (And no, it doesn't even compile... working on that ;)
> 

The full tree which more Actually Compiles[TM] at:

git://git.kernel.org/pub/scm/linux/kernel/git/hpa/linux-x32.git x86/x32

I won't patchbomb it tonight, though.

	-hpa

-- 
H. Peter Anvin, Intel Open Source Technology Center
I work for Intel.  I don't speak on their behalf.

Re: [PATCH 06/30] sysinfo: Use explicit types in <linux/sysinfo.h>

[Geert Uytterhoeven]

On Mon, Feb 20, 2012 at 01:07, H. Peter Anvin <hpa@zytor.com> wrote:
> --- a/include/linux/sysinfo.h
> +++ b/include/linux/sysinfo.h
> @@ -1,22 +1,24 @@
>  #ifndef _LINUX_SYSINFO_H
>  #define _LINUX_SYSINFO_H
>
> +#include <linux/types.h>
> +
>  #define SI_LOAD_SHIFT  16
>  struct sysinfo {
> -       long uptime;                    /* Seconds since boot */
> -       unsigned long loads[3];         /* 1, 5, and 15 minute load averages */
> -       unsigned long totalram;         /* Total usable main memory size */
> -       unsigned long freeram;          /* Available memory size */
> -       unsigned long sharedram;        /* Amount of shared memory */
> -       unsigned long bufferram;        /* Memory used by buffers */
> -       unsigned long totalswap;        /* Total swap space size */
> -       unsigned long freeswap;         /* swap space still available */
> -       unsigned short procs;           /* Number of current processes */
> -       unsigned short pad;             /* explicit padding for m68k */
> -       unsigned long totalhigh;        /* Total high memory size */
> -       unsigned long freehigh;         /* Available high memory size */
> -       unsigned int mem_unit;          /* Memory unit size in bytes */
> -       char _f[20-2*sizeof(long)-sizeof(int)]; /* Padding: libc5 uses this.. */
> +       __kernel_long_t uptime;         /* Seconds since boot */
> +       __kernel_ulong_t loads[3];      /* 1, 5, and 15 minute load averages */
> +       __kernel_ulong_t totalram;      /* Total usable main memory size */
> +       __kernel_ulong_t freeram;       /* Available memory size */
> +       __kernel_ulong_t sharedram;     /* Amount of shared memory */
> +       __kernel_ulong_t bufferram;     /* Memory used by buffers */
> +       __kernel_ulong_t totalswap;     /* Total swap space size */
> +       __kernel_ulong_t freeswap;      /* swap space still available */
> +       __u16 procs;                    /* Number of current processes */
> +       __u16 pad;                      /* Explicit padding for m68k */

Fueling the discussion about natural vs. 4-byte alignment?

> +       __kernel_ulong_t totalhigh;     /* Total high memory size */
> +       __kernel_ulong_t freehigh;      /* Available high memory size */
> +       __u32 mem_unit;                 /* Memory unit size in bytes */
> +       char _f[20-2*sizeof(__kernel_ulong_t)-sizeof(__u32)];   /* Padding: libc5 uses this.. */
>  };

Gr{oetje,eeting}s,

                        Geert

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
                                -- Linus Torvalds

Re: [PATCH 06/30] sysinfo: Use explicit types in <linux/sysinfo.h>

[H. Peter Anvin]

Not really... it comes down to "implicit padding in kernel ABI structures is bad".  They can easily become security holes.



Geert Uytterhoeven <geert@linux-m68k.org> wrote:

>On Mon, Feb 20, 2012 at 01:07, H. Peter Anvin <hpa@zytor.com> wrote:
>> --- a/include/linux/sysinfo.h
>> +++ b/include/linux/sysinfo.h
>> @@ -1,22 +1,24 @@
>>  #ifndef _LINUX_SYSINFO_H
>>  #define _LINUX_SYSINFO_H
>>
>> +#include <linux/types.h>
>> +
>>  #define SI_LOAD_SHIFT  16
>>  struct sysinfo {
>> -       long uptime;                    /* Seconds since boot */
>> -       unsigned long loads[3];         /* 1, 5, and 15 minute load
>averages */
>> -       unsigned long totalram;         /* Total usable main memory
>size */
>> -       unsigned long freeram;          /* Available memory size */
>> -       unsigned long sharedram;        /* Amount of shared memory */
>> -       unsigned long bufferram;        /* Memory used by buffers */
>> -       unsigned long totalswap;        /* Total swap space size */
>> -       unsigned long freeswap;         /* swap space still available
>*/
>> -       unsigned short procs;           /* Number of current
>processes */
>> -       unsigned short pad;             /* explicit padding for m68k
>*/
>> -       unsigned long totalhigh;        /* Total high memory size */
>> -       unsigned long freehigh;         /* Available high memory size
>*/
>> -       unsigned int mem_unit;          /* Memory unit size in bytes
>*/
>> -       char _f[20-2*sizeof(long)-sizeof(int)]; /* Padding: libc5
>uses this.. */
>> +       __kernel_long_t uptime;         /* Seconds since boot */
>> +       __kernel_ulong_t loads[3];      /* 1, 5, and 15 minute load
>averages */
>> +       __kernel_ulong_t totalram;      /* Total usable main memory
>size */
>> +       __kernel_ulong_t freeram;       /* Available memory size */
>> +       __kernel_ulong_t sharedram;     /* Amount of shared memory */
>> +       __kernel_ulong_t bufferram;     /* Memory used by buffers */
>> +       __kernel_ulong_t totalswap;     /* Total swap space size */
>> +       __kernel_ulong_t freeswap;      /* swap space still available
>*/
>> +       __u16 procs;                    /* Number of current
>processes */
>> +       __u16 pad;                      /* Explicit padding for m68k
>*/
>
>Fueling the discussion about natural vs. 4-byte alignment?
>
>> +       __kernel_ulong_t totalhigh;     /* Total high memory size */
>> +       __kernel_ulong_t freehigh;      /* Available high memory size
>*/
>> +       __u32 mem_unit;                 /* Memory unit size in bytes
>*/
>> +       char _f[20-2*sizeof(__kernel_ulong_t)-sizeof(__u32)];   /*
>Padding: libc5 uses this.. */
>>  };
>
>Gr{oetje,eeting}s,
>
>                        Geert
>
>--
>Geert Uytterhoeven -- There's lots of Linux beyond ia32 --
>geert@linux-m68k.org
>
>In personal conversations with technical people, I call myself a
>hacker. But
>when I'm talking to journalists I just say "programmer" or something
>like that.
>                                -- Linus Torvalds

