Mirror of the gdb-patches mailing list
 help / color / mirror / Atom feed
* [patch] Fix CLONE_VM vs. TLS  [Re: Is CLONE_VM really needed in  gdbserver?]
       [not found] ` <20100127000821.GA29862@caradoc.them.org>
@ 2010-01-27 22:12   ` Jan Kratochvil
  2010-01-28 16:22     ` H.J. Lu
  2010-01-28 17:01     ` Daniel Jacobowitz
  0 siblings, 2 replies; 7+ messages in thread
From: Jan Kratochvil @ 2010-01-27 22:12 UTC (permalink / raw)
  To: gdb-patches; +Cc: H.J. Lu, GDB

On Wed, 27 Jan 2010 01:08:26 +0100, Daniel Jacobowitz wrote:
> On Tue, Jan 26, 2010 at 03:51:38PM -0800, H.J. Lu wrote:
> > Hi,
> > 
> > There is a race condition between gdbserver and ld.so on Linux/x86-64:
> > 
> > http://www.sourceware.org/bugzilla/show_bug.cgi?id=11214
> > 
> > Is CLONE_VM really needed? In general, CLONE_VM is a very bad
> > idea if there is any symbol lookup in both parent and child processes.
> 
> It is necessary because gdbserver supports uClinux.  However, on Linux
> we might be able to get away with fork (see linux_tracefork_child in
> gdb/linux-nat.c).

Coded it as suggested.

No regressions on {x86_64,x86_64-m32,i686}-fedora12-linux-gnu and on
x86_64-fedora12-linux-gnu via gdbserver.

Verified linux_supports_tracefork_flag gets still set with the patch.

Verified unsetting HAVE_FORK for gdb/gdbserver/ still works the same.

gdb/ already tests HAVE_FORK in config.in but it is brought in by other
macros, therefore rather added an explicit configure.ac test for it.

Have not found an easy enough uClinux disk image of some arch for qemu-*.


Regards,
Jan


gdb/
2010-01-27  Jan Kratochvil  <jan.kratochvil@redhat.com>

	PR libc/11214:
	* configure.ac (AC_CHECK_FUNCS): Check for fork.
	* configure: Regenerate.

gdb/gdbserver/
2010-01-27  Jan Kratochvil  <jan.kratochvil@redhat.com>

	PR libc/11214:
	* configure.ac (AC_CHECK_FUNCS): Check for fork.
	* configure: Regenerate.
	* config.in: Regenerate.
	* linux-low.c (linux_tracefork_child) [HAVE_FORK]: New.
	(linux_test_for_tracefork): Move `stack' in [!HAVE_FORK].
	(linux_test_for_tracefork) [HAVE_FORK]: New.

gdb/testsuite/
2010-01-27  Jan Kratochvil  <jan.kratochvil@redhat.com>

	PR libc/11214:
	* gdb.threads/current-lwp-dead.c (fn, main): Move CLONE_VM in [UCLINUX].

--- a/gdb/configure
+++ b/gdb/configure
@@ -11234,7 +11234,7 @@ for ac_func in canonicalize_file_name realpath getrusage getuid \
 		getgid pipe poll pread64 sbrk setpgid setpgrp setsid \
 		sigaction sigprocmask sigsetmask socketpair syscall \
 		ttrace wborder setlocale iconvlist libiconvlist btowc \
-		setrlimit getrlimit
+		setrlimit getrlimit fork
 do :
   as_ac_var=`$as_echo "ac_cv_func_$ac_func" | $as_tr_sh`
 ac_fn_c_check_func "$LINENO" "$ac_func" "$as_ac_var"
--- a/gdb/configure.ac
+++ b/gdb/configure.ac
@@ -802,7 +802,7 @@ AC_CHECK_FUNCS([canonicalize_file_name realpath getrusage getuid \
 		getgid pipe poll pread64 sbrk setpgid setpgrp setsid \
 		sigaction sigprocmask sigsetmask socketpair syscall \
 		ttrace wborder setlocale iconvlist libiconvlist btowc \
-		setrlimit getrlimit])
+		setrlimit getrlimit fork])
 AM_LANGINFO_CODESET
 
