Add a couple of bogons people sent.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@37

diff --git a/tests/stpcpy.c b/tests/stpcpy.c
new file mode 100644 (file)
index 0000000..54464be
--- /dev/null
+++ b/tests/stpcpy.c
@@ -0,0 +1,51 @@
+
+#if 0
+/*
+Subject:  valgrind glibc suppression
+   Date:  Fri, 22 Mar 2002 23:54:44 -0500 (EST)
+   From:  Alex Larsson <alexl@redhat.com>
+     To:  jseward@acm.org
+
+
+Hi.
+
+I'm getting a lot of errors in __stpcpy(). I think this may be a bug in 
+glibc. I didn't analyze the stpcpy asm in detail, so it might still be a 
+valgrind bug, but it's probably a glibc bug.
+
+Here is a test case:
+*/
+#endif
+
+#include <string.h>
+#include <stdlib.h>
+
+int main()
+{
+  char *string;
+  char buffer[10];
+  
+  string = malloc (1);
+  string[0] = '\0';
+  
+  stpcpy (buffer, string);
+}
+
+#if 0
+/*
+Gives warnings like:
+==10941== Use of uninitialised CPU condition code
+==10941==    at 0x4034B9DA: __stpcpy (__stpcpy:36)
+==10941==    by 0x402DF627: __libc_start_main (../sysdeps/generic/libc-start.c:129)
+==10941==    by 0x80483D1: __libc_start_main@@GLIBC_2.0 (in /home/alex/other_src/valgrind-20020320/a.out)
+==10941==    by <bogus frame pointer> ???
+
+Here is the supression i use:
+{
+   __stpcpy(Value0)
+   Value0
+   fun:__stpcpy
+   fun:*
+}
+*/
+#endif

diff --git a/tests/trivialleak.c b/tests/trivialleak.c
new file mode 100644 (file)
index 0000000..002a48a
--- /dev/null
+++ b/tests/trivialleak.c
@@ -0,0 +1,12 @@
+static void test()
+  {
+    void* leak;
+    int i;
+    for (i = 0; i < 1000; i++)
+       leak = (void*)malloc( 100 );
+  }
+  int main()
+  {
+    test();
+    return 0;
+  }