| Commit message (Collapse) | Author | Age |
|---|
| |
|
|
|
|
|
|
|
| |
This reverts commit b2f8f9edd5c1bb0a9c8c4f4b049381b979d3e385, reversing
changes made to da99355b4d6de31aec5a660f7afe100190a8e683.
Hugo asked for more discussion on this topic, and it was not in the roadmap. I
merged it prematurely because I thought there was a consensus. Also, I missed
that it was changing coq_makefile. Sorry about that.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
On a machine for which ocamlopt is available, the make world will now
perform bytecode compilation only in grammar/ (up to the syntax
extension grammar.cma), and then exclusively use ocamlopt.
In particular, make world do not build bin/coqtop.byte.
A separate rule 'make byte' does it, as well as bytecode plugins and
things like dev/printers.cma.
'make install' deals only with the part built by 'make', while a new
rule 'make install-byte' installs the part built by 'make byte'.
IMPORTANT: PLEASE AVOID doing things like 'make -j world byte' or any
parallel mix of native and byte rules. These are known to crash sometimes,
see below. Instead, do rather 'make -j && make -j byte'.
Indeed, apart from marginal compilation speed-up for users not interested
in byte versions, the main reason for this commit is to discourage any
simultaneous use of OCaml native and byte compilers. Indeed, ocamlopt and
ocamlc will both happily destroy and recreate .cmi for .ml files with no .mli,
and in case of parallel build this may happen at the very moment another
ocaml(c|opt) is accessing this .cmi. Until now, this issue has been
handled via nasty hacks (see the former MLWITHOUTMLI and HACKMLI vars in
Makefile.build). But these hacks weren't obvious to extend to ocamlopt
-pack vs. ocamlopt -pack.
coqdep_boot takes a "-dyndep" option to control precisely how a Declare ML
Module influences the .v.d dependency file. Possible values are:
-dyndep opt : regular situation now, depends only on .cmxs
-dyndep byte : no ocamlopt, or compilation forced to bytecode, depends on .cm(o|a)
-dyndep both : earlier behavior, dependency over both .cm(o|a) and .cmxs
-dyndep none : interesting for coqtop with statically linked plugins
-dyndep var : place Makefile variables $(DYNLIB) and $(DYNOBJ) in .v.d
instead of extensions .cm*, so that the choice is made in the rest of the
makefile (see next commit about coq_makedile)
NB: two extra mli added to avoid building unecessary .cmo during 'make world',
without having to use the ocamldep -native option.
NB: we should state somewhere that coqmktop -top won't work unless
'make byte' was done first
|
| | |
|
| |
|
|
|
|
|
|
|
|
| |
Now that the plugins are packed, a plugin forms now a unique
compilation unit, and we only need to install the main cmi file of
this plugin (foo_plugin.cmi).
Btw, better variable names (e.g. OMEGACMO instead of OMEGACMA) and
some other cleanup in Makefile.common (no more INITPLUGINS variable,
for instance).
|
|
|
General idea : Makefile.build was far too big to be easy to grasp or
maintain, with information scattered everywhere. Let's try to tidy that!
Normally, this commit is transparent for the user. We simply regroup
some parts of Makefile.build in several new dedicated files:
- Makefile.ide
- Makefile.checker
- Makefile.dev (for printers, revision, extra partial targets, otags)
- Makefile.install
These new files are "included" at the start of Makefile.build, to provide
the same behavior as before, but with a Makefile.build shrinked by 50%
(to approx 600 lines). Makefile.build now handles in priority the build
of coqtop, minor tools, theories and plugins.
Note: this is *not* a separate build system for coqchk nor coqide,
even if this can be seen as a first step in this direction (won't be easy
anyway to continue, due to the sharing of various stuff in lib and more).
In particular Makefile.{coqchk,ide} may rely here and there on some generic
rules left in Mafefile.build. Conversely, be sure to prefix rules in
Makefile.{coqchk,ide} by checker/... or ide/... in order to avoid
interferences with generic rules.
Makefile.common is still there, but quite simplified. For instance,
some variables that were used only once (e.g. lists of cmo files to link
in the various tools) are now defined in Makefile.build, directly
where they're needed. THEORIESVO and PLUGINSVO are made directly out of
the theories/*/vo.itarget and plugins/*/vo.itarget files, no long manual
list of subdirs anymore. Specific sub-targets such as 'reals' still
exist, but in Makefile.dev, and they aren't mandatory.
Makefile.doc is augmented by the rules building the documentation of
the sources via ocamldoc.
This classification attempt could probably be improved. For instance,
the install rules for coqide are currently in Makefile.ide, but could
also go in Makefile.install. Note that I've removed install-library-light
which was broken anyway (arith isn't self-contained anymore).
|
Maxime Dénès
Pierre Letouzey