Re: ✘gpsd .23.2~rc1

[Gary E. Miller]

Yo Hal!

On Fri, 15 Apr 2022 02:12:00 -0700
Hal Murray <halmurray@sonic.net> wrote:

> gem@rellim.com said:
> > I think I'm caught up.  So all yours.  Fred wants to slip a macro
> > into driver_ubx.c, but that should be cosmetic.   
> 
> Builds all work.

Gerat.

> My setup uses clang on OpenBSD and FreeBSD.  Both provide a blizzard
> of cc: warning: optimization flag '-fexcess-precision=standard' is
> not supported [-Wignored-optimization-argument]
> 
> OpenBSD clang version 11.1.0
> OpenBSD clang version 11.1.0 (32 bit)
> FreeBSD clang version 11.0.1
> FreeBSD clang version 10.0.0
> FreeBSD clang version 11.0.1 (32 bit)

Yeah, for some reason the scons option checker misses that one.  I'll
play with it.

> longjmp warning:
> gpsd-3.23.2~rc1/gpsmon/gpsmon.c:1421:24: warning: variable 
> \u2018ts_timeout\u2019 might be clobbered by \u2018longjmp\u2019 or 
> \u2018vfork\u2019 [-Wclobbered]
> gpsd-3.23.2~rc1/gpsd/gpsd.c:2651:26: warning: variable
> \u2018ts_timeout\u2019 might be clobbered by \u2018longjmp\u2019 or
> \u2018vfork\u2019 [-Wclobbered]

As previously discussed. that is a totally invalid warning.  Id
like to just rip out the setjmp/longjump stuff, but that is too big a
job for an rc1.

> Rutime checks are mostly clean.

Good.

> 32 bit OpenBSD gets timeouts.  I tried increasing WRITE_PAD.  It's
> currently 0.001.  I got timeouts with 0.002.  With 0.005 I got a big
> diff error.  I'm putting that on the back burner and typing this in.

The timeouts have been a PITA for a long time.

So, one mior nit to fix this cycle.

RGDS
GARY
---------------------------------------------------------------------------
Gary E. Miller Rellim 109 NW Wilmington Ave., Suite E, Bend, OR 97703
        gem@rellim.com  Tel:+1 541 382 8588

            Veritas liberabit vos. -- Quid est veritas?
    "If you can't measure it, you can't improve it." - Lord Kelvin

pgp4pVYYTNnoF.pgp
Description: OpenPGP digital signature