hohn/codeql
1
0
Fork 0
mirror of https://github.com/github/codeql.git synced 2026-05-02 20:25:13 +02:00
Code Issues Packages Projects Releases Wiki Activity

Python: Rewrite attribute lookup helpers for better performance

Browse Source 
Not that they actually had a huge problem right now, just that using the old
pattern HAS lead to bad performance in the past. See
https://github.com/github/codeql/pull/4361
This commit is contained in:
Rasmus Wriedt Larsen
2020-09-30 17:24:50 +02:00
parent 4adc26eb62
commit c4a2e1d6d1
1 changed files with 38 additions and 10 deletions
Download Patch File Download Diff File Expand all files Collapse all files

48
python/ql/src/experimental/semmle/python/frameworks/Stdlib.qll
View File

@@ -29,7 +29,7 @@ private module Stdlib {
*
* For example, using `attr_name = "system"` will get all uses of `os.system`.
*/
private DataFlow::Node os_attr(string attr_name, DataFlow::TypeTracker t) {
private DataFlow::Node os_attr(DataFlow::TypeTracker t, string attr_name) {
attr_name in ["system", "popen",
// exec
"execl", "execle", "execlp", "execlpe", "execv", "execve", "execvp", "execvpe",
@@ -41,10 +41,24 @@ private module Stdlib {
result = DataFlow::importMember("os", attr_name)
or
t.startInAttr(attr_name) and
result = os()
or
exists(DataFlow::TypeTracker t2 | result = os_attr(attr_name, t2).track(t2, t))
result = DataFlow::importModule("os")
)
or
// Due to bad performance when using normal setup with `os_attr(t2, attr_name).track(t2, t)`
// we have inlined that code and forced a join
exists(DataFlow::TypeTracker t2 |
exists(DataFlow::StepSummary summary |
os_attr_first_join(t2, attr_name, result, summary) and
t = t2.append(summary)
)
)
}
pragma[nomagic]
private predicate os_attr_first_join(
DataFlow::TypeTracker t2, string attr_name, DataFlow::Node res, DataFlow::StepSummary summary
) {
DataFlow::StepSummary::step(os_attr(t2, attr_name), res, summary)
}
/**
@@ -54,7 +68,7 @@ private module Stdlib {
* For example, using `"system"` will get all uses of `os.system`.
*/
private DataFlow::Node os_attr(string attr_name) {
result = os_attr(attr_name, DataFlow::TypeTracker::end())
result = os_attr(DataFlow::TypeTracker::end(), attr_name)
}
/**
@@ -148,17 +162,31 @@ private module Stdlib {
*
* For example, using `attr_name = "Popen"` will get all uses of `subprocess.Popen`.
*/
private DataFlow::Node subprocess_attr(string attr_name, DataFlow::TypeTracker t) {
private DataFlow::Node subprocess_attr(DataFlow::TypeTracker t, string attr_name) {
attr_name in ["Popen", "call", "check_call", "check_output", "run"] and
(
t.start() and
result = DataFlow::importMember("subprocess", attr_name)
or
t.startInAttr(attr_name) and
result = subprocess()
or
exists(DataFlow::TypeTracker t2 | result = subprocess_attr(attr_name, t2).track(t2, t))
result = DataFlow::importModule("subprocess")
)
or
// Due to bad performance when using normal setup with `subprocess_attr(t2, attr_name).track(t2, t)`
// we have inlined that code and forced a join
exists(DataFlow::TypeTracker t2 |
exists(DataFlow::StepSummary summary |
subprocess_attr_first_join(t2, attr_name, result, summary) and
t = t2.append(summary)
)
)
}
pragma[nomagic]
private predicate subprocess_attr_first_join(
DataFlow::TypeTracker t2, string attr_name, DataFlow::Node res, DataFlow::StepSummary summary
) {
DataFlow::StepSummary::step(subprocess_attr(t2, attr_name), res, summary)
}
/**
@@ -168,7 +196,7 @@ private module Stdlib {
* For example, using `attr_name = "Popen"` will get all uses of `subprocess.Popen`.
*/
private DataFlow::Node subprocess_attr(string attr_name) {
result = subprocess_attr(attr_name, DataFlow::TypeTracker::end())
result = subprocess_attr(DataFlow::TypeTracker::end(), attr_name)
}
/**