[PATCH] gcore fix for native gdb on solaris

[PATCH] gcore fix for native gdb on solaris

[Daniel Gutson]

[-- Attachment #1: Type: text/plain, Size: 978 bytes --]

The attached patch solves a bug that caused GDB to wrongly read core 
dump files generated with the gcore command, running on solaris.

A define (_STRUCTURE_PROC) was needed before including procfs.h.
Additionally, fp regs are properly read.

I tested this with the gdb test suite, comparing results before and 
after (the only differences were caused by timeouts).

Please let me know any comments, or commit it for me since I don't have 
write access.

Thanks,
	Daniel.

2009-11-10  Daniel Gutson  <dgutson@codesourcery.com>

	bfd/
	* bfd.m4 (BFD_HAVE_SYS_PROCFS_TYPE): Define _STRUCTURE_PROC
	before including procfs.h.
	(BFD_HAVE_SYS_PROCFS_TYPE_MEMBER): Likewise.
	* configure.in: Added autoconf probe for the pr_fpreg member.
	* configure: Regenerated.
	* config.in: Regenerated.
	* elf.c: Define _STRUCTURE_PROC before including procfs.h.

	gdb/
	* procfs.c: (procfs_do_thread_registers) Use
	elfcore_write_lwpstatus.

-- 
Daniel Gutson
CodeSourcery
www.codesourcery.com

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #2: procfs.patch --]
[-- Type: text/x-diff; name="procfs.patch", Size: 6703 bytes --]

Index: bfd/bfd.m4
===================================================================
RCS file: /cvs/src/src/bfd/bfd.m4,v
retrieving revision 1.6
diff -u -p -r1.6 bfd.m4
--- bfd/bfd.m4	7 Sep 2009 09:15:27 -0000	1.6
+++ bfd/bfd.m4	10 Nov 2009 22:06:31 -0000
@@ -8,6 +8,8 @@ AC_DEFUN([BFD_HAVE_SYS_PROCFS_TYPE],
  AC_CACHE_VAL(bfd_cv_have_sys_procfs_type_$1,
    [AC_TRY_COMPILE([
 #define _SYSCALL32
+/* Needed for new procfs interface on sparc-solaris.  */
+#define _STRUCTURED_PROC 1
 #include <sys/procfs.h>],
       [$1 avar],
       bfd_cv_have_sys_procfs_type_$1=yes,
@@ -27,6 +29,8 @@ AC_DEFUN([BFD_HAVE_SYS_PROCFS_TYPE_MEMBE
  AC_CACHE_VAL(bfd_cv_have_sys_procfs_type_member_$1_$2,
    [AC_TRY_COMPILE([
 #define _SYSCALL32
+/* Needed for new procfs interface on sparc-solaris.  */
+#define _STRUCTURED_PROC 1
 #include <sys/procfs.h>],
       [$1 avar; void* aref = (void*) &avar.$2],
       bfd_cv_have_sys_procfs_type_member_$1_$2=yes,
Index: bfd/config.in
===================================================================
RCS file: /cvs/src/src/bfd/config.in,v
retrieving revision 1.45
diff -u -p -r1.45 config.in
--- bfd/config.in	7 Sep 2009 09:15:27 -0000	1.45
+++ bfd/config.in	10 Nov 2009 22:06:31 -0000
@@ -114,6 +114,9 @@
 /* Define if <sys/procfs.h> has lwpstatus_t.pr_context. */
 #undef HAVE_LWPSTATUS_T_PR_CONTEXT
 
+/* Define if <sys/procfs.h> has lwpstatus_t.pr_fpreg. */
+#undef HAVE_LWPSTATUS_T_PR_FPREG
+
 /* Define if <sys/procfs.h> has lwpstatus_t.pr_reg. */
 #undef HAVE_LWPSTATUS_T_PR_REG
 
Index: bfd/configure
===================================================================
RCS file: /cvs/src/src/bfd/configure,v
retrieving revision 1.317
diff -u -p -r1.317 configure
--- bfd/configure	29 Sep 2009 14:17:03 -0000	1.317
+++ bfd/configure	10 Nov 2009 22:06:31 -0000
@@ -13923,6 +13923,8 @@ else
 /* end confdefs.h.  */
 
