[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [PATCH] shuffle: disable by default
From: |
Yann E. MORIN |
Subject: |
Re: [PATCH] shuffle: disable by default |
Date: |
Sun, 16 Oct 2022 15:09:09 +0200 |
User-agent: |
Mutt/1.5.22 (2013-10-16) |
Paul, All,
On 2022-10-16 08:39 -0400, Paul Smith spake thusly:
> On Sun, 2022-10-16 at 12:17 +0200, Yann E. MORIN wrote:
> > Commit 621d3196fae9 ([SV 62100] Add '--shuffle' option support)
> > introduced randomisation of prerequisites, so as to detect more
> > build ordering issues, especially with parallel builds. It also
> > made randomisation the default.
>
> I'm not sure I understand the goal of this patch.
>
> If you don't want randomization, just don't add the --shuffle option
> and you won't get any randomization.
>
> The behavior of make without the --shuffle option hasn't changed from
> what it was before: it will never be the case that shuffle is the
> default "always on" behavior
I looked back at my tests, and indeed I messed up between the tests
outputs, confusing one with --shuffle as a test without it...
I redid the tests in a clean session, paying extra attention, and the
ordering is un-modified without --shuffle, which is as I expected.
> because POSIX mandates the order in which
> prerequisites are considered.
Thanks, I hadn't have a look before, but this is indeed explicitly
stated:
https://pubs.opengroup.org/onlinepubs/9699919799.2018edition/utilities/make.html#tag_20_76_13
The make utility shall treat all prerequisites [...],processing them
in the order in which they appear in the rule.
Good to know, I was always wondering if that was an implicit behaviour,
or if that was mandated. Thanks for having me look at it. :-)
> So I don't see why we would say that both "make" and "make --shuffle"
> do the same thing (no shuffle). If you've added --shuffle then I'd
> expect you'd want to see shuffling...?
>
> Am I wrong and there is some way to get shuffle mode enabled even
> without the --shuffle option? That would definitely be a bug!
No, there is no bug; I messed up.
Sorry for causing confusion; I'll get my second coffee before I post
patches next time.
Regards,
Yann E. MORIN.
--
.-----------------.--------------------.------------------.--------------------.
| Yann E. MORIN | Real-Time Embedded | /"\ ASCII RIBBON | Erics' conspiracy: |
| +33 662 376 056 | Software Designer | \ / CAMPAIGN | ___ |
| +33 561 099 427 `------------.-------: X AGAINST | \e/ There is no |
| http://ymorin.is-a-geek.org/ | _/*\_ | / \ HTML MAIL | v conspiracy. |
'------------------------------^-------^------------------^--------------------'