Mirror of the gdb-patches mailing list
 help / color / mirror / Atom feed
* RFA: ARM stub unwinder
@ 2005-03-03 14:46 Daniel Jacobowitz
  2005-03-03 14:50 ` Richard Earnshaw
  0 siblings, 1 reply; 2+ messages in thread
From: Daniel Jacobowitz @ 2005-03-03 14:46 UTC (permalink / raw)
  To: rearnsha, gdb-patches

This patch adds a stub unwinder for ARM, and just like other architectures
with a stub unwinder, uses it for PLT entries and unreadable code addresses.
Richard tested this on arm-netbsd and it fixed a hundred failures; I tested
it on arm-linux some time ago, and it did the same there.

OK?

-- 
Daniel Jacobowitz
CodeSourcery, LLC

2005-03-02  Daniel Jacobowitz  <dan@codesourcery.com>

	* arm-tdep.c (arm_make_stub_cache, arm_stub_this_id)
	(arm_stub_unwind, arm_stub_unwind_sniffer): New.
	(arm_gdbarch_init): Add arm_stub_unwind_sniffer.

Index: arm-tdep.c
===================================================================
RCS file: /big/fsf/rsync/src-cvs/src/gdb/arm-tdep.c,v
retrieving revision 1.193
diff -u -p -r1.193 arm-tdep.c
--- arm-tdep.c	21 Feb 2005 07:15:47 -0000	1.193
+++ arm-tdep.c	2 Mar 2005 20:38:55 -0000
@@ -1013,6 +1013,56 @@ arm_prologue_unwind_sniffer (struct fram
   return &arm_prologue_unwind;
 }
 
+static struct arm_prologue_cache *
+arm_make_stub_cache (struct frame_info *next_frame)
+{
+  int reg;
+  struct arm_prologue_cache *cache;
+  CORE_ADDR unwound_fp;
+
+  cache = frame_obstack_zalloc (sizeof (struct arm_prologue_cache));
+  cache->saved_regs = trad_frame_alloc_saved_regs (next_frame);
+
+  cache->prev_sp = frame_unwind_register_unsigned (next_frame, ARM_SP_REGNUM);
+
+  return cache;
+}
+
+/* Our frame ID for a stub frame is the current SP and LR.  */
+
+static void
+arm_stub_this_id (struct frame_info *next_frame,
+		  void **this_cache,
+		  struct frame_id *this_id)
+{
+  struct arm_prologue_cache *cache;
+
+  if (*this_cache == NULL)
+    *this_cache = arm_make_stub_cache (next_frame);
+  cache = *this_cache;
+
+  *this_id = frame_id_build (cache->prev_sp,
+			     frame_pc_unwind (next_frame));
+}
+
+struct frame_unwind arm_stub_unwind = {
+  NORMAL_FRAME,
+  arm_stub_this_id,
+  arm_prologue_prev_register
+};
+
+static const struct frame_unwind *
+arm_stub_unwind_sniffer (struct frame_info *next_frame)
+{
+  char dummy[4];
+
+  if (in_plt_section (frame_unwind_address_in_block (next_frame), NULL)
+      || target_read_memory (frame_pc_unwind (next_frame), dummy, 4) != 0)
+    return &arm_stub_unwind;
+
+  return NULL;
+}
+
 static CORE_ADDR
 arm_normal_frame_base (struct frame_info *next_frame, void **this_cache)
 {
@@ -2737,6 +2787,7 @@ arm_gdbarch_init (struct gdbarch_info in
   gdbarch_init_osabi (info, gdbarch);
 
   /* Add some default predicates.  */
+  frame_unwind_append_sniffer (gdbarch, arm_stub_unwind_sniffer);
   frame_unwind_append_sniffer (gdbarch, arm_sigtramp_unwind_sniffer);
   frame_unwind_append_sniffer (gdbarch, arm_prologue_unwind_sniffer);
 


^ permalink raw reply	[flat|nested] 2+ messages in thread

* Re: RFA: ARM stub unwinder
  2005-03-03 14:46 RFA: ARM stub unwinder Daniel Jacobowitz
@ 2005-03-03 14:50 ` Richard Earnshaw
  0 siblings, 0 replies; 2+ messages in thread
From: Richard Earnshaw @ 2005-03-03 14:50 UTC (permalink / raw)
  To: Daniel Jacobowitz; +Cc: rearnsha, gdb-patches

On Thu, 2005-03-03 at 14:45, Daniel Jacobowitz wrote:
> This patch adds a stub unwinder for ARM, and just like other architectures
> with a stub unwinder, uses it for PLT entries and unreadable code addresses.
> Richard tested this on arm-netbsd and it fixed a hundred failures; I tested
> it on arm-linux some time ago, and it did the same there.
> 
> OK?

OK.

I forgot last night that it would be me approving this :-)

R.


^ permalink raw reply	[flat|nested] 2+ messages in thread

end of thread, other threads:[~2005-03-03 14:50 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2005-03-03 14:46 RFA: ARM stub unwinder Daniel Jacobowitz
2005-03-03 14:50 ` Richard Earnshaw

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