 #define _SYSCALL32
+/* Needed for new procfs interface on sparc-solaris.  */
+#define _STRUCTURED_PROC 1
 #include <sys/procfs.h>
 int
 main ()
@@ -13958,6 +13960,8 @@ else
 /* end confdefs.h.  */
 
 #define _SYSCALL32
+/* Needed for new procfs interface on sparc-solaris.  */
+#define _STRUCTURED_PROC 1
 #include <sys/procfs.h>
 int
 main ()
@@ -13993,6 +13997,8 @@ else
 /* end confdefs.h.  */
 
 #define _SYSCALL32
+/* Needed for new procfs interface on sparc-solaris.  */
+#define _STRUCTURED_PROC 1
 #include <sys/procfs.h>
 int
 main ()
@@ -14028,6 +14034,8 @@ else
 /* end confdefs.h.  */
 
 #define _SYSCALL32
+/* Needed for new procfs interface on sparc-solaris.  */
+#define _STRUCTURED_PROC 1
 #include <sys/procfs.h>
 int
 main ()
@@ -14063,6 +14071,8 @@ else
 /* end confdefs.h.  */
 
 #define _SYSCALL32
+/* Needed for new procfs interface on sparc-solaris.  */
+#define _STRUCTURED_PROC 1
 #include <sys/procfs.h>
 int
 main ()
@@ -14098,6 +14108,8 @@ else
 /* end confdefs.h.  */
 
 #define _SYSCALL32
+/* Needed for new procfs interface on sparc-solaris.  */
+#define _STRUCTURED_PROC 1
 #include <sys/procfs.h>
 int
 main ()
@@ -14133,6 +14145,8 @@ else
 /* end confdefs.h.  */
 
 #define _SYSCALL32
+/* Needed for new procfs interface on sparc-solaris.  */
+#define _STRUCTURED_PROC 1
 #include <sys/procfs.h>
 int
 main ()
@@ -14168,6 +14182,8 @@ else
 /* end confdefs.h.  */
 
 #define _SYSCALL32
+/* Needed for new procfs interface on sparc-solaris.  */
+#define _STRUCTURED_PROC 1
 #include <sys/procfs.h>
 int
 main ()
@@ -14203,6 +14219,8 @@ else
 /* end confdefs.h.  */
 
 #define _SYSCALL32
+/* Needed for new procfs interface on sparc-solaris.  */
+#define _STRUCTURED_PROC 1
 #include <sys/procfs.h>
 int
 main ()
@@ -14238,6 +14256,8 @@ else
 /* end confdefs.h.  */
 
 #define _SYSCALL32
+/* Needed for new procfs interface on sparc-solaris.  */
+#define _STRUCTURED_PROC 1
 #include <sys/procfs.h>
 int
 main ()
@@ -14273,6 +14293,8 @@ else
 /* end confdefs.h.  */
 
 #define _SYSCALL32
+/* Needed for new procfs interface on sparc-solaris.  */
+#define _STRUCTURED_PROC 1
 #include <sys/procfs.h>
 int
 main ()
@@ -14308,6 +14330,8 @@ else
 /* end confdefs.h.  */
 
 #define _SYSCALL32
+/* Needed for new procfs interface on sparc-solaris.  */
+#define _STRUCTURED_PROC 1
 #include <sys/procfs.h>
 int
 main ()
@@ -14343,6 +14367,8 @@ else
 /* end confdefs.h.  */
 
 #define _SYSCALL32
+/* Needed for new procfs interface on sparc-solaris.  */
+#define _STRUCTURED_PROC 1
 #include <sys/procfs.h>
 int
 main ()
@@ -14378,6 +14404,8 @@ else
 /* end confdefs.h.  */
 
 #define _SYSCALL32
+/* Needed for new procfs interface on sparc-solaris.  */
+#define _STRUCTURED_PROC 1
 #include <sys/procfs.h>
 int
 main ()
@@ -14413,6 +14441,8 @@ else
 /* end confdefs.h.  */
 
 #define _SYSCALL32
+/* Needed for new procfs interface on sparc-solaris.  */
+#define _STRUCTURED_PROC 1
 #include <sys/procfs.h>
 int
 main ()
@@ -14448,6 +14478,8 @@ else
 /* end confdefs.h.  */
 