 # Check the return and argument types of ptrace.  No canned test for
--- a/gdb/gdbserver/config.in
+++ b/gdb/gdbserver/config.in
@@ -30,6 +30,9 @@
 /* Define to 1 if you have the <fcntl.h> header file. */
 #undef HAVE_FCNTL_H
 
+/* Define to 1 if you have the `fork' function. */
+#undef HAVE_FORK
+
 /* Define to 1 if you have the <inttypes.h> header file. */
 #undef HAVE_INTTYPES_H
 
--- a/gdb/gdbserver/configure
+++ b/gdb/gdbserver/configure
@@ -3783,7 +3783,7 @@ fi
 
 done
 
-for ac_func in pread pwrite pread64
+for ac_func in pread pwrite pread64 fork
 do :
   as_ac_var=`$as_echo "ac_cv_func_$ac_func" | $as_tr_sh`
 ac_fn_c_check_func "$LINENO" "$ac_func" "$as_ac_var"
@@ -4153,6 +4153,8 @@ else
 /* end confdefs.h.  */
 
 #define _SYSCALL32
+/* Needed for new procfs interface on sparc-solaris.  */
+#define _STRUCTURED_PROC 1
 #include <sys/procfs.h>
 int
 main ()
@@ -4188,6 +4190,8 @@ else
 /* end confdefs.h.  */
 
 #define _SYSCALL32
+/* Needed for new procfs interface on sparc-solaris.  */
+#define _STRUCTURED_PROC 1
 #include <sys/procfs.h>
 int
 main ()
@@ -4223,6 +4227,8 @@ else
 /* end confdefs.h.  */
 
 #define _SYSCALL32
+/* Needed for new procfs interface on sparc-solaris.  */
+#define _STRUCTURED_PROC 1
 #include <sys/procfs.h>
 int
 main ()
@@ -4258,6 +4264,8 @@ else
 /* end confdefs.h.  */
 
 #define _SYSCALL32
+/* Needed for new procfs interface on sparc-solaris.  */
+#define _STRUCTURED_PROC 1
 #include <sys/procfs.h>
 int
 main ()
--- a/gdb/gdbserver/configure.ac
+++ b/gdb/gdbserver/configure.ac
@@ -42,7 +42,7 @@ AC_CHECK_HEADERS(sgtty.h termio.h termios.h sys/reg.h string.h dnl
 		 errno.h fcntl.h signal.h sys/file.h malloc.h dnl
 		 sys/ioctl.h netinet/in.h sys/socket.h netdb.h dnl
 		 netinet/tcp.h arpa/inet.h sys/wait.h)
-AC_CHECK_FUNCS(pread pwrite pread64)
+AC_CHECK_FUNCS(pread pwrite pread64 fork)
 AC_REPLACE_FUNCS(memmem)
 
 dnl dladdr is glibc-specific.  It is used by thread-db.c but only for
--- a/gdb/gdbserver/linux-low.c
+++ b/gdb/gdbserver/linux-low.c
@@ -2586,6 +2586,14 @@ linux_tracefork_child (void *arg)
 {
   ptrace (PTRACE_TRACEME, 0, 0, 0);
   kill (getpid (), SIGSTOP);
+
+#ifdef HAVE_FORK
+
+  if (fork () == 0)
+    linux_tracefork_grandchild (NULL);
+
+#else /* !HAVE_FORK */
+
 #ifdef __ia64__
   __clone2 (linux_tracefork_grandchild, arg, STACK_SIZE,
 	    CLONE_VM | SIGCHLD, NULL);
@@ -2593,6 +2601,9 @@ linux_tracefork_child (void *arg)
   clone (linux_tracefork_grandchild, arg + STACK_SIZE,
 	 CLONE_VM | SIGCHLD, NULL);
 #endif
+
+#endif /* !HAVE_FORK */
+
   _exit (0);
 }
 
@@ -2605,18 +2616,31 @@ linux_test_for_tracefork (void)
 {
   int child_pid, ret, status;
   long second_pid;
+#ifndef HAVE_FORK
   char *stack = xmalloc (STACK_SIZE * 4);
+#endif /* !HAVE_FORK */
 
   linux_supports_tracefork_flag = 0;
 
