standalone: -s to set codegen size_t & ./check_demo.sh to test both 32-bits and 64-bits #468
No reviewers
Labels
No Milestone
No Assignees
3 Participants
Notifications
Due Date
No due date set.
Dependencies
No dependencies set.
Reference: M-Labs/nac3#468
Loading…
Reference in New Issue
No description provided.
Delete Branch "standalone-any-size-t"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
Check NAC3 code generator in both 32-bits and 64-bits.
A lot of new 32-bit specific errors are revealed with
./check_demos.sh
. New issue on each of them later.standalone: -s to set codegen size_t & ./check_demo.sh to test both 32-bits and 64-bitsto WIP: standalone: -s to set codegen size_t & ./check_demo.sh to test both 32-bits and 64-bitsWIP: standalone: -s to set codegen size_t & ./check_demo.sh to test both 32-bits and 64-bitsto standalone: -s to set codegen size_t & ./check_demo.sh to test both 32-bits and 64-bitsThought there is a critical error in the PR.
Please ignore the title changes.
@ -249,0 +261,4 @@
// Check `size_t_bits`
if !(size_t_bits == 32 || size_t_bits == 64) {
println!("size_t_bits cannot be {size_t_bits}. Must be 32 or 64.");
return; // Terminate
These two comments are obvious, remove.
@ -25,1 +18,4 @@
# Run the .py with an actual Python interpreter.
# This is the test oracle.
echo ">>>>>> Running $demo with the Python interpreter to get test answers"
reference output
@ -26,0 +21,4 @@
echo ">>>>>> Running $demo with the Python interpreter to get test answers"
./interpret_demo.py "$demo" > interpreted.log
# Test NAC3 on 32-bits and 64-bits code output, with and without --lli,
Obviously
@ -74,2 +74,4 @@
mcpu: Option<String>,
/// The number of bits of `size_t` of the NAC3 code generator.
// NOTE: Don't set a default value (e.g., `default_value = usize::BITS`). Force the user to specify to make it clear to them the `size_t` they will be using.
No, the default should be host machine dependent. GCC doesn't force you to set one.
38b1567e10
to15f4cc79c4
Revised
I would suggest using
-m32
and-m64
instead to conform with GCC/Clang.Can we close this?
Sure
NOTE: #477 has made this PR redundant.
Pull request closed