 #define _SYSCALL32
+/* Needed for new procfs interface on sparc-solaris.  */
+#define _STRUCTURED_PROC 1
 #include <sys/procfs.h>
 int
 main ()
Index: bfd/configure.in
===================================================================
RCS file: /cvs/src/src/bfd/configure.in,v
retrieving revision 1.272
diff -u -p -r1.272 configure.in
--- bfd/configure.in	29 Sep 2009 14:17:03 -0000	1.272
+++ bfd/configure.in	10 Nov 2009 22:06:31 -0000
@@ -500,6 +500,7 @@ changequote([,])dnl
     BFD_HAVE_SYS_PROCFS_TYPE(lwpxstatus_t)
     BFD_HAVE_SYS_PROCFS_TYPE_MEMBER(lwpstatus_t, pr_context)
     BFD_HAVE_SYS_PROCFS_TYPE_MEMBER(lwpstatus_t, pr_reg)
+    BFD_HAVE_SYS_PROCFS_TYPE_MEMBER(lwpstatus_t, pr_fpreg)
     BFD_HAVE_SYS_PROCFS_TYPE(win32_pstatus_t)
   fi
 fi
Index: bfd/elf.c
===================================================================
RCS file: /cvs/src/src/bfd/elf.c,v
retrieving revision 1.493
diff -u -p -r1.493 elf.c
--- bfd/elf.c	28 Sep 2009 09:45:33 -0000	1.493
+++ bfd/elf.c	10 Nov 2009 22:06:31 -0000
@@ -7420,6 +7420,8 @@ _bfd_elf_rel_vtable_reloc_fn
    out details about the corefile.  */
 
 #ifdef HAVE_SYS_PROCFS_H
+/* Needed for new procfs interface on sparc-solaris.  */
+# define _STRUCTURED_PROC 1
 # include <sys/procfs.h>
 #endif
 
Index: gdb/procfs.c
===================================================================
RCS file: /cvs/src/src/gdb/procfs.c,v
retrieving revision 1.120
diff -u -p -r1.120 procfs.c
--- gdb/procfs.c	10 Nov 2009 21:00:41 -0000	1.120
+++ gdb/procfs.c	10 Nov 2009 22:06:31 -0000
@@ -6074,7 +6074,7 @@ procfs_do_thread_registers (bfd *obfd, p
   target_fetch_registers (regcache, -1);
 
   fill_gregset (regcache, &gregs, -1);
-#if defined (UNIXWARE)
+#if defined (UNIXWARE) || defined (NEW_PROC_API)
   note_data = (char *) elfcore_write_lwpstatus (obfd,
 						note_data,
 						note_size,

Re: [PATCH] gcore fix for native gdb on solaris

[Tom Tromey]

>>>>> "Daniel" == Daniel Gutson <dgutson@codesourcery.com> writes:

Daniel> 2009-11-10  Daniel Gutson  <dgutson@codesourcery.com>
Daniel> 	bfd/
Daniel> 	* bfd.m4 (BFD_HAVE_SYS_PROCFS_TYPE): Define _STRUCTURE_PROC
Daniel> 	before including procfs.h.
Daniel> 	(BFD_HAVE_SYS_PROCFS_TYPE_MEMBER): Likewise.
Daniel> 	* configure.in: Added autoconf probe for the pr_fpreg member.
Daniel> 	* configure: Regenerated.
Daniel> 	* config.in: Regenerated.
Daniel> 	* elf.c: Define _STRUCTURE_PROC before including procfs.h.

I think BFD changes have to go to the binutils list.

Tom

Re: [PATCH] gcore fix for native gdb on solaris

[Daniel Gutson]

Hi Tom,

Tom Tromey wrote:
>>>>>> "Daniel" == Daniel Gutson <dgutson@codesourcery.com> writes:
> 
> Daniel> 2009-11-10  Daniel Gutson  <dgutson@codesourcery.com>
> Daniel> 	bfd/
> Daniel> 	* bfd.m4 (BFD_HAVE_SYS_PROCFS_TYPE): Define _STRUCTURE_PROC
> Daniel> 	before including procfs.h.
> Daniel> 	(BFD_HAVE_SYS_PROCFS_TYPE_MEMBER): Likewise.
> Daniel> 	* configure.in: Added autoconf probe for the pr_fpreg member.
> Daniel> 	* configure: Regenerated.
> Daniel> 	* config.in: Regenerated.
> Daniel> 	* elf.c: Define _STRUCTURE_PROC before including procfs.h.
> 
> I think BFD changes have to go to the binutils list.

