-
Notifications
You must be signed in to change notification settings - Fork 4.4k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
fix test_new_from_file_crafted_executable for m1 #26009
fix test_new_from_file_crafted_executable for m1 #26009
Conversation
Wasn't sure if this issue was due to the m1 architecture itself or the compiler. To test, I wrote a simple c++ program:
Compiling with clang led to a failure (see womp). |
Codecov Report
@@ Coverage Diff @@
## master #26009 +/- ##
=========================================
- Coverage 82.1% 82.0% -0.2%
=========================================
Files 628 631 +3
Lines 171471 172920 +1449
=========================================
+ Hits 140878 141874 +996
- Misses 30593 31046 +453 |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
lgtm
runtime/src/append_vec.rs
Outdated
const FALSE: bool = false; // keep clippy happy | ||
if *executable_bool == FALSE { | ||
panic!("This didn't occur if this test passed."); | ||
} |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
could you restore this with #[cfg(!m1)]
?
the intent isn't clear but this test is checking the existence of undefined behavior around bool inside rustc (or llvm): https://doc.rust-lang.org/reference/types/boolean.html:
An object with the boolean type has a size and alignment of 1 each. The value false has the bit pattern 0x00 and the value true has the bit pattern 0x01. It is undefined behavior for an object with the boolean type to have any other bit pattern.
as you know, we're not relying on the UB (bit pattern other than 0x00 and 0x01 for bool
) in any production code. however, this assertion is needed to validate that this test is still exposing the UB to make sure we're not relaying on it accidentally. So, if possible, i want to the test to check the existence of UB at least for x64.
any ad-hoc cfg
should be fine as long as our CI is happy.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Gotcha, re-added the if/panic, but disabled it for aarch64.
@@ -899,16 +899,16 @@ pub mod tests { | |||
let accounts = av.accounts(0); | |||
let account = accounts.first().unwrap(); | |||
|
|||
// upper 7-bits are not 0, so sanitization should fail | |||
assert!(!account.sanitize_executable()); |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
👍
my memory is vague. but i think i tried to add this assertion, but i couldn't because the ub assertion below stopped working correctly back then. (welcome to UB land. lol)
if today's build toolchain doesn't negatively affect the UB detection code, please leave to assert!
here.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
^passing CI with the assert left there.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
@apfitzge lgtm!
Problem
Test fails on m1 mac's w/ current rust compiler.
Summary of Changes
Assert that sanitize_executable fails (what we really care about).
Remove the check that
*executable_bool == FALSE
, which changes based on compilers.Fixes #