[PATCH 1/3] um: Disable scan_elf_aux() on x86_64

[PATCH 1/3] um: Disable scan_elf_aux() on x86_64

[Richard Weinberger]

Reusing the host's vsyscall area makes only sense on x86_32.

Signed-off-by: Richard Weinberger <richard@nod.at>
---
 arch/um/os-Linux/Makefile  |    4 +++-
 arch/um/os-Linux/elf_aux.c |    7 +------
 arch/um/os-Linux/main.c    |    4 ++++
 3 files changed, 8 insertions(+), 7 deletions(-)

diff --git a/arch/um/os-Linux/Makefile b/arch/um/os-Linux/Makefile
index d66f038..b33f4df 100644
--- a/arch/um/os-Linux/Makefile
+++ b/arch/um/os-Linux/Makefile
@@ -3,10 +3,12 @@
 # Licensed under the GPL
 #
 
-obj-y = aio.o elf_aux.o execvp.o file.o helper.o irq.o main.o mem.o process.o \
+obj-y = aio.o execvp.o file.o helper.o irq.o main.o mem.o process.o \
 	registers.o sigio.o signal.o start_up.o time.o tty.o uaccess.o \
 	umid.o tls.o user_syms.o util.o drivers/ sys-$(SUBARCH)/ skas/
 
+obj-$(CONFIG_ARCH_REUSE_HOST_VSYSCALL_AREA) += elf_aux.o
+
 USER_OBJS := $(user-objs-y) aio.o elf_aux.o execvp.o file.o helper.o irq.o \
 	main.o mem.o process.o registers.o sigio.o signal.o start_up.o time.o \
 	tty.o tls.o uaccess.o umid.o util.o
diff --git a/arch/um/os-Linux/elf_aux.c b/arch/um/os-Linux/elf_aux.c
index 608784d..9533237 100644
--- a/arch/um/os-Linux/elf_aux.c
+++ b/arch/um/os-Linux/elf_aux.c
@@ -14,16 +14,11 @@
 #include "mem_user.h"
 #include <kern_constants.h>
 
-/* Use the one from the kernel - the host may miss it, if having old headers. */
-#if UM_ELF_CLASS == UM_ELFCLASS32
 typedef Elf32_auxv_t elf_auxv_t;
-#else
-typedef Elf64_auxv_t elf_auxv_t;
-#endif
 
 /* These are initialized very early in boot and never changed */
 char * elf_aux_platform;
-long elf_aux_hwcap;
+extern long elf_aux_hwcap;
 unsigned long vsyscall_ehdr;
 unsigned long vsyscall_end;
 unsigned long __kernel_vsyscall;
diff --git a/arch/um/os-Linux/main.c b/arch/um/os-Linux/main.c
index fb2a97a..8471b81 100644
--- a/arch/um/os-Linux/main.c
+++ b/arch/um/os-Linux/main.c
@@ -21,6 +21,8 @@
 #define STACKSIZE (8 * 1024 * 1024)
 #define THREAD_NAME_LEN (256)
 
+long elf_aux_hwcap;
+
 static void set_stklim(void)
 {
 	struct rlimit lim;
@@ -143,7 +145,9 @@ int __init main(int argc, char **argv, char **envp)
 	install_fatal_handler(SIGINT);
 	install_fatal_handler(SIGTERM);
 
+#ifdef CONFIG_ARCH_REUSE_HOST_VSYSCALL_AREA
 	scan_elf_aux(envp);
+#endif
 
 	do_uml_initcalls();
 	ret = linux_main(argc, argv);
-- 
1.7.6

[PATCH 2/3] um: Set __HAVE_ARCH_GATE_AREA

[Richard Weinberger]

Implement arch_vma_name() and make get_gate_vma(),
in_gate_area() and in_gate_area_no_mm() a nop.

We need arch_vma_name() to support vDSO.

Signed-off-by: Richard Weinberger <richard@nod.at>
---
 arch/um/sys-x86_64/asm/elf.h |    1 +
 arch/um/sys-x86_64/mem.c     |   22 ++++++++++++++++++++++
 2 files changed, 23 insertions(+), 0 deletions(-)

diff --git a/arch/um/sys-x86_64/asm/elf.h b/arch/um/sys-x86_64/asm/elf.h
index d6d5af3..74eb0ac 100644
--- a/arch/um/sys-x86_64/asm/elf.h
+++ b/arch/um/sys-x86_64/asm/elf.h
@@ -119,4 +119,5 @@ extern long elf_aux_hwcap;
 
 #define SET_PERSONALITY(ex) do ; while(0)
 
+#define __HAVE_ARCH_GATE_AREA 1
 #endif
diff --git a/arch/um/sys-x86_64/mem.c b/arch/um/sys-x86_64/mem.c
index 3f8df8a..1bf301e 100644
--- a/arch/um/sys-x86_64/mem.c
+++ b/arch/um/sys-x86_64/mem.c
@@ -14,3 +14,25 @@ unsigned long vm_data_default_flags = __VM_DATA_DEFAULT_FLAGS;
 unsigned long vm_data_default_flags32 = __VM_DATA_DEFAULT_FLAGS;
 unsigned long vm_force_exec32 = PROT_EXEC;
 
+const char *arch_vma_name(struct vm_area_struct *vma)
+{
+	if (vma->vm_mm && vma->vm_start == um_vdso_addr)
+		return "[vdso]";
+
+	return NULL;
+}
+
+struct vm_area_struct *get_gate_vma(struct mm_struct *mm)
+{
+	return 0;
+}
+
+int in_gate_area(struct mm_struct *mm, unsigned long addr)
+{
+	return 0;
+}
+
+int in_gate_area_no_mm(unsigned long addr)
+{
+	return 0;
+}
-- 
1.7.6

[uml-devel] [PATCH 3/3] um: Implement a x86_64 vDSO

[Richard Weinberger]

Until now UML had no x86_64 vDSO.
So glibc always used the vsyscall page for gettimeday()
and friends.
Calls to gettimeday() returned falsely the host time and
confused some programs.

This patch adds a vDSO which turns all __vdso_* calls into
a system call so that UML can trap them.

As glibc still uses the vsyscall page for static binaries this
patch improves the situation only for dynamic binaries.