I've just committed the binutils part; please let me know if the gdb 
part is OK to commit.

Thanks!
	Daniel.

-- 
Daniel Gutson
CodeSourcery
www.codesourcery.com

Re: [PATCH] gcore fix for native gdb on solaris

[Tom Tromey]

>>>>> "Daniel" == Daniel Gutson <dgutson@codesourcery.com> writes:

Daniel> 2009-11-10  Daniel Gutson  <dgutson@codesourcery.com>
Daniel> bfd/
Daniel> * bfd.m4 (BFD_HAVE_SYS_PROCFS_TYPE): Define _STRUCTURE_PROC
Daniel> before including procfs.h.
Daniel> (BFD_HAVE_SYS_PROCFS_TYPE_MEMBER): Likewise.
Daniel> * configure.in: Added autoconf probe for the pr_fpreg member.
Daniel> * configure: Regenerated.
Daniel> * config.in: Regenerated.
Daniel> * elf.c: Define _STRUCTURE_PROC before including procfs.h.

Tom> I think BFD changes have to go to the binutils list.

Daniel> I've just committed the binutils part; please let me know if the gdb
Daniel> part is OK to commit.

Just to be clear, I'm afraid someone else will have to review this, I
don't know anything about this area.

Tom

Re: [PATCH] gcore fix for native gdb on solaris

[Joel Brobecker]

For the record:

> 	gdb/
> 	* procfs.c: (procfs_do_thread_registers) Use
> 	elfcore_write_lwpstatus.

I checked in the following patch as an emergency fix, because the bfd
change caused a build error in GDB (missing elfcore_write_prstatus).
It's your change, with a slight modification. See:

    http://www.sourceware.org/ml/gdb-patches/2010-01/msg00191.html
    [commit] Fix build failure on sparc-solaris.

-- 
Joel

Re: [PATCH] gcore fix for native gdb on solaris

[Jan Kratochvil]

Hi Daniel,

On Tue, 10 Nov 2009 23:14:21 +0100, Daniel Gutson wrote:
> A define (_STRUCTURE_PROC) was needed before including procfs.h.
...
> 2009-11-10  Daniel Gutson  <dgutson@codesourcery.com>
> 
> 	bfd/
> 	* bfd.m4 (BFD_HAVE_SYS_PROCFS_TYPE): Define _STRUCTURE_PROC
> 	before including procfs.h.
...
> Index: bfd/bfd.m4
> ===================================================================
> RCS file: /cvs/src/src/bfd/bfd.m4,v
> retrieving revision 1.6
> diff -u -p -r1.6 bfd.m4
> --- bfd/bfd.m4	7 Sep 2009 09:15:27 -0000	1.6
> +++ bfd/bfd.m4	10 Nov 2009 22:06:31 -0000
> @@ -8,6 +8,8 @@ AC_DEFUN([BFD_HAVE_SYS_PROCFS_TYPE],
>   AC_CACHE_VAL(bfd_cv_have_sys_procfs_type_$1,
>     [AC_TRY_COMPILE([
>  #define _SYSCALL32
> +/* Needed for new procfs interface on sparc-solaris.  */
> +#define _STRUCTURED_PROC 1
>  #include <sys/procfs.h>],
>        [$1 avar],
>        bfd_cv_have_sys_procfs_type_$1=yes,
> @@ -27,6 +29,8 @@ AC_DEFUN([BFD_HAVE_SYS_PROCFS_TYPE_MEMBE
>   AC_CACHE_VAL(bfd_cv_have_sys_procfs_type_member_$1_$2,
>     [AC_TRY_COMPILE([
>  #define _SYSCALL32
> +/* Needed for new procfs interface on sparc-solaris.  */
> +#define _STRUCTURED_PROC 1
>  #include <sys/procfs.h>],
>        [$1 avar; void* aref = (void*) &avar.$2],
>        bfd_cv_have_sys_procfs_type_member_$1_$2=yes,


This change has caused the following change in generated gdb/configure which
is currently not checked in.

