Peter LaFosse

This isn't fixed. Repro steps 1. Open a new mapped file 2. Add a memory region 3. Go to the end of that memory region 4. try to assemble at...

This is not related to hexagon at all this is simply a missing HLIL simplification. This can be demonstrated on x86 as well:  Example test bndb [test_2596.bndb.zip](https://github.com/user-attachments/files/17333617/test_2596.bndb.zip)

I think we need a new token type for IL index. I believe currently we just use an AddressTokenType

I have what is largely a fix for this that I'm hoping to get pushed in this release. Currently there is an issue that needs to be resolved which is...

I'm putting it on the current milestone but no promises that we'll get it done by then. We are however looking into this.

It is heavily pipelined. You need to be able to lift multiple instructions at once in order to appropriately translate the semantics.

I've read through all the code and I generally am in favor. I guess my one thought that I haven't fully thought through is default enum signedness on clang vs...

In general I'm a huge fan of this PR. I appreciate all the static asserts which should help this code from breaking in the future. I don't see downsides to...

I can see the desire for this in the UI but having the `remove_memory_region` API do that seems a little heavy handed as you should just be able to remove...

Is there another issue tracking this? Its not hard to come up with a good size. sizeof(paste_buffer), or page-aligned(sizeof(paste_buffer)) would be reasonable decisions. zero isn't a reasonable decision.