Re: [PATCH] allow disabling fsync everywhere

From: Junio C Hamano <gitster@pobox.com>
Date: Oct 28, 2021, 19:29 UTC
Message-ID: <xmqqilxhq774.fsf@gitster.g>
In-Reply-To: <20211028182824.GA1307@dcvr>

Eric Wong <e@80x24.org> writes:

Show 25 quoted lines

> Jeff King <peff@peff.net> wrote:
>> On Thu, Oct 28, 2021 at 12:21:02AM +0000, Eric Wong wrote:
>> 
>> > "core.fsync" and the "GIT_FSYNC" environment variable now exist
>> > for disabling fsync() even on packfiles and other critical data.
>
> Fwiw, I'm questioning the need for core.fsync.  GIT_FSYNC alone
> may be sufficient.
>
>> > This will also be useful for 3rd-party tools which create
>> > throwaway git repositories of temporary data.
>> 
>> Do you mostly just care about the tests, or is the third-party tool
>> support important to you? I ask because most of us switched to running
>> the tests with --root=/some/tmpfs long ago to achieve the same speedup.
>
> Third-party tools and OSes which don't have a tmpfs mounted by
> default (I don't think most *BSDs have tmpfs enabled by
> default).
>
> I try to use libeatmydata everywhere I can; but that's not
> always installed.
>
> I'm also strongly considering making GIT_FSYNC=0 the default for
> our own test suite since it's less setup for newbies.

If the primary focus is for testing, perhaps GIT_TEST_FSYNC would be better? I do not think we want to even advertise an option for not syncing at all for end users working with any real repositories. Not even when they promise that the end user accepts all the responsibility and will keep both halves when things got broken.

Thanks.