Assuming it is OK to update gdb/configure.


Regards,
Jan


--- a/gdb/configure
+++ b/gdb/configure
@@ -11897,6 +11898,8 @@ else
 /* end confdefs.h.  */
 
 #define _SYSCALL32
+/* Needed for new procfs interface on sparc-solaris.  */
+#define _STRUCTURED_PROC 1
 #include <sys/procfs.h>
 int
 main ()
@@ -11932,6 +11935,8 @@ else
 /* end confdefs.h.  */
 
 #define _SYSCALL32
+/* Needed for new procfs interface on sparc-solaris.  */
+#define _STRUCTURED_PROC 1
 #include <sys/procfs.h>
 int
 main ()
@@ -11967,6 +11972,8 @@ else
 /* end confdefs.h.  */
 
 #define _SYSCALL32
+/* Needed for new procfs interface on sparc-solaris.  */
+#define _STRUCTURED_PROC 1
 #include <sys/procfs.h>
 int
 main ()
@@ -12002,6 +12009,8 @@ else
 /* end confdefs.h.  */
 
 #define _SYSCALL32
+/* Needed for new procfs interface on sparc-solaris.  */
+#define _STRUCTURED_PROC 1
 #include <sys/procfs.h>
 int
 main ()
@@ -12037,6 +12046,8 @@ else
 /* end confdefs.h.  */
 
 #define _SYSCALL32
+/* Needed for new procfs interface on sparc-solaris.  */
+#define _STRUCTURED_PROC 1
 #include <sys/procfs.h>
 int
 main ()
@@ -12072,6 +12083,8 @@ else
 /* end confdefs.h.  */
 
 #define _SYSCALL32
+/* Needed for new procfs interface on sparc-solaris.  */
+#define _STRUCTURED_PROC 1
 #include <sys/procfs.h>
 int
 main ()
@@ -12107,6 +12120,8 @@ else
 /* end confdefs.h.  */
 
 #define _SYSCALL32
+/* Needed for new procfs interface on sparc-solaris.  */
+#define _STRUCTURED_PROC 1
 #include <sys/procfs.h>
 int
 main ()
@@ -12142,6 +12157,8 @@ else
 /* end confdefs.h.  */
 
 #define _SYSCALL32
+/* Needed for new procfs interface on sparc-solaris.  */
+#define _STRUCTURED_PROC 1
 #include <sys/procfs.h>
 int
 main ()
@@ -12177,6 +12194,8 @@ else
 /* end confdefs.h.  */
 
 #define _SYSCALL32
+/* Needed for new procfs interface on sparc-solaris.  */
+#define _STRUCTURED_PROC 1
 #include <sys/procfs.h>
 int
 main ()
@@ -12212,6 +12231,8 @@ else
 /* end confdefs.h.  */
 
 #define _SYSCALL32
+/* Needed for new procfs interface on sparc-solaris.  */
+#define _STRUCTURED_PROC 1
 #include <sys/procfs.h>
 int
 main ()
@@ -12247,6 +12268,8 @@ else
 /* end confdefs.h.  */
 
 #define _SYSCALL32
+/* Needed for new procfs interface on sparc-solaris.  */
+#define _STRUCTURED_PROC 1
 #include <sys/procfs.h>
 int
 main ()
@@ -12282,6 +12305,8 @@ else
 /* end confdefs.h.  */
 
 #define _SYSCALL32
+/* Needed for new procfs interface on sparc-solaris.  */
+#define _STRUCTURED_PROC 1
 #include <sys/procfs.h>
 int
 main ()
@@ -12317,6 +12342,8 @@ else
 /* end confdefs.h.  */
 
 #define _SYSCALL32
+/* Needed for new procfs interface on sparc-solaris.  */
+#define _STRUCTURED_PROC 1
 #include <sys/procfs.h>
 int
 main ()
@@ -12352,6 +12379,8 @@ else
 /* end confdefs.h.  */
 
 #define _SYSCALL32
+/* Needed for new procfs interface on sparc-solaris.  */
+#define _STRUCTURED_PROC 1
 #include <sys/procfs.h>
 int
 main ()

Re: [PATCH] gcore fix for native gdb on solaris

[Joel Brobecker]

> This change has caused the following change in generated gdb/configure which
> is currently not checked in.
> 
> Assuming it is OK to update gdb/configure.