-- 
Sent from my mobile phone. Please excuse my brevity and lack of formatting.

Re: [PATCH 06/30] sysinfo: Use explicit types in <linux/sysinfo.h>

[Geert Uytterhoeven]

On Mon, Feb 20, 2012 at 18:29, H. Peter Anvin <hpa@zytor.com> wrote:
> Not really... it comes down to "implicit padding in kernel ABI structures is bad".  They can easily become security holes.

On 64-bit platforms with natural alignment of long, there's an implicit padding.

> Geert Uytterhoeven <geert@linux-m68k.org> wrote:
>
>>On Mon, Feb 20, 2012 at 01:07, H. Peter Anvin <hpa@zytor.com> wrote:
>>> --- a/include/linux/sysinfo.h
>>> +++ b/include/linux/sysinfo.h
>>> @@ -1,22 +1,24 @@
>>>  #ifndef _LINUX_SYSINFO_H
>>>  #define _LINUX_SYSINFO_H
>>>
>>> +#include <linux/types.h>
>>> +
>>>  #define SI_LOAD_SHIFT  16
>>>  struct sysinfo {
>>> -       long uptime;                    /* Seconds since boot */
>>> -       unsigned long loads[3];         /* 1, 5, and 15 minute load
>>averages */
>>> -       unsigned long totalram;         /* Total usable main memory
>>size */
>>> -       unsigned long freeram;          /* Available memory size */
>>> -       unsigned long sharedram;        /* Amount of shared memory */
>>> -       unsigned long bufferram;        /* Memory used by buffers */
>>> -       unsigned long totalswap;        /* Total swap space size */
>>> -       unsigned long freeswap;         /* swap space still available
>>*/
>>> -       unsigned short procs;           /* Number of current
>>processes */
>>> -       unsigned short pad;             /* explicit padding for m68k
>>*/
>>> -       unsigned long totalhigh;        /* Total high memory size */
>>> -       unsigned long freehigh;         /* Available high memory size
>>*/
>>> -       unsigned int mem_unit;          /* Memory unit size in bytes
>>*/
>>> -       char _f[20-2*sizeof(long)-sizeof(int)]; /* Padding: libc5
>>uses this.. */
>>> +       __kernel_long_t uptime;         /* Seconds since boot */
>>> +       __kernel_ulong_t loads[3];      /* 1, 5, and 15 minute load
>>averages */
>>> +       __kernel_ulong_t totalram;      /* Total usable main memory
>>size */
>>> +       __kernel_ulong_t freeram;       /* Available memory size */
>>> +       __kernel_ulong_t sharedram;     /* Amount of shared memory */
>>> +       __kernel_ulong_t bufferram;     /* Memory used by buffers */
>>> +       __kernel_ulong_t totalswap;     /* Total swap space size */
>>> +       __kernel_ulong_t freeswap;      /* swap space still available
>>*/
>>> +       __u16 procs;                    /* Number of current
>>processes */
>>> +       __u16 pad;                      /* Explicit padding for m68k
>>*/
>>
>>Fueling the discussion about natural vs. 4-byte alignment?
>>
>>> +       __kernel_ulong_t totalhigh;     /* Total high memory size */
>>> +       __kernel_ulong_t freehigh;      /* Available high memory size
>>*/
>>> +       __u32 mem_unit;                 /* Memory unit size in bytes
>>*/
>>> +       char _f[20-2*sizeof(__kernel_ulong_t)-sizeof(__u32)];   /*
>>Padding: libc5 uses this.. */
>>>  };

Gr{oetje,eeting}s,

                        Geert

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
                                -- Linus Torvalds

Re: [PATCH 06/30] sysinfo: Use explicit types in <linux/sysinfo.h>

[H. Peter Anvin]

On 02/20/2012 12:01 PM, Geert Uytterhoeven wrote:
> On Mon, Feb 20, 2012 at 18:29, H. Peter Anvin <hpa@zytor.com> wrote:
>> Not really... it comes down to "implicit padding in kernel ABI structures is bad".  They can easily become security holes.
> 
> On 64-bit platforms with natural alignment of long, there's an implicit padding.
> 

We should try to make those explicit, but it sometimes gets hard
retroactively.  Note that the particular patch you're pointing to is
just an identity patch.

	-hpa

-- 
H. Peter Anvin, Intel Open Source Technology Center
I work for Intel.  I don't speak on their behalf.

Re: [PATCH 30/30] x32: Add x32 VDSO support

[Andy Lutomirski]

On 02/19/2012 04:08 PM, H. Peter Anvin wrote:
> From: "H. J. Lu" <hjl.tools@gmail.com>
> 
> Add support for the x32 VDSO.  The x32 VDSO takes advantage of the
> similarity between the x86-64 and the x32 ABIs to contain the same
> content, only the container is different, as the x32 VDSO obviously is
> an x32 shared object.

