add workshop codeql-dataflow-sql-injection

This commit is contained in:
2025-06-25 00:05:50 -07:00
committed by =Michael Hohn
parent ab7fba9007
commit 4f4ef1b84d
19 changed files with 1574 additions and 0 deletions

2
WHERE-FROM Normal file
View File

@@ -0,0 +1,2 @@
curl -L -O https://github.com/github/codeql-cli-binaries/releases/download/v2.22.0/codeql-osx64.zip
unzip -q assets/codeql-osx64.zip

BIN
assets/codeql-osx64.zip (Stored with Git LFS) Normal file

Binary file not shown.

View File

@@ -0,0 +1,229 @@
[[https://imgs.xkcd.com/comics/exploits_of_a_mom.png]]
(from https://xkcd.com/327/)
* SQL injection example
** Setup and sample run
#+BEGIN_SRC sh
# Use a simple headline prompt
PS1='
\033[32m---- SQL injection demo ----\[\033[33m\033[0m\]
$?:$ '
# Build
./build.sh
# Prepare db
./admin -r
./admin -c
./admin -s
# Add regular user interactively
./add-user 2>> users.log
First User
# Regular user via "external" process
echo "User Outside" | ./add-user 2>> users.log
# Check
./admin -s
# Add Johnny Droptable
./add-user 2>> users.log
Johnny'); DROP TABLE users; --
# And the problem:
./admin -s
# Check the log
tail users.log
#+END_SRC
** Identify the problem
=./add-user= is reading from =STDIN=, and writing to a database; looking at the code in
[[./add-user.c]] leads to
: count = read(STDIN_FILENO, buf, BUFSIZE - 1);
for the read and
: rc = sqlite3_exec(db, query, NULL, 0, &zErrMsg);
for the write.
This problem is thus a dataflow problem; in codeql terminology we have
- a /source/ at the =read(STDIN_FILENO, buf, BUFSIZE - 1);=
- a /sink/ at the =sqlite3_exec(db, query, NULL, 0, &zErrMsg);=
We write codeql to identify these two, and then connect them via
- a /dataflow configuration/ -- for this problem, the more general /taintflow
configuration/.
** Build codeql database
To get started, build the codeql database (adjust paths to your setup):
#+BEGIN_SRC sh
# Build the db with source commit id.
# export PATH=$HOME/local/vmsync/codeql250:"$PATH"
SRCDIR=$(pwd)
DB=$SRCDIR/cpp-sqli-$(cd $SRCDIR && git rev-parse --short HEAD)
echo $DB
test -d "$DB" && rm -fR "$DB"
mkdir -p "$DB"
cd $SRCDIR && codeql database create --language=cpp -s . -j 8 -v $DB --command='./build.sh'
#+END_SRC
Then add this database directory to your VS Code =DATABASES= tab.
** Build codeql database in steps
For larger projects, using a single command to build everything is costly when
any part of the build fails.
To build a database in steps, use the following sequence, adjusting paths to
your setup:
#+BEGIN_SRC sh
# Build the db with source commit id.
export PATH=$HOME/local/vmsync/codeql250:"$PATH"
SRCDIR=$HOME/local/codeql-training-material.cpp-sqli/cpp/codeql-dataflow-sql-injection
DB=$SRCDIR/cpp-sqli-$(cd $SRCDIR && git rev-parse --short HEAD)
# Check paths
echo $DB
echo $SRCDIR
# Prepare db directory
test -d "$DB" && rm -fR "$DB"
mkdir -p "$DB"
# Run the build
cd $SRCDIR
codeql database init --language=cpp -s . -v $DB
# Repeat trace-command as needed to cover all targets
codeql database trace-command -v $DB -- make
codeql database finalize -j4 $DB
#+END_SRC
Then add this database directory to your VS Code =DATABASES= tab.
** Develop the query bottom-up
1. Identify the /source/ part of the
: read(STDIN_FILENO, buf, BUFSIZE - 1);
expression, the =buf= argument.
Start from a =from..where..select=, then convert to a predicate.
2. Identify the /sink/ part of the
: sqlite3_exec(db, query, NULL, 0, &zErrMsg);
expression, the =query= argument. Again start from =from..where..select=,
then convert to a predicate.
3. Fill in the /taintflow configuration/ boilerplate
#+BEGIN_SRC java
class CppSqli extends TaintTracking::Configuration {
CppSqli() { this = "CppSqli" }
override predicate isSource(DataFlow::Node node) {
none()
}
override predicate isSink(DataFlow::Node node) {
none()
}
}
#+END_SRC
Note that an inout-argument in C/C++ (the =buf= pointer is passed to =read=
and points to updated data after the return) is accessed as a codeql source
via
: source.(DataFlow::PostUpdateNode).getPreUpdateNode().asExpr()
instead of the usual
: source.asExpr()
The final query (without =isAdditionalTaintStep=) is
#+BEGIN_SRC java
/**
,* @name SQLI Vulnerability
,* @description Using untrusted strings in a sql query allows sql injection attacks.
,* @kind path-problem
,* @id cpp/SQLIVulnerable
,* @problem.severity warning
,*/
import cpp
import semmle.code.cpp.dataflow.TaintTracking
import DataFlow::PathGraph
class SqliFlowConfig extends TaintTracking::Configuration {
SqliFlowConfig() { this = "SqliFlow" }
override predicate isSource(DataFlow::Node source) {
// count = read(STDIN_FILENO, buf, BUFSIZE);
exists(FunctionCall read |
read.getTarget().getName() = "read" and
read.getArgument(1) = source.(DataFlow::PostUpdateNode).getPreUpdateNode().asExpr()
)
}
override predicate isSink(DataFlow::Node sink) {
// rc = sqlite3_exec(db, query, NULL, 0, &zErrMsg);
exists(FunctionCall exec |
exec.getTarget().getName() = "sqlite3_exec" and
exec.getArgument(1) = sink.asExpr()
)
}
}
from SqliFlowConfig conf, DataFlow::PathNode source, DataFlow::PathNode sink
where conf.hasFlowPath(source, sink)
select sink, source, sink, "Possible SQL injection"
#+END_SRC
** Optional: sarif file review of the results
Query results are available in several output formats using the cli. The
following produces the sarif format, a json-based result description.
#+BEGIN_SRC sh
# The setup information from before
export PATH=$HOME/local/vmsync/codeql250:"$PATH"
SRCDIR=$HOME/local/codeql-training-material.cpp-sqli/cpp/codeql-dataflow-sql-injection
DB=$SRCDIR/cpp-sqli-$(cd $SRCDIR && git rev-parse --short HEAD)
# Check paths
echo $DB
echo $SRCDIR
# To see the help
codeql database analyze -h
# Run a query
codeql database analyze \
-v \
--ram=14000 \
-j12 \
--rerun \
--search-path ~/local/vmsync/ql \
--format=sarif-latest \
--output cpp-sqli.sarif \
-- \
$DB \
$SRCDIR/SqlInjection.ql
# Examine the file in an editor
edit cpp-sqli.sarif
#+END_SRC
An example of using the sarif data is in the the jq script [[./sarif-summary.jq]].
When run against the sarif input via
#+BEGIN_SRC sh
jq --raw-output --join-output -f sarif-summary.jq < cpp-sqli.sarif > cpp-sqli.txt
#+END_SRC
it produces output in a form close to that of compiler error messages:
#+BEGIN_SRC text
query-id: message line
Path
...
Path
...
#+END_SRC