Signed-off-by: Richard Weinberger <richard@nod.at>
---
 arch/um/sys-x86_64/Makefile               |    2 +
 arch/um/sys-x86_64/asm/elf.h              |    9 +++
 arch/um/sys-x86_64/vdso/Makefile          |   90 +++++++++++++++++++++++++++++
 arch/um/sys-x86_64/vdso/checkundef.sh     |   10 +++
 arch/um/sys-x86_64/vdso/um_vdso.c         |   71 +++++++++++++++++++++++
 arch/um/sys-x86_64/vdso/vdso-layout.lds.S |   64 ++++++++++++++++++++
 arch/um/sys-x86_64/vdso/vdso-note.S       |   12 ++++
 arch/um/sys-x86_64/vdso/vdso.S            |   10 +++
 arch/um/sys-x86_64/vdso/vdso.lds.S        |   32 ++++++++++
 arch/um/sys-x86_64/vdso/vma.c             |   67 +++++++++++++++++++++
 10 files changed, 367 insertions(+), 0 deletions(-)
 create mode 100644 arch/um/sys-x86_64/vdso/Makefile
 create mode 100755 arch/um/sys-x86_64/vdso/checkundef.sh
 create mode 100644 arch/um/sys-x86_64/vdso/um_vdso.c
 create mode 100644 arch/um/sys-x86_64/vdso/vdso-layout.lds.S
 create mode 100644 arch/um/sys-x86_64/vdso/vdso-note.S
 create mode 100644 arch/um/sys-x86_64/vdso/vdso.S
 create mode 100644 arch/um/sys-x86_64/vdso/vdso.lds.S
 create mode 100644 arch/um/sys-x86_64/vdso/vma.c

diff --git a/arch/um/sys-x86_64/Makefile b/arch/um/sys-x86_64/Makefile
index c1ea9eb..cc70f7a 100644
--- a/arch/um/sys-x86_64/Makefile
+++ b/arch/um/sys-x86_64/Makefile
@@ -8,6 +8,8 @@ obj-y = bug.o bugs.o delay.o fault.o ldt.o mem.o ptrace.o ptrace_user.o \
 	setjmp.o signal.o stub.o stub_segv.o syscalls.o syscall_table.o \
 	sysrq.o ksyms.o tls.o
 
+obj-y += vdso/
+
 subarch-obj-y = lib/csum-partial_64.o lib/memcpy_64.o lib/thunk_64.o \
 		lib/rwsem_64.o
 subarch-obj-$(CONFIG_MODULES) += kernel/module.o
diff --git a/arch/um/sys-x86_64/asm/elf.h b/arch/um/sys-x86_64/asm/elf.h
index 74eb0ac..11a2bfb 100644
--- a/arch/um/sys-x86_64/asm/elf.h
+++ b/arch/um/sys-x86_64/asm/elf.h
@@ -120,4 +120,13 @@ extern long elf_aux_hwcap;
 #define SET_PERSONALITY(ex) do ; while(0)
 