> +
> +/*
> + * This controls what userland symbols we export from the vDSO.
> + */
> +VERSION {
> +	LINUX_2.6 {
> +	global:
> +		clock_gettime;
> +		__vdso_clock_gettime;
> +		gettimeofday;
> +		__vdso_gettimeofday;
> +		getcpu;
> +		__vdso_getcpu;
> +		time;
> +		__vdso_time;
> +	local: *;
> +	};
> +}
> +

Would it make sense to remove the non-__vdso-prefixed weak symbols?
AFAICT they are somewhere between useless (because the __vdso symbols
are unambiguous), confusing (has anyone not read this and said "huh?"),
and wrong (they are not interchangeable with glibc's symbols as they
return different values).

We're stuck with them on x86-64, but x32 is new and has no
backwards-compatibility issues.

--Andy

Re: [PATCH 30/30] x32: Add x32 VDSO support

[Andy Lutomirski]

On 02/19/2012 04:08 PM, H. Peter Anvin wrote:
> From: "H. J. Lu" <hjl.tools@gmail.com>
> 
> Add support for the x32 VDSO.  The x32 VDSO takes advantage of the
> similarity between the x86-64 and the x32 ABIs to contain the same
> content, only the container is different, as the x32 VDSO obviously is
> an x32 shared object.

> +
> +/*
> + * This controls what userland symbols we export from the vDSO.
> + */
> +VERSION {
> +	LINUX_2.6 {
> +	global:
> +		clock_gettime;
> +		__vdso_clock_gettime;
> +		gettimeofday;
> +		__vdso_gettimeofday;
> +		getcpu;
> +		__vdso_getcpu;
> +		time;
> +		__vdso_time;
> +	local: *;
> +	};
> +}
> +

Would it make sense to remove the non-__vdso-prefixed weak symbols?
AFAICT they are somewhere between useless (because the __vdso symbols
are unambiguous), confusing (has anyone not read this and said "huh?"),
and wrong (they are not interchangeable with glibc's symbols as they
return different values).

We're stuck with them on x86-64, but x32 is new and has no
backwards-compatibility issues.

--Andy

Re: [PATCH 30/30] x32: Add x32 VDSO support

[H. Peter Anvin]

On 02/20/2012 04:12 PM, Andy Lutomirski wrote:
> 
> Would it make sense to remove the non-__vdso-prefixed weak symbols?
> AFAICT they are somewhere between useless (because the __vdso symbols
> are unambiguous), confusing (has anyone not read this and said "huh?"),
> and wrong (they are not interchangeable with glibc's symbols as they
> return different values).
> 
> We're stuck with them on x86-64, but x32 is new and has no
> backwards-compatibility issues.
> 

What about non-glibc?

	-hpa


-- 
H. Peter Anvin, Intel Open Source Technology Center
I work for Intel.  I don't speak on their behalf.

Re: [PATCH 30/30] x32: Add x32 VDSO support

[Andrew Lutomirski]

On Mon, Feb 20, 2012 at 7:58 PM, H. Peter Anvin <hpa@zytor.com> wrote:
> On 02/20/2012 04:12 PM, Andy Lutomirski wrote:
>>
>> Would it make sense to remove the non-__vdso-prefixed weak symbols?
>> AFAICT they are somewhere between useless (because the __vdso symbols
>> are unambiguous), confusing (has anyone not read this and said "huh?"),
>> and wrong (they are not interchangeable with glibc's symbols as they
>> return different values).
>>
>> We're stuck with them on x86-64, but x32 is new and has no
>> backwards-compatibility issues.
>>
>
> What about non-glibc?

IMO non-glibc users should just call __vdso_clock_gettime, etc.
Currently, code like:

if (clock_gettime(whatever) == -1)
  handle_the_error();

is correct when linked against glibc but incorrect when linked
directly against the vdso.

--Andy

Re: [PATCH 30/30] x32: Add x32 VDSO support

[H. Peter Anvin]

On 02/21/2012 08:52 AM, Andrew Lutomirski wrote:
>>
>> What about non-glibc?
>
> IMO non-glibc users should just call __vdso_clock_gettime, etc.
> Currently, code like:
>
> if (clock_gettime(whatever) == -1)
>    handle_the_error();
>
> is correct when linked against glibc but incorrect when linked
> directly against the vdso.
>

The issue is what uclibc, Bionic, etc. actually do.

	-hpa

Re: [PATCH 30/30] x32: Add x32 VDSO support

[Andrew Lutomirski]

On Tue, Feb 21, 2012 at 9:51 AM, H. Peter Anvin <hpa@zytor.com> wrote:
> On 02/21/2012 08:52 AM, Andrew Lutomirski wrote:
>>>
>>>
>>> What about non-glibc?
>>
>>
>> IMO non-glibc users should just call __vdso_clock_gettime, etc.
>> Currently, code like:
>>
>> if (clock_gettime(whatever) == -1)
>>   handle_the_error();
>>
>> is correct when linked against glibc but incorrect when linked
>> directly against the vdso.
>>
>
> The issue is what uclibc, Bionic, etc. actually do.

AFAICS Bionic only works on x86-32 and calls clock_gettime via
hardcoded int 0x80, written in assembly (!).  uclibc calls
__vdso_getcpu and does not seem to use the other vdso calls.  On a
cursory inspection, klibc uses neither the vsyscall page nor the vdso.