View File

@@ -0,0 +1,46 @@
/**
* @name SQLI Vulnerability
* @description Using untrusted strings in a sql query allows sql injection attacks.
* @ kind path-problem
* @id cpp/sqlivulnerable
* @problem.severity warning
*/
import cpp
import semmle.code.cpp.dataflow.new.TaintTracking
module SqliFlowConfig implements DataFlow::ConfigSig {
predicate isSource(DataFlow::Node source) {
// count = read(STDIN_FILENO, buf, BUFSIZE);
exists(FunctionCall read |
read.getTarget().getName() = "read" and
(
read.getArgument(1) = source.asDefiningArgument()
or
read.getArgument(1) = source.asExpr()
)
)
}
predicate isBarrier(DataFlow::Node sanitizer) { none() }
predicate isSink(DataFlow::Node sink) {
// rc = sqlite3_exec(db, query, NULL, 0, &zErrMsg);
exists(FunctionCall exec |
exec.getTarget().getName() = "sqlite3_exec" and
exec.getArgument(1) = sink.asIndirectArgument()
)
}
}
module MyFlow = TaintTracking::Global<SqliFlowConfig>;
// import MyFlow::PathGraph
from DataFlow::Node thing
where SqliFlowConfig::isSource(thing)
select thing, thing.getAQlClass()
// from MyFlow::PathNode source, MyFlow::PathNode sink
// where MyFlow::flowPath(source, sink)
// select sink, source, sink, "Possible SQL injection"

View File

@@ -0,0 +1,100 @@
#include <stdio.h>
#include <stdlib.h>
#include <unistd.h>
#include <ctype.h>
#include <sqlite3.h>
#include <time.h>
void write_log(const char* fmt, ...) {
time_t t;
char tstr[26];
va_list args;
va_start(args, fmt);
t = time(NULL);
ctime_r(&t, tstr);
tstr[24] = 0; /* no \n */
fprintf(stderr, "[%s] ", tstr);
vfprintf(stderr, fmt, args);
va_end(args);
fflush(stderr);
}
void abort_on_error(int rc, sqlite3 *db) {
if( rc ) {
fprintf(stderr, "Can't open database: %s\n", sqlite3_errmsg(db));
sqlite3_close(db);
fflush(stderr);
abort();
}
}
void abort_on_exec_error(int rc, sqlite3 *db, char* zErrMsg) {
if( rc!=SQLITE_OK ){
fprintf(stderr, "SQL error: %s\n", zErrMsg);
sqlite3_free(zErrMsg);
sqlite3_close(db);
fflush(stderr);
abort();
}
}
char* get_user_info() {
#define BUFSIZE 1024
char* buf = (char*) malloc(BUFSIZE * sizeof(char));
if(buf==NULL) abort();
int count;
// Disable buffering to avoid need for fflush
// after printf().
setbuf( stdout, NULL );
printf("*** Welcome to sql injection ***\n");
printf("Please enter name: ");
count = read(STDIN_FILENO, buf, BUFSIZE - 1);
if (count <= 0) abort();
// ensure the buffer is zero-terminated
buf[count] = '\0';
/* strip trailing whitespace */
while (count && isspace(buf[count-1])) {
buf[count-1] = 0; --count;
}
return buf;
}
int get_new_id() {
int id = getpid();
return id;
}
void write_info(int id, char* info) {
sqlite3 *db;
int rc;
int bufsize = 1024;
char *zErrMsg = 0;
char query[bufsize];
/* open db */
rc = sqlite3_open("users.sqlite", &db);
abort_on_error(rc, db);
/* Format query */
snprintf(query, bufsize, "INSERT INTO users VALUES (%d, '%s')", id, info);
write_log("query: %s\n", query);
/* Write info */
rc = sqlite3_exec(db, query, NULL, 0, &zErrMsg);
abort_on_exec_error(rc, db, zErrMsg);
sqlite3_close(db);
}
int main(int argc, char* argv[]) {
char* info;
int id;
info = get_user_info();
id = get_new_id();
write_info(id, info);
free(info);
/*
* show_info(id);
*/
}

View File

@@ -0,0 +1,27 @@
#!/bin/bash
get-user-info () {
echo "*** Welcome to sql injection ***"
read -r -p "Please enter name: " NAME
}
get-new-id () {
ID=$(/bin/bash -c 'echo $$')
}
add-user-info () {
echo "
INSERT INTO users VALUES ($ID, '$NAME')
" | sqlite3 users.sqlite
}
show-user-info () {
echo "We have the following information for you:"
echo "
select * FROM users where user_id=$ID
" | sqlite3 users.sqlite
}
get-user-info
get-new-id
add-user-info
show-user-info

