https://gcc.gnu.org/bugzilla/show_bug.cgi?id=127598

--- Comment #2 from kingdomkind <kingdomkind at protonmail dot com> ---
(In reply to Jonathan Wakely from comment #1)
> Please see https://gcc.gnu.org/ai-policy.html

Good morning,

My apologies for not having followed the standards.
I did not initially intend to make a commit (or atleast, not one that I am
aware of, reading further it seems commits are done via email so I may have
accidentally done, I mainly just supplied the patch, because why not), so I am
unsure if the "Assisted-by:" tag applies (or where to put it), however, it was
Assisted-by: GPT-5.6-Sol.

As so for the explanation itself, it was not generated / written by an LLM, and
is my own understanding. However, GPT-5.6-Sol was used to help explain the
codebase and the bug, during this (the part in quotations is quoting myself if
unclear, I just put it in quotes because I pasted it from my github where I
initially wrote about it).

As the latter would fall under the "Personal Use" section (or atleast, that was
my reading of it), and the former I am not particularly sure what section it
falls under as I did not intend to commit, I am more than happy to remove the
patch file from the bug report, although I'm unsure how to edit the original
message.

I believe the explanation largely explains the rationale behind the patch,
although, I did this a little over 2 weeks ago and have not looked at it since
(I was waiting for account approval). If any further explanation for anything
is needed, I am more than happy to look over it again, and explain each line
specifically. I have not added any additional test cases / changelog (reading
further from the ai-policy to the "usual pre-requisites" / contribute section)
as I was unaware of the requirement at the time (and I am unsure if this counts
as a submission of a patch). As per the same section, I have also performed a
complete build of GCC (with disabled multilib), and additionally ran the C++
test suite. There were quite a few failed tests, although I believe they are
unrelated to my changes, as the same tests also failed for me in a normal build
(and searching online, it seems like many tests are expected to fail). If
required, I am more than willing to write a test case & changelog to match, and
provide .sum files for the existing test summaries.

I apologise for not having read this initially (and for missing the tag, if it
falls under this), and for a delayed response (running the tests took a while).
Thank you for your time.

Reply via email to