I doubt that there's any existing libc replacement that uses the
non-prefixed vdso entries and that already works on x32 -- that would
be impressive.  I'm not suggesting changing anything in the x86-64
vdso.

uclibc hardcodes a call to the vsyscall gettimeofday implementation in
its locking primitives, which probably gives terrible performance, but
that's a separate issue.  I think do_emulate_vsyscall should send a
segfault if called by an x32 task -- there's some security benefit to
doing so, and there's unlikely to be any downside.

--Andy

Re: [PATCH 30/30] x32: Add x32 VDSO support

[H. Peter Anvin]

On 02/21/2012 10:54 AM, Andrew Lutomirski wrote:
> 
> uclibc hardcodes a call to the vsyscall gettimeofday implementation in
> its locking primitives, which probably gives terrible performance, but
> that's a separate issue.  I think do_emulate_vsyscall should send a
> segfault if called by an x32 task -- there's some security benefit to
> doing so, and there's unlikely to be any downside.
> 

The vsyscall page shouldn't be mapped for x32 tasks...

	-hpa

Re: [PATCH 10/30] compat: Use COMPAT_USE_64BIT_TIME in the Bluetooth subsystem

[Gustavo Padovan]

* H. Peter Anvin <hpa@zytor.com> [2012-02-19 16:07:48 -0800]:

> From: "H. J. Lu" <hjl.tools@gmail.com>
> 
> Enable the Bluetooth subsystem to be used with a compat ABI with
> 64-bit time.
> 
> Signed-off-by: H. Peter Anvin <hpa@zytor.com>
> Cc: Marcel Holtmann <marcel@holtmann.org>
> Cc: Gustavo F. Padovan <padovan@profusion.mobi>
> Cc: David S. Miller <davem@davemloft.net>
> ---
>  net/bluetooth/hci_sock.c |    3 ++-
>  1 files changed, 2 insertions(+), 1 deletions(-)

Applied, thanks.

	Gustavo

Re: [PATCH 10/30] compat: Use COMPAT_USE_64BIT_TIME in the Bluetooth subsystem

[H. Peter Anvin]

On 02/21/2012 11:05 AM, Gustavo Padovan wrote:
> 
> * H. Peter Anvin <hpa@zytor.com> [2012-02-19 16:07:48 -0800]:
> 
>> From: "H. J. Lu" <hjl.tools@gmail.com>
>>
>> Enable the Bluetooth subsystem to be used with a compat ABI with
>> 64-bit time.
>>
>> Signed-off-by: H. Peter Anvin <hpa@zytor.com>
>> Cc: Marcel Holtmann <marcel@holtmann.org>
>> Cc: Gustavo F. Padovan <padovan@profusion.mobi>
>> Cc: David S. Miller <davem@davemloft.net>
>> ---
>>  net/bluetooth/hci_sock.c |    3 ++-
>>  1 files changed, 2 insertions(+), 1 deletions(-)
> 
> Applied, thanks.
> 

OK, I have it in my x86/x32 tree; you don't want to have this patch
without the prior patch for obvious reasons.

	-hpa

Re: [PATCH 10/30] compat: Use COMPAT_USE_64BIT_TIME in the Bluetooth subsystem

[H. Peter Anvin]

On 02/21/2012 11:05 AM, Gustavo Padovan wrote:
> 
> * H. Peter Anvin <hpa@zytor.com> [2012-02-19 16:07:48 -0800]:
> 
>> From: "H. J. Lu" <hjl.tools@gmail.com>
>>
>> Enable the Bluetooth subsystem to be used with a compat ABI with
>> 64-bit time.
>>
>> Signed-off-by: H. Peter Anvin <hpa@zytor.com>
>> Cc: Marcel Holtmann <marcel@holtmann.org>
>> Cc: Gustavo F. Padovan <padovan@profusion.mobi>
>> Cc: David S. Miller <davem@davemloft.net>
>> ---
>>  net/bluetooth/hci_sock.c |    3 ++-
>>  1 files changed, 2 insertions(+), 1 deletions(-)
> 
> Applied, thanks.
> 

OK, I have it in my x86/x32 tree; you don't want to have this patch
without the prior patch for obvious reasons.

	-hpa

Re: [PATCH 30/30] x32: Add x32 VDSO support

[Andrew Lutomirski]

On Tue, Feb 21, 2012 at 11:03 AM, H. Peter Anvin <hpa@zytor.com> wrote:
> On 02/21/2012 10:54 AM, Andrew Lutomirski wrote:
>>
>> uclibc hardcodes a call to the vsyscall gettimeofday implementation in
>> its locking primitives, which probably gives terrible performance, but
>> that's a separate issue.  I think do_emulate_vsyscall should send a
>> segfault if called by an x32 task -- there's some security benefit to
>> doing so, and there's unlikely to be any downside.
>>
>
> The vsyscall page shouldn't be mapped for x32 tasks...

How is that possible?  It lives in the fixmap and is presumably
visible from any 64-bit code.

Admittedly, x32 tasks are probably somewhat difficult to trick into
calling addresses with high bits set, but it's not necessarily
impossible.

--Andy

Re: [PATCH 30/30] x32: Add x32 VDSO support

[H. Peter Anvin]

On 02/21/2012 11:29 AM, Andrew Lutomirski wrote:
>>
>> The vsyscall page shouldn't be mapped for x32 tasks...
> 
> How is that possible?  It lives in the fixmap and is presumably
> visible from any 64-bit code.
> 
> Admittedly, x32 tasks are probably somewhat difficult to trick into
> calling addresses with high bits set, but it's not necessarily
> impossible.
> 

Fair enough, and it's not necessarily all that hard either.