View File

@@ -0,0 +1,60 @@
#!/bin/bash
set -e
script=$(basename "$0")
GREEN='\033[0;32m'
MAGENTA='\033[0;95m'
NC='\033[0m'
RED='\033[0;31m'
YELLOW='\033[0;33m'
help() {
echo -e "Usage: ./${script} [options]" \
"\n${YELLOW}Options: ${NC}" \
"\n\t -h ${GREEN}Show Help ${NC}" \
"\n\t -c ${MAGENTA}Creates a users table ${NC}" \
"\n\t -s ${MAGENTA}Shows all records in the users table ${NC}" \
"\n\t -r ${RED}Removes users table ${NC}"
}
remove-db () {
rm users.sqlite
}
create-db () {
echo '
CREATE TABLE users (
user_id INTEGER not null,
name TEXT NOT NULL
);
' | sqlite3 users.sqlite
}
show-db () {
echo '
SELECT * FROM users;
' | sqlite3 users.sqlite
}
if [ $# == 0 ]; then
help
exit 0
fi
while getopts "h?csr" option
do
case "${option}"
in
h|\?)
help
exit 0
;;
c) create-db
;;
s) show-db
;;
r) remove-db
;;
esac
done

View File

@@ -0,0 +1,2 @@
#!/bin/bash
clang -Wall add-user.c -lsqlite3 -o add-user

View File

@@ -0,0 +1,12 @@
{
"folders": [
{
"path": "."
}
],
"settings": {
"codeQL.runningQueries.autoSave": true,
"makefile.configureOnOpen": false,
"codeQL.githubDatabase.download": "never"
}
}