 #define __HAVE_ARCH_GATE_AREA 1
+#define ARCH_HAS_SETUP_ADDITIONAL_PAGES 1
+struct linux_binprm;
+extern int arch_setup_additional_pages(struct linux_binprm *bprm,
+	int uses_interp);
+
+extern unsigned long um_vdso_addr;
+#define AT_SYSINFO_EHDR 33
+#define ARCH_DLINFO	NEW_AUX_ENT(AT_SYSINFO_EHDR, um_vdso_addr)
+
 #endif
diff --git a/arch/um/sys-x86_64/vdso/Makefile b/arch/um/sys-x86_64/vdso/Makefile
new file mode 100644
index 0000000..5dffe6d
--- /dev/null
+++ b/arch/um/sys-x86_64/vdso/Makefile
@@ -0,0 +1,90 @@
+#
+# Building vDSO images for x86.
+#
+
+VDSO64-y		:= y
+
+vdso-install-$(VDSO64-y)	+= vdso.so
+
+
+# files to link into the vdso
+vobjs-y := vdso-note.o um_vdso.o
+
+# files to link into kernel
+obj-$(VDSO64-y)			+= vdso.o vma.o
+
+vobjs := $(foreach F,$(vobjs-y),$(obj)/$F)
+
+$(obj)/vdso.o: $(obj)/vdso.so
+
+targets += vdso.so vdso.so.dbg vdso.lds $(vobjs-y)
+
+export CPPFLAGS_vdso.lds += -P -C
+
+VDSO_LDFLAGS_vdso.lds = -m64 -Wl,-soname=linux-vdso.so.1 \
+       -Wl,-z,max-page-size=4096 -Wl,-z,common-page-size=4096
+
+$(obj)/vdso.o: $(src)/vdso.S $(obj)/vdso.so
+
+$(obj)/vdso.so.dbg: $(src)/vdso.lds $(vobjs) FORCE
+	$(call if_changed,vdso)
+
+$(obj)/%.so: OBJCOPYFLAGS := -S
+$(obj)/%.so: $(obj)/%.so.dbg FORCE
+	$(call if_changed,objcopy)
+
+#
+# Don't omit frame pointers for ease of userspace debugging, but do
+# optimize sibling calls.
+#
+CFL := $(PROFILING) -mcmodel=small -fPIC -O2 -fasynchronous-unwind-tables -m64 \
+       $(filter -g%,$(KBUILD_CFLAGS)) $(call cc-option, -fno-stack-protector) \
+       -fno-omit-frame-pointer -foptimize-sibling-calls
+
+$(vobjs): KBUILD_CFLAGS += $(CFL)
+
+#
+# vDSO code runs in userspace and -pg doesn't help with profiling anyway.
+#
+CFLAGS_REMOVE_vdso-note.o = -pg
+CFLAGS_REMOVE_um_vdso.o = -pg
+
+targets += vdso-syms.lds
+obj-$(VDSO64-y)			+= vdso-syms.lds
+
+#
+# Match symbols in the DSO that look like VDSO*; produce a file of constants.
+#
+sed-vdsosym := -e 's/^00*/0/' \
+	-e 's/^\([0-9a-fA-F]*\) . \(VDSO[a-zA-Z0-9_]*\)$$/\2 = 0x\1;/p'
+quiet_cmd_vdsosym = VDSOSYM $@
+define cmd_vdsosym
+	$(NM) $< | LC_ALL=C sed -n $(sed-vdsosym) | LC_ALL=C sort > $@
+endef
+
+$(obj)/%-syms.lds: $(obj)/%.so.dbg FORCE
+	$(call if_changed,vdsosym)
+
+#
+# The DSO images are built using a special linker script.
+#
+quiet_cmd_vdso = VDSO    $@
+      cmd_vdso = $(CC) -nostdlib -o $@ \
+		       $(VDSO_LDFLAGS) $(VDSO_LDFLAGS_$(filter %.lds,$(^F))) \
+		       -Wl,-T,$(filter %.lds,$^) $(filter %.o,$^) && \
+		 sh $(srctree)/$(src)/checkundef.sh '$(NM)' '$@'
+
+VDSO_LDFLAGS = -fPIC -shared $(call cc-ldoption, -Wl$(comma)--hash-style=sysv)
+GCOV_PROFILE := n
+
+#
+# Install the unstripped copy of vdso*.so listed in $(vdso-install-y).
+#
+quiet_cmd_vdso_install = INSTALL $@
+      cmd_vdso_install = cp $(obj)/$@.dbg $(MODLIB)/vdso/$@
+$(vdso-install-y): %.so: $(obj)/%.so.dbg FORCE
+	@mkdir -p $(MODLIB)/vdso
+	$(call cmd,vdso_install)
+
+PHONY += vdso_install $(vdso-install-y)
+vdso_install: $(vdso-install-y)
diff --git a/arch/um/sys-x86_64/vdso/checkundef.sh b/arch/um/sys-x86_64/vdso/checkundef.sh
new file mode 100755
index 0000000..7ee90a9
--- /dev/null
+++ b/arch/um/sys-x86_64/vdso/checkundef.sh
@@ -0,0 +1,10 @@
+#!/bin/sh
+nm="$1"
+file="$2"
+$nm "$file" | grep '^ *U' > /dev/null 2>&1
+if [ $? -eq 1 ]; then
+    exit 0
+else
+    echo "$file: undefined symbols found" >&2
+    exit 1
+fi
diff --git a/arch/um/sys-x86_64/vdso/um_vdso.c b/arch/um/sys-x86_64/vdso/um_vdso.c
new file mode 100644
index 0000000..7c441b5
--- /dev/null
+++ b/arch/um/sys-x86_64/vdso/um_vdso.c
@@ -0,0 +1,71 @@
+/*
+ * Copyright (C) 2011 Richard Weinberger <richrd@nod.at>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ *
+ * This vDSO turns all calls into a syscall so that UML can trap them.
+ */
+
+
+/* Disable profiling for userspace code */
+#define DISABLE_BRANCH_PROFILING
+
+#include <linux/time.h>
+#include <linux/getcpu.h>
+#include <asm/unistd.h>
+
+int __vdso_clock_gettime(clockid_t clock, struct timespec *ts)
+{
+	long ret;
+
+	asm("syscall" : "=a" (ret) :
+		"0" (__NR_clock_gettime), "D" (clock), "S" (ts) : "memory");
+
+	return ret;
+}
+int clock_gettime(clockid_t, struct timespec *)
+	__attribute__((weak, alias("__vdso_clock_gettime")));
+
+int __vdso_gettimeofday(struct timeval *tv, struct timezone *tz)
+{
+	long ret;
+
+	asm("syscall" : "=a" (ret) :
+		"0" (__NR_gettimeofday), "D" (tv), "S" (tz) : "memory");
+
+	return ret;
+}
+int gettimeofday(struct timeval *, struct timezone *)
+	__attribute__((weak, alias("__vdso_gettimeofday")));
+
+time_t __vdso_time(time_t *t)
+{
+	long secs;
+
+	asm volatile("syscall"
+		: "=a" (secs)
+		: "0" (__NR_time), "D" (t) : "cc", "r11", "cx", "memory");
+
+	return secs;
+}
+int time(time_t *t) __attribute__((weak, alias("__vdso_time")));
+
+long
+__vdso_getcpu(unsigned *cpu, unsigned *node, struct getcpu_cache *unused)