And it's visible even in a 32-bit task, although a 32-bit task has to
switch into 64-bit mode.  Yet another reason the vsyscall page needs to die.

I was having delusions that we could have a task-owned PDT in negative
space, but that would require unsharing the third level, too, which is
just way too messy.

	-hpa

Re: [PATCH 30/30] x32: Add x32 VDSO support

[Andrew Lutomirski]

On Tue, Feb 21, 2012 at 11:37 AM, H. Peter Anvin <hpa@zytor.com> wrote:
> On 02/21/2012 11:29 AM, Andrew Lutomirski wrote:
>>>
>>> The vsyscall page shouldn't be mapped for x32 tasks...
>>
>> How is that possible?  It lives in the fixmap and is presumably
>> visible from any 64-bit code.
>>
>> Admittedly, x32 tasks are probably somewhat difficult to trick into
>> calling addresses with high bits set, but it's not necessarily
>> impossible.
>>
>
> Fair enough, and it's not necessarily all that hard either.
>
> And it's visible even in a 32-bit task, although a 32-bit task has to
> switch into 64-bit mode.  Yet another reason the vsyscall page needs to die.
>
> I was having delusions that we could have a task-owned PDT in negative
> space, but that would require unsharing the third level, too, which is
> just way too messy.

I'd like to do that, too, and I'd also like to have a per-cpu
kernel-only page in there, but that's even worse.  If we had a
separate cr3-like register for negative addresses, life would be good
:)

--Andy

Re: [PATCH 30/30] x32: Add x32 VDSO support

[Andrew Lutomirski]

On Tue, Feb 21, 2012 at 11:37 AM, H. Peter Anvin <hpa@zytor.com> wrote:
> On 02/21/2012 11:29 AM, Andrew Lutomirski wrote:
>>>
>>> The vsyscall page shouldn't be mapped for x32 tasks...
>>
>> How is that possible?  It lives in the fixmap and is presumably
>> visible from any 64-bit code.
>>
>> Admittedly, x32 tasks are probably somewhat difficult to trick into
>> calling addresses with high bits set, but it's not necessarily
>> impossible.
>>
>
> Fair enough, and it's not necessarily all that hard either.
>
> And it's visible even in a 32-bit task, although a 32-bit task has to
> switch into 64-bit mode.  Yet another reason the vsyscall page needs to die.
>
> I was having delusions that we could have a task-owned PDT in negative
> space, but that would require unsharing the third level, too, which is
> just way too messy.

I'd like to do that, too, and I'd also like to have a per-cpu
kernel-only page in there, but that's even worse.  If we had a
separate cr3-like register for negative addresses, life would be good
:)

--Andy

Re: [PATCH 30/30] x32: Add x32 VDSO support

[H. Peter Anvin]

On 02/21/2012 11:40 AM, Andrew Lutomirski wrote:
>>
>> I was having delusions that we could have a task-owned PDT in negative
>> space, but that would require unsharing the third level, too, which is
>> just way too messy.
> 
> I'd like to do that, too, and I'd also like to have a per-cpu
> kernel-only page in there, but that's even worse.  If we had a
> separate cr3-like register for negative addresses, life would be good
> :)
> 

No, that wouldn't help.  The situation is actually quite similar to the
current situation where we have an unshared fourth level, but since the
fourth entries are 512G per entry, we would have to push unsharing of
the kernel address space at least one more level (1G), possibly two
(2M).  Painful.

The main advantage of a separate cr3 would be that we wouldn't need the
unshared top level for the kernel side.

	-hpa

Re: [PATCH 30/30] x32: Add x32 VDSO support

[Andrew Lutomirski]

On Tue, Feb 21, 2012 at 11:49 AM, H. Peter Anvin <hpa@zytor.com> wrote:
> On 02/21/2012 11:40 AM, Andrew Lutomirski wrote:
>>>
>>> I was having delusions that we could have a task-owned PDT in negative
>>> space, but that would require unsharing the third level, too, which is
>>> just way too messy.
>>
>> I'd like to do that, too, and I'd also like to have a per-cpu
>> kernel-only page in there, but that's even worse.  If we had a
>> separate cr3-like register for negative addresses, life would be good
>> :)
>>
>
> No, that wouldn't help.  The situation is actually quite similar to the
> current situation where we have an unshared fourth level, but since the
> fourth entries are 512G per entry, we would have to push unsharing of
> the kernel address space at least one more level (1G), possibly two
> (2M).  Painful.
>
> The main advantage of a separate cr3 would be that we wouldn't need the
> unshared top level for the kernel side.

Also, as is, if the top level wants to be per-cpu *and* per-task,
that's a big explosion of page tables that all need to stay in sync.

Oh well.

--Andy

Re: [PATCH 30/30] x32: Add x32 VDSO support

[Andrew Lutomirski]

On Tue, Feb 21, 2012 at 11:49 AM, H. Peter Anvin <hpa@zytor.com> wrote:
> On 02/21/2012 11:40 AM, Andrew Lutomirski wrote:
>>>
>>> I was having delusions that we could have a task-owned PDT in negative
>>> space, but that would require unsharing the third level, too, which is
>>> just way too messy.
>>
>> I'd like to do that, too, and I'd also like to have a per-cpu
>> kernel-only page in there, but that's even worse.  If we had a
>> separate cr3-like register for negative addresses, life would be good
>> :)
>>
>
> No, that wouldn't help.  The situation is actually quite similar to the
> current situation where we have an unshared fourth level, but since the
> fourth entries are 512G per entry, we would have to push unsharing of
> the kernel address space at least one more level (1G), possibly two
> (2M).  Painful.
>
> The main advantage of a separate cr3 would be that we wouldn't need the
> unshared top level for the kernel side.