Yes, of course. Regenerating the configure script can be considered
obvious, particularly in this case.

Thanks for spotting this, Jan.

-- 
Joel

Re: [PATCH] gcore fix for native gdb on solaris

[Jan Kratochvil]

On Tue, 12 Jan 2010 04:39:05 +0100, Joel Brobecker wrote:
> > This change has caused the following change in generated gdb/configure which
> > is currently not checked in.
> > 
> > Assuming it is OK to update gdb/configure.
> 
> Yes, of course. Regenerating the configure script can be considered
> obvious, particularly in this case.

I was not somehow sure whether it is expected and I also wanted to make some
heads-up whether _STRUCTURED_PROC #defines aren't now obsoleted in proc*.c
files (I do not know if they are or not).


Thanks,
Jan


http://sourceware.org/ml/gdb-cvs/2010-01/msg00105.html

--- src/gdb/ChangeLog	2010/01/12 08:50:44	1.11238
+++ src/gdb/ChangeLog	2010/01/12 12:22:49	1.11239
@@ -1,3 +1,7 @@
+2010-01-12  Jan Kratochvil  <jan.kratochvil@redhat.com>
+
+	* configure: Regenerate (for _STRUCTURED_PROC).
+
 2010-01-12  Joel Brobecker  <brobecker@adacore.com>
 
 	Delete dead function.
--- src/gdb/configure	2010/01/08 07:16:43	1.295
+++ src/gdb/configure	2010/01/12 12:22:52	1.296
@@ -11897,6 +11897,8 @@
 /* end confdefs.h.  */
 
 #define _SYSCALL32
+/* Needed for new procfs interface on sparc-solaris.  */
+#define _STRUCTURED_PROC 1
 #include <sys/procfs.h>
 int
 main ()
@@ -11932,6 +11934,8 @@
 /* end confdefs.h.  */
 
 #define _SYSCALL32
+/* Needed for new procfs interface on sparc-solaris.  */
+#define _STRUCTURED_PROC 1
 #include <sys/procfs.h>
 int
 main ()
@@ -11967,6 +11971,8 @@
 /* end confdefs.h.  */
 
 #define _SYSCALL32
+/* Needed for new procfs interface on sparc-solaris.  */
+#define _STRUCTURED_PROC 1
 #include <sys/procfs.h>
 int
 main ()
@@ -12002,6 +12008,8 @@
 /* end confdefs.h.  */
 
 #define _SYSCALL32
+/* Needed for new procfs interface on sparc-solaris.  */
+#define _STRUCTURED_PROC 1
 #include <sys/procfs.h>
 int
 main ()
@@ -12037,6 +12045,8 @@
 /* end confdefs.h.  */
 
 #define _SYSCALL32
+/* Needed for new procfs interface on sparc-solaris.  */
+#define _STRUCTURED_PROC 1
 #include <sys/procfs.h>
 int
 main ()
@@ -12072,6 +12082,8 @@
 /* end confdefs.h.  */
 
 #define _SYSCALL32
+/* Needed for new procfs interface on sparc-solaris.  */
+#define _STRUCTURED_PROC 1
 #include <sys/procfs.h>
 int
 main ()
@@ -12107,6 +12119,8 @@
 /* end confdefs.h.  */
 
 #define _SYSCALL32
+/* Needed for new procfs interface on sparc-solaris.  */
+#define _STRUCTURED_PROC 1
 #include <sys/procfs.h>
 int
 main ()
@@ -12142,6 +12156,8 @@
 /* end confdefs.h.  */
 
 #define _SYSCALL32
+/* Needed for new procfs interface on sparc-solaris.  */
+#define _STRUCTURED_PROC 1
 #include <sys/procfs.h>
 int
 main ()
@@ -12177,6 +12193,8 @@
 /* end confdefs.h.  */
 
 #define _SYSCALL32
+/* Needed for new procfs interface on sparc-solaris.  */
+#define _STRUCTURED_PROC 1
 #include <sys/procfs.h>
 int
 main ()
@@ -12212,6 +12230,8 @@
 /* end confdefs.h.  */
 