View File

@@ -0,0 +1,983 @@
<!-- -*- coding: utf-8 -*- -->
<!-- https://gist.github.com/hohn/
-->
# CodeQL Tutorial for C/C++: Data Flow and SQL Injection
<!--
!-- xx:
!-- md_toc github < codeql-dataflow-sql-injection.md
-->
- [CodeQL Tutorial for C/C++: Data Flow and SQL Injection](#codeql-tutorial-for-cc-data-flow-and-sql-injection)
- [Setup Instructions](#setup-instructions)
- [Documentation Links](#documentation-links)
- [Codeql Recap](#codeql-recap)
- [from, where, select](#from-where-select)
- [Predicates](#predicates)
- [Existential quantifiers (local variables in queries)](#existential-quantifiers-local-variables-in-queries)
- [Classes](#classes)
- [The Problem in Action](#the-problem-in-action)
- [Problem Statement](#problem-statement)
- [Data flow overview and illustration](#data-flow-overview-and-illustration)
- [Tutorial: Sources, Sinks and Flow Steps](#tutorial-sources-sinks-and-flow-steps)
- [The Data Sink](#the-data-sink)
- [The Data Source](#the-data-source)
- [The Extra Flow Step](#the-extra-flow-step)
- [The CodeQL Taint Flow Configuration](#the-codeql-taint-flow-configuration)
- [Taint Flow Configuration](#taint-flow-configuration)
- [Path Problem Setup](#path-problem-setup)
- [Path Problem Query Format](#path-problem-query-format)
- [Tutorial: Taint Flow Details](#tutorial-taint-flow-details)
- [The isSink Predicate](#the-issink-predicate)
- [The isSource Predicate](#the-issource-predicate)
- [The isAdditionalTaintStep Predicate](#the-isadditionaltaintstep-predicate)
- [Appendix](#appendix)
- [The complete Query: SqlInjection.ql](#the-complete-query-sqlinjectionql)
- [The Database Writer: add-user.c](#the-database-writer-add-userc)
## Setup Instructions
To run CodeQL queries on dotnet/coreclr, follow these steps:
1. Install the Visual Studio Code IDE.
2. Download and install the [CodeQL extension for Visual Studio Code](https://help.semmle.com/codeql/codeql-for-vscode.html). Full setup instructions are [here](https://help.semmle.com/codeql/codeql-for-vscode/procedures/setting-up.html).
3. [Set up the starter workspace](https://help.semmle.com/codeql/codeql-for-vscode/procedures/setting-up.html#using-the-starter-workspace).
- **Important**: Don't forget to `git clone --recursive` or `git submodule update --init --remote`, so that you obtain the standard query libraries.
4. Open the starter workspace: File > Open Workspace > Browse to `vscode-codeql-starter/vscode-codeql-starter.code-workspace`.
5. Download the sample database [`codeql-dataflow-sql-injection-d5b28fb.zip`](https://drive.google.com/file/d/1eBZ69ZQx6YnnZu41iUL0m8_e9qyMCZ9B/view?usp=sharing)
6. Unzip the database.
7. Import the unzipped database into Visual Studio Code:
- Click the **CodeQL** icon in the left sidebar.
- Place your mouse over **Databases**, and click the + sign that appears on
the right.
- Choose the unzipped database directory on your filesystem.
8. Create a new file, name it `SqliInjection.ql`, save it under `codeql-custom-queries-cpp`.
## Documentation Links
If you get stuck, try searching our documentation and blog posts for help and ideas. Below are a few links to help you get started:
- [Learning CodeQL](https://help.semmle.com/QL/learn-ql)
- [Learning CodeQL for C/C++](https://help.semmle.com/QL/learn-ql/cpp/ql-for-cpp.html)
- [Using the CodeQL extension for VS Code](https://help.semmle.com/codeql/codeql-for-vscode.html)
## Codeql Recap
This is a brief review of CodeQL taken from the [full
introduction](https://git.io/JJqdS). For more details, see the [documentation
links](#documentation-links). We will revisit all of this during the tutorial.
### from, where, select
Recall that codeql is a declarative language and a basic query is defined by a
_select_ clause, which specifies what the result of the query should be. For
example:
```ql
import cpp
select "hello world"
```
More complicated queries look like this:
```ql
from /* ... variable declarations ... */
where /* ... logical formulas ... */
select /* ... expressions ... */
```
The `from` clause specifies some variables that will be used in the query. The
`where` clause specifies some conditions on those variables in the form of logical
formulas. The `select` clauses specifies what the results should be, and can refer
to variables defined in the `from` clause.
The `from` clause is defined as a series of variable declarations, where each
declaration has a _type_ and a _name_. For example:
```ql
from IfStmt ifStmt
select ifStmt
```
We are declaring a variable with the name `ifStmt` and the type `IfStmt` (from the
CodeQL standard library for analyzing C/C++). Variables represent a **set of
values**, initially constrained by the type of the variable. Here, the variable
`ifStmt` represents the set of all `if` statements in the C/C++ program, as we can
see if we run the query.
A query using all three clauses to find empty blocks:
```ql
from IfStmt ifStmt, Block block
where
ifStmt.getThen() = block and
block.getNumStmt() = 0
select ifStmt, "Empty if statement"
```
### Predicates
The other feature we will use are _predicates_. These provide a way to encapsulate
portions of logic in the program so that they can be reused. You can think of
them as a mini `from`-`where`-`select` query clause. Like a select clause they
also produce a set of "tuples" or rows in a result table.
We can introduce a new predicate in our query that identifies the set of empty
blocks in the program (for example, to reuse this feature in another query):
```ql
predicate isEmptyBlock(Block block) {
block.getNumStmt() = 0
}
from IfStmt ifStmt
where isEmptyBlock(ifStmt.getThen())
select ifStmt, "Empty if statement"
```
### Existential quantifiers (local variables in queries)
Although the terminology may sound scary if you are not familiar with logic and
logic programming, *existential quantifiers* are simply ways to introduce
temporary variables with some associated conditions. The syntax for them is:
```ql
exists(<variable declarations> | <formula>)
```
They have a similar structure to the `from` and `where` clauses, where the first
part allows you to declare one or more variables, and the second formula
("conditions") that can be applied to those variables.
For example, we can use this to refactor the query
```ql
from IfStmt ifStmt, Block block
where
ifStmt.getThen() = block and
block.getNumStmt() = 0
select ifStmt, "Empty if statement"
```
to use a temporary variable for the empty block:
```ql
from IfStmt ifStmt
where
exists(Block block |
ifStmt.getThen() = block and
block.getNumStmt() = 0
)
select ifStmt, "Empty if statement"
```
This is frequently used to convert a query into a predicate.
### Classes
Classes are a way in which you can define new types within CodeQL, as well as
providing an easy way to reuse and structure code.
Like all types in CodeQL, classes represent a set of values. For example, the
`Block` type is, in fact, a class, and it represents the set of all blocks in the
program. You can also think of a class as defining a set of logical conditions
that specifies the set of values for that class.
For example, we can define a new CodeQL class to represent empty blocks:
```ql
class EmptyBlock extends Block {
EmptyBlock() {
this.getNumStmt() = 0
}
}
```
and use it in a query:
```ql
from IfStmt ifStmt, EmptyBlock block
where ifStmt.getThen() = block
select ifStmt, "Empty if statement"
```
## The Problem in Action
Running the code is a great way to see the problem and check whether the code is
vulnerable.