Also, as is, if the top level wants to be per-cpu *and* per-task,
that's a big explosion of page tables that all need to stay in sync.

Oh well.

--Andy

Re: [PATCH 30/30] x32: Add x32 VDSO support

[H. Peter Anvin]

On 02/21/2012 11:51 AM, Andrew Lutomirski wrote:
>>
>> No, that wouldn't help.  The situation is actually quite similar to the
>> current situation where we have an unshared fourth level, but since the
>> fourth entries are 512G per entry, we would have to push unsharing of
>> the kernel address space at least one more level (1G), possibly two
>> (2M).  Painful.
>>
>> The main advantage of a separate cr3 would be that we wouldn't need the
>> unshared top level for the kernel side.
> 
> Also, as is, if the top level wants to be per-cpu *and* per-task,
> that's a big explosion of page tables that all need to stay in sync.
> 

Fair enough.

	-hpa

Re: [PATCH 30/30] x32: Add x32 VDSO support

[H. Peter Anvin]

On 02/21/2012 11:51 AM, Andrew Lutomirski wrote:
>>
>> No, that wouldn't help.  The situation is actually quite similar to the
>> current situation where we have an unshared fourth level, but since the
>> fourth entries are 512G per entry, we would have to push unsharing of
>> the kernel address space at least one more level (1G), possibly two
>> (2M).  Painful.
>>
>> The main advantage of a separate cr3 would be that we wouldn't need the
>> unshared top level for the kernel side.
> 
> Also, as is, if the top level wants to be per-cpu *and* per-task,
> that's a big explosion of page tables that all need to stay in sync.
> 

Fair enough.

	-hpa

Re: [PATCH 02/30] x86-64: Use explicit sizes in sigcontext.h, prepare for x32

[Arnd Bergmann]

On Monday 20 February 2012, H. Peter Anvin wrote:
> We are using __u64 as x86-32 compatible since we are sharing most of the
> really complex path (like ioctl) with i386 much more so than x86-64.  So
> it is defined in userspace as:
> 
> typedef unsigned long long __u64 __attribute__((aligned(4)));
> 
> __aligned_u64 obviously is naturally aligned, which matches uint64_t is
> userspace.

Has someone audited the interfaces to check if there are data structures that
use a plain signed or unsigned "long long" instead of __s64/__u64 in places
where i386 differs from the other compat implementations?

I found DRM_IOCTL_UPDATE_DRAW, but there could be more like this one.

	Arnd

Re: [PATCH 02/30] x86-64: Use explicit sizes in sigcontext.h, prepare for x32

[Arnd Bergmann]

On Monday 20 February 2012, H. Peter Anvin wrote:
> We are using __u64 as x86-32 compatible since we are sharing most of the
> really complex path (like ioctl) with i386 much more so than x86-64.  So
> it is defined in userspace as:
> 
> typedef unsigned long long __u64 __attribute__((aligned(4)));
> 
> __aligned_u64 obviously is naturally aligned, which matches uint64_t is
> userspace.

Has someone audited the interfaces to check if there are data structures that
use a plain signed or unsigned "long long" instead of __s64/__u64 in places
where i386 differs from the other compat implementations?

I found DRM_IOCTL_UPDATE_DRAW, but there could be more like this one.

	Arnd

Re: [PATCH 10/30] compat: Use COMPAT_USE_64BIT_TIME in the Bluetooth subsystem

[Jiri Kosina]

On Tue, 21 Feb 2012, Gustavo Padovan wrote:

> > From: "H. J. Lu" <hjl.tools@gmail.com>
> > 
> > Enable the Bluetooth subsystem to be used with a compat ABI with
> > 64-bit time.
> > 
> > Signed-off-by: H. Peter Anvin <hpa@zytor.com>
> > Cc: Marcel Holtmann <marcel@holtmann.org>
> > Cc: Gustavo F. Padovan <padovan@profusion.mobi>
> > Cc: David S. Miller <davem@davemloft.net>
> > ---
> >  net/bluetooth/hci_sock.c |    3 ++-
> >  1 files changed, 2 insertions(+), 1 deletions(-)
> 
> Applied, thanks.

How much sense does it make to apply this single patch without the 
previous x32 infrastructure changes?

-- 
Jiri Kosina
SUSE Labs

Re: [PATCH 10/30] compat: Use COMPAT_USE_64BIT_TIME in the Bluetooth subsystem

[Jiri Kosina]

On Tue, 21 Feb 2012, Gustavo Padovan wrote:

> > From: "H. J. Lu" <hjl.tools@gmail.com>
> > 
> > Enable the Bluetooth subsystem to be used with a compat ABI with
> > 64-bit time.
> > 
> > Signed-off-by: H. Peter Anvin <hpa@zytor.com>
> > Cc: Marcel Holtmann <marcel@holtmann.org>
> > Cc: Gustavo F. Padovan <padovan@profusion.mobi>
> > Cc: David S. Miller <davem@davemloft.net>
> > ---
> >  net/bluetooth/hci_sock.c |    3 ++-
> >  1 files changed, 2 insertions(+), 1 deletions(-)
> 
> Applied, thanks.

How much sense does it make to apply this single patch without the 
previous x32 infrastructure changes?

