Re: [Texmacs-dev] TeXmacs 1.0.0.7

[David Allouche]

On Wednesday 12 June 2002 18:03, Stéphane Payrard wrote:
>
> TeXmacs 1.0.0.7 compiles and runs fine with gcc 3.1.1 modulo the
> following patch that remove a spurious default affectation to a formal
> parameter in a .cc It also warns about bogus volatile but that will be
> fixed by the nogencc patch from David.

I do not think that nogencc will fix the volatile problem. If I am right 
it a problem with fatal_error. Joris told me that this function is marked 
volatile because:

1. it does not return;

2. where it is used, it prevent the compiler from warning because there 
is an execution path which does not return a value.

Actually, I think we could find a way around that problem in looking in 
macros like "assert_not_reached" (or something like that) in glib. I have 
not tried, but there is the same problem there so the glib folks 
hopefully fixed is the Right Way. Anyway that is not going in before 
nogencc is merged with mainline.

Since Joris has accepted the patch, I will merge it in the next nogencc 
release so it can be merged with mainline in the next unstable. I think 
you can expect the last nogencc (with a minor fix in hashmap.hh, proper 
support for TeXmacs-misc, and a good test round on available compile 
farms) at the end of next week.
-- 

                                  -- David --