Mirror of the gdb-patches mailing list
 help / color / mirror / Atom feed
* Re: [patch/rfa] Handle unwinding through export stubs on hpux
@ 2004-05-17 18:46 John David Anglin
  0 siblings, 0 replies; 4+ messages in thread
From: John David Anglin @ 2004-05-17 18:46 UTC (permalink / raw)
  To: gdb-patches; +Cc: randolph, cagney

+  if (gdbarch_osabi (gdbarch) == GDB_OSABI_HPUX_ELF 
+      || gdbarch_osabi (gdbarch) == GDB_OSABI_HPUX_SOM)
+    {
+      /* HPUX uses export stubs in function calls; the export stub clobbers
+         the return value of the caller, and, later restores it from the
+        stack.  */
+      u = find_unwind_entry (frame_pc_unwind (next_frame));
+
+      if (u && u->stub_unwind.stub_type == EXPORT)
+       {
+          info->saved_regs[HPPA_PCOQ_HEAD_REGNUM].addr = info->base - 24;

I don't believe that the 64-bit HP-UX runtime uses export stubs which
clobber the return value of the caller.  This is only true for the SOM
runtime.

Dave
-- 
J. David Anglin                                  dave.anglin@nrc-cnrc.gc.ca
National Research Council of Canada              (613) 990-0752 (FAX: 952-6602)


^ permalink raw reply	[flat|nested] 4+ messages in thread
* [patch/rfa] Handle unwinding through export stubs on hpux
@ 2004-05-15 23:58 Randolph Chung
  2004-05-17 15:09 ` Andrew Cagney
  0 siblings, 1 reply; 4+ messages in thread
From: Randolph Chung @ 2004-05-15 23:58 UTC (permalink / raw)
  To: gdb-patches

This fixes a problem reported by Dave Anglin; when doing a backtrace
through a function that is called through an export stub, the unwinder
fails. The problem is that the stub unwinder assumes that the return
pointer doesn't change inside the stub, but the export stub changes it.

Possibly something also needs to be changed for an argument relocation
stub. Will investigate and fix that separately if needed.

tested on hppa2.0w-hp-hpux11.11.

ok to apply?

randolph


2004-05-15  Randolph Chung  <tausq@debian.org>

	* hppa-tdep.c (hppa_stub_frame_unwind_cache): For HPUX stub unwinding,
	check to see if we are unwinding an export stub.  Export stubs clobber
	the %rp in the call path and stores the original rp in a stack slot.

Index: hppa-tdep.c
===================================================================
RCS file: /cvs/src/src/gdb/hppa-tdep.c,v
retrieving revision 1.156
diff -u -p -u -r1.156 hppa-tdep.c
--- hppa-tdep.c	8 May 2004 03:59:34 -0000	1.156
+++ hppa-tdep.c	15 May 2004 20:54:57 -0000
@@ -2023,6 +2023,7 @@ hppa_stub_frame_unwind_cache (struct fra
 {
   struct gdbarch *gdbarch = get_frame_arch (next_frame);
   struct hppa_stub_unwind_cache *info;
+  struct unwind_table_entry *u;
 
   if (*this_cache)
     return *this_cache;
@@ -2031,8 +2032,19 @@ hppa_stub_frame_unwind_cache (struct fra
   *this_cache = info;
   info->saved_regs = trad_frame_alloc_saved_regs (next_frame);
 
-  info->saved_regs[HPPA_PCOQ_HEAD_REGNUM].realreg = HPPA_RP_REGNUM;
-  info->base = frame_unwind_register_unsigned (next_frame, HPPA_SP_REGNUM);
+  if ((gdbarch_osabi (gdbarch) == GDB_OSABI_HPUX_ELF 
+       || gdbarch_osabi (gdbarch) == GDB_OSABI_HPUX_SOM)
+      && (u = find_unwind_entry (frame_pc_unwind (next_frame)))
+      && (u->stub_unwind.stub_type == EXPORT))
+    {
+      info->base = frame_unwind_register_unsigned (next_frame, HPPA_SP_REGNUM);
+      info->saved_regs[HPPA_PCOQ_HEAD_REGNUM].addr = info->base - 24;
+    }
+  else
+    {
+      info->base = frame_unwind_register_unsigned (next_frame, HPPA_SP_REGNUM);
+      info->saved_regs[HPPA_PCOQ_HEAD_REGNUM].realreg = HPPA_RP_REGNUM;
+    }
 
   return info;
 }

-- 
Randolph Chung
Debian GNU/Linux Developer, hppa/ia64 ports
http://www.tausq.org/


^ permalink raw reply	[flat|nested] 4+ messages in thread

end of thread, other threads:[~2004-05-17 18:46 UTC | newest]

Thread overview: 4+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2004-05-17 18:46 [patch/rfa] Handle unwinding through export stubs on hpux John David Anglin
  -- strict thread matches above, loose matches on Subject: below --
2004-05-15 23:58 Randolph Chung
2004-05-17 15:09 ` Andrew Cagney
2004-05-17 16:47   ` Randolph Chung

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