This program can be compiled and linked, and a simple sqlite db created via
```sh
# Build
./build.sh
# Prepare db
./admin -r
./admin -c
./admin -s
```
Users can be added via `stdin` in several ways; the second is a pretend "server"
using the `echo` command.
```sh
# Add regular user interactively
./add-user 2>> users.log
First User
# Regular user via "external" process
echo "User Outside" | ./add-user 2>> users.log
```
Check the db and log:
```
# Check
./admin -s
tail -4 users.log
```
Looks ok:
```
0:$ ./admin -s
87797|First User
87808|User Outside
0:$ tail -4 users.log
[Tue Jul 21 14:15:46 2020] query: INSERT INTO users VALUES (87797, 'First User')
[Tue Jul 21 14:17:07 2020] query: INSERT INTO users VALUES (87808, 'User Outside')
```
But there may be bad input; this one guesses the table name and drops it:
```sh
# Add Johnny Droptable
./add-user 2>> users.log
Johnny'); DROP TABLE users; --
```
And then we have this:
```sh
# And the problem:
./admin -s
0:$ ./admin -s
Error: near line 2: no such table: users
```
What happened? The log shows that data was treated as command:
```
1:$ tail -4 users.log
[Tue Jul 21 14:15:46 2020] query: INSERT INTO users VALUES (87797, 'First User')
[Tue Jul 21 14:17:07 2020] query: INSERT INTO users VALUES (87808, 'User Outside')
[Tue Jul 21 14:18:25 2020] query: INSERT INTO users VALUES (87817, 'Johnny'); DROP TABLE users; --')
```
Looking ahead, we now *know* that there is unsafe external data (source)
which reaches (flow path) a database-writing command (sink). Thus, a query
written against this code should find at least one taint flow path.
## Problem Statement
Many security problems can be phrased in terms of _information flow_:
_Given a (problem-specific) set of sources and sinks, is there a path in the data
flow graph from some source to some sink?_
The example we look at is SQL injection: sources are user-input, sinks are SQL
queries processing a string formed at runtime.
When parts of the string can be specified by the user, they allow an attacker to
insert arbitrary sql statements; these could erase a table or extract internal
data etc.
We will use CodeQL to analyze the source code constructing a SQL
query using string concatenation and then executing that query
string. The following example uses the `sqlite3` library; it
- receives user-provided data from `stdin` and keeps it in `buf`
- uses environment data and stores it in `id`,
- runs a query in `sqlite3_exec`
This is intentionally simple code, but it has all the elements that have to be
considered in real code and illustrates the QL features.
```c
#include <stdio.h>
#include <stdlib.h>
#include <unistd.h>
#include <ctype.h>
#include <sqlite3.h>
#include <time.h>
void write_log(const char* fmt, ...);
void abort_on_error(int rc, sqlite3 *db);
void abort_on_exec_error(int rc, sqlite3 *db, char* zErrMsg);
char* get_user_info() {
#define BUFSIZE 1024
char* buf = (char*) malloc(BUFSIZE * sizeof(char));
int count;
// Disable buffering to avoid need for fflush
// after printf().
setbuf( stdout, NULL );
printf("*** Welcome to sql injection ***\n");
printf("Please enter name: ");
count = read(STDIN_FILENO, buf, BUFSIZE);
if (count <= 0) abort();
/* strip trailing whitespace */
while (count && isspace(buf[count-1])) {
buf[count-1] = 0; --count;
}
return buf;
}
int get_new_id() {
int id = getpid();
return id;
}
void write_info(int id, char* info) {
sqlite3 *db;
int rc;
int bufsize = 1024;
char *zErrMsg = 0;
char query[bufsize];
/* open db */
rc = sqlite3_open("users.sqlite", &db);
abort_on_error(rc, db);
/* Format query */
snprintf(query, bufsize, "INSERT INTO users VALUES (%d, '%s')", id, info);
write_log("query: %s\n", query);
/* Write info */
rc = sqlite3_exec(db, query, NULL, 0, &zErrMsg);
abort_on_exec_error(rc, db, zErrMsg);
sqlite3_close(db);
}
int main(int argc, char* argv[]) {
char* info;
int id;
info = get_user_info();
id = get_new_id();
write_info(id, info);
/*
* show_info(id);
*/
}
```
In terms of sources, sinks, and information flow, the concrete problem for codeql is:
1. specifying `buf` as **source**,
2. specifying the `query` argument to `sqlite3_exec()` as **sink**,
3. specifying some code-specific data flow steps for the codeql library,
3. using the codeql taint flow library find taint flow paths (if there are any)
between the source and the sink.
In the following, we go into more concrete detail and develop codedql scripts to
solve this problem.
## Data flow overview and illustration
In the previous sections we identified the sources of problematic strings
(accesses of `info` etc.), and the sink that their data may flow to (the argument
to `sqlite3_exec`).
We need to see if there is data flow between the source(s) and this sink.
The solution here is to use the data flow library. Data flow is, as the name
suggests, about tracking the flow of data through the program. It helps answers
questions like: does this expression ever hold a value that originates from a
particular other place in the program?
We can visualize the data flow problem as one of finding paths through a directed
graph, where the nodes of the graph are elements in program, and the edges
represent the flow of data between those elements. If a path exists, then the data
flows between those two nodes.
This graph represents the flow of data from the tainted parameter. The nodes of
graph represent program elements that have a value, such as function parameters
and expressions. The edges of this graph represent flow through these nodes.
There are two variants of data flow available in CodeQL:
- Local (“intra-procedural”) data flow models flow within one function; feasible
to compute for all functions in a CodeQL database.
- Global (“inter-procedural”) data flow models flow across function calls; not
feasible to compute for all functions in a CodeQL database.
While local data flow is feasible to compute for all functions in a CodeQL
database, global data flow is not. This is because the number of paths becomes
_exponentially_ larger for global data flow.
The global data flow (and taint tracking) library avoids this problem by requiring
that the query author specifies which _sources_ and _sinks_ are applicable. This
allows the implementation to compute paths only between the restricted set of
nodes, rather than for the full graph.
To illustrate the dataflow for this problem, we have a [collection of slides](https://drive.google.com/file/d/1eEG0eGVDVEQh0C-0_4UIMcD23AWwnGtV/view?usp=sharing)
for this workshop.
## Tutorial: Sources, Sinks and Flow Steps
<!--
XX:
!-- The complete project can be downloaded via this
!-- [drive](https://drive.google.com/file/d/1-6c3S-e4FKa_IsuuzhhXupiAwCzzPgD-/view?usp=sharing)
!-- link.
-->
The tutorial is split into several steps and introduces concepts as they are
needed. Experimentation with the presented queries is encouraged, and the
autocomplete suggestions (Ctrl + Space) and the jump-to-definition command (F12 in
VS Code) are good ways explore the libraries.
### The Data Sink
Now let's find the function `sqlite3_exec`. In CodeQL, this uses `Function`
and a `getName()` attribute.
```ql
from Function f
where f.getName() = "sqlite3_exec"
select f
```
This should find one result,
```ql
SQLITE_API int sqlite3_exec(
sqlite3*, /* An open database */
const char *sql, /* SQL to be evaluated */
int (*callback)(void*,int,char**,char**), /* Callback function */
void *, /* 1st argument to callback */
char **errmsg /* Error msg written here */
);
```
in the header `sqlite3.h`.