+#ifdef HAVE_FORK
+
+  child_pid = fork ();
+  if (child_pid == 0)
+    linux_tracefork_child (NULL);
+
+#else /* !HAVE_FORK */
+
   /* Use CLONE_VM instead of fork, to support uClinux (no MMU).  */
 #ifdef __ia64__
   child_pid = __clone2 (linux_tracefork_child, stack, STACK_SIZE,
 			CLONE_VM | SIGCHLD, stack + STACK_SIZE * 2);
-#else
+#else /* !__ia64__ */
   child_pid = clone (linux_tracefork_child, stack + STACK_SIZE,
 		     CLONE_VM | SIGCHLD, stack + STACK_SIZE * 2);
-#endif
+#endif /* !__ia64__ */
+
+#endif /* !HAVE_FORK */
+
   if (child_pid == -1)
     perror_with_name ("clone");
 
@@ -2685,7 +2709,9 @@ linux_test_for_tracefork (void)
     }
   while (WIFSTOPPED (status));
 
+#ifndef HAVE_FORK
   free (stack);
+#endif /* !HAVE_FORK */
 }
 
 
--- a/gdb/testsuite/gdb.threads/current-lwp-dead.c
+++ b/gdb/testsuite/gdb.threads/current-lwp-dead.c
@@ -51,8 +51,11 @@ fn (void *unused)
   stack = malloc (STACK_SIZE);
   assert (stack != NULL);
 
-  new_pid = clone (fn_return, stack + STACK_SIZE, CLONE_FILES | CLONE_VM, NULL,
-		   NULL, NULL, NULL);
+  new_pid = clone (fn_return, stack + STACK_SIZE, CLONE_FILES
+#ifdef UCLINUX
+		   | CLONE_VM
+#endif /* UCLINUX */
+		   , NULL, NULL, NULL, NULL);
   assert (new_pid > 0);
 
   return 0;
@@ -67,8 +70,11 @@ main (int argc, char **argv)
   stack = malloc (STACK_SIZE);
   assert (stack != NULL);
 
-  new_pid = clone (fn, stack + STACK_SIZE, CLONE_FILES | CLONE_VM, NULL, NULL,
-		   NULL, NULL);
+  new_pid = clone (fn, stack + STACK_SIZE, CLONE_FILES
+#ifdef UCLINUX
+		   | CLONE_VM
+#endif /* UCLINUX */
+		   , NULL, NULL, NULL, NULL);
   assert (new_pid > 0);
 
   return 0;


^ permalink raw reply	[flat|nested] 7+ messages in thread

* Re: [patch] Fix CLONE_VM vs. TLS [Re: Is CLONE_VM really needed in   gdbserver?]
  2010-01-27 22:12   ` [patch] Fix CLONE_VM vs. TLS [Re: Is CLONE_VM really needed in gdbserver?] Jan Kratochvil
@ 2010-01-28 16:22     ` H.J. Lu
  2010-01-28 17:01     ` Daniel Jacobowitz
  1 sibling, 0 replies; 7+ messages in thread
From: H.J. Lu @ 2010-01-28 16:22 UTC (permalink / raw)
  To: Jan Kratochvil; +Cc: gdb-patches, GDB

On Wed, Jan 27, 2010 at 2:12 PM, Jan Kratochvil
<jan.kratochvil@redhat.com> wrote:
> On Wed, 27 Jan 2010 01:08:26 +0100, Daniel Jacobowitz wrote:
>> On Tue, Jan 26, 2010 at 03:51:38PM -0800, H.J. Lu wrote:
>> > Hi,
>> >
>> > There is a race condition between gdbserver and ld.so on Linux/x86-64:
>> >
>> > http://www.sourceware.org/bugzilla/show_bug.cgi?id=11214
>> >
>> > Is CLONE_VM really needed? In general, CLONE_VM is a very bad
>> > idea if there is any symbol lookup in both parent and child processes.
>>
>> It is necessary because gdbserver supports uClinux.  However, on Linux
>> we might be able to get away with fork (see linux_tracefork_child in
>> gdb/linux-nat.c).
>
> Coded it as suggested.
>
> No regressions on {x86_64,x86_64-m32,i686}-fedora12-linux-gnu and on
> x86_64-fedora12-linux-gnu via gdbserver.
>
> Verified linux_supports_tracefork_flag gets still set with the patch.
>
> Verified unsetting HAVE_FORK for gdb/gdbserver/ still works the same.
>
> gdb/ already tests HAVE_FORK in config.in but it is brought in by other
> macros, therefore rather added an explicit configure.ac test for it.
>
> Have not found an easy enough uClinux disk image of some arch for qemu-*.
>
>
>

