public inbox for binutils@sourceware.org
 help / color / mirror / Atom feed
From: Nick Alcock <nick.alcock@oracle.com>
To: lkcl via Binutils <binutils@sourceware.org>
Cc: Nick Alcock <nick.alcock@oracle.com>,
	Dmitry Selyutin <ghostmansd@gmail.com>,
	lkcl <luke.leighton@gmail.com>
Subject: Re: Are big tests allowed in binutils?
Date: Tue, 21 Jun 2022 14:43:37 +0100	[thread overview]
Message-ID: <87sfnyxhbq.fsf@esperi.org.uk> (raw)
In-Reply-To: <5222EC1F-A9D2-48BA-A4F2-6141CA88C687@gmail.com> (lkcl via Binutils's message of "Mon, 20 Jun 2022 12:31:17 +0100")

On 20 Jun 2022, lkcl via Binutils spake thusly:

> sorry nick i am stuck with a rubbish mailer, rather than top-post i'm repeating your question: why not submit the full source of the autogenerator?
>
> A: because the list of dependencies, whilst entirely FOSS, is absolutely mental and is completely outside of our control, copyright-wise.  it's also 4 levels deep dependencies and also includes *another* repository comprising CSV and markdown files extracted from the Power ISA Specification PDF [1] and microwatt [2]
>
> now, Dmitry did write a small stand-alone program to autogenerate the test cases but he had to write that by hand.
>
> our plan is to *autogenerate the autogenerator* [based on information that ultimately comes directly from the Power ISA Specification].

That sounds like a good idea to me.

> 1) because if mistakes are found they can be tracked back directly to
> the spec. [we have found appx ... 15 corrections to the published spec
> because of this]
>
> 2) because (no implied criticism of anyone else that does do manual
> unit tests) we simply don't have time money or resources to throw
> brute-force at this so out of sheer forced pragmatism are doing more
> with less.
> 
> i have no view on whether the autogenerated-autogenerator should be
> submitted or whether the automated unit tests themselves should be
> submitted but i can say for sure you *really* don't want the full
> source code of the full software stack we have been developing for 2+
> years as a hard critical build dependency of binutils.

While it's perfectly acceptable to commit the generated output if
generating the output is annoying (and having a huge stack of deps
certainly sounds like it would be annoying), you've got to commit enough
that others can also modify the testcases the same way you do: the
preferred form for modification in this case is pretty clearly not the
generated testcases but the autogenerators themselves (and any input to
them), so the GPL requires you to provide the generators too. They don't
have to be in the binutils repo, but honestly putting them in there so
people can manually run them if needed seems like a good idea. (Stick
in simple pointers to the deps as well, particularly if just figuring ou
what the deps *are* would otherwise be difficult, as "absolutely mental"
suggests it would be).

-- 
NULL && (void)

  reply	other threads:[~2022-06-21 13:43 UTC|newest]

Thread overview: 12+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-06-17  5:11 Dmitry Selyutin
2022-06-17  6:19 ` Dmitry Selyutin
2022-06-20 10:42   ` Nick Alcock
2022-06-20 11:31     ` lkcl
2022-06-21 13:43       ` Nick Alcock [this message]
2022-06-21 17:26         ` lkcl
2022-06-21 17:29           ` Dmitry Selyutin
2022-06-21 19:23             ` lkcl
2022-06-17  9:14 ` Andreas Schwab
2022-06-21 15:32 ` Richard Earnshaw
2022-06-21 16:28   ` Dmitry Selyutin
2022-06-22  6:55     ` Jan Beulich

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=87sfnyxhbq.fsf@esperi.org.uk \
    --to=nick.alcock@oracle.com \
    --cc=binutils@sourceware.org \
    --cc=ghostmansd@gmail.com \
    --cc=luke.leighton@gmail.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).