Next, let's find the calls to `sqlite3_exec` using the `FunctionCall` type
```ql
from FunctionCall exec
where exec.getTarget().getName() = "sqlite3_exec"
select exec
```
This finds our call in `add-user.c`,
rc = sqlite3_exec(db, query, NULL, 0, &zErrMsg);
We are interested in the `query` argument, which we can get using `.getArgument`:
```ql
from FunctionCall exec, Expr query
where
exec.getTarget().getName() = "sqlite3_exec" and
query = exec.getArgument(1)
select exec, query
```
### The Data Source
The external data enters through the call
count = read(STDIN_FILENO, buf, BUFSIZE);
We thus want the `buf` argument to the call of the `read` function. Together, this is
```ql
from FunctionCall read, Expr buf
where
read.getTarget().getName() = "read" and
buf = read.getArgument(1)
select read, buf
```
### The Extra Flow Step
The codeql data flow library traverses *visible* source code fairly well, but flow
through opaque functions requires additional support (more on this later).
Functions for which only a headers is available are opaque, and we have one of
these here: the call to `snprintf`. Once we locate this call, there are *two* nodes
to identify: the inflow and outflow.
Let's start with `snprintf`. If we try
```ql
from FunctionCall printf
where printf.getTarget().getName() = "snprintf"
select printf
```
we get zero results. This is puzzling; if we visit the `add-user.c` source and
follow the definition of `snprintf`, it turns out to be a macro on MacOS:
```c
#undef snprintf
#define snprintf(str, len, ...) \
__builtin___snprintf_chk (str, len, 0, __darwin_obsz(str), __VA_ARGS__)
#endif
```
Fortunately, the underlying function `__builtin___snprintf_chk` has `snprintf` in
the name. So instead of working with C macros from codeql, we generalize our
query using a name pattern with `.matches`:
```ql
from FunctionCall printf
where printf.getTarget().getName().matches("%snprintf%")
select printf
```
This identifies our call
snprintf(query, bufsize, "INSERT INTO users VALUES (%d, '%s')", id, info);
and we need the inflow and outflow nodes next. `query` is the outflow, `info` is
the inflow.
In the `snprintf` macro call, those have indices 0 and 4. In the underlying function
`__builtin___snprintf_chk`, the indices are 0 and 6. Using the latter:
```ql
from FunctionCall printf, Expr out, Expr into
where
printf.getTarget().getName().matches("%snprintf%") and
printf.getArgument(0) = out and
printf.getArgument(6) = into
select printf, out, into
```
This correctly identifies the call and the extra flow arguments.
<!-- !-- Practice exercise: !-- Very specific: shifted index for macro.
Generalize this to consider !-- all trailing arguments as sources. -->
Practice exercise: If you are using linux or windows, generalize this query for
the `snprintf` arguments found there. One way to do this is using `or`:
```ql
printf.getTarget().getName().matches("%snprintf%") and
(
// mac version
or
// linux version
or
// windows version
)
```
## The CodeQL Taint Flow Configuration
The previous queries identify our source, sink and one additional flow step. To
use global data flow and taint tracking we need some additional codeql setup:
- a taint flow configuration
- the path problem header and imports
- a query formatted for path problems.
These are done next.
### Taint Flow Configuration
The way we configure global taint flow is by creating a custom extension of the
`TaintTracking::Configuration` class, and speciyfing `isSource`, `isSink`, and
`isAdditionalTaintStep` predicates.
The sources and sinks were explained earlier. Data flow and taint tracking
configuration classes support a number of additional features that help configure
the process of building and exploring the data flow path.
One such feature is adding additional taint steps. This is useful if you use
libraries which are not modelled by the default taint tracking. You can implement
this by overriding `isAdditionalTaintStep` predicate. This has two parameters, the
`from` and the `to` node, and it essentially allows you to add extra edges into the
taint tracking or data flow graph.
A starting configuration can look like the following, with details to be filled
in.
```ql
class SqliFlowConfig extends TaintTracking::Configuration {
SqliFlowConfig() { this = "SqliFlow" }
override predicate isSource(DataFlow::Node source) {
// count = read(STDIN_FILENO, buf, BUFSIZE);
}
override predicate isSanitizer(DataFlow::Node sanitizer) { none() }
override predicate isAdditionalTaintStep(DataFlow::Node into, DataFlow::Node out) {
// Extra taint step for
// snprintf(query, bufsize, "INSERT INTO users VALUES (%d, '%s')", id, info);
}
override predicate isSink(DataFlow::Node sink) {
// rc = sqlite3_exec(db, query, NULL, 0, &zErrMsg);
}
}
```
`TaintTracking::Configuration` is a _configuration_ class. In this case, there will be
a single instance of the class, identified by a unique string specified in the
characteristic predicate. We then override the `isSource` predicates to represent
the set of possible sources in the program, and `isSink` to represent the possible
set of sinks in the program.
### Path Problem Setup
Queries will only list sources and sinks by default. To inspect these results and
work with them, we also need the data paths from source to sink. For this, the
query needs to have the form of a _path problem_ query.
This requires a modifications to the query header and an extra import:
- The `@kind` comment has to be `path-problem`. This tells the CodeQL toolchain
to interpret the results of this query as path results.
- A new import `DataFlow::PathGraph`, which will report the path data
alongside the query results.
Together, this looks like
```ql
/**
* @name SQLI Vulnerability
* @description Using untrusted strings in a sql query allows sql injection attacks.
* @kind path-problem
* @id cpp/SQLIVulnerable
* @problem.severity warning
*/
import cpp
import semmle.code.cpp.dataflow.TaintTracking
import DataFlow::PathGraph
```
### Path Problem Query Format
To use this new configuration and `PathGraph` support, we call the
`hasFlowPath(source, sink)` predicate, which will compute a reachability table
between the defined sources and sinks. Behind the scenes, you can think of this as
performing a graph search algorithm from sources to sinks. The query will look
like this:
```ql
from SqliFlowConfig conf, DataFlow::PathNode source, DataFlow::PathNode sink
where conf.hasFlowPath(source, sink)
select sink, source, sink, "Possible SQL injection"
```
## Tutorial: Taint Flow Details
With the dataflow configuration in place, we just need to provide the details for
source(s), sink(s), and taint step(s).
Some more steps are required to convert our previous queries for use in data
flow. These are covered here.
### The isSink Predicate
Note that our previous queries used `Expr` nodes, but the taint query requires
`DataFlow::Node` nodes.
We have identified arguments to the call of the `sqlite3_exec` function via the
query
```ql
from FunctionCall exec, Expr query
where
exec.getTarget().getName() = "sqlite3_exec" and
query = exec.getArgument(1)
select exec, query
```
First, we need to incorporate the `DataFlow::Node`. The key to this is
`node.asExpr()`, which yields the `node`'s expression. Adding this we get
```ql
import cpp
import semmle.code.cpp.dataflow.TaintTracking
from FunctionCall exec, Expr query, DataFlow::Node sink
where
exec.getTarget().getName() = "sqlite3_exec" and
query = exec.getArgument(1) and
sink.asExpr() = query