I think it is a glibc regression. I will try to fix it in glibc.


-- 
H.J.


^ permalink raw reply	[flat|nested] 7+ messages in thread

* Re: [patch] Fix CLONE_VM vs. TLS  [Re: Is CLONE_VM really needed  in  gdbserver?]
  2010-01-27 22:12   ` [patch] Fix CLONE_VM vs. TLS [Re: Is CLONE_VM really needed in gdbserver?] Jan Kratochvil
  2010-01-28 16:22     ` H.J. Lu
@ 2010-01-28 17:01     ` Daniel Jacobowitz
  2010-01-29  1:26       ` Jan Kratochvil
  1 sibling, 1 reply; 7+ messages in thread
From: Daniel Jacobowitz @ 2010-01-28 17:01 UTC (permalink / raw)
  To: Jan Kratochvil; +Cc: gdb-patches, H.J. Lu, GDB

On Wed, Jan 27, 2010 at 11:12:36PM +0100, Jan Kratochvil wrote:
> No regressions on {x86_64,x86_64-m32,i686}-fedora12-linux-gnu and on
> x86_64-fedora12-linux-gnu via gdbserver.
> 
> Verified linux_supports_tracefork_flag gets still set with the patch.
> 
> Verified unsetting HAVE_FORK for gdb/gdbserver/ still works the same.
> 
> gdb/ already tests HAVE_FORK in config.in but it is brought in by other
> macros, therefore rather added an explicit configure.ac test for it.
> 
> Have not found an easy enough uClinux disk image of some arch for qemu-*.

I don't recommend testing HAVE_FORK; uClibc may provide a dummy fork
that returns ENOSYS.  There's a HAS_NOMMU in linux-low.c that should work.

For the test case, what defines UCLINUX?

-- 
Daniel Jacobowitz
CodeSourcery


^ permalink raw reply	[flat|nested] 7+ messages in thread

* Re: [patch] Fix CLONE_VM vs. TLS  [Re: Is CLONE_VM really needed  in   gdbserver?]
  2010-01-28 17:01     ` Daniel Jacobowitz
@ 2010-01-29  1:26       ` Jan Kratochvil
  2010-01-29  2:30         ` Daniel Jacobowitz
  0 siblings, 1 reply; 7+ messages in thread
From: Jan Kratochvil @ 2010-01-29  1:26 UTC (permalink / raw)
  To: gdb-patches; +Cc: H.J. Lu, GDB

On Thu, 28 Jan 2010 18:01:07 +0100, Daniel Jacobowitz wrote:
> I don't recommend testing HAVE_FORK; uClibc may provide a dummy fork
> that returns ENOSYS.  There's a HAS_NOMMU in linux-low.c that should work.

OK, copied the current usage of HAS_NOMMU.


> For the test case, what defines UCLINUX?

Seen in some Googled source but it is true nobody sets it.

CodeSourcery arm-2009q3-66-arm-uclinuxeabi-i686-pc-linux-gnu.tar.bz2 gcc sets
__uClinux__ although gdb/gdbserver/linux-low.c depends on __UCLIBC__.
CLONE_VM is a kernel feature so gcc should be more appropriate but maybe it
does not matter.


Thanks,
Jan


gdb/gdbserver/
2010-01-29  Jan Kratochvil  <jan.kratochvil@redhat.com>

	PR libc/11214:
	* linux-low.c (linux_tracefork_child) [!(__UCLIBC__ && HAS_NOMMU)]: New.
	(linux_test_for_tracefork): Move `stack' into [__UCLIBC__ && HAS_NOMMU].
	(linux_test_for_tracefork) [!(__UCLIBC__ && HAS_NOMMU)]: New.

gdb/testsuite/
2010-01-29  Jan Kratochvil  <jan.kratochvil@redhat.com>

	PR libc/11214:
	* gdb.threads/current-lwp-dead.c (fn, main): Move CLONE_VM into
	[__uClinux__].