-- 
Jiri Kosina
SUSE Labs

Re: [PATCH 10/30] compat: Use COMPAT_USE_64BIT_TIME in the Bluetooth subsystem

[Gustavo Padovan]

Hi Jiri,

* Jiri Kosina <jkosina@suse.cz> [2012-02-22 14:47:29 +0100]:

> On Tue, 21 Feb 2012, Gustavo Padovan wrote:
> 
> > > From: "H. J. Lu" <hjl.tools@gmail.com>
> > > 
> > > Enable the Bluetooth subsystem to be used with a compat ABI with
> > > 64-bit time.
> > > 
> > > Signed-off-by: H. Peter Anvin <hpa@zytor.com>
> > > Cc: Marcel Holtmann <marcel@holtmann.org>
> > > Cc: Gustavo F. Padovan <padovan@profusion.mobi>
> > > Cc: David S. Miller <davem@davemloft.net>
> > > ---
> > >  net/bluetooth/hci_sock.c |    3 ++-
> > >  1 files changed, 2 insertions(+), 1 deletions(-)
> > 
> > Applied, thanks.
> 
> How much sense does it make to apply this single patch without the 
> previous x32 infrastructure changes?

No sense, I already stepped back with this changes. That's what happens when
you try to apply patches in the middle of the Brazilian Carnival. :)

	Gustavo

Re: [PATCH 02/30] x86-64: Use explicit sizes in sigcontext.h, prepare for x32

[H. Peter Anvin]

On 02/22/2012 04:22 AM, Arnd Bergmann wrote:
> On Monday 20 February 2012, H. Peter Anvin wrote:
>> We are using __u64 as x86-32 compatible since we are sharing most of the
>> really complex path (like ioctl) with i386 much more so than x86-64.  So
>> it is defined in userspace as:
>>
>> typedef unsigned long long __u64 __attribute__((aligned(4)));
>>
>> __aligned_u64 obviously is naturally aligned, which matches uint64_t is
>> userspace.
> 
> Has someone audited the interfaces to check if there are data structures that
> use a plain signed or unsigned "long long" instead of __s64/__u64 in places
> where i386 differs from the other compat implementations?
> 
> I found DRM_IOCTL_UPDATE_DRAW, but there could be more like this one.
> 

Has someone audited every single ioctl in the kernel?  Definitely not,
which is why x32 is marked EXPERIMENTAL.  I think it is still time for
this work to switch to happening in the upstream, however.

	-hpa


-- 
H. Peter Anvin, Intel Open Source Technology Center
I work for Intel.  I don't speak on their behalf.

Re: [PATCH 02/30] x86-64: Use explicit sizes in sigcontext.h, prepare for x32

[Arnd Bergmann]

On Wednesday 22 February 2012, H. Peter Anvin wrote:
> On 02/22/2012 04:22 AM, Arnd Bergmann wrote:
> > On Monday 20 February 2012, H. Peter Anvin wrote:
> >> We are using __u64 as x86-32 compatible since we are sharing most of the
> >> really complex path (like ioctl) with i386 much more so than x86-64.  So
> >> it is defined in userspace as:
> >>
> >> typedef unsigned long long __u64 __attribute__((aligned(4)));
> >>
> >> __aligned_u64 obviously is naturally aligned, which matches uint64_t is
> >> userspace.
> > 
> > Has someone audited the interfaces to check if there are data structures that
> > use a plain signed or unsigned "long long" instead of __s64/__u64 in places
> > where i386 differs from the other compat implementations?
> > 
> > I found DRM_IOCTL_UPDATE_DRAW, but there could be more like this one.
> > 
> 
> Has someone audited every single ioctl in the kernel?  Definitely not,
> which is why x32 is marked EXPERIMENTAL.  I think it is still time for
> this work to switch to happening in the upstream, however.

Depends on how you want to do it. In some cases, the easiest answer
would be to change the data structure to use __u64 and be compatible
with i386. Once there are distros built using data structure with
padding around a long long, you have to use a run-time conditional
in the compat handler.

I'd say we should fix at least the ones that are easy to spot because
they already use compat_u64 or have an #ifdef CONFIG_X86_64 in compat
code. I've looked at everything I could find that fits into that category
and found only two locations. My expectation is that all other data
structures that would fall into this category are already broken
for 32 bit emulation on x86.

Signed-off-by: Arnd Bergmann <arnd@arndb.de>

diff --git a/include/drm/drm.h b/include/drm/drm.h
index 49d94ed..73b7c33 100644
--- a/include/drm/drm.h
+++ b/include/drm/drm.h
@@ -438,7 +438,7 @@ struct drm_update_draw {
 	drm_drawable_t handle;
 	unsigned int type;
 	unsigned int num;
-	unsigned long long data;
+	__u64 data;
 };
 
 /**
diff --git a/include/sound/asound.h b/include/sound/asound.h
index a2e4ff5..a17e96c 100644
--- a/include/sound/asound.h
+++ b/include/sound/asound.h
@@ -824,8 +824,8 @@ struct snd_ctl_elem_value {
 			long *value_ptr;	/* obsoleted */
 		} integer;
 		union {
-			long long value[64];
-			long long *value_ptr;	/* obsoleted */
+			__s64 value[64];
+			__s64 *value_ptr;	/* obsoleted */
 		} integer64;
 		union {
 			unsigned int item[128];
	Arnd

Re: [PATCH 02/30] x86-64: Use explicit sizes in sigcontext.h, prepare for x32

[H. Peter Anvin]

On 02/22/2012 01:23 PM, Arnd Bergmann wrote:
> 
> Depends on how you want to do it. In some cases, the easiest answer
> would be to change the data structure to use __u64 and be compatible
> with i386. Once there are distros built using data structure with
> padding around a long long, you have to use a run-time conditional
> in the compat handler.
> 

That we'd like to avoid, but as you said, they'd be broken or might
already be broken.