select exec, query, sink
```
Notice that `query` is now redundant, so this simplifies to
```ql
from FunctionCall exec, DataFlow::Node sink
where
exec.getTarget().getName() = "sqlite3_exec" and
sink.asExpr() = exec.getArgument(1)
select exec, sink
```
Second, we need this as a predicate of a single argument, `predicate
isSink(DataFlow::Node sink)`. For this we introduce the `exists()`
[quantifier](https://help.semmle.com/QL/ql-handbook/formulas.html?highlight=exists#exists)
to move the `FunctionCall exec` into the body of the query and remove it from the
result:
```ql
from DataFlow::Node sink
where
exists(FunctionCall exec |
exec.getTarget().getName() = "sqlite3_exec" and
sink.asExpr() = exec.getArgument(1)
)
select sink
```
To turn this into a predicate, `from` contents become arguments, the `where`
becomes the body, and the `select` is dropped:
```ql
predicate isSink(DataFlow::Node sink) {
// rc = sqlite3_exec(db, query, NULL, 0, &zErrMsg);
exists(FunctionCall exec |
exec.getTarget().getName() = "sqlite3_exec" and
sink.asExpr() = exec.getArgument(1)
)
}
```
### The isSource Predicate
Recall that the external data enters through the `buf` argument to the call
count = read(STDIN_FILENO, buf, BUFSIZE);
and we got this via the query
```ql
from FunctionCall read, Expr buf
where
read.getTarget().getName() = "read" and
buf = read.getArgument(1)
select read, buf
```
As for the `isSink` predicate in the previous section, we need to convert this to
a predicate of a single argument, `predicate isSource(DataFlow::Node source)`.
Following the same steps, we introduce a `DataFlow::Node` and an `exists()`:
```ql
import cpp
import semmle.code.cpp.dataflow.TaintTracking
from DataFlow::Node source
where
exists(FunctionCall read |
read.getTarget().getName() = "read" and
read.getArgument(1) = source.asExpr()
)
select source
```
There is one more adjustment needed for this to work. The `buf` argument is both
read by and written to by the `snprintf` function call. Because we are specifying
it as a *source*, the value of interest is the value *after* the call. We get
this value by
[casting](https://help.semmle.com/QL/ql-handbook/expressions.html#casts) to the
post-update node. Instead of `source.asExpr()`, we use
`source.(DataFlow::PostUpdateNode).getPreUpdateNode().asExpr()`
Last, we incorporate this into a predicate:
```ql
predicate isSource(DataFlow::Node source) {
// count = read(STDIN_FILENO, buf, BUFSIZE);
exists(FunctionCall read |
read.getTarget().getName() = "read" and
read.getArgument(1) = source.(DataFlow::PostUpdateNode).getPreUpdateNode().asExpr()
)
}
```
If you quick-eval this predicate, you will see that `source` is now `ref arg buf`
instead of `buf`.
### The isAdditionalTaintStep Predicate
Our previous query identifies the call to `snprintf` and the extra flow arguments:
```ql
from FunctionCall printf, Expr out, Expr into
where
printf.getTarget().getName().matches("%snprintf%") and
printf.getArgument(0) = out and
printf.getArgument(6) = into
select printf, out, into
```
As for the `isSource` and `isSink` predicates, we need to
- change from `Expr` to a `DataFlow::Node`
- change the outflow (`out`) type to a `PostUpdateNode`
- convert this to a predicate
Put together:
```ql
import cpp
import semmle.code.cpp.dataflow.TaintTracking
predicate isAdditionalTaintStep(DataFlow::Node into, DataFlow::Node out) {
// Extra taint step for
// snprintf(query, bufsize, "INSERT INTO users VALUES (%d, '%s')", id, info);
exists(FunctionCall printf |
printf.getTarget().getName().matches("%snprintf%") and
printf.getArgument(0) = out.(DataFlow::PostUpdateNode).getPreUpdateNode().asExpr() and
printf.getArgument(6) = into.asExpr()
)
}
```
## Appendix
This appendix has the complete C source and codeql query.
### The complete Query: SqlInjection.ql
The full query is
```ql
/**
* @name SQLI Vulnerability
* @description Using untrusted strings in a sql query allows sql injection attacks.
* @kind path-problem
* @id cpp/SQLIVulnerable
* @problem.severity warning
*/
import cpp
import semmle.code.cpp.dataflow.TaintTracking
import DataFlow::PathGraph
class SqliFlowConfig extends TaintTracking::Configuration {
SqliFlowConfig() { this = "SqliFlow" }
override predicate isSource(DataFlow::Node source) {
// count = read(STDIN_FILENO, buf, BUFSIZE);
exists(FunctionCall read |
read.getTarget().getName() = "read" and
read.getArgument(1) = source.(DataFlow::PostUpdateNode).getPreUpdateNode().asExpr()
)
}
override predicate isSanitizer(DataFlow::Node sanitizer) { none() }
override predicate isAdditionalTaintStep(DataFlow::Node into, DataFlow::Node out) {
// Extra taint step
// snprintf(query, bufsize, "INSERT INTO users VALUES (%d, '%s')", id, info);
// But snprintf is a macro on mac os. The actual function's name is
// #undef snprintf
// #define snprintf(str, len, ...) \
// __builtin___snprintf_chk (str, len, 0, __darwin_obsz(str), __VA_ARGS__)
// #endif
exists(FunctionCall printf |
printf.getTarget().getName().matches("%snprintf%") and
printf.getArgument(0) = out.(DataFlow::PostUpdateNode).getPreUpdateNode().asExpr() and
// very specific: shifted index for macro.
printf.getArgument(6) = into.asExpr()
)
}
override predicate isSink(DataFlow::Node sink) {
// rc = sqlite3_exec(db, query, NULL, 0, &zErrMsg);
exists(FunctionCall exec |
exec.getTarget().getName() = "sqlite3_exec" and
exec.getArgument(1) = sink.asExpr()
)
}
}
from SqliFlowConfig conf, DataFlow::PathNode source, DataFlow::PathNode sink
where conf.hasFlowPath(source, sink)
select sink, source, sink, "Possible SQL injection"
```
### The Database Writer: add-user.c
The complete source for the sqlite database writer
```c
#include <stdio.h>
#include <stdlib.h>
#include <unistd.h>
#include <ctype.h>
#include <sqlite3.h>
#include <time.h>
void write_log(const char* fmt, ...) {
time_t t;
char tstr[26];
va_list args;
va_start(args, fmt);
t = time(NULL);
ctime_r(&t, tstr);
tstr[24] = 0; /* no \n */
fprintf(stderr, "[%s] ", tstr);
vfprintf(stderr, fmt, args);
va_end(args);
fflush(stderr);
}
void abort_on_error(int rc, sqlite3 *db) {
if( rc ) {
fprintf(stderr, "Can't open database: %s\n", sqlite3_errmsg(db));
sqlite3_close(db);
fflush(stderr);
abort();
}
}
void abort_on_exec_error(int rc, sqlite3 *db, char* zErrMsg) {
if( rc!=SQLITE_OK ){
fprintf(stderr, "SQL error: %s\n", zErrMsg);
sqlite3_free(zErrMsg);
sqlite3_close(db);
fflush(stderr);
abort();
}
}
char* get_user_info() {
#define BUFSIZE 1024
char* buf = (char*) malloc(BUFSIZE * sizeof(char));
int count;
// Disable buffering to avoid need for fflush
// after printf().
setbuf( stdout, NULL );
printf("*** Welcome to sql injection ***\n");
printf("Please enter name: ");
count = read(STDIN_FILENO, buf, BUFSIZE);
if (count <= 0) abort();
/* strip trailing whitespace */
while (count && isspace(buf[count-1])) {
buf[count-1] = 0; --count;
}
return buf;
}
int get_new_id() {
int id = getpid();
return id;
}
void write_info(int id, char* info) {
sqlite3 *db;
int rc;
int bufsize = 1024;
char *zErrMsg = 0;
char query[bufsize];
/* open db */
rc = sqlite3_open("users.sqlite", &db);
abort_on_error(rc, db);
/* Format query */
snprintf(query, bufsize, "INSERT INTO users VALUES (%d, '%s')", id, info);
write_log("query: %s\n", query);
/* Write info */
rc = sqlite3_exec(db, query, NULL, 0, &zErrMsg);
abort_on_exec_error(rc, db, zErrMsg);
sqlite3_close(db);
}
int main(int argc, char* argv[]) {
char* info;
int id;
info = get_user_info();
id = get_new_id();
write_info(id, info);
/*
* show_info(id);
*/
}
```