--- a/gdb/gdbserver/linux-low.c
+++ b/gdb/gdbserver/linux-low.c
@@ -2586,6 +2586,14 @@ linux_tracefork_child (void *arg)
 {
   ptrace (PTRACE_TRACEME, 0, 0, 0);
   kill (getpid (), SIGSTOP);
+
+#if !(defined(__UCLIBC__) && defined(HAS_NOMMU))
+
+  if (fork () == 0)
+    linux_tracefork_grandchild (NULL);
+
+#else /* defined(__UCLIBC__) && defined(HAS_NOMMU) */
+
 #ifdef __ia64__
   __clone2 (linux_tracefork_grandchild, arg, STACK_SIZE,
 	    CLONE_VM | SIGCHLD, NULL);
@@ -2593,6 +2601,9 @@ linux_tracefork_child (void *arg)
   clone (linux_tracefork_grandchild, arg + STACK_SIZE,
 	 CLONE_VM | SIGCHLD, NULL);
 #endif
+
+#endif /* defined(__UCLIBC__) && defined(HAS_NOMMU) */
+
   _exit (0);
 }
 
@@ -2605,18 +2616,31 @@ linux_test_for_tracefork (void)
 {
   int child_pid, ret, status;
   long second_pid;
+#if defined(__UCLIBC__) && defined(HAS_NOMMU)
   char *stack = xmalloc (STACK_SIZE * 4);
+#endif /* defined(__UCLIBC__) && defined(HAS_NOMMU) */
 
   linux_supports_tracefork_flag = 0;
 
+#if !(defined(__UCLIBC__) && defined(HAS_NOMMU))
+
+  child_pid = fork ();
+  if (child_pid == 0)
+    linux_tracefork_child (NULL);
+
+#else /* defined(__UCLIBC__) && defined(HAS_NOMMU) */
+
   /* Use CLONE_VM instead of fork, to support uClinux (no MMU).  */
 #ifdef __ia64__
   child_pid = __clone2 (linux_tracefork_child, stack, STACK_SIZE,
 			CLONE_VM | SIGCHLD, stack + STACK_SIZE * 2);
-#else
+#else /* !__ia64__ */
   child_pid = clone (linux_tracefork_child, stack + STACK_SIZE,
 		     CLONE_VM | SIGCHLD, stack + STACK_SIZE * 2);
-#endif
+#endif /* !__ia64__ */
+
+#endif /* defined(__UCLIBC__) && defined(HAS_NOMMU) */
+
   if (child_pid == -1)
     perror_with_name ("clone");
 
@@ -2685,7 +2709,9 @@ linux_test_for_tracefork (void)
     }
   while (WIFSTOPPED (status));
 
+#if defined(__UCLIBC__) && defined(HAS_NOMMU)
   free (stack);
+#endif /* defined(__UCLIBC__) && defined(HAS_NOMMU) */
 }
 
 
--- a/gdb/testsuite/gdb.threads/current-lwp-dead.c
+++ b/gdb/testsuite/gdb.threads/current-lwp-dead.c
@@ -51,8 +51,11 @@ fn (void *unused)
   stack = malloc (STACK_SIZE);
   assert (stack != NULL);
 
-  new_pid = clone (fn_return, stack + STACK_SIZE, CLONE_FILES | CLONE_VM, NULL,
-		   NULL, NULL, NULL);
+  new_pid = clone (fn_return, stack + STACK_SIZE, CLONE_FILES
+#ifdef __uClinux__
+		   | CLONE_VM
+#endif /* __uClinux__ */
+		   , NULL, NULL, NULL, NULL);
   assert (new_pid > 0);
 
   return 0;
@@ -67,8 +70,11 @@ main (int argc, char **argv)
   stack = malloc (STACK_SIZE);
   assert (stack != NULL);
 
