patches/gdb/6.8a/130-reg-no-longer-active.patch
author Michael Hope <michael.hope@linaro.org>
Wed Oct 19 15:27:32 2011 +1300 (2011-10-19)
changeset 2739 f320e22f2cba
parent 1343 e6f6153d318c
permissions -rw-r--r--
arch: add softfp support

Some architectures support a mixed hard/soft floating point, where
the compiler emits hardware floating point instructions, but passes
the operands in core (aka integer) registers.

For example, ARM supports this mode (to come in the next changeset).

Add support for softfp cross compilers to the GCC and GLIBC
configuration. Needed for Ubuntu and other distros that are softfp.

Signed-off-by: Michael Hope <michael.hope@linaro.org>
[yann.morin.1998@anciens.enib.fr: split the original patch]
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@anciens.enib.fr>
yann@1343
     1
Fix bug reported by Thomas Petazzoni <thomas.petazzoni@free-electrons.com>:
yann@1343
     2
See: http://sourceware.org/ml/crossgcc/2009-05/msg00055.html
yann@1343
     3
     https://bugzilla.redhat.com/show_bug.cgi?id=436037
yann@1343
     4
yann@1343
     5
Fix from: http://cvs.fedoraproject.org/viewvc/devel/gdb/gdb-6.8-bz436037-reg-no-longer-active.patch?revision=1.1
yann@1343
     6
yann@1343
     7
diff -d -urpN src.0/gdb/valops.c src.1/gdb/valops.c
yann@1343
     8
--- src.0/gdb/valops.c	2008-07-27 04:00:03.000000000 +0200
yann@1343
     9
+++ src.1/gdb/valops.c	2008-07-31 15:17:42.000000000 +0200
yann@1343
    10
@@ -813,10 +813,18 @@ value_assign (struct value *toval, struc
yann@1343
    11
 	struct frame_info *frame;
yann@1343
    12
 	int value_reg;
yann@1343
    13
 
yann@1343
    14
-	/* Figure out which frame this is in currently.  */
yann@1343
    15
-	frame = frame_find_by_id (VALUE_FRAME_ID (toval));
yann@1343
    16
 	value_reg = VALUE_REGNUM (toval);
yann@1343
    17
 
yann@1343
    18
+	/* Figure out which frame this is in currently.  */
yann@1343
    19
+	frame = frame_find_by_id (VALUE_FRAME_ID (toval));
yann@1343
    20
+	/* "set $reg+=1" should work on programs with no debug info,
yann@1343
    21
+	   but frame_find_by_id returns NULL here (RH bug 436037).
yann@1343
    22
+	   Use current frame, it represents CPU state in this case.
yann@1343
    23
+	   If frame_find_by_id is changed to do it internally
yann@1343
    24
+	   (it is contemplated there), remove this.  */
yann@1343
    25
+	if (!frame)
yann@1343
    26
+	  frame = get_current_frame ();
yann@1343
    27
+	/* Probably never happens.  */
yann@1343
    28
 	if (!frame)
yann@1343
    29
 	  error (_("Value being assigned to is no longer active."));
yann@1343
    30