View File

@@ -0,0 +1,14 @@
---
lockVersion: 1.0.0
dependencies:
codeql/cpp-all:
version: 0.9.1
codeql/dataflow:
version: 0.0.2
codeql/ssa:
version: 0.1.3
codeql/tutorial:
version: 0.1.3
codeql/util:
version: 0.1.3
compiled: false

BIN
codeql-dataflow-sql-injection/cpp-sqli-3fe610d-1.zip (Stored with Git LFS) Normal file

Binary file not shown.

Binary file not shown.

Binary file not shown.

Binary file not shown.

View File

@@ -0,0 +1,4 @@
name: codeql-workshop/cpp-sql-injection
version: 0.0.1
dependencies:
codeql/cpp-all: "*"

View File

@@ -0,0 +1,60 @@
# -*- sh -*-
.runs | .[] | .results | .[] |
( (.ruleId, ": ",
(.message.text | split("\n") | ( .[0], " [", length-1 , " more]")),
"\n")
,
(if (.codeFlows != null) then
(.codeFlows | .[] |
(" Path\n"
,
( .threadFlows | .[] | .locations | .[] | .location | " "
,
( .physicalLocation | ( .artifactLocation.uri, ":", .region.startLine, ":"))
,
(.message.text, " ")
,
"\n"
)))
else
(.locations | .[] |
( " "
,
(.physicalLocation | ( .artifactLocation.uri, ":", .region.startLine, ":"))
))
,
# .message.text,
"\n"
end)
) | tostring
# This script extracts the following parts of the sarif output:
#
# # problem
# "runs" : [ {
# "results" : [ {
# "ruleId" : "cpp/UncheckedErrorCode",
# # path problem
# "runs" : [ {
# "tool" : {
# "driver" : {
# "rules" : [ {
# "properties" : {
# "kind" : "path-problem",
# "runs" : [ {
# "results" : [ {
# "ruleId" : "cpp/DangerousArithmetic",
# "ruleIndex" : 6,
# "message" : {
# "text" : "Potential overflow (conversion: int -> unsigned int)\nPotential overflow (con
# "runs" : [ {
# "results" : [ {
# "codeFlows" : [ {
# "threadFlows" : [ {
# "locations" : [ {
# "location" : {
# "message" : {
# "text" : "buff"

View File

@@ -0,0 +1,29 @@
import cpp
// 1. invalid input -- source
// count = read(STDIN_FILENO, buf, BUFSIZE - 1);
//
// 2. gets to a sql statement -- flow
// flow config
//
// 3. drops table -- sink
// rc = sqlite3_exec(db, query, NULL, 0, &zErrMsg);
// All predicates and classes are using one of:
// AST Abstract syntax tree
// CFG Control flow graph
// DFG Data flow graph
// Type hierarchy
class DataSource extends VariableAccess {
DataSource() {
exists(FunctionCall read |
read.getTarget().getName() = "read" and
read.getArgument(1) = this
)
}
}
from FunctionCall read, VariableAccess buf
where
read.getTarget().getName() = "read" and
read.getArgument(1) = buf
select buf