 #define _SYSCALL32
+/* Needed for new procfs interface on sparc-solaris.  */
+#define _STRUCTURED_PROC 1
 #include <sys/procfs.h>
 int
 main ()
@@ -12247,6 +12267,8 @@
 /* end confdefs.h.  */
 
 #define _SYSCALL32
+/* Needed for new procfs interface on sparc-solaris.  */
+#define _STRUCTURED_PROC 1
 #include <sys/procfs.h>
 int
 main ()
@@ -12282,6 +12304,8 @@
 /* end confdefs.h.  */
 
 #define _SYSCALL32
+/* Needed for new procfs interface on sparc-solaris.  */
+#define _STRUCTURED_PROC 1
 #include <sys/procfs.h>
 int
 main ()
@@ -12317,6 +12341,8 @@
 /* end confdefs.h.  */
 
 #define _SYSCALL32
+/* Needed for new procfs interface on sparc-solaris.  */
+#define _STRUCTURED_PROC 1
 #include <sys/procfs.h>
 int
 main ()
@@ -12352,6 +12378,8 @@
 /* end confdefs.h.  */
 
 #define _SYSCALL32
+/* Needed for new procfs interface on sparc-solaris.  */
+#define _STRUCTURED_PROC 1
 #include <sys/procfs.h>
 int
 main ()

Re: [PATCH] gcore fix for native gdb on solaris

[Joel Brobecker]

> I was not somehow sure whether it is expected and I also wanted to make some
> heads-up whether _STRUCTURED_PROC #defines aren't now obsoleted in proc*.c
> files (I do not know if they are or not).

It isn't, as far as I know. I thought I had seen on Solaris 10
that _STRUCTURED_PROC was always assumed, but upon double-checking,
it appears not so.  In any case, the only risk is with old versions
of Solaris where _STRUCTURED_PROC is ignored - we might have a build
failure because we expected to have elfcore_write_lwpstatus, whereas
bfd only provided elfcore_write_lpstatus... I doubt it; my guess is
NEW_PROC_API would not be defined on these systems.

-- 
Joel

Re: [PATCH] gcore fix for native gdb on solaris

[Daniel Gutson]

Hi Joel,

Joel Brobecker wrote:
>> I was not somehow sure whether it is expected and I also wanted to make some
>> heads-up whether _STRUCTURED_PROC #defines aren't now obsoleted in proc*.c
>> files (I do not know if they are or not).
> 
> It isn't, as far as I know. I thought I had seen on Solaris 10
> that _STRUCTURED_PROC was always assumed, but upon double-checking,
> it appears not so.  In any case, the only risk is with old versions
> of Solaris where _STRUCTURED_PROC is ignored - we might have a build
> failure because we expected to have elfcore_write_lwpstatus, whereas
> bfd only provided elfcore_write_lpstatus... I doubt it; my guess is
> NEW_PROC_API would not be defined on these systems.
> 

I'm becoming up to date with this thread.
Is there anything pending from me to do, any commit, any validation?
FWIW, I already checked it for sparc-solaris.

Thanks!
	Daniel.

-- 
Daniel Gutson
CodeSourcery
www.codesourcery.com

Re: [PATCH] gcore fix for native gdb on solaris

[Jan Kratochvil]

On Tue, 12 Jan 2010 22:44:28 +0100, Daniel Gutson wrote:
> Is there anything pending from me to do, any commit, any validation?
> FWIW, I already checked it for sparc-solaris.

Just I see you agree with this change so everything is OK.

On Tue, 12 Jan 2010 14:23:35 +0100, Joel Brobecker wrote:
> my guess is NEW_PROC_API would not be defined on these systems.

Joel Brobecker already examined the issue I had suspection about but I do not
have experience with these systems at all.


Thanks,
Jan

Re: [PATCH] gcore fix for native gdb on solaris

[Daniel Gutson]

Jan Kratochvil wrote:
> On Tue, 12 Jan 2010 22:44:28 +0100, Daniel Gutson wrote:
>> Is there anything pending from me to do, any commit, any validation?
>> FWIW, I already checked it for sparc-solaris.
> 
> Just I see you agree with this change so everything is OK.

I think it's OK too.

-- 
Daniel Gutson
CodeSourcery
www.codesourcery.com