Re: [lmi] [lmi-commits] master a6eb063a: Adapt to wine-7.0.1

[Vadim Zeitlin]

On Sat, 21 Jan 2023 14:54:32 +0000 Greg Chicares <gchicares@sbcglobal.net> 
wrote:

GC> On 1/20/23 00:43, Vadim Zeitlin wrote:
GC> > On Thu, 19 Jan 2023 19:29:49 -0500 (EST) Greg Chicares 
<gchicares@sbcglobal.net> wrote:
GC> [...]
GC> > GC>     Upgrading from
GC> > GC>       wine-7.0 (Debian 7.0~repack-8)
GC> > GC>     to
GC> > GC>       wine-7.0.1 (Debian 7.0.1~repack-1)
GC> > GC>     causes some unwanted diagnostics to appear. Filtered them out, 
because
GC> > GC>     there seems to be no safe and easy way to suppress them with 
$WINEDEBUG.
GC> [...]
GC> >  I'm somewhat surprised that these messages happen now if they didn't
GC> > happen before because this typically just indicates that there is no X
GC> > server running and this is not something that can possibly change with a
GC> > Wine upgrade.
GC> 
GC> Correct. I tried running while connected to an actual TTY, and did not
GC> see these messages. The cause was not a 'wine' upgrade, but rather
GC> running with no TTY, e.g.:
GC>   nohup ./lmi_setup_00.sh >log_"$(date -u +"%Y%m%dT%H%MZ")" 2>&1 & disown; 
exit
GC> 
GC> These messages occur only when building for msw--never for GNU/Linux.
GC> It would seem that 'wine' expects a TTY (and diagnoses its absence)

 Interesting, I wouldn't expect this, but at least it answers my original
question, thank you.

GC> Incidentally, these messages appear only for command-line tasks,
GC> but not for the GUI test that is run under xvfb.

 My hypothesis is that without the TTY Wine fails to connect to the local X
server for whatever reason (e.g. perhaps it considers that it's not running
interactively in this case), resulting in these messages. But when we
explicitly use it with Xvfb, it does use it, and so the problem doesn't
occur.

 This is pure speculation so far, but I should be able to check it, if you
think it's worth doing it, it just would take more time because my way of
checking is rather time-consuming as it consists in first looking at Wine
code and then, after invariably failing to understand what it does, running
Wine under gdb and looking at what it's actually doing.

 Anyhow, please let me know if I should look at this in more details,
VZ

pgpia5OeB2184.pgp
Description: PGP signature