+{
+	/*
+	 * UML does not support SMP, we can cheat here. :)
+	 */
+
+	if (cpu)
+		*cpu = 0;
+	if (node)
+		*node = 0;
+
+	return 0;
+}
+
+long getcpu(unsigned *cpu, unsigned *node, struct getcpu_cache *tcache)
+	__attribute__((weak, alias("__vdso_getcpu")));
diff --git a/arch/um/sys-x86_64/vdso/vdso-layout.lds.S b/arch/um/sys-x86_64/vdso/vdso-layout.lds.S
new file mode 100644
index 0000000..634a2cf
--- /dev/null
+++ b/arch/um/sys-x86_64/vdso/vdso-layout.lds.S
@@ -0,0 +1,64 @@
+/*
+ * Linker script for vDSO.  This is an ELF shared object prelinked to
+ * its virtual address, and with only one read-only segment.
+ * This script controls its layout.
+ */
+
+SECTIONS
+{
+	. = VDSO_PRELINK + SIZEOF_HEADERS;
+
+	.hash		: { *(.hash) }			:text
+	.gnu.hash	: { *(.gnu.hash) }
+	.dynsym		: { *(.dynsym) }
+	.dynstr		: { *(.dynstr) }
+	.gnu.version	: { *(.gnu.version) }
+	.gnu.version_d	: { *(.gnu.version_d) }
+	.gnu.version_r	: { *(.gnu.version_r) }
+
+	.note		: { *(.note.*) }		:text	:note
+
+	.eh_frame_hdr	: { *(.eh_frame_hdr) }		:text	:eh_frame_hdr
+	.eh_frame	: { KEEP (*(.eh_frame)) }	:text
+
+	.dynamic	: { *(.dynamic) }		:text	:dynamic
+
+	.rodata		: { *(.rodata*) }		:text
+	.data		: {
+	      *(.data*)
+	      *(.sdata*)
+	      *(.got.plt) *(.got)
+	      *(.gnu.linkonce.d.*)
+	      *(.bss*)
+	      *(.dynbss*)
+	      *(.gnu.linkonce.b.*)
+	}
+
+	.altinstructions	: { *(.altinstructions) }
+	.altinstr_replacement	: { *(.altinstr_replacement) }
+
+	/*
+	 * Align the actual code well away from the non-instruction data.
+	 * This is the best thing for the I-cache.
+	 */
+	. = ALIGN(0x100);
+
+	.text		: { *(.text*) }			:text	=0x90909090
+}
+
+/*
+ * Very old versions of ld do not recognize this name token; use the constant.
+ */
+#define PT_GNU_EH_FRAME	0x6474e550
+
+/*
+ * We must supply the ELF program headers explicitly to get just one
+ * PT_LOAD segment, and set the flags explicitly to make segments read-only.
+ */
+PHDRS
+{
+	text		PT_LOAD		FLAGS(5) FILEHDR PHDRS; /* PF_R|PF_X */
+	dynamic		PT_DYNAMIC	FLAGS(4);		/* PF_R */
+	note		PT_NOTE		FLAGS(4);		/* PF_R */
+	eh_frame_hdr	PT_GNU_EH_FRAME;
+}
diff --git a/arch/um/sys-x86_64/vdso/vdso-note.S b/arch/um/sys-x86_64/vdso/vdso-note.S
new file mode 100644
index 0000000..79a071e
--- /dev/null
+++ b/arch/um/sys-x86_64/vdso/vdso-note.S
@@ -0,0 +1,12 @@
+/*
+ * This supplies .note.* sections to go into the PT_NOTE inside the vDSO text.
+ * Here we can supply some information useful to userland.
+ */
+
+#include <linux/uts.h>
+#include <linux/version.h>
+#include <linux/elfnote.h>
+
+ELFNOTE_START(Linux, 0, "a")
+	.long LINUX_VERSION_CODE
+ELFNOTE_END
diff --git a/arch/um/sys-x86_64/vdso/vdso.S b/arch/um/sys-x86_64/vdso/vdso.S
new file mode 100644
index 0000000..ec82c16
--- /dev/null
+++ b/arch/um/sys-x86_64/vdso/vdso.S
@@ -0,0 +1,10 @@
+#include <linux/init.h>
+
+__INITDATA
+
+	.globl vdso_start, vdso_end
+vdso_start:
+	.incbin "arch/um/sys-x86_64/vdso/vdso.so"
+vdso_end:
+
+__FINIT
diff --git a/arch/um/sys-x86_64/vdso/vdso.lds.S b/arch/um/sys-x86_64/vdso/vdso.lds.S
new file mode 100644
index 0000000..b96b267
--- /dev/null
+++ b/arch/um/sys-x86_64/vdso/vdso.lds.S
@@ -0,0 +1,32 @@
+/*
+ * Linker script for 64-bit 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 0xffffffffff700000
+#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: *;
+	};
+}
+
+VDSO64_PRELINK = VDSO_PRELINK;
diff --git a/arch/um/sys-x86_64/vdso/vma.c b/arch/um/sys-x86_64/vdso/vma.c
new file mode 100644
index 0000000..b62827a
--- /dev/null
+++ b/arch/um/sys-x86_64/vdso/vma.c
@@ -0,0 +1,67 @@
+/*
+ * Copyright (C) 2011 Richard Weinberger <richrd@nod.at>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#include <linux/slab.h>
+#include <linux/sched.h>
+#include <linux/mm.h>
+#include <asm/page.h>
+#include <linux/init.h>
+
+unsigned int __read_mostly vdso_enabled = 1;
+unsigned long um_vdso_addr;
+
+extern unsigned long task_size;
+extern char vdso_start[], vdso_end[];
+
+static struct page **vdsop;
+
+static int __init init_vdso(void)
+{
+	struct page *um_vdso;
+
+	BUG_ON(vdso_end - vdso_start > PAGE_SIZE);
+
+	um_vdso_addr = task_size - PAGE_SIZE;
+
+	vdsop = kmalloc(GFP_KERNEL, sizeof(struct page *));
+	if (!vdsop)
+		goto oom;
+
+	um_vdso = alloc_page(GFP_KERNEL);
+	if (!um_vdso) {
+		kfree(vdsop);
+
+		goto oom;
+	}
+
+	copy_page(page_address(um_vdso), vdso_start);
+	*vdsop = um_vdso;
+
+	return 0;
+
+oom:
+	printk(KERN_ERR "Cannot allocate vdso\n");
+	vdso_enabled = 0;
+
+	return -ENOMEM;
+}
+subsys_initcall(init_vdso);
+
+int arch_setup_additional_pages(struct linux_binprm *bprm, int uses_interp)
+{
+	struct mm_struct *mm = current->mm;
+
+	if (!vdso_enabled)
+		return 0;
+
+	return install_special_mapping(mm, um_vdso_addr, PAGE_SIZE,
+		VM_READ|VM_EXEC|
+		VM_MAYREAD|VM_MAYWRITE|VM_MAYEXEC|
+		VM_ALWAYSDUMP,
+		vdsop);
+}
-- 
1.7.6