-  new_pid = clone (fn, stack + STACK_SIZE, CLONE_FILES | CLONE_VM, NULL, NULL,
-		   NULL, NULL);
+  new_pid = clone (fn, stack + STACK_SIZE, CLONE_FILES
+#ifdef __uClinux__
+		   | CLONE_VM
+#endif /* __uClinux__ */
+		   , NULL, NULL, NULL, NULL);
   assert (new_pid > 0);
 
   return 0;


^ permalink raw reply	[flat|nested] 7+ messages in thread

* Re: [patch] Fix CLONE_VM vs. TLS  [Re: Is CLONE_VM really needed   in   gdbserver?]
  2010-01-29  1:26       ` Jan Kratochvil
@ 2010-01-29  2:30         ` Daniel Jacobowitz
  2010-01-29 12:41           ` Jan Kratochvil
  0 siblings, 1 reply; 7+ messages in thread
From: Daniel Jacobowitz @ 2010-01-29  2:30 UTC (permalink / raw)
  To: Jan Kratochvil; +Cc: gdb-patches, H.J. Lu, GDB

On Fri, Jan 29, 2010 at 02:26:25AM +0100, Jan Kratochvil wrote:
> CodeSourcery arm-2009q3-66-arm-uclinuxeabi-i686-pc-linux-gnu.tar.bz2 gcc sets
> __uClinux__ although gdb/gdbserver/linux-low.c depends on __UCLIBC__.
> CLONE_VM is a kernel feature so gcc should be more appropriate but maybe it
> does not matter.

It's the __UCLIBC_HAS_MMU__ / __ARCH_HAS_MMU__ that make this OK; the end
result is functionally equivalent to a __uClinux__ check.  uClibc can
be used on full Linux too, and there you have fork.

> 
> 
> Thanks,
> Jan
> 
> 
> gdb/gdbserver/
> 2010-01-29  Jan Kratochvil  <jan.kratochvil@redhat.com>
> 
> 	PR libc/11214:
> 	* linux-low.c (linux_tracefork_child) [!(__UCLIBC__ && HAS_NOMMU)]: New.
> 	(linux_test_for_tracefork): Move `stack' into [__UCLIBC__ && HAS_NOMMU].
> 	(linux_test_for_tracefork) [!(__UCLIBC__ && HAS_NOMMU)]: New.
> 
> gdb/testsuite/
> 2010-01-29  Jan Kratochvil  <jan.kratochvil@redhat.com>
> 
> 	PR libc/11214:
> 	* gdb.threads/current-lwp-dead.c (fn, main): Move CLONE_VM into
> 	[__uClinux__].

OK.

-- 
Daniel Jacobowitz
CodeSourcery


^ permalink raw reply	[flat|nested] 7+ messages in thread

* Re: [patch] Fix CLONE_VM vs. TLS  [Re: Is CLONE_VM really needed in    gdbserver?]
  2010-01-29  2:30         ` Daniel Jacobowitz
@ 2010-01-29 12:41           ` Jan Kratochvil
  2010-02-01 20:20             ` Jan Kratochvil
  0 siblings, 1 reply; 7+ messages in thread
From: Jan Kratochvil @ 2010-01-29 12:41 UTC (permalink / raw)
  To: gdb-patches; +Cc: H.J. Lu, GDB

On Fri, 29 Jan 2010 03:30:43 +0100, Daniel Jacobowitz wrote:
> On Fri, Jan 29, 2010 at 02:26:25AM +0100, Jan Kratochvil wrote:
> > CodeSourcery arm-2009q3-66-arm-uclinuxeabi-i686-pc-linux-gnu.tar.bz2 gcc sets
> > __uClinux__ although gdb/gdbserver/linux-low.c depends on __UCLIBC__.
> > CLONE_VM is a kernel feature so gcc should be more appropriate but maybe it
> > does not matter.
> 
> It's the __UCLIBC_HAS_MMU__ / __ARCH_HAS_MMU__ that make this OK; the end
> result is functionally equivalent to a __uClinux__ check.  uClibc can
> be used on full Linux too, and there you have fork.

In such case __uClinux__ can be unified with what is already in use in
linux-low.c.

