False reports from rand_s

[derekbruening]

From timurrrr@google.com on July 22, 2010 08:14:06

On Win XP, r378 / r24 give false reports on the following code:

#define _CRT_RAND_S
#include <windows.h>
#include <assert.h>
#include <stdlib.h>
#include <stdio.h>

unsigned int Rand() {
unsigned int v = 1;
assert(rand_s(&v) == 0);
return v;
}

int main(void) {
printf("%d\n", Rand());
return 0;
}

gives the following report:

Error #1: UNINITIALIZED READ: reading 0x0012ff1c-0x0012ff20 4 byte(s)
@0:00:02.016 in thread 4772
0x00808315 <ADVAPI32.dll+0x8315> ADVAPI32.dll!SystemFunction036
??:0
0x008082f3 <ADVAPI32.dll+0x82f3> ADVAPI32.dll!SystemFunction036
??:0
0x008082b6 <ADVAPI32.dll+0x82b6> ADVAPI32.dll!SystemFunction036
??:0
0x00401a1f <test.exe+0x1a1f> test.exe!rand_s
f:\dd\vctools\crt_bld\self_x86\crt\src\rand_s.c:107
0x00401068 <test.exe+0x1068> test.exe!main
z:\dr-sandbox\issues\new\test.cpp:14
0x00401d88 <test.exe+0x1d88> test.exe!__tmainCRTStartup
f:\dd\vctools\crt_bld\self_x86\crt\src\crt0.c:327
0x7c817077 <KERNEL32.dll+0x17077> KERNEL32.dll!RegisterWaitForInputIdle
??:0

Error #2: UNINITIALIZED READ: reading 0x0012ff18-0x0012ff19 1 byte(s)
@0:00:02.016 in thread 4772
0x00808334 <ADVAPI32.dll+0x8334> ADVAPI32.dll!SystemFunction036
??:0
0x008082f3 <ADVAPI32.dll+0x82f3> ADVAPI32.dll!SystemFunction036
??:0
0x008082b6 <ADVAPI32.dll+0x82b6> ADVAPI32.dll!SystemFunction036
??:0
0x00401a1f <test.exe+0x1a1f> test.exe!rand_s
f:\dd\vctools\crt_bld\self_x86\crt\src\rand_s.c:107
0x00401068 <test.exe+0x1068> test.exe!main
z:\dr-sandbox\issues\new\test.cpp:14
0x00401d88 <test.exe+0x1d88> test.exe!__tmainCRTStartup
f:\dd\vctools\crt_bld\self_x86\crt\src\crt0.c:327
0x7c817077 <KERNEL32.dll+0x17077> KERNEL32.dll!RegisterWaitForInputIdle
??:0

These can be suppressed by the following suppression:
UNINITIALIZED READ
ADVAPI32.dll!SystemFunction036
ADVAPI32.dll!SystemFunction036
ADVAPI32.dll!SystemFunction036
*!rand_s

Original issue: http://code.google.com/p/drmemory/issues/detail?id=15

[derekbruening]

From timurrrr@google.com on August 04, 2010 07:20:27

Labels: Bug-FalsePositive

[derekbruening]

From timurrrr@google.com on June 21, 2011 02:52:09

On Win7 this gives UNADDRs instead (as of r343 ):
Error #1: UNADDRESSABLE ACCESS: writing 0x0018fb6c-0x0018fb70 4 byte(s)
@0:00:04.058 in thread 4000
0x74f71759 <CRYPTBASE.dll+0x1759> CRYPTBASE.dll!SystemFunction036
0x74f72b25 <CRYPTBASE.dll+0x2b25> CRYPTBASE.dll!SystemFunction036
0x74f74187 <CRYPTBASE.dll+0x4187> CRYPTBASE.dll!SystemFunction036
0x74f740d9 <CRYPTBASE.dll+0x40d9> CRYPTBASE.dll!SystemFunction036
0x74f7141d <CRYPTBASE.dll+0x141d> CRYPTBASE.dll!SystemFunction036
0x74f71376 <CRYPTBASE.dll+0x1376> CRYPTBASE.dll!SystemFunction036
0x74f71343 <CRYPTBASE.dll+0x1343> CRYPTBASE.dll!SystemFunction036
0x74f7131f <CRYPTBASE.dll+0x131f> CRYPTBASE.dll!SystemFunction036
0x74f71304 <CRYPTBASE.dll+0x1304> CRYPTBASE.dll!SystemFunction036
0x767d194b <ADVAPI32.dll+0x194b> ADVAPI32.dll!SystemFunction036
0x0040191c <test.exe+0x191c> test.exe!rand_s
f:\dd\vctools\crt_bld\self_x86\crt\src\rand_s.c:107
0x00401034 <test.exe+0x1034> test.exe!Rand
test.cpp:9
0x00401068 <test.exe+0x1068> test.exe!main
test.cpp:14

[derekbruening]

From timurrrr@google.com on June 21, 2011 03:28:56

The uninits are explained in issue #65 The Win 7 unaddrs are splitted as issue #462

Status: Duplicate
Mergedinto: 65