summaryrefslogtreecommitdiffstats
path: root/config/gnu-warnings/noerror-5
diff options
context:
space:
mode:
authorAllen Byrne <50328838+byrnHDF@users.noreply.github.com>2020-11-25 17:50:21 (GMT)
committerGitHub <noreply@github.com>2020-11-25 17:50:21 (GMT)
commit0b54c556a02f67a611638c47d5f92364da8eadd2 (patch)
tree2ba1eef16dd30a98b11ffa753fd8df28c3c469ca /config/gnu-warnings/noerror-5
parentffcb3542377c62cf972ff683c101e7268332f8a0 (diff)
downloadhdf5-0b54c556a02f67a611638c47d5f92364da8eadd2.zip
hdf5-0b54c556a02f67a611638c47d5f92364da8eadd2.tar.gz
hdf5-0b54c556a02f67a611638c47d5f92364da8eadd2.tar.bz2
Add else choice to error flags option (#128)
Diffstat (limited to 'config/gnu-warnings/noerror-5')
-rw-r--r--config/gnu-warnings/noerror-512
1 files changed, 12 insertions, 0 deletions
diff --git a/config/gnu-warnings/noerror-5 b/config/gnu-warnings/noerror-5
new file mode 100644
index 0000000..a7d40dd
--- /dev/null
+++ b/config/gnu-warnings/noerror-5
@@ -0,0 +1,12 @@
+-Wincompatible-pointer-types
+#
+# In GCC 4.4.7, the compiler gripes about shadowed global
+# declarations when a local variable uses the name of a
+# function that's in a system header file. For some reason,
+# later versions of GCC (e.g., 5.2.0) don't complain about
+# the shadowed globals. Maybe later versions are less fussy?
+# Anyway, the shadowing seems to be harmless, and GCC 4.4.7
+# is not a supported compiler, so let us promote shadowed globals
+# warnings to errors only for GCC 5 and later.
+#
+-Wshadow