diff options
author | Eric Blake <ebb9@byu.net> | 2009-06-15 11:04:21 -0600 |
---|---|---|
committer | Eric Blake <ebb9@byu.net> | 2009-06-15 21:05:08 -0600 |
commit | dcc2f67b6ffab6e9def088ccbf7627edcda4bbac (patch) | |
tree | f2e6e7f96060b3b996611dbe2a6b27b365483eeb /lib | |
parent | e06226e32481672ab9321d5caddee63fe1dab720 (diff) | |
download | gnulib-dcc2f67b6ffab6e9def088ccbf7627edcda4bbac.tar.gz |
memchr, memchr2: add valgrind exception
* lib/memchr.valgrind: New file.
* lib/memchr2.valgrind: New file.
* modules/memchr (Files): Distribute valgrind file.
* modules/memchr2 (Files): Likewise.
Signed-off-by: Eric Blake <ebb9@byu.net>
Diffstat (limited to 'lib')
-rw-r--r-- | lib/memchr.valgrind | 14 | ||||
-rw-r--r-- | lib/memchr2.valgrind | 14 |
2 files changed, 28 insertions, 0 deletions
diff --git a/lib/memchr.valgrind b/lib/memchr.valgrind new file mode 100644 index 0000000000..60f247e10d --- /dev/null +++ b/lib/memchr.valgrind @@ -0,0 +1,14 @@ +# Suppress a valgrind message about use of uninitialized memory in memchr(). +# POSIX states that when the character is found, memchr must not read extra +# bytes in an overestimated length (for example, where memchr is used to +# implement strnlen). However, we use a safe word read to provide a speedup. +{ + memchr-value4 + Memcheck:Value4 + fun:rpl_memchr +} +{ + memchr-value8 + Memcheck:Value8 + fun:rpl_memchr +} diff --git a/lib/memchr2.valgrind b/lib/memchr2.valgrind new file mode 100644 index 0000000000..778fe867dc --- /dev/null +++ b/lib/memchr2.valgrind @@ -0,0 +1,14 @@ +# Suppress a valgrind message about use of uninitialized memory in memchr2(). +# Like memchr, it is safe to overestimate the length when the terminator +# is guaranteed to be found. In this case, we may end up reading a word +# that is partially uninitialized, but this use is OK for a speedup. +{ + memchr2-value4 + Memcheck:Value4 + fun:memchr2 +} +{ + memchr2-value8 + Memcheck:Value8 + fun:memchr2 +} |