Going to check it in in some times as this change is IMO pre-approved.


Thanks,
Jan


gdb/gdbserver/
2010-01-29  Jan Kratochvil  <jan.kratochvil@redhat.com>

	PR libc/11214:
	* linux-low.c (linux_tracefork_child) [!(__UCLIBC__ && HAS_NOMMU)]: New.
	(linux_test_for_tracefork): Move `stack' into [__UCLIBC__ && HAS_NOMMU].
	(linux_test_for_tracefork) [!(__UCLIBC__ && HAS_NOMMU)]: New.

gdb/testsuite/
2010-01-29  Jan Kratochvil  <jan.kratochvil@redhat.com>

	PR libc/11214:
	* gdb.threads/current-lwp-dead.c: Include features.h.
	(HAS_NOMMU): New.
	(fn, main): Move CLONE_VM into [__UCLIBC__ && HAS_NOMMU].

--- a/gdb/gdbserver/linux-low.c
+++ b/gdb/gdbserver/linux-low.c
@@ -2586,6 +2586,14 @@ linux_tracefork_child (void *arg)
 {
   ptrace (PTRACE_TRACEME, 0, 0, 0);
   kill (getpid (), SIGSTOP);
+
+#if !(defined(__UCLIBC__) && defined(HAS_NOMMU))
+
+  if (fork () == 0)
+    linux_tracefork_grandchild (NULL);
+
+#else /* defined(__UCLIBC__) && defined(HAS_NOMMU) */
+
 #ifdef __ia64__
   __clone2 (linux_tracefork_grandchild, arg, STACK_SIZE,
 	    CLONE_VM | SIGCHLD, NULL);
@@ -2593,6 +2601,9 @@ linux_tracefork_child (void *arg)
   clone (linux_tracefork_grandchild, arg + STACK_SIZE,
 	 CLONE_VM | SIGCHLD, NULL);
 #endif
+
+#endif /* defined(__UCLIBC__) && defined(HAS_NOMMU) */
+
   _exit (0);
 }
 
@@ -2605,18 +2616,31 @@ linux_test_for_tracefork (void)
 {
   int child_pid, ret, status;
   long second_pid;
+#if defined(__UCLIBC__) && defined(HAS_NOMMU)
   char *stack = xmalloc (STACK_SIZE * 4);
+#endif /* defined(__UCLIBC__) && defined(HAS_NOMMU) */
 
   linux_supports_tracefork_flag = 0;
 
+#if !(defined(__UCLIBC__) && defined(HAS_NOMMU))
+
+  child_pid = fork ();
+  if (child_pid == 0)
+    linux_tracefork_child (NULL);
+
+#else /* defined(__UCLIBC__) && defined(HAS_NOMMU) */
+
   /* Use CLONE_VM instead of fork, to support uClinux (no MMU).  */
 #ifdef __ia64__
   child_pid = __clone2 (linux_tracefork_child, stack, STACK_SIZE,
 			CLONE_VM | SIGCHLD, stack + STACK_SIZE * 2);
-#else
+#else /* !__ia64__ */
   child_pid = clone (linux_tracefork_child, stack + STACK_SIZE,
 		     CLONE_VM | SIGCHLD, stack + STACK_SIZE * 2);
-#endif
+#endif /* !__ia64__ */
+
+#endif /* defined(__UCLIBC__) && defined(HAS_NOMMU) */
+
   if (child_pid == -1)
     perror_with_name ("clone");
 
@@ -2685,7 +2709,9 @@ linux_test_for_tracefork (void)
     }
   while (WIFSTOPPED (status));
 
+#if defined(__UCLIBC__) && defined(HAS_NOMMU)
   free (stack);
+#endif /* defined(__UCLIBC__) && defined(HAS_NOMMU) */
 }
 
 
--- a/gdb/testsuite/gdb.threads/current-lwp-dead.c
+++ b/gdb/testsuite/gdb.threads/current-lwp-dead.c
@@ -30,6 +30,13 @@
 #include <unistd.h>
 #include <stdlib.h>
 
