aboutsummaryrefslogtreecommitdiffstats
path: root/meta/recipes-devtools/gcc/files/gfortran.patch
diff options
context:
space:
mode:
authorRichard Purdie <rpurdie@linux.intel.com>2010-08-27 15:14:24 +0100
committerRichard Purdie <rpurdie@linux.intel.com>2010-08-27 15:29:45 +0100
commit29d6678fd546377459ef75cf54abeef5b969b5cf (patch)
tree8edd65790e37a00d01c3f203f773fe4b5012db18 /meta/recipes-devtools/gcc/files/gfortran.patch
parentda49de6885ee1bc424e70bc02f21f6ab920efb55 (diff)
downloadopenembedded-core-contrib-29d6678fd546377459ef75cf54abeef5b969b5cf.tar.gz
Major layout change to the packages directory
Having one monolithic packages directory makes it hard to find things and is generally overwhelming. This commit splits it into several logical sections roughly based on function, recipes.txt gives more information about the classifications used. The opportunity is also used to switch from "packages" to "recipes" as used in OpenEmbedded as the term "packages" can be confusing to people and has many different meanings. Not all recipes have been classified yet, this is just a first pass at separating things out. Some packages are moved to meta-extras as they're no longer actively used or maintained. Signed-off-by: Richard Purdie <rpurdie@linux.intel.com>
Diffstat (limited to 'meta/recipes-devtools/gcc/files/gfortran.patch')
-rw-r--r--meta/recipes-devtools/gcc/files/gfortran.patch40
1 files changed, 40 insertions, 0 deletions
diff --git a/meta/recipes-devtools/gcc/files/gfortran.patch b/meta/recipes-devtools/gcc/files/gfortran.patch
new file mode 100644
index 0000000000..96905e5d7d
--- /dev/null
+++ b/meta/recipes-devtools/gcc/files/gfortran.patch
@@ -0,0 +1,40 @@
+The patch below fixes a crash building libgfortran on arm-linux-gnueabi.
+
+This target doesn't really have a 128-bit integer type, however it does use
+TImode to represent the return value of certain special ABI defined library
+functions. This results in type_for_size(TImode) being called.
+
+Because TImode deosn't correspond to any gfortran integer kind
+gfc_type_for_size returns NULL and we segfault shortly after.
+
+The patch below fixes this by making gfc_type_for_size handle TImode in the
+same way as the C frontend.
+
+Tested on x86_64-linux and arm-linux-gnueabi.
+Applied to trunk.
+
+Paul
+
+2007-05-15 Paul Brook <paul@codesourcery.com>
+
+ gcc/fortran/
+ * trans-types.c (gfc_type_for_size): Handle signed TImode.
+
+Index: gcc-4.2.1/gcc/fortran/trans-types.c
+===================================================================
+--- gcc-4.2.1/gcc/fortran/trans-types.c (revision 170435)
++++ gcc-4.2.1/gcc/fortran/trans-types.c (working copy)
+@@ -1800,6 +1800,13 @@ gfc_type_for_size (unsigned bits, int un
+ if (type && bits == TYPE_PRECISION (type))
+ return type;
+ }
++
++ /* Handle TImode as a special case because it is used by some backends
++ (eg. ARM) even though it is not available for normal use. */
++#if HOST_BITS_PER_WIDE_INT >= 64
++ if (bits == TYPE_PRECISION (intTI_type_node))
++ return intTI_type_node;
++#endif
+ }
+ else
+ {