------------------------------------------------------------------------------
AppSumo Presents a FREE Video for the SourceForge Community by Eric 
Ries, the creator of the Lean Startup Methodology on "Lean Startup 
Secrets Revealed." This video shows you how to validate your ideas, 
optimize your ideas and identify your business strategy.
http://p.sf.net/sfu/appsumosfdev2dev
_______________________________________________
User-mode-linux-devel mailing list
User-mode-linux-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-devel

[PATCH 3/3] um: Implement a x86_64 vDSO

[Richard Weinberger]

Until now UML had no x86_64 vDSO.
So glibc always used the vsyscall page for gettimeday()
and friends.
Calls to gettimeday() returned falsely the host time and
confused some programs.

This patch adds a vDSO which turns all __vdso_* calls into
a system call so that UML can trap them.

As glibc still uses the vsyscall page for static binaries this
patch improves the situation only for dynamic binaries.

Signed-off-by: Richard Weinberger <richard@nod.at>
---
 arch/um/sys-x86_64/Makefile               |    2 +
 arch/um/sys-x86_64/asm/elf.h              |    9 +++
 arch/um/sys-x86_64/vdso/Makefile          |   90 +++++++++++++++++++++++++++++
 arch/um/sys-x86_64/vdso/checkundef.sh     |   10 +++
 arch/um/sys-x86_64/vdso/um_vdso.c         |   71 +++++++++++++++++++++++
 arch/um/sys-x86_64/vdso/vdso-layout.lds.S |   64 ++++++++++++++++++++
 arch/um/sys-x86_64/vdso/vdso-note.S       |   12 ++++
 arch/um/sys-x86_64/vdso/vdso.S            |   10 +++
 arch/um/sys-x86_64/vdso/vdso.lds.S        |   32 ++++++++++
 arch/um/sys-x86_64/vdso/vma.c             |   67 +++++++++++++++++++++
 10 files changed, 367 insertions(+), 0 deletions(-)
 create mode 100644 arch/um/sys-x86_64/vdso/Makefile
 create mode 100755 arch/um/sys-x86_64/vdso/checkundef.sh
 create mode 100644 arch/um/sys-x86_64/vdso/um_vdso.c
 create mode 100644 arch/um/sys-x86_64/vdso/vdso-layout.lds.S
 create mode 100644 arch/um/sys-x86_64/vdso/vdso-note.S
 create mode 100644 arch/um/sys-x86_64/vdso/vdso.S
 create mode 100644 arch/um/sys-x86_64/vdso/vdso.lds.S
 create mode 100644 arch/um/sys-x86_64/vdso/vma.c

diff --git a/arch/um/sys-x86_64/Makefile b/arch/um/sys-x86_64/Makefile
index c1ea9eb..cc70f7a 100644
--- a/arch/um/sys-x86_64/Makefile
+++ b/arch/um/sys-x86_64/Makefile
@@ -8,6 +8,8 @@ obj-y = bug.o bugs.o delay.o fault.o ldt.o mem.o ptrace.o ptrace_user.o \
 	setjmp.o signal.o stub.o stub_segv.o syscalls.o syscall_table.o \
 	sysrq.o ksyms.o tls.o
 
+obj-y += vdso/
+
 subarch-obj-y = lib/csum-partial_64.o lib/memcpy_64.o lib/thunk_64.o \
 		lib/rwsem_64.o
 subarch-obj-$(CONFIG_MODULES) += kernel/module.o
diff --git a/arch/um/sys-x86_64/asm/elf.h b/arch/um/sys-x86_64/asm/elf.h
index 74eb0ac..11a2bfb 100644
--- a/arch/um/sys-x86_64/asm/elf.h
+++ b/arch/um/sys-x86_64/asm/elf.h
@@ -120,4 +120,13 @@ extern long elf_aux_hwcap;
 #define SET_PERSONALITY(ex) do ; while(0)
 