+#include <features.h>
+#ifdef __UCLIBC__
+#if !(defined(__UCLIBC_HAS_MMU__) || defined(__ARCH_HAS_MMU__))
+#define HAS_NOMMU
+#endif
+#endif
+
 #define STACK_SIZE 0x1000
 
 static int
@@ -51,8 +58,11 @@ fn (void *unused)
   stack = malloc (STACK_SIZE);
   assert (stack != NULL);
 
-  new_pid = clone (fn_return, stack + STACK_SIZE, CLONE_FILES | CLONE_VM, NULL,
-		   NULL, NULL, NULL);
+  new_pid = clone (fn_return, stack + STACK_SIZE, CLONE_FILES
+#if defined(__UCLIBC__) && defined(HAS_NOMMU)
+		   | CLONE_VM
+#endif /* defined(__UCLIBC__) && defined(HAS_NOMMU) */
+		   , NULL, NULL, NULL, NULL);
   assert (new_pid > 0);
 
   return 0;
@@ -67,8 +77,11 @@ main (int argc, char **argv)
   stack = malloc (STACK_SIZE);
   assert (stack != NULL);
 
-  new_pid = clone (fn, stack + STACK_SIZE, CLONE_FILES | CLONE_VM, NULL, NULL,
-		   NULL, NULL);
+  new_pid = clone (fn, stack + STACK_SIZE, CLONE_FILES
+#if defined(__UCLIBC__) && defined(HAS_NOMMU)
+		   | CLONE_VM
+#endif /* defined(__UCLIBC__) && defined(HAS_NOMMU) */
+		   , NULL, NULL, NULL, NULL);
   assert (new_pid > 0);
 
   return 0;


^ permalink raw reply	[flat|nested] 7+ messages in thread

* Re: [patch] Fix CLONE_VM vs. TLS  [Re: Is CLONE_VM really needed in     gdbserver?]
  2010-01-29 12:41           ` Jan Kratochvil
@ 2010-02-01 20:20             ` Jan Kratochvil
  0 siblings, 0 replies; 7+ messages in thread
From: Jan Kratochvil @ 2010-02-01 20:20 UTC (permalink / raw)
  To: gdb-patches; +Cc: H.J. Lu, GDB

On Fri, 29 Jan 2010 13:41:28 +0100, Jan Kratochvil wrote:
> Going to check it in in some times as this change is IMO pre-approved.
[...]
> gdb/gdbserver/
> 2010-01-29  Jan Kratochvil  <jan.kratochvil@redhat.com>
> 
> 	PR libc/11214:
> 	* linux-low.c (linux_tracefork_child) [!(__UCLIBC__ && HAS_NOMMU)]: New.
> 	(linux_test_for_tracefork): Move `stack' into [__UCLIBC__ && HAS_NOMMU].
> 	(linux_test_for_tracefork) [!(__UCLIBC__ && HAS_NOMMU)]: New.
> 
> gdb/testsuite/
> 2010-01-29  Jan Kratochvil  <jan.kratochvil@redhat.com>
> 
> 	PR libc/11214:
> 	* gdb.threads/current-lwp-dead.c: Include features.h.
> 	(HAS_NOMMU): New.
> 	(fn, main): Move CLONE_VM into [__UCLIBC__ && HAS_NOMMU].

Checked-in:
	http://sourceware.org/ml/gdb-cvs/2010-02/msg00010.html


Regards,
Jan


^ permalink raw reply	[flat|nested] 7+ messages in thread

end of thread, other threads:[~2010-02-01 20:20 UTC | newest]

Thread overview: 7+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
     [not found] <6dc9ffc81001261551j6221db6v88e96713d6dd9497@mail.gmail.com>
     [not found] ` <20100127000821.GA29862@caradoc.them.org>
2010-01-27 22:12   ` [patch] Fix CLONE_VM vs. TLS [Re: Is CLONE_VM really needed in gdbserver?] Jan Kratochvil
2010-01-28 16:22     ` H.J. Lu
2010-01-28 17:01     ` Daniel Jacobowitz
2010-01-29  1:26       ` Jan Kratochvil
2010-01-29  2:30         ` Daniel Jacobowitz
2010-01-29 12:41           ` Jan Kratochvil
2010-02-01 20:20             ` Jan Kratochvil

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox