Charlie Gordon
Charlie Gordon
The bug is in the `JS_GetOwnPropertyNamesInternal` function: we use `max_int` in atom_count = num_keys_count + str_keys_count + sym_keys_count + exotic_keys_count; /* avoid allocating 0 bytes */ tab_atom = js_malloc(ctx, sizeof(tab_atom[0])...
Not yet, currently focusing on other issues: ropes, bug fixes and support for recent tc39 extensions. Chqrlie. > On 11 Apr 2022, at 05:07, Hongbo Wan ***@***.***> wrote: > >...
There is nothing wrong with the code flagged by your compiler, and there is not much that can be done to "improve" the code. Compiling the module as C instead...
This is certainly a bug. It is unusual to label some thing that is not a loop statement, which might explain why it went unnoticed until now. Thank you for...
Releasing the contexts would also allow for builtin memory leak detection using `DUMP_LEAKS`.
> I don't think it's possible to do that reliably since a bug might have corrupted the memory. I agree. Yet tests could be run in batches of 1000 and...
@kenballus: the example is compelling! A blatant case of compiler perversion. I did not find such a potential problems in the source code, but we should probably either: * find...
@andrjohns: sorry about this, not familiar with current practice for this, shall improve soon.
Why use jemalloc by default on NetBSD?
Hello @zwarrior1, Thank you for contributing. I cannot merge your patch as posted but will apply your patches with attribution after some adjustments. See my comments inline. Chqrlie.