Re: Potential spiped 1.6.1 release -- please test

[Graham Percival <gperciva@tarsnap.com>]

On Sat, Dec 28, 2019 at 02:00:44PM +0200, Peter Pentchev wrote:
> On Sat, Dec 28, 2019 at 01:45:52PM +0200, Peter Pentchev wrote:
> > Thanks for your work on spiped! What do you think about the attached
> > trivial patches to the new pushbits test?

At first glance they look great, thanks!  I'm doing more testing on them, and
will submit as a PR soon.

I wish that gcc had an -Weverything option.  Turns out -Wstrict-prototypes
isn't enabled by -Wall -Wextra, and I hadn't (yet) dug into finding & enabling
ad-hoc warnings.  (I much prefer the "blacklist" method with clang's
-Weverything, and then disabling any spurious warnings.)

> Just to note: I came to these two patches as part of making a
> preliminary Debian package of spiped-1.6.1a and trying to figure out why
> the pushbits test was failing.

If the test suite is failing on any OS, I'm 99% certain that Colin will want
to make a 1.6.1b at a later date.

Could you please send me (perhaps off-list) which Debian distro and
architecture?  The tests work ok on various Ubuntu distros and buster-i386
(inside Virtualbox on a FreeBSD machine).  I'm very willing to believe that I
wrote something that isn't standard-compliant, and also willing to believe
that glibc is doing something not standard-compliant but for which we can work
around with our "POSIX compatibility failure" tests.

Cheers,
- Graham