This is a summary of discussions relative to the merge request created by
Antoni Boucher (antoyo) titled
libgccjit: Add _Float16, _Float32, _Float64 and __float128 support for jit
since its creation.
Description: ```
gcc/jit/ChangeLog:
* docs/topics/types.rst: Document new types.
* dummy-frontend.cc: Support new types in tree_type_to_jit_type.
* jit-common.h: Update NUM_GCC_JIT_TYPES.
* jit-playback.cc: Support new types in get_tree_node_for_type.
* jit-recording.cc: Support new types.
* libgccjit.h (gcc_jit_types): Add new types.
gcc/testsuite/ChangeLog:
* jit.dg/test-types.c: Test new types.
```
* [x] Add the types like GCC_JIT_TYPE_FLOAT16 as supported. (DONE in #108).
CC: David Malcolm <[email protected]>, <[email protected]>
The full and up to date discussion can be found at
https://forge.sourceware.org/gcc/gcc-TEST/pulls/82
The merge request has been closed without being merged directly on the forge
repository.
On 2025-09-19 14:00:43+00:00, Antoni Boucher (antoyo) commented on the code:
> +++ gcc/testsuite/jit.dg/test-types.c
> @@ -157,6 +161,29 @@ create_code (gcc_jit_context *ctxt, void *user_data)
> gcc_jit_field *field_m_long_double =
> CREATE_FIELD (GCC_JIT_TYPE_LONG_DOUBLE, "m_long_double");
>
> + gcc_jit_target_info *target_info = gcc_jit_context_get_target_info(ctxt);
This PR also depends on [this one](https://github.com/antoyo/libgccjit/pull/6)
for the `target_info` stuff.
On 2025-09-19 14:03:41+00:00, Lichenor Forgejo Bot (forge-bot) wrote:
<!-- pr-new-version -->
Version 1 of this pull request has been stored. It includes the following
commits:
- libgccjit: Add _Float16, _Float32, _Float64 and __float128 support for jit -
7ecd0158db98bfdf4b6971bd875fe84bd6a1a95e
On 2025-09-23 09:10:57+00:00, Claudio Bantaloukas (rdfm) wrote:
Hi @antoyo ,
it looks like the bot posted the patch correctly but had issues posting the
comment pointing [at the patch in
inbox](https://inbox.sourceware.org/test-list/175743424336.5.12081261745369171928.batrachomyomachia.gcc.gcc-test.74....@forge-test.rdfm.org/T/#t).
Thank you for trying this out and please let me know if something is not
working as expected.
On 2025-09-23 12:45:53+00:00, Antoni Boucher (antoyo) wrote:
@rdfm: Yeah, I just remembered about telling you this.
Also, it seems the CC didn't work either: did I use the wrong formatting?
On 2025-10-21 21:11:09+00:00, David Malcolm (dmalcolm) wrote:
Thanks for the patch.
I was a bit worried at first when I saw the cases added in the middle of the
switch statements, but I checked and the patch adds the new values at the end
of `enum gcc_jit_types`, so it's not an ABI break.
Given that, the patch is good to push to trunk. Thanks!
On 2025-10-22 00:46:02+00:00, Antoni Boucher (antoyo) wrote:
Merged.
Thanks for the review!
