Re: [PATCH 6/6] maintenance: use random minute in systemd scheduler

From: Taylor Blau <me@ttaylorr.com>
Date: Aug 8, 2023, 20:05 UTC
Message-ID: <ZNKgBMg+bU4k+9fU@nand.local>
In-Reply-To: <2db943c3-2ec2-4b7e-a161-a7c9b3f04d36@github.com>

On Tue, Aug 08, 2023 at 09:49:40AM -0400, Derrick Stolee wrote:

Show 7 quoted lines

> > But I figured I'd raise the question in case you did mean for them to
> > both run on either Sunday or Monday.
>
> I don't intend to change the day that is run as part of this change.
>
> I think all other schedulers run on Sunday, but systemd running on Monday
> is a particular detail of its "weekly" schedule.

No problem -- I figured that you didn't intend on changing the days around here, just wanted to make sure it was known that the systemd scheduler picks a different day of the week for its weekly schedule than the others do.

Show 17 quoted lines

> >> -static int systemd_timer_write_unit_templates(const char *exec_path)
> >> +static int systemd_timer_write_unit_template(enum schedule_priority schedule,
> >> +					     const char *exec_path,
> >> +					     int minute)
> >>  {
> >>  	char *filename;
> >>  	FILE *file;
> >>  	const char *unit;
> >> +	char *schedule_pattern = NULL;
> >
> > You should be able to drop the NULL initialization, since you assign
> > this value unconditionally in the switch statement below (or BUG() on an
> > unknown schedule type).
>
> Unfortunately, GCC complained about a possibly-unassigned value when I
> left this unset during development, so this actually is necessary for
> that compiler.

Ah, I would have thought that GCC would be smart enough to figure out that schedule_pattern is unconditionally initialized via the switch statement, but I guess not. Makes sense.

Thanks, Taylor