> I'd say we should fix at least the ones that are easy to spot because
> they already use compat_u64 or have an #ifdef CONFIG_X86_64 in compat
> code. I've looked at everything I could find that fits into that category
> and found only two locations. My expectation is that all other data
> structures that would fall into this category are already broken
> for 32 bit emulation on x86.
> 
> Signed-off-by: Arnd Bergmann <arnd@arndb.de>
> 
> diff --git a/include/drm/drm.h b/include/drm/drm.h
> index 49d94ed..73b7c33 100644
> --- a/include/drm/drm.h
> +++ b/include/drm/drm.h
> @@ -438,7 +438,7 @@ struct drm_update_draw {
>  	drm_drawable_t handle;
>  	unsigned int type;
>  	unsigned int num;
> -	unsigned long long data;
> +	__u64 data;
>  };
>  
>  /**
> diff --git a/include/sound/asound.h b/include/sound/asound.h
> index a2e4ff5..a17e96c 100644
> --- a/include/sound/asound.h
> +++ b/include/sound/asound.h
> @@ -824,8 +824,8 @@ struct snd_ctl_elem_value {
>  			long *value_ptr;	/* obsoleted */
>  		} integer;
>  		union {
> -			long long value[64];
> -			long long *value_ptr;	/* obsoleted */
> +			__s64 value[64];
> +			__s64 *value_ptr;	/* obsoleted */
>  		} integer64;
>  		union {
>  			unsigned int item[128];
> 	Arnd

Right, those are good starts.

	-hpa

-- 
H. Peter Anvin, Intel Open Source Technology Center
I work for Intel.  I don't speak on their behalf.

Re: [PATCH 02/30] x86-64: Use explicit sizes in sigcontext.h, prepare for x32

[H. Peter Anvin]

On 02/22/2012 01:23 PM, Arnd Bergmann wrote:
> 
> Depends on how you want to do it. In some cases, the easiest answer
> would be to change the data structure to use __u64 and be compatible
> with i386. Once there are distros built using data structure with
> padding around a long long, you have to use a run-time conditional
> in the compat handler.
> 

That we'd like to avoid, but as you said, they'd be broken or might
already be broken.

> I'd say we should fix at least the ones that are easy to spot because
> they already use compat_u64 or have an #ifdef CONFIG_X86_64 in compat
> code. I've looked at everything I could find that fits into that category
> and found only two locations. My expectation is that all other data
> structures that would fall into this category are already broken
> for 32 bit emulation on x86.
> 
> Signed-off-by: Arnd Bergmann <arnd@arndb.de>
> 
> diff --git a/include/drm/drm.h b/include/drm/drm.h
> index 49d94ed..73b7c33 100644
> --- a/include/drm/drm.h
> +++ b/include/drm/drm.h
> @@ -438,7 +438,7 @@ struct drm_update_draw {
>  	drm_drawable_t handle;
>  	unsigned int type;
>  	unsigned int num;
> -	unsigned long long data;
> +	__u64 data;
>  };
>  
>  /**
> diff --git a/include/sound/asound.h b/include/sound/asound.h
> index a2e4ff5..a17e96c 100644
> --- a/include/sound/asound.h
> +++ b/include/sound/asound.h
> @@ -824,8 +824,8 @@ struct snd_ctl_elem_value {
>  			long *value_ptr;	/* obsoleted */
>  		} integer;
>  		union {
> -			long long value[64];
> -			long long *value_ptr;	/* obsoleted */
> +			__s64 value[64];
> +			__s64 *value_ptr;	/* obsoleted */
>  		} integer64;
>  		union {
>  			unsigned int item[128];
> 	Arnd

Right, those are good starts.

	-hpa

-- 
H. Peter Anvin, Intel Open Source Technology Center
I work for Intel.  I don't speak on their behalf.

Re: [PATCH 02/30] x86-64: Use explicit sizes in sigcontext.h, prepare for x32

[H. Peter Anvin]

On 02/22/2012 01:23 PM, Arnd Bergmann wrote:
> 
> I'd say we should fix at least the ones that are easy to spot because
> they already use compat_u64 or have an #ifdef CONFIG_X86_64 in compat
> code. I've looked at everything I could find that fits into that category
> and found only two locations. My expectation is that all other data
> structures that would fall into this category are already broken
> for 32 bit emulation on x86.
> 
> Signed-off-by: Arnd Bergmann <arnd@arndb.de>
> 

Would you mind sending these as patches to the respective maintainers?
Since they don't depend on the x32 patchset per se it's probably the
best way to handle this.

Feel free to include my Acked-by: H. Peter Anvin <hpa@zytor.com>.

	-hpa

-- 
H. Peter Anvin, Intel Open Source Technology Center
I work for Intel.  I don't speak on their behalf.

Re: [PATCH 02/30] x86-64: Use explicit sizes in sigcontext.h, prepare for x32

[H. Peter Anvin]

On 02/23/2012 06:33 PM, H. Peter Anvin wrote:
> On 02/22/2012 01:23 PM, Arnd Bergmann wrote:
>>
>> I'd say we should fix at least the ones that are easy to spot because
>> they already use compat_u64 or have an #ifdef CONFIG_X86_64 in compat
>> code. I've looked at everything I could find that fits into that category
>> and found only two locations. My expectation is that all other data
>> structures that would fall into this category are already broken
>> for 32 bit emulation on x86.
>>
>> Signed-off-by: Arnd Bergmann <arnd@arndb.de>
>>
> 
> Would you mind sending these as patches to the respective maintainers?
> Since they don't depend on the x32 patchset per se it's probably the
> best way to handle this.
> 
> Feel free to include my Acked-by: H. Peter Anvin <hpa@zytor.com>.
> 

Ping?

	-hpa