 #define __HAVE_ARCH_GATE_AREA 1
+#define ARCH_HAS_SETUP_ADDITIONAL_PAGES 1
+struct linux_binprm;
+extern int arch_setup_additional_pages(struct linux_binprm *bprm,
+	int uses_interp);
+
+extern unsigned long um_vdso_addr;
+#define AT_SYSINFO_EHDR 33
+#define ARCH_DLINFO	NEW_AUX_ENT(AT_SYSINFO_EHDR, um_vdso_addr)
+
 #endif
diff --git a/arch/um/sys-x86_64/vdso/Makefile b/arch/um/sys-x86_64/vdso/Makefile
new file mode 100644
index 0000000..5dffe6d
--- /dev/null
+++ b/arch/um/sys-x86_64/vdso/Makefile
@@ -0,0 +1,90 @@
+#
+# Building vDSO images for x86.
+#
+
+VDSO64-y		:= y
+
+vdso-install-$(VDSO64-y)	+= vdso.so
+
+
+# files to link into the vdso
+vobjs-y := vdso-note.o um_vdso.o
+
+# files to link into kernel
+obj-$(VDSO64-y)			+= vdso.o vma.o
+
+vobjs := $(foreach F,$(vobjs-y),$(obj)/$F)
+
+$(obj)/vdso.o: $(obj)/vdso.so
+
+targets += vdso.so vdso.so.dbg vdso.lds $(vobjs-y)
+
+export CPPFLAGS_vdso.lds += -P -C
+
+VDSO_LDFLAGS_vdso.lds = -m64 -Wl,-soname=linux-vdso.so.1 \
+       -Wl,-z,max-page-size=4096 -Wl,-z,common-page-size=4096
+
+$(obj)/vdso.o: $(src)/vdso.S $(obj)/vdso.so
+
+$(obj)/vdso.so.dbg: $(src)/vdso.lds $(vobjs) FORCE
+	$(call if_changed,vdso)
+
+$(obj)/%.so: OBJCOPYFLAGS := -S
+$(obj)/%.so: $(obj)/%.so.dbg FORCE
+	$(call if_changed,objcopy)
+
+#
+# Don't omit frame pointers for ease of userspace debugging, but do
+# optimize sibling calls.
+#
+CFL := $(PROFILING) -mcmodel=small -fPIC -O2 -fasynchronous-unwind-tables -m64 \
+       $(filter -g%,$(KBUILD_CFLAGS)) $(call cc-option, -fno-stack-protector) \
+       -fno-omit-frame-pointer -foptimize-sibling-calls
+
+$(vobjs): KBUILD_CFLAGS += $(CFL)
+
+#
+# vDSO code runs in userspace and -pg doesn't help with profiling anyway.
+#
+CFLAGS_REMOVE_vdso-note.o = -pg
+CFLAGS_REMOVE_um_vdso.o = -pg
+
+targets += vdso-syms.lds
+obj-$(VDSO64-y)			+= vdso-syms.lds
+
+#
+# Match symbols in the DSO that look like VDSO*; produce a file of constants.
+#
+sed-vdsosym := -e 's/^00*/0/' \
+	-e 's/^\([0-9a-fA-F]*\) . \(VDSO[a-zA-Z0-9_]*\)$$/\2 = 0x\1;/p'
+quiet_cmd_vdsosym = VDSOSYM $@
+define cmd_vdsosym
+	$(NM) $< | LC_ALL=C sed -n $(sed-vdsosym) | LC_ALL=C sort > $@
+endef
+
+$(obj)/%-syms.lds: $(obj)/%.so.dbg FORCE
+	$(call if_changed,vdsosym)
+
+#
+# The DSO images are built using a special linker script.
+#
+quiet_cmd_vdso = VDSO    $@
+      cmd_vdso = $(CC) -nostdlib -o $@ \
+		       $(VDSO_LDFLAGS) $(VDSO_LDFLAGS_$(filter %.lds,$(^F))) \
+		       -Wl,-T,$(filter %.lds,$^) $(filter %.o,$^) && \
+		 sh $(srctree)/$(src)/checkundef.sh '$(NM)' '$@'
+
+VDSO_LDFLAGS = -fPIC -shared $(call cc-ldoption, -Wl$(comma)--hash-style=sysv)
+GCOV_PROFILE := n
+
+#
+# Install the unstripped copy of vdso*.so listed in $(vdso-install-y).
+#
+quiet_cmd_vdso_install = INSTALL $@
+      cmd_vdso_install = cp $(obj)/$@.dbg $(MODLIB)/vdso/$@
+$(vdso-install-y): %.so: $(obj)/%.so.dbg FORCE
+	@mkdir -p $(MODLIB)/vdso
+	$(call cmd,vdso_install)
+
+PHONY += vdso_install $(vdso-install-y)
+vdso_install: $(vdso-install-y)
diff --git a/arch/um/sys-x86_64/vdso/checkundef.sh b/arch/um/sys-x86_64/vdso/checkundef.sh
new file mode 100755
index 0000000..7ee90a9
--- /dev/null
+++ b/arch/um/sys-x86_64/vdso/checkundef.sh
@@ -0,0 +1,10 @@
+#!/bin/sh
+nm="$1"
+file="$2"
+$nm "$file" | grep '^ *U' > /dev/null 2>&1
+if [ $? -eq 1 ]; then
+    exit 0
+else
+    echo "$file: undefined symbols found" >&2
+    exit 1
+fi
diff --git a/arch/um/sys-x86_64/vdso/um_vdso.c b/arch/um/sys-x86_64/vdso/um_vdso.c
new file mode 100644
index 0000000..7c441b5
--- /dev/null
+++ b/arch/um/sys-x86_64/vdso/um_vdso.c
@@ -0,0 +1,71 @@
+/*
+ * Copyright (C) 2011 Richard Weinberger <richrd@nod.at>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ *
+ * This vDSO turns all calls into a syscall so that UML can trap them.
+ */
+
+
+/* Disable profiling for userspace code */
+#define DISABLE_BRANCH_PROFILING
+
+#include <linux/time.h>
+#include <linux/getcpu.h>
+#include <asm/unistd.h>
+
+int __vdso_clock_gettime(clockid_t clock, struct timespec *ts)
+{
+	long ret;
+
+	asm("syscall" : "=a" (ret) :
+		"0" (__NR_clock_gettime), "D" (clock), "S" (ts) : "memory");
+
+	return ret;
+}
+int clock_gettime(clockid_t, struct timespec *)
+	__attribute__((weak, alias("__vdso_clock_gettime")));
+
+int __vdso_gettimeofday(struct timeval *tv, struct timezone *tz)
+{
+	long ret;
+
+	asm("syscall" : "=a" (ret) :
+		"0" (__NR_gettimeofday), "D" (tv), "S" (tz) : "memory");
+
+	return ret;
+}
+int gettimeofday(struct timeval *, struct timezone *)
+	__attribute__((weak, alias("__vdso_gettimeofday")));
+
+time_t __vdso_time(time_t *t)
+{
+	long secs;
+
+	asm volatile("syscall"
+		: "=a" (secs)
+		: "0" (__NR_time), "D" (t) : "cc", "r11", "cx", "memory");
+
+	return secs;
+}
+int time(time_t *t) __attribute__((weak, alias("__vdso_time")));
+
+long
+__vdso_getcpu(unsigned *cpu, unsigned *node, struct getcpu_cache *unused)
+{
+	/*
+	 * UML does not support SMP, we can cheat here. :)
+	 */
+
+	if (cpu)
+		*cpu = 0;
+	if (node)
+		*node = 0;
+
+	return 0;
+}
+
+long getcpu(unsigned *cpu, unsigned *node, struct getcpu_cache *tcache)
+	__attribute__((weak, alias("__vdso_getcpu")));
diff --git a/arch/um/sys-x86_64/vdso/vdso-layout.lds.S b/arch/um/sys-x86_64/vdso/vdso-layout.lds.S
new file mode 100644
index 0000000..634a2cf
--- /dev/null
+++ b/arch/um/sys-x86_64/vdso/vdso-layout.lds.S
@@ -0,0 +1,64 @@
+/*
+ * Linker script for vDSO.  This is an ELF shared object prelinked to
+ * its virtual address, and with only one read-only segment.
+ * This script controls its layout.
+ */
+
+SECTIONS
+{
+	. = VDSO_PRELINK + SIZEOF_HEADERS;
+
+	.hash		: { *(.hash) }			:text
+	.gnu.hash	: { *(.gnu.hash) }
+	.dynsym		: { *(.dynsym) }
+	.dynstr		: { *(.dynstr) }
+	.gnu.version	: { *(.gnu.version) }
+	.gnu.version_d	: { *(.gnu.version_d) }
+	.gnu.version_r	: { *(.gnu.version_r) }
+
+	.note		: { *(.note.*) }		:text	:note
+
+	.eh_frame_hdr	: { *(.eh_frame_hdr) }		:text	:eh_frame_hdr
+	.eh_frame	: { KEEP (*(.eh_frame)) }	:text
+
+	.dynamic	: { *(.dynamic) }		:text	:dynamic
+
+	.rodata		: { *(.rodata*) }		:text
+	.data		: {
+	      *(.data*)
+	      *(.sdata*)
+	      *(.got.plt) *(.got)
+	      *(.gnu.linkonce.d.*)
+	      *(.bss*)
+	      *(.dynbss*)
+	      *(.gnu.linkonce.b.*)
+	}
+
+	.altinstructions	: { *(.altinstructions) }
+	.altinstr_replacement	: { *(.altinstr_replacement) }
+
+	/*
+	 * Align the actual code well away from the non-instruction data.
+	 * This is the best thing for the I-cache.
+	 */
+	. = ALIGN(0x100);
+
+	.text		: { *(.text*) }			:text	=0x90909090
+}
+
+/*
+ * Very old versions of ld do not recognize this name token; use the constant.
+ */
+#define PT_GNU_EH_FRAME	0x6474e550
+
+/*
+ * We must supply the ELF program headers explicitly to get just one
+ * PT_LOAD segment, and set the flags explicitly to make segments read-only.
+ */
+PHDRS
+{
+	text		PT_LOAD		FLAGS(5) FILEHDR PHDRS; /* PF_R|PF_X */
+	dynamic		PT_DYNAMIC	FLAGS(4);		/* PF_R */
+	note		PT_NOTE		FLAGS(4);		/* PF_R */
+	eh_frame_hdr	PT_GNU_EH_FRAME;
+}
diff --git a/arch/um/sys-x86_64/vdso/vdso-note.S b/arch/um/sys-x86_64/vdso/vdso-note.S
new file mode 100644
index 0000000..79a071e
--- /dev/null
+++ b/arch/um/sys-x86_64/vdso/vdso-note.S
@@ -0,0 +1,12 @@
+/*
+ * This supplies .note.* sections to go into the PT_NOTE inside the vDSO text.
+ * Here we can supply some information useful to userland.
+ */
+
+#include <linux/uts.h>
+#include <linux/version.h>
+#include <linux/elfnote.h>
+
+ELFNOTE_START(Linux, 0, "a")
+	.long LINUX_VERSION_CODE
+ELFNOTE_END
diff --git a/arch/um/sys-x86_64/vdso/vdso.S b/arch/um/sys-x86_64/vdso/vdso.S
new file mode 100644
index 0000000..ec82c16
--- /dev/null
+++ b/arch/um/sys-x86_64/vdso/vdso.S
@@ -0,0 +1,10 @@
+#include <linux/init.h>
+
+__INITDATA
+
+	.globl vdso_start, vdso_end
+vdso_start:
+	.incbin "arch/um/sys-x86_64/vdso/vdso.so"
+vdso_end:
+
+__FINIT
diff --git a/arch/um/sys-x86_64/vdso/vdso.lds.S b/arch/um/sys-x86_64/vdso/vdso.lds.S
new file mode 100644
index 0000000..b96b267
--- /dev/null
+++ b/arch/um/sys-x86_64/vdso/vdso.lds.S
@@ -0,0 +1,32 @@
+/*
+ * Linker script for 64-bit 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 0xffffffffff700000
+#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: *;
+	};
+}
+
+VDSO64_PRELINK = VDSO_PRELINK;
diff --git a/arch/um/sys-x86_64/vdso/vma.c b/arch/um/sys-x86_64/vdso/vma.c
new file mode 100644
index 0000000..b62827a
--- /dev/null
+++ b/arch/um/sys-x86_64/vdso/vma.c
@@ -0,0 +1,67 @@
+/*
+ * Copyright (C) 2011 Richard Weinberger <richrd@nod.at>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#include <linux/slab.h>
+#include <linux/sched.h>
+#include <linux/mm.h>
+#include <asm/page.h>
+#include <linux/init.h>
+
+unsigned int __read_mostly vdso_enabled = 1;
+unsigned long um_vdso_addr;
+
+extern unsigned long task_size;
+extern char vdso_start[], vdso_end[];
+
+static struct page **vdsop;
+
+static int __init init_vdso(void)
+{
+	struct page *um_vdso;
+
+	BUG_ON(vdso_end - vdso_start > PAGE_SIZE);
+
+	um_vdso_addr = task_size - PAGE_SIZE;
+
+	vdsop = kmalloc(GFP_KERNEL, sizeof(struct page *));
+	if (!vdsop)
+		goto oom;
+
+	um_vdso = alloc_page(GFP_KERNEL);
+	if (!um_vdso) {
+		kfree(vdsop);
+
+		goto oom;
+	}
+
+	copy_page(page_address(um_vdso), vdso_start);
+	*vdsop = um_vdso;
+
+	return 0;
+
+oom:
+	printk(KERN_ERR "Cannot allocate vdso\n");
+	vdso_enabled = 0;
+
+	return -ENOMEM;
+}
+subsys_initcall(init_vdso);
+
+int arch_setup_additional_pages(struct linux_binprm *bprm, int uses_interp)
+{
+	struct mm_struct *mm = current->mm;
+
+	if (!vdso_enabled)
+		return 0;
+
+	return install_special_mapping(mm, um_vdso_addr, PAGE_SIZE,
+		VM_READ|VM_EXEC|
+		VM_MAYREAD|VM_MAYWRITE|VM_MAYEXEC|
+		VM_ALWAYSDUMP,
+		vdsop);
+}
-- 
1.7.6

Re: [PATCH 3/3] um: Implement a x86_64 vDSO

[Andy Lutomirski]

On 07/18/2011 09:07 AM, Richard Weinberger wrote:
> Until now UML had no x86_64 vDSO.
> So glibc always used the vsyscall page for gettimeday()
> and friends.
> Calls to gettimeday() returned falsely the host time and
> confused some programs.
>
> This patch adds a vDSO which turns all __vdso_* calls into
> a system call so that UML can trap them.
>
> As glibc still uses the vsyscall page for static binaries this
> patch improves the situation only for dynamic binaries.
>
> Signed-off-by: Richard Weinberger<richard@nod.at>

> +	asm("syscall" : "=a" (ret) :
> +		"0" (__NR_clock_gettime), "D" (clock), "S" (ts) : "memory");
> +

The x86-64 ABI and glibc think that rcx, and r11 are clobbered.  glibc 
also thinks that cc is clobbered.  I personally think that rdx and 
r8-r15 ought to be clobbered under the principle that syscall should 
look like a function call.

It's hard to imagine this causing a problem in a function this trivial, 
though.

This applies to the other two syscall instructions as well.

--Andy

P.S.  If you really care, with the changes in tip/x86/vdso, it ought to 
be possible for UML to trap vsyscalls as well.  This would need a change 
to do_emulate_vsyscall to honor TIF_SYSCALL_EMU with intelligent 
semantics.  I know nothing about ptrace, so I'm not volunteering :)

Re: [PATCH 3/3] um: Implement a x86_64 vDSO

[Richard Weinberger]

Am Mittwoch 20 Juli 2011, 19:54:02 schrieb Andy Lutomirski:
> On 07/18/2011 09:07 AM, Richard Weinberger wrote:
> > Until now UML had no x86_64 vDSO.
> > So glibc always used the vsyscall page for gettimeday()
> > and friends.
> > Calls to gettimeday() returned falsely the host time and
> > confused some programs.
> > 
> > This patch adds a vDSO which turns all __vdso_* calls into
> > a system call so that UML can trap them.
> > 
> > As glibc still uses the vsyscall page for static binaries this
> > patch improves the situation only for dynamic binaries.
> > 
> > Signed-off-by: Richard Weinberger<richard@nod.at>
> > 
> > +	asm("syscall" : "=a" (ret) :
> > +		"0" (__NR_clock_gettime), "D" (clock), "S" (ts) : "memory");
> > +
> 
> The x86-64 ABI and glibc think that rcx, and r11 are clobbered.  glibc
> also thinks that cc is clobbered.  I personally think that rdx and
> r8-r15 ought to be clobbered under the principle that syscall should
> look like a function call.

Hmm, I thought my implementation is fine because it follows 
arch/x86/vdso/vclock_gettime.c (Linus' tree).

notrace static long vdso_fallback_gettime(long clock, struct timespec *ts)
{
  long ret;
  asm("syscall" : "=a" (ret) :
      "0" (__NR_clock_gettime),"D" (clock), "S" (ts) : "memory");
  return ret;
}

Am I missing something or needs vdso_fallback_gettime() also some updates?

> It's hard to imagine this causing a problem in a function this trivial,
> though.
> 
> This applies to the other two syscall instructions as well.
> 
> --Andy
> 
> P.S.  If you really care, with the changes in tip/x86/vdso, it ought to
> be possible for UML to trap vsyscalls as well.  This would need a change
> to do_emulate_vsyscall to honor TIF_SYSCALL_EMU with intelligent
> semantics.  I know nothing about ptrace, so I'm not volunteering :)

Ok, I'll have at look at it.

Thanks,
//richard

Re: [PATCH 3/3] um: Implement a x86_64 vDSO

[Andrew Lutomirski]

On Wed, Jul 20, 2011 at 2:09 PM, Richard Weinberger <richard@nod.at> wrote:
> Am Mittwoch 20 Juli 2011, 19:54:02 schrieb Andy Lutomirski:
>> On 07/18/2011 09:07 AM, Richard Weinberger wrote:
>> > Until now UML had no x86_64 vDSO.
>> > So glibc always used the vsyscall page for gettimeday()
>> > and friends.
>> > Calls to gettimeday() returned falsely the host time and
>> > confused some programs.
>> >
>> > This patch adds a vDSO which turns all __vdso_* calls into
>> > a system call so that UML can trap them.
>> >
>> > As glibc still uses the vsyscall page for static binaries this
>> > patch improves the situation only for dynamic binaries.
>> >
>> > Signed-off-by: Richard Weinberger<richard@nod.at>
>> >
>> > +   asm("syscall" : "=a" (ret) :
>> > +           "0" (__NR_clock_gettime), "D" (clock), "S" (ts) : "memory");
>> > +
>>
>> The x86-64 ABI and glibc think that rcx, and r11 are clobbered.  glibc
>> also thinks that cc is clobbered.  I personally think that rdx and
>> r8-r15 ought to be clobbered under the principle that syscall should
>> look like a function call.
>
> Hmm, I thought my implementation is fine because it follows
> arch/x86/vdso/vclock_gettime.c (Linus' tree).
>
> notrace static long vdso_fallback_gettime(long clock, struct timespec *ts)
> {
>  long ret;
>  asm("syscall" : "=a" (ret) :
>      "0" (__NR_clock_gettime),"D" (clock), "S" (ts) : "memory");
>  return ret;
> }
>
> Am I missing something or needs vdso_fallback_gettime() also some updates?

vdso_fallback_gettime is going away :)

I haven't read all of the sysret code paths, but it looks like at
least r11 is used in a comment.  That being said, your code is an
uninlinable function, and functions already clobber all those
registers, so the code shouldn't cause problems in real life.

--Andy
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

Re: [PATCH 3/3] um: Implement a x86_64 vDSO

[Andrew Lutomirski]

On Wed, Jul 20, 2011 at 2:09 PM, Richard Weinberger <richard@nod.at> wrote:
> Am Mittwoch 20 Juli 2011, 19:54:02 schrieb Andy Lutomirski:
>> On 07/18/2011 09:07 AM, Richard Weinberger wrote:
>> > Until now UML had no x86_64 vDSO.
>> > So glibc always used the vsyscall page for gettimeday()
>> > and friends.
>> > Calls to gettimeday() returned falsely the host time and
>> > confused some programs.
>> >
>> > This patch adds a vDSO which turns all __vdso_* calls into
>> > a system call so that UML can trap them.
>> >
>> > As glibc still uses the vsyscall page for static binaries this
>> > patch improves the situation only for dynamic binaries.
>> >
>> > Signed-off-by: Richard Weinberger<richard@nod.at>
>> >
>> > +   asm("syscall" : "=a" (ret) :
>> > +           "0" (__NR_clock_gettime), "D" (clock), "S" (ts) : "memory");
>> > +
>>
>> The x86-64 ABI and glibc think that rcx, and r11 are clobbered.  glibc
>> also thinks that cc is clobbered.  I personally think that rdx and
>> r8-r15 ought to be clobbered under the principle that syscall should
>> look like a function call.
>
> Hmm, I thought my implementation is fine because it follows
> arch/x86/vdso/vclock_gettime.c (Linus' tree).
>
> notrace static long vdso_fallback_gettime(long clock, struct timespec *ts)
> {
>  long ret;
>  asm("syscall" : "=a" (ret) :
>      "0" (__NR_clock_gettime),"D" (clock), "S" (ts) : "memory");
>  return ret;
> }
>
> Am I missing something or needs vdso_fallback_gettime() also some updates?

vdso_fallback_gettime is going away :)

I haven't read all of the sysret code paths, but it looks like at
least r11 is used in a comment.  That being said, your code is an
uninlinable function, and functions already clobber all those
registers, so the code shouldn't